Proefschrift gho by Nicole Nijhuis

Opportunities in the Failing Heart Johannes Michael Ing Han Gho

Opportunities in the Failing Heart

Johannes Michael Ing Han Gho

Opportunities in the Failing Heart ÂŠ 2015 JMIH Gho, Utrecht Layout:

wenz iD.nl / Wendy Schoneveld

Cover design:

Robert van Sluis / eyefordetail.nl

Printed by:

Gildeprint - Enschede

Financial support by the Dutch Heart Foundation and the Heart & Lung Foundation Utrecht for the publication of this thesis is gratefully acknowledged. This research forms part of the Project P1.04 SMARTCARE of the research program of the BioMedical Materials Institute, co-funded by the Dutch Ministry of Economic Affairs, Agriculture and Innovation. Publication of this thesis was financially supported by Stichting Genetische Hartspierziekte PLN (http://stichtingpln.nl, Middenmeer, the Netherlands). Further financial support for the publication of this thesis by Stichting Cardiovasculaire Biologie and ChipSoft B.V. is gratefully acknowledged. ISBN 978-94-6233-100-6

Opportunities in the Failing Heart Mogelijkheden in het falende hart (met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof.dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op donderdag 29 oktober 2015 des middags te 12.45 uur

door

Johannes Michael Ing Han Gho geboren op 29 september 1986 te Leusden

Promotoren:

Prof.dr. P.A.F.M. Doevendans

Prof.dr. F.W. Asselbergs

Copromotoren: Dr. S.A.J. Chamuleau

Dr. A. Vink

TABLE OF CONTENTS

Chapter 1

PART ONE

Introduction

Chapter 2

Clinical Course Portrayed Heart Failure Following Myocardial Infarction: a Cohort Study of Incidence

and Prognostic Factors in 24 745 Patients Using Linked Electronic Records Chapter 3

Heart Failure Following ST-elevation Myocardial Infarction: an AGNES

Cohort Study of Incidence and Prognostic Factors

PART TWO Chapter 4

Finding Fibrosis Patterns A Systematic Comparison of Cardiovascular Magnetic Resonance and

High Resolution Histological Fibrosis Quantification in a Chronic Porcine Infarct Model Chapter 5

Endogenous Contrast MRI of Cardiac Fibrosis: Beyond Late Gadolinium

Enhancement Chapter 6

High Resolution Systematic Digital Histological Quantification of Cardiac Fibrosis and Adipose Tissue in Phospholamban p.Arg14del Mutation

Associated Cardiomyopathy Chapter 7

The Distribution Pattern of Fibrosis in Genetic Cardiomyopathy is Related to the Type of Pathogenic Mutation

109

PART THREE Chapter 8

Elucidating (Epi)genetic and Translating Therapeutic Pathways Chromatin Regulation in Phospholamban R14del Mutation Associated

127

Cardiomyopathy Chapter 9

Cell Therapy, a Novel Remedy for Dilated Cardiomyopathy?

175

A Systematic Review Chapter 10

Xenotransplantation of Human Cardiomyocyte Progenitor Cells Does Not

199

Improve Cardiac Function in a Porcine Model of Chronic Ischemic Heart Failure Chapter 11

General Discussion

221

Samenvatting in het Nederlands

232

List of publications

236

Acknowledgements / Dankwoord

238

Curriculum vitae

242

Chapter

Introduction

CHAPTER 1

INTRODUCTION Heart failure (HF) is a syndrome with typical signs and symptoms that can result from abnormal cardiac structure or function.1 It can lead to impaired quality of life, decreased functional capacity, hospital admissions and mortality. The historical terminology used to describe HF has been based on measurement of left ventricular ejection fraction (EF). Currently the diagnosis of HF has been divided in HF with reduced (HFrEF) or preserved EF (HFpEF). Causes of HF in the western world include hypertension, valvular heart disease and (non-ischaemic) cardiomyopathies and the leading cause is ischaemic heart disease (Figure 1).2, 3 Epidemiology of heart failure Heart failure incidence increases with age4 and in a previous Dutch community based middleaged cohort study 4.4% of the individuals were diagnosed with HF during a median follow-up of 11.5 years of which 66% with HFrEF and 34% with HFpEF.5 Research studying the incidence of HF following myocardial infarction (MI) is scarce and mainly stems from the thrombolytic era. There is conflicting evidence whether the long-term incidence of HF following MI is increasing or decreasing. An analysis of the Framingham Heart Study regarding long-term trends following MI demonstrated an increase in incident HF between the 1970s and 1990s, possibly due to improved survival after MI.6 The 5-year incidence of HF following MI rose from 27.6% (1970 to 1979) to 31.9% (1990 to 1999). In a different study using the Swedish hospital discharge register with MI patients from 1993-2004, the 3-year risk of HF remained high despite a decrease observed in hospitalisations for HF.7 In MI patients 65 years or older, they found a 3-year hospital HF diagnosis of 31.5% (1993 to 1995) and 28.0% (2002 to 2004). Portraying the incidence of and prognostic factors associated with HF following MI in the current era is important for future observational studies, clinical trials, measuring health outcomes and guideline development.

Figure 1. Causes of heart failure. Figure created with data derived from 3.

INTRODUCTION

Pathophysiology Heart failure has been associated with myocardial fibrosis and arrhythmias and different

pathophysiological processes can result in focal and diffuse myocardial fibrosis.8 After myocardial injury, maladaptive changes can lead to pathologic remodelling of the left ventricle with dilatation, impaired contractility and relaxation.9 In patients with left ventricular systolic dysfunction, systemic responses such as an elevated sympathetic tone and renin-angiotensin-aldosterone system can have detrimental effects leading to the clinical syndrome of HF. In a previous study using the Framingham Offspring Study cohort, approximately 18% of the risk of HF was attributable to parental HF.10 An other important cause of HF are non-ischaemic cardiomyopathies which are mainly inherited and during the last decades developments in sequencing allowed that an increasing amount of genetic causes are being unravelled.11 Dilated cardiomyopathy (DCM) is characterised by left ventricular dilation associated with cardiac dysfunction and ventricular arrhythmias.12 Causative genes in DCM predominantly encode cytoskeletal and sarcomeric proteins, but disturbance of calcium cycling also seems to play an important role (Table 1).13, 14 For the purpose of this thesis, we will discuss cardiac calcium homeostasis in more detail. Calcium cycling in human cardiomyocytes is mainly regulated by sarcoplasmic reticulum (SR) Ca2+ ATPase (SERCA2a), a Ca2+ pump which can replenish Ca2+ stores in the SR (Figure 2).15 Lowering of the Ca2+ concentration in the cytosol by SERCA2a causes relaxation of the cardiomyocyte and influences cardiac contractility by the availability of Ca2+ stores for the next beat. The decreased SERCA2a expression in HF has been associated with impaired cardiac contractility and relaxation.16 Phospholamban (PLN) is a small transmembrane protein in the SR, a reversible inhibitor of SERCA2a and in its dephosphorylated state PLN inhibits SERCA2a activity.15 Phospholamban has been derived from phosphate and the Greek word λαμβανειν, which means to receive or to seize.17 Upon phosphorylation, PLN dissociates from SERCA2a, thereby relieving Ca2+ pump inhibition and enhancing relaxation and contractility. If PLN becomes chronically inhibitory, it can lead to diminished contractility and several mutations in PLN have been associated with DCM including the R14del (p.Arg14del) mutation.18 Within the Netherlands a relatively large population of patients exists with the PLN R14del mutation causing HF and it is likely that this mutation arose around 575–825 years ago.19 Estimated 1 in 1400 people is carrier of the R14del mutation in the North of the Netherlands and the total number of Dutch R14del carriers is likely to be more than 2000. The R14del

Table 1. Genes associated with familial dilated cardiomyopathy. Adapted from 14. Mutation group

Gene(s)

Cytoskeletal

DES, DMD, ILK, LAMA4, LDB3, PDLIM3, SGCD, VCL

Sarcomeric

ACTC1, ACTN2, ANKRD1, CSRP3, MYBPC3, MYH6, MYH7, MYPN, TCAP, TNNC1, TNNI3, TNNT2, TPM1, TTN

Calcium cycling

PLN

Nuclear envelope

LMNA, TMPO

Desmosomal

DSC2, DSG2, DSP

Other

PSEN1, PSEN2, ABCC9, SCN5A, TAZ (G4.5), BAG3, CRYAB, EYA4

CHAPTER 1

Figure 2. Calcium cycling in cardiomyocytes. In response to depolarization due to sodium (Na+) influx, calcium (Ca2+) enters the cytosol through the L-type calcium channels (LTCC) in the plasma membrane. This Ca2+ influx leads to calcium-induced calcium release from the sarcoplasmic reticulum mediated by the Ca2+ release channels (ryanodine receptors; RyRs). Calcium binds to troponin in the thin filaments of myofibrils to activate muscle contraction and Ca2+ is removed from the cytosol by sarcoplasmic reticulum Ca2+ ATPase (SERCA2a) and the sodium-calcium exchanger (NCX) on the plasma membrane. Phospholamban (PLN) is a reversible inhibitor of SERCA2a and in its dephosphorylated state PLN inhibits SERCA2a activity. Upon phosphorylation by protein kinase A (PKA), through the Î˛-adrenergic receptor pathway, or Ca2+/calmodulindependent protein kinase (CaMKII), PLN dissociates from SERCA2a, thereby relieving Ca2+ pump inhibition and enhancing cardiac relaxation and contractility. Phospholamban is dephosphorylated by protein phosphatase (PP1), which ends the stimulation phase and PP1 is regulated by inhibitor-1 (I-1).

mutation has been associated with DCM and arrhythmogenic cardiomyopathy. Insights regarding the distribution pattern of fibrosis and fatty replacement in R14del cardiomyopathy patients are limited. Furthermore, several other non-ischaemic cardiomyopathies have been associated with causal mutations and fibrosis is frequently observed in this group. Revealing distribution patterns of cardiac fibrosis in cardiomyopathies is important to elucidate mechanisms of pathophysiology and for cardiac imaging purposes.

INTRODUCTION

Since the discovery of deoxyribonucleic acid (DNA)20 and introduction of methods for DNA sequencing21 novel methods have emerged for high-throughput next-generation sequencing

(NGS).22 State of the art genetic research has the potential to lead to a more fundamental understanding of the underlying pathophysiology related to HF and to identify new therapeutic targets and diagnostic biomarkers. Despite these advances limited genetic loci associated with risk of incident HF have been identified. Epigenetics is defined as the study of heritable changes in gene expression which are not due changes in DNA sequence.23 Epigenetic modifications can make regulatory elements of genes more or less permissive to transcription factors, thereby changing gene expression. Recently, NGS technologies have been developed to profile the â&#x20AC;&#x2DC;regulomeâ&#x20AC;&#x2122;.24 The regulome is a set of DNA transcriptional regulatory elements, e.g., promotors and enhancers. Identification of differentially regulated gene regions in HF could lead to improved understanding of disease mechanisms, risk stratification and targeted therapies. Diagnosis While clinical assessment is the keystone of patient management, clinicians often utilize additional laboratory tests or imaging studies to aid differential diagnosis.25 Cardiac function can be assessed using different imaging modalities, e.g., echocardiography or cardiovascular magnetic resonance imaging (CMR). Precise assessment of cardiac fibrosis is important for diagnosis, predicting prognosis and to guide and monitor therapy.26 The non-invasive reference standard to assess cardiac anatomy and function is CMR26 and local cardiac function can be assessed for example using wall thickening27 or strain analyses28. Fibrosis detection using late gadolinium enhancement (LGE) CMR is possible due to prolonged retention of contrast agent in regions of myocardial fibrosis compared to healthy myocardium, resulting in an increased signal intensity in fibrotic regions using T1 weighted CMR.29 While LGE CMR provides an accurate qualitative measure of replacement fibrosis, it has several disadvantages. Use of gadolinium has potential adverse effects and does not provide a quantitative or direct measurement of cardiac collagen. The LGE method is restricted in assessment of diffuse fibrosis, results differ between different imaging studies and it is less suitable for longitudinal studies by factors affecting reproducibility.26 Possible adverse effects include rare but potentially life-threatening anaphylaxis, contrast induced nephropathy and nephrogenic systemic fibrosis.30 Currently, there is no consensus on the preferred method for LGE quantification although this would be useful for interpreting diagnostic studies, determining prognostic value and measuring therapeutic outcomes. Therefore it would be of interest to compare CMR parameters with histological fibrosis. Recently, several novel CMR techniques have been proposed for fibrosis detection. Contrastenhanced T1-mapping can generate a map representing the T1 relaxation times by directly quantifying T1 values per voxel.31 This method provides quantitative assessment of extra cellular volume (ECV) and could be an improvement over LGE, especially in patients with diffuse fibrosis. Direct detection of myocardial fibrosis might be possible when the acquisition is tuned to the changes in collagen content, could obviate the need for use of contrast agents, reduce patient burden and health costs. Endogenous contrast methods to assess fibrosis with CMR include non-contrast T132 and T1rho mapping33.

CHAPTER 1

Advanced therapy for HF: cardiac regeneration One of the most important modalities to regenerate the failing heart is heart transplantation.34 Unfortunately, the availability of donor organs is a major challenge and heart transplantation comes at a price with accompanying morbidity and mortality.35 Several other therapies have been of interest to induce cardiac regeneration, including cell therapy.36 In theory, cell therapy can be used to repair the failing heart and proposed working mechanisms include formation of new muscle cells and blood vessels.37 Given the lack of evidence regarding transdifferentiation of cells to cardiomyocytes led to the hypothesis that paracrine mechanisms contribute to improvement in cardiac function.38 While stem cell therapy has been extensively studied in preclinical39 and clinical40 studies of ischaemic HF, evidence in the setting of non-ischaemic DCM with myocardial dysfunction seems scant. Thus far, multiple cell types have been studied for cardiac regeneration, but the optimal cardiac regenerative therapy has yet to be discovered. Our research group previously reported the successful isolation of human cardiomyocyte progenitor cells.41 These cells could form functional cardiomyocytes and while this might be a promising regenerative therapy for HF patients, further research is needed before translation to the clinic. Outline of this thesis To overcome the challenges in HF the intent of this thesis is to provide opportunities from bench-to-bedside to elucidate pathways associated within the failing heart with the ultimate aim to improve clinical outcome for patients. Chapter 2 and 3 investigate the incidence of and prognostic factors associated with HF following MI in an observational cohort study, using linked electronic health record (EHR) sources from primary and secondary care in England (Chapter 2) and using patients with follow-up for incident HF after a first STEMI in the Arrhythmia Genetics in the NEtherlandS (AGNES) study (Chapter 3). In Chapter 4, we systematically compare in vivo CMR techniques with myocardial fibrosis assessment. The CMR parameters of interest are LGE CMR, myocardial strain (derived from feature tracking) and wall thickening. In Chapter 5, an overview is presented of novel methods for cardiac fibrosis detection on CMR beyond LGE, in particular methods with endogenous contrast. A novel high resolution systematic digital histological quantification method is proposed in Chapter 6 where patterns of cardiac fibrosis and adipose tissue in PLN R14del cardiomyopathy hearts are studied. In Chapter 7, we use the method previously described in Chapter 6 to determine the distribution pattern of fibrosis in multiple genetic non-ischaemic cardiomyopathies. In Chapter 8, we compare the activity of transcriptional regulatory elements between hearts of patients with the PLN R14del mutation and controls. Finally, Chapter 9 and 10 depict studies related to the translation of cell therapy towards the clinic. Chapter 9 describes a systematic literature review on cell therapy in preclinical and clinical studies for non-ischaemic DCM. In Chapter 10, we study the effects of human cardiomyocyte progenitor cells in a large animal model of chronic ischaemic HF. In summary, we tried to unravel the basic mechanisms leading to HF, but also the pathways relevant for cardiac regeneration.

INTRODUCTION

REFERENCES 1.

2. 3.

6. 7.

8. 9. 10. 11. 12. 13.

14. 15. 16.

17. 18.

19.

20. 21. 22. 23. 24.

McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al. ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC. Eur Heart J. 2012;33:1787-847. Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137-46. Baldasseroni S, Opasich C, Gorini M, Lucci D, Marchionni N, Marini M, et al. Left bundle-branch block is associated with increased 1-year sudden and total mortality rate in 5517 outpatients with congestive heart failure: a report from the Italian network on congestive heart failure. Am Heart J. 2002;143:398-405. Bleumink GS, Knetsch AM, Sturkenboom MC, Straus SM, Hofman A, Deckers JW, et al. Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk and prognosis of heart failure The Rotterdam Study. Eur Heart J. 2004;25:1614-9. Brouwers FP, de Boer RA, van der Harst P, Voors AA, Gansevoort RT, Bakker SJ, et al. Incidence and epidemiology of new onset heart failure with preserved vs. reduced ejection fraction in a community-based cohort: 11-year follow-up of PREVEND. Eur Heart J. 2013;34:1424-31. Velagaleti RS, Pencina MJ, Murabito JM, Wang TJ, Parikh NI, Dâ&#x20AC;&#x2122;Agostino RB, et al. Long-term trends in the incidence of heart failure after myocardial infarction. Circulation. 2008;118:2057-62. Shafazand M, Rosengren A, Lappas G, Swedberg K, Schaufelberger M. Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993-2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden. Eur J Heart Fail. 2011;13:135-41. Weber KT, Sun Y, Bhattacharya SK, Ahokas RA, Gerling IC. Myofibroblast-mediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol. 2013;10:15-26. McMurray JJ. Clinical practice. Systolic heart failure. N Engl J Med. 2010;362:228-38. Lee DS, Pencina MJ, Benjamin EJ, Wang TJ, Levy D, Oâ&#x20AC;&#x2122;Donnell CJ, et al. Association of parental heart failure with risk of heart failure in offspring. N Engl J Med. 2006;355:138-47. Towbin JA, Bowles NE. The failing heart. Nature. 2002;415:227-33. Jefferies JL, Towbin JA. Dilated cardiomyopathy. Lancet. 2010;375:752-62. Harakalova M, Kummeling G, Sammani A, Linschoten M, Baas AF, van der Smagt J, et al. A systematic analysis of genetic dilated cardiomyopathy reveals numerous ubiquitously expressed and muscle-specific genes. Eur J Heart Fail. 2015;17:484-93. Hershberger RE, Hedges DJ, Morales A. Dilated cardiomyopathy: the complexity of a diverse genetic architecture. Nat Rev Cardiol. 2013;10:531-47. MacLennan DH, Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol. 2003;4:566-77. Hasenfuss G, Reinecke H, Studer R, Meyer M, Pieske B, Holtz J, et al. Relation between myocardial function and expression of sarcoplasmic reticulum Ca(2+)-ATPase in failing and nonfailing human myocardium. Circ Res. 1994;75:434-42. Katz AM. Discovery of phospholamban. A personal history. Ann N Y Acad Sci. 1998;853:9-19. Haghighi K, Kolokathis F, Gramolini AO, Waggoner JR, Pater L, Lynch RA, et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc Natl Acad Sci U S A. 2006;103:1388-93. van der Zwaag PA, van Rijsingen IA, de Ruiter R, Nannenberg EA, Groeneweg JA, Post JG, et al. Recurrent and founder mutations in the Netherlands-Phospholamban p.Arg14del mutation causes arrhythmogenic cardiomyopathy. Neth Heart J. 2013;21:286-93. Watson JD, Crick FH. Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature. 1953;171:737-8. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A. 1977;74:5463-7. Metzker ML. Sequencing technologies - the next generation. Nat Rev Genet. 2010;11:31-46. Eccleston A, DeWitt N, Gunter C, Marte B, Nath D. Epigenetics. Nature. 2007;447:395-. Blow MJ, McCulley DJ, Li Z, Zhang T, Akiyama JA, Holt A, et al. ChIP-Seq identification of weakly conserved heart enhancers. Nat Genet. 2010;42:806-10.

CHAPTER 1

25. Vasan RS. Biomarkers of cardiovascular disease: molecular basis and practical considerations. Circulation. 2006;113:2335-62. 26. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA. Assessment of myocardial fibrosis with cardiovascular magnetic resonance. J Am Coll Cardiol. 2011;57:891-903. 27. Holman ER, Vliegen HW, van der Geest RJ, Reiber JH, van Dijkman PR, van der Laarse A, et al. Quantitative analysis of regional left ventricular function after myocardial infarction in the pig assessed with cine magnetic resonance imaging. Magn Reson Med. 1995;34:161-9. 28. Hor KN, Gottliebson WM, Carson C, Wash E, Cnota J, Fleck R, et al. Comparison of magnetic resonance feature tracking for strain calculation with harmonic phase imaging analysis. JACC Cardiovasc Imaging. 2010;3:144-51. 29. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, et al. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation. 1999;100:1992-2002. 30. Bellin MF, Van Der Molen AJ. Extracellular gadolinium-based contrast media: an overview. Eur J Radiol. 2008;66:160-7. 31. Moon JC, Messroghli DR, Kellman P, Piechnik SK, Robson MD, Ugander M, et al. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR Working Group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson. 2013;15:92. 32. Bull S, White SK, Piechnik SK, Flett AS, Ferreira VM, Loudon M, et al. Human non-contrast T1 values and correlation with histology in diffuse fibrosis. Heart. 2013;99:932-7. 33. Witschey WR, Zsido GA, Koomalsingh K, Kondo N, Minakawa M, Shuto T, et al. In vivo chronic myocardial infarction characterization by spin locked cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2012;14:37. 34. Chien KR. Regenerative medicine and human models of human disease. Nature. 2008;453:302-5. 35. Yacoub M. Heart transplantation: the end of the beginning. Am J Transplant. 2008;8:1767-8. 36. Segers VF, Lee RT. Stem-cell therapy for cardiac disease. Nature. 2008;451:937-42. 37. Bartunek J, Dimmeler S, Drexler H, Fernandez-Aviles F, Galinanes M, Janssens S, et al. The consensus of the task force of the European Society of Cardiology concerning the clinical investigation of the use of autologous adult stem cells for repair of the heart. Eur Heart J. 2006;27:1338-40. 38. Maxeiner H, Krehbiehl N, Muller A, Woitasky N, Akinturk H, Muller M, et al. New insights into paracrine mechanisms of human cardiac progenitor cells. Eur J Heart Fail. 2010;12:730-7. 39. van der Spoel TI, Jansen of Lorkeers SJ, Agostoni P, van Belle E, Gyongyosi M, Sluijter JP, et al. Human relevance of pre-clinical studies in stem cell therapy: systematic review and metaanalysis of large animal models of ischaemic heart disease. Cardiovasc Res. 2011;91:649-58. 40. Jeevanantham V, Butler M, Saad A, Abdel-Latif A, Zuba-Surma EK, Dawn B. Adult bone marrow cell therapy improves survival and induces long-term improvement in cardiac parameters: a systematic review and meta-analysis. Circulation. 2012;126:551-68. 41. Goumans MJ, de Boer TP, Smits AM, van Laake LW, van Vliet P, Metz CH, et al. TGF-beta1 induces efficient differentiation of human cardiomyocyte progenitor cells into functional cardiomyocytes in vitro. Stem Cell Res. 2007;1:138-49.

PART ONE

Clinical Course Portrayed

Chapter

Heart Failure Following Myocardial Infarction: a Cohort Study of Incidence and Prognostic Factors in 24 745 Patients Using Linked Electronic Records In preparation

Johannes M.I.H. Gho1,2, Amand F. Schmidt1, Laura Pasea1, Stefan Koudstaal1,2, Mar Pujades-Rodriguez1, Spiros Denaxas1, Anoop D. Shah1, Riyaz S. Patel1,3, Chris P. Gale4, Arno W. Hoes5, John G. Cleland6, Harry Hemingway1, Folkert W. Asselbergs1,2,3,7

Farr Institute of Health Informatics Research, UCL Institute of Health Informatics,

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht,

Institute of Cardiovascular Science, University College London, London, UK

University of Leeds, Leeds, UK

Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht,

Imperial College London, London, UK

Durrer Center for Cardiogenetic Research, ICIN-Netherlands Heart Institute, Utrecht,

University College London, London, UK Utrecht, the Netherlands

Utrecht, the Netherlands

the Netherlands

PART ONE CHAPTER 2

ABSTRACT Aims To investigate the incidence of and prognostic factors associated with heart failure (HF) after myocardial infarction (MI). Methods We used primary care data from England, linked to national datasets providing information on hospital admissions, an MI registry and mortality as part of the CALIBER (CArdiovascular research using LInked Bespoke studies and Electronic health Records) programme. All patients aged 18 years or older who experienced a first MI between 1 January 1998 and 25 March 2010 were included. Heart failure was defined as a HF diagnosis in any of the CALIBER sources and used as outcome in Cox proportional hazard models. Results Of 24,745 patients with a first MI and without a prior history of HF (median follow-up of 3.7 years), 6005 (24.3%) patients developed HF, which represents a crude incidence rate of HF following a first MI of 66.1 cases (95% CI: 64.4 – 67.8) per 1000 person-years. Independent baseline factors associated with subsequent HF were: age [HR per 10 years increase: 1.45 (95%CI 1.41 – 1.49)], higher socioeconomic deprivation (5th [HR 1.27 (95%CI 1.13 – 1.42)] vs. 1st quintile), a history of hypertension [HR 1.16 (95%CI 1.09 – 1.23)], diabetes [HR 1.44 (95%CI 1.34 – 1.55)], atrial fibrillation [HR 1.63 (95%CI 1.51 – 1.75)], peripheral arterial disease [HR 1.38 (95%CI 1.26 – 1.51)], COPD [HR 1.28 (95%CI 1.17 – 1.40)] and a STEMI at presentation [HR 1.21 (95%CI 1.11 – 1.32)]. Conclusion One in 4 people developed HF within 4 years of experiencing a first MI. Key clinical characteristics readily identified patients at risk of HF following MI which may, therefore, be used to guide interventions to reduce cardiovascular disease burden.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

INTRODUCTION In the western world heart failure (HF) is a major medical problem, carrying a high morbidity and mortality risk and is in most cases secondary to ischaemic heart disease.1 Research studying the incidence of HF following myocardial infarction (MI) is limited, mainly originating from the

thrombolytic era. Torabi et al. described the natural history and prognosis of HF after MI in 1998 among 1000 British patients to show that 20.3% developed HF during their index MI admission and 33% after discharge preceding death.2 A Danish nationwide cohort study reported a temporal decrease in 90-day HF incidence after MI, from 23.6% in 1997-98 to 19.6% in 2009-10.3 A more recent Swedish study found a decrease in HF incidence from 46% in 1996-1997 to 28% in 2008 during hospitalization for index MI.4 The availability of prospectively collected electronic health records (EHRs) offers a unique opportunity to conduct clinical research.5 Moreover, in England linked EHRs provide a more accurate estimation of incident cardiovascular events.6 Prospective collected data of large cohorts using linked EHRs from primary and secondary care is lacking regarding the incidence and prognostic factors associated with HF after a first MI in the current era. The aim of this research was, therefore, to determine using linked EHRs of population-based data from primary and secondary care the incidence of and prognostic factors associated with HF following MI.

METHODS Study design Record-linkage cohort study of data sources within the CALIBER (CArdiovascular disease research using LInked Bespoke studies and Electronic health Records) dataset, which includes linked data from (i) primary care EHR data using the diagnoses coded with the Read system (Clinical Practice Research Datalink, CPRD) from 244 consenting general practices,7 (ii) secondary care administrative records (information coded using the International Classification of Diseases (ICD) and the Office of Population Censuses and Surveys Classification of Interventions and Procedures (OPCS-4), hospital records (Hospital Episode Statistics, HES)), the Myocardial Ischaemia National Audit Project (MINAP) and (iii) the mortality register (from UK Office for National Statistics, ONS).8 Setting and participants All patients aged â&#x2030;Ľ18 years old, registered in CPRD practices in England consenting to data linkage, with at least one year of up-to-standard pre-study follow-up, and who experienced a first MI (see Supplementary for MI definition) between Jan 1, 1998, to March 25, 2010 were potentially eligible for inclusion.6 The first record of MI during the patientâ&#x20AC;&#x2122;s study period and across the linked data sources was considered as the index event and subsequent MI records within 30 days in the other data sources were considered as representing the same event. Patients with a fatal index MI or a history of HF before their index MI were excluded. Patients were censored at de-registration from the general practice, the date of death or the administrative censoring of the dataset (March 25, 2010).

PART ONE CHAPTER 2

Variables and data sources The variables used in this study have been previously defined, were derived from Read, ICD-9 or ICD-10, MINAP, Multilex drug or procedure codes, and the EHR phenotyping algorithms can be found online (https://caliberresearch.org/portal). For continuous variables, the most recent measurement recorded in CPRD in the year before study entry was used as a baseline value. Data before study entry were used to determine prognostic factors, including age (years), sex, ethnicity, social deprivation (index of multiple deprivation, IMD score as recorded in CPRD, in quintiles), smoking, alcohol use, history of cardiovascular disease, previous revascularization, history of diabetes, history of thyroid disease, history of chronic obstructive pulmonary disease (COPD) and history of non-metastatic cancer. Data in the year before study entry was collected, including body mass index (BMI), vital signs before hospitalisation, biomarkers and medications prescribed before index MI. Data from MINAP were used to determine infarct characteristics, including: type of MI, site of infarction, type of reperfusion therapy, delay from symptom onset to delivery of reperfusion therapy and peak cardiac biomarkers. The primary study outcome was the first recorded HF event following MI as previously defined and published online (https:// www.caliberresearch.org/portal/show/phenotype_hf). Events were defined in primary care by a diagnosis of HF or notification of left ventricular dysfunction on echocardiogram, in secondary care by a diagnosis of HF or mortality due to HF or in ONS by mortality due to HF (see Supplementary for HF definition). The first recorded event date that satisfied the pre-defined definition was used as the incident date of HF from different sources. Statistical methods Incidence rates (cases per 1000 person-years) and Kaplan-Meier cumulative incidence (cumulative percentage) of HF were calculated. Kaplan-Meier curves were used to assess unadjusted survival differences between incident HF after index MI, by age (<50, 50-65, ≥65 years) and type of MI. Cumulative incidence of HF was then adjusted for competing risk of mortality as first event.9 The associations of baseline variables with the onset of HF following MI were explored using Cox proportional hazard models. The proportional hazards assumption was checked graphically and using a chi-square test, throughout slight deviations were seen at the end of follow-up which could readily be approximated by additive time. All models were stratified for general practice and calendar period (3 year periods between 1998 and 2010). Models were sequentially adjusted for: (1) age and sex, (2) cardiovascular risk factors, (3) type of MI, and (4) co-morbidities and prescribed medication. Associations are presented as hazard ratio’s (HRs) with 95% confidence intervals (95%CI) and two-sided p-values using an alpha of 0.05. Analyses were performed using R (version 3.1.2). Sensitivity analyses Missing data for baseline smoking, BMI, and systolic and diastolic blood pressure were multiply imputed using the mice package,10 and results across 20 imputed datasets were pooled using Rubin’s rules (Supplementary).11 Finally, to account for the fact that patients may have died before developing HF after a first MI (e.g., competing risk by mortality), all analyses were repeated using Fine and Gray models.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Ethical approval CALIBER was approved by the Lewisham Local Research Ethics Committee (ref: 09/H0810/16 date: 08/04/2009) and the Ethics and Confidentiality Committee (ECC) (ref: ECC 2-06(b)/2009 CALIBER dataset). CALIBER has been registered with the University College London Data Protection Officer (ref: Z6364106/2009/2/26). CALIBER EHR data are anonymised; individual

informed consent was not sought from study participants. This study was approved by the ISAC (Independent Scientific Advisory Committee) for MHRA Database Research (protocol no. 14_198R) and the MINAP Academic Group. This study is registered with ClinicalTrials.gov, number NCT02384213.

RESULTS Subjects and baseline characteristics We identified 52,270 patients with a first MI recorded in CALIBER between 1998 and 2010 (Figure 1). After exclusion of patients with a fatal index MI (n = 15,104), a prior history of HF (n = 3392) or those with insufficient baseline data (n = 9029) a cohort of (n = 24,745) index MI patients were included in primary analyses. This cohort comprised 4547 patients with STsegment-elevation MI (STEMI), 6892 with Non-STEMI (NSTEMI) and 13,306 with unclassified MI. Baseline characteristics are described in Table 1. Of all the MI patients 16,134 (65.2%) were men, 4594 (42.2%) were current smokers, 12,409 (50.1%) had a history of hypertension and 3056 (12.3%) had a history of diabetes at baseline. In total, 6154 (24.9%) were prescribed betablockers and 5062 (20.5%) ACE inhibitors before their index MI. Incidence Patients contributed 90,817 person-years of follow-up after MI to the study. During a median time of 3.7 years (IQR: 1.4 â&#x20AC;&#x201C; 6.7), 6005 patients (24.3%) developed HF. The crude incidence rate of HF following a first MI was 66.1 cases (95% CI: 64.4 - 67.8) per 1000 person-years. During the first 30 days from index MI 2581 (10.4%) patients developed HF (Figure 2A). In patients event free during the first 30 days, from day 30 onwards 2779/21286 (13.1%) of MI patients developed HF during 5-year follow-up (Figure 2B). Death occurred at the time of first HF presentation for 137 (2.3%) patients. The 30-day HF incidence was 5.2% (126/2409) in patients younger than 50 years, 6.5% (498/7624) in 50-65 year olds and 13.3% (1957/14712) in those 65 years and older (Figure 3A). During 5-year follow-up in patients event free during the first 30 days, subsequent cumulative estimates per age group were 5.9% (131/2208), 7.6% (526/6948) and 17.5% (2122/12130), respectively (Figure 3B). The 30-day incidence of HF was similar between NSTEMI (922/6892, 13.4%) and STEMI (610/4547, 13.4%) patients, but lower in the unclassified MI subgroup (1049/13306, 7.9%) (Figure 3C). Among patients who were event free during the first 30 days, a higher incidence of HF at 5 years was observed amongst NSTEMI (692/5746, 12.0%) compared with STEMI (384/3812, 10.1%) subgroup (Figure 3D).

PART ONE CHAPTER 2

Figure 1. Flowchart of study population CALIBER = Cardiovascular disease research using Linked Bespoke studies and Electronic health Records; CPRD = Clinical Practice Research Datalink; HF = heart failure; MI = myocardial infarction; NSTEMI = NonST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction.

Figure 2. Crude incidence of HF after index MI Kaplan-Meier curves for crude incidence of HF after MI. A. 30 day follow-up after index MI. B. 5 year followup in patients who survived the first 30 days and did not develop HF during the first 30 days (30 days event free). HF = heart failure; MI = myocardial infarction.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Figure 3. Crude incidence of HF after index MI by subgroup Kaplan-Meier curves for crude incidence of HF after MI stratified by subgroup. A. 30 day follow-up after index MI stratified by age group. B. 5 year follow-up in 30 day event free patients stratified by age group. Log-rank test of >=65 year old patients compared to patients aged 50-65 years p < 0.001. C. 30 day follow-up after index MI stratified by MI subtype. D. 5 year follow-up in patients who survived the first 30 days and did not develop HF during the first 30 days stratified by MI subtype. Log-rank test of NSTEMI compared to STEMI p < 0.001. HF = heart failure; MI = myocardial infarction; NSTEMI = Non-ST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction.

Cumulative incidence in competing risks data At 30 days, 10.4% (2581/24735) developed HF and only one patient died without developing HF. Of patients event free during the first 30 days, from day 30 onwards 12.7% (2823/22147) developed HF and 4.7% (1048) died without first being diagnosed with HF (Figure 4). Of the patients with an index MI between 1998-2001 17.9% (938/5229) developed HF compared with 21.6% (1496/6939) of patients with MI between 2004-2007 during a 3-year follow-up. The estimated 3-year cumulative HF incidence in MI patients age 65 years and over between 19982001 was 23.5% (2977) compared with 27.2% (4219) between 2004-2007.

PART ONE CHAPTER 2

Table 1. Baseline characteristics at index myocardial infarction STEMI patients NSTEMI n = 4547 patients n = 6892

MI patients not Total MI further patients specified n = 24745 n = 13306

Unknown (%)

Follow-up time (years), median (IQR)

2.8 (1.2–4.6)

2.3 (0.9–4.1)

5.9 (2.3–8.7)

3.7 (1.4–6.7)

Mean age, years (SD)

65.5 (13.4)

70.5 (13.6)

67.8 (12.8)

68.1 (13.2)

Male sex

3243 (71.3%)

4160 (60.4%)

8731 (65.6%)

16134 (65.2%)

White

3407 (77.4%)

5281 (79.1%)

9240 (74%)

17928 (76.1%)

4.8%

Asian

85 (1.9%)

104 (1.6%)

214 (1.7%)

403 (1.7%)

Black

19 (0.4%)

29 (0.4%)

43 (0.3%)

91 (0.4%)

Other

334 (7.6%)

456 (6.8%)

718 (5.8%)

1508 (6.4%)

Underweight

25 (1.8%)

60 (2.3%)

58 (1.7%)

143 (1.9%)

Normal

360 (26.6%)

720 (27.2%)

920 (27.2%)

2000 (27.1%)

Overweight

557 (41.1%)

1043 (39.4%)

1400 (41.4%)

3000 (40.7%)

Obese

412 (30.4%)

824 (31.1%)

1000 (29.6%)

2236 (30.3%)

918 (20.3%)

1282 (18.7%)

2731 (20.6%)

4931 (20%)

0.4%

Current smoker

1360 (50.5%)

1380 (33%)

1854 (46.3%)

4594 (42.2%)

56%

Excess alcohol

63 (9.4%)

113 (8.5%)

120 (8.6%)

296 (8.7%)

86.3%

History of atrial fibrillation

353 (7.8%)

1058 (15.4%)

1208 (9.1%)

2619 (10.6%)

n/a

History of hypertension

2095 (46.1%)

4060 (58.9%)

6254 (47%)

12409 (50.1%)

n/a

History of peripheral arterial disease

233 (5.1%)

598 (8.7%)

911 (6.8%)

1742 (7%)

n/a

PCI

757 (16.6%)

715 (10.4%)

1295 (9.7%)

2767 (11.2%)

n/a

CABG

116 (2.6%)

320 (4.6%)

538 (4%)

974 (3.9%)

n/a

Previous TIA

135 (3%)

401 (5.8%)

632 (4.7%)

1168 (4.7%)

n/a

Previous stroke

78 (1.7%)

185 (2.7%)

234 (1.8%)

497 (2%)

n/a

History of diabetes

511 (11.2%)

1056 (15.3%)

1489 (11.2%)

3056 (12.3%)

n/a

History of thyroid disease

254 (5.6%)

597 (8.7%)

776 (5.8%)

1627 (6.6%)

n/a

History of COPD

302 (6.6%)

683 (9.9%)

973 (7.3%)

1958 (7.9%)

n/a

History of non-metastatic cancer

483 (10.6%)

970 (14.1%)

1335 (10%)

2788 (11.3%)

n/a

Heart rate, beats/min

73.0 (64–84)

72.0 (64–84)

75.5 (66–84)

73 (64–84)

91.1%

Systolic blood pressure, mmHg

140 (130–150)

140 (129–150)

140 (130–158)

140 (130–154)

34.6%

Diastolic blood pressure, mmHg

80 (74–87)

80 (70–85)

80 (73–90)

80 (72–88)

34.6%

Ethnicity

Body mass index (BMI) 70.2%

Index of multiple deprivation Most deprived quintile Risk factors before index MI

Previous revascularization

Vital signs before admission median (IQR)

Biomarkers median (IQR)

Troponin I (maximum)

19.38 (3.8–50.0) 1.73 (0.29–7.4)

Troponin T (maximum)

1.86 (0.6–4.8)

CK (maximum)

740 (226.0–1687) 189 (89.0–512)

0.30 (0.19–1.02) 3.88 (0.62–21.4) 88.1%

0.33 (0.13–0.87) 0.23 (0.12–0.24) 0.65 (0.18–2) 122 (63.2–148)

331 (123–1068)

90.6% 79.7%

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Table 1. Continued STEMI patients NSTEMI n = 4547 patients n = 6892

MI patients not Total MI further patients specified n = 24745 n = 13306

Unknown (%)

Haemoglobin, g/dL

14.1 (1.72)

13.5 (1.87)

13.7 (1.81)

13.7 (1.83)

67.3%

White blood cell count

8.06 (2.96)

7.83 (2.79)

7.93 (3.04)

7.91 (2.92)

69.3%

Neutrophil count

5.03 (2.13)

4.97 (2.32)

5.01 (2.48)

5 (2.35)

72.1%

Platelets

272 (84.6)

267 (89.5)

264 (86.6)

267 (87.5)

69.3%

Erythrocyte sedimentation rate

18.6 (18.2)

22.4 (21.7)

22.4 (21.8)

21.8 (21.2)

89.5%

Creatinine, Âľmol/L

99.3 (40.9)

102.1 (47.4)

102.6 (47.1)

102 (46.1)

59%

eGFR - CKD-EPI

70.6 (19.6)

66.1 (20.3)

66.2 (19.6)

67 (20)

60.7%

Random glucose concentration, mmol/L

6.90 (3.36)

6.95 (3.34)

7.61 (4.33)

7.23 (3.82)

76.7%

Fasting glucose concentration, mmol/L 6.07 (2.02)

6.00 (1.99)

6.50 (2.62)

6.18 (2.22)

92.8%

HbA1c, mmol/mol

60.4 (19.8)

57.6 (16.8)

59.0 (17.8)

58.6 (17.8)

90%

Total cholesterol

5.33 (1.42)

5.04 (1.25)

5.43 (1.29)

5.27 (1.31)

63.7%

LDL cholesterol

3.24 (1.11)

2.93 (1.08)

3.23 (1.11)

3.1 (1.11)

81.4%

Biomarkers before index MI mean (SD)

STEMI Site of infarction Anterior

518 (40.5%)

71.8%

Primary PCI

332 (19.4%)

62.4%

Prehospital fibrinolysis

432 (9.5%)

n/a

In-hospital fibrinolysis

1030 (60.3%)

62.4%

Median (IQR) delay from symptom to reperfusion (min)

150 (97â&#x20AC;&#x201C;280)

37.5%

Prescribed medication before index MI Antiplatelet

815 (17.9%)

2516 (36.5%)

4555 (34.2%)

7886 (31.9%)

n/a

Oral anticoagulant

69 (1.5%)

260 (3.8%)

363 (2.7%)

692 (2.8%)

n/a

Statin

882 (19.4%)

2269 (32.9%)

3006 (22.6%)

6157 (24.9%)

n/a

ACE inhibitor

692 (15.2%)

1643 (23.8%)

2727 (20.5%)

5062 (20.5%)

n/a

Angiotensin receptor blocker (ARB)

250 (5.5%)

611 (8.9%)

575 (4.3%)

1436 (5.8%)

n/a

Beta-blocker

729 (16%)

1782 (25.9%)

3643 (27.4%)

6154 (24.9%)

n/a

Calcium channel blocker

786 (17.3%)

1877 (27.2%)

2791 (21%)

5454 (22%)

n/a

Loop diuretic

227 (5%)

850 (12.3%)

1380 (10.4%)

2457 (9.9%)

n/a

Aldosterone antagonist

25 (0.5%)

77 (1.1%)

121 (0.9%)

223 (0.9%)

n/a

Digoxin

50 (1.1%)

210 (3%)

309 (2.3%)

569 (2.3%)

n/a

IQR = inter-quartile range; SD = standard deviation; MI = myocardial infarction; NSTEMI = non-ST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction; PCI = Percutaneous Coronary Intervention; CABG = Coronary Artery Bypass Grafting; TIA = Transient Ischaemic Attack; COPD = Chronic Obstructive Pulmonary Disease; CK = Creatine kinase; ACE = Angiotensin-converting-enzyme. Prehospital and in-hospital fibrinolysis are not mutually exclusive.

PART ONE CHAPTER 2

Table 2. Hazard ratio for heart failure in patients following a first myocardial infarction using multivariable Cox regression Model 1 n = 24458 n events = 5961

Model 2 n = 24458 n events = 5961

Model 3 n = 24458 n events = 5961

Model 4 n = 24458 n events = 5961

HR (95% CI)

Age, per 10 years

1.51 (1.48 – 1.55)

1.49 (1.46 – 1.53)

1.46 (1.43 – 1.50)

1.45 (1.41 – 1.49)

Men

1.06 (1.00 – 1.12)

1.08 (1.02 – 1.14)

1.07 (1.01 – 1.13)

1.06 (1.00 – 1.12)

Index of multiple deprivation

overall p = <0.001 overall p = <0.001 overall p = <0.001

Q1 (least deprived)

Reference

1.09 (0.99 – 1.20)

1.08 (0.99 – 1.19)

1.08 (0.98 – 1.19)

1.20 (1.09 – 1.33)

1.19 (1.07 – 1.31)

1.20 (1.08 – 1.33)

1.17 (1.06 – 1.30)

Q5 (most deprived)

1.30 (1.16 – 1.45)

1.30 (1.17 – 1.46)

1.27 (1.13 – 1.42)

History of hypertension

1.19 (1.12 – 1.26)

1.17 (1.10 – 1.23)

1.16 (1.09 – 1.23)

History of diabetes

1.47 (1.37 – 1.58)

1.47 (1.37 – 1.59)

1.44 (1.34 – 1.55)

History of atrial fibrillation

1.65 (1.53 – 1.77)

1.63 (1.51 – 1.75)

Type of MI

overall p = <0.001 overall p = <0.001

NSTEMI

Reference

Unclassified MI vs. NSTEMI

0.89 (0.82 – 0.97)

STEMI vs. NSTEMI

1.19 (1.10 – 1.30)

1.21 (1.11 – 1.32)

History of peripheral arterial disease

1.38 (1.26 – 1.51)

History of COPD

1.28 (1.17 – 1.40)

Prescribed ACE inhibitor before MI

1.07 (1.00 – 1.14)

Prescribed Angiotensin receptor blocker (ARB) use before MI

1.00 (0.89 – 1.12)

Prescribed Beta-blocker before MI

0.94 (0.88 – 1.00)

HR = hazard ratio; CI = confidence interval; MI = myocardial infarction; NSTEMI = non-ST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction; COPD = chronic obstructive pulmonary disease; ACE = Angiotensin-converting enzyme.

Prognostic factors In multivariable Cox regression with complete-case analysis (n = 24,458), increasing age [HR per 10 years increase: 1.45 (95%CI 1.41 – 1.49)], higher socioeconomic deprivation (3rd [HR 1.19 (95%CI 1.07 – 1.31)] – 5th [HR 1.27 (95%CI 1.13 – 1.42)] vs. 1st quintile), a history of hypertension [HR 1.16 (95%CI 1.09 – 1.23)], diabetes [HR 1.44 (95%CI 1.34 – 1.55)], atrial fibrillation [HR 1.63 (95%CI 1.51 – 1.75)], peripheral arterial disease [HR 1.38 (95%CI 1.26 – 1.51)], COPD [HR 1.28 (95%CI 1.17 – 1.40)] and an index STEMI presentation [HR 1.21 (95%CI 1.11 – 1.32)] were all associated with increased ratios of HF (Table 2). These clinical factors also predicted outcomes in competing risks analysis, with the difference of an unclassified MI [HR 1.17 (95%CI 1.10 – 1.24)] and beta-blocker prescriptions in the year prior to index MI [HR 0.93 (95%CI 0.87 – 0.99)].

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

In sensitivity analyses, the results of the adjusted complete-case analysis (n = 4911) were similar to those obtained from the multiple imputation analysis except that we found a negative association of being overweight at index MI [HR: 0.82 (95%CI 0.75 – 0.90)], a positive association of smoking [HR: 1.17 (95%CI 1.07 – 1.28)] and no association of systolic [HR: 1.00 (95%CI 1.00 – 1.00)] or diastolic blood pressure [HR: 1.00 (95%CI 1.00 – 1.00)] on admission

(Supplementary).

Figure 4. Cumulative incidence curves for HF adjusted for mortality as first event after index MI Cumulative incidence curves adjusted for competing risk of mortality as first event. 5-year follow-up in patients who survived the first 30 days and did not develop HF during the first 30 days (30 days event free).

DISCUSSION In this large population-based longitudinal study using linked EHRs data from primary care, hospital admissions, a MI registry and the death register, one quarter of patients who experienced a first MI developed HF during a median follow-up of 3.7 years. The incidence rate of HF following a first MI was 66 cases per 1000 person-years. Moreover, the 3-year incidence was high - over a quarter of patients aged 65 years or older developed HF. Increasing age, diabetes and atrial fibrillation, were important prognostic factors associated with the risk of incident HF following MI. HF incidence We found that HF incidence following MI is high in the first 30 days after index MI (10.4%). From day 30 onwards, there was a more gradual increase in the proportion of HF patients to 13.1% during the 5-year follow-up. The cumulative incidence of HF following MI derived from competing risk analyses was similar. These findings are comparable with the decreasing trend

PART ONE CHAPTER 2

seen in the incidence of HF patients after myocardial infarction from another study by comparing within 30 days of admission, 30-days to 1 year and 1-3 years follow-up.12 There is uncertainty whether the HF incidence is increasing13, 14 or decreasing3, 4, 12, 15, 16 in calendar time, the majority of these studies had short-term follow-up and did not include patients managed in primary care. Ezekowitz et al. found a relative 25% increase comparing 1994 to 2000 5-year in-hospital HF rate in an elderly (â&#x2030;Ľ65 years of age) MI patient cohort.14 Using a national sample of Medicare beneficiaries in the United States of America, the HF hospitalization following acute MI decreased from 16.1 per 100 person-years in 1998 to 14.2 per 100 person years in 2010 (p<0.001).16 We did not observe an altered HF incidence in this study that might be elucidated due to capturing milder HF cases, managed in primary care, from multiple linked sources. The low cumulative incidence of mortality as first event during the first 30 days can be attributed to exclusion of patients with a fatal index MI. We found that 28.9% (n = 15,104) of total MIâ&#x20AC;&#x2122;s were fatal which was in line with our broader definition, in comparison to death within seven days of the index MI6 we used a 30-day interval to define fatal MI. Our findings regarding 3-year cumulative HF incidence are in line with the overall HF incidence after a first MI from 2002-2004 (35-64 year: 11.52% and 65-84 year: 27.96%) in a Swedish study (1993-2004).12 They also found a higher incidence of HF in women (6%) compared to men, where we did not found a significant association of gender with HF incidence. We know that the uptake of effective treatment for STEMI, such as primary PCI was slower in the UK than in Sweden.17 This could explain the relatively low percentage of primary PCI for STEMI in our study (19.4%). The higher cumulative incidence in NSTEMI compared to STEMI patients might be explained by the on average higher age with co-morbidities in NSTEMI compared to STEMI patients. Prognostic factors In a previous UK study incidence of HF after MI during 1998 also steeply increased with age.18 In the latter study, of patients <65 years of age half developed HF during a 6-year follow-up compared to 73% of patients aged 65-75 years and 87% of patients aged >75 years. Socioeconomic deprivation has previously been shown to be an independent predictor of HF development and associated with an increased incidence of HF.19-21 Socioeconomic deprivation has also been associated with more frequent hospital admission and higher mortality in patients with HF.22, 23 In a previous CALIBER study with linked EHRs the hazards of HF increased linearly with higher socioeconomic deprivation in patients initially free of CVD.24 In the current study, we observed that higher socioeconomic deprivation in patients with MI also increased the risk for incident HF. Several other studies also previously linked co-morbidities before admission hypertension, diabetes, atrial fibrillation, peripheral arterial disease and COPD to an increased risk of HF following MI.12, 14, 25-27 In a previous multivariable analysis, users of beta-blockers after the index MI were less likely to develop HF.14 In competing risks analyses, we found that betablocker use in the year before MI is an independent predictor for a lower risk of HF after MI. In summary, these prognostic factors can be used to identify patients suitable for clinical trials of novel therapies to target HF following MI.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Strengths and limitations The linkage of multiple EHR sources from primary and secondary care leads to a more accurate estimation regarding incidence and prognostic factors associated with HF following MI. This allowed for â&#x20AC;&#x2DC;real worldâ&#x20AC;&#x2122; data from a relatively large sample size with contemporary therapy and to adjust for a large range of possible confounders. The CPRD consists of a sample of English

general practices, but seem representative of the UK population.28 The ascertainment of cardiovascular outcomes was not based on clinical criteria (e.g., validated questionnaires and properly conducted physical examinations) and there may have been changes in medical coding over time. However, we used stringent code lists to determine variables and outcomes, which have been previously defined, validated and published (https://caliberresearch.org/portal).6-8 Using the first MI recorded in the database without a prior history of HF, might have introduced bias due to left truncation or selection bias.29 There is a possible delay between primary care and secondary care records, as we know from previous research that CPRD tends to record MI after HES or MINAP admission dates.6 The lower 30-day HF incidence in unclassified MI patients primarily derived from primary care could be partly explained because of a delay in coding. Therefore we showed cumulative incidence rates in patients (alive and HF event free within the first 30 days) from 30 days after index MI to account for a delay in recording of MI in primary care. We were unable to differentiate between HF with preserved ejection fraction (HFpEF) and HF with reduced ejection fraction (HFrEF) as we had no access to detailed (echocardiographic) parameters to assess diastolic dysfunction, but it is likely that the majority of our HF patients had developed systolic dysfunction after MI.

CONCLUSION In this large cohort study using linked EHRs in the UK from primary and secondary care we observed that about 1 in 4 people will develop HF within 4 years after the occurrence of a first MI and increasing age, higher socioeconomic deprivation, a history of hypertension, diabetes, atrial fibrillation, peripheral arterial disease or COPD and a STEMI at presentation were important risk factors. Identified amendable prognostic factors can be used to decrease the incidence of HF following MI.

PART ONE CHAPTER 2

REFERENCES 1. 2.

5. 6.

7. 8.

9. 10. 11.

12.

13.

14.

15. 16.

17.

18. 19.

Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137-46. Torabi A, Cleland JG, Khan NK, Loh PH, Clark AL, Alamgir F, et al. The timing of development and subsequent clinical course of heart failure after a myocardial infarction. Eur Heart J. 2008;29:859-70. Gjesing A, Gislason GH, Kober L, Gustav Smith J, Christensen SB, Gustafsson F, et al. Nationwide trends in development of heart failure and mortality after first-time myocardial infarction 1997-2010: A Danish cohort study. Eur J Intern Med. 2014;25:731-8. Desta L, Jernberg T, Lofman I, Hofman-Bang C, Hagerman I, Spaak J, et al. Incidence, Temporal Trends, and Prognostic Impact of Heart Failure Complicating Acute Myocardial Infarction: The SWEDEHEART Registry (Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies): A Study of 199,851 Patients Admitted With Index Acute Myocardial Infarctions, 1996 to 2008. JACC Heart Fail. 2015;3:234-42. Hripcsak G, Albers DJ. Next-generation phenotyping of electronic health records. J Am Med Inform Assoc. 2013;20:117-21. Herrett E, Shah AD, Boggon R, Denaxas S, Smeeth L, van Staa T, et al. Completeness and diagnostic validity of recording acute myocardial infarction events in primary care, hospital care, disease registry, and national mortality records: cohort study. BMJ. 2013;346:f2350. Herrett E, Gallagher AM, Bhaskaran K, Forbes H, Mathur R, van Staa T, et al. Data Resource Profile: Clinical Practice Research Datalink (CPRD). Int J Epidemiol. 2015. Denaxas SC, George J, Herrett E, Shah AD, Kalra D, Hingorani AD, et al. Data resource profile: cardiovascular disease research using linked bespoke studies and electronic health records (CALIBER). Int J Epidemiol. 2012;41:1625-38. Satagopan JM, Ben-Porat L, Berwick M, Robson M, Kutler D, Auerbach AD. A note on competing risks in survival data analysis. Br J Cancer. 2004;91:1229-35. van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res. 2007;16:219-42. Marshall A, Altman DG, Holder RL, Royston P. Combining estimates of interest in prognostic modelling studies after multiple imputation: current practice and guidelines. BMC Med Res Methodol. 2009;9:57. Shafazand M, Rosengren A, Lappas G, Swedberg K, Schaufelberger M. Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993-2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden. Eur J Heart Fail. 2011;13:135-41. Velagaleti RS, Pencina MJ, Murabito JM, Wang TJ, Parikh NI, Dâ&#x20AC;&#x2122;Agostino RB, et al. Long-term trends in the incidence of heart failure after myocardial infarction. Circulation. 2008;118:205762. Ezekowitz JA, Kaul P, Bakal JA, Armstrong PW, Welsh RC, McAlister FA. Declining in-hospital mortality and increasing heart failure incidence in elderly patients with first myocardial infarction. J Am Coll Cardiol. 2009;53:13-20. Fox KA, Steg PG, Eagle KA, Goodman SG, Anderson FA, Jr., Granger CB, et al. Decline in rates of death and heart failure in acute coronary syndromes, 1999-2006. JAMA. 2007;297:1892-900. Chen J, Hsieh AF, Dharmarajan K, Masoudi FA, Krumholz HM. National trends in heart failure hospitalization after acute myocardial infarction for Medicare beneficiaries: 1998-2010. Circulation. 2013;128:2577-84. Chung SC, Gedeborg R, Nicholas O, James S, Jeppsson A, Wolfe C, et al. Acute myocardial infarction: a comparison of short-term survival in national outcome registries in Sweden and the UK. Lancet. 2014;383:1305-12. Torabi A, Cleland JG, Rigby AS, Sherwi N. Development and course of heart failure after a myocardial infarction in younger and older people. J Geriatr Cardiol. 2014;11:1-12. McAlister FA, Murphy NF, Simpson CR, Stewart S, MacIntyre K, Kirkpatrick M, et al. Influence of socioeconomic deprivation on the primary care burden and treatment of patients with a diagnosis of heart failure in general practice in Scotland: population based study. BMJ. 2004;328:1110.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

20. Hawkins NM, Jhund PS, McMurray JJ, Capewell S. Heart failure and socioeconomic status: accumulating evidence of inequality. Eur J Heart Fail. 2012;14:138-46. 21. Ramsay SE, Whincup PH, Papacosta O, Morris RW, Lennon LT, Wannamethee SG. Inequalities in heart failure in older men: prospective associations between socioeconomic measures and heart failure incidence in a 10-year follow-up study. Eur Heart J. 2014;35:442-7. 22. Philbin EF, Dec GW, Jenkins PL, DiSalvo TG. Socioeconomic status as an independent risk factor for hospital readmission for heart failure. Am J Cardiol. 2001;87:1367-71. 23. Struthers AD, Anderson G, Donnan PT, MacDonald T. Social deprivation increases cardiac hospitalisations in chronic heart failure independent of disease severity and diuretic nonadherence. Heart. 2000;83:12-6. 24. Pujades-Rodriguez M, Timmis A, Stogiannis D, Rapsomaniki E, Denaxas S, Shah A, et al. Socioeconomic deprivation and the incidence of 12 cardiovascular diseases in 1.9 million women and men: implications for risk prediction and prevention. PLoS ONE. 2014;9:e104671. 25. Ali AS, Rybicki BA, Alam M, Wulbrecht N, Richer-Cornish K, Khaja F, et al. Clinical predictors of heart failure in patients with first acute myocardial infarction. Am Heart J. 1999;138:1133-9. 26. Lewis EF, Moye LA, Rouleau JL, Sacks FM, Arnold JM, Warnica JW, et al. Predictors of late development of heart failure in stable survivors of myocardial infarction: the CARE study. J Am Coll Cardiol. 2003;42:1446-53. 27. Lewis EF, Velazquez EJ, Solomon SD, Hellkamp AS, McMurray JJ, Mathias J, et al. Predictors of the first heart failure hospitalization in patients who are stable survivors of myocardial infarction complicated by pulmonary congestion and/or left ventricular dysfunction: a VALIANT study. Eur Heart J. 2008;29:748-56. 28. Parkinson JP, Davis S, van Staa T. The General Practice Research Database: now and the future. Mann R, Andrews EB, eds Pharmacovigilance: John Wiley; 2007. p. 341-8. 29. Hazelbag CM, Klungel OH, van Staa TP, de Boer A, Groenwold RH. Left truncation results in substantial bias of the relation between time-dependent exposures and adverse events. Ann Epidemiol. 2015;25:590-6.

PART ONE CHAPTER 2

SUPPLEMENTAL MATERIAL Table S1. Hazard ratio for heart failure in patients following a first myocardial infarction using the competing risks regression based on Fine and Gray’s proportional subhazards model to account for the competing risk of all cause mortality Model 1 n = 24458

Model 2 n = 24458

Model 3 n = 24458

Model 4 n = 24458

HR (95% CI)

Age, per 10 years increase

1.41 (1.38 – 1.44)

1.39 (1.36 – 1.42)

1.37 (1.34 – 1.40)

1.36 (1.33 – 1.39)

Men

1.04 (0.99 – 1.10)

1.06 (1.01 – 1.12)

1.06 (1.00 – 1.12)

1.05 (1.00 – 1.11)

overall p = <0.001

Reference

1.11 (1.02 – 1.21)

1.11 (1.02 – 1.20)

1.10 (1.02 – 1.20)

1.21 (1.12 – 1.32)

1.21 (1.11 – 1.31)

1.19 (1.10 – 1.30)

1.24 (1.14 – 1.35)

1.22 (1.12 – 1.32)

1.35 (1.25 – 1.47)

1.35 (1.24 – 1.46)

1.31 (1.20 – 1.42)

History of hypertension

1.15 (1.09 – 1.21)

1.14 (1.08 – 1.20)

1.15 (1.08 – 1.22)

History of diabetes

1.38 (1.29 – 1.48)

1.39 (1.29 – 1.49)

1.37 (1.27 – 1.47)

History of atrial fibrillation

1.52 (1.41 – 1.63)

1.51 (1.40 – 1.62)

Type of MI

overall p = <0.001

NSTEMI

Reference

Unclassified MI vs. NSTEMI

1.16 (1.09 – 1.23)

1.17 (1.10 – 1.24)

STEMI vs. NSTEMI

1.14 (1.05 – 1.24)

1.15 (1.06 – 1.24)

Index of multiple deprivation

History of peripheral arterial disease

1.29 (1.19 – 1.41)

History of COPD

1.19 (1.09 – 1.29)

ACE inhibitor use before MI

1.01 (0.95 – 1.08)

Angiotensin receptor blocker (ARB) use before MI

0.96 (0.87 – 1.07)

Beta-blocker use before MI

0.93 (0.87 – 0.99)

Overall p-values are shown for categorical covariates. ACE = Angiotensin-converting enzyme; CI, confidence interval; COPD = chronic obstructive pulmonary disease; MI = myocardial infarction; NSTEMI = non-ST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction; Q = quintile; HR, ratio of the subdistribution hazards.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Table S2. Hazard ratio for heart failure in patients following a first myocardial infarction in sensitivity analyses Model 5 without multiple imputation

Model 5 with multiple imputation

Model 5 competing risks

HR (95% CI)

Age, per 10 years increase

1.49 (1.38 – 1.60)

1.48 (1.44 – 1.52)

1.45 (1.36 – 1.54)

Men

1.04 (0.90 – 1.20)

1.07 (1.01 – 1.13)

1.07 (0.90 – 1.20)

Body mass index (BMI)

p = 0.006

p = <0.001

p = < 0.001

Underweight

0.79 (0.48 – 1.32)

0.74 (0.56 – 0.99)

0.71 (0.48 – 1.32)

Normal

Reference

Overweight

0.84 (0.71 – 0.99)

0.82 (0.75 – 0.90)

0.84 (0.71 – 0.99)

Obese

1.12 (0.94 – 1.34)

0.97 (0.88 – 1.07)

1.17 (0.94 – 1.34)

p = 0.040

p = 0.001

p = < 0.001

Index of multiple deprivation Q1

Reference

1.22 (0.94 – 1.58)

1.08 (0.98 – 1.19)

1.35 (0.94 – 1.58)

1.34 (1.03 – 1.75)

1.18 (1.06 – 1.30)

1.39 (1.03 – 1.75)

1.24 (0.94 – 1.63)

1.16 (1.05 – 1.29)

1.31 (0.94 – 1.63)

1.46 (1.09 – 1.94)

1.25 (1.12 – 1.40)

1.61 (1.09 – 1.94)

Smoking status

p = 0.85

p = 0.002

p = 0.93

Non-smoker

Reference

Current smoker

0.95 (0.78 – 1.15)

1.17 (1.07 – 1.28)

1.03 (0.78 – 1.15)

Ex-smoker

0.98 (0.83 – 1.15)

1.04 (0.95 – 1.13)

0.99 (0.83 – 1.15)

History of hypertension

1.13 (0.95 – 1.35)

1.16 (1.09 – 1.24)

1.19 (0.95 – 1.35)

History of diabetes

1.26 (1.08 – 1.47)

1.47 (1.36 – 1.58)

1.24 (1.08 – 1.47)

History of atrial fibrillation

1.48 (1.25 – 1.76)

1.64 (1.52 – 1.76)

1.37 (1.25 – 1.76)

Type of MI

p = 0.024

p = <0.001

p = 0.69

NSTEMI

Reference

Unclassified MI vs. NSTEMI

0.83 (0.70 – 0.99)

0.89 (0.81 – 0.96)

1.02 (0.70 – 0.99)

STEMI vs. NSTEMI

1.06 (0.89 – 1.27)

1.20 (1.10 – 1.31)

1.07 (0.89 – 1.27)

History of peripheral arterial disease

1.55 (1.27 – 1.89)

1.36 (1.24 – 1.49)

1.35 (1.27 – 1.89)

History of COPD

1.19 (0.97 – 1.46)

1.24 (1.13 – 1.36)

1.22 (0.97 – 1.46)

Systolic blood pressure, mmHg

1.00 (0.99 – 1.00)

1.00 (1.00 – 1.00)

1.00 (0.99 – 1.00)

Diastolic blood pressure, mmHg

1.00 (0.99 – 1.01)

1.00 (1.00 – 1.00)

1.00 (0.99 – 1.01)

ACE inhibitor use before MI

1.01 (0.87 – 1.18)

1.08 (1.00 – 1.15)

1.06 (0.87 – 1.18)

Angiotensin receptor blocker (ARB) use before MI

0.95 (0.77 – 1.18)

1.01 (0.91 – 1.13)

0.93 (0.77 – 1.18)

Beta-blocker use before MI

0.87 (0.75 – 1.01)

0.95 (0.89 – 1.01)

0.90 (0.75 – 1.01)

Using multivariable Cox regression without and with multiple imputation (n = 4911, n events = 1160) and the competing risks regression based on Fine and Gray’s proportional subhazards model to account for the competing risk of all cause mortality Additional prognostic covariates: body mass index, smoking, systolic and diastolic blood pressure. Overall p-values are shown for categorical covariates. HR = hazard ratio; CI = confidence interval; MI = myocardial infarction; NSTEMI = non-ST-segment-elevation myocardial infarction; STEMI = ST-segment-elevation myocardial infarction; COPD = chronic obstructive pulmonary disease; ACE = Angiotensin-converting enzyme; Q = quintile.

PART ONE CHAPTER 2

Multiple imputation algorithm Multiple imputation was implemented using the â&#x20AC;&#x2DC;miceâ&#x20AC;&#x2122; package for R, to replace missing values in prognostic covariates (body mass index, smoking, systolic and diastolic blood pressure).1 Imputation models included general practice and 3-year periods from 1998 to 2010 of index myocardial infarction (MI) and: 1. All the baseline covariates as used in the multivariable analysis (age, sex, index of multiple deprivation, hypertension, diabetes, type of MI, body mass index, smoking, systolic blood pressure, diastolic blood pressure); 2. Baseline measurements of covariates not considered in the multivariable analysis (alcohol intake, haemoglobin, white blood cell count, creatinine, alanine aminotransferase (ALAT), total cholesterol, HDL cholesterol); 3. Baseline medications prior to index MI (antiplatelet, oral anticoagulant, statin, ACE inhibitor, angiotensin receptor blocker (ARB), beta-blocker, calcium channel blocker, loop diuretic, aldosterone antagonist, digoxin); 4. Coexisting medical conditions (history of atrial fibrillation, peripheral arterial disease, depression, cancer, chronic obstructive pulmonary disease); 5. The Nelson-Aalen hazard and the event status for the outcomes analysed.

REFERENCES 1.

van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res. 2007;16:219-42.

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

Phenotype coding lists CALIBER myocardial infarction phenotype Fatal or non-fatal myocardial infarction (MI) as recorded in any of the constituent CALIBER data sources. The type of myocardial infarction (ST elevation MI, STEMI; or non ST elevation MI, NSTEMI) is recorded only in MINAP and GPRD. Read codes used to identify myocardial infarction in primary care records (Clinical Research Practice Datalink). Category

Clinical code

Clinical term

Datasource lookup

STEMI (3)

G30X000

Acute ST segment elevation myocardial infarction

NSTEMI (4)

G307100

Acute non-ST segment elevation myocardial infarction

10562

Acute MI not further specified (5)

323..00

ECG: myocardial infarction

7783

12229

Acute MI not further specified (5)

3233.00

ECG: antero-septal infarct.

26975

Acute MI not further specified (5)

3234.00

ECG:posterior/inferior infarct

26972 55401

Acute MI not further specified (5)

3235.00

ECG: subendocardial infarct

Acute MI not further specified (5)

3236.00

ECG: lateral infarction

52705

Acute MI not further specified (5)

323Z.00

ECG: myocardial infarct NOS

59032

Acute MI not further specified (5)

889A.00

Diab mellit insulin-glucose infus acute myocardial infarct

61670

Acute MI not further specified (5)

G30..00

Acute myocardial infarction

241

Acute MI not further specified (5)

G30..12

Coronary thrombosis

2491

Acute MI not further specified (5)

G30..13

Cardiac rupture following myocardial infarction (MI)

30421

Acute MI not further specified (5)

G30..15

MI - acute myocardial infarction

1677

Acute MI not further specified (5)

G30..16

Thrombosis - coronary

13571 12139

Acute MI not further specified (5)

G300.00

Acute anterolateral infarction

Acute MI not further specified (5)

G301.00

Other specified anterior myocardial infarction

5387

Acute MI not further specified (5)

G301000

Acute anteroapical infarction

40429

Acute MI not further specified (5)

G301100

Acute anteroseptal infarction

17872

Acute MI not further specified (5)

G301z00

Anterior myocardial infarction NOS

14897 8935

Acute MI not further specified (5)

G302.00

Acute inferolateral infarction

Acute MI not further specified (5)

G303.00

Acute inferoposterior infarction

29643

Acute MI not further specified (5)

G304.00

Posterior myocardial infarction NOS

23892 14898

Acute MI not further specified (5)

G305.00

Lateral myocardial infarction NOS

Acute MI not further specified (5)

G306.00

True posterior myocardial infarction

63467

Acute MI not further specified (5)

G307.00

Acute subendocardial infarction

3704

Acute MI not further specified (5)

G307000

Acute non-Q wave infarction

9507

Acute MI not further specified (5)

G308.00

Inferior myocardial infarction NOS

1678

Acute MI not further specified (5)

G309.00

Acute Q-wave infarct

30330

Acute MI not further specified (5)

G30B.00

Acute posterolateral myocardial infarction

32854

Acute MI not further specified (5)

G30X.00

Acute transmural myocardial infarction of unspecif site

29758

Acute MI not further specified (5)

G30y.00

Other acute myocardial infarction

34803 28736

Acute MI not further specified (5)

G30y000

Acute atrial infarction

Acute MI not further specified (5)

G30y100

Acute papillary muscle infarction

62626

Acute MI not further specified (5)

G30y200

Acute septal infarction

41221

Acute MI not further specified (5)

G30yz00

Other acute myocardial infarction NOS

46017

Acute MI not further specified (5)

G30z.00

Acute myocardial infarction NOS

14658

Acute MI not further specified (5)

G31y100

Microinfarction of heart

68357

Acute MI not further specified (5)

G38..00

Postoperative myocardial infarction

32272

Acute MI not further specified (5)

G380.00

Postoperative transmural myocardial infarction anterior wall

46112

Acute MI not further specified (5)

G381.00

Postoperative transmural myocardial infarction inferior wall

46276

Acute MI not further specified (5)

G384.00

Postoperative subendocardial myocardial infarction

41835

Acute MI not further specified (5)

G38z.00

Postoperative myocardial infarction, unspecified

68748

Acute MI not further specified (5)

Gyu3400

[X]Acute transmural myocardial infarction of unspecif site

96838

PART ONE CHAPTER 2

Hospital Episode Statistics Category (code)

ICD10 code

ICD10 term

Acute MI not further specified (5)

I21

Acute myocardial infarction

Category (code)

OPCS 4

OPCS 4 term

Transluminal coronary thrombolysis (2)

K50.2

Percutaneous transluminal coronary thrombolysis using streptokinase

Transluminal coronary thrombolysis (2)

K50.3

Percutaneous transluminal injection of therapeutic substance into coronary artery NEC

OPCS

Office for National Statistics (ONS) Category (code)

ICD10 code

ICD10 term

Death from MI (3)

I21

Acute myocardial infarction

Death from MI (3)

I22

Subsequent myocardial infarction

Death from MI (3)

I23

Certain current complications following acute myocardial infarction

Office for National Statistics (ONS) prior to 2000 (ICD9) Category (code)

ICD9 code

ICD9 term

Death from myocardial infarction (3)

410

Acute myocardial infarction

Death from myocardial infarction (3)

4110

Other acute and subacute forms of ischemic heart disease ; Postmyocardial infarction syndrome

Death from myocardial infarction (3)

4297

Ill-defined descriptions and complications of heart disease ; Certain sequelae of myocardial infarction, not elsewhere classified

Myocardial Ischaemia National Audit Project (MINAP) 1 Acute STEMI 2 Acute NSTEMI

HEART FAILURE FOLLOWING MI: A CALIBER STUDY

CALIBER heart failure phenotype Read codes used to identify myocardial infarction in primary care records (Clinical Research Practice Datalink). Category (code)

Clinical code

Clinical term

Datasource lookup

Heart failure due to valvular disease (3) G580400

Congestive heart failure due to valvular disease

94870

Heart failure due to hypertension (4)

G210.00

Malignant hypertensive heart disease

50157

Heart failure due to hypertension (4)

G210000

Malignant hypertensive heart disease without CCF

95334

Heart failure due to hypertension (4)

G210100

Malignant hypertensive heart disease with CCF

72668

Heart failure due to hypertension (4)

G211100

Benign hypertensive heart disease with CCF

52127

Heart failure due to hypertension (4)

G21z100

Hypertensive heart disease NOS with CCF

62718

Heart failure due to hypertension (4)

G230.00

Malignant hypertensive heart and renal disease

67232

Heart failure due to hypertension (4)

G232.00

Hypertensive heart&renal dis wth (congestive) heart failure

21837

Heart failure due to hypertension (4)

G234.00

Hyperten heart&renal dis+both(congestv)heart and renal fail

57987

Heart failure due to other cause (5)

G1yz100

Rheumatic left ventricular failure

22262

Heart failure NOS (6)

1O1..00

Heart failure confirmed

9913

Heart failure NOS (6)

662W.00

Heart failure annual review

30779

Heart failure NOS (6)

662p.00

Heart failure 6 month review

83502 24503

Heart failure NOS (6)

8B29.00

Cardiac failure therapy

Heart failure NOS (6)

8H2S.00

Admit heart failure emergency

32898

Heart failure NOS (6)

9Or0.00

Heart failure review completed

19380

Heart failure NOS (6)

G400.00

Acute cor pulmonale

8464

Heart failure NOS (6)

G41z.11

Chronic cor pulmonale

5695 5141

Heart failure NOS (6)

G554000

Congestive cardiomyopathy

Heart failure NOS (6)

G554011

Congestive obstructive cardiomyopathy

68766

Heart failure NOS (6)

G58..00

Heart failure

2062

Heart failure NOS (6)

G58..11

Cardiac failure

1223

Heart failure NOS (6)

G580.00

Congestive heart failure

398

Heart failure NOS (6)

G580.11

Congestive cardiac failure

2906

Heart failure NOS (6)

G580.12

Right heart failure

10079

Heart failure NOS (6)

G580.13

Right ventricular failure

10154 9524

Heart failure NOS (6)

G580.14

Biventricular failure

Heart failure NOS (6)

G580000

Acute congestive heart failure

23707

Heart failure NOS (6)

G580100

Chronic congestive heart failure

32671 27884

Heart failure NOS (6)

G580200

Decompensated cardiac failure

Heart failure NOS (6)

G580300

Compensated cardiac failure

11424

Heart failure NOS (6)

G581.00

Left ventricular failure

884

Heart failure NOS (6)

G581.11

Asthma - cardiac

23481

Heart failure NOS (6)

G581.13

Impaired left ventricular function

5942 5255

Heart failure NOS (6)

G581000

Acute left ventricular failure

Heart failure NOS (6)

G582.00

Acute heart failure

27964

Heart failure NOS (6)

G58z.00

Heart failure NOS

4024 17278

Heart failure NOS (6)

G58z.12

Cardiac failure NOS

Heart failure NOS (6)

G5yy900

Left ventricular systolic dysfunction

8966

Heart failure NOS (6)

G5yyA00

Left ventricular diastolic dysfunction

12550

Heart failure NOS (6)

R2y1000

[D]Cardiorespiratory failure

20324

Left ventricular dysfunction (3)

585f.00

Echocardiogram shows left ventricular systolic dysfunction

11284

Left ventricular dysfunction (3)

585g.00

Echocardiogram shows left ventricular diastolic dysfunction

11351

PART ONE CHAPTER 2

Hospital Episode Statistics Category (code)

ICD10 code

ICD10 term

HF - hypertension cause (4)

I110

Hypertensive heart disease with (congestive) heart failure

HF - hypertension cause (4)

I130

Hypertensive heart and renal disease with (congestive) heart failure

HF - hypertension cause (4)

I132

Hypertensive heart and renal disease with both (congestive) heart failure and renal failure

HF - other unspecified cause (5)

I260

Pulmonary embolism with mention of acute cor pulmonale

HF - not otherwise specified (6)

I50

Heart failure

Office for National Statistics (ONS) Category (code)

ICD10 code

ICD10 term

Hypertensive heart failure (4)

I110

Hypertensive heart disease with (congestive) heart failure

Hypertensive heart failure (4)

I130

Hypertensive heart and renal disease with (congestive) heart failure

Hypertensive heart failure (4)

I132

Hypertensive heart and renal disease with both (congestive) heart failure and renal failure

Heart failure with other specified cause (5)

I260

Pulmonary embolism with mention of acute cor pulmonale

Heart failure with unspecified cause (6)

I50

Heart failure

Office for National Statistics (ONS) prior to 2000 (ICD9) Category (code)

ICD9 code

ICD9 term

Heart failure with other specified cause (5)

4151

Acute pulmonary heart disease ; Pulmonary embolism and infarction

Heart failure with unspecified cause (6)

428

Heart failure

Myocardial Ischaemia National Audit Project (MINAP) CCF 1

PART ONE

Clinical Course Portrayed

Chapter

Heart Failure Following ST-elevation Myocardial Infarction: an AGNES Cohort Study of Incidence and Prognostic Factors In preparation

Johannes M.I.H. Gho1, Pieter G. Postema2, Maartje Conijn1, Nienke Bruinsma2, Jonas S.S.G. de Jong2, Connie R. Bezzina3, Arthur A.M. Wilde2, Folkert W. Asselbergs1,4,5

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Cardiology, Heart Center, Academic Medical Center, Amsterdam, The Netherlands

Department of Clinical and Experimental Cardiology, Academic Medical Center, Amsterdam, The Netherlands.

Durrer Center for Cardiogenetic Research, ICIN-Netherlands Heart Institute, Utrecht, The Netherlands

Institute of Cardiovascular Science, Faculty of Population Health Sciences, University College London, London, United Kingdom

PART ONE CHAPTER 3

ABSTRACT Background Aim of the current study is to determine the contemporary incidence, risk factors and prognosis of heart failure (HF) after ST-elevation myocardial infarction (STEMI). Methods We used the Arrhythmia Genetics in the NEtherlandS (AGNES) observational cohort study to identify patients with a first STEMI from 2001 onwards (n=1459). Heart failure during follow-up was defined as a hospitalisation for HF or an outpatient clinic visit for HF. Cox regression was performed to estimate the relationship between baseline covariates and the onset of HF. Results Follow-up was completed for 1360 patients with an overall median follow-up time of 6.7 years. A total of 85 patients (6.3%) developed HF during follow-up. In multivariable analysis, peak CKMB levels (HR 1.11 per 100U/L, 95%CI 1.11 â&#x20AC;&#x201C; 1.22) and an LAD culprit lesion (HR 2.82, 95%CI 1.50 â&#x20AC;&#x201C; 5.30) were risk factors associated with HF. Conclusion We found a relatively low long-term contemporary incidence of HF after a first STEMI in comparison to other reports.

HEART FAILURE FOLLOWING STEMI: AN AGNES STUDY

INTRODUCTION Heart failure (HF) is a major medical problem in the western world.1 HF is associated with substantial morbidity and mortality and contributes in a significant extent to national healthcare costs.2 The incidence of HF could be influenced by an increasing survival following an ischaemic event due to primary percutaneous coronary intervention (PCI) and because of demographic changes (ageing population).1 Data on the long-term incidence of HF after ST-elevation myocardial infarction (STEMI) is lacking and varies among different studies. In a recent nationwide Swedish

registry study (SWEDEHEART), HF incidence during index hospitalisation declined from 50% to 28% in the subgroup including STEMI patients between 1996 and 2008.3 In a Canadian cohort study 13.6% of STEMI patients were diagnosed with HF during index MI hospitalisation with a 1-year cumulative HF incidence of 23.4% between 2002 and 2008.4 In the same study, development of HF was associated with a 1-year mortality of 9.9% in STEMI patients.4 Here, we aim to determine the contemporary incidence and prognostic factors associated with HF after STEMI in the Netherlands percutaneous coronary intervention (PCI) era.

METHODS Study design The Arrhythmia Genetics in the NEtherlandS (AGNES) study is a Dutch multicenter observational study of which details have been published before.5-7 The AGNES cohort consisted of individuals with a first STEMI. Setting and participants This study complies with the Declaration of Helsinki. All patients or their legal representative gave written informed consent for inclusion and collection of information during follow-up was approved by the institutional ethics committee. Patients with first STEMI were recruited at seven heart centers in The Netherlands from 2001 onwards. Excluded were individuals with an actual nonâ&#x20AC;&#x201C;ST-elevation myocardial infarction, prior myocardial infarction or revascularisation prior to inclusion, congenital heart defects, known structural heart disease, severe comorbidity at discretion of the operator which would result in severely decreased prognosis, severe electrolyte disturbances, trauma at presentation, recent surgery, previous coronary artery bypass graft or use of class I and III antiarrhythmic drugs. The study cohort consisted of patients with a first STEMI. Patients who died within 30 days of the index infarct and patients with an incomplete follow-up were excluded from this analysis. Data sources Data on age, gender, cardiovascular risk factors, medication use before STEMI, infarct characteristics (e.g., type of reperfusion therapy, biomarkers and culprit lesion) were recorded at admission for index MI. Data on risk factors and family history in a subcohort were previously defined.7 Family history included sudden death and cardiovascular disease. A positive family history of sudden death was defined as a self-reported parent or sibling who died suddenly and

PART ONE CHAPTER 3

unexpectedly before the age of 80 years. We collected follow-up data from the hospital of index MI admission and the referral center. Collected endpoints included subsequent MI, stroke, HF, revascularization (CABG and PCI), ICD implantation, aborted cardiac arrest and all-cause mortality. In case of insufficient data we contacted general practitioners for additional information. The end points were verified by an adjudication committee (P.G.P. and F.W.A.). In this study we defined HF cases as patients with a hospital admission for HF or patients visiting the outpatient clinic for HF during follow-up. Patients without HF during follow-up served as control cases. Mortality data were verified using the Dutch Municipal Personal Records Database. Statistical methods Baseline patient characteristics, index MI characteristics and outcomes were described using descriptive statistics. Differences in continuous variables between cases with HF and controls without HF were tested using an independent t-test when data were normally distributed or otherwise using a Mann-Whitney test. Differences in categorical variables between cases and controls were determined using a chi-square test. We calculated cumulative incidence (cumulative percentage) of HF. Kaplan-Meier curves were generated to describe unadjusted cumulative HF incidence and all-cause mortality in STEMI patients using 5-year follow-up. Cox regression The associations of exposures of interest with the onset of HF were explored using Cox proportional hazard models. The primary outcome measure was patients admitted to the hospital with HF or visiting the outpatient clinic for HF. Cox regression with complete case analysis was performed to estimate the relationship between baseline prognostic factors and the development of heart failure after myocardial infarction and to determine hazard ratios (HR) with 95% confidence intervals (95%CI). The proportional hazards assumption was investigated by partial residual plots. In all analyses a p-value of <0.05 was considered as significant. Statistics were performed using IBM SPSS Statistics (Version 22, IBM Corporation, Armonk, New York, United States).

RESULTS In the AGNES cohort (n=1459), follow-up was incomplete for 74 patients due to insufficient available information (inability to retrieve follow-up information) and 25 patients died within 30 days of index MI (Figure 1). Follow-up was completed for 1360 patients. In total, 85 HF cases and 1275 controls without HF were included. Baseline and infarct characteristics Heart failure cases were significantly older at their index MI (59.9 vs. 57.2 years, p=0.001), more commonly had a positive family history (FH) of cardiovascular disease (CVD) (77.5% vs. 65.8%, p=0.032), a positive FH of sudden death (SD) (41% vs. 26.1%, p=0.006), a history of atrial fibrillation (6.1% vs. 1.4%, p=0.001), and were less likely to use statins prior to MI compared to controls (3.6% vs. 10.6%, p=0.041) (Table 1). In patients with HF median peak CK-MB levels at index STEMI were significantly higher (422.5 U/L, IQR 333.0) than in controls (181.0 U/L, IQR

HEART FAILURE FOLLOWING STEMI: AN AGNES STUDY

Figure 1. Flow chart

259.0) (p<0.001) (Table 2). Patients with HF were more likely to have suffered an anterior MI (79.8% vs. 48.9%, p<0.001) or LAD as culprit lesion (82.3% vs. 49.8%, p<0.001), whereas controls were more likely to have an RCX (3.8% vs. 15.9%, p=0.003) or RCA as culprit lesion (12.7% vs. 33.3%, p=0.001) at index STEMI. Other baseline characteristics were comparable between the two groups. There was no significant difference in ventricular fibrillation at index MI between cases (48.2%) and controls (41.4%). Heart failure incidence after MI and survival A total of 85 (6.3%) patients developed HF during an overall median follow-up time of 6.7 years (Table 3). Median follow-up was longer in HF cases than in controls (9.4 vs. 6.4 years, p<0.001). The median time until the onset of HF was 2.1 years. During follow-up, HF cases were more likely to have been subjected to revascularisation by additional PCI of the initial culprit artery (14.1% versus 7.4%, p = 0.025) or CABG (10.6% versus 5.3%, p = 0.042) than controls. We also found a higher rate of ICD implantation (44.7% versus 3.5%, p<0.001) and stroke (7.1% versus 2.6%, p = 0.017) in cases compared to controls. Overall 4.3% of the patients developed HF during 5-year follow-up (Figure 2). During follow-up 23.5% of HF cases versus 11.1% controls died (p=0.001). Overall 5-year mortality after a first STEMI was 6.7% (Figure 3). Risk factors associated with the onset of heart failure following index STEMI The proportional hazards assumption was violated and therefore the dataset was administratively censored at 10 years follow-up to ensure goodness-of-fit. A multivariable analysis (n=833) for

PART ONE CHAPTER 3

the outcome of HF was performed using age at index STEMI, gender, maximal measured Creatine Kinase (CK-MB) levels, the LAD culprit lesion and the calendar year of the index STEMI as predictors. Predictors associated with HF were peak CK-MB levels (HR 1.11 per 100U/L, 95%CI 1.11 – 1.22), a LAD culprit lesion (HR 2.82, 95% CI 1.50 – 5.30) and the year of index STEMI (HR 0.89 per year, 95%CI 0.83 – 0.96). Furthermore we found a trend towards an association with higher age at STEMI (HR 1.02, 95%CI 1.00 – 1.05). We did not find an association of gender with the outcome HF (male HR 0.77, 95%CI 0.40 – 1.48).

Table 1. Baseline characteristics With HF (n=85)

Without HF (n=1275)

Number of missings

p-value

Gender (female)

20 (23.5%)

261 (20.5%)

0.509

Mean age in years at index infarction (SD)

59.9 (10.3)

57.2 (10.7)

0.001

VF at index MI

41 (48.2%)

527 (41.4%)

0.212

Cardiovascular risk profile Mean BMI (SD)

26.6 (3.9)

0.702

FH of CVD

62 (77.5%)

817 (65.8%)

0.032

Current smoker

50 (50.5%)

767 (61.5%)

0.718

Diabetes

8 (9.6%)

94 (7.6%)

0.502

Hypertension

32 (39%)

378 (31.3%)

0.148

Atrial fibrillation

5 (6.1%)

17 (1.4%)

0.001

17 (22.7%)

378 (32.1)

110

0.087

FH of sudden death

High cholesterol

34 (41%)

330 (26.1%)

0.006

Angina 48h before STEMI

37 (45.1%)

455 (36.6%)

0.122

Cardiac medication before STEMI β-Blocker

13 (15.7%)

120 (9.6%)

0.073

Statins

3 (3.6%)

133 (10.6%)

0.041

Diuretics

5 (6.1%)

79 (6.3%)

0.943

ACE inhibitors/ARB

11 (13.3%)

98 (7.8%)

0.080

Aspirin/oral anticoagulation

9 (10.8%)

103 (8.2%)

0.401

Numbers in the columns with and without HF denote n= (%), unless specified otherwise. Abbreviations: ACE = angiotensin converting enzyme; ARB = Angiotensin receptor blocker; BMI = body mass index; FH = family history; HF = heart failure; MI = myocardial infarction; SD = standard deviation; STEMI = STelevation myocardial infarction; VF = ventricular fibrillation.

HEART FAILURE FOLLOWING STEMI: AN AGNES STUDY

Table 2. Characteristics of index STEMI With HF (n=85)

Without HF (n=1275)

Number of missings

p-value

73 (85.9%)

1159 (90.9%)

0.125

Reperfusion therapy PCI CABG

3 (3.5%)

19 (1.5%)

0.149

Thrombolysis

4 (4.7%)

34 (2.7%)

0.269

Non (spontaneous reperfusion on angiography or medication)

5 (5.9%)

53 (4.2%)

0.446

Time between symptoms and PCI (min) 180 (IQR 112)

180 (IQR 119)

0.670

Maximal Troponin T value (median) Îźg/L 6.26 (IQR 8.74)

2.10 (IQR 4.03)

1050

0.280

Maximal CK-MB value (median) U/L

422.5 (IQR 333.0)

181.0 (IQR 259.0)

243

<0.001

Location MI (anterior)

67 (79.8%)

617 (48.9%)

<0.001

65 (82.3%)

610 (49.8%)

Culprit lesion LAD

55 <0.001

RCX

3 (3.8%)

195 (15.9%)

0.003

RCA

10 (12.7%)

409 (33.3%)

0.001

1 (1.3%)

12 (1.0%)

30 (38.5%)

450 (36.9%)

LM Multivessel disease

0.828 62

0.780

Numbers in the columns with and without HF denote n= (%), unless specified otherwise. Abbreviations: CABG = coronary artery bypass grafting; HF = heart failure; LAD = left anterior descending artery; LM = left main artery; MI = myocardial infarction. PCI = percutaneous coronary intervention; RCA = right coronary artery; RCX = ramus circumflexus.

Table 3. Outcomes after STEMI Total (n=1360)

With HF (n=85)

Without HF (n=1275) p-value

6.7 (2.5-10.9)

9.4 (6.5-12.6)

6.4 (0.2-10.8)

<0.001

PCI culprit

106 (7.8%)

12 (14.1%)

94 (7.4%)

0.025

PCI non culprit

117 (8.6%)

10 (11.8%)

107 (8.4%)

0.283

CABG

Median follow-up in years (IQR) Revascularisations

77 (5.7%)

9 (10.6%)

68 (5.3%)

0.042

ICD

83 (6.1%)

38 (44.7%)

45 (3.5%)

<0.001

Second MI

130 (9.6%)

10 (11.8%)

120 (9.4%)

0.475

Stroke

39 (2.9%)

6 (7.1%)

33 (2.6%)

0.017

Deaths

162 (11.9%)

20 (23.5%)

143 (11.2%)

0.001

Numbers in the columns with and without HF denote n= (%), unless specified otherwise. Abbreviations: The p-value is derived from the comparison between HF cases and controls. CABG = coronary artery bypass grafting; HF = heart failure; ICD = implantable cardioverter-defibrillator; MI = myocardial infarction. PCI = percutaneous coronary intervention.

PART ONE CHAPTER 3

Figure 2. Kaplan-Meier curve showing the time until onset of heart failure in 5-year follow-up. The inset shows the portion of the graph with 100-90% free of heart failure. STEMI = ST-elevation myocardial infarction.

Figure 3. Kaplan-Meier curve showing the overall unadjusted 5-year survival after a first STEMI. The inset shows the portion of the graph with 100-90% cumulative survival. STEMI = ST-elevation myocardial infarction.

HEART FAILURE FOLLOWING STEMI: AN AGNES STUDY

DISCUSSION Using a patient cohort recruited in the Netherlands with a first STEMI we observed that 6.3% of patients developed HF during a median follow-up of 6.7 years. The median time to onset of HF was 2.1 years. The overall mortality was twice as high in the HF group versus the control group after index MI. Higher CK-MB levels and an LAD culprit lesion at index STEMI were important risk factors for the outcome of HF.

Incidence Studies researching the contemporary long-term HF incidence in STEMI subgroups report an incidence between 23 and 28%.3, 4 When comparing different studies, variability in reperfusion therapy changing over the years and HF definition should be taken into account. In the current study overall 90.6% of patients underwent primary PCI and 2.8% thrombolysis, which resembles a unique contemporary STEMI cohort and comparable population studies undergoing similar reperfusion therapy are lacking. Of the SWEDEHEART study patients with STEMI or left bundle branch block admitted in 2002-2003, 72.2% underwent thrombolysis and 2.78% primary PCI compared to 8.5% thrombolysis and 91.5% primary PCI in 2008.3 In the Canadian study subgroup of STEMI patients, PCI was performed in 46.8% of patients with HF during index hospitalisation, 52.5% of patients who developed HF after discharge and 60.2% in those who did not develop HF.4 Other studies were not limited to STEMI, but also included patients with non-ST-elevation infarction (NSTEMI) and were predominantly from the thrombolytic era. In the Framingham Heart Study, Velagaleti et al. found that 14.8% (21/142) of patients surviving 30 days after index MI developed congestive HF during 5-year follow-up from 1990 to 1999.8 Data of Torabi et al. showed a HF incidence of 33% after discharge from hospital admission for index MI.9 In a study by Najafi et al. an 22.4% developed HF within 28 days of index admission (after exclusion of patients who died within 28 days) and from these patients 12.4% had at least one subsequent admission with HF after 10 years follow-up.10 Epidemiologic research in Olmsted County (Minnesota, USA) found a HF incidence of 41% during a median follow-up of 6.6 years after MI.11 A more recent Danish study found a 90-day HF incidence of 19.6% in 2009-2010.12 Prognostic factors Infarct size and anterior MI are well known risk factors for adverse prognosis in STEMI patients.13 In this study, maximal CK-MB value as a proxy for infarct size was an adverse prognostic factor for the development of HF. Cardiac biomarkers, including peak CK-MB significantly correlated in previous research with infarct size and LVEF.14 We also found a LAD culprit lesion to be a risk factor for development of HF after STEMI (HR 2.82). Our study demonstrated a negative association between the year in which the index event occurred and the development of HF. This is in accordance to other studies where decreasing trends in the development of HF after MI are reported.11, 12, 15, 16 We also found a trend for higher age at the time of the index MI associated with the development of HF after a first STEMI. This was expected as age is a well-known risk factor for HF development.3, 10, 11, 15 For example, Desta et al. found an odds ratio of 1.05 (95%CI 1.05 â&#x20AC;&#x201C; 1.06) for every additional year increase in age.3 A previous study found a significant association of gender with HF incidence after MI, but we could not replicate this finding in our STEMI cohort.15

PART ONE CHAPTER 3

Survival We found that about 1 in 4 HF patients died during a median follow-up of 6.7 years. This is consistent with results reported in other studies in which the poor prognosis after onset of HF is described.10, 11, 16, 17 In the SWEDEHEART study, Desta et al. found a decrease in 1-year mortality for patients with clinical HF following MI from 36% to 31% between 1996 and 2008.3 Kaul et al. found a 1-year mortality rate of 10.6% in patients with HF and 2.4% mortality in patients without HF after STEMI.4 Limitations Cardiovascular endpoints might have been missed as follow-up was performed in the hospital of the index admission, for example (predominantly milder) HF cases managed by the general practitioner or HF cases managed in other hospitals. In this study we defined HF as a hospitalisation or outpatient clinic visit for HF, which is of relevance to external validity. Other studies also used administration of diuretics or symptoms of HF to define HF.3,

9, 12

The study did not systematically include consecutive STEMI patients, patients with primary VF were selectively included and patients without primary VF were random included but the cohort represents a contemporary STEMI population.

CONCLUSION This research demonstrates a low contemporary incidence of HF after a first PCI-treated STEMI in the Netherlands in comparison to other reports. Higher CK-MB levels and a LAD culprit lesion at index STEMI were important risk factors for the development of HF after STEMI.

HEART FAILURE FOLLOWING STEMI: AN AGNES STUDY

REFERENCES 1. 2. 3.

4. 5.

10. 11. 12.

13. 14.

15.

16.

17.

Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137-46. McMurray JJ, Stewart S. Epidemiology, aetiology, and prognosis of heart failure. Heart. 2000;83:596-602. Desta L, Jernberg T, Lofman I, Hofman-Bang C, Hagerman I, Spaak J, et al. Incidence, Temporal Trends, and Prognostic Impact of Heart Failure Complicating Acute Myocardial Infarction: The SWEDEHEART Registry (Swedish Web-System for Enhancement and Development of Evidence-Based Care in Heart Disease Evaluated According to Recommended Therapies): A Study of 199,851 Patients Admitted With Index Acute Myocardial Infarctions, 1996 to 2008. JACC Heart Fail. 2015;3:234-42. Kaul P, Ezekowitz JA, Armstrong PW, Leung BK, Savu A, Welsh RC, et al. Incidence of heart failure and mortality after acute coronary syndromes. Am Heart J. 2013;165:379-85 e2. Bezzina CR, Pazoki R, Bardai A, Marsman RF, de Jong JS, Blom MT, et al. Genome-wide association study identifies a susceptibility locus at 21q21 for ventricular fibrillation in acute myocardial infarction. Nat Genet. 2010;42:688-91. de Jong JS, Marsman RF, Henriques JP, Koch KT, de Winter RJ, Tanck MW, et al. Prognosis among survivors of primary ventricular fibrillation in the percutaneous coronary intervention era. Am Heart J. 2009;158:467-72. Dekker LR, Bezzina CR, Henriques JP, Tanck MW, Koch KT, Alings MW, et al. Familial sudden death is an important risk factor for primary ventricular fibrillation: a case-control study in acute myocardial infarction patients. Circulation. 2006;114:1140-5. Velagaleti RS, Pencina MJ, Murabito JM, Wang TJ, Parikh NI, Dâ&#x20AC;&#x2122;Agostino RB, et al. Long-term trends in the incidence of heart failure after myocardial infarction. Circulation. 2008;118:205762. Torabi A, Cleland JG, Khan NK, Loh PH, Clark AL, Alamgir F, et al. The timing of development and subsequent clinical course of heart failure after a myocardial infarction. Eur Heart J. 2008;29:859-70. Najafi F, Dobson AJ, Hobbs M, Jamrozik K. Late-onset heart failure after myocardial infarction: trends in incidence and survival. Eur J Heart Fail. 2008;10:765-71. Hellermann JP, Jacobsen SJ, Redfield MM, Reeder GS, Weston SA, Roger VL. Heart failure after myocardial infarction: clinical presentation and survival. Eur J Heart Fail. 2005;7:119-25. Gjesing A, Gislason GH, Kober L, Gustav Smith J, Christensen SB, Gustafsson F, et al. Nationwide trends in development of heart failure and mortality after first-time myocardial infarction 1997-2010: A Danish cohort study. Eur J Intern Med. 2014;25:731-8. Sutton MG, Sharpe N. Left ventricular remodeling after myocardial infarction: pathophysiology and therapy. Circulation. 2000;101:2981-8. Chia S, Senatore F, Raffel OC, Lee H, Wackers FJ, Jang IK. Utility of cardiac biomarkers in predicting infarct size, left ventricular function, and clinical outcome after primary percutaneous coronary intervention for ST-segment elevation myocardial infarction. JACC Cardiovasc Interv. 2008;1:415-23. Shafazand M, Rosengren A, Lappas G, Swedberg K, Schaufelberger M. Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993-2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden. Eur J Heart Fail. 2011;13:135-41. Chen J, Hsieh AF, Dharmarajan K, Masoudi FA, Krumholz HM. National trends in heart failure hospitalization after acute myocardial infarction for Medicare beneficiaries: 1998-2010. Circulation. 2013;128:2577-84. Spencer FA, Meyer TE, Goldberg RJ, Yarzebski J, Hatton M, Lessard D, et al. Twenty year trends (1975-1995) in the incidence, in-hospital and long-term death rates associated with heart failure complicating acute myocardial infarction: a community-wide perspective. J Am Coll Cardiol. 1999;34:1378-87.

PART TWO

Finding Fibrosis Patterns

Chapter

A Systematic Comparison of Cardiovascular Magnetic Resonance and High Resolution Histological Fibrosis Quantification in a Chronic Porcine Infarct Model Submitted

Johannes M.I.H. Gho1*, RenĂŠ van Es1*, Frebus J. van Slochteren2, Sanne J. Jansen of Lorkeers1, Allard J. Hauer1, Joep W.M. van Oorschot3, Pieter A. Doevendans1, Tim Leiner3, Aryan Vink4, Folkert W. Asselbergs1,5,6, Steven A.J. Chamuleau1 *Contributed equally to this article and are shared first authors

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

Interuniversity Cardiology Institute of the Netherlands, Utrecht, the Netherlands

Department of Radiology, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Pathology, University Medical Center Utrecht, Utrecht, the Netherlands

Durrer Center for Cardiogenetic Research, ICIN-Netherlands Heart Institute, Utrecht, the Netherlands

Institute of Cardiovascular Science, Faculty of Population Health Sciences, University College London, London, United Kingdom

PART TWO CHAPTER 4

ABSTRACT Background The noninvasive reference standard for myocardial fibrosis detection on cardiovascular magnetic resonance imaging (CMR) is late gadolinium enhancement (LGE). Currently there is no consensus on the preferred method for LGE quantification. Moreover myocardial wall thickening (WT) and strain are measures of regional deformation and can provide insight into local mechanical effects of myocardial fibrosis. In this article we present a high resolution systematic digital histological quantification of cardiac fibrosis and provide a detailed comparison to various in vivo CMR imaging techniques. Aim of this research was to systematically compare in vivo CMR parameters, such as LGE, WT and strain, with histological fibrosis quantification. Methods Eight weeks after 90 minutes ischemia/reperfusion of the LAD artery, 16 pigs (79.8Âą5.8 kg) underwent in vivo CMR on a clinical 3T scanner. Cine imaging and LGE sequences were performed. Histological sections from transverse heart slices were digitally analysed for fibrosis quantification. Mean percentage of fibrosis of each digitally analysed section was related to the different CMR techniques (using Segment or feature tracking software) for that slice using a linear mixed model analysis. For each CMR derived variable, R2 values were calculated as a measure of the percentage of variability in myocardial fibrosis that can be explained by a linear association with the CMR parameters. Results The full width at half maximum (FWHM) technique for quantification of LGE yielded the highest R2 of 60%. Cine derived myocardial WT explained 16 to 36% of the histological myocardial fibrosis. The peak circumferential and radial strain measured by feature tracking could explain 15% and 10% of the variance of myocardial fibrosis, respectively. Conclusions This novel method to systematically compare CMR image data with digital histological images is feasible. Myocardial WT and strain were only modestly related with the amount of fibrosis. The fully automatic FWHM analysis technique is the preferred method to detect myocardial fibrosis.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

BACKGROUND Myocardial fibrosis has been associated with heart failure and can act as a substrate for cardiac arrhythmias.1 Following myocardial infarction (MI), loss of cardiomyocytes leads to reparative fibrosis with replacement by connective tissue. Noninvasive assessment of cardiac fibrosis is important for diagnosis, predicting prognosis and treatment planning.2 The noninvasive reference standard for fibrosis detection is late gadolinium enhancement (LGE) on cardiovascular magnetic resonance imaging (CMR). Since there is no consensus on the LGE quantification techniques,3 a detailed comparison with the reference standard of histological analysis is important. For example, accurate fibrosis quantification can predict reversible myocardial dysfunction after revascularization.4, 5 Thus far, mainly correlation studies have been performed with small endomyocardial biopsies, triphenyl tetrazolium chloride stained or ex vivo hearts.6-8 Studies using

whole heart slices are scarce,9 especially for focal fibrosis. While LGE provides an accurate qualitative measure of fibrosis, it requires contrast administration with potential adverse effects and does not provide a quantitative or direct measurement of cardiac collagen.2, 10 The result of LGE differs between different imaging studies and by variable intensity threshold settings and thus relies on an adequate imaging protocol. Functional assessment of the local myocardium is typically performed visually on cine CMR images. Quantitative assessment of local myocardial function (e.g., wall thickening) can also be performed with (semi-)automatic segmentation software packages.11, 12 More recently, feature tracking (FT) has been introduced as a method to assess local myocardial deformation (strain) using cine images without the need for tagged CMR scans.13 Feature tracking is relatively quick in post processing, has shown reasonable agreement with tagging CMR when looking at global strain from complete slices and might be usable between different field strengths.14-16 However, there is debate about the agreement between strain derived from FT and tagging CMR at a segmental level, as multiple studies found poor intra- and interobserver variability for segmental strain.14, 17-19 We recently developed a method for high resolution systematic digital histological quantification of (diffuse and focal) cardiac fibrosis in a whole heart slice,20 which can provide a detailed reference for comparing different CMR imaging techniques. The aim of this study was to systematically analyse in vivo CMR derived parameters and high spatial resolution digital fibrosis quantification in a chronic porcine infarct model to compare different CMR techniques with myocardial fibrosis assessment. Parameters of interest are: LGE CMR, myocardial strain and wall thickening (WT). Myocardial strain and myocardial WT are respectively assessed by FT and manual segmentation on cine MRI.

METHODS Animal model All in vivo experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources. Experiments were approved (protocol no.: 2012.II.09.145) by the local Animal Experiments Committee (DEC) (Utrecht, the Netherlands).

PART TWO CHAPTER 4

Our protocol regarding a porcine chronic MI model has been described in detail before.21 Eight weeks after 90 minutes ischemia/reperfusion of the proximal left anterior descending artery (LAD), 16 Dalland Landrace pigs (79.8Âą5.8 kg) under continuous anesthesia underwent in vivo CMR on a clinical 3T scanner (Achieva TX, Software Release 3.2.1, Philips Healthcare, Best, the Netherlands). CMR Pigs were positioned supine with a dedicated 32-channel phased-array receiver coil over the chest and scanned using a standardized protocol. In short, for image planning scout images were obtained in short-axis and two-chamber long-axis views. ECG-gated steady-state free precession (SSFP) cine of short-axis (from apex to base of LV) and two chamber long-axis views were acquired. Thirty frames were acquired per RR cycle. Cine parameters: echo time (TE)/ repetition time (TR) 1.6/3.2ms, 13 slices, slice thickness 8 mm, resolution = 2x2mm, field of view (FOV) = 320x320mm2, bandwidth = 1200Hz and flip angle = 45Â°.

Figure 1. Cardiovascular Magnetic Resonance Analysis Methods A-C. The endo- and epicardial segmentations of the wall thickening (Segment), feature tracking (Image-Arena) and LGE (Segment) analyses respectively. The asterisk (*) indicates the point in the mid-lateral wall that was used for registration between CMR and histology. The automatic FWHM infarct delineation is shown in C (yellow). D. Resulting WT patterns of the four segments shown in A. The actual WT analysis used 60 sections. E. Radial strain analysis of 48 segments (Image-Arena); the result of the segment indicated with the asterisk in B is shown in red. F. The %TM data was exported in 360 segments. The digitized annotation map of the LV, as shown in Figure 2B, is then projected onto the LGE exported data for analysis. FWHM = full width at half maximum; LGE = late gadolinium enhancement; LV = left ventricle; %TM = fraction of transmurality; WT = wall thickening.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

LGE Late gadolinium enhancement CMR was performed using an inversion recovery 3D-turbogradient-echo-technique 15 minutes after double-dose intravenous bolus injection of a gadolinium based contrast agent (Gadovist, Bayer Healthcare, Berlin, Germany). First, a look-locker scout was performed for the optimal inversion time. Acquisition parameters for the LGE scan: inversion time (TI) = 200-270ms, TE/TR = 1.5/4.7ms, slice thickness = 6mm, spatial resolution = 1.5x1.5mm2, FOV = 300x300mm2, flip angle = 25°, 63 TFE shots, bandwidth = 300Hz, number of signals averaged = 2, SENSE acceleration = 2. CMR imaging analysis Segment Offline image analysis to derive WT and LGE was performed using Segment version v1.9 R3590

(http://segment.heiberg.se, Medviso AB, Lund, Sweden).12 In all datasets, the short-axis image corresponding to the histological slice was selected based on its location, and used for further analysis (Figure 1). In the short-axis cine images, LV endo- and epicardial borders were manually segmented in all frames. The segmentations were copied to the corresponding LGE slice (Figure 1C). Manual adjustment was performed if necessary. From the short-axis cine dataset the absolute WT (mm) per image frame of 60 LV segments was exported using the reportSlice function of Segment. The end systolic absolute WT of each segment was used for further analysis. Feature Tracking Cardiac strain analysis by FT was performed using the Image-Arena 2D Cardiac Performance Analysis software toolbox version 1.2 (TomTec Imaging Systems, Unterschleissheim, Germany). The end-diastolic endo- and epicardial contours of the slice were manually traced (Figure 1B). The mid-lateral point along the endocardial contour was marked as an anatomical reference and data was exported for registration purposes. For all 48 segments, raw data containing circumferential strain (εcc), radial strain (εrr) and WT (endo to epi distance), was exported for further analysis (Figure 1E). Viability analysis Segment For viability analysis of short-axis LGE datasets, scar was delineated using automatic full width at half maximum (FWHM) (Figure 1C), automatic uncorrected and manually corrected standard deviation (SD) from remote (2, 3 and 5SD) algorithms. For the SD methods, the remote healthy myocardium of the lateral wall was selected as remote. Manual corrections of the infarct area were performed in the subgroups. The fraction of area based transmurality (%TM), the mean infarct size fraction of the wall thickness, was analysed in 360 equal segments over the LV wall. For myocardial signal intensity (MSI) and each scar delineation method the %TM was exported using Segment’s reportBullsEye function. Histology Following CMR, the animals were sacrificed by exsanguination under general anaesthesia and the hearts were excised and cut into transverse (short-axis) 1cm thick slices beginning from

PART TWO CHAPTER 4

apex to base. Each third transverse slice was fixed in formalin, cut into smaller sections and a map of the heart slice was drawn to annotate the origin of each tissue specimen (Figure 2A). These sections were embedded in paraffin and stained with Massonâ&#x20AC;&#x2122;s trichrome. The slides were scanned at 20X magnification as described before.22 Images were extracted using Aperio ImageScope v.12.0.0.5039 (Aperio, Vista, CA, USA) and resized to 10% for digital analysis. Histological analysis Digital histological analysis was performed systematically as previously described, using the in house developed software package Fibroquant (http://sourceforge.net/projects/fibroquant).20 The epicardium, defined as the outer region of fatty tissue bordered by the first row of cardiomyocytes, was excluded from further analysis. The remaining myocardium, including the compact and trabeculated region was analysed as a whole. The percentage of connective tissue (blue), cardiomyocytes (red) and adipose tissue (cells with non-stained cytoplasm) was digitally quantified using Fibroquant. The results were annotated to their corresponding heart region using an automated algorithm and transformed to a standardized schematic overview (Figure 2B). Matching MRI with histology After the construction of the standardized schematic overview, the exported high detail MRI data was averaged over the regions delineated by the histological sections (e.g., Figure 1F). A landmark in the mid-lateral wall of all datasets was used as a reference point to assure optimal registration. This reference was defined as the point opposite both hinge points of the right ventricle. Statistical analysis Statistics were performed using IBM SPSS Statistics (Version 20.0, IBM Corporation, Armonk, New York, United States). We compared different CMR techniques with percentages of fibrosis per section using a linear mixed model analysis. For fibrosis, the amount of residual variance (Ď&#x192;2) within the animals and the variance (intercept, Ď&#x201E;) between animals were calculated

Figure 2. Digital Histological Analysis Methods A. Map containing the origin of the histological sections. B. The digitized left ventricular annotation map of image A. C. The manual tracings of the four regions across the histological myocardial section, corresponding to location 5 in A and B.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

(null model). Subsequently, CMR parameters were added to the model (full model). R2 (Snijders & Bosker, 2012), was calculated as 100% minus the ratio of the full and null models (equation 1),23, 24 representing the explained variance. (eq. 1)

Figure 3. Results of Histological Analysis A-B. Results of the histological analyses of sections 5 and 7 respectively (Figure 2A&B). C. Analysis of the whole heart of this animal. D. Mean fibrosis content in the left ventricle of all animals (n=15). LV = left ventricle; RV = right ventricle.

RESULTS Histology A total of 116 histological sections (16 animals) were successfully stained, segmented (Figure 2C) and analysed (Figure 3A&B). Digital maps with the annotated origins of each histological section were used to construct schematic histological overviews for each animal (e.g., Figure 3C). One animal (8 sections) was excluded from further analysis because of lack of histological

PART TWO CHAPTER 4

fibrosis due to a sampling error (Supplementary Figure 1), 108 sections remained. Mean fibrosis percentages (15 animals) were log transformed to reduce right-skewness and heterogeneity of variance. Myocardial fibrosis was mainly observed in the anteroseptal wall (Figure 3D), corresponding to the LAD territory. The 8-week old LAD infarct model used in this study yielded fractional fibrosis values of maximally 57%. Cine CMR analysis Segment Myocardial WT curves were successfully extracted for 15 animals in 60 segments. After matching the CMR segments according to the histological sections, mean end systolic absolute WT was 2.4±2.4mm (range -2.1 to 7.2mm) (Figure 4A, 5A). The explained variance of WT for histological myocardial fibrosis was 36% (Table 1). Feature tracking Data of 15 animals was successfully extracted and analysed in 48 segments. After matching the data according to histological sections the mean εcc, εrr and WT were -12.1±12.1% (range -38.5 to 24.3%), 26.9±34.7% (range -26.0 to 171.9%) and -1.6±2.5mm (range -4.6 to 7.7mm) respectively (Figure 4B, 5B-D). Feature tracking derived εcc, εrr and WT explained 15%, 10% and 16% of myocardial fibrosis respectively (Table 1).

Table 1. Comparison of Wall Thickening, Strain and LGE with Histological Fibrosis n = 15

σ2full

τ00full

Cine WT (Segment)

1.07

0.02

36%

Cine WT (Image-Arena)

1.44

16%

Radial Strain

1.54

10%

Circumferential Strain

1.45

15%

2SD

0.94

0.07

41%

3SD

0.82

0.11

46%

5SD

0.65

0.12

55%

2SD Manual Subgroup

0.66

0.03

60%

3SD Manual Subgroup

0.63

0.05

60%

5SD Manual Subgroup

0.65

0.07

58%

FWHM

0.62

0.07

60%

Myocardial Signal Intensity

0.68

0.14

52%

The null model used in the statistical analysis containing the variance in myocardial fibrosis was expressed as σ2null+ τ00null=1.71*. The full model containing the unexplained variance in myocardial fibrosis after adding explanatory CMR parameters is expressed as σ2full+ τfull. The resulting R2 value was calculated as 100% - the fraction of unexplained variance (eq. 1). *= Although all convergence criteria were satisfied, the hessian matrix was not positive definite, as a consequence the residual and random intercept variance were pooled. FWHM = full width at half maximum; LGE = late gadolinium enhancement; R2 = explained variance of myocardial fibrosis; SD = standard deviation; σ2 = intra animal variance; τ = inter animal variance; WT = wall thickening.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

Figure 4. Comparisons of Cardiovascular Magnetic Resonance Parameters and Fibrosis (n = 15) The dots represent 108 separate sections from 15 animals. R2 values (explained variance) are derived from a separate linear mixed model analysis. Fibrosis was compared with: A. Wall thickening results of the cine analysis (Segment); B. Circumferential strain analysis (Image-Arena); C. Myocardial signal intensity analysis (Segment); D. LGE FWHM analysis (Segment). FWHM = full width at half maximum; LGE = late gadolinium enhancement.

Viability CMR analysis Segment The mean LGE scar transmurality and MSI (15 animals) is shown in Figure 6. The variance between animals was small (Ď&#x201E; range 0 to 0.14) (Table 1). The automatic FWHM algorithm resulted in 60% explained variance for histological myocardial fibrosis. Manually corrected SD from remote methods yielded an explained variance between 58 - 60%. The variance within animals (Ď&#x192;2) decreased (0.94 to 0.65) with more stringent automatic segmentation methods (2-5 SD), while it remained similar in the manually edited SD subgroup. The MSI of the LGE images, in which no additional interaction is required, explained the myocardial fibrosis for 52%. Scar transmurality compared to fibrosis and functional parameters In the boxplots (Figure 7), the median fibrosis ranged from 1.4% in the segments without LGE to 22.7% in the segments with 75-100% LGE transmurality using FWHM. Median WT ranged

PART TWO CHAPTER 4

from 4.3 to 0.0mm and εcc and εrr, ranged from -15.0% to -7.7% and 40.0% to -1.0% respectively between the different increasing amounts of LGE transmurality.

Figure 5. Mean Wall Thickening and Strain Analysis (n = 15) A. The averaged results of the cine WT analysis (Segment). B-D. The averaged results of the feature tracking analysis, the WT, radial (εrr) and circumferential (εcc) strain respectively (Image-Arena). εcc = circumferential strain; εrr = radial strain; WT = wall thickening.

Figure 6. Mean Late Gadolinium Enhancement Analysis (n = 15) Averaged results of the fraction of transmurality in 360 segments for the 2, 3 and 5SD automatic methods, the manually corrected subgroup and the fully automatic FWHM analysis. In the bottom right corner, the averaged myocardial signal intensity is shown. FWHM = full width at half maximum; SD = standard deviation.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

Figure 7. Boxplots of Late Gadolinium Enhancement Scar Transmurality and Fibrosis or Functional Parameters (n = 15) A-D. The dots represent outliers. Late gadolinium enhancement fraction of transmurality based on FWHM divided in different subgroups (0, 0-25, 26-50, 51-75 and 76-100%) compared to fibrosis percentage, wall thickening from feature tracking, and radial or circumferential strain from FT. FT = feature tracking; FWHM = full width at half maximum; WT = wall thickening.

DISCUSSION This study demonstrates a novel systematic method to compare different in vivo CMR techniques with high spatial resolution histological analyses in a chronic porcine infarct model. To the best of our knowledge, this is the first time such a systematic comparison has been performed and this method can be readily adapted for use with other imaging modalities such as SPECT, CT, PET and echocardiography. For LGE imaging, we found that the FWHM and the 2, 3 or 5SD methods with manual correction were the best methods to quantify the amount of myocardial fibrosis. Although it is expected that measures of myocardial deformation would be affected in infarcted myocardium, myocardial WT and strain were only modestly related with the amount of myocardial fibrosis.

PART TWO CHAPTER 4

Myocardial WT The reference standard for the noninvasive assessment of cardiac anatomy and function is CMR.2 While cine imaging is not intended for fibrosis quantification, MI is associated with a lower amount of WT in the infarct area.6 We measured myocardial WT using two different software analysis tools, Segment and Image-Arena. Our analysis shows that WT derived from Segment performs better in explaining myocardial fibrosis compared to Image-Arena (R2 = 36% vs. 16%). In Segment, both the endo- and epicardial contours of the selected slice were manually traced in every frame. In Image-Arena the segmentation was performed in only one frame and the tracing in subsequent frames were calculated automatically by the software algorithm. The FT workflow could limit accuracy and cause variability, due to the lack to perform specific adjustments in multiple frames which is inherent to the algorithm. Furthermore, through plane motion and filtering algorithms resulting in loss of detail are likely to introduce erroneously tracked features, thereby affecting the FT results. Cardiac deformation In our analysis of the complete LV slices, both radial and circumferential strain based on FT were poorly associated with myocardial fibrosis (R2 = 10% and 15% respectively). As previously described by Cowan et al., differences in strain could occur between two regions indicated as healthy by LGE.19 From our study it can be concluded that FT based strain imaging is less applicable to identify regional myocardial fibrosis in this experimental infarct model. This might be caused by the fact that strain is not a strictly local phenomenon, but is the resulting deformation caused by the contraction of cardiomyocytes in the vicinity of the infarct. Although local myocardial deformation assessment is not a direct measurement of fibrosis, strain analysis might supply new insights in local cardiac biomechanics after ischemic injury, but rigorous clinical validation is required. Viability Compared to the reference standard of fibrosis on histology, FWHM and the 2, 3 or 5SD methods with manual correction were the best explanatory variables for variance in fibrosis. The FWHM technique uses half of the maximal signal intensity within the scar region as a threshold to determine the infarct area.25, 26 The used SD from remote methods delineate scar by pixels with an image intensity higher than the mean plus 2, 3 or 5SD from the mean in a non-infarcted remote region.6 Late gadolinium enhancement has been previously validated ex vivo using a MSI analysis compared to quantitative histopathological sections with fibrosis to discriminate between healthy myocardium, dense scar and the border zone.8 A previous clinical study showed limited reproducibility of the SD from remote techniques with manual correction and found the FWHM technique the most reproducible method.3 In our analysis, the MSI measured from LGE images resulted in worse explanatory values for myocardial fibrosis compared to FWHM (52% vs. 60%). The between animal variance of LGE MSI compared to FWHM was higher (Ď&#x201E; 0.14 vs. 0.07) which could be explained by MSI variations due to small variations in gadolinium injection time to acquisition. Because of the improved explanation for myocardial fibrosis and reduced manual interaction needed, FWHM would be the preferred method of choice from this study for infarct quantification on LGE CMR.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

In diffuse fibrosis, with less distinct LGE or remote areas, the FWHM method could be technically more difficult and T1-mapping might be more promising.9 Limitations The final resolution of the comparison between histology and MRI is determined by the method of histological sections cut from the LV, typically 7 or 8 sections per heart slice in this study. Only one short-axis slice of the CMR images was used and the high detail MRI data was averaged to match exactly with the histology data, leading to a loss of detail, especially along the infarct border zone. For future studies dissection of the entire heart using a microtome could be considered to preserve the gross cardiac anatomy. Resulting high resolution images could later be analyzed and subdivided using automated algorithms to allow for a more detailed comparison with high detail imaging modalities.

In the assessment of local cardiac function (e.g., WT, Îľcc and Îľrr) it is likely that a higher fibrosis percentage would have led to a stronger relation with worsening of functional parameters. The right ventricle was excluded from all comparisons because the conventional analyses applied on the cine and LGE images in this study did not allow an accurate analysis of the right ventricle. Future implications The comparison method used in this study could be applied on data from other imaging modalities such as SPECT, PET, CT, echocardiography and other CMR sequences (e.g., T1mapping) and can also be translated for use with a 3D model. While the proximal LAD ischemia/ reperfusion model used in this study produced consequent isolated anteroseptal infarctions, it would be in the interest of external validity to study other infarct sizes and locations using this method. Improved fibrosis detection with CMR will be applicable to a broad clinical spectrum ranging from diagnostic to therapeutic outcomes, including cell therapy, ablation therapy, valvular diseases, cardiomyopathies and ischemic heart disease. For example, precise identification of the fibrotic region allows for accurate therapy guidance to the target area with the aim to ultimately improve clinical outcome for patients.

CONCLUSIONS In conclusion, the novel systematic method to compare high resolution in vivo CMR imaging with detailed histological fibrosis data was feasible and can readily be applied to other imaging data. Locally measured functional parameters such as WT, and measures of myocardial deformation derived from FT: radial and circumferential strain related modestly with local myocardial fibrosis, yet can be used to gain insight into local cardiac mechanics. The fully automatic FWHM algorithm applied on the reference standard LGE CMR showed to be preferred to detect myocardial fibrosis in a chronic in vivo infarct model.

PART TWO CHAPTER 4

Acknowledgments We kindly acknowledge Grace Croft, Marlijn Jansen, Evelyn Velema, Joyce Visser, Merel Schurink, Martijn van Nieuwburg and Gerard Marchal for their assistance with the animal experiments. We would like to thank Jaco Zwanenburg and Martijn Froeling for performing the CMR experiments. We are also grateful to Rebecca Stellato for statistical advice.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

REFERENCES 1. 2. 3.

10. 11. 12.

13.

14.

15.

16.

17.

18.

Weber KT, Sun Y, Bhattacharya SK, Ahokas RA, Gerling IC. Myofibroblast-mediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol. 2013;10:15-26. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA. Assessment of myocardial fibrosis with cardiovascular magnetic resonance. J Am Coll Cardiol. 2011;57:891-903. Flett AS, Hasleton J, Cook C, Hausenloy D, Quarta G, Ariti C, et al. Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging. 2011;4:150-6. Kim RJ, Wu E, Rafael A, Chen EL, Parker MA, Simonetti O, et al. The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med. 2000;343:1445-53. Bondarenko O, Beek AM, McCann GP, van Rossum AC. Revascularization in patients with chronic ischaemic myocardial dysfunction: insights from cardiovascular magnetic resonance imaging. Eur Heart J Cardiovasc Imaging. 2012;13:985-90. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, et al. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation. 1999;100:1992-2002. Malliaras K, Smith RR, Kanazawa H, Yee K, Seinfeld J, Tseliou E, et al. Validation of contrastenhanced magnetic resonance imaging to monitor regenerative efficacy after cell therapy in a porcine model of convalescent myocardial infarction. Circulation. 2013;128:2764-75. Pop M, Ghugre NR, Ramanan V, Morikawa L, Stanisz G, Dick AJ, et al. Quantification of fibrosis in infarcted swine hearts by ex vivo late gadolinium-enhancement and diffusion-weighted MRI methods. Phys Med Biol. 2013;58:5009-28. Iles LM, Ellims AH, Llewellyn H, Hare JL, Kaye DM, McLean CA, et al. Histological validation of cardiac magnetic resonance analysis of regional and diffuse interstitial myocardial fibrosis. Eur Heart J Cardiovasc Imaging. 2015;16:14-22. Bellin MF, Van Der Molen AJ. Extracellular gadolinium-based contrast media: an overview. Eur J Radiol. 2008;66:160-7. Attili AK, Schuster A, Nagel E, Reiber JH, van der Geest RJ. Quantification in cardiac MRI: advances in image acquisition and processing. Int J Cardiovasc Imaging. 2010;26 Suppl 1:27-40. Heiberg E, Sjogren J, Ugander M, Carlsson M, Engblom H, Arheden H. Design and validation of Segment--freely available software for cardiovascular image analysis. BMC Med Imaging. 2010;10:1. Hor KN, Gottliebson WM, Carson C, Wash E, Cnota J, Fleck R, et al. Comparison of magnetic resonance feature tracking for strain calculation with harmonic phase imaging analysis. JACC Cardiovasc Imaging. 2010;3:144-51. Augustine D, Lewandowski AJ, Lazdam M, Rai A, Francis J, Myerson S, et al. Global and regional left ventricular myocardial deformation measures by magnetic resonance feature tracking in healthy volunteers: comparison with tagging and relevance of gender. J Cardiovasc Magn Reson. 2013;15:8. Schuster A, Morton G, Hussain ST, Jogiya R, Kutty S, Asrress KN, et al. The intra-observer reproducibility of cardiovascular magnetic resonance myocardial feature tracking strain assessment is independent of field strength. Eur J Radiol. 2013;82:296-301. Khan JN, Singh A, Nazir SA, Kanagala P, Gershlick AH, McCann GP. Comparison of cardiovascular magnetic resonance feature tracking and tagging for the assessment of left ventricular systolic strain in acute myocardial infarction. Eur J Radiol. 2015. Morton G, Schuster A, Jogiya R, Kutty S, Beerbaum P, Nagel E. Inter-study reproducibility of cardiovascular magnetic resonance myocardial feature tracking. J Cardiovasc Magn Reson. 2012;14:43. Wu L, Germans T, Guclu A, Heymans MW, Allaart CP, van Rossum AC. Feature tracking compared with tissue tagging measurements of segmental strain by cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2014;16:10.

PART TWO CHAPTER 4

19. Cowan BR, Peereboom SM, Greiser A, Guehring J, Young AA. Image Feature Determinants of Global and Segmental Circumferential Ventricular Strain From Cine CMR. JACC Cardiovasc Imaging. 2014. 20. Gho JM, van Es R, Stathonikos N, Harakalova M, te Rijdt WP, Suurmeijer AJ, et al. High resolution systematic digital histological quantification of cardiac fibrosis and adipose tissue in phospholamban p.Arg14del mutation associated cardiomyopathy. PLoS ONE. 2014;9:e94820. 21. Koudstaal S, Jansen of Lorkeers S, Gho JM, van Hout GP, Jansen MS, Grundeman PF, et al. Myocardial infarction and functional outcome assessment in pigs. J Vis Exp. 2014. 22. Huisman A, Looijen A, van den Brink SM, van Diest PJ. Creation of a fully digital pathology slide archive by high-volume tissue slide scanning. Hum Pathol. 2010;41:751-7. 23. Snijders TAB, Bosker RJ. Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. 2nd ed. London: Sage Publications; 2012. 24. LaHuis DM, Hartman MJ, Hakoyama S, Clark PC. Explained Variance Measures for Multilevel Models. Organizational Research Methods. 2014;17:433-51. 25. Amado LC, Gerber BL, Gupta SN, Rettmann DW, Szarf G, Schock R, et al. Accurate and objective infarct sizing by contrast-enhanced magnetic resonance imaging in a canine myocardial infarction model. J Am Coll Cardiol. 2004;44:2383-9. 26. Hsu LY, Natanzon A, Kellman P, Hirsch GA, Aletras AH, Arai AE. Quantitative myocardial infarction on delayed enhancement MRI. Part I: Animal validation of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging. 2006;23:298-308.

COMPARISON OF CMR AND HISTOLOGICAL FIBROSIS

SUPPLEMENTAL MATERIAL

Supplementary Figure 1. Whole heart slice fibrosis analysis, one animal (excluded from analysis due to absence of left ventricular fibrosis). The percentage of fibrosis is shown using a color scale. Ant. = anterior; LV = left ventricle; Post. = posterior; RV = right ventricle.

PART TWO

Finding Fibrosis Patterns

Chapter

Endogenous Contrast MRI of Cardiac Fibrosis: Beyond Late Gadolinium Enhancement Published as J Magn Reson Imaging. 2015;41:1181-9

Johannes M.I.H. Gho1*, Joep W.M. van Oorschot2*, Gerardus P.J. van Hout1, Martijn Froeling2, Sanne J. Jansen of Lorkeers1, Imo E. Hoefer1, Pieter A. Doevendans1, Peter R. Luijten2, Steven A.J. Chamuleau1, Jaco J.M. Zwanenburg2 *Both authors contributed equally to this manuscript 1

Department of Cardiology, University Medical Center Utrecht, Utrecht, The Netherlands

Department of Radiology, University Medical Center Utrecht, Utrecht, The Netherlands

PART TWO CHAPTER 5

ABSTRACT The aim of this review is to provide an overview of detection of cardiac fibrosis with MRI using current standards and novel endogenous MRI techniques. Assessment of cardiac fibrosis is important for diagnosis, prediction of prognosis and follow up after therapy. During the last years, progress has been made in fibrosis detection using MRI. Cardiac infarct size can be assessed non-invasively with late gadolinium enhancement (LGE). Several methods for fibrosis detection using endogenous contrast have been developed, such as native T1-mapping, T1Ď -mapping, Magnetization Transfer Imaging and T2*-mapping. Each of these methods will be described, providing the basic methodology, showing potential applications from applied studies, and discussing the potential and challenges or pitfalls. We will also identify future steps and developments that are needed for bringing these methods to the clinical practice.

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

INTRODUCTION In myocardial tissue, the interstitial collagen matrix surrounds and supports the cardiac myocytes. An excessive increase in myocardial collagen is defined by thickening and scarring of connective tissue, usually as a result of injury. Scar tissue can lead to heart failure by decreasing ventricular systolic function, induction of diastolic dysfunction and adverse cardiac remodeling. There are different types of myocardial fibrosis that result from different etiologies. The first type is replacement fibrosis, that results from the replacement of dead cardiomyocytes. Replacement fibrosis occurs after injury such as myocardial infarction, and leads to a localized distribution of collagen, which mainly consists of type 1 collagen. The second type is interstitial reactive fibrosis, which is a progressive form of fibrosis resulting from an increase in collagen synthesis by myofibroblasts. Interstitial fibrosis normally shows a diffuse distribution. It can be seen in almost any cardiac disease, and is common in idiopathic dilated cardiomyopathy and hypertrophic cardiomyopathy,1â&#x20AC;&#x201C;3 but other pathologies, such as amyloidosis and hypertension have also been associated with infiltrative interstitial fibrosis in

the heart and expansion of the extracellular matrix.4,5 Regardless of etiology, maladaptive cardiac fibrosis increases the risk of potentially lethal complications such as arrhythmias and heart failure, underlining the importance of effective treatment. In addition to existing treatments, there is an emerging focus on regenerative medicine. One example is stem cell therapy, which aims to restore damaged myocardium by stem cell injection and as a result myocardial function. 6 Adequate diagnosis and precise identification of the fibrotic region will improve the effectiveness of the treatment, but also allows for better prediction of the prognosis and outcome after therapy.7 Hence, precise detection of myocardial fibrosis becomes more important.8 Current techniques such as endomyocardial biopsies can provide a qualitative and quantitative assessment of collagen. However they do require an invasive procedure and provide only regional information.2 Cardiac infarct size can be assessed non-invasively with Cardiac Magnetic Resonance Imaging (MR). The most common technique is late gadolinium enhancement (LGE), which allows for indirect detection of fibrosis using an exogenous contrast agent.9 Currently, there is an increased interest for direct detection of myocardial fibrosis without the use of an exogenous contrast agent, to overcome the limitations associated with contrast agents and to obtain a quantitative standardized measurement to assess myocardial fibrosis. This is possible by exploiting the various endogenous contrast mechanisms available using MRI. Several methods for direct detection of myocardial fibrosis have been proposed. In this review the current standard MRI method, LGE, will be described, followed by an overview of novel contrast mechanisms, such as T1, T2, T2*, T1Ď and magnetization transfer imaging (Table 1). These different endogenous MRI techniques will be illustrated with data from previous published literature and a pilot study in a porcine model of chronic ischemic heart failure (see Materials and MRI methods).

76 - Assessment of myocardial edema in acute phase - Direct method without need for exogenous contrast - Detection of intramyocardial hemorrhage

- Direct method without need for exogenous contrast

- T2 depends on tissue composition and is mainly influenced by edema

- T2* reflects both local tissue content and structure, which are both changed when collagen content changes

- T1Ď depends mainly on macromolecular interactions in the tissue

- Based on the transfer of magnetization from the hydrogen nuclei with restricted motion to hydrogen nuclei of free water, associated with macromolecular sensitivity

T2 mapping

T2*

T1Ď

Magnetization transfer imaging

- Direct method without need for exogenous contrast

- T1 depends on tissue composition and could change with varying collagen content

Native T1 mapping

- Assessment of diffuse fibrosis

- Not yet (pre-)clinically validated

- Homogeneity of the spin lock pulse is important - High Specific absorption rate (SAR) of spin lock pulses - Long acquisition time to obtain a T1Ď map - Specificity for fibrosis not yet known - Not yet clinically validated

- Influence of macroscopic field inhomogeneities - Shimming is very important - Unclear whether hemorrhage can be distinguished from fibrosis - Not yet clinically validated

- Only proven to work in acute phase - Minor tissue contrast in chronic myocardial fibrosis

- Only minor contrast differences (low sensitivity) - Not yet clinically validated - Specificity for collagen not yet known

- Contrast administration, causing variability and possible toxic effects - Require long acquisition time to acquire pre- and post contrast T1 map - Blood sample analysis necessary for hematocrit - Not yet clinically validated

- Quantitative assessment Extra Cellular Volume (ECV)

- Contrast enhanced T1 mapping relies on the same principle as LGE, but provides absolute T1 relaxation times after contrast injection.

Contrast enhanced T1 mapping (ECV mapping)

Methods using endogenous contrast (without contrast agents)

- Limited quantitative assessment of fibrosis - Restricted assessment of diffuse fibrosis - Contrast administration, causing possible toxic effects / variability due to lack of standardized protocol

- Clinically validated - Qualitative assessment of fibrosis - Prognostic value

Disadvantages

- Relies on difference in contrast agent washout between normal and diseased myocardium

Advantages

Late gadolinium enhancement (LGE)

Contrast-enhanced methods

Methodology

Table 1. Overview of Cardiac MR fibrosis detection methods

PART TWO CHAPTER 5

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

Contrast enhanced MRI of myocardial fibrosis Late gadolinium enhancement Late gadolinium enhancement is the non-invasive reference standard for detection of myocardial infarct size and location. The technique relies on the difference in contrast agent washout between normal and diseased myocardium.10 Gadolinium contrast agents reduce the T1 relaxation time of tissue. In regions of myocardial fibrosis or necrosis, the retention of gadolinium contrast agent is prolonged compared to healthy myocardium. This results in an increased signal intensity of this region on a T1 weighted MRI image, by adding an inversion pulse with a well-chosen inversion time to null the signal of healthy myocardium. Replacement fibrosis, e.g., due to ischemic heart disease can be clearly distinguished as it is well-delineated on LGE images. LGE cardiac MR provides prognostic value as has been demonstrated in several ischemic and non-ischemic cardiomyopathies.11–14 LGE also provides guidance to therapeutic strategies, usually in combination with other modalities (e.g., echocardiography and electromechanical mapping). This has been demonstrated for different therapeutic interventions. LGE in

combination with electromechanical mapping allows to guide cell injections to the target area in stem cell therapy experimentally. The use of LGE prior to cardiac resynchronization therapy 15

(CRT) can improve appropriateness of CRT, and implantable cardioverter-defibrillator (ICD) implantation. This reduces the rate of ‘non-responders’ and improves arrhythmia risk stratification by guidance of left ventricular lead deployment away from scarred myocardium for an improved clinical outcome.16–18 Despite the clinical value of LGE in determining infarct size, there are some limitations and drawbacks to this method. Adverse reactions after intravenous administration of gadoliniumbased contrast agents are rare but potentially life threatening, including anaphylactic reactions and especially in patients with renal failure contrast induced nephropathy and nephrogenic systemic fibrosis.19,20 Using data from the Food and Drug Administrations Adverse Event Reporting system between 1988 and 2012, 614 anaphylaxis cases have been reported, leading to mortality in 7.2% of the reports related to gadolinium based contrast agents (GBCA). 21 GBCAs have also been associated with nephrotoxicity, especially when given at high doses (>0.3 mmol/kg).22 The incidence nephrogenic systemic fibrosis (NSF) prior to 2008 varied from 0.26% in patients on dialysis up to 8.8% in patients with a eGFR smaller than 15 ml/min/1.73m2 without hemodialysis.23 Since the connection between NSF and GBCAs has become known, MRI protocols have changed and incorporated a contra-indication for using GBCAs in patients with severe renal failure. MRI methods based on endogenous contrast may be particularly important for these patients. Next to adverse reactions, the results from LGE vary between different imaging studies. The infarct size can be significantly overestimated, depending on the applied intensity threshold settings, or on the human observer if manual outlining is performed.9,24 The sensitivity to the intensity threshold is even higher in patients with in various cardiomyopathies that result in less focal areas of fibrosis.2 Furthermore, precise timing of the MRI exam after contrast agent administration is important to prevent over- or underestimation of scarred tissue and thus an adequate, standardized protocol is needed. Because of this variability, the LGE method is less suitable for longitudinal studies, such as follow up after treatment. Finally, the most important

PART TWO CHAPTER 5

limitation is the difficulty to apply LGE for the detection of diffuse interstitial myocardial fibrosis. LGE relies on the difference in signal intensity between fibrotic and normal areas to generate contrast. Lack of an area with clearly unaffected myocardium will impair detection of diffuse fibrosis by LGE. Contrast enhanced T1 mapping (ECV mapping) Recent developments have enabled the measurement of a high resolution T1 map of the human myocardium in a single breath hold, with the so called Modified Look-Locker Inversion Recovery (MOLLI) sequence.25 Similar to the LGE method, a T1 map acquired after gadolinium injection can show differences in local gadolinium concentration. T1 mapping however provides a quantitative T1 relaxation time per voxel, which alleviates the need for a remote or healthy area in the analysis. By measuring the quantitative T1 relaxation time constants instead of using qualitative T1 contrast, it is possible to overcome the limitations of LGE in patients with diffuse myocardial fibrosis.26 With the pre- and post-contrast T1 of myocardium and blood, adjusted for hematocrit, the extracellular volume (ECV) can be calculated,27 using the following formula: ECV = (1 - hematocrit) (Î&#x201D;R1myocardium/Î&#x201D;R1blood). It was shown that contrast-enhanced T1 mapping allows non-invasive quantification of diffuse myocardial fibrosis by the calculation of an ECV map.28,29 Using equilibrium contrast assessment, increased diffuse myocardial fibrosis was found in case of LV impairment and diastolic dysfunction in case of patients with severe aortic stenosis.30 In a large group of 793 patients, the ECV (in areas not containing an infarct) was found to be associated with mortality and/or the need for a heart transplant or cardiac assist device.31 The equilibrium contrast method requires 30 to 45 minutes to reach contrast equilibrium, making this a complex and time-consuming technique.28 Recent studies showed that a single bolus approach is a suitable alternative in clinical practice sufficient for most myocardial ECV applications,32 as equally accurate estimates of the ECV were found.33 Although the application of contrast enhanced myocardial T1 mapping and ECV mapping provides a valuable tool to study cardiac fibrosis, it still needs to be validated in further clinical studies.34 Nonetheless, there are still limitations that stimulate the search for other imaging techniques. Calculation of an ECV map requires the acquisition of 2 different T1 maps, one before and one 15-20 minutes after contrast injection. Furthermore, the hematocrit value should be determined, which requires the withdrawal of blood before the MRI exam. This can be done by the intravenous line that is already in place for the gadolinium injection, but creates a time-consuming procedure in clinical workflow, as the blood sample needs to be analyzed, and the resulting ECV map can only be calculated afterwards with the hematocrit value, pre- and post T1 map combined. To overcome these problems and further simplify the procedure, a single post-contrast T 1 measurement was proposed. It was shown that an isolated post-contrast T1 measurement is insufficient to provide reliable results.35 Endogenous MRI contrast of myocardial fibrosis Native T1 mapping Where the native T1 map is needed for the calculation of an ECV map, the information that can be obtained from an isolated native T1 map can also be used. It has been shown that T1 mapping

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

without contrast agent can provide information on diffuse fibrosis.36,37 Native T1 values provide diagnostic accuracy to discriminate between normal and diffuse fibrosis in patients with nonischemic dilated cardiomyopathy and hypertrophic cardiomyopathy.38 An example of T1 mapping to detect myocardial fibrosis in patients with aortic stenosis is shown in Figure 1.37 Benefits of native T1 mapping without the use of a contrast agent would be the independence of renal function and timing of the measurement of the outcome of the quantitative T1 map. Main drawback of the method is that reported changes in T1 after fibrosis are very small, which leads to weak sensitivity of the method.37 Moreover, since the resulting T1 values in the T1 map are sensitive to the pulse sequence and scanner used in the experiment, use of native T1 mapping for diagnosis of diffuse fibrosis might be challenging. A standardized imaging protocols should be used to overcome these problems.32 A recent study in a canine animal model has shown the first evidence that native T1 mapping at 3T can be used to determine the location, size and transmurality of chronic myocardial infarction, with high sensitivity (95 %) and high specificity (97%).39 Future work should focus on the translation and validation of this technique in patient studies.

Figure 1. Color maps of T1 values using shortened modified Lookâ&#x20AC;&#x201C;Locker inversion in a mid-ventricular short-axis slice with the corresponding slice with late gadolinium enhancement (LGE) imaging. The left-hand panel shows a normal volunteer (T1=944 ms). The middle panels show moderate aortic stenosis (AS) with moderate left ventricular hypertrophy (T1=951 ms). The right-hand panel shows severe AS with severe left ventricular hypertrophy (T1=1020 ms).37 (Adapted from Ref. 37, with permission)

PART TWO CHAPTER 5

Figure 2. T2 maps with corresponding LGE images of three animals with acute myocardial infarction. The area of the myocardium with a higher T2 correlates with the enhanced area on the LGE images. (Adapted from Ref. 41, with permission)

T2 mapping The T2 relaxation parameter is prolonged in regions of edema. Edema measured by T2 mapping in animal studies has been shown to correlate with the region of acute ischemic injury, which led to interest in using T2-weighted imaging to assess myocardial damage in the acute stage of ischemia.40 An example of detection of acute myocardial edema in an animal model with T2 mapping is shown in Figure 2.41 Also in patients the T2 relaxation time provides an effective parameter to detect myocardial edema in the case of acute myocardial infarction.42 For chronic myocardial fibrosis, T足2 is less effective to generate contrast with healthy myocardium.42 In a study in diabetic mice at ultrahigh field (11.75 T) however, a significant relation was found between collagen fractional area and T2 relaxation times.43 Further research is needed to investigate if these results can be translated to application in a clinical setting at lower field strengths. T2* mapping The T2* relaxation time constant characterizes the relaxation of the transverse magnetization, as a result of static magnetic field inhomogeneities.44 Therefore, T2* mapping enables the measurement of macroscopic and mesoscopic magnetic field inhomogeneities. Macroscopic inhomogeneities are mainly caused by inhomogeneity of the main magnetic field such as imperfect shimming or to air-tissue transitions. Mesoscopic field inhomogeneities, however, are induced by the biological structure of the tissue. For example, it has been shown that T2* mapping enables the imaging of myocardial fiber structure.45 Currently the main use of T2* mapping cardiac MR is for assessment of myocardial iron. It has been proven that T2* mapping can provide insight on cardiac iron overload in thalassemia major patients and other iron overload conditions.46,47

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

From cartilage, it is known that the T2* relaxation time in collagen rich tissue becomes very short.48 A short T2* means that the signal decays very quickly after excitation. To enable detection of such ‘short living’ signal, Ultra Short Echo Time (UTE) MRI can be used.49 Ex vivo results in a rat model of myocardial infarction show that the area with a short T2* as detected with UTE MRI, correlates with the area of myocardial fibrosis as determined by histology using a picrosirius red staining.50 Iron staining was negative in this study, so the T2* shortening could be attributed to the presence of collagen only. In addition, it has been shown in an in vivo mouse model that the T2* value decreases during the initial weeks, when the infarction progresses into the chronic phase.51 An example of T2* mapping in a porcine model of chronic myocardial infarction is shown in Figure 3 (see Materials and MRI methods section). Extravasation of erythrocytes, during reperfusion after acute coronary occlusion, may lead to chronic iron deposition and chronic inflammation. Recently, Kali et al. have shown a relationship between postreperfusion intramyocardial hemorrhage iron deposits and changes in signal intensity on T2* weighted images.52 Further research is necessary to investigate to what extent

shortening of T2* can be attributed to collagen, and what the influence is of hemorrhage on T2* shortening in myocardial fibrosis detection after infarction. A limitation for the use of T2* mapping MRI for detection of myocardial fibrosis is the influence of macroscopic field inhomogeneities. These field differences, mainly induced at the lung-heart transition, influence the T2* map and indicate false positive areas.53 Therefore, good shimming of the heart region is very important, or sophisticated image analysis software is needed that aims to correct for the effects of macroscopic inhomogeneity with the use of the phase image.54

Figure 3. Detection of chronic myocardial infarction (4 weeks post infarction) with T2* mapping in a porcine animal model, see Materials and MRI methods section. A: post mortem TTC staining of the infarction area. B: late gadolinium enhancement (LGE) image. C: native R2* (1/T2*) map. The arrow indicates the infarcted area. Artifacts arising from macroscopic field inhomogeneities can be observed by the lungtissue interface.

T1 ρ mapping T1ρ is a less known and less used MRI relaxation mechanism, which is commonly applied to study articular cartilage.55,56 The T1ρ relaxation time describes relaxation while the magnetization is in the transverse plane, in the presence of a so-called spin-lock pulse. A spin-lock pulse is a low amplitude radiofrequency (RF) pulse on-resonance with the precessing transverse

PART TWO CHAPTER 5

magnetization, and the relaxation rate constant of the transverse magnetization under influence of this spin-lock pulse is the T1ρ relaxation. In conventional T1 relaxation, energy is exchanged from spins to the surrounding lattice in processes occurring near the Larmor frequency. However in the presence of a spin-lock pulse, the interactions between water and other molecules (exchange and rotation correlation times) have to be close to the frequency of the spin-lock pulse in order to contribute to the relaxation. As the rotation times of macromolecules are in the order of the frequency with which the magnetizations precesses around the spin-lock pulse, T1ρ relaxation is sensitive to watermacromolecule interactions. This culminated in the hypothesis that T1ρ is sensitive to the increase in collagen macromolecules content in case of myocardial fibrosis. An increase of T1ρ relaxation time was found in a swine model of chronic myocardial infarction. The area with a significantly higher T1ρ relaxation time constant (compared to healthy myocardium) correlated with the infarcted area confirmed by histology.57,58 It was also shown that the area of the myocardium with a significant higher T1ρ relaxation time constant correlates with the infarct area indicated by the gold standard LGE, both in a swine model of chronic myocardial infarction, and a mouse model of chronic myocardial infarction.58,59 Figure 4 shows an example of the difference in T1ρ between an infarcted area and the remote myocardium in a porcine infarct model (see Materials and MRI methods section). It has been shown that with increasing spin-lock pulse amplitude, the contrast between infarct, border zone and remote tissue is increased.57,60 However, for clinical application of T1ρ MRI, the specific absorption rate (SAR) delivered to the subjects limits the maximum achievable spin-lock amplitudes. As T1ρ mapping is a relative new technique in the field of cardiac MR, future research should aim at further elucidating the potential of this technique regarding the assessment of cardiac fibrosis. Additionally, it is important that the underlying mechanism of the effect of myocardial fibrosis on T1ρ is studied further. Where in myocardial infarction tissue consistently a higher T1ρ time is found compared to healthy myocardium,61–63 studies in cartilage and protein solutions show an opposite effect with a decrease of T1ρ time corresponding with a higher macromolecule content.64

Figure 4. Detection of chronic myocardial infarction (4 weeks post infarction) with T1ρ mapping in a porcine animal model, see Materials and MRI methods section. A: post mortem TTC staining of the infarction area. B: late gadolinium enhancement (LGE) image. C: native T1ρ map. The arrow indicates the infarcted area.

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

Magnetization transfer imaging Magnetization transfer imaging (MTI) is a MRI technique which is based on the transfer of magnetization from the hydrogen nuclei with restricted motion to hydrogen nuclei of free water. The hydrogen nuclei with restricted motion are mainly associated to be bound to macromolecules through hydrogen bonds. The amount of magnetization transfer can be assessed, amongst other methods, by measuring the difference in signal of free water obtained with and without the application of an off-resonance RF pulse prior to imaging. Hydrogen in bound water has a broad frequency distribution, and will be partly saturated by the application of an off-resonance RF pulse. Due to the continuous exchange between bound and free water, the saturated hydrogen will exchange with the free water, thereby reducing the signal intensity of the image obtained with the application of an off-resonance pulse. Thus, MTI provides an alternative contrast compared to the commonly used mechanisms based on T1 and T2 relaxation times, but is assumed to have similarity with T1Ď relaxation.65 It has been shown ex vivo in a rat model that the magnetization transfer rate is decreased in acute and chronically infarcted myocardium compared to healthy tissue.66 It was suggested that

this decrease in magnetization transfer is induced by a change in macromolecular interactions with the surrounding water as a result of myocardial infarction. The use of magnetization transfer imaging has been tested in vivo in patients with acute myocardial infarction.67 By acquiring two balanced steady-state free precession images with different levels of RF power deposition, a magnetization transfer ratio (MTR) image could be calculated (Figure 5). A significantly reduced magnetization transfer ratio was found in the infarct region in patients with subacute (2-5 days) myocardial infarction, corresponding well with the LGE images.

Figure 5. Short-axis view of a 59-year old male patient with an anterior-septal infarct. Left is the late enhancement (LGE) image. Right is an magnetization transfer (MTR) map calculated from two SSFP images with different levels of RF power deposition. Reproduced from Ref 67, with permission.

PART TWO CHAPTER 5

Magnetization transfer imaging for detection of myocardial infarction is a promising technique. However, it needs to be investigated if this technique can also provide information on chronic myocardial infarction and the formation of fibrosis. Since T1ρ relaxation and magnetization transfer imaging are both associated with macromolecular sensitivity, it is important to investigate the underlying principle on how the concentration of myocardial fibrosis influences T1ρ and MTI.68 Materials and MRI methods All in vivo experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources. Experiments were approved by the Animal Experimentation Committee of the University Medical Center Utrecht. Two Dalland landrace pigs (69 ± 5 kg) underwent a 90-minute percutaneous balloon occlusion of the left anterior descending (LAD) coronary artery distal to the second diagonal branch, followed by reperfusion. Four weeks after myocardial infarction, an in vivo MRI was performed under anesthesia on a clinical 3T scanner (Achieva TX, Philips Healthcare, Best, the Netherlands) with a 32-channel receive coil. The pigs were sacrificed for triphenyl tetrazolium chloride (TTC) staining of serial sectioned heart slices and histology. For the T1ρ map a single slice T1ρ-prepared, multishot gradient echo sequence with Cartesian readout was used (TE/TR = 2.1 / 4.2, slice thickness = 10 mm, resolution = 1.5 x 1.5 mm2, FOV = 350 x 350 mm2, flip angle = 15 degrees, echo train length = 15. Five different pulse lengths were set for the T1ρ preparation pulse (3, 10, 20, 30, 40 ms), with a spin lock amplitude of 500 Hz (total scan time = 7 minutes). The T2* map was acquired with a multi-echo gradient echo sequence with Cartesian readout (10 echoes, first TE = 1.46 ms, ΔTE = 0.9 ms, TR = 1875 ms, 6 slices, slice thickness = 6 mm, resolution = 2x2 mm2, FOV = 300 x 300 mm2, flip angle = 90 degrees, total scan time = 5 minutes). The T1ρ and T2* maps were calculated by pixelwise fitting of an exponential decay function. LGE MRI was performed using a 3D gradient echo sequence, fifteen minutes after injection of a gadolinium based contrast agent (TE/TR = 1.38/4.3, 20 slices, slice thickness = 6 mm, resolution = 1.75 x 1.75 mm2, FOV = 320 x 320 mm2, flip angle = 25 degrees, inversion delay = 330 ms, total scan time = 3 minutes). Future implications The need for detecting diffuse fibrosis in a non-invasive quantitative way has driven the search for new methods to detect fibrosis with or without using contrast agents, e.g., T1 mapping, T2 mapping, T2* mapping, UTE imaging, T1ρ mapping, and MTI. These MRI techniques show promising initial results, but further research is needed to better understand how the contrast is formed. ECV mapping shows promising initial results in clinical studies, but has the downside that it requires contrast administration and blood sampling. Currently, in the field of non-invasive endogenous MRI detection of myocardial fibrosis, native T1, T1ρ mapping and MTI look most promising. Further development for clinical application of these techniques is required, for instance development of a clinically applicable protocol, further validation in clinical trials and correlation with histology. Furthermore, standardized measurement protocols to assess these contrasts will also be necessary. Once established, these techniques might be applicable to a

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

broad clinical spectrum, including ablation therapy, valvular diseases, cardiomyopathies, ischemic heart disease and stem cell therapy. Acknowledgments We kindly acknowledge Grace Croft, Marlijn Jansen, Merel Schurink, Evelyn Velema, Cees Verlaan and Joyce Visser for their assistance with the animal experiments.

PART TWO CHAPTER 5

REFERENCES 1.

2. 3. 4.

10.

11.

12.

13.

14. 15.

16.

17. 18. 19.

Marijianowski MM, Teeling P, Mann J, Becker AE. Dilated cardiomyopathy is associated with an increase in the type I/type III collagen ratio: a quantitative assessment. J Am Coll Cardiol. 1995;25:1263–72. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA. Assessment of myocardial fibrosis with cardiovascular magnetic resonance. J Am Coll Cardiol. 2011;57:891–903. Maron BJ. Hypertrophic Cardiomyopathy. JAMA. 2002;287:1308–20. Robbers LF, Baars EN, Brouwer WP, Beek AM, Hofman MB, Niessen HW, et al. T1 mapping shows increased extracellular matrix size in the myocardium due to amyloid depositions. Circ Cardiovasc Imaging. 2012;5:423–6. Díez J, González A, López B, Querejeta R. Mechanisms of disease: pathologic structural remodeling is more than adaptive hypertrophy in hypertensive heart disease. Nat Clin Pract Cardiovasc Med. 2005;2:209–16. Koudstaal S, Jansen Of Lorkeers SJ, Gaetani R, Gho JM, van Slochteren FJ, Sluijter JP, et al. Concise review: heart regeneration and the role of cardiac stem cells. Stem Cells Transl Med. 2013;2:434-43. Gulati A, Jabbour A, Ismail TF, Guha K, Khwaja J, Raza S, et al. Association of fibrosis with mortality and sudden cardiac death in patients with nonischemic dilated cardiomyopathy. JAMA. 2013;309:896–908. Van Slochteren FJ, Teske AJ, van der Spoel TI, Koudstaal S, Doevendans PA, Sluijter JP, et al. Advanced measurement techniques of regional myocardial function to assess the effects of cardiac regenerative therapy in different models of ischaemic cardiomyopathy. Eur Heart J Cardiovasc Imaging. 2012;13:808–18. Bondarenko O, Beek A, Hofman M, Kühl H, Twisk J, van Dockum W, et al. Standardizing the Definition of Hyperenhancement in the Quantitative Assessment of Infarct Size and Myocardial Viability Using Delayed Contrast-Enhanced CMR. J Cardiovasc Magn Reson. 2005;7:481–5. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, et al. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation. 1999;100:1992–2002. Kwong RY, Chan AK, Brown KA, Chan CW, Reynolds HG, Tsang S, et al. Impact of unrecognized myocardial scar detected by cardiac magnetic resonance imaging on event-free survival in patients presenting with signs or symptoms of coronary artery disease. Circulation. 2006;113:2733–43. Assomull RG, Prasad SK, Lyne J, Smith G, Burman ED, Khan M, et al. Cardiovascular magnetic resonance, fibrosis, and prognosis in dilated cardiomyopathy. J Am Coll Cardiol. 2006;48:1977– 85. Müller KA, Müller I, Kramer U, Kandolf R, Gawaz M, Bauer A, et al. Prognostic Value of Contrastenhanced Cardiac Magnetic Resonance Imaging in Patients with Newly Diagnosed NonIschemic Cardiomyopathy: Cohort Study. PLoS One. 2013;8:e57077. Karamitsos TD, Francis JM, Myerson S, Selvanayagam JB, Neubauer S. The role of cardiovascular magnetic resonance imaging in heart failure. J Am Coll Cardiol. 2009;54:1407–24. van Slochteren FJ, van Es R, Koudstaal S, van der Spoel TI, Sluijter JP, Verbree J, et al. Multimodality infarct identification for optimal image-guided intramyocardial cell injections. Neth Heart J. 2014;22:493-500. Leyva F, Foley PW, Chalil S, Ratib K, Smith RE, Prinzen F, et al. Cardiac resynchronization therapy guided by late gadolinium-enhancement cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2011;13:29. Aggarwal NR, Martinez MW, Gersh BJ, Chareonthaitawee P. Role of cardiac MRI and nuclear imaging in cardiac resynchronization therapy. Nat Rev Cardiol. 2009;6:759–70. Vernooy K, van Deursen CJ, Strik M, Prinzen FW. Strategies to improve cardiac resynchronization therapy. Nat Rev Cardiol. 2014;11:481–93. Bellin MF, Van Der Molen AJ. Extracellular gadolinium-based contrast media: an overview. Eur J Radiol. 2008;66:160–7.

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

20. Murphy KP, Szopinski KT, Cohan RH, Mermillod B, Ellis JH. Occurrence of adverse reactions to gadolinium-based contrast material and management of patients at increased risk: a survey of the American Society of Neuroradiology Fellowship Directors. Acad Radiol. 1999;6:656–64. 21. Raisch DW, Garg V, Arabyat R, Shen X, Edwards BJ, Miller FH, et al. Anaphylaxis associated with gadolinium-based contrast agents: data from the Food and Drug Administration’s adverse event reporting system and review of case reports in the literature. Expert Opin Drug Saf. 2014;13:15–23. 22. Perazella MA. Gadolinium-contrast toxicity in patients with kidney disease: nephrotoxicity and nephrogenic systemic fibrosis. Curr Drug Saf. 2008;3:67–75. 23. Reiter T, Ritter O, Prince MR, Nordbeck P, Wanner C, Nagel E, et al. Minimizing risk of nephrogenic systemic fibrosis in cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2012;14:31. 24. Beek AM, Bondarenko O, Afsharzada F, van Rossum AC. Quantification of late gadolinium enhanced CMR in viability assessment in chronic ischemic heart disease: a comparison to functional outcome. J Cardiovasc Magn Reson. 2009;11:6. 25. Messroghli DR, Greiser A, Fröhlich M, Dietz R, Schulz-Menger J. Optimization and validation of a fully-integrated pulse sequence for modified look-locker inversion-recovery (MOLLI) T1 mapping of the heart. J Magn Reson Imaging. 2007;26:1081–6. 26. Sibley CT, Noureldin RA, Gai N, Nacif MS, Liu S, Turkbey EB, et al. T1 Mapping in cardiomyopathy at cardiac MR: comparison with endomyocardial biopsy. Radiology. 2012;265:724–32. 27. Messroghli DR, Nordmeyer S, Dietrich T, Dirsch O, Kaschina E, Savvatis K, et al. Assessment of diffuse myocardial fibrosis in rats using small-animal Look-Locker inversion recovery T1 mapping. Circ Cardiovasc Imaging. 2011;4:636–40. 28. Flett AS, Hayward MP, Ashworth MT, Hansen MS, Taylor AM, Elliott PM, et al. Equilibrium contrast cardiovascular magnetic resonance for the measurement of diffuse myocardial fibrosis: preliminary validation in humans. Circulation. 2010;122:138–44. 29. Kellman P, Wilson JR, Xue H, Ugander M, Arai AE. Extracellular volume fraction mapping in the myocardium, part 1: evaluation of an automated method. J Cardiovasc Magn Reson. 2012;14:63. 30. Flett AS, Sado DM, Quarta G, Mirabel M, Pellerin D, Herrey AS, et al. Diffuse myocardial fibrosis in severe aortic stenosis: an equilibrium contrast cardiovascular magnetic resonance study. Eur Heart J Cardiovasc Imaging. 2012;13:819–26. 31. Wong TC, Piehler K, Meier CG, Testa SM, Klock AM, Aneizi AA, et al. Association between extracellular matrix expansion quantified by cardiovascular magnetic resonance and short-term mortality. Circulation. 2012;126:1206–16. 32. Moon JC, Messroghli DR, Kellman P, Piechnik SK, Robson MD, Ugander M, et al. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR Working Group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson. 2013;15:92. 33. White SK, Sado DM, Fontana M, Banypersad SM, Maestrini V, Flett AS, et al. T1 mapping for myocardial extracellular volume measurement by CMR: bolus only versus primed infusion technique. JACC Cardiovasc Imaging. 2013;6:955–62. 34. Kellman P, Wilson JR, Xue H, Bandettini WP, Shanbhag SM, Druey KM, et al. Extracellular volume fraction mapping in the myocardium, part 2: initial clinical experience. J Cardiovasc Magn Reson. 2012;14:64. 35. Miller CA, Naish JH, Bishop P, Coutts G, Clark D, Zhao S, et al. Comprehensive validation of cardiovascular magnetic resonance techniques for the assessment of myocardial extracellular volume. Circ Cardiovasc Imaging. 2013;6:373–83. 36. Dass S, Suttie JJ, Piechnik SK, Ferreira VM, Holloway CJ, Banerjee R, et al. Myocardial tissue characterization using magnetic resonance noncontrast t1 mapping in hypertrophic and dilated cardiomyopathy. Circ Cardiovasc Imaging. 2012;5:726–33. 37. Bull S, White SK, Piechnik SK, Flett AS, Ferreira VM, Loudon M, et al. Human non-contrast T1 values and correlation with histology in diffuse fibrosis. Heart. 2013;99:932-7. 38. Puntmann VO, Voigt T, Chen Z, Mayr M, Karim R, Rhode K, et al. Native t1 mapping in differentiation of normal myocardium from diffuse disease in hypertrophic and dilated cardiomyopathy. JACC Cardiovasc Imaging. 2013;6:475–84.

PART TWO CHAPTER 5

39. Kali A, Cokic I, Tang RL, Yang HJ, Sharif B, Marbán E, et al. Determination of location, size, and transmurality of chronic myocardial infarction without exogenous contrast media by using cardiac magnetic resonance imaging at 3 T. Circ Cardiovasc Imaging. 2014;7:471–81. 40. Salerno M, Kramer CM. Advances in parametric mapping with CMR imaging. JACC Cardiovasc Imaging. 2013;6:806–22. 41. Giri S, Chung YC, Merchant A, Mihai G, Rajagopalan S, Raman S V, et al. T2 quantification for improved detection of myocardial edema. J Cardiovasc Magn Reson. 2009;11:56. 42. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, et al. Delayed enhancement and T2-weighted cardiovascular magnetic resonance imaging differentiate acute from chronic myocardial infarction. Circulation. 2004;109:2411–6. 43. Bun SS, Kober F, Jacquier A, Espinosa L, Kalifa J, Bonzi MF, et al. Value of in vivo T2 measurement for myocardial fibrosis assessment in diabetic mice at 11.75 T. Invest Radiol. 2012;47:319–23. 44. Reeder SB, Faranesh AZ, Boxerman JL, McVeigh ER. In vivo measurement of T*2 and field inhomogeneity maps in the human heart at 1.5 T. Magn Reson Med. 1998;39:988–98. 45. Köhler S, Hiller KH, Waller C, Jakob PM, Bauer WR, Haase A. Visualization of myocardial microstructure using high-resolution T*2 imaging at high magnetic field. Magn Reson Med. 2003;49:371–5. 46. Carpenter JP, Roughton M, Pennell DJ. International survey of T2* cardiovascular magnetic resonance in β-thalassemia major. Haematologica. 2013;98:1368–74. 47. Baksi AJ, Pennell DJ. T2* imaging of the heart: methods, applications, and outcomes. Top Magn Reson Imaging. 2014;23:13–20. 48. Reiter DA, Lin PC, Fishbein KW, Spencer RG. Multicomponent T2 relaxation analysis in cartilage. Magn Reson Med. 2009;61:803–9. 49. Holmes JE, Bydder GM. MR imaging with ultrashort TE (UTE) pulse sequences: Basic principles. Radiography. 2005;11:163–74. 50. de Jong S, Zwanenburg JJ, Visser F, van der Nagel R, van Rijen HV, Vos MA, et al. Direct detection of myocardial fibrosis by MRI. J Mol Cell Cardiol. 2011;51:974–9. 51. Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, van Echteld CJ, et al. Quantitative T 2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice. MAGMA. 2012;25:369–79. 52. Kali A, Kumar A, Cokic I, Tang RL, Tsaftaris SA, Friedrich MG, et al. Chronic manifestation of postreperfusion intramyocardial hemorrhage as regional iron deposition: a cardiovascular magnetic resonance study with ex vivo validation. Circ Cardiovasc Imaging. 2013;6:218–28. 53. Atalay MK, Poncelet BP, Kantor HL, Brady TJ, Weisskoff RM. Cardiac susceptibility artifacts arising from the heart-lung interface. Magn Reson Med. 2001;45:341–5. 54. de Leeuw H, Bakker CJ. Correction of gradient echo images for first and second order macroscopic signal dephasing using phase derivative mapping. Neuroimage. 2012;60:818–29. 55. Menezes NM, Gray ML, Hartke JR, Burstein D. T2 and T1rho MRI in articular cartilage systems. Magn Reson Med. 2004;51:503–9. 56. Li X, Cheng J, Lin K, Saadat E, Bolbos RI, Jobke B, et al. Quantitative MRI using T1ρ and T2 in human osteoarthritic cartilage specimens: Correlation with biochemical measurements and histology. Magn Reson Imaging. 2011;29:324–34. 57. Witschey WR, Pilla JJ, Ferrari G, Koomalsingh K, Haris M, Hinmon R, et al. Rotating frame spin lattice relaxation in a swine model of chronic, left ventricular myocardial infarction. Magn Reson Med. 2010;64:1453–60. 58. Musthafa HS, Dragneva G, Lottonen L, Merentie M, Petrov L, Heikura T, et al. Longitudinal rotating frame relaxation time measurements in infarcted mouse myocardium in vivo. Magn Reson Med. 2013;69:1389–95. 59. Witschey WR, Zsido GA, Koomalsingh K, Kondo N, Minakawa M, Shuto T, et al. In vivo chronic myocardial infarction characterization by spin locked cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2012;14:37. 60. Oorschot J, Kaplan M, Baldus M, Luijten P, Zwanenburg J. Analysis of T1 , T2 and T1rho spin lock field dependency in myocardial infarction tissue using HRMAS spectroscopy at 11.7T. Proc Int Soc Magn Reson Med. 2012;3069.

ENDOGENOUS CONTRAST MRI OF CARDIAC FIBROSIS

61. Han Y, Liimatainen T, Gorman RC, Witschey WR. Assessing Myocardial Disease Using T1ρ MRI. Curr Cardiovasc Imaging Rep. 2014;7:9248. 62. Muthupillai R, Flamm SD, Wilson JM, Pettigrew RI. Radiology Acute Myocardial Infarction : Tissue Characterization with T1rho -weighted MR Imaging — Initial Experience. Radiology. 2004;606–10. 63. Musthafa HS, Dragneva G, Lottonen L, Merentie M, Petrov L, Heikura T, et al. Longitudinal rotating frame relaxation time measurements in infarcted mouse myocardium in vivo. Magn Reson Med. 2013;69:1389–95. 64. Andrasko J. Water in agarose gels studied by nuclear magnetic resonance relaxation in the rotating frame. Biophys J. 1975;15:1235–43. 65. Ulmer JL, Mathews VP, Hamilton CA, Elster AD, Moran PR. Magnetization transfer or spinlock? An investigation of off-resonance saturation pulse imaging with varying frequency offsets. AJNR Am J Neuroradiol. 1996;17:805–19. 66. Scholz TD, Hoyt RF, DeLeonardis JR, Ceckler TL, Balaban RS. Water-macromolecular proton magnetization transfer in infarcted myocardium: a method to enhance magnetic resonance image contrast. Magn Reson Med. 1995;33:178–84. 67. Weber OM, Speier P, Scheffler K, Bieri O. Assessment of magnetization transfer effects in myocardial tissue using balanced steady-state free precession (bSSFP) cine MRI. Magn Reson Med. 2009;62:699–705. 68. Scholz TD, Ceckler TL, Balaban RS. Magnetization transfer characterization of hypertensive cardiomyopathy: significance of tissue water content. Magn Reson Med. 1993;29:352–7.

PART TWO

Finding Fibrosis Patterns

Chapter

High Resolution Systematic Digital Histological Quantification of Cardiac Fibrosis and Adipose Tissue in Phospholamban p.Arg14del Mutation Associated Cardiomyopathy Published as PloS ONE 2014;9:e94820

Johannes M.I.H. Gho1, RenĂŠ van Es1, Nikolas Stathonikos2, Magdalena Harakalova1,2, Wouter P. te Rijdt3, Albert J.H. Suurmeijer4, Jeroen F. van der Heijden1, Nicolaas de Jonge1, Steven A.J. Chamuleau1, Roel A. de Weger2, Folkert W. Asselbergs1,5, Aryan Vink2

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Pathology, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Cardiology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands

Department of Pathology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands

Interuniversity Cardiology Institute of the Netherlands (ICIN), Utrecht, the Netherlands

PART TWO CHAPTER 6

ABSTRACT Myocardial fibrosis can lead to heart failure and act as a substrate for cardiac arrhythmias. In dilated cardiomyopathy diffuse interstitial reactive fibrosis can be observed, whereas arrhythmogenic cardiomyopathy is characterized by fibrofatty replacement in predominantly the right ventricle. The p.Arg14del mutation in the phospholamban (PLN) gene has been associated with dilated cardiomyopathy and recently also with arrhythmogenic cardiomyopathy. Aim of the present study is to determine the exact pattern of fibrosis and fatty replacement in PLN p.Arg14del mutation positive patients, with a novel method for high resolution systematic digital histological quantification of fibrosis and fatty tissue in cardiac tissue. Transversal mid-ventricular slices (n = 8) from whole hearts were collected from patients with the PLN p.Arg14del mutation (age 48Âą16 years; 4 (50%) male). An in-house developed open source MATLAB script was used for digital analysis of Massonâ&#x20AC;&#x2122;s trichrome stained slides (http://sourceforge.net/projects/ fibroquant/). Slides were divided into trabecular, inner and outer compact myocardium. Per region the percentage of connective tissue, cardiomyocytes and fatty tissue was quantified. In PLN p.Arg14del mutation associated cardiomyopathy, myocardial fibrosis is predominantly present in the left posterolateral wall and to a lesser extent in the right ventricular wall, whereas fatty changes are more pronounced in the right ventricular wall. No difference in distribution pattern of fibrosis and adipocytes was observed between patients with a clinical predominantly dilated and arrhythmogenic cardiomyopathy phenotype. In the future, this novel method for quantifying fibrosis and fatty tissue can be used to assess cardiac fibrosis and fatty tissue in animal models and a broad range of human cardiomyopathies.

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

INTRODUCTION A network of extracellular matrix maintains the structural integrity of the myocardium. Due to several etiologies increased deposition of collagen and other extracellular matrix proteins can occur leading to cardiac fibrosis.1 After myocardial infarction, cardiomyocytes are replaced by connective tissue leading to reparative fibrosis. In contrast, in non-ischemic cardiomyopathies, an increase in collagen synthesis by myofibroblasts results in diffuse interstitial reactive fibrosis. In arrhythmogenic cardiomyopathy (AC), fibrosis is accompanied by an increase of adipocytes leading to so-called fibrofatty replacement.2 Myocardial fibrosis is an important part of the histological characteristics in heart failure (HF) with preserved and reduced ejection fraction and may act as a substrate for cardiac arrhythmias. Adequate detection of the amount and distribution of fibrosis in the heart is important for diagnosis, predicting prognosis, treatment planning and follow-up after therapy.3, 4 The reference noninvasive standard for indirect detection of myocardial fibrosis is late gadolinium enhancement on cardiac magnetic resonance imaging (MRI).3 Thus far, detailed histological correlation studies to validate this MRI technique are scarce. Histological assessment of cardiac fibrosis is mostly limited by the small amount of tissue available in diagnostic endomyocardial biopsies that only provides regional information.2 In addition, quantification of histological fibrosis is usually

performed semi-quantitatively, classifying the fibrosis in limited categories. Phospholamban is a protein in the sarcoplasmic reticulum and acts as a (reversible) inhibitor of the Ca2+ pump: sarcoplasmic reticulum Ca2+-ATPase 2a (SERCA2a). On phosphorylation it dissociates from SERCA2a and thereby activates the Ca2+ pump. This cascade regulates cardiac relaxation and contractility. Several causal phospholamban (PLN) mutations have been described in humans.5-8 The p.Arg14del (c.40_42delAGA) founder mutation in the PLN gene has been associated with dilated cardiomyopathy (DCM) and recently also with AC.9 Detailed histologic analysis of the pattern of fibrosis and fatty changes in PLN mutation associated cardiomyopathies has not been extensively studied and to the best of our knowledge has not been performed on transverse heart slices. The aim of this study was to determine the exact patterns of fibrosis and fatty changes in the myocardium of patients with the PLN p.Arg14del mutation associated cardiomyopathy in relation to their clinical phenotype. This study population was used as proof-of-principle for a novel method of high resolution systematic digital quantification of fibrosis and fatty tissue in transversal cardiac slides. In the future this method may be used for detailed histological quantification and determination of the distribution pattern of cardiac fibrosis in different types of heart disease, in addition it provides a detailed high resolution reference for imaging techniques of cardiac fibrosis.

PART TWO CHAPTER 6

METHODS Ethics statement The study met the criteria of the code of proper use of human tissue that is used in the Netherlands. The study was approved by the scientific advisory board of the biobank of the University Medical Center Utrecht, Utrecht, the Netherlands (protocol no. 12/387). Written informed consent was obtained or in certain cases waived by the ethics committee when obtaining informed consent was not possible due to death of the patient. Hearts obtained at autopsy (n = 2) or explantation (n = 6) were collected from patients with the PLN p.Arg14del mutation. Based on their initial clinical presentation, patients were divided in two categories: predominantly DCM or AC. Three control hearts, two donor hearts not-used for transplantation and one heart obtained at autopsy of a road accident victim, were used as reference. We used a systematic methodology for high resolution digital cardiac fibrosis quantification (Figure 1). Hearts were cut in transverse (short-axis) slices of 1 cm thick starting at the apex including both ventricles. Each fourth transverse slice was fixed in formalin and divided into smaller pieces. A map of the heart slice was drawn to annotate the origin of each tissue specimen. Subsequently the samples were embedded in paraffin and Masson’s trichrome staining was performed. The slides were scanned at 20X magnification as described previously.10 Images were extracted using Aperio ImageScope v12.0.0.5039 (Aperio, Vista, CA, USA) as a TIFF file with lossless compression. The images were resized to 10% of their original size for digital analysis. To analyze the cardiac tissue, slides were divided into several layers: the epicardial area, compact myocardium and non-compact (trabeculated) myocardium (http://sourceforge.net/projects/ fibroquant/). The epicardial area was defined as the outer region of fatty tissue bordered by the first row of cardiomyocytes. The non-compact myocardial area was defined as the endocardial trabeculated region. The compact myocardium was defined by the area between the trabeculated area and the epicardial area and was artificially divided with an equidistant line in two halves. An equidistant line is one for which every point on the line is equidistant from the nearest points on both the epicardial and trabecular segmentation. In case a midline division was not feasible, e.g., due to a thin wall or originating from the interventricular septum (without epicardium), mean values of the total myocardium were used. Thus, the myocardium was divided in four layers: trabecular myocardium, inner or outer compact myocardium and the epicardium. In all four layers the percentage of connective tissue (blue), cardiomyocytes (red) and adipose tissue (cells with non stained cytoplasm) was digitally quantified using MATLAB (Release R2012a, The MathWorks, Inc., Natick, Massachusetts, United States). The percentage per region was calculated by separating the Masson’s trichrome stained slides into its constituent stains of methyl blue and ponceau-fuchsin by performing colour deconvolution.11 This produces two grayscale images depicting the concentration of the two stainings. The resulting images are filtered using a 2D median filter and further processed using morphological dilation, thresholding and closing operations. The adipose tissue is quantified using a separate function in order to highlight the “chain-link” structure of adipocytes as seen on glass slides. The original image is

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

Figure 1. Overview of methodology. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior. A. gross showing a transverse heart slice of arrhythmogenic cardiomyopathy. B. transverse slice dissection scheme. C-F. examples of digital slide processing. Regions of interest are shown in orange lines (defining the epicardium, compact myocardium divided by an equidistant midline and trabeculated part). C. slide from the left ventricle posterior wall. D. slide C after digital processing. Red: cardiomyocytes. Blue: connective tissue. Pseudo green: adipose tissue. E. slide from the right ventricle lateral wall. F. slide E after digital processing. Red: cardiomyocytes. Blue: connective tissue. Pseudo green: adipose tissue.

converted to grayscale and filtered using a 2D median filter. After that a threshold is applied and morphological opening is performed. The median filtered image is then subtracted from the morphologically opened image and the resulting image is further morphologically processed by performing consecutive closing and opening operations. This converts the â&#x20AC;&#x153;chain-linkâ&#x20AC;? structure into a black and white image where only the adipose tissue is left. The area of each constituent is determined and a percentage is calculated based on the total area that was processed.12 The epicardial region was excluded from analysis. The resulting values (percentage fibrosis and fatty tissue) were annotated to the corresponding region in the heart using an automated algorithm (Supplementary Figure S1).

PART TWO CHAPTER 6

Subsequently the annotated map of the transverse slice is automatically transformed to a standardized schematic overview. The annotated map was translated to a schematic overview by determining the angular properties of each separate section from the middle of the ventricles with a precision of one degree. The results of the quantification (percentage fibrosis or fatty tissue) are displayed using an easily interpretable color scale. Statistics were performed using IBM SPSS Statistics version 20. We compared mean percentages of fibrosis and adipose tissue in different areas between the two clinical phenotypes of AC and DCM and control hearts. The left ventricle was divided into a septal, posterior, posterolateral, lateral, anterolateral and anterior part. The right ventricle was divided into a posterior and anterior part. To compare mean percentages of fibrosis and adipose tissue corrected for surface area per region and condition a repeated measures analysis was performed. After Greenhouse-Geisser correction, interactions between region and condition were explored. Post hoc tests (Tukey HSD) were performed in the absence of a significant interaction.

RESULTS Clinical characteristics of the patients are summarized in Table 1. Mean age was 48Âą16 years; 4 (50%) patients were male. Five patients were known with a clinical phenotype of DCM and 3 patients were known with a clinical phenotype of AC. The schematic overviews depict distribution and percentage of fibrosis and adipose tissue for DCM (Figure 2) and arrhythmogenic patients (Figure 3). Mean values of fibrosis and adipose tissue with standard deviations per region and condition (control, AC or DCM) are presented in Supplementary Table S1. In the 8 heart slices of PLN mutation carriers, myocardial fibrosis was mainly observed in the trabecular part of the posterolateral wall of the right ventricle and in the posterolateral (mean >38%) and in lesser extent anterolateral (mean >26%) wall of the left ventricle. In the left ventricle, fibrosis

Table 1. Patient characteristics of the PLN p.Arg14del mutation carriers (n = 8) Age; mean (SD)

48 (16)

Male sex

4 (50%)

Device ICD

4 (50%)

CRT-D

3 (38%)

Left ventricular assist device

4 (50%)

First presenting symptom Arrhythmia

4 (50%)

Heart failure

4 (50%)

Heart transplantation

6 (75%)

SD, standard deviation; ICD, implantable cardioverter-defibrillator; CRT-D, cardiac resynchronization therapy defibrillator.

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

was more pronounced in the outer layer of compact myocardium than in the myocardial layers more closely to the lumen. Mean percentage of fibrosis and adipose tissue is shown in the combined schematic overviews (Figure 4).

6 Figure 2. Schematic heart slice overview of patients with a clinical phenotype of dilated cardiomyopathy. The results of the digital quantification in heart slices in percentage of fibrosis or adipose tissue are shown using a color scale. The epicardial fat has been excluded from this overview. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior.

Figure 3. Schematic heart slice overview of patients with a clinical phenotype of arrhythmogenic cardiomyopathy. The results of the digital quantification in heart slices in percentage of fibrosis or adipose tissue are shown using a color scale. The epicardial fat has been excluded from this overview. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior.

PART TWO CHAPTER 6

The septum and ventral wall of the right ventricle revealed the least amount of interstitial fibrosis. Fatty changes of myocardium were predominantly observed in the entire right ventricle wall (mean 37.2±14% and 28.9±4% in AC and 26.1±20% and 24.3±12% in DCM respectively in regions 7 and 8) and in the epicardial side of the compact myocardium of the posterolateral wall of the left ventricle (mean 6.9±5% in AC and 5.8±6% in DCM in region 1). Overall fatty infiltration of myocardium was more pronounced in the right (mean adipose tissue >24%) than in the left (mean adipose tissue <11%) ventricle. Mean fibrosis in control hearts was less than 6%, mean adipose tissue was 2% or less in the left ventricle myocardium and less than 15% in the right ventricular wall (Figure 5). Mean connective and adipose tissue per condition and region (n = 8, corrected for surface area) is shown in boxplots (Figure 6). Mean values of fibrosis and adipose tissue were log transformed to reduce right-skewness and heterogeneity of variance and a repeated measures analysis was performed. Comparing fibrosis we found a significant interaction between condition and region (p<0.001), therefore post-hoc testing was not performed. For adipose tissue there was no significant interaction between condition and region (p=0.470). We found no significant difference between AC and DCM in pattern of adipose tissue (p=0.382). Compared to controls, we found a higher percentage of adipose tissue in AC (p=0.028) and a trend for a higher percentage of adipose tissue in DCM (p=0.126).

Figure 4. Schematic heart slice overview of patients with the PLN p.Arg14del mutation (dilated and arrhythmogenic cardiomyopathy). The results of the mean percentage of fibrosis or adipose tissue are shown using a color scale. In total 102 heart slides of 8 heart slices were used. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior.

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

Figure 5. Schematic overview of fibrosis and adipose tissue in heart slices of control hearts. The results of the digital quantification in percentage of fibrosis or adipose tissue are shown using a color scale. The epicardial fat has been excluded from the analysis. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior.

DISCUSSION To the best of our knowledge this is the first study that provides the exact and detailed pattern of fibrosis and fatty changes in PLN p.Arg14del mutation associated cardiomyopathy hearts. For this study we developed a novel method for systematic high resolution digital quantification of different tissue types in the heart. This quantification has been applied to Massonâ&#x20AC;&#x2122;s trichrome stained slides of transverse cardiac slices in PLN p.Arg14del mutation associated cardiomyopathies. Interestingly we found an overlap in fibrosis and fatty changes between DCM and AC in PLN p.Arg14del mutation carriers. Myocardial fibrosis was mainly observed in the posterolateral wall of the left ventricle and in less extent in the posterolateral wall of the right ventricle. Fatty tissue was more pronounced in the myocardium bordering the epicardium of the right ventricle. We found a significant higher percentage of adipose tissue in AC compared to control hearts (p=0.028). Phospholamban is a regulator of the SERCA2a pump, important for maintaining Ca2+ homeostasis and crucial for cardiac contractility.6,13 Phosphorylation of PLN increases SERCA2a activity, leading to increased cardiac relaxation and contractility for the next beat. The precise pathophysiological mechanism in PLN p.Arg14del mutation carriers leading to cardiac fibrosis and heart failure remains unknown. Transgenic mice overexpressing the mutant PLN-R14del showed extensive myocardial fibrosis, myocyte disarray, ventricular dilation and premature death, recapitulating human cardiomyopathy.8 Co-expression of the normal and mutant protein in HEK-293 cells

PART TWO CHAPTER 6

resulted in SERCA2a super inhibition. From these results it was inferred that the PLN p.Arg14del mutation causes inhibition of the SERCA2a pump and thereby leads to disturbed calcium metabolism and subsequently cardiac dysfunction. These data show that this process of reactive fibrosis develops according to a specific pattern, irrespective of the phenotype of the patients. Our small sample size should thereby be taken into account and because of a significant interaction post-hoc testing was not feasible for comparing the mean fibrosis values. This varying phenotype might by influenced by effect modifiers, e.g., epigenetics or intense endurance exercise.14 In addition, it has been postulated that the pattern of reactive fibrosis is determined by myocardial stress, microvascular dysfunction and sustained activation of neurohormonal and cytokine systems.15 Future research is needed to elucidate underlying pathophysiology in PLN mutation carriers related to the phenotype. Previous (not PLN mutation specific) histopathological studies have shown similar areas of predilection in AC.16 At first, affected areas in the right ventricle (RV) were described as the classic triangle of dysplasia, including the RV inflow tract, the apex and the RV outflow tract.17

Figure 6. Boxplots of mean fibrosis and mean adipose tissue in the different regions. Boxplots of mean percentage of fibrosis (A) and adipose tissue (B) per condition in 8 regions (C) corrected for surface area. Outliers are represented by small circles or stars. AC, arrhythmogenic cardiomyopathy; DCM, dilated cardiomyopathy; LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior; 1 = LV posterolateral wall; 2 = LV posterior wall; 3 = interventricular septum; 4 = LV anterior wall; 5 = LV anterolateral wall; 6 = LV lateral wall; 7 = RV dorsal wall; 8 = RV ventral wall.

100

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

Further research revealed that AC is not isolated to the RV. In a clinicopathologic study by Corrado et al. left ventricular involvement was found in 76% of cases with AC affecting both the septum and LV free wall, with a predilection for the posteroseptal and posterolateral areas.18 A recent MRI study in desmosomal mutation positive patients with AC showed involvement of the basal inferior and anterior right ventricle and the posterolateral left ventricle in AC, supporting the presence of a new biventricular triangle in early AC.19 Our results support the evidence from experimental animal models that the disease process in AC starts on the epicardial side and extends as a wave-front from the epicardium towards the endocardium.20 Clinical diagnosis of AC is made using International Task Force (revised) criteria, including structural (MRI and echocardiogram), histological (e.g. endomyocardial biopsy), electrocardiographic, arrhythmic and genetic features.2 The sensitivity of endomyocardial biopsies from the right ventricular septum in AC is low, according to our findings this might also be the case in PLN mutation associated cardiomyopathies.21 The method of fibrosis quantification presented here can be used for several applications. First, determination of the fibrosis pattern in the heart could provide an important link for genotypephenotype relationships in genetic cardiomyopathies. Previous studies have proposed morphometric evaluation of either fibrosis or adipocytes in different tissues.22-24 In our analysis fibrosis and fibrofatty replacement can be assessed simultaneously in layer specific detail. We

defined different regions of interest to divide the heart in an epicardial, compact and trabeculated layer. In addition, the pattern of fibrosis can be studied in the different regions of both ventricles. By studying the exact fibrosis pattern throughout the heart, patterns of disease might be discovered thereby elucidating mechanisms of pathophysiology. Numerous disease-causing genes for different cardiomyopathies have been identified during the past two decades and the challenge for the future is to link these genetic mutations to specific patterns of disease in the heart.20, 25 In future, AC with different underlying causal mutations could be compared to the PLN p.Arg14del mutation carriers with AC. The second application could be validation of myocardial imaging techniques. The reference noninvasive standard for fibrosis detection is late gadolinium enhancement on MRI.3 Adequate correlation to the gold standard of histology is important. Thus far, correlation studies are mostly done with small endomyocardial biopsies that only represent a fraction of the total myocardium or with triphenyl tetrazolium chloride (TTC) stained heart slides.26 Recently, several novel techniques for fibrosis detection have been proposed, including T1-mapping, that also require adequate correlation to histology.3, 27 Cardiac MRI images obtained before autopsy or heart transplantation, indicating fibrosis could be divided in similar segments to produce a bullseye for comparison with histopathological quantification. However, some heart failure patients are ineligible for MRI because of implanted devices, such as implantable cardioverter-defibrillator, cardiac resynchronization therapy and left ventricular assist devices. A third potential application could be systematic fibrosis quantification in animal models, for example in models of ischemic28 or non-ischemic cardiomyopathy29. Effects of novel therapies on myocardial fibrosis can be examined in randomized preclinical trials, by systematically comparing the amount of fibrosis on histology. A standardized preclinical model with induced myocardial infarction can be used to study new therapeutics, such as cell therapy, as a strategy to attenuate cardiac fibrosis and stop progression towards heart failure.30, 31

101

PART TWO CHAPTER 6

In conclusion, in PLN p.Arg14del mutation associated cardiomyopathy myocardial fibrosis is predominantly present in the left posterolateral wall, whereas fatty changes are more pronounced in the wall of the right ventricle. In the analysed heart slices from PLN p.Arg14del mutation carriers with non-ischemic cardiomyopathy we found an overlap in distribution pattern between patients with DCM and AC and a significant higher percentage of adipose tissue in AC compared to control hearts (p=0.028). We developed a novel method for systematic high resolution digital histological quantification of fibrosis and fatty tissue in the heart. This method can be used to assess cardiac fibrosis and fatty tissue in a broad range of human cardiomyopathies, animal models and can serve as gold standard for noninvasive imaging techniques. Acknowledgments We thank Petra van der Kraak-Homoet for technical assistance. We are also grateful to R.K. Stellato, MSc for statistical advice.

102

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

REFERENCES 1. 2.

3. 4. 5. 6. 7.

10. 11. 12. 13.

14.

15.

16. 17. 18.

19.

20. 21.

Weber KT, Sun Y, Bhattacharya SK, Ahokas RA, Gerling IC. Myofibroblast-mediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol. 2013;10:15-26. Marcus FI, McKenna WJ, Sherrill D, Basso C, Bauce B, Bluemke DA, et al. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the task force criteria. Circulation. 2010;121:1533-41. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA. Assessment of myocardial fibrosis with cardiovascular magnetic resonance. J Am Coll Cardiol. 2011;57:891-903. de Jong S, van Veen TA, de Bakker JM, van Rijen HV. Monitoring cardiac fibrosis: a technical challenge. Neth Heart J. 2012;20:44-8. Schmitt JP, Kamisago M, Asahi M, Li GH, Ahmad F, Mende U, et al. Dilated cardiomyopathy and heart failure caused by a mutation in phospholamban. Science. 2003;299:1410-3. MacLennan DH, Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol. 2003;4:566-77. Haghighi K, Kolokathis F, Pater L, Lynch RA, Asahi M, Gramolini AO, et al. Human phospholamban null results in lethal dilated cardiomyopathy revealing a critical difference between mouse and human. J Clin Invest. 2003;111:869-76. Haghighi K, Kolokathis F, Gramolini AO, Waggoner JR, Pater L, Lynch RA, et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc Natl Acad Sci U S A. 2006;103:1388-93. van der Zwaag PA, van Rijsingen IA, Asimaki A, Jongbloed JD, van Veldhuisen DJ, Wiesfeld AC, et al. Phospholamban R14del mutation in patients diagnosed with dilated cardiomyopathy or arrhythmogenic right ventricular cardiomyopathy: evidence supporting the concept of arrhythmogenic cardiomyopathy. Eur J Heart Fail. 2012;14:1199-207. Huisman A, Looijen A, van den Brink SM, van Diest PJ. Creation of a fully digital pathology slide archive by high-volume tissue slide scanning. Hum Pathol. 2010;41:751-7. Ruifrok AC, Johnston DA. Quantification of histochemical staining by color deconvolution. Anal Quant Cytol Histol. 2001;23:291-9. Pratt WK. Digital image processing: PIKS Scientific inside. 4th ed. Hoboken, New Jersey: Wiley; 2007. Gustavsson M, Verardi R, Mullen DG, Mote KR, Traaseth NJ, Gopinath T, et al. Allosteric regulation of SERCA by phosphorylation-mediated conformational shift of phospholamban. Proc Natl Acad Sci U S A. 2013;110:17338-43. La Gerche A, Burns AT, Mooney DJ, Inder WJ, Taylor AJ, Bogaert J, et al. Exercise-induced right ventricular dysfunction and structural remodelling in endurance athletes. Eur Heart J. 2012;33:998-1006. Masci PG, Barison A, Aquaro GD, Pingitore A, Mariotti R, Balbarini A, et al. Myocardial delayed enhancement in paucisymptomatic nonischemic dilated cardiomyopathy. Int J Cardiol. 2012;157:43-7. Basso C, Corrado D, Marcus FI, Nava A, Thiene G. Arrhythmogenic right ventricular cardiomyopathy. Lancet. 2009;373:1289-300. Marcus FI, Fontaine GH, Guiraudon G, Frank R, Laurenceau JL, Malergue C, et al. Right ventricular dysplasia: a report of 24 adult cases. Circulation. 1982;65:384-98. Corrado D, Basso C, Thiene G, McKenna WJ, Davies MJ, Fontaliran F, et al. Spectrum of clinicopathologic manifestations of arrhythmogenic right ventricular cardiomyopathy/dysplasia: a multicenter study. J Am Coll Cardiol. 1997;30:1512-20. Te Riele AS, James CA, Philips B, Rastegar N, Bhonsale A, Groeneweg JA, et al. Mutationpositive arrhythmogenic right ventricular dysplasia/cardiomyopathy: the triangle of dysplasia displaced. J Cardiovasc Electrophysiol. 2013;24:1311-20. Basso C, Bauce B, Corrado D, Thiene G. Pathophysiology of arrhythmogenic cardiomyopathy. Nat Rev Cardiol. 2012;9:223-33. Basso C, Ronco F, Marcus F, Abudureheman A, Rizzo S, Frigo AC, et al. Quantitative assessment of endomyocardial biopsy in arrhythmogenic right ventricular cardiomyopathy/dysplasia: an in vitro validation of diagnostic criteria. Eur Heart J. 2008;29:2760-71.

103

PART TWO CHAPTER 6

22. Krajewska M, Smith LH, Rong J, Huang X, Hyer ML, Zeps N, et al. Image analysis algorithms for immunohistochemical assessment of cell death events and fibrosis in tissue sections. J Histochem Cytochem. 2009;57:649-63. 23. Farris AB, Adams CD, Brousaides N, Della Pelle PA, Collins AB, Moradi E, et al. Morphometric and visual evaluation of fibrosis in renal biopsies. J Am Soc Nephrol. 2011;22:176-86. 24. Osman OS, Selway JL, Kepczynska MA, Stocker CJ, Oâ&#x20AC;&#x2122;Dowd JF, Cawthorne MA, et al. A novel automated image analysis method for accurate adipocyte quantification. Adipocyte. 2013;2:160-4. 25. Jacoby D, McKenna WJ. Genetics of inherited cardiomyopathy. Eur Heart J. 2012;33:296-304. 26. Schalla S, Bekkers SC, Dennert R, van Suylen RJ, Waltenberger J, Leiner T, et al. Replacement and reactive myocardial fibrosis in idiopathic dilated cardiomyopathy: comparison of magnetic resonance imaging with right ventricular biopsy. Eur J Heart Fail. 2010;12:227-31. 27. Moon JC, Messroghli DR, Kellman P, Piechnik SK, Robson MD, Ugander M, et al. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR Working Group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson. 2013;15:92. 28. Pop M, Ghugre NR, Ramanan V, Morikawa L, Stanisz G, Dick AJ, et al. Quantification of fibrosis in infarcted swine hearts by ex vivo late gadolinium-enhancement and diffusion-weighted MRI methods. Phys Med Biol. 2013;58:5009-28. 29. Gho JM, Kummeling GJ, Koudstaal S, Jansen of Lorkeers SJ, Doevendans PA, Asselbergs FW, et al. Cell therapy, a novel remedy for dilated cardiomyopathy? A systematic review. J Card Fail. 2013;19:494-502. 30. Elnakish MT, Kuppusamy P, Khan M. Stem cell transplantation as a therapy for cardiac fibrosis. J Pathol. 2013;229:347-54. 31. Koudstaal S, Jansen of Lorkeers S, Gho JM, van Hout GP, Jansen MS, Grundeman PF, et al. Myocardial infarction and functional outcome assessment in pigs. J Vis Exp. 2014.

104

CARDIAC DIGITAL HISTOLOGICAL QUANTIFICATION

SUPPLEMENTARY INFORMATION

Supplementary Figure 1. Schematic overview of Case 6 Arrhythmogenic Cardiomyopathy with representative images. A,B, schematic heart slice overview of mean fibrosis (A) and adipose tissue (B). Corresponding slide numbers are shown inside the inner rings (n = 11). Mean percentage of fibrosis (A) or adipose tissue (B) per area have been superimposed on the overview (values rounded to the nearest whole number, values <20% are shown in white to improve readability). C,E,G, raw microscopic slide images after Massonâ&#x20AC;&#x2122;s trichrome staining, respectively corresponding to slide 1, 5 and 7. D,F,H, corresponding slides after digital processing. Red: cardiomyocytes. Blue: connective tissue. Pseudo green: adipose tissue. The slides depicted in Figure 1C-F are also derived from the same heart, 1C,D correspond to slide 3 and 1E,F correspond to slide 10.

105

PART TWO CHAPTER 6

Supplementary Table 1. Mean percentage of fibrosis and adipose tissue per region and condition Control (n=3) Fibrosis

Adipose tissue

Region 1

AC (n=3)

DCM (n=5)

Mean (%)

2.5

0.3

40.1

8.8

38.9

10.5

Region 2

2.2

0.6

25.4

9.7

31.2

11.3

Region 3

2.3

0.4

17.7

4.3

33.8

12.0

Region 4

1.8

0.7

18.5

10.8

34.0

14.4

Region 5

2.4

0.8

26.4

8.2

30.6

10.2

Region 6

2.4

0.4

37.6

2.3

34.5

13.5

Region 7

4.6

1.6

24.9

3.6

23.1

9.3

Region 8

5.1

2.0

19.7

1.7

19.3

9.2

Region 1

0.7

6.9

4.7

5.8

6.0

Region 2

0.6

10.1

10.8

5.7

9.7

Region 3

0.7

0.4

3.6

3.9

1.8

1.1

Region 4

2.0

2.5

4.9

2.7

2.5

2.2

Region 5

0.7

0.4

6.4

7.5

1.9

0.7

Region 6

0.7

0.2

7.7

5.6

2.4

1.1

Region 7

7.8

6.1

37.2

14.1

26.1

20.0

Region 8

14.4

5.9

28.9

4.1

24.3

12.0

Regions correspond to the depicted regions in Figure 6C. AC, Arrhythmogenic Cardiomyopathy; DCM, Dilated Cardiomyopathy; SD, standard deviation.

106

107

PART TWO

Finding Fibrosis Patterns

Chapter

The Distribution Pattern of Fibrosis in Genetic Cardiomyopathy is Related to the Type of Pathogenic Mutation In preparation

Johannes M.I.H. Gho1*, Shahrzad Sepherkhouy2*, RenĂŠ van Es1, Magdalena Harakalova1,2, Nicolaas de Jonge1, Jasper J. van der Smagt3, Roel Goldschmeding2, Roel A. de Weger2, Folkert W. Asselbergs1, Aryan Vink2 *Both authors contributed equally

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Pathology, University Medical Center Utrecht, Utrecht, the Netherlands

Department of Medical Genetics, University Medical Center Utrecht, Utrecht, the Netherlands

PART TWO CHAPTER 7

ABSTRACT Aims Genetic cardiomyopathies form a heterogeneous group in which myocardial fibrosis is frequently observed. We aimed to unravel the exact distribution pattern of cardiac fibrosis in cardiomyopathies caused by different specific pathogenically mutated genes. Methods and results A complete transversal slice was obtained from hearts of 28 patients with end stage heart failure due to non-ischemic cardiomyopathy and a known mutation and 4 controls. Fibrosis and fatty changes were quantitatively analyzed using digital microscopy. Hearts with mutations in the nuclear envelope protein lamin A/C (n=4) and hearts with sarcomeric mutations (n=10) showed circumferential subendocardial and midmyocardial fibrosis in left and right ventricle, except for 2 hearts from patients that initially had a hypertrophic cardiomyopathy. In hearts with a desmosomal (n=3) or a calcium pump (phospholamban; n=8) mutation fibrosis and fibrofatty replacement were observed in the left ventricle outer myocardium, mainly in the posterolateral wall, and in the right ventricle. In phospholamban mutated hearts the left ventricle was significantly more affected than in the desmosomal group. In hearts with a desminopathy (n=3) the fibrosis was mainly located in the outer myocardium of the left ventricle in combination with fibrosis and minor fibrofatty replacement in the right ventricle. Conclusion Distribution of fibrosis in genetic cardiomyopathies is related to the type of pathogenic mutation. These specific patterns may provide a roadmap for cardiac imaging interpretation and add to our understanding of disease patterns.

110

FIBROSIS IN GENETIC CARDIOMYOPATHIES

INTRODUCTION Genetic cardiomyopathies are disorders leading to declined functionality of the heart muscle and have been associated with different pathogenic mutations.1 Currently patients with cardiomyopathy are categorized based on their clinical presentation and clinical phenotype. However, the clinical presentation and clinical phenotype of a cardiomyopathy caused by a specific pathogenic mutation can vary enormously among patients. For example the phospholamban (PLN) p.Arg14del mutation can lead to a clinical presentation of heart failure, but rhythm disturbances can also be the primary presentation.2 In genetic cardiomyopathies the clinical presentation can also vary greatly due to variable penetrance, even in siblings with the same specific gene mutation. 3 This heterogenous clinical presentation of genetic cardiomyopathies, even among patients with the same genetic mutation, makes it challenging for clinicians to adequately classify these patients. Therefore more tools are needed to distinguish the different genetic cardiomyopathies and their underlying mechanisms. Although genetic cardiomyopathies are heterogeneous, histologically observed cardiac fibrosis is a hallmark feature seen in most cardiomyopathies. A multitude of mechanisms accompanied with stress and injury can cause changes in the extracellular matrix leading to increased production of collagen by interstitial fibroblasts. Fibrosis is a major cause of cardiac stiffness, diastolic and systolic impairment, arrhythmias and can eventually lead to heart failure.4 Knowledge about the pattern of fibrosis in relation to the pathogenic mutation will improve diagnosis and may eventually lead to better understanding of disease mechanisms. We recently demonstrated that PLN p.Arg14del cardiomyopathy has a specific pattern of fibrosis in the heart that is

independent of clinical presentation.5 The aim of this study was to unravel the amount and distribution pattern of cardiac fibrosis in cardiomyopathies caused by different specific pathogenic mutations.

METHODS The study met the criteria of the code of proper use of human tissue that is used in the Netherlands. The study was approved by the scientific advisory board of the biobank of the University Medical Center Utrecht, Utrecht, the Netherlands (protocol no. 12/387). Thirty-one hearts with cardiomyopathy and a known cardiac pathogenic mutation were obtained at transplantation (n=25) or autopsy (n=6). At the time of transplantation or death all patients had severe heart failure with reduced ejection fraction. Initially, at the time of clinical presentation, the cardiomyopathy of these patients had been classified as dilated (DCM), arrhythmogenic (ACM) or hypertrophic (HCM) cardiomyopathy by their cardiologist. For the initial diagnosis ACM the Revised Task Force criteria were used.6 Four non-cardiomyopathy hearts served as control group: three donor hearts rejected for transplantation and one heart obtained at autopsy of a patients with a non-cardiac cause of death. We used a systematic methodology for high resolution digital cardiac fibrosis quantification as previously described.5 Hearts were cut in transverse (short-axis) slices of 1 cm thick starting at the apex including both ventricles. A complete transverse heart slice of 1 cm thick was obtained

111

PART TWO CHAPTER 7

at the mid-ventricular level (halfway between the atrioventricular valves and the apex) and fixed in formalin. This heart slice was cut in smaller pieces and embedded in paraffin. A map of the heart slice was drawn to annotate the origin of each tissue specimen. Microscopic slides were stained with Masson’s trichrome (Artisan Link Pro, Dako, Glostrup, Denmark) and the slides were scanned at 20x magnification (Aperio, Leica Biosystems, Nussloch, Germany). The cardiac tissue was digitally divided into three layers: trabeculated myocardium, inner compact myocardium and outer compact myocardium. The epicardial fatty tissue was not analyzed. In case of fibrofatty replacement and in the right ventricle (RV), the border between epicardial fatty tissue and the myocardium was drawn by a line that connected the most outer parts (remnants) of the myocardium. For the interventricular septum and in some cases for the RV it was not possible to divide the myocardium into an inner and outer portion so in that case the whole compact myocardium was analyzed as one layer. In all three layers the percentage of connective tissue (blue), cardiomyocytes (red) and adipose tissue (cells with non-stained cytoplasm) was digitally quantified using MATLAB (Release R2012a, The MathWorks, Inc., Natick, Massachusetts, United States; http://sourceforge.net/projects/fibroquant). The annotated map of the transversal slice is automatically transformed to a standardized schematic overview and results in percentage fibrosis and fatty tissue are displayed using a color scale (Addendum Figure 1). For the comparison between the different mutation groups the heart was divided into 6 regions for the left ventricle (posterolateral, posterior, septal, anterior, anterolateral, and lateral part) and 2 regions for the right ventricle (posterior and anterior). Statistics were performed using IBM SPSS Statistics (Version 20.0, IBM Corporation, Armonk, New York, United States). Categorical data were compared using a Fisher’s exact test. A MannWhitney test was used to compare continuous variables between 2 groups. A Kruskal-Wallis test was used to compare continuous variables between multiple groups. Data are presented as median [interquartile range, IQR] or mean ± standard error of the mean.

RESULTS Patients were grouped in functional groups of gene mutations: sarcomeric, nuclear envelope, desmin filament network, calcium pump and desmosomal mutations (Table 1). Three patients had two pathogenic mutations in different functional groups and were excluded for further analysis. Patient characteristics are summarized in Table 1. The mean age was 36 ± 14 years; 16 patients (57%) were male. During heart transplantation or death all patients had severe reduced ejection fraction heart failure. The initial clinical diagnosis that was noted in the medical file by the treating cardiologist at the time of the diagnosis of the cardiomyopathy varied among groups. All 4 lamin A/C mutation patients presented with DCM and all 3 desmosomal mutation patients presented with ACM. The sarcomeric patients presented with DCM or HCM. Sarcomeric patients with initial HCM phenotype that progressed to a DCM phenotype had the following mutations: myosin heavy chain 7 and myosin binding protein C3. In the desminopathy and calcium pump groups patients presented with DCM or ACM. Half of the patients had previous implantation of a left ventricle assist device (LVAD). On routine coronary angiography for pre-heart transplant / LVAD

112

FIBROSIS IN GENETIC CARDIOMYOPATHIES

implantation evaluation two patients had coronary artery disease for which single-vessel percutaneous coronary intervention had been performed (1 phospholamban mutation and 1 lamin A/C mutation). The total amount of fibrosis The total amount of fibrosis at the midventricular level per condition in left and right ventricle is shown in Figure 1. The four controls revealed almost no fibrosis (median [IQR] 3% [2-4]). On average the hearts with a calcium pump mutation (25% [21-38]) and the hearts with a desminopathy (25% [24-26]) had the most fibrosis, followed by the sarcomeric gene mutations (23% [16-28]), desmosomal mutations (14% [14-17]) and the nuclear envelope gene mutations (11% [7-24]; p=0.022 between mutation groups).

Table 1. Functional groups of mutated genes

Functional group of mutated genes

Gene mutations in group

Sarcomeric

5 titin, 2 myosin binding protein C, 1 myosin heavy chain 7, 1 troponin T, 1 troponin I3

Nuclear envelope

4 lamin A/C

Desmin filament network

2 desmin, 1 alpha-B-crystallin

Calcium pump

8 phospholamban

Desmosomal

2 plakophilin-2, 1 desmoplakin

Figure 1. Boxplots of mean percentage of fibrosis in both ventricles in the different mutation groups. P=0.022 between mutation groups.

113

PART TWO CHAPTER 7

Distribution of fibrosis and adipose tissue Examples of Massonâ&#x20AC;&#x2122;s trichrome stains of the myocardium of the left ventricle (LV) per condition are shown in Figure 2. The stains in Figure 2 show that the fibrosis was not evenly distributed throughout the wall of the left ventricle. Schematic overviews of the distribution of fibrosis and adipose tissue for each layer (trabecular layer, inner compact myocardium, outer compact myocardium) per condition are shown in Figure 3. The average percentage of fibrosis per layer in the different mutation groups is shown in Addendum Figure 2.

Figure 2. Histological findings of the posterolateral wall in the control heart and pathogenic mutations. The slides were stained with Massonâ&#x20AC;&#x2122;s trichrome: control heart with almost no fibrosis; hearts with a desminopathy, calcium pump (phospholamban) mutation and desmosomal mutation with fibrosis in the outer layer of the myocardium; hearts with a mutation in the genes encoding nuclear envelope (lamin A/C) and sarcomeric proteins with subendocardial fibrosis and fibrosis in the inner layer of the compact myocardium.

114

FIBROSIS IN GENETIC CARDIOMYOPATHIES

Figure 3. Schematic overview of fibrosis (A) and adipose tissue (B) in the myocardium of heart slices of control hearts and the different groups of mutations. The results of the digital quantification in percentage of fibrosis and adipose tissue are shown using a colorscale. LV, left ventricle; RV, right ventricle; Ant., anterior; Post., posterior.

Three different distribution patterns were observed: Pattern 1: classical arrhythmogenic pattern with left ventricular involvement •

Left ventricle: fibrosis and some fatty changes where the outer compact myocardium of the posterolateral wall is most affected.

•

Right ventricle: replacement of myocardium by adipose tissue accompanied with fibrosis

•

(fibrofatty replacement). Mutation groups: calcium pump (PLN) and desmosomal mutations. Left ventricular fibrosis was more pronounced in the PLN mutation group (27% [22-39]) than in the desmosomal group (15% [14-17], p=0.024). A trend towards more adipose tissue in the RV of the desmosomal mutation group (46% [43-53]) than in the PLN group (28% [26-39], p=0.07) was found.

Pattern 2: circumferential subendocardial fibrosis LV and RV •

Left ventricle: predominantly circumferential trabecular (subendocardial) and midmyocardial fibrosis without adipose tissue.

•

Right ventricle: minor to moderate fibrosis without increased fatty infiltration.

•

Mutation groups: sarcomeric and nuclear envelope mutations. All lamin A/C mutation patients revealed pattern 2 where the RV fibrosis was most pronounced in the ventral wall.

115

PART TWO CHAPTER 7

The sarcomeric group is a heterogenous group consisting of patients presenting with DCM and HCM. Patients in the sarcomeric group with DCM (mutations in titin (n=5), troponin T2 (n=1) and troponin I3 (n=1)) all had pattern 2. In the titin subgroup 2/5 hearts also parts of the outer compact myocardium revealed fibrosis. Sarcomeric mutation patients presenting with HCM: the patient with a myosin heavy chain 7 mutation (n=1) also presented as pattern 2 with severe replacement fibrosis of the septum, whereas both patients with a myosin binding protein C3 mutation (n=2) showed a more diffuse distribution with both interstitial fibrosis and replacement fibrosis (Addendum Figure 3). Pattern 3: variable predominantly subepicardial fibrosis LV with fibrosis and minor fatty changes RV •

Left ventricle: fibrosis predominantly most pronounced in the outer compact myocardium, location varying per patient with involvement of the anterior, lateral and posterior wall. Minor fatty changes in the LV of one patient.

•

Right ventricle: moderate fibrosis and minor fibrofatty changes.

•

Mutation group: desminopathies.

Posterolateral wall of the left ventricle The posterolateral wall of the left ventricle was the most distinctive area when trabecular (endocardial), inner compact and outer compact (epicardial) myocardium were compared. Therefore subgroup analysis was performed in this region where these three layers of the wall were compared and divided into 3 groups: fibrosis/fatty changes epicardial > endocardial, epicardial < endocardial and epicardial = endocardial (difference at least 5%). Strikingly, all hearts in the desmosomal (3/3) and PLN (8/8) groups showed more fibrosis/fatty changes in the epicardial than in the endocardial area. In all hearts of the nuclear envelope group (4/4) the opposite pattern was observed with more fibrosis in the endocardial area than in the epicardial area. 6/8 (75%) of the hearts with a sarcomeric mutation also had more endocardial than

Sarcomeric n=10

Lamin A/C n=4

Desminopathy n=3

Calcium pump n=8

Desmosomal n=3

Total n=28

Age at diagnosis (yrs)

Control n=4

Table 2. Patient characteristics per mutation group

Unknown

29±15

48±8

40±2

36±13

34±7

36±14

Sex (m / f)

Unknown

6/4

3/1

2/1

3/5

2/1

16/12

Initial clinical diagnosis

7 DCM 3 HCM

4 DCM

2 DCM 1 ACM

6 DCM 2 ACM

3 ACM

19 DCM 6 ACM 3 HCM

LVAD

4/10

3/4

2/3

5/8

0/3

14/28

PM/ICD/CRT-D

6/10

3/4

3/3

7/8

3/3

22/28

DCM = dilated cardiomyopathy, HCM = hypertrophic cardiomyopathy, ACM = arrhythmogenic cardiomyopathy, LVAD = left ventricular assist device, PM = pacemaker, ICD = Implantable Cardioverter Defibrillator, CRT-D = Cardiac Resynchronization Therapy Defibrillator.

116

FIBROSIS IN GENETIC CARDIOMYOPATHIES

epicardial fibrosis. In the desminopathy group 1/3 (33%) hearts showed more epicardial than endocardial alterations, whereas the other 2 hearts did not reveal differences between both wall layers. The latter can be explained by the fact that fibrosis in the desminopathy is more pronounced in the anterior and posterior wall than in the posterolateral wall. Differences among mutation groups were significant (p<0.001), results are shown in Figure 4.

7 Figure 4. Pattern of fibrosis and / or adipose tissue in the posterolateral wall of the left ventricle is associated with mutation group. Trabecular (endocardial), inner compact and outer compact (epicardial) myocardium were compared and divided into 3 groups: fibrosis/fatty changes epicardial > endocardial, epicardial < endocardial and epicardial = endocardial (difference at least 5%).

DISCUSSION By applying our previously developed method for high resolution systematic digital histological quantification of cardiac fibrosis and adipose tissue, we aimed to unravel the exact distribution pattern of cardiac fibrosis in cardiomyopathies caused by different specific pathogenically mutated genes. We found that distribution pattern of fibrosis in genetic cardiomyopathies is related to the type of pathogenic genetic mutation. Especially the posterolateral wall of the left ventricle was found to be highly discriminating between groups of mutations. Desmosomal and calcium pump mutations Desmosomes are located in the intercalated disks that connect cardiomyocytes. Desmosomal mutations have been associated with ACM that was first described as fibrofatty replacement of the right ventricle in the classical triangle of dysplasia consisting of the right ventricle inflow tract, outflow tract and apex. Later it was recognized that ACM is not limited to the right ventricle

117

PART TWO CHAPTER 7

and that the left ventricle is frequently involved not only in end stage disease, but also in earlier stages.7 In the present study patients with desmosomal mutations (2 plakophilin-2 and 1 desmoplakin mutation) revealed right ventricular fibrofatty changes and fibrosis with fatty changes in the left ventricular wall, most pronounced in the posterolateral wall. Our observations in the desmosomal group are in line with earlier observations in autopsy studies from ACM patients with unknown mutations.8, 9 In addition our results in the left ventricle confirm late gadolinium enhancement in cardiovascular magnetic resonance studies that typically involves the subepicardial and midwall layers of the inferolateral and inferoseptal regions of the left ventricle in ACM.10 We recently demonstrated that hearts from patients with a PLN p.Arg14del mutation also have a pattern of right ventricle fibrofatty replacement and left ventricle fibrosis with fatty changes mostly in the posterior wall, independently of clinical presentation.5 To the best of our knowledge the present study is the first histological study that shows that PLN hearts have significantly more fibrosis in the left ventricle and a trend towards less fibrofatty replacement in the right ventricle as compared to hearts with desmosomal mutations. Our results confirm recent observations in a cohort of 142 Dutch AC patients and in a combined USA and Dutch cohort of 577 patients. In these cohorts also more left ventricular involvement in the PLN mutated patients was found compared to desmosomal mutations using electrocardiographic and cardiovascular magnetic resonance criteria.11, 12 Nuclear envelope mutations The lamin A/C gene encodes by alternative splicing for the nuclear envelope proteins lamin A and C, which are major components of the nuclear lamina that contribute to the structural integrity of the nuclear envelope and provide mechanical support for the nucleus. Mutations in the lamin A/C gene can cause several disorders including dilated cardiomyopathy accompanied by conduction abnormalities/arrhythmias.13 We observed a pattern of left ventricular subendocardial and middle layer fibrosis and subtle right ventricular fibrosis more in the ventral than dorsal wall of the right ventricle. Previous histological postmortem examinations of fullthickness myocardium in 5 patients demonstrated extensive areas of interstitial and replacement fibrosis throughout the left and right ventricular myocardium and extensive fibrosis of the cardiac conduction system without further details of the exact location of the fibrosis.14 In another study of a large Japanese family with DCM and AV block due to a lamin A/C mutation, fibrosclerosic degeneration in the AV node and interstitial fibrosis in RV endocardial biopsies were described.15 In our study, hearts with lamin A/C mutations revealed the least amount of fibrosis at the mid ventricular level as compared to the other mutations groups. This might be partly explained by the observations in a previous cardiovascular magnetic resonance study of lamin A/C mutation carriers where late gadolinium enhancement was predominantly located in the basal or midventricular septal wall, including the area around the AV node, and diffuse enhancement of LV myocardium was observed in advanced cases of the disease.16 So we add to previous literature that the fibrosis in lamin A/C mutation induced cardiomyopaty is predominantly present in the subendocardial and midmyocardial layer of the left ventricle and in the ventral wall of the right ventricle in addition to the previously established fibrosis in the AV node area.

118

FIBROSIS IN GENETIC CARDIOMYOPATHIES

Sarcomeric mutations including titin Titin is the largest human protein and two titin molecules together span the entire sarcomere and are anchored at the Z-line and M-line.17 It was reported 2012 that truncating mutations in TTN, the gene encoding titin, account for a significant portion of about 25% of the genetic etiology in familial DCM.18 Since this report there has been some debate about the pathogenicity due to observations of TTN mutations in apparently healthy controls and in the general population.19 We observed that TTN mutations induce predominantly subendocardial and midmyocardial left ventricular fibrosis and in 2/5 hearts also some fibrosis in the outer myocardial layer next fibrosis in the right ventricular wall. In a previous case report of one explanted heart with a novel mutation, c.58880insA, in exon 326 of the titin gene, some interstitial fibrosis was described without further specification of the location of the fibrosis.20 Another study reported that histopathological characteristics of cardiac specimens from subjects with TTN truncating mutations were typical of idiopathic dilated cardiomyopathy, without further specification of the fibrosis pattern.18 In a titin knock-in mouse model mimicking the c.43628insAT allele of the TTN 2-bp insertion mutation diffuse myocardial fibrosis was observed, also without further specification.21 Sarcomeric cardiomyopathies are caused by a number of mutated sarcomeric genes and can lead to a variety of phenotypes, including HCM, DCM and restrictive cardiomyopathy.22 In our series we have hearts from 3 patients that initially presented as HCM and progressed to DCM. Two of these patients had a MYBC3 mutation and hearts from these patients showed fibrosis in the wall of both ventricles without specific pattern consisting of both interstitial fibrosis and replacement type of fibrosis. One patient with a MYH7 mutation showed severe fibrosis of the septum with subendocardial and midventricular fibrosis in the left ventricle. These observations

are in line with previous observations in HCM patients with unknown genetic mutations where in addition to interstitial fibrosis areas of replacement fibrosis were observed, probably induced by ischemic myocardial damage due to insufficient oxygen supply to the hypertrophic myocardium.23 Two hearts in our series had a mutation in the thin sarcomeric filaments troponin I3 and troponin T2. In both hearts the fibrosis was predominantly present in the subendocardial and midmyocardial layers of both ventricles, but in the TNNT2 heart changes were only subtle. In a recent case report of hearts from two cousins with the same p.Arg186Gln (CGG>GAG)(R186Q) mutation in exon 8 of the troponin I (TNN13) gene both explanted hearts revealed increased interstitial fibrosis without further specification of the pattern and one of the two hearts showed severe scars throughout the wall of the left ventricle.24 In another report of two cases of elderly individuals (74 and 92 years old) with the Lys183del mutation in the TNNI3 gene transmural diffuse scarring of the left ventricle was observed in one case and in the other case the scarring was predominantly present in the septum and anterior wall.25 In two cases reports of the Arg92Trp mutation in the TNNT gene with clinical DCM-like features the anterior walls and interventricular septa of the hearts were replaced with extensive fibrosis and showed thinning. 26 In these case reports diffuse interstitial fibrosis was not mentioned. Thus we observed fibrosis mainly in the subendocardial and midmyocardial areas of the left and right ventricles in hearts with mutations in the sarcomere. In hearts with first presentation of HCM this pattern was not found, which was caused by large areas of replacement fibrosis.

119

PART TWO CHAPTER 7

Desminopathies Desmin is a muscle-specific intermediate filament essential for proper muscular structure and function. Mutations affecting desmin expression or promoting its aggregation lead to skeletal myopathies and/or cardiomyopathies. We observed fibrosis most pronounced in the outer compact myocardium of the left ventricle of which the location varied per patient and fibrosis and minor fibrofatty changes in the right ventricle. Our results are in line with findings of delayed enhancement in a cardiovascular MR imaging study in patients with desmin mutations with skeletal muscle weakness in which delayed enhancement was detected in a midwall position of the left ventricle without signs of fibrosis in the subendocardial region.27 In this imaging study the location of fibrosis varied per patient and involved the anterolateral, lateral and the inferoseptal LV wall. In our study the patient with the most fibrofatty replacement in the RV initially presented as AMC. It is known that atrioventricular conduction abnormalities are a frequent feature of desminopathy attributed to the fact that the heart conduction system is rich in desmin.28 However ventricular tachycardiaâ&#x20AC;&#x2122;s have also been described, which might be partly explained by the fibrofatty changes that we observed in the RV.29 Clinical implications and limitations Our results may have clinical implications. Techniques of CMR with late gadolinium enhancement have improved dramatically in recent years. In addition novel MR techniques with or without using contrast agents look promising for cardiac fibrosis detection.30 We observed that in the LV posterolateral wall the pattern of fibrosis and fatty changes is highly distinctive for groups of mutations. Therefore we hope the present study will provide a roadmap for further in vivo cardiac fibrosis assessment in patients. Our study is limited by the fact that groups of hearts with mutated genes are small. However, our study is a start of systematic analysis of hearts with known mutations. We think that in coming years cardiovascular pathologists around the world can work in collaboration with cardiologists and clinical geneticists to create databases and biobanks of cardiomyopathy patients that include clinical variables, pathogenic mutations and detailed examination of cardiac tissue. For this it is important that pathologists work according to standardized protocols when examining hearts and initiatives to create such a protocol, e.g., under the auspices of the Association of European Cardiovascular Pathology are urgently needed. In conclusion, we observed that the pattern of fibrosis and fatty changes in cardiomyopathy hearts with known pathogenic mutations as assessed by digital histological analysis is associated with the mutation group. The posterolateral wall of the LV appeared to be the most discriminating area for the identification of the different groups of mutations.

120

FIBROSIS IN GENETIC CARDIOMYOPATHIES

REFERENCES 1. 2.

3. 4. 5.

7. 8. 9.

10.

11.

12.

13.

14.

15.

16.

17. 18. 19. 20.

Watkins H, Ashrafian H, Redwood C. Inherited cardiomyopathies. N Engl J Med. 2011;364:164356. van der Zwaag PA, van Rijsingen IA, Asimaki A, Jongbloed JD, van Veldhuisen DJ, Wiesfeld AC, et al. Phospholamban R14del mutation in patients diagnosed with dilated cardiomyopathy or arrhythmogenic right ventricular cardiomyopathy: evidence supporting the concept of arrhythmogenic cardiomyopathy. Eur J Heart Fail. 2012;14:1199-207. Jacoby D, McKenna WJ. Genetics of inherited cardiomyopathy. Eur Heart J. 2012;33:296-304. Weber KT, Sun Y, Bhattacharya SK, Ahokas RA, Gerling IC. Myofibroblast-mediated mechanisms of pathological remodelling of the heart. Nat Rev Cardiol. 2013;10:15-26. Gho JM, van Es R, Stathonikos N, Harakalova M, te Rijdt WP, Suurmeijer AJ, et al. High resolution systematic digital histological quantification of cardiac fibrosis and adipose tissue in phospholamban p.Arg14del mutation associated cardiomyopathy. PloS One. 2014;9:e94820. Marcus FI, McKenna WJ, Sherrill D, Basso C, Bauce B, Bluemke DA, et al. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the Task Force Criteria. Eur Heart J. 2010;31:806-14. Basso C, Bauce B, Corrado D, Thiene G. Pathophysiology of arrhythmogenic cardiomyopathy. Nat Rev Cardiol. 2012;9:223-33. Basso C, Thiene G, Corrado D, Angelini A, Nava A, Valente M. Arrhythmogenic right ventricular cardiomyopathy. Dysplasia, dystrophy, or myocarditis? Circulation. 1996;94:983-91. Corrado D, Basso C, Thiene G, McKenna WJ, Davies MJ, Fontaliran F, et al. Spectrum of clinicopathologic manifestations of arrhythmogenic right ventricular cardiomyopathy/dysplasia: a multicenter study. J Am Coll Cardiol. 1997;30:1512-20. Sen-Chowdhry S, Syrris P, Prasad SK, Hughes SE, Merrifield R, Ward D, et al. Left-dominant arrhythmogenic cardiomyopathy: an under-recognized clinical entity. J Am Coll Cardiol. 2008;52:2175-87. Groeneweg JA, van der Zwaag PA, Olde Nordkamp LR, Bikker H, Jongbloed JD, Jongbloed R, et al. Arrhythmogenic right ventricular dysplasia/cardiomyopathy according to revised 2010 task force criteria with inclusion of non-desmosomal phospholamban mutation carriers. Am J Cardiol. 2013;112:1197-206. Bhonsale A, Groeneweg JA, James CA, Dooijes D, Tichnell C, Jongbloed JD, et al. Impact of genotype on clinical course in arrhythmogenic right ventricular dysplasia/cardiomyopathyassociated mutation carriers. Eur Heart J. 2015;36:847-55. Fatkin D, MacRae C, Sasaki T, Wolff MR, Porcu M, Frenneaux M, et al. Missense mutations in the rod domain of the lamin A/C gene as causes of dilated cardiomyopathy and conductionsystem disease. N Engl J Med. 1999;341:1715-24. van Tintelen JP, Tio RA, Kerstjens-Frederikse WS, van Berlo JH, Boven LG, Suurmeijer AJ, et al. Severe myocardial fibrosis caused by a deletion of the 5' end of the lamin A/C gene. J Am Coll Cardiol. 2007;49:2430-9. Otomo J, Kure S, Shiba T, Karibe A, Shinozaki T, Yagi T, et al. Electrophysiological and histopathological characteristics of progressive atrioventricular block accompanied by familial dilated cardiomyopathy caused by a novel mutation of lamin A/C gene. J Cardiovasc Electrophysiol. 2005;16:137-45. Holmstrom M, Kivisto S, Helio T, Jurkko R, Kaartinen M, Antila M, et al. Late gadolinium enhanced cardiovascular magnetic resonance of lamin A/C gene mutation related dilated cardiomyopathy. J Cardiovasc Magn Reson. 2011;13:30. Liversage AD, Holmes D, Knight PJ, Tskhovrebova L, Trinick J. Titin and the sarcomere symmetry paradox. J Mol Biol. 2001;305:401-9. Herman DS, Lam L, Taylor MR, Wang L, Teekakirikul P, Christodoulou D, et al. Truncations of titin causing dilated cardiomyopathy. N Engl J Med. 2012;366:619-28. Golbus JR, Puckelwartz MJ, Fahrenbach JP, Dellefave-Castillo LM, Wolfgeher D, McNally EM. Population-based variation in cardiomyopathy genes. Circ Cardiovasc Genet. 2012;5:391-9. Yoskovitz G, Peled Y, Gramlich M, Lahat H, Resnik-Wolf H, Feinberg MS, et al. A novel titin mutation in adult-onset familial dilated cardiomyopathy. Am J Cardiol. 2012;109:1644-50.

121

PART TWO CHAPTER 7

21. Gramlich M, Michely B, Krohne C, Heuser A, Erdmann B, Klaassen S, et al. Stress-induced dilated cardiomyopathy in a knock-in mouse model mimicking human titin-based disease. J Mol Cell Cardiol. 2009;47:352-8. 22. Olivotto I, d'Amati G, Basso C, Van Rossum A, Patten M, Emdin M, et al. Defining phenotypes and disease progression in sarcomeric cardiomyopathies: contemporary role of clinical investigations. Cardiovasc Res. 2015;105:409-23. 23. Basso C, Thiene G, Corrado D, Buja G, Melacini P, Nava A. Hypertrophic cardiomyopathy and sudden death in the young: pathologic evidence of myocardial ischemia. Hum Pathol. 2000;31:988-98. 24. Roberts WC, Roberts CC, Ko JM, Grayburn PA, Tandon A, Kuiper JJ, et al. Dramatically different phenotypic expressions of hypertrophic cardiomyopathy in male cousins undergoing cardiac transplantation with identical disease-causing gene mutation. Am J Cardiol. 2013;111:1818-22. 25. Funada A, Masuta E, Fujino N, Hayashi K, Ino H, Kita Y, et al. Heterogeneity of clinical manifestation of hypertrophic cardiomyopathy caused by deletion of lysine 183 in cardiac troponin I gene. Int Heart J. 2010;51:214-7. 26. Shimizu M, Ino H, Yamaguchi M, Terai H, Uchiyama K, Inoue M, et al. Autopsy findings in siblings with hypertrophic cardiomyopathy caused by Arg92Trp mutation in the cardiac troponin T gene showing dilated cardiomyopathy-like features. Clin Cardiol. 2003;26:536-9. 27. Strach K, Sommer T, Grohe C, Meyer C, Fischer D, Walter MC, et al. Clinical, genetic, and cardiac magnetic resonance imaging findings in primary desminopathies. Neuromuscul Disord. 2008;18:475-82. 28. Goldfarb LG, Dalakas MC. Tragedy in a heartbeat: malfunctioning desmin causes skeletal and cardiac muscle disease. J Clin Invest. 2009;119:1806-13. 29. Luethje LG, Boennemann C, Goldfarb L, Goebel HH, Halle M. Prophylactic implantable cardioverter defibrillator placement in a sporadic desmin related myopathy and cardiomyopathy. Pacing Clin Electrophysiol. 2004;27:559-60. 30. van Oorschot JW, Gho JM, van Hout GP, Froeling M, Jansen Of Lorkeers SJ, Hoefer IE, et al. Endogenous contrast MRI of cardiac fibrosis: beyond late gadolinium enhancement. J Magn Reson Imaging. 2015;41:1181-9.

122

FIBROSIS IN GENETIC CARDIOMYOPATHIES

ADDENDUM

Addendum Figure 1. Transversal heart slice stained with Massonâ&#x20AC;&#x2122;s trichrome (A) is automatically transformed to a standardized schematic overview of fibrosis (B) and adipose tissue (C) to determine the exact pattern of fibrosis and fatty changes. The results of the digital quantification in percentage of fibrosis (B) or adipose tissue (C) is shown using a color scale. Ant. = anterior; LV = left ventricle; Post. = posterior; RV = right ventricle.

Addendum Figure 2. Average value of fibrosis of the left ventricular free wall in the trabecular, inner compact and outer compact myocardial layer.

123

PART TWO CHAPTER 7

Addendum Figure 3. Schematic overview of fibrosis in the different mutated genes in the sarcomeric group. The results of the digital quantification in percentage of fibrosis is shown using a color scale. Ant. = anterior; LV = left ventricle; Post. = posterior; RV = right ventricle.

124

125

PART THREE

Elucidating (Epi)genetic and Translating

Therapeutic Pathways

Chapter

Chromatin Regulation in Phospholamban R14del Mutation Associated Cardiomyopathy In preparation

Johannes M.I.H. Gho1*, Magdalena Harakalova1,2,4*, Shahrzad Sepehrkhouy2, Joyce van Kuik2, Noortje A.M. van den Dungen1, Erica Sierra-de Koning2, RenĂŠ van Es1, Arjan H. Schoneveld1, Pieter A. Doevendans1, Nicolaas de Jonge1, Edward E.S. Nieuwenhuis3, Hester M. den Ruijter1, Gerard Pasterkamp1, Roel A. de Weger2, Aryan Vink2, Michal Mokry3*, Folkert W. Asselbergs1,4,5* *equal contribution

Department of Cardiology, University Medical Center Utrecht, Utrecht, The Netherlands

Department of Pathology, University Medical Center Utrecht, Utrecht, The Netherlands

Laboratory of translational immunology, Wilhelmina Childrenâ&#x20AC;&#x2122;s Hospital, Utrecht, The Netherlands

ICIN-Netherlands Heart Institute, Utrecht, The Netherlands

Institute of Cardiovascular Science, Faculty of Population Health Sciences, University College London, London, United Kingdom

PART THREE CHAPTER 8

ABSTRACT Background The phospholamban (PLN) R14del founder mutation is associated with dilated and arrhythmogenic cardiomyopathy with high mortality and morbidity in hundreds of Dutch mutation carriers. DNA occupancy of H3K27ac histone mark gives valuable information about transcriptional regulation and can be used for candidate gene or biomarker discovery. For the first time we used ChIP-seq to assay disease-specific fingerprint of H3K27ac in cardiac tissue from PLN R14del end-stage cardiomyopathy to better understand the underlying pathophysiological mechanisms. Methods and results Firstly, we catalogued DNA regulatory elements by producing H3K27ac ChIP-seq data in cardiac tissue from PLN R14del cardiomyopathy carriers (n=6) compared to healthy controls (n=4). Increased H3K27ac occupancy was detected within a 5kb window near genes involved in fibrosis, developmental programs and chromatin assembly, while decreased H3K27ac occupancy was linked to genes involved in mitochondrial function and beta-oxidation of fatty acids. Secondly, we performed tissue specificity analysis of differentially regulated genes in PLN R14del cardiomyopathy using RNA data from various tissues. We observed that expression of all genes with differential H3K27ac occupancy was enriched in fetal and adult heart and adult skeletal muscle. Thirdly, in order to exclude general pathways linked to end-stage heart failure, we included ChIP-seq data from patients with other genetic types of non-ischemic cardiomyopathy (n=6) and ischemic cardiomyopathy (n=4). K-mean cluster analysis revealed 208 genes annotated to PLN R14del-specific regulatory regions. Among these, 55 genes are involved in (phospho) lipid and lipoprotein synthesis and metabolism and can be considered for candidate genes explaining the lipid storage evident in PLN R14del hearts. Conclusion Using an integrative chromatin analysis we identified the major effector genes and pathways involved in PLN R14del cardiomyopathy. Ubiquitously expressed genes with increased H3K27ac occupancy in patients are in line with general fibrotic pathways and developmental programs in end-stage organ failure. The suppression of mitochondrial lipid metabolism needed for muscle function is underlined by cardiac- and skeletal-muscle specific genes with increased H3K27ac occupancy. The detection of 208 PLN R14del-specific genes might provide valuable clues to recognize carriers at a high risk of developing cardiomyopathy using a plasma detectable biomarker.Â

128

CHROMATIN REGULATION IN CARDIOMYOPATHIES

INTRODUCTION Heart failure (HF) is a major medical problem in the western world, with a high morbidity and mortality risk.1 Cardiomyopathies are a leading cause of HF and numerous disease causing genes have been found.2, 3 The phospholamban (PLN) R14del mutation has been associated with dilated and arrhythmogenic right ventricular cardiomyopathies and is often accompanied with life threatening ventricular arrhythmias.4, 5 Phospholamban is a regulator of the sarcoplasmic reticulum Ca2+ ATPase (SERCA2a) pump, important for maintaining Ca2+ homeostasis and crucial for cardiac contractility. Within the Netherlands a relatively large population of patients exist that have the PLN R14del mutation.6 Previous experimental research showed possible superinhibition by mutant PLN leading to disturbed Ca2+ cycling.4 However, the precise pathophysiological mechanism in PLN R14del mutation carriers leading to cardiac fibrosis and HF remains unknown and age of onset and symptom severity differs greatly between patients. Previously, in a human histopathological study of PLN R14del cardiomyopathy, we found that myocardial fibrosis was predominantly located in the left posterolateral wall and fatty changes were more pronounced in the right ventricular wall.7 Induced pluripotent stem cells (iPSCs) derived cardiomyocytes from a patient with the R14del mutation showed Ca2+ handling abnormalities, abnormal cytoplasmic distribution of PLN protein and increased expression of cardiac hypertrophy markers (ANF, BNP) and a decreased MYH6/MYH7 ratio.8 Recently, next-generation sequencing (NGS) technologies have emerged to profile the DNA regulatory elements controlling the timing, localization and the level of gene transcription.9 Using a hypothesis free genome-wide approach associations with disease can be found by profiling histone marks related to active promoters and enhancers. We hypothesize that changes of the profiles of these epigenetic programs can serve as an indicator of disease pathogenesis.10 The

aim of this project is to profile the active chromatin histone mark H3K27ac assayed by chromatin immunoprecipitation and sequencing (ChIP-seq) to detect the regulome of human cardiac tissue from PLN R14del cardiomyopathy patients and healthy controls. To exclude chromatin activity related to end-stage heart failure and cardiomyopathies in general, we included also two distinct types of cardiomyopathies, such as ischemic cardiomyopathy and non-ischemic cardiomyopathy based on mutations in genes encoding the proteins of the sarcomere. Identification of differentially active regions in cardiac disease has the potential to lead to a more fundamental understanding of the underlying pathophysiological mechanisms related to HF and identify new treatment targets and diagnostic biomarkers.

METHODS Study design and samples This study was approved by the Biobank Research Ethics Committee, University Medical Center Utrecht, Utrecht, the Netherlands (protocol number 12/387). Written informed consent was obtained or in certain cases waived by the ethics committee when obtaining informed consent was not possible due to death of the subject. Heart samples collected at autopsy or explantation

129

PART THREE CHAPTER 8

were obtained from patients with the PLN R14del mutation (n=6). Control hearts were used as a reference (n=4). To further elaborate on R14del-specific changes hearts obtained from patients with ischemic cardiomyopathy (n=4) and from non-ischemic cardiomyopathy based on mutations in genes encoding the protein of the sarcomere (n=6) were also included. Samples used for ChIP-seq were obtained from the posterolateral part of the left ventricle or from the septum in a region halfway between the atrioventricular valves and the apex, and were stored at -80°C (Table 1, Table S1). Histopathological characterization For each individual, an adjacent part of tissue block was paraffin-embedded and stained with Masson’s trichrome. In case sections from whole heart transversal slices were available, we used high resolution systematic digital histological quantification of fibrosis and fatty tissue in trichrome stained slides to create schematic overviews showing mean fibrosis or adipose tissue per group.7 The digital histological images from the paraffin-embedded slides stained with Masson’s trichrome of adjacent regions used for ChIP-seq can be seen in Figure S1. The results of the mean percentage of fibrosis or adipose tissue in whole heart slices per group are shown in schematic overviews (Figure S2). Chromatin H3K27ac immunoprecipitation and sequencing For each sample, 10 x 10 µm thick frozen cardiac tissue slices were used for chromatin isolation using the MAGnify™ Chromatin Immunoprecipitation System kit (Life Technologies) according to manufacturer’s instructions. In short, the tissue was crosslinked with 1% formaldehyde and the crosslinking was stopped by adding 1.25 M glycine. Cells were lysed using the kit-provided lysis buffer and nuclei were sonicated using Covaris microTUBE (duty cycle 5%, intensity 2, 200 cycles per burst, 60s cycle time, 10 cycles). Sheared chromatin was diluted based on the expected number of isolated cells and was incubated with antihistone H3 acetyl K27 antibody (H3K27ac; ab4729, Abcam) pre-coupled to magnetic beads for 2 hours at 4°C. Beads were extensively washed and crosslinking was reversed by the kit-provided reverse crosslinking buffer and 20 mg/mL Proteinase K. DNA was purified using ChIP DNA Clean & Concentrator kit (Zymo Research). Isolated DNA was additionally sheared, end-repaired, sequencing adaptors were ligated and the library was amplified by PCR using primers with sample-specific barcodes according to our modification to manufacturer’s recommendations.11 After PCR, the library was purified and checked for the proper size range and for the absence of adaptor dimers on a 2% agarose gel and sequenced on SOLiD Wildfire sequencer. ChIP-seq data processing Sequencing reads were mapped against the reference genome (hg19 assembly, NCBI37) using BWA package12 (-c, -l 25, -k 2, -n 10). Multiple reads mapping to same location and strand have been collapsed to single read and only uniquely placed reads were used for peak-calling. Peaks were called using Cisgenome 2.013 ( -e 150 -maxgap 200 -minlen 200).

130

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Table 1. Detailed clinical overview of the included individuals PLN R14del (n=6)

Sarcomeric (n=6)

ICM (n=4)

Controls (n=4)

Age at explant, mean±SD

43±7.6*

35±4.5

56±7.9

unknown

Female gender, count [%]

2 (33.3%)

4 (66.7%)

1 (25%)

unknown

Cardiac device, count [%] - ICD

6 (100%)

4 (66.7%)

2 (50%)

unknown

- CRT-D

2 (33.3%)

1 (16.7%)

2 (50%)

unknown

- LVAD

3 (50%)

2 (33.3%)

1 (25%)

unknown

Last known EF, mean

21%#

20%

unknown

Heart weight [g], mean±SD

403±76.6*

349±135.6

470±165.9§

339±112.5§

*n=5; #n=4; §n=3. Abbreviations: PLN = phospholamban, ICM = ischaemic cardiomyopathy, SD = standard deviation, ICD = Implantable Cardioverter Defibrillator, CRT-D = Cardiac Resynchronization Therapy Defibrillator, LVAD = left ventricular assist device, EF = ejection fraction.

Identification of regions with differential H3K27ac occupancy (Analysis 1) Peak coordinates from all PLN and Control samples were stretched to at least 2000 base pairs and collapsed into a single common list. Overlapping peaks were merged based on their outmost coordinates. Only peaks supported by at least 2 independent datasets were further analyzed. Sequencing reads from each ChIP-seq sample were overlapped back with the common peak list to set the H3K27ac occupancy for every peak-sample pair. Sex chromosomes were excluded from analysis. Peaks with differential H3K27ac occupancy between PLN samples and controls were identified using DESeq2 using the standard settings14 (p < 0.05). Supervised hierarchical clustering was performed with quantile normalized (limma::normalizeQiantiles() function in R),

log2 transformed and median centered read counts per common peak. To avoid log2 transformation of zero values, one read was added to each peak. Gene annotation was performed for peaks with the peak summit located within 5kb from the transcription start site (TSS) of a gene. ToppGene Suite tool ToppFun was used for gene list enrichment analysis and candidate gene prioritization based on functional annotations and protein interactions network.15 The list of hyper- and hypoacetylated genes was tested separately using Bonferroni correction, p-value cut-off of 0.05 and gene limit of 1-2,000 genes per pathway. Tissue specificity of annotated genes (Analysis 2) To assess the tissue specificity of collected genes, publically available RNA sequencing (RNAseq) data were downloaded from two sources. Expression values represented by reads per kilobase per million (RPKM), were obtained from the RNA-Seq Atlas, a reference database for gene expression profiling in eleven healthy human tissue types (adipose, colon, heart, hypothalamus, kidney, liver, lung, ovary, skeletal muscle, spleen and testes) pooled from multiple donors.16 In addition, raw RNA-seq data from 91 and 120 days old fetal hearts were downloaded from The NIH Roadmap Epigenomics Mapping Consortium under accession numbers GSM1059495 and GSM1059494, respectively. Data were subsequently mapped to the human reference genome GRCh37/hg19 using BWA mapper and RPKM values were calculated. If genes

131

PART THREE CHAPTER 8

had more than one annotated transcript, RPKM values were averaged out. RPKM values from both sources (adult and fetal tissues) were quantile normalized, log2 transformed and mediancentered per gene. A small number (0.1) was added to each RPKM to avoid logarithmic transformation of zero values. Relative expression values higher than 4.5 and lower than -4.5 were set to 4.5 and -4.5, respectively. k-mean clustering (Analysis 3) Sequencing reads from each ChIP-seq sample (PLN (n=6), controls (n=4), ischemic (n=4) and Sarcomeric (n=6)) were overlapped back with the differentially acetylated peak list to set the H3K27ac occupancy for every peak-sample pair. Raw read counts were quantile normalized (limma::normalizeQiantiles() function in R), log2 transformed and median centered (to avoid log2 transformation of zero values, one read was added to each peak). Median value from each sample group was used to construct an n x k table where n = 4 (one value per each sample type) and k represent the number of differentially acetylated regions. The k-means (nstart = 200) function in R was user to partition the peaks into 12 different clusters. To enable the reproducibility of identified clusters set.seed(10) R command was called before the clustering. Clusters with PLN R14del-specific patters were used for gene functional annotations using ToppGene Suite (ToppFun) and GeneMANIA prediction server.17 The list of hyper- and hypoacetylated genes was tested separately using Bonferroni correction, a p-value cut-off of 0.05 and gene limit of 1-2,000 genes per pathway (ToppFun) and default automatically selected weighting settings (GeneMANIA).

RESULTS Regions with differential H3K27ac occupancy in PLN group versus control group (Analysis 1) On average, we identified 28,149±9,538 and 25,721±8,460 peaks in the PLN and the control group, respectively (Table S2). When comparing both groups 2107 peaks were differentially regulated (Figure 1 and Figure S3). Out of those, 958 peaks were hyperacetylated in the PLN group and 225 genes were annotated in their proximity including 212 protein coding, 1 antisense RNA, 3 miRNA, 2 lncRNA and 5 snoRNA genes, and 2 pseudogenes (Table S3). Gene enrichment analysis resulted in several enriched GO biological processes related to fibrosis, (cardiac) development and chromatin assembly. Significantly enriched KEGG and REACTOME pathways as well as the GO processes related to chromatin assembly were mostly based on the histone H1 cluster on chromosome 6 based on a single significantly hyperacetylated peak. Two GO cellular components were overrepresented: extracellular matrix and actin cytoskeleton (Table S4). Hyperacetylated H3K27ac peak was detected within 5kb from TSS of the PDLIM3 gene previously linked to dilated cardiomyopathy.18 Hypoacetylation in the PLN group was detected in 1149 peaks and annotated to 482 genes including 454 protein coding, 3 antisense RNA, 6 miRNA, and 6 lncRNA genes, and 2 intronic transcripts (Table S5). Gene enrichment analysis resulted in two GO biological processes related to acyltransferase activity. GO cellular components showed enrichment related to the mitochondrial matrix. Three pathways were significantly enriched: isoleucine degradation

132

CHROMATIN REGULATION IN CARDIOMYOPATHIES

(BIOCYC), metabolism (REACTOME) and fatty acid beta oxidation (WikiPathways). Hypoacetylated H3K27ac peaks in the PLN group were detected within 5kb from TSS of CDKN1C, CHKB, DES, HADHA, HADHB, KCNQ1, MLYCD, MYH7, MYL3, NDUFV1, NEU1, PIGT, PNPLA2, SLC25A20 (Human Phenotype Ontology, Cardiomyopathy, HP:0001638). RNA specificity of genes located near peaks with differential acetylation (Analysis 2) Secondly, we performed tissue specificity analysis of differentially regulated genes in PLN R14del cardiomyopathy (Table S3 and Table S5) using RNA data from various tissues. While there was no significant tissue specificity observed in the group of genes with increased H3K27ac occupancy (Figure 2A), genes with decreased H3K27ac occupancy were more expressed in cardiac and skeletal muscle, including fetal heart (Figure 2B). Genes with increased H3K27ac occupancy have on average higher expression than the mean expression of all genes expressed in human adult heart while this was not the case for the hypoacetylated genes. PLN-specific fingerprint of chromatin regulation (Analysis 3) In order to exclude general pathways linked to end-stage heart failure, we included ChIP-seq data from patients with other genetic types of non-ischemic cardiomyopathy (n=6) and ischemic cardiomyopathy (n=4). K-mean analysis revealed PLN R14del-specific clusters of genes, among others, with PLN R14del-specific up- and down acetylation or clusters of peaks linked to cardiomyopathies in general (Figure S4). Next, we focused on the main clusters showing a PLN

Figure 1. Supervised clustering of differentially acetylated regions in PLN patients and healthy controls. The clustering is based on the mean of H3K27ac signal over differentially acetylated regions (p<0.05) in the heart.

133

PART THREE CHAPTER 8

R14del-specific pattern. We have identified two clusters with hyperacetylation (cluster 1 and 12) including 50 genes and two clusters with hypoacetylation (cluster 3 and 4) resulting in 158 genes in PLN R14del when compared to the other groups (Figure 3, Table 2). Interestingly, 13 genes from the hyperacetylated group and 55 genes from the hypoacetylated group are known to be involved in various pathways related to (phospho)lipid and lipoprotein synthesis and metabolism (Table 2).

Figure 2. Box plot representing tissue specificity of detected genes. Public RNA sequencing data were used to assess the tissue specificity of genes located in the proximity of hyperacetylated peaks (A) and hypoacetylated peaks (B). Only the regions differentially acetylated between healthy controls and PLN patients (p<0.05) are used in this analysis. The first column indicates the mean expression of all genes expressed in adult cardiac tissue. Columns 2-13 indicate the selection of genes based on this study. Y-axis represents log2 RPKM (reads per kilobase per million).

DISCUSSION This is the first study using ChIP-seq to assay disease-specific fingerprint of H3K27ac in cardiac tissue from PLN R14del end-stage cardiomyopathy to better understand the underlying pathophysiological mechanisms. We found differentially regulated gene regions in PLN R14del hearts mainly involved in fibrosis, (cardiac) development and chromatin assembly and PLN R14del cardiomyopathy specific pathways related to lipid metabolism. Epigenetic modification (DNA methylation and histone modifications) can make regulatory elements of genes more or less permissive to transcription factors.19 Several noncoding RNAs (e.g., microRNA and long noncoding RNA) are involved in the regulation of gene expression integrated in a network with epigenetic modifications and disregulation has been associated

134

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Figure 3. Comparison PLN R14del vs. other types of cardiomyopathies. K-mean clustering of PLN R14del cardiomyopathy (1-6), healthy controls (7-10), ischemic cardiomyopathy (11-14) and sarcomeric non-ischemic cardiomyopathy (15-20) based on H3K27ac signal was user to partition the peaks into 12 different clusters. Selected were clusters considered to be specifically upregulated in PLN R14del (A) and downregulated in PLN R14del (B) as compared to other conditions. (n=number of detected peaks)

with cardiac failure. In a previous study from Yang et al. RNA sequencing detected differential expression in heart failure and long noncoding RNA (lncRNA) could distinguish ischemic cardiomyopathy from non-ischemic cardiomyopathy.20 Furthermore, pathological expression of lncNRAs improved in response to LVAD support. Herrer et al. showed altered cytoskeletal processes in human tissue RNA-Seq data of ischemic and dilated cardiomyopathy. They also found changes in MYLK4, RHOU, and ANKRD1 with qRT-PCR.21 The transcriptional activator p300 has been associated with hypertrophy and heart failure through transcription by myocyte enhancer factor-2 and GATA-4.22, 23 In addition, several stress related pathways have been implicated in DCM, including biomechanical, cytoskeletal, myocyte survival, apoptosis and sarcoplasmic reticulum calcium cycling pathways.24

135

PART THREE CHAPTER 8

Table 2. Genes specific for PLN R14del cardiomyopathy Hyperacetylated genes (cluster 1,12)

136

Gene symbol Gene name

ABL1

ABL proto-oncogene 1, non-receptor tyrosine kinase

ITPRIPL2

inositol 1,4,5-trisphosphate receptor interacting protein-like 2

ARHGEF28

Rho guanine nucleotide exchange factor (GEF) 28

KCNE4

potassium channel, voltage gated subfamily E regulatory beta subunit 4

BFSP1

beaded filament structural protein 1, filensin

L3MBTL3

l(3)mbt-like 3 (Drosophila)

CCDC80

coiled-coil domain containing 80

LUM

lumican

CMKLR1

chemerin chemokine-like receptor 1

LY96

lymphocyte antigen 96

CNN3

calponin 3, acidic

MAML2

mastermind-like 2 (Drosophila)

COL14A1

collagen, type XIV, alpha 1

MMP2

matrix metallopeptidase 2

COL6A3

collagen, type VI, alpha 3

MSMO1

methylsterol monooxygenase 1

COL8A1

collagen, type VIII, alpha 1

NNMT

nicotinamide N-methyltransferase

CREB3L2

cAMP responsive element binding protein 3-like 2

NOD1

nucleotide-binding oligomerization domain containing 1

CYP1B1

cytochrome P450, family 1, subfamily B, polypeptide 1

NOX4

NADPH oxidase 4

EDNRA

endothelin receptor type A

NT5E

5â&#x20AC;&#x2122;-nucleotidase, ecto (CD73)

EMP1

epithelial membrane protein 1

PDGFD

platelet derived growth factor D

EPSTI1

epithelial stromal interaction 1 (breast)

PECAM1

platelet/endothelial cell adhesion molecule 1

ESR1

estrogen receptor 1

PPAP2A

phosphatidic acid phosphatase type 2A

FIBIN

fin bud initiation factor homolog (zebrafish)

PRICKLE2

prickle homolog 2 (Drosophila)

FLRT2

fibronectin leucine rich transmembrane protein 2

RNF138P1

ring finger protein 138, E3 ubiquitin protein ligase pseudogene 1

GLIPR1

GLI pathogenesis-related 1

SEC22B

SEC22 vesicle trafficking protein homolog B (S. cerevisiae) (gene/ pseudogene)

HIST1H1B

histone cluster 1, H1b

SLCO3A1

solute carrier organic anion transporter family, member 3A1

HIST1H2AM

histone cluster 1, H2am

SNORA74B

small nucleolar RNA, H/ACA box 74B

HIST1H2BO

histone cluster 1, H2bo

SNTB1

syntrophin, beta 1 (dystrophinassociated protein A1, 59kDa, basic component 1)

HIST1H3I

histone cluster 1, H3i

SPRY4

sprouty homolog 4 (Drosophila)

HIST1H3J

histone cluster 1, H3j

UACA

uveal autoantigen with coiled-coil domains and ankyrin repeats

HIST1H4L

histone cluster 1, H4l

WIPF1

WAS/WASL interacting protein family, member 1

IFFO2

intermediate filament family orphan 2

ZNF217

zinc finger protein 217

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Table 2. Continued Hypoacetylated genes (cluster 3,4) Gene symbol Gene name

Gene symbol Gene name

ABCB6

ATP-binding cassette, sub-family B (MDR/TAP), member 6 (Langereis blood group)

MADD

MAP-kinase activating death domain

ABHD6

abhydrolase domain containing 6

MAP4K1

mitogen-activated protein kinase kinase kinase kinase 1

ACBD4

acyl-CoA binding domain containing 4

MAPK7

mitogen-activated protein kinase 7

ADSSL1

adenylosuccinate synthase like 1

myoglobin

AGPAT2

1-acylglycerol-3-phosphate O-acyltransferase 2

MFN2

mitofusin 2

ALS2CL

ALS2 C-terminal like

MFSD3

major facilitator superfamily domain containing 3

ANGPTL4

angiopoietin-like 4

MIR486-1

microRNA 486-1

ANK1

ankyrin 1, erythrocytic

MLPH

melanophilin

AP5S1

adaptor-related protein complex 5, sigma 1 subunit

MLYCD

malonyl-CoA decarboxylase

AQP1

aquaporin 1 (Colton blood group)

MRPL21

mitochondrial ribosomal protein L21

AQP7

aquaporin 7

MYEOV

myeloma overexpressed

ARAP1

ArfGAP with RhoGAP domain, ankyrin repeat and PH domain 1

MYH7

myosin, heavy chain 7, cardiac muscle, beta

ARHGAP23

Rho GTPase activating protein 23

MYL3

myosin, light chain 3, alkali; ventricular, skeletal, slow

ARHGAP27

Rho GTPase activating protein 27

MYL7

myosin, light chain 7, regulatory

ASAP3

ArfGAP with SH3 domain, ankyrin repeat and PH domain MYLK3 3

myosin light chain kinase 3

ASB18

ankyrin repeat and SOCS box containing 18

MYLPF

myosin light chain, phosphorylatable, fast skeletal muscle

ASS1

argininosuccinate synthase 1

MZT2A

mitotic spindle organizing protein 2A

ATG2A

autophagy related 2A

NCBP2

nuclear cap binding protein subunit 2, 20kDa

ATP6V1F

ATPase, H+ transporting, lysosomal 14kDa, V1 subunit F NEU1

sialidase 1 (lysosomal sialidase)

BAHD1

bromo adjacent homology domain containing 1

NMRK2

nicotinamide riboside kinase 2

BAP1

BRCA1 associated protein-1 (ubiquitin carboxy-terminal hydrolase)

NUDT8

nudix (nucleoside diphosphate linked moiety X)-type motif 8

BCL7C

B-cell CLL/lymphoma 7C

P2RY2

purinergic receptor P2Y, G-protein coupled, 2

BRICD5

BRICHOS domain containing 5

PACSIN3

protein kinase C and casein kinase substrate in neurons 3

C16orf86

chromosome 16 open reading frame 86

PBXIP1

pre-B-cell leukemia homeobox interacting protein 1

C1orf228

chromosome 1 open reading frame 228

PCBP4

poly(rC) binding protein 4

C1orf86

chromosome 1 open reading frame 86

PDCD6

programmed cell death 6

CABLES2

Cdk5 and Abl enzyme substrate 2

PDE4A

phosphodiesterase 4A, cAMP-specific

CASP2

caspase 2, apoptosis-related cysteine peptidase

PDK2

pyruvate dehydrogenase kinase, isozyme 2

CCDC57

coiled-coil domain containing 57

PGP

phosphoglycolate phosphatase

CDK18

cyclin-dependent kinase 18

PHF7

PHD finger protein 7

137

PART THREE CHAPTER 8

Table 2. Continued

138

CHAC1

ChaC glutathione-specific gammaglutamylcyclotransferase 1

PIGT

phosphatidylinositol glycan anchor biosynthesis, class T

CLIP2

CAP-GLY domain containing linker protein 2

PITPNM2

phosphatidylinositol transfer protein, membrane-associated 2

CLRN2

clarin 2

PLCD3

phospholipase C, delta 3

CLUH

clustered mitochondria (cluA/CLU1) homolog

PPAPDC3

phosphatidic acid phosphatase type 2 domain containing 3

CLUHP3

clustered mitochondria (cluA/CLU1) homolog pseudogene 3

PRR26

proline rich 26

CNIH4

cornichon family AMPA receptor auxiliary protein 4

PSME2

proteasome (prosome, macropain) activator subunit 2 (PA28 beta)

CRIP3

cysteine-rich protein 3

PTDSS2

phosphatidylserine synthase 2

CTF1

cardiotrophin 1

PTGDS

prostaglandin D2 synthase 21kDa (brain)

CTNNBIP1

catenin, beta interacting protein 1

PWWP2B

PWWP domain containing 2B

DES

desmin

PYROXD2

pyridine nucleotide-disulphide oxidoreductase domain 2

DGKZ

diacylglycerol kinase, zeta

QARS

glutaminyl-tRNA synthetase

DIRAS1

DIRAS family, GTP-binding RAS-like 1

QDPR

quinoid dihydropteridine reductase

DNAAF1

dynein, axonemal, assembly factor 1

RANBP1

RAN binding protein 1

DNHD1

dynein heavy chain domain 1

RGS3

regulator of G-protein signaling 3

DYSF

dysferlin

RNF31

ring finger protein 31

EIF3K

eukaryotic translation initiation factor 3, subunit K

SELENBP1

selenium binding protein 1

ELP6

elongator acetyltransferase complex subunit 6

SGSM2

small G protein signaling modulator 2

ENKD1

enkurin domain containing 1

SLC25A20

solute carrier family 25 (carnitine/ acylcarnitine translocase), member 20

FAHD1

fumarylacetoacetate hydrolase domain containing 1

SLC25A34

solute carrier family 25, member 34

FAM110D

family with sequence similarity 110, member D

SLC4A3

solute carrier family 4 (anion exchanger), member 3

FAM65A

family with sequence similarity 65, member A

SLCO2B1

solute carrier organic anion transporter family, member 2B1

FDXR

ferredoxin reductase

SPAG7

sperm associated antigen 7

FPGS

folylpolyglutamate synthase

SPAG8

sperm associated antigen 8

GBAS

glioblastoma amplified sequence

SRF

serum response factor (c-fos serum response element-binding transcription factor)

GPSM1

G-protein signaling modulator 1

STK11

serine/threonine kinase 11

GPT

glutamic-pyruvate transaminase (alanine aminotransferase)

STXBP6

syntaxin binding protein 6 (amisyn)

GTF3C2

general transcription factor IIIC, polypeptide 2, beta 110kDa

SYNPO2L

synaptopodin 2-like

HADHA

hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA TBC1D10B thiolase/enoyl-CoA hydratase (trifunctional protein), alpha subunit

TBC1 domain family, member 10B

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Table 2. Continued HADHB

hydroxyacyl-CoA dehydrogenase/3-ketoacyl-CoA thiolase/enoyl-CoA hydratase (trifunctional protein), beta subunit

TINAGL1

tubulointerstitial nephritis antigen-like 1

HAGH

hydroxyacylglutathione hydrolase

TMEM120B

transmembrane protein 120B

HES4

hes family bHLH transcription factor 4

TMEM139

transmembrane protein 139

HINT2

histidine triad nucleotide binding protein 2

TMEM154

transmembrane protein 154

HRH2

histamine receptor H2

TMEM200B

transmembrane protein 200B

HSD17B1

hydroxysteroid (17-beta) dehydrogenase 1

TMEM229B

transmembrane protein 229B

HSDL1

hydroxysteroid dehydrogenase like 1

TMEM53

transmembrane protein 53

IGHMBP2

immunoglobulin mu binding protein 2

TMEM82

transmembrane protein 82

ILVBL

ilvB (bacterial acetolactate synthase)-like

TNFAIP8L1

tumor necrosis factor, alpha-induced protein 8-like 1

INPP5J

inositol polyphosphate-5-phosphatase J

TP73-AS1

TP73 antisense RNA 1

IPPK

inositol 1,3,4,5,6-pentakisphosphate 2-kinase

TRMT2A

tRNA methyltransferase 2 homolog A (S. cerevisiae)

KCNJ4

potassium channel, inwardly rectifying subfamily J, member 4

TSR1

TSR1, 20S rRNA accumulation, homolog (S. cerevisiae)

KLHDC8B

kelch domain containing 8B

TTYH3

tweety family member 3

KLHL42

kelch-like family member 42

TUB

tubby bipartite transcription factor

LAMB2

laminin, beta 2 (laminin S)

UQCRC1

ubiquinol-cytochrome c reductase core protein I

LCNL1

lipocalin-like 1

USP19

ubiquitin specific peptidase 19

LIN9

lin-9 DREAM MuvB core complex component

VWA7

von Willebrand factor A domain containing 7

LOC150776

sphingomyelin phosphodiesterase 4, neutral membrane (neutral sphingomyelinase-3) pseudogene

ZNF629

zinc finger protein 629

LOC645166

lymphocyte-specific protein 1 pseudogene

ZNF747

zinc finger protein 747

LPAR5

lysophosphatidic acid receptor 5

ZNF8

zinc finger protein 8

LRP5L

low density lipoprotein receptor-related protein 5-like

ZNRF3

zinc and ring finger 3

Bold letters indicate genes involved in pathways/networks related to lipid metabolism (GO:molecular function, GO:biological process, GO:cellular component, Human phenotype, Mouse phenotype, Pathway as based on the ToppGene Suite).

139

PART THREE CHAPTER 8

We detected several PLN R14del-specific genes with differential chromatin regulation. The KCNE4 gene encodes the potassium channel, voltage gated subfamily E regulatory beta subunit 4.25 Potassium channels are crucial for generation and propagation of action potentials in the heart and related to heart rate and arrhythmias. KCNE4 knockout mice showed significantly dilated left ventricular internal diameters and reduced fractional shortening and ejection fraction.26 COL14A1 encodes the alpha chain of collagen involved in fibrillogenesis27 and RNA-Seq analysis has previously identified differential expression of this gene in tissue samples of patients with ischemic cardiomyopathy and controls pointing to involvement of collagen-related genes in remodelling.28 MMP2 (matrix metalloprotease 2) is also related to collagen turnover and has been proposed as a biomarker to predict prognosis of patients with HF.29 The identified PLN R14del specific pathways related to lipid metabolism could be studied in future for candidate genes related to the characteristic fibrofatty replacement in PLN R14del hearts and arrhythmogenic cardiomyopathy.7, 30 The criteria for annotation of differentially acetylated peaks to potentially relevant genes have consequences on data interpretation. In this study we used a stringent 5 kb window to decrease the chance of false positive gene annotations. Interestingly, a hypoacetylated peak 5,036 bp from a TSS of MIR21 gene, a crucial player upregulated in cardiac fibrosis31, was identified in the PLN group. Given the orientation of the surrounding genes VMP1, TUBD1, and MIR21, the detected peak at chr17:57,922,150-57,925,229 is located downstream of the three genes and unlikely has a promoter function. The peak was detected as 223/958 sorted by significance (p=0.0038) and taking together the unknown role and low expression of the other two genes in cardiac tissue it indicates a possible role of an enhancer region of MIR21. Therefore we enlarged the window used for gene annotation from 5kb to 20kb from TSS and increased the number of resulting annotated genes from 225 to 416. When performing pathway analysis using ToppFun we increase the number of genes in biological processes, such as GO:0030198 (extracellular matrix organization) from 18 to 32 genes, in GO:0007507 (heart development) from 19 to 32 genes, and GO:0015629 (actin cytoskeleton) from 15 to 24 genes. This indicates that a right window size needs to be taken into account for in silico annotation of potentially relevant genes. However, a deeper functional annotation, such as using a chromatin conformation capture method32, is needed to prove the functional relationship between the differentially acetylated region and its regulated gene. Taken together, using an integrative chromatin analysis we identified the major effector genes and pathways involved in PLN R14del cardiomyopathy. Ubiquitously expressed genes with increased H3K27ac occupancy in patients are in line with general fibrotic pathways and developmental programs in end-stage organ failure. The suppression of mitochondrial lipid metabolism needed for muscle function is underlined by cardiac and skeletal muscle-specific genes with increased H3K27ac occupancy. The detection of PLN R14del-specific genes might provide valuable clues to recognize carriers at a high risk of developing cardiomyopathy using a plasma detectable biomarker.Â

140

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Acknowledgments This work was supported by “Stichting Genetische Hartspierziekte PLN” (http://stichtingpln.nl, Middenmeer, the Netherlands). Magdalena Harakalova is supported by NIH Ro1 grant LM010098 and the OZF/14 WKZ-funds. Michal Mokry is supported by OZF/2012 WKZ-fund. Folkert W. Asselbergs is supported by a Dekker scholarship-Junior Staff Member 2014T001 – Netherlands Heart Foundation and UCL Hospitals NIHR Biomedical Research Centre.

141

PART THREE CHAPTER 8

REFERENCES 1. 2.

3. 4.

9. 10. 11.

12. 13. 14. 15. 16.

17.

18.

19. 20.

21.

142

Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137-46. Arbustini E, Narula N, Dec GW, Reddy KS, Greenberg B, Kushwaha S, et al. The MOGE(S) classification for a phenotype-genotype nomenclature of cardiomyopathy: endorsed by the World Heart Federation. J Am Coll Cardiol. 2013;62:2046-72. Jacoby D, McKenna WJ. Genetics of inherited cardiomyopathy. Eur Heart J. 2012;33:296-304. Haghighi K, Kolokathis F, Gramolini AO, Waggoner JR, Pater L, Lynch RA, et al. A mutation in the human phospholamban gene, deleting arginine 14, results in lethal, hereditary cardiomyopathy. Proc Natl Acad Sci U S A. 2006;103:1388-93. van Rijsingen IA, van der Zwaag PA, Groeneweg JA, Nannenberg EA, Jongbloed JD, Zwinderman AH, et al. Outcome in phospholamban R14del carriers: results of a large multicentre cohort study. Circ Cardiovasc Genet. 2014;7:455-65. van der Zwaag PA, van Rijsingen IA, Asimaki A, Jongbloed JD, van Veldhuisen DJ, Wiesfeld AC, et al. Phospholamban R14del mutation in patients diagnosed with dilated cardiomyopathy or arrhythmogenic right ventricular cardiomyopathy: evidence supporting the concept of arrhythmogenic cardiomyopathy. Eur J Heart Fail. 2012;14:1199-207. Gho JM, van Es R, Stathonikos N, Harakalova M, te Rijdt WP, Suurmeijer AJ, et al. High resolution systematic digital histological quantification of cardiac fibrosis and adipose tissue in phospholamban p.Arg14del mutation associated cardiomyopathy. PLoS ONE. 2014;9:e94820. Karakikes I, Stillitano F, Nonnenmacher M, Tzimas C, Sanoudou D, Termglinchan V, et al. Correction of human phospholamban R14del mutation associated with cardiomyopathy using targeted nucleases and combination therapy. Nat Commun. 2015;6:6955. Levo M, Segal E. In pursuit of design principles of regulatory sequences. Nat Rev Genet. 2014;15:453-68. Kathiriya IS, Nora EP, Bruneau BG. Investigating the transcriptional control of cardiovascular development. Circ Res. 2015;116:700-14. Harakalova M, Mokry M, Hrdlickova B, Renkens I, Duran K, van Roekel H, et al. Multiplexed array-based and in-solution genomic enrichment for flexible and cost-effective targeted nextgeneration sequencing. Nat Protoc. 2011;6:1870-86. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25:1754-60. Ji H, Jiang H, Ma W, Johnson DS, Myers RM, Wong WH. An integrated software system for analyzing ChIP-chip and ChIP-seq data. Nat Biotechnol. 2008;26:1293-300. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550. Chen J, Bardes EE, Aronow BJ, Jegga AG. ToppGene Suite for gene list enrichment analysis and candidate gene prioritization. Nucleic Acids Res. 2009;37:W305-11. Krupp M, Marquardt JU, Sahin U, Galle PR, Castle J, Teufel A. RNA-Seq Atlas--a reference database for gene expression profiling in normal tissue by next-generation sequencing. Bioinformatics. 2012;28:1184-5. Warde-Farley D, Donaldson SL, Comes O, Zuberi K, Badrawi R, Chao P, et al. The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res. 2010;38:W214-20. Harakalova M, Kummeling G, Sammani A, Linschoten M, Baas AF, van der Smagt J, et al. A systematic analysis of genetic dilated cardiomyopathy reveals numerous ubiquitously expressed and muscle-specific genes. Eur J Heart Fail. 2015;17:484-93. Greco CM, Condorelli G. Epigenetic modifications and noncoding RNAs in cardiac hypertrophy and failure. Nat Rev Cardiol. 2015. Yang KC, Yamada KA, Patel AY, Topkara VK, George I, Cheema FH, et al. Deep RNA sequencing reveals dynamic regulation of myocardial noncoding RNAs in failing human heart and remodeling with mechanical circulatory support. Circulation. 2014;129:1009-21. Herrer I, Rosello-Lleti E, Rivera M, Molina-Navarro MM, Tarazon E, Ortega A, et al. RNAsequencing analysis reveals new alterations in cardiomyocyte cytoskeletal genes in patients with heart failure. Lab Invest. 2014;94:645-53.

CHROMATIN REGULATION IN CARDIOMYOPATHIES

22. Morimoto T, Sunagawa Y, Kawamura T, Takaya T, Wada H, Nagasawa A, et al. The dietary compound curcumin inhibits p300 histone acetyltransferase activity and prevents heart failure in rats. J Clin Invest. 2008;118:868-78. 23. Yanazume T, Hasegawa K, Morimoto T, Kawamura T, Wada H, Matsumori A, et al. Cardiac p300 is involved in myocyte growth with decompensated heart failure. Mol Cell Biol. 2003;23:3593606. 24. Chien KR. Stress pathways and heart failure. Cell. 1999;98:555-8. 25. Abbott GW, Sesti F, Splawski I, Buck ME, Lehmann MH, Timothy KW, et al. MiRP1 forms IKr potassium channels with HERG and is associated with cardiac arrhythmia. Cell. 1999;97:17587. 26. Ciampa EJ. Investigating the Function of KCNE4 in Cardiac Physiology [Dissertation]. Nashville, Tennessee: Vanderbilt University; 2011. 27. Ansorge HL, Meng X, Zhang G, Veit G, Sun M, Klement JF, et al. Type XIV Collagen Regulates Fibrillogenesis: PREMATURE COLLAGEN FIBRIL GROWTH AND TISSUE DYSFUNCTION IN NULL MICE. J Biol Chem. 2009;284:8427-38. 28. Gil-Cayuela C, Rivera M, Ortega A, Tarazon E, Trivino JC, Lago F, et al. RNA sequencing analysis identifies new human collagen genes involved in cardiac remodeling. J Am Coll Cardiol. 2015;65:1265-7. 29. Sanchis L, Andrea R, Falces C, Llopis J, Morales-Ruiz M, Lopez-Sobrino T, et al. Prognosis of new-onset heart failure outpatients and collagen biomarkers. Eur J Clin Invest. 2015;45:842-9. 30. Marcus FI, McKenna WJ, Sherrill D, Basso C, Bauce B, Bluemke DA, et al. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the task force criteria. Circulation. 2010;121:1533-41. 31. Lorenzen JM, Schauerte C, Hubner A, Kolling M, Martino F, Scherf K, et al. Osteopontin is indispensible for AP1-mediated angiotensin II-related miR-21 transcription during cardiac fibrosis. Eur Heart J. 2015. 32. Stadhouders R, Kolovos P, Brouwer R, Zuin J, van den Heuvel A, Kockx C, et al. Multiplexed chromosome conformation capture sequencing for rapid genome-scale high-resolution detection of long-range chromatin interactions. Nat Protoc. 2013;8:509-24.

143

PART THREE CHAPTER 8

SUPPLEMENTARY INFORMATION

Supplementary Figure 1. Histological slides derived from the same tissue region as used for ChIP-seq. The slides were stained with Massonâ&#x20AC;&#x2122;s trichrome. A. PLN R14del mutation patients heart sections (n = 6). B. Control heart sections (n = 4). LV = left ventricle; Sept = Septum.

144

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Figure 2. Schematic overview of mean fibrosis and adipose tissue in heart slices from hearts used for ChIP-seq. The results of mean percentage of fibrosis or adipose tissue are shown using a color scale. A-B. heart slices of PLN R14del mutation patients (n = 6/6). C-D. heart slices of controls (n = 3/4). LV = left ventricle; RV = right ventricle; Post. = Posterior; Ant. = Anterior.

Supplementary Figure 3. Supervised clustering of differentially acetylated regions in PLN patients and healthy controls. The clustering is based on the mean of H3K27ac signal over differentially acetylated regions (p<0.05) in the heart.

145

Supplementary Figure 4. K-mean clustering of PLN R14del cardiomyopathy (1-6), healthy controls (7-10), ischemic cardiomyopathy (11-14) and sarcomeric non-ischemic cardiomyopathy (15-20) based on H3K27ac signal was user to partition the peaks into 12 different clusters. Selected were clusters considered to be specifically upregulated in PLN R14del (A) and downregulated in PLN R14del (B) as compared to other conditions. (n=number of detected peaks)

PART THREE CHAPTER 8

146

road accident victim

DCM

DCM with marks of hypertrophy

505 174* 257 403

35 39

266

34 36

486

450 NA

52 61

645 315*

64 47

360 NA

NA NA

217 439

NA NA

390 495

43 34

NA 385

NA 51

293 452

37 50

Yes

Sex Age at Age at Heart ICD diagnosis explant weight (years) (years) (g)

Yes

19%

13%

25%

21%

17%

20%

25%

18%

20%

31%

15%

25%

Yes

PLLV

SEPT

PLLV

SEPT

PLLV

SEPT

PLLV

CRT-D LVAD Second Significant Last Schematic Location HTx CAD known EF overview1 of sample available before explant

Schematic overview of mean percentage of fibrosis or adipose tissue from transversal heart slice; *after removal of heart slice. Abbreviations: LVAD = left ventricular assist device; ICD = implantable cardioverter-defibrillator; CRT-D = cardiac resynchronization therapy defibrillator; HTx = heart transplantation; CAD = coronary artery disease; EF = ejection fraction; PLLV = posterolateral left ventricle; SEPT = septum; RZ = remote zone to myocardiac infarction; ICM = ischemic cardiomyopathy; HCM = hypertrophic cardiomyopathy; HOCM = hypertrophic obstructive cardiomyopathy; DCM = dilated cardiomyopathy; ACM = arrhythmogenic cardiomyopathy; F = female (Sex); M = male (Sex); NA = not available.

Sarcomeric

TTN_1

TTN_2

DCM with marks of hypertrophy

HOCM in a dilated phase

Sarcomeric

MYH7_2

HCM

MYH7_1

HCM in a dilated phase

ICM

rejected donor

MYBPC3_2 Sarcomeric

Ischemic

ICM_3

ICM_4

rejected donor

MYBPC3_1 Sarcomeric

Ischemic

ICM_1

ICM_2

Control

Control_3

Control_4

Control

Control_1

Control_2

Phospholamban ACM

PLN_5

PLN_6

Phospholamban DCM

PLN_3

PLN_4

Phospholamban DCM

PLN_1

Diagnosis

PLN_2

Group

Type

Supplementary Table 1A. Detailed clinical overview of the included individuals

CHROMATIN REGULATION IN CARDIOMYOPATHIES

147

148

Phospholamban

Control

Ischemic

Sarcomeric

PLN_1

PLN_2

PLN_3

PLN_4

PLN_5

PLN_6

Control_1

Control_2

Control_3

Control_4

ICM_1

ICM_2

ICM_3

ICM_4

MYBPC3_1

MYBPC3_2

MYH7_1

MYH7_2

TTN_1

TTN_2 TTN

5xTTN

MYH7

MYBPC3

PLN

PLN/MYH6*

PLN

Mutated gene

c.32041G>A

c.5905G>T, c.82004C>G, c.10733C>T, c.53744G>A, c.87949G>A

c.704C>G

c.2827C>T

c.40_42delAGA

CDS change

p.Val10681Ile

p.Asp1969Tyr, p.Thr27335Ser, p.Pro3578Leu, p.Arg17915His, p.Ala29317Thr

p.Thr235Ser

p.Arg943X

p.Arg14del

Protein change

*variant of uncertain significance (VUS). Abbreviations: CDS = CoDing Sequence; NA = not available.

Group

Type

Supplementary Table 1B. Detailed clinical genetic overview of the included individuals

ACTC1, CRSP3, DES, GLA, LAMP2, LDB3, MYBPC3, MYH7, MYL2, NYL3, PLN, PRKAG2, SVNA, SGTC, STA, TAZ, TCHP, TNNC1, TNNI3, TNNT2, TPM1, TCL

MYH7, LMNA, PKP2, DSP, DSC2, DSG2, PLN

mutation detected in 1981

PKP2

DSC2, DSP DSG2, JUP, TMEM43, PKP2

Cardiome 1.0 - 440 genes - research

Negative DNA assay

PART THREE CHAPTER 8

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 2. ChIP-seq libraries peak number and selection for Analysis 1-3 Sample ID

Number of detected peaks

Analysis 1

Analysis 2

Analysis 3

PLN_1

40003

PLN_2

30041

PLN_3

34430

PLN_4

27167

PLN_5

25300

PLN_6

11954

Mean±SD

28,149±9,538

Control_1

26032

Control_2

18731

Control_3

39941

Control_4

18178

Mean±SD

25,721±8,460

ICM_1

12613

ICM_2

25387

ICM_3

26081

ICM_4

46978

Mean±SD

27,765±8,139

MYBPC3_1

30736

MYBPC3_2

40574

MYH7_1

30374

MYH7_2

38189

TTN_1

25761

TTN_2

41838

Mean±SD

34,579±6,429

SD = standard deviation.

149

PART THREE CHAPTER 8

Supplementary Table 3. Annotation of genes with increased H3K27ac occupancy within 5kb from transcription start site (TSS)

150

Chromosome

Peak start

Peak end

Peak length

Distance to TSS

Distance to Gene TES Symbol

Gene Name

chr9

133708430

133713699

5269

234

-51993

ABL1

ABL proto-oncogene 1, non-receptor tyrosine kinase

chr11

34376480

34378599

2119

1262

-205005

ABTB2

ankyrin repeat and BTB (POZ) domain containing 2

chr5

156997500

156999629

2129

4203

-94252

ADAM19

ADAM metallopeptidase domain 19

chr15

89164200

89166449

2249

798

-10187

AEN

apoptosis enhancing nuclease

chr2

100720930

100723029

2099

-558262

AFF3

AF4/FMR2 family, member 3

chr6

151560030

151563649

3619

706

-117852

AKAP12

A kinase (PRKA) anchor protein 12

chr3

105087030

105089199

2169

2402

-207628

ALCAM

activated leukocyte cell adhesion molecule

chr11

94499880

94504049

4169

457

-107952

AMOTL1

angiomotin like 1

chr22

38243750

38246779

3029

-4962

-18402

ANKRD54 ankyrin repeat domain 54

chr12

45608280

45612229

3949

379

-223932

ANO6

anoctamin 6

chr15

60682100

60690599

8499

3835

-46999

ANXA2

annexin A2

chr5

72920350

72923599

3249

-8

-285552

ARHGEF28

Rho guanine nucleotide exchange factor (GEF) 28

chr4

114898700

114901699

2999

678

-78760

ARSJ

arylsulfatase family, member J

chr12

54068250

54072479

4229

-256

-11420

ATP5G2

ATP synthase, H+ transporting, mitochondrial Fo complex, subunit C2 (subunit 9)

chr6

56818050

56820129

2079

-683

-73048

BEND6

BEN domain containing 6

chr20

17537400

17540699

3299

555

-64499

BFSP1

beaded filament structural protein 1, filensin

chr20

11870100

11873499

3399

323

-35442

BTBD3

BTB (POZ) domain containing 3

chr21

18982300

18985429

3129

1403

-17895

BTG3

BTG family, member 3

chr18

47017100

47019279

2179

-4589

-10160

C18orf32

chromosome 18 open reading frame 32

chr1

230993350

230996099

2749

-2943

-21859

C1orf198

chromosome 1 open reading frame 198

chr12

7167130

7169999

2869

585

-9768

C1S

complement component 1, s subcomponent

chr9

89762080

89765099

3019

-11051

C9orf170

chromosome 9 open reading frame 170

chr9

117372150

117375429

3279

-34906

C9orf91

chromosome 9 open reading frame 91

chr10

115437500

115441529

4029

-51147

CASP7

caspase 7, apoptosis-related cysteine peptidase

chr2

202121780

202124479

2699

-2093

-29304

CASP8

caspase 8, apoptosis-related cysteine peptidase

chr3

112358550

112361749

3199

-173

-36739

CCDC80

coiled-coil domain containing 80

chr19

41813200

41817729

4529

-629

-15322

CCDC97

coiled-coil domain containing 97

chr9

21993950

21996999

3049

-985

-27723

CDKN2A

cyclin-dependent kinase inhibitor 2A

chr9

21993950

21996999

3049

685

-125616

CDKN2BAS1

CDKN2B antisense RNA 1

chr10

106111880

106114349

2469

-407

-101724

CFAP58

cilia and flagella associated protein 58

chr7

89873400

89876049

2649

237

-65650

CFAP69

cilia and flagella associated protein 69

chr4

54928400

54934199

5799

-512

-55342

CHIC2

cysteine-rich hydrophobic domain 2

chr3

87275130

87278149

3019

227

-28057

CHMP2B

charged multivesicular body protein 2B

chr11

123063850

123067399

3549

382

-122590

CLMP

CXADR-like membrane protein

chr12

108730630

108733129

2499

1214

-50058

CMKLR1

chemerin chemokine-like receptor 1

chr1

95388530

95396299

7769

320

-29906

CNN3

calponin 3, acidic

chr6

75911500

75917579

6079

1083

-120497

COL12A1

collagen, type XII, alpha 1

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 3. Continued chr8

121135930

121138849

2919

-246876

COL14A1

collagen, type XIV, alpha 1

chr17

48281500

48284299

2799

-3900

-21441

COL1A1

collagen, type I, alpha 1

chr2

238319130

238323499

4369

1535

-88660

COL6A3

collagen, type VI, alpha 3

chr3

99355480

99361399

5919

986

-156718

COL8A1

collagen, type VIII, alpha 1

chr7

137675900

137688499

12599

4646

-122473

CREB3L2

cAMP responsive element binding protein 3-like 2

chr16

84853480

84859029

5549

2668

-86861

CRISPLD2 cysteine-rich secretory protein LCCL domain containing 2

chr11

118757180

118762149

4969

-4436

-7305

CXCR5

chemokine (C-X-C motif) receptor 5

chr2

38299030

38302099

3069

2758

-5819

CYP1B1

cytochrome P450, family 1, subfamily B, polypeptide 1

chr5

39420680

39425729

5049

2130

-51425

DAB2

Dab, mitogen-responsive phosphoprotein, homolog 2 (Drosophila)

chr1

85928750

85932249

3499

389

-146331

DDAH1

dimethylarginine dimethylaminohydrolase 1

chr18

67066750

67069899

3149

-447500

DOK6

docking protein 6

chr1

168695080

168697729

2649

2023

-31698

DPT

dermatopontin

chr6

116690650

116693799

3149

115

-67215

DSE

dermatan sulfate epimerase

chr18

65183480

65186129

2649

-838

-10986

DSEL

dermatan sulfate epimerase-like

chr6

56818050

56820129

2079

323

-496303

DST

dystonin

chr2

44000530

44002779

2249

473

-20657

DYNC2LI1 dynein, cytoplasmic 2, light intermediate chain 1

chr20

33102250

33105699

3449

-229

-24786

DYNLRB1 dynein, light chain, roadblock-type 1

chr8

25895730

25903479

7749

2787

-198032

EBF2

early B-cell factor 2

chr6

12291000

12293599

2599

1771

-5126

EDN1

endothelin 1

chr4

148401030

148405129

4099

1011

-63026

EDNRA

endothelin receptor type A

chr22

38243750

38246779

3029

-105

-39372

EIF3L

eukaryotic translation initiation factor 3, subunit L

chr19

18629100

18634699

5599

1037

-78425

ELL

elongation factor RNA polymerase II

chr12

13348000

13354999

6999

1898

-18207

EMP1

epithelial membrane protein 1

chr13

43564880

43568299

3419

-213

-104468

EPSTI1

epithelial stromal interaction 1 (breast)

chr6

152011080

152013599

2519

709

-412066

ESR1

estrogen receptor 1

chr5

76010400

76015449

5049

1057

-18669

F2R

coagulation factor II (thrombin) receptor

chr2

187558150

187560249

2099

411

-69310

FAM171B

family with sequence similarity 171, member B

chr10

120100350

120102399

2049

464

-32803

FAM204A

family with sequence similarity 204, member A

chr2

106011650

106017129

5479

1185

-37105

FHL2

four and a half LIM domains 2

chr11

27013400

27018629

5229

387

-2615

FIBIN

fin bud initiation factor homolog (zebrafish)

chr14

85994900

85997029

2129

-523

-98304

FLRT2

fibronectin leucine rich transmembrane protein 2

chr12

50096950

50100929

3979

2257

-67216

FMNL3

formin-like 3

chr16

86607800

86609849

2049

-3290

-6478

FOXL1

forkhead box L1

chr1

100233150

100235199

2049

-2826

-59916

FRRS1

ferric-chelate reductase 1

chr5

52774380

52779599

5219

395

-4912

FST

follistatin

chr17

9936450

9940679

4229

1499

-124639

GAS7

growth arrest-specific 7

chr15

45668850

45670949

2099

1080

-16576

GATM

glycine amidinotransferase (Larginine:glycine amidinotransferase)

chr5

154319300

154322049

2749

-2899

-53699

GEMIN5

gem (nuclear organelle) associated protein 5

chr12

75872950

75876599

3649

262

-20940

GLIPR1

GLI pathogenesis-related 1

151

PART THREE CHAPTER 8

Supplementary Table 3. Continued

152

chr19

47141050

47143379

2329

-4276

-4881

GNG8

guanine nucleotide binding protein (G protein), gamma 8

chr7

50857150

50863399

6249

884

-202515

GRB10

growth factor receptor-bound protein 10

chr19

48900500

48902499

1999

3368

-46687

GRIN2D

glutamate receptor, ionotropic, N-methyl D-aspartate 2D

chr4

54968530

54970579

2049

3307

1433

GSX2

GS homeobox 2

chr1

40096930

40107599

10669

3083

-13161

HEYL

hes-related family bHLH transcription factor with YRPW motif-like

chr6

27838500

27841899

3399

-4841

-5629

HIST1H1B histone cluster 1, H1b

chr6

27112480

27116399

3919

-468

-905

HIST1H2AH

histone cluster 1, H2ah

chr6

27803550

27808499

4949

-366

HIST1H2AK

histone cluster 1, H2ak

chr6

27861130

27864299

3169

-1752

-2238

HIST1H2AM

histone cluster 1, H2am

chr6

26271350

26275549

4199

246

-189

HIST1H2BI

histone cluster 1, H2bi

chr6

27112480

27116399

3919

179

-8367

HIST1H2BK

histone cluster 1, H2bk

chr6

27803550

27808499

4949

-415

-862

HIST1H2BN

histone cluster 1, H2bn

chr6

27861130

27864299

3169

1512

1046

HIST1H2BO

histone cluster 1, H2bo

chr6

26271350

26275549

4199

-1838

-2303

HIST1H3G histone cluster 1, H3g

chr6

27838500

27841899

3399

-101

-576

HIST1H3I

chr6

27861130

27864299

3169

-4145

-4621

HIST1H3J histone cluster 1, H3j

chr6

27838500

27841899

3399

1089

727

HIST1H4L histone cluster 1, H4l

chr6

26535750

26540379

4629

-507

-9097

HMGN4

high mobility group nucleosomal binding domain 4

chr1

19280300

19282649

2349

1351

-50701

IFFO2

intermediate filament family orphan 2

chr10

91172850

91176549

3699

375

-6058

IFIT5

interferon-induced protein with tetratricopeptide repeats 5

chr9

27527050

27529899

2849

4163

1981

IFNK

interferon, kappa

chr7

45959050

45961079

2029

806

-8215

IGFBP3

insulin-like growth factor binding protein 3

chr7

41738030

41742429

4399

2476

-11627

INHBA

inhibin, beta A

chr12

66580200

66586079

5879

162

-65252

IRAK3

interleukin-1 receptor-associated kinase 3

chr16

19126650

19132299

5649

4221

-3473

ITPRIPL2

inositol 1,4,5-trisphosphate receptor interacting protein-like 2

chr2

223914350

223923129

8779

1878

-1613

KCNE4

potassium channel, voltage gated subfamily E regulatory beta subunit 4

chr20

36886500

36889449

2949

1199

-49068

KIAA1755 KIAA1755

chr4

38663350

38677349

13999

4560

-32778

KLF3

Kruppel-like factor 3 (basic)

chr1

32572400

32575499

3099

306

-68216

KPNA6

karyopherin alpha 6 (importin alpha 7)

chr6

130339150

130343049

3899

1366

-121484

L3MBTL3

l(3)mbt-like 3 (Drosophila)

chr7

107639600

107645049

5449

1479

-78079

LAMB1

laminin, beta 1

chr6

25279050

25281649

2599

702

-340406

LRRC16A

leucine rich repeat containing 16A

chr17

45907450

45909549

2099

-600

-6190

LRRC46

leucine rich repeat containing 46

chr15

101457850

101460099

2249

-485

-151340

LRRK1

leucine-rich repeat kinase 1

chr12

91501150

91506449

5299

1742

-6567

LUM

lumican

chr8

74902130

74905999

3869

478

-37240

LY96

lymphocyte antigen 96

chr3

184432950

184435649

2699

-4464

-6144

MAGEF1

melanoma antigen family F1

chr11

96070200

96077249

7049

2619

-362285

MAML2

mastermind-like 2 (Drosophila)

histone cluster 1, H3i

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 3. Continued chr2

210442980

210444999

2019

-413

-154844

MAP2

microtubule-associated protein 2

chr3

187008000

187012429

4429

-463

-74277

MASP1

mannan-binding lectin serine peptidase 1 (C4/C2 activating component of Rareactive factor)

chr1

46267830

46271029

3199

145

-232367

MAST2

microtubule associated serine/threonine kinase 2

chr18

47806350

47808629

2279

654

-12273

MBD1

methyl-CpG binding domain protein 1

chr18

74839080

74844599

5519

2934

-151050

MBP

myelin basic protein

chr12

116713380

116717199

3819

-299

-318907

MED13L

mediator complex subunit 13-like

chr12

31810630

31813549

2919

-519

-9924

METTL20

methyltransferase like 20

chr2

170679980

170682099

2119

313

-12770

METTL5

methyltransferase like 5

chr6

31461050

31466129

5079

-2265

-15308

MICB

MHC class I polypeptide-related sequence B

chr21

17959050

17961729

2679

-2167

-2254

MIR125B2 microRNA 125b-2

chr22

38243750

38246779

3029

-4887

-4986

MIR658

microRNA 658

chr22

38243750

38246779

3029

-1484

-1580

MIR659

microRNA 659

chr16

55508150

55511249

3099

-3381

-30884

MMP2

matrix metallopeptidase 2

chr9

27527050

27529899

2849

1375

-203267

MOB3B

MOB kinase activator 3B

chr17

16943400

16952699

9299

1943

-140820

MPRIP

myosin phosphatase Rho interacting protein

chr17

45907450

45909549

2099

403

-7861

MRPL10

mitochondrial ribosomal protein L10

chr5

154319300

154322049

2749

-28295

MRPL22

mitochondrial ribosomal protein L22

chr2

17993530

17995749

2219

-3146

-3726

MSGN1

mesogenin 1

chr4

166247200

166250149

2949

-143

-15549

MSMO1

methylsterol monooxygenase 1

chr5

79286750

79288749

1999

-662

-15209

MTX3

metaxin 3

chr8

128746080

128752599

6519

1025

-4338

MYC

v-myc avian myelocytomatosis viral oncogene homolog

chr20

50176580

50181229

4649

263

-171139

NFATC2

nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 2

chr11

114161600

114169679

8079

-895

-17598

NNMT

nicotinamide N-methyltransferase

chr7

30506700

30521899

15199

4093

-50156

NOD1

nucleotide-binding oligomerization domain containing 1

chr14

27065800

27068999

3199

-440

-152310

NOVA1

neuro-oncological ventral antigen 1

chr11

89223200

89225999

2799

-167076

NOX4

NADPH oxidase 4

chr6

86158280

86161829

3549

753

-45441

NT5E

5â&#x20AC;&#x2122;-nucleotidase, ecto (CD73)

chr1

114520880

114524579

3699

700

-2145

OLFML3

olfactomedin-like 3

chr12

56210030

56213579

3549

-1

-3154

ORMDL2

ORMDL sphingolipid biosynthesis regulator 2

chr2

19562050

19564099

2049

-4703

-11828

OSR1

odd-skipped related transciption factor 1

chr9

112541880

112545699

3819

1201

-390999

PALM2AKAP2

PALM2-AKAP2 readthrough

chr11

93860300

93871049

10749

3581

-49460

PANX1

pannexin 1

chr4

120546050

120550579

4529

127

-132763

PDE5A

phosphodiesterase 5A, cGMP-specific

chr11

104033350

104036349

2999

177

-256935

PDGFD

platelet derived growth factor D

chr4

186452830

186459649

6819

472

-33388

PDLIM3

PDZ and LIM domain 3

chr17

62403900

62407599

3699

-4545

-5886

PECAM1

platelet/endothelial cell adhesion molecule 1

chr12

76421480

76425399

3919

2116

-4212

PHLDA1

pleckstrin homology-like domain, family A, member 1

chr15

68346550

68348779

2229

1093

-132737

PIAS1

protein inhibitor of activated STAT, 1

chr4

88928730

88930749

2019

920

-69188

PKD2

polycystic kidney disease 2 (autosomal dominant)

chr10

95750980

95752979

1999

-1766

-336166

PLCE1

phospholipase C, epsilon 1

153

PART THREE CHAPTER 8

Supplementary Table 3. Continued

154

chr8

38756580

38768499

11919

3787

-68888

PLEKHA2

pleckstrin homology domain containing, family A (phosphoinositide binding specific) member 2

chr12

45608280

45612229

3949

-466

-43408

PLEKHA8P1

pleckstrin homology domain containing, family A member 8 pseudogene 1

chr1

145609600

145612399

2799

-116

-18394

POLR3C

polymerase (RNA) III (DNA directed) polypeptide C (62kD)

chr5

54827750

54832679

4929

658

-109530

PPAP2A

phosphatidic acid phosphatase type 2A

chr4

81104650

81107449

2799

-374

-19430

PRDM8

PR domain containing 8

chr3

64205650

64212849

7199

1881

-129723

PRICKLE2 prickle homolog 2 (Drosophila)

chr2

219695200

219697229

2029

297

-9109

PRKAG3

protein kinase, AMP-activated, gamma 3 non-catalytic subunit

chr3

93691200

93693999

2799

334

-100718

PROS1

protein S (alpha)

chr1

214160280

214164029

3749

295

-47605

PROX1

prospero homeobox 1

chr10

129844950

129847429

2479

377

-37972

PTPRE

protein tyrosine phosphatase, receptor type, E

chr1

31536250

31538879

2629

999

-133212

PUM1

pumilio RNA-binding family member 1

chr2

37570400

37572599

2199

-253

-28964

QPCT

glutaminyl-peptide cyclotransferase

chr18

9706200

9710449

4249

-154228

RAB31

RAB31, member RAS oncogene family

chr7

4918480

4920479

1999

3855

-80740

RADIL

Ras association and DIL domains

chr15

93615750

93617799

2049

-386

-30138

RGMA

repulsive guidance molecule family member a

chr20

19866100

19872549

6449

-885

-113775

RIN2

Ras and Rab interactor 2

chr1

182556350

182559799

3449

-1967

-15303

RNASEL

ribonuclease L (2â&#x20AC;&#x2122;,5â&#x20AC;&#x2122;-oligoisoadenylate synthetase-dependent)

chr1

145609600

145612399

2799

-36

-77774

RNF115

ring finger protein 115

chr5

54827750

54832679

4929

155

-5545

RNF138P1 ring finger protein 138, E3 ubiquitin protein ligase pseudogene 1

chr18

47017100

47019279

2179

716

-3331

RPL17

ribosomal protein L17

chr4

152020180

152022349

2169

511

-4537

RPS3A

ribosomal protein S3A

chr21

36256180

36261749

5569

2022

-98866

RUNX1

runt-related transcription factor 1

chr1

165412150

165415149

2999

780

-43299

RXRG

retinoid X receptor, gamma

chr12

56210030

56213579

3549

-265

-65558

SARNP

SAP domain containing ribonucleoprotein

chr1

145091450

145093499

2049

-3932

-24447

SEC22B

SEC22 vesicle trafficking protein homolog B (S. cerevisiae) (gene/pseudogene)

chr6

2845930

2848029

2099

-4899

-13244

SERPINB1 serpin peptidase inhibitor, clade B (ovalbumin), member 1

chr18

42260150

42263649

3499

1037

-386573

SETBP1

SET binding protein 1

chr10

10837450

10842499

5049

-3098

-13573

SFTA1P

surfactant associated 1, pseudogene

chr19

16934930

16936999

2069

-4253

-55199

SIN3B

SIN3 transcription regulator family member B

chr20

1873280

1881199

7919

1414

-43299

SIRPA

signal-regulatory protein alpha

chr14

61190150

61192249

2099

-408

-14944

SIX4

SIX homeobox 4

chr1

211752550

211756379

3829

-2366

-6084

SLC30A1

solute carrier family 30 (zinc transporter), member 1

chr3

14442430

14448579

6149

1399

-44186

SLC6A6

solute carrier family 6 (neurotransmitter transporter), member 6

chr15

92393230

92401849

8619

602

-311597

SLCO3A1

solute carrier organic anion transporter family, member 3A1

chr17

33568050

33572649

4599

264

-24404

SLFN5

schlafen family member 5

chr9

72870450

72876579

6129

-363

-96274

SMC5

structural maintenance of chromosomes 5

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 3. Continued chr17

62656680

62659999

3319

-117605

SMURF2

SMAD specific E3 ubiquitin protein ligase 2

chr5

121647450

121649529

2079

670

-151304

SNCAIP

synuclein, alpha interacting protein

chr5

172442450

172446299

3849

-3354

-3557

SNORA74B

small nucleolar RNA, H/ACA box 74B

chr18

47017100

47019279

2179

-473

-537

SNORD58A

small nucleolar RNA, C/D box 58A

chr18

47017100

47019279

2179

-91

-156

SNORD58B

small nucleolar RNA, C/D box 58B

chr18

47017100

47019279

2179

-2496

-2585

SNORD58C

small nucleolar RNA, C/D box 58C

chr4

152020180

152022349

2169

-3714

-3778

SNORD73A

small nucleolar RNA, C/D box 73A

chr8

121820400

121825749

5349

1234

-275089

SNTB1

syntrophin, beta 1 (dystrophin-associated protein A1, 59kDa, basic component 1)

chr12

93964430

93969879

5449

3557

-2823

SOCS2

suppressor of cytokine signaling 2

chr5

141695980

141708449

12469

2405

-12223

SPRY4

sprouty homolog 4 (Drosophila)

chr14

35449630

35453499

3869

-539

-47206

SRP54

signal recognition particle 54kDa

chr12

26348100

26351849

3749

1469

-37731

SSPN

sarcospan

chr8

70403850

70407399

3549

597

-167522

SULF1

sulfatase 1

chr19

39934750

39937949

3199

-30956

SUPT5H

suppressor of Ty 5 homolog (S. cerevisiae)

chr9

113341100

113343129

2029

-214584

SVEP1

sushi, von Willebrand factor type A, EGF and pentraxin domain containing 1

chr10

30021930

30027449

5519

-278413

SVIL

supervillin

chr8

38641650

38646629

4979

-582

-66405

TACC1

transforming, acidic coiled-coil containing protein 1

chr6

134208730

134213179

4449

695

-2436

TCF21

transcription factor 21

chr1

218517250

218523249

5999

859

-97709

TGFB2

transforming growth factor, beta 2

chr20

30696380

30699129

2749

446

-57306

TM9SF4

transmembrane 9 superfamily protein member 4

chr8

116677100

116682699

5599

1328

-259175

TRPS1

trichorhinophalangeal syndrome I

chr15

71059230

71062099

2869

-4815

-113770

UACA

uveal autoantigen with coiled-coil domains and ankyrin repeats

chr2

170679980

170682099

2119

-2978

-259597

UBR3

ubiquitin protein ligase E3 component n-recognin 3 (putative)

chr1

165795450

165799879

4429

775

-79672

UCK2

uridine-cytidine kinase 2

chr12

109535600

109537749

2149

668

-12123

UNG

uracil-DNA glycosylase

chr7

48124050

48131299

7249

-1071

-20654

UPP1

uridine phosphorylase 1

chr5

82769000

82771779

2779

2897

-107732

VCAN

versican

chr10

104531680

104547499

15819

3702

-36424

WBP1L

WW domain binding protein 1-like

chr2

175497100

175500379

3279

567

-74437

WIPF1

WAS/WASL interacting protein family, member 1

chr1

37936150

37946979

10829

1446

-8411

ZC3H12A

zinc finger CCCH-type containing 12A

chr16

73075550

73094829

19279

-2916

-268402

ZFHX3

zinc finger homeobox 3

chr6

28108700

28110749

2049

-15511

ZKSCAN8 zinc finger with KRAB and SCAN domains 8

chr20

52201150

52203849

2699

-2864

-18888

ZNF217

zinc finger protein 217

chr16

49858550

49863349

4799

-32

-336428

ZNF423

zinc finger protein 423

chr7

63505180

63507199

2019

369

-32735

ZNF727

zinc finger protein 727

TSS = transcription start site; TES = transcription end site.

155

156

GO:0030199

GO:0030198

GO:0043062

GO:0072358

GO:0072359

GO:0048584

GO:0051094

GO:0072075

GO:0060537

GO:0009888

GO:0031497

GO:0072074

GO:0007507

GO:0030335

GO:2000147

GO:0006334

GO:0051272

GO:0071822

GO:0065004

GO:0040017

GO:0006333

GO:0048729

GO:0009967

GO:2000145

GO:0043933

GO:0010647

GO:0006357

GO:0014706

No.

GO: Biological Process

striated muscle tissue development

regulation of transcription from RNA polymerase II promoter

positive regulation of cell communication

macromolecular complex subunit organization

regulation of cell motility

positive regulation of signal transduction

tissue morphogenesis

chromatin assembly or disassembly

positive regulation of locomotion

protein-DNA complex assembly

protein complex subunit organization

positive regulation of cellular component movement

nucleosome assembly

positive regulation of cell motility

positive regulation of cell migration

heart development

kidney mesenchyme development

chromatin assembly

tissue development

muscle tissue development

metanephric mesenchyme development

positive regulation of developmental process

positive regulation of response to stimulus

circulatory system development

cardiovascular system development

extracellular structure organization

extracellular matrix organization

collagen fibril organization

Name

1,42E-5

1,35E-5

1,28E-5

1,27E-5

9,84E-6

9,25E-6

8,24E-6

7,35E-6

7,16E-6

6,62E-6

5,81E-6

5,57E-6

5,33E-6

3,98E-6

3,17E-6

2,11E-6

1,80E-6

1,76E-6

1,59E-6

1,30E-6

4,83E-7

4,61E-7

1,48E-7

1,20E-7

8,79E-8

8,44E-8

1,62E-8

p-value

4,84E-2

4,61E-2

4,37E-2

4,34E-2

3,35E-2

3,15E-2

2,81E-2

2,51E-2

2,44E-2

2,26E-2

1,98E-2

1,90E-2

1,82E-2

1,36E-2

1,08E-2

7,18E-3

6,14E-3

6,00E-3

5,41E-3

4,43E-3

1,65E-3

1,57E-3

5,05E-4

4,08E-4

3,00E-4

2,88E-4

5,53E-5

q-value Bonferroni

1,73E-3

1,71E-3

1,68E-3

1,40E-3

1,37E-3

1,28E-3

1,19E-3

1,10E-3

9,05E-4

7,70E-4

5,52E-4

5,12E-4

4,92E-4

2,06E-4

8,42E-5

8,16E-5

5,53E-5

q-value FDR B&H

1,51E-2

1,49E-2

1,46E-2

1,22E-2

1,19E-2

1,11E-2

1,04E-2

1,03E-2

9,58E-3

7,88E-3

6,71E-3

4,81E-3

4,46E-3

4,28E-3

1,79E-3

7,33E-4

7,11E-4

4,81E-4

q-value FDR B&Y

Genes from Input

357

1629

1196

1752

579

1118

572

182

337

180

1570

330

143

321

315

476

157

1794

374

911

1584

905

386

385

Genes in Annotation

development

chromatin assembly

fibrosis

chromatin assembly

fibrosis

development

chromatin assembly

fibrosis

chromatin assembly

fibrosis

chromatin assembly

fibrosis

development

chromatin assembly

development

fibrosis

development

fibrosis

Concluding remark

Supplementary Table 4. Gene enrichment analysis related to genes with increased H3K27ac occupancy within 5kb from transcription start site (TSS)

PART THREE CHAPTER 8

GO:0044420

GO:0005578

GO:0031012

GO:0015629

KEGG: 585563

KEGG: 83122

REACTOME: 106552

REACTOME: 106555

REACTOME: 905996

REACTOME: 366238

REACTOME: 205242

REACTOME: 106551

REACTOME: 106550

REACTOME: 905993

REACTOME: 160948

7,93E-4 2,24E-3 2,75E-2

q-value Bonferroni 6,64E-5 1,80E-4 1,24E-3 8,51E-3 1,01E-2 1,22E-2 1,45E-2 1,72E-2 2,02E-2 3,25E-2

2,16E-6 6,08E-6 7,48E-5

p-value 6,61E-8 1,79E-7 1,24E-6 8,47E-6 1,01E-5 1,21E-5 1,44E-5 1,71E-5 2,01E-5 3,23E-5 4,40E-5

Oxidative Stress Induced Senescence

RNA Polymerase I, RNA Polymerase III, and Mitochondrial Transcription

RNA Polymerase I Transcription

RNA Polymerase I Promoter Clearance

Meiotic Recombination

Amyloids

Senescence-Associated Secretory Phenotype (SASP)

RNA Polymerase I Chain Elongation

RNA Polymerase I Promoter Opening

Systemic lupus erythematosus

Alcoholism

Name

actin cytoskeleton

extracellular matrix

4,42E-2

1,70E-4

4,61E-7

proteinaceous extracellular matrix

extracellular matrix component

q-value Bonferroni

p-value

Name

4,02E-3

3,25E-3

2,25E-3

2,14E-3

2,07E-3

2,03E-3

2,02E-3

4,14E-4

8,99E-5

6,64E-5

q-value FDR B&H

6,88E-3

7,46E-4

3,97E-4

1,70E-4

q-value FDR B&H

3,01E-2

2,43E-2

1,68E-2

1,61E-2

1,55E-2

1,52E-2

1,51E-2

3,10E-3

6,73E-4

4,98E-4

q-value FDR B&Y

4,46E-2

4,84E-3

2,57E-3

1,10E-3

q-value FDR B&Y

Genes from Input

GO = Gene Ontology; FDR B&H = False discovery rate, Benjamini and Hochberg concept; FDR B&Y = False discovery rate, Benjamini and Yekutieli concept.

No.

Pathway

No.

GO: Cellular Component

131

126

109

138

180

Genes in Annotation

415

422

348

138

Genes in Annotation

chromatin assembly

Concluding remark

fibrosis

Concluding remark

CHROMATIN REGULATION IN CARDIOMYOPATHIES

157

PART THREE CHAPTER 8

Supplementary Table 5. Annotation of genes with decreased H3K27ac occupancy within 5kb from transcription start site (TSS) Chromo足 Peak start some

158

Peak end

Peak length

Distance to TSS

Distance to TES

Gene Symbol

Gene Name

chr17

74439750

74449399

9649

-4858

-21623

AANAT

aralkylamine N-acetyltransferase

chr2

220081700

220083849

2149

897

-8281

ABCB6

ATP-binding cassette, sub-family B (MDR/TAP), member 6 (Langereis blood group)

chr3

183732100

183737229

5129

1062

-96939

ABCC5

ATP-binding cassette, sub-family C (CFTR/MRP), member 5

chr3

183902130

183905599

3469

-7930

ABCF3

ATP-binding cassette, sub-family F (GCN20), member 3

chr20

62487700

62491949

4249

-2741

-4516

ABHD16B

abhydrolase domain containing 16B

chr3

58221830

58225429

3599

371

-56829

ABHD6

abhydrolase domain containing 6

chr19

17412600

17418449

5849

-1243

-12584

ABHD8

abhydrolase domain containing 8 actin binding Rho activating protein

chr8

107780580

107784049

3469

157

-10603

ABRA

chr17

43203380

43211299

7919

-2627

-14201

ACBD4

acyl-CoA binding domain containing 4

chr22

41863350

41867199

3849

146

-59717

ACO2

aconitase 2, mitochondrial

chr19

39132800

39136779

3979

-3537

-86380

ACTN4

actinin, alpha 4

chr3

52015700

52019599

3899

-5567

ACY1

aminoacylase 1

chr19

50190100

50192799

2699

-492

-2797

ADM5

adrenomedullin 5 (putative)

chr22

24819650

24822699

3049

-2355

-17149

ADORA2A

adenosine A2a receptor

chr14

105192780

105194929

2149

-2373

-19790

ADSSL1

adenylosuccinate synthase like 1

chr12

58119150

58121199

2049

152

-1963

AGAP2-AS1

AGAP2 antisense RNA 1

chr9

139582850

139586629

3779

-2829

-17145

AGPAT2

1-acylglycerol-3-phosphate O-acyltransferase 2

chr21

45342180

45350899

8719

1261

-60934

AGPAT3

1-acylglycerol-3-phosphate O-acyltransferase 3

chr3

46730700

46739949

9249

-154

-24646

ALS2CL

ALS2 C-terminal like

chr9

140081500

140084979

3479

-183

-14004

ANAPC2

anaphase promoting complex subunit 2

chr19

8430330

8434329

3999

3319

-6927

ANGPTL4

angiopoietin-like 4

chr8

41511200

41530879

19679

1764

-10295

ANK1

ankyrin 1, erythrocytic

chr20

3798150

3802979

4829

-638

-5389

AP5S1

adaptor-related protein complex 5, sigma 1 subunit

chr11

57001650

57003999

2349

2102

-1771

APLNR

apelin receptor

chr7

30944450

30966699

22249

4107

-9556

AQP1

aquaporin 1 (Colton blood group)

chr9

33400630

33403099

2469

652

-16916

AQP7

aquaporin 7

chr11

72428900

72435899

6999

1003

-36285

ARAP1

ArfGAP with RhoGAP domain, ankyrin repeat and PH domain 1

chr17

36602800

36617799

14999

-3344

-58327

ARHGAP23

Rho GTPase activating protein 23

chr17

43485880

43490729

4849

-4856

-17036

ARHGAP27

Rho GTPase activating protein 27

chr8

1770700

1774149

3449

276

-134381

ARHGEF10

Rho guanine nucleotide exchange factor (GEF) 10

chr1

3371530

3374849

3319

2043

-24485

ARHGEF16

Rho guanine nucleotide exchange factor (GEF) 16

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr12

58001480

58008899

7419

-28

-5835

ARHGEF25

Rho guanine nucleotide exchange factor (GEF) 25

chr1

23804600

23812099

7499

2400

-53294

ASAP3

ArfGAP with SH3 domain, ankyrin repeat and PH domain 3

chr7

150884380

150886499

2119

-962

-12655

ASB10

ankyrin repeat and SOCS box containing 10

chr17

42247200

42250329

3129

691

-7686

ASB16

ankyrin repeat and SOCS box containing 16

chr2

237168450

237173579

5129

1973

-67499

ASB18

ankyrin repeat and SOCS box containing 18

chr17

79933380

79936049

2669

-711

-40565

ASPSCR1

alveolar soft part sarcoma chromosome region, candidate 1

chr9

133319850

133323549

3699

1606

-54960

ASS1

argininosuccinate synthase 1

chr11

64682930

64687299

4369

-393

-23110

ATG2A

autophagy related 2A

chr7

128498250

128501749

3499

-2898

-5902

ATP6V1F

ATPase, H+ transporting, lysosomal 14kDa, V1 subunit F

chr17

42277350

42279399

2049

-2846

-9201

ATXN7L3

ataxin 7-like 3

chr1

1307630

1312749

5119

372

-1079

AURKAIP1

aurora kinase A interacting protein 1

chr6

33243600

33248429

4829

1098

-589

B3GALT4

UDP-Gal:betaGlcNAc beta 1,3-galactosyltransferase, polypeptide 4

chr15

40730530

40733399

2869

-1446

-28476

BAHD1

bromo adjacent homology domain containing 1

chr17

79012250

79014249

1999

4303

-71062

BAIAP2

BAI1-associated protein 2

chr6

33544230

33549349

5119

1280

-6466

BAK1

BCL2-antagonist/killer 1

chr3

52442050

52446799

4749

-416

-9398

BAP1

BRCA1 associated protein-1 (ubiquitin carboxy-terminal hydrolase)

chr19

45310400

45315649

5249

687

-10585

BCAM

basal cell adhesion molecule (Lutheran blood group)

chr19

49312350

49317799

5449

-897

-16755

BCAT2

branched chain amino-acid transaminase 2, mitochondrial

chr16

31122030

31126299

4269

4503

BCKDK

branched chain ketoacid dehydrogenase kinase

chr16

30903680

30909349

5669

-1116

-7398

BCL7C

B-cell CLL/lymphoma 7C

chr16

2262050

2266549

4499

-3289

-5046

BRICD5

BRICHOS domain containing 5

chr19

2013430

2017449

4019

262

-29993

BTBD2

BTB (POZ) domain containing 2

chr5

180470430

180473549

3119

4765

-16532

BTNL9

butyrophilin-like 9

chr12

7050980

7056099

5119

337

-1624

C12orf57

chromosome 12 open reading frame 57

chr15

36867750

36870399

2649

-2737

-233373

C15orf41

chromosome 15 open reading frame 41

chr16

685600

694399

8799

-3653

-5571

C16orf13

chromosome 16 open reading frame 13

chr16

67700130

67702549

2419

623

-1319

C16orf86

chromosome 16 open reading frame 86

chr17

76138880

76142799

3919

-1594

-21523

C17orf99

chromosome 17 open reading frame 99

chr1

45138330

45141879

3549

-289

-51158

C1orf228

chromosome 1 open reading frame 228

159

PART THREE CHAPTER 8

Supplementary Table 5. Continued

160

chr20

3744750

3749799

5049

1177

-13117

C20orf27

chromosome 20 open reading frame 27

chr15

62457530

62459779

2249

-1173

-2918

C2CD4B

C2 calcium-dependent domain containing 4B

chr2

241836580

241839499

2919

-2467

-12574

C2orf54

chromosome 2 open reading frame 54

chr9

139832150

139847999

15849

377

-1351

C8G

complement component 8, gamma polypeptide

chr8

22453030

22464799

11769

1793

-2745

C8orf58

chromosome 8 open reading frame 58

chr9

139883230

139889449

6219

-530

-2087

C9orf142

chromosome 9 open reading frame 142

chr9

140144030

140149999

5969

1285

-919

C9orf173

chromosome 9 open reading frame 173

chr1

150227300

150233699

6399

282

-6976

CA14

carbonic anhydrase XIV

chr20

60980500

60982679

2179

749

-17904

CABLES2

Cdk5 and Abl enzyme substrate 2

chr17

4888830

4894049

5219

-509

-20147

CAMTA2

calmodulin binding transcription activator 2

chr16

574980

577699

2719

-1516

-28294

CAPN15

calpain 15

chr11

67181830

67188699

6869

2116

-7811

CARNS1

carnosine synthase 1

chr17

73503400

73511679

8279

-1582

-11197

CASKIN2

CASK interacting protein 2

chr7

142981380

142987279

5899

-1072

-20454

CASP2

caspase 2, apoptosis-related cysteine peptidase

chr1

10854430

10856949

2519

1017

-159022

CASZ1

castor zinc finger 1

chr17

77750250

77752249

1999

-743

-5081

CBX2

chromobox homolog 2

chr8

22453030

22464799

11769

-3636

-19060

CCAR2

cell cycle and apoptosis regulator 2

chr19

56159080

56163729

4649

2451

-3119

CCDC106

coiled-coil domain containing 106

chr19

11544680

11547879

3199

-300

-15007

CCDC151

coiled-coil domain containing 151

chr9

139685050

139687199

2149

-4677

-16067

CCDC183

coiled-coil domain containing 183

chr15

74606200

74608699

2499

-3450

-21031

CCDC33

coiled-coil domain containing 33

chr17

80170500

80172549

2049

-836

-112179

CCDC57

coiled-coil domain containing 57

chr3

49201400

49205299

3899

435

-3381

CCDC71

coiled-coil domain containing 71

chr20

30597300

30599829

2529

320

-21419

CCM2L

cerebral cavernous malformation 2-like

chr17

41923180

41927749

4569

918

-15529

CD300LG

CD300 molecule-like family member g

chr19

49841400

49845099

3699

4573

-609

CD37

CD37 molecule

chr6

14117500

14120079

2579

925

-18356

CD83

CD83 molecule

chr11

65077050

65079229

2179

-4191

-11759

CDC42EP2

CDC42 effector protein (Rho GTPase binding) 2

chr1

205471350

205477149

5799

523

-27665

CDK18

cyclin-dependent kinase 18

chr11

67270450

67277679

7229

2037

-96

CDK2AP2

cyclin-dependent kinase 2 associated protein 2

chr11

2910230

2912229

1999

-4235

-6780

CDKN1C

cyclin-dependent kinase inhibitor 1C (p57, Kip2)

chr16

66965730

66971329

5599

183

-10462

CES2

carboxylesterase 2

chr16

56994150

56997349

3199

-85

-22006

CETP

cholesteryl ester transfer protein, plasma

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr15

41247600

41251249

3649

3789

708

CHAC1

ChaC glutathione-specific gamma-glutamylcyclotransferase 1

chr22

51014300

51022929

8629

2813

-1228

CHKB

choline kinase beta

chr22

51014300

51022929

8629

2813

-11324

CHKB-CPT1B

CHKB-CPT1B readthrough (NMD candidate)

chr11

45682650

45685499

2849

3097

-13648

CHST1

carbohydrate (keratan sulfate Gal-6) sulfotransferase 1

chr1

16337830

16349649

11819

-4746

-16804

CLCNKA

chloride channel, voltage-sensitive Ka

chr22

19507780

19514699

6919

1620

-693

CLDN5

claudin 5

chr9

139883230

139889449

6219

4684

2720

CLIC3

chloride intracellular channel 3

chr7

73701450

73706329

4879

-116382

CLIP2

CAP-GLY domain containing linker protein 2

chr4

17512050

17515299

3249

-3113

-15052

CLRN2

clarin 2

chr17

2609200

2611349

2149

4652

-17595

CLUH

clustered mitochondria (cluA/CLU1) homolog

chr16

81475680

81482179

6499

155

-266435

CMIP

c-Maf inducing protein

chr16

66638980

66641179

2199

1513

-7713

CMTM3

CKLF-like MARVEL transmembrane domain containing 3

chr1

224544200

224546429

2229

720

-18376

CNIH4

cornichon family AMPA receptor auxiliary protein 4

chr19

17666050

17668899

2849

964

-26490

COLGALT1

collagen beta(1-O) galactosyltransferase 1

chr17

27945730

27951749

6019

-299

-6966

CORO6

coronin 6

chr22

51014300

51022929

8629

-2109

-11324

CPT1B

carnitine palmitoyltransferase 1B (muscle)

chr19

50190100

50192799

2699

-2957

-25538

CPT1C

carnitine palmitoyltransferase 1C

chr14

105936030

105952849

16819

3309

-2060

CRIP2

cysteine-rich protein 2

chr6

43273300

43276279

2979

1740

-1579

CRIP3

cysteine-rich protein 3

chr8

19533450

19536879

3429

4929

-273492

CSGALNACT1 chondroitin sulfate N-acetylgalactosaminyltransferase 1

chr4

1242630

1244849

2219

-832

-38512

CTBP1

C-terminal binding protein 1

chr4

1242630

1244849

2219

-437

-2876

CTBP1-AS2

CTBP1 antisense RNA 2 (head to head)

chr16

30903680

30909349

5669

-1413

-8365

CTF1

cardiotrophin 1

chr1

9960150

9973749

13599

3366

-58616

CTNNBIP1

catenin, beta interacting protein 1

chr9

140114550

140116549

1999

-4094

-5212

CYSRT1

cysteine-rich tail protein 1

chr1

100713980

100716899

2919

-31

-62961

DBT

dihydrolipoamide branched chain transacylase E2

chr19

17412600

17418449

5849

-4812

-18579

DDA1

DET1 and DDB1 associated 1

chr1

20985900

20990479

4579

-153

-9930

DDOST

dolichyl-diphosphooligosaccharide-protein glycosyltransferase subunit (non-catalytic)

chr20

3183500

3185779

2279

655

-13622

DDRGK1

DDRGK domain containing 1

chr1

153915700

153922549

6849

-17148

DENND4B

DENN/MADD domain containing 4B

chr2

220279700

220288849

9149

1176

-7184

DES

desmin

chr8

145546230

145550849

4619

2027

-10293

DGAT1

diacylglycerol O-acyltransferase 1

chr22

18889580

18893529

3949

-2181

-8045

DGCR6

DiGeorge syndrome critical region gene 6

161

PART THREE CHAPTER 8

Supplementary Table 5. Continued chr22

162

20307150

20310579

3429

-1257

-7063

DGCR6L

DiGeorge syndrome critical region gene 6-like

chr11

46359280

46364779

5499

-4957

-40074

DGKZ

diacylglycerol kinase, zeta

chr19

2721400

2725129

3729

-1875

-8700

DIRAS1

DIRAS family, GTP-binding RAS-like 1

chr16

84175700

84179199

3499

-1415

-34073

DNAAF1

dynein, axonemal, assembly factor 1

chr11

6517630

6520149

2519

364

-23206

DNHD1

dynein heavy chain domain 1

chr3

16304800

16310129

5329

-969

-8896

DPH3

diphthamide biosynthesis 3

chr2

162929150

162931229

2079

862

-81433

DPP4

dipeptidyl-peptidase 4

chr18

32072330

32075829

3499

826

-335212

DTNA

dystrobrevin, alpha

chr1

1279200

1283179

3979

3302

-10531

DVL1

dishevelled segment polarity protein 1

chr2

71693380

71698379

4999

2048

-218012

DYSF

dysferlin

chr10

135185580

135188849

3269

-307

-11227

ECHS1

enoyl CoA hydratase, short chain, 1, mitochondrial

chr16

2299730

2303279

3549

-11606

ECI1

enoyl-CoA delta isomerase 1

chr20

62127280

62136049

8769

-1160

-12299

EEF1A2

eukaryotic translation elongation factor 1 alpha 2

chr3

127871100

127876849

5749

1662

-253513

EEFSEC

eukaryotic elongation factor, selenocysteine-tRNA-specific

chr11

65637800

65641799

3999

540

-5887

EFEMP2

EGF containing fibulin-like extracellular matrix protein 2

chr9

139549230

139560549

11319

-2492

-12239

EGFL7

EGF-like-domain, multiple 7

chr19

39108050

39111379

3329

-7

-17880

EIF3K

eukaryotic translation initiation factor 3, subunit K

chr3

47553250

47555429

2179

859

-17209

ELP6

elongator acetyltransferase complex subunit 6

chr17

48449130

48451499

2369

-266

-8505

EME1

essential meiotic structure-specific endonuclease 1

chr16

67700130

67702549

2419

-712

-4490

ENKD1

enkurin domain containing 1

chr6

132127250

132129929

2679

-566

-87703

ENPP1

ectonucleotide pyrophosphatase/ phosphodiesterase 1

chr19

45925350

45928979

3629

-345

-10472

ERCC1

excision repair crosscomplementation group 1

chr16

68263750

68268249

4499

4136

-3550

ESRP2

epithelial splicing regulatory protein 2

chr11

64070830

64074629

3799

-314

-11480

ESRRA

estrogen-related receptor alpha

chr1

2123680

2131879

8199

-1566

-6791

FAAP20

Fanconi anemia core complex associated protein 20

chr16

1875050

1878249

3199

-575

-2258

FAHD1

fumarylacetoacetate hydrolase domain containing 1

chr1

26480880

26493149

12269

1504

-2103

FAM110D

family with sequence similarity 110, member D

chr9

140144030

140149999

5969

-4793

-8978

FAM166A

family with sequence similarity 166, member A

chr16

685600

694399

8799

-1849

-8474

FAM195A

family with sequence similarity 195, member A

chr17

79789250

79793849

4599

-383

-11257

FAM195B

family with sequence similarity 195, member B

chr3

49839150

49847149

7999

2463

687

FAM212A

family with sequence similarity 212, member A

chr1

2513830

2518479

4649

-2094

-6747

FAM213B

family with sequence similarity 213, member B

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr16

67557730

67568399

10669

311

-17623

FAM65A

family with sequence similarity 65, member A

chr9

134153400

134155549

2149

-2569

-21009

FAM78A

family with sequence similarity 78, member A

chr16

66965730

66971329

5599

-210

-2571

FAM96B

family with sequence similarity 96, member B

chr17

80056450

80060499

4049

-2369

-22260

FASN

fatty acid synthase

chr9

139832150

139847999

15849

-902

-5187

FBXW5

F-box and WD repeat domain containing 5

chr1

27699250

27701399

2149

990

-4722

FCN3

ficolin (collagen/fibrinogen domain containing) 3

chr17

72872300

72875879

3579

-4934

-15470

FDXR

ferredoxin reductase

chr11

64006730

64011049

4319

-698

-2716

FKBP2

FK506 binding protein 2, 13kDa

chr17

43299000

43307899

8899

4158

-21231

FMNL1

formin-like 1

chr9

130563450

130566979

3529

-293

-11341

FPGS

folylpolyglutamate synthase

chr19

50311200

50314199

2999

3768

-2551

FUZ

fuzzy planar cell polarity protein

chr5

170213550

170217499

3949

4802

-25523

GABRP

gamma-aminobutyric acid (GABA) A receptor, pi

chr22

29700700

29709399

8699

2053

-3724

GAS2L1

growth arrest-specific 2 like 1

chr7

56030430

56037579

7149

1709

-33866

GBAS

glioblastoma amplified sequence

chr6

33544230

33549349

5119

-4686

-10012

GGNBP1

gametogenetin binding protein 1 (pseudogene)

chr1

228333250

228335279

2029

-3288

-13260

GJC2

gap junction protein, gamma 2, 47kDa

chr20

62257800

62260899

3099

-969

-40395

GMEB2

glucocorticoid modulatory element binding protein 2

chr19

34854350

34858199

3849

206

-34961

GPI

glucose-6-phosphate isomerase

chr8

144289550

144292349

2799

-4118

-8093

GPIHBP1

glycosylphosphatidylinositol anchored high density lipoprotein binding protein 1

chr14

105530000

105532449

2449

529

-15487

GPR132

G protein-coupled receptor 132

chr7

1087930

1097449

9519

-4451

-6207

GPR146

G protein-coupled receptor 146

chr2

128400730

128402849

2119

-1649

-8422

GPR17

G protein-coupled receptor 17

chr9

139217600

139224349

6749

-957

-15312

GPSM1

G-protein signaling modulator 1

chr8

145725930

145730229

4299

-1385

-4474

GPT

glutamic-pyruvate transaminase (alanine aminotransferase)

chr22

47022480

47025799

3319

1482

-51547

GRAMD4

GRAM domain containing 4

chr8

144640180

144642199

2019

713

-4039

GSDMD

gasdermin D

chr4

106628350

106631429

3079

-2080

-137018

GSTCD

glutathione S-transferase, C-terminal domain containing

chr2

27577800

27581299

3499

318

-30829

GTF3C2

general transcription factor IIIC, polypeptide 2, beta 110kDa

chr1

228333250

228335279

2029

1861

-2388

GUK1

guanylate kinase 1

chr2

26463700

26470749

7049

369

-53720

HADHA

hydroxyacyl-CoA dehydrogenase/3ketoacyl-CoA thiolase/enoyl-CoA hydratase (trifunctional protein), alpha subunit

chr2

26463700

26470749

7049

-391

-46107

HADHB

hydroxyacyl-CoA dehydrogenase/3ketoacyl-CoA thiolase/enoyl-CoA hydratase (trifunctional protein), beta subunit

163

PART THREE CHAPTER 8

Supplementary Table 5. Continued

164

chr16

1875050

1878249

3199

545

-17544

HAGH

chr1

937850

941599

3749

-4173

-5383

HES4

hes family bHLH transcription factor 4

chr9

35810700

35815449

4749

1967

-116

HINT2

histidine triad nucleotide binding protein 2

chr19

37824430

37826979

2549

125

-29650

HKR1

HKR1, GLI-Kruppel zinc finger family member

hydroxyacylglutathione hydrolase

chr5

175105900

175116349

10449

2661

-433

HRH2

histamine receptor H2

chr17

40705030

40708349

3319

2706

-534

HSD17B1

hydroxysteroid (17-beta) dehydrogenase 1

chr16

84175700

84179199

3499

1350

-21706

HSDL1

hydroxysteroid dehydrogenase like 1

chr1

16337830

16349649

11819

1545

-3217

HSPB7

heat shock 27kDa protein family, member 7 (cardiovascular)

chr3

50332550

50342399

9849

3529

-154

HYAL1

hyaluronoglucosaminidase 1

chr3

50332550

50342399

9849

-576

-7213

HYAL3

hyaluronoglucosaminidase 3

chr3

50327080

50332249

5169

361

-4500

IFRD2

interferon-related developmental regulator 2

chr11

68669900

68673799

3899

531

-36219

IGHMBP2

immunoglobulin mu binding protein 2

chr19

14141200

14144229

3029

453

-21310

IL27RA

interleukin 27 receptor, alpha

chr19

15232930

15237899

4969

1162

-9628

ILVBL

ilvB (bacterial acetolactate synthase)-like

chr2

86420630

86422849

2219

1153

-50684

IMMT

inner membrane protein, mitochondrial

chr14

105153800

105158049

4249

-33

-15211

INF2

inverted formin, FH2 and WH2 domain containing

chr22

31513550

31520899

7349

-1736

-13457

INPP5J

inositol polyphosphate-5phosphatase J

chr4

106628350

106631429

3079

-9

-26103

INTS12

integrator complex subunit 12

chr6

33712580

33716449

3869

247

-25072

IP6K3

inositol hexakisphosphate kinase 3

chr1

44410650

44415499

4849

597

-20618

IPO13

importin 13

chr9

95430680

95434079

3399

167

-56914

IPPK

inositol 1,3,4,5,6-pentakisphosphate 2-kinase

chr16

19726930

19731429

4499

1402

-139677

IQCK

IQ motif containing K

chr19

49246330

49251199

4869

1401

-4618

IZUMO1

izumo sperm-egg fusion 1

chr19

8405250

8409529

4279

756

-19921

KANK3

KN motif and ankyrin repeat domains 3

chr16

31122030

31126299

4269

-4820

-18548

KAT8

K(lysine) acetyltransferase 8

chr22

38844330

38850849

6519

3613

-25257

KCNJ4

potassium channel, inwardly rectifying subfamily J, member 4

chr11

2478500

2486549

8049

-159

-387814

KCNQ1

potassium channel, voltage gated KQT-like subfamily Q, member 1

chr3

49207300

49215029

7729

2097

-2753

KLHDC8B

kelch domain containing 8B

chr12

27931400

27934829

3429

-72

-22857

KLHL42

kelch-like family member 42

chr16

19726930

19731429

4499

312

-11504

KNOP1

lysine-rich nucleolar protein 1

chr12

52583730

52590249

6519

-1206

-24209

KRT80

keratin 80, type II

chr3

49167180

49174629

7449

-306

-12357

LAMB2

laminin, beta 2 (laminin S)

chr9

139641350

139644029

2679

290

-4221

LCN6

lipocalin 6

chr9

139873150

139879049

5899

-1345

-4110

LCNL1

lipocalin-like 1

chr17

76975600

76977629

2029

-554

-9279

LGALS3BP

lectin, galactoside-binding, soluble, 3 binding protein

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr2

128400730

128402849

2119

3963

-5793

LIMS2

LIM and senescent cell antigen-like domains 2

chr1

226492350

226498679

6329

1683

-76654

LIN9

lin-9 DREAM MuvB core complex component

chr16

574980

577699

2719

1067

507

LINC00235

long intergenic non-protein coding RNA 235

chr1

204010500

204015049

4549

-2383

-11200

LINC00303

long intergenic non-protein coding RNA 303

chr21

44894180

44902029

7849

-2

-16131

LINC00313

long intergenic non-protein coding RNA 313

chr17

8128700

8131129

2429

-2554

-5966

LINC00324

long intergenic non-protein coding RNA 324

chr6

33552330

33560199

7869

4850

-2382

LINC00336

long intergenic non-protein coding RNA 336

chr17

79280000

79288399

8399

-1152

-7576

LINC00482

long intergenic non-protein coding RNA 482 lemur tyrosine kinase 2

chr7

97736250

97738279

2029

1068

-101677

LMTK2

chr16

427600

430899

3299

-2991

-13710

LOC100134368 uncharacterized LOC100134368

chr2

132247880

132249929

2049

-1481

-30244

LOC150776

sphingomyelin phosphodiesterase 4, neutral membrane (neutral sphingomyelinase-3) pseudogene

chr1

148927800

148930399

2599

-1305

-23954

LOC645166

lymphocyte-specific protein 1 pseudogene

chr12

6739500

6743499

3999

-685

-13499

LPAR5

lysophosphatidic acid receptor 5

chr22

25759850

25762399

2549

-2601

-13739

LRP5L

low density lipoprotein receptorrelated protein 5-like

chr5

192450

196749

4299

2974

-868

LRRC14B

leucine rich repeat containing 14B

chr18

7232780

7235749

2969

3128

2224

LRRC30

leucine rich repeat containing 30

chr17

79978800

79981299

2499

-1230

-8976

LRRC45

leucine rich repeat containing 45

chr6

43478000

43481749

3749

-1451

-5168

LRRC73

leucine rich repeat containing 73

chr20

60691250

60697399

6149

-3192

-16109

LSM14B

LSM14B, SCD6 homolog B (S. cerevisiae)

chr11

1883600

1887029

3429

-4588

-28177

LSP1

lymphocyte-specific protein 1

chr3

46503580

46506349

2769

1430

-27468

LTF

lactotransferrin

chr11

47289380

47293829

4449

-66

-59977

MADD

MAP-kinase activating death domain

chr17

79881300

79885699

4399

-2092

-7353

MAFG

v-maf avian musculoaponeurotic fibrosarcoma oncogene homolog G

chr19

39108050

39111379

3329

-1072

-31434

MAP4K1

mitogen-activated protein kinase kinase kinase kinase 1

chr11

64566650

64572699

6049

1038

-13065

MAP4K2

mitogen-activated protein kinase kinase kinase kinase 2

chr17

19279500

19282399

2899

-824

-5906

MAPK7

mitogen-activated protein kinase 7

chr5

65890530

65892729

2199

-546

-390720

MAST4

microtubule associated serine/ threonine kinase family member 4

chr22

36005130

36020099

14969

769

-9803

myoglobin

chr11

119184500

119193429

8929

-1125

-9730

MCAM

melanoma cell adhesion molecule

chr13

113623380

113625479

2099

1673

-128432

MCF2L

MCF.2 cell line derived transforming sequence-like

chr1

12037930

12047429

9499

2442

-30891

MFN2

mitofusin 2

chr8

145732030

145734849

2819

-1112

-3149

MFSD3

major facilitator superfamily domain containing 3

165

PART THREE CHAPTER 8

Supplementary Table 5. Continued chr4

166

679900

684749

4849

648

-6706

MFSD7

major facilitator superfamily domain containing 7

chr17

37884730

37887429

2699

708

-669

MIEN1

migration and invasion enhancer 1

chr1

12078850

12081379

2529

603

-11991

MIIP

migration and invasion inhibitory protein

chr12

7065200

7074549

9349

-3385

-3479

MIR141

microRNA 141

chr22

20019100

20021679

2579

-272

-352

MIR185

microRNA 185

chr11

64657150

64663099

5949

-1407

-1516

MIR192

microRNA 192

chr11

64657150

64663099

5949

-1214

-1298

MIR194-2

microRNA 194-2

chr12

7065200

7074549

9349

-2987

-3054

MIR200C

microRNA 200c

chr2

219268450

219272979

4529

3346

3270

MIR26B

microRNA 26b

chr8

41511200

41530879

19679

-3014

-3081

MIR486-1

microRNA 486-1

chr2

238393950

238399729

5779

962

-67120

MLPH

melanophilin

chr16

83930830

83940479

9649

2925

-14130

MLYCD

malonyl-CoA decarboxylase

chr10

88712330

88722199

9869

160

-21965

MMRN2

multimerin 2

chr6

29621000

29627249

6249

-684

-16022

MOG

myelin oligodendrocyte glycoprotein

chr16

118550

130449

11899

-3669

-11341

MPG

N-methylpurine-DNA glycosylase

chr17

41907450

41912249

4799

688

-31681

MPP3

membrane protein, palmitoylated 3 (MAGUK p55 subfamily member 3)

chr20

61422600

61429579

6979

-1715

-5853

MRGBP

MRG/MORF4L binding protein

chr11

68669900

68673799

3899

-547

-13103

MRPL21

mitochondrial ribosomal protein L21

chr11

1965980

1972229

6249

603

-8732

MRPL23

mitochondrial ribosomal protein L23

chr11

2007080

2010879

3799

2170

-4540

MRPL23-AS1

MRPL23 antisense RNA 1

chr17

48449130

48451499

2369

247

-5086

MRPL27

mitochondrial ribosomal protein L27

chr19

17412600

17418449

5849

-952

-2127

MRPL34

mitochondrial ribosomal protein L34

chr2

86420630

86422849

2219

-4816

-18736

MRPL35

mitochondrial ribosomal protein L35

chr5

68512880

68516199

3319

967

-11443

MRPS36

mitochondrial ribosomal protein S36

chr14

105884430

105890499

6069

1279

-49583

MTA1

metastasis associated 1

chr9

129087400

129091479

4079

312

-70472

MVB12B

multivesicular body subunit 12B

chr4

2264730

2269349

4619

-3301

-17880

MXD4

MAX dimerization protein 4

chr17

4456380

4460729

4349

126

-16363

MYBBP1A

MYB binding protein (P160) 1a

chr11

69061700

69069049

7349

3753

622

MYEOV

myeloma overexpressed

chr14

23902280

23915729

13449

-4135

-27057

MYH7

myosin, heavy chain 7, cardiac muscle, beta

chr3

46901750

46917629

15879

-4717

-10333

MYL3

myosin, light chain 3, alkali; ventricular, skeletal, slow

chr7

44176550

44180699

4149

2291

-160

MYL7

myosin, light chain 7, regulatory

chr16

46776200

46786679

10479

781

-45244

MYLK3

myosin light chain kinase 3

chr16

30381630

30385179

3549

-2718

-5905

MYLPF

myosin light chain, phosphorylatable, fast skeletal muscle

chr17

73585450

73588399

2949

2786

-36000

MYO15B

myosin XVB pseudogene

chr17

1384900

1400779

15879

-3789

-25358

MYO1C

myosin IC

chr2

132247880

132249929

2049

1159

-7371

MZT2A

mitotic spindle organizing protein 2A

chr11

63705000

63708099

3099

108

-17894

NAA40

N(alpha)-acetyltransferase 40, NatD catalytic subunit

chr8

144659150

144662599

3449

-362

-3919

NAPRT

nicotinate phosphoribosyltransferase

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr3

50332550

50342399

9849

-763

-3641

NAT6

N-acetyltransferase 6 (GCN5-related)

chr3

47016580

47030849

14269

2542

-27478

NBEAL2

neurobeachin-like 2

chr3

196667800

196671699

3899

-286

-7477

NCBP2

nuclear cap binding protein subunit 2, 20kDa

chr11

67372880

67376099

3219

167

-5522

NDUFV1

NADH dehydrogenase (ubiquinone) flavoprotein 1, 51kDa

chr9

140144030

140149999

5969

-2744

-20984

NELFB

negative elongation factor complex member B

chr6

31829100

31832549

3449

-116

-3995

NEU1

sialidase 1 (lysosomal sialidase)

chr19

3357000

3359979

2979

-1126

-105112

NFIC

nuclear factor I/C (CCAAT-binding transcription factor)

chr19

3930650

3932849

2199

-1351

-10662

NMRK2

nicotinamide riboside kinase 2

chr9

139440700

139445029

4329

-2627

-53968

NOTCH1

notch 1

chr16

1828450

1830549

2099

-3433

-9691

NUBP2

nucleotide binding protein 2

chr6

34358350

34362149

3799

191

-104248

NUDT3

nudix (nucleoside diphosphate linked moiety X)-type motif 3

chr11

67397380

67400999

3619

-1789

-3780

NUDT8

nudix (nucleoside diphosphate linked moiety X)-type motif 8

chr12

57604630

57608729

4099

-3898

-13551

NXPH4

neurexophilin 4

chr8

145115030

145119379

4349

-1621

-11038

OPLAH

5-oxoprolinase (ATP-hydrolysing)

chr11

4663080

4669349

6269

1059

-10499

OR51E1

olfactory receptor, family 51, subfamily E, member 1

chr3

16304800

16310129

5329

751

-40128

OXNAD1

oxidoreductase NAD-binding domain containing 1

chr11

72926850

72931279

4429

-437

-18328

P2RY2

purinergic receptor P2Y, G-protein coupled, 2

chr17

79820600

79822599

1999

-3056

-20566

P4HB

prolyl 4-hydroxylase, beta polypeptide

chr11

47196880

47211899

15019

3568

-5305

PACSIN3

protein kinase C and casein kinase substrate in neurons 3

chr8

145055750

145062349

6599

1585

-7729

PARP10

poly (ADP-ribose) polymerase family, member 10

chr1

154925950

154931229

5279

-23

-12029

PBXIP1

pre-B-cell leukemia homeobox interacting protein 1

chr3

51995350

52003949

8599

-3998

-8176

PCBP4

poly(rC) binding protein 4

chr4

698200

702999

4799

1027

-63825

PCGF3

polycomb group ring finger 3

chr2

242798300

242803129

4829

343

-8681

PDCD1

programmed cell death 1

chr5

269850

272399

2549

-611

-43964

PDCD6

programmed cell death 6

chr19

10539800

10545299

5499

-561

-37757

PDE4A

phosphodiesterase 4A, cAMPspecific

chr17

48171180

48176399

5219

1094

-14632

PDK2

pyruvate dehydrogenase kinase, isozyme 2

chr7

95223880

95231629

7749

-1830

-14940

PDK4

pyruvate dehydrogenase kinase, isozyme 4

chr7

44101430

44106729

5299

1106

-1753

PGAM2

phosphoglycerate mutase 2 (muscle)

chr16

2262050

2266549

4499

522

-2697

PGP

phosphoglycolate phosphatase

chr22

41863350

41867199

3849

-567

-9553

PHF5A

PHD finger protein 5A

chr3

52442050

52446799

4749

-102

-13231

PHF7

PHD finger protein 7

chr11

803650

807549

3899

-1136

-6421

PIDD1

p53-induced death domain protein 1

167

PART THREE CHAPTER 8

Supplementary Table 5. Continued

168

chr20

44043350

44050849

7499

2376

-7784

PIGT

phosphatidylinositol glycan anchor biosynthesis, class T

chr22

50346600

50353799

7199

-3943

-7519

PIM3

Pim-3 proto-oncogene, serine/ threonine kinase

chr11

67270450

67277679

7229

-1222

-14825

PITPNM1

phosphatidylinositol transfer protein, membrane-associated 1

chr12

123575300

123606029

30729

4310

-122638

PITPNM2

phosphatidylinositol transfer protein, membrane-associated 2

chr9

131464700

131469279

4579

2188

-16208

PKN3

protein kinase N3

chr16

68277400

68283829

6429

1368

-14344

PLA2G15

phospholipase A2, group XV

chr17

43203380

43211299

7919

2551

-18331

PLCD3

phospholipase C, delta 3

chr8

145041550

145051029

9479

1407

-56969

PLEC

plectin

chr19

4517900

4520749

2849

-1609

-17133

PLIN4

perilipin 4

chr3

142313680

142316179

2499

-299

-117573

PLS1

plastin 1

chr22

50746300

50752549

6249

-3423

-36017

PLXNB2

plexin B2

chr17

37819630

37837829

18199

4223

2002

PNMT

phenylethanolamine N-methyltransferase

chr11

817800

821749

3949

873

-5441

PNPLA2

patatin-like phospholipase domain containing 2

chr6

43478000

43481749

3749

-4916

-9369

POLR1C

polymerase (RNA) I polypeptide C, 30kDa

chr11

842750

847999

5249

-2846

-5653

POLR2L

polymerase (RNA) II (DNA directed) polypeptide L, 7.6kDa

chr9

134161100

134177199

16099

4069

-15499

PPAPDC3

phosphatidic acid phosphatase type 2 domain containing 3

chr17

79789250

79793849

4599

1376

-182

PPP1R27

protein phosphatase 1, regulatory subunit 27

chr8

9011450

9014329

2879

-4670

-19116

PPP1R3B

protein phosphatase 1, regulatory subunit 3B

chr11

64083730

64089799

6069

1196

-2517

PRDX5

peroxiredoxin 5

chr19

11544680

11547879

3199

-15500

PRKCSH

protein kinase C substrate 80K-H

chr19

50091980

50096699

4719

-572

-35355

PRR12

proline rich 12

chr10

694600

697729

3129

148

-14943

PRR26

proline rich 26

chr17

62074250

62078049

3799

439

-5492

PRR29

proline rich 29

chr14

24613880

24617999

4119

-85

-3365

PSME2

proteasome (prosome, macropain) activator subunit 2 (PA28 beta)

chr11

450730

452949

2219

1560

-39546

PTDSS2

phosphatidylserine synthase 2

chr9

139867530

139871599

4069

-2391

-6628

PTGDS

prostaglandin D2 synthase 21kDa (brain)

chr12

7050980

7056099

5119

-2200

-16939

PTPN6

protein tyrosine phosphatase, non-receptor type 6

chr10

134210300

134215799

5499

2348

-18301

PWWP2B

PWWP domain containing 2B

chr10

100173400

100176299

2899

128

-31527

PYROXD2

pyridine nucleotide-disulphide oxidoreductase domain 2

chr3

49140100

49143649

3549

296

-8509

QARS

glutaminyl-tRNA synthetase

chr4

17512050

17515299

3249

182

-25655

QDPR

quinoid dihydropteridine reductase

chr3

49129400

49133149

3749

229

-64131

QRICH1

glutamine-rich 1

chr11

67156500

67160899

4399

-723

-7182

RAD9A

RAD9 homolog A (S. pombe)

chr9

135987050

135999829

12779

3121

-20332

RALGDS

ral guanine nucleotide dissociation stimulator

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr7

45195050

45197099

2049

-1292

-27772

RAMP3

receptor (G protein-coupled) activity modifying protein 3

chr22

20103200

20105499

2299

-674

-10354

RANBP1

RAN binding protein 1

chr12

48154830

48160049

5219

-4551

-28985

RAPGEF3

Rap guanine nucleotide exchange factor (GEF) 3

chr11

64510600

64517099

6499

-2220

-19466

RASGRP2

RAS guanyl releasing protein 2 (calcium and DAG-regulated)

chr19

49246330

49251199

4869

-4795

-24922

RASIP1

Ras interacting protein 1

chr11

63684250

63691129

6879

-3374

-8986

RCOR2

REST corepressor 2

chr9

116324450

116338849

14399

4289

-28367

RGS3

regulator of G-protein signaling 3

chr16

118550

130449

11899

-1871

-16441

RHBDF1

rhomboid 5 homolog 1 (Drosophila)

chr19

14141200

14144229

3029

3698

932

RLN3

relaxin 3

chr9

140116700

140119679

2979

-2415

-3483

RNF208

ring finger protein 208

chr14

24613880

24617999

4119

-719

-13930

RNF31

ring finger protein 31

chr11

124765550

124772199

6649

-1044

-14760

ROBO4

roundabout, axon guidance receptor, homolog 4 (Drosophila)

chr1

151803080

151806949

3869

-667

-26467

RORC

RAR-related orphan receptor C

chr8

146015880

146019829

3949

-126

-2700

RPL8

ribosomal protein L8

chr11

803650

807549

3899

-4336

-7276

RPLP2

ribosomal protein, large, P2

chr11

64120350

64128099

7749

-2400

-15461

RPS6KA4

ribosomal protein S6 kinase, 90kDa, polypeptide 4

chr3

9884280

9887099

2819

-6156

RPUSD3

RNA pseudouridylate synthase domain containing 3

chr6

42987230

42990699

3469

-420

-8372

RRP36

ribosomal RNA processing 36 homolog (S. cerevisiae)

chr22

42980000

42982049

2049

-3008

-11757

RRP7BP

ribosomal RNA processing 7 homolog B (S. cerevisiae), pseudogene

chr1

151339980

151346099

6119

2124

-6259

SELENBP1

selenium binding protein 1

chr2

97535400

97542429

7029

-3180

-13441

SEMA4C

sema domain, immunoglobulin domain (Ig), transmembrane domain (TM) and short cytoplasmic domain, (semaphorin) 4C

chr17

2238380

2242929

4549

-153

-43692

SGSM2

small G protein signaling modulator 2

chr9

130538250

130540529

2279

1658

-38793

SH2D3C

SH2 domain containing 3C

chr4

2787230

2795149

7919

-3560

-51632

SH3BP2

SH3-domain binding protein 2

chr19

457930

462449

4519

806

-43605

SHC2

SHC (Src homology 2 domain containing) transforming protein 2

chr3

48538100

48542099

3999

1561

-30903

SHISA5

shisa family member 5

chr11

65403430

65410199

6769

-777

-11576

SIPA1

signal-induced proliferationassociated 1

chr17

80185000

80189079

4079

-3104

-10328

SLC16A3

solute carrier family 16 (monocarboxylate transporter), member 3

chr17

79676980

79679049

2069

-1356

-10026

SLC25A10

solute carrier family 25 (mitochondrial carrier; dicarboxylate transporter), member 10

chr3

48933350

48937749

4399

852

-41193

SLC25A20

solute carrier family 25 (carnitine/ acylcarnitine translocase), member 20

chr1

16061050

16063329

2279

-619

-5694

SLC25A34

solute carrier family 25, member 34

chr2

220487900

220503449

15549

3383

-11026

SLC4A3

solute carrier family 4 (anion exchanger), member 3

169

PART THREE CHAPTER 8

Supplementary Table 5. Continued chr11

170

74870080

74873249

3169

820

-45779

SLCO2B1

solute carrier organic anion transporter family, member 2B1

chr1

214448850

214457949

9099

-1165

-57077

SMYD2

SET and MYND domain containing 2

chr10

88712330

88722199

9869

-1023

-5752

SNCG

synuclein, gamma (breast cancer-specific protein 1)

chr16

29299600

29304949

5349

-266

-83321

SNX29P2

sorting nexin 29 pseudogene 2

chr12

53494300

53500099

5799

-74

-21122

SOAT2

sterol O-acyltransferase 2

chr20

62683150

62687899

4749

-4546

-6444

SOX18

SRY (sex determining region Y) -box 18

chr17

4868900

4879399

10499

-3018

-11627

SPAG7

sperm associated antigen 7

chr9

35810700

35815449

4749

-816

-5291

SPAG8

sperm associated antigen 8

chr17

48619730

48621949

2219

-3722

-12371

SPATA20

spermatogenesis associated 20

chr16

1828450

1830549

2099

3081

-2786

SPSB3

splA/ryanodine receptor domain and SOCS box containing 3

chr6

43137430

43143849

6419

1720

-8603

SRF

serum response factor (c-fos serum response element-binding transcription factor)

chr9

140081500

140084979

3479

186

-1582

SSNA1

Sjogren syndrome nuclear autoantigen 1

chr7

116588580

116591829

3249

-3176

-273750

ST7

suppression of tumorigenicity 7

chr7

116588580

116591829

3249

4183

2298

ST7-AS1

ST7 antisense RNA 1

chr7

116588580

116591829

3249

-3748

-9662

ST7-OT4

ST7 overlapping transcript 4

chr3

52525050

52530899

5849

-1381

-30534

STAB1

stabilin 1

chr19

1204180

1210999

6819

1792

-20844

STK11

serine/threonine kinase 11

chr17

79978800

79981299

2499

723

-3470

STRA13

stimulated by retinoic acid 13

chr14

25514800

25521199

6399

1095

-236692

STXBP6

syntaxin binding protein 6 (amisyn)

chr14

53196100

53198649

2549

492

-44328

STYX

serine/threonine/tyrosine interacting protein

chr7

854830

858799

3969

563

-57724

SUN1

Sad1 and UNC84 domain containing 1

chr22

24572450

24574579

2129

-3929

-11559

SUSD2

sushi domain containing 2

chr19

15216650

15220329

3679

276

-7299

SYDE1

synapse defective 1, Rho GTPase, homolog 1 (C. elegans)

chr10

75407750

75417129

9379

-1653

-7795

SYNPO2L

synaptopodin 2-like

chr11

61343900

61345999

2099

3348

-62162

SYT7

synaptotagmin VII

chr22

20007100

20014429

7329

2134

-42682

TANGO2

transport and golgi organization 2 homolog (Drosophila)

chr16

30381630

30385179

3549

-1883

-14983

TBC1D10B

TBC1 domain family, member 10B

chr17

45769930

45775129

5199

-100

-16897

TBKBP1

TBK1 binding protein 1

chr16

2020650

2023879

3229

201

-6486

TBL3

transducin (beta)-like 3

chr20

62683150

62687899

4749

-2914

-18174

TCEA2

transcription elongation factor A (SII), 2

chr12

117534350

117536479

2129

1836

-58685

TESC

tescalcin

chr11

64070830

64074629

3799

4867

492

TEX40

testis expressed 40

chr17

76917250

76924629

7379

532

-71877

TIMP2

TIMP metallopeptidase inhibitor 2

chr1

32033050

32056249

23199

2564

-8635

TINAGL1

tubulointerstitial nephritis antigen-like 1

chr16

84536800

84540049

3249

-137

-28457

TLDC1

TBC/LysM-associated domain containing 1

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 5. Continued chr19

3027000

3031149

4149

-31438

TLE2

transducin-like enhancer of split 2

chr17

76122080

76127979

5899

-169

-16031

TMC6

transmembrane channel-like 6

chr17

76122080

76127979

5899

-1829

-14019

TMC8

transmembrane channel-like 8

chr7

75615600

75625479

9879

3452

-4384

TMEM120A

transmembrane protein 120A

chr12

122149000

122153979

4979

832

-68482

TMEM120B

transmembrane protein 120B

chr7

142981380

142987279

5899

2267

-810

TMEM139

transmembrane protein 139

chr9

139685050

139687199

2149

348

-1643

TMEM141

transmembrane protein 141

chr4

153603800

153605849

2049

-3634

-57554

TMEM154

transmembrane protein 154

chr19

47548900

47552299

3399

1282

-1432

TMEM160

transmembrane protein 160

chr1

29448600

29453049

4449

-418

-4885

TMEM200B

transmembrane protein 200B

chr1

9647400

9657229

9829

3338

-12691

TMEM201

transmembrane protein 201

chr14

67979550

67982249

2699

1121

-43916

TMEM229B

transmembrane protein 229B

chr1

45138330

45141879

3549

-6

-20603

TMEM53

transmembrane protein 53

chr1

16064180

16066199

2019

-3727

-9285

TMEM82

transmembrane protein 82

chr1

1361650

1363849

2199

1242

-416

TMEM88B

transmembrane protein 88B

chr16

427600

430899

3299

2700

-8473

TMEM8A

transmembrane protein 8A

chr3

46741050

46744679

3629

-9548

TMIE

transmembrane inner ear

chr14

103586900

103589979

3079

-4224

-15336

TNFAIP2

tumor necrosis factor, alpha-induced protein 2

chr19

4638100

4640149

2049

-405

-14827

TNFAIP8L1

tumor necrosis factor, alpha-induced protein 8-like 1

chr1

1149280

1152279

2999

-1268

-4073

TNFRSF4

tumor necrosis factor receptor superfamily, member 4

chr11

1859750

1865449

5699

1168

-310

TNNI2

troponin I type 2 (skeletal, fast)

chr11

1940500

1949349

8849

4126

-15010

TNNT3

troponin T type 3 (skeletal, fast)

chr9

140167930

140173429

5499

-1600

-6412

TOR4A

torsin family 4, member A

chr1

3660900

3663879

2979

1496

-9840

TP73-AS1

TP73 antisense RNA 1

chr11

64083730

64089799

6069

-1732

-2597

TRMT112

tRNA methyltransferase 11-2 homolog (S. cerevisiae)

chr8

125461880

125464499

2619

142

-2074

TRMT12

tRNA methyltransferase 12 homolog (S. cerevisiae)

chr22

20103200

20105499

2299

418

-4952

TRMT2A

tRNA methyltransferase 2 homolog A (S. cerevisiae)

chr14

103992800

103997049

4249

-584

-8483

TRMT61A

tRNA methyltransferase 61A

chr11

842750

847999

5249

929

-21741

TSPAN4

tetraspanin 4

chr17

2238380

2242929

4549

-14663

TSR1

TSR1, 20S rRNA accumulation, homolog (S. cerevisiae)

chr7

2665930

2668029

2099

-4623

-37447

TTYH3

tweety family member 3

chr11

8106800

8108899

2099

4941

-19803

TUB

tubby bipartite transcription factor

chr15

41849380

41851899

2519

-592

-20886

TYRO3

TYRO3 protein tyrosine kinase

chr19

56159080

56163729

4649

-4011

-24676

U2AF2

U2 small nuclear RNA auxiliary factor 2

chr17

74439750

74449399

9649

4713

-58961

UBE2O

ubiquitin-conjugating enzyme E2O

chr3

48645280

48647579

2299

668

-9998

UQCRC1

ubiquinol-cytochrome c reductase core protein I

chr3

49156450

49159799

3349

-12017

USP19

ubiquitin specific peptidase 19

chr8

145652030

145657279

5249

-728

-5655

VPS28

vacuolar protein sorting 28 homolog (S. cerevisiae)

chr11

64862300

64865579

3279

257

-15240

VPS51

vacuolar protein sorting 51 homolog (S. cerevisiae)

171

PART THREE CHAPTER 8

Supplementary Table 5. Continued chr3

184532400

184534899

2499

3719

-236752

VPS8

vacuolar protein sorting 8 homolog (S. cerevisiae)

chr6

31742700

31745649

2949

933

-10803

VWA7

von Willebrand factor A domain containing 7

chr6

43478000

43481749

3749

4827

-310

YIPF3

Yip1 domain family, member 3

chr22

20115300

20119029

3729

-2200

-16793

ZDHHC8

zinc finger, DHHC-type containing 8

chr19

58985900

58989149

3249

-270

-5071

ZNF446

zinc finger protein 446

chr20

62598150

62605699

7549

-707

-13870

ZNF512B

zinc finger protein 512B

chr19

55986350

55992199

5849

1576

-6579

ZNF628

zinc finger protein 628

chr16

30795330

30800349

5019

683

-8070

ZNF629

zinc finger protein 629

chr16

30549450

30552249

2799

-4656

-8070

ZNF747

zinc finger protein 747

chr16

30536350

30540449

4099

-490

-3077

ZNF768

zinc finger protein 768

chr7

150079080

150082479

3399

4374

-14939

ZNF775

zinc finger protein 775

chr19

58788250

58791979

3729

-203

-17139

ZNF8

zinc finger protein 8

chr22

29277680

29279749

2069

-1175

-174760

ZNRF3

zinc and ring finger 3

chr7

76020330

76025699

5369

-3826

-48372

ZP3

zona pellucida glycoprotein 3 (sperm receptor)

TSS = transcription start site; TES = transcription end site.

172

CHROMATIN REGULATION IN CARDIOMYOPATHIES

Supplementary Table 6. Gene enrichment analysis related to genes with decreased H3K27ac occupancy within 5kb from transcription start site (TSS) GO: Biological Process No. ID

Name

p-value

q-value q-value q-value Genes Bonferroni FDR B&H FDR B&Y from Input

Genes in Concluding Annotation remark

GO:0008374

O-acyltransferase activity

1,87E-5

1,96E-2

0,148

mitochondrial metabolism

GO:0016747

transferase activity, 4,59E-5 transferring acyl groups other than amino-acyl groups

4,83E-2

2,41E-2

0,182

208

mitochondrial metabolism

GO: Cellular Component No. ID

Name

p-value

q-value q-value q-value Genes Bonferroni FDR B&H FDR B&Y from Input

Genes in Concluding Annotation remark

mitochondrial matrix

6,76E-6

3,38E-3

339

No. ID

Name

p-value

q-value q-value q-value Genes Bonferroni FDR B&H FDR B&Y from Input

Genes in Concluding Annotation remark

BIOCYC: 142358

isoleucine degradation I 7,55E-6

1,01E-2

5,05E-3

3,93E-2

mitochondrial metabolism

REACTOME: 477135

Metabolism

1,75E-5

2,35E-2

7,82E-3

6,08E-2

1575

mitochondrial metabolism

WikiPathways: Fatty Acid Beta 198865 Oxidation

8,80E-7

1,18E-3

9,17E-3

mitochondrial metabolism

GO:0005759

3,38E-3

0,023

mitochondrial metabolism

Pathway

GO = Gene Ontology; FDR B&H = False discovery rate, Benjamini and Hochberg concept; FDR B&Y = False discovery rate, Benjamini and Yekutieli concept.

173

PART THREE

Elucidating (Epi)genetic and Translating

Therapeutic Pathways

Chapter

Cell Therapy, a Novel Remedy for Dilated Cardiomyopathy? A Systematic Review Published as J Card Fail. 2013;19:494-502

Johannes M.I.H. Gho1, Gijs J.M. Kummeling1,2, Stefan Koudstaal1,2, Sanne J. Jansen of Lorkeers1, Pieter A. Doevendans1,2, Folkert W. Asselbergs1, Steven A.J. Chamuleau1,2 1

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

Interuniversity Cardiology Institute of the Netherlands (ICIN), Utrecht, the Netherlands

PART THREE CHAPTER 9

ABSTRACT Background Dilated cardiomyopathy (DCM) is the most common form of non-ischemic cardiomyopathy worldwide and can lead to sudden cardiac death and heart failure. Despite ongoing advances made in the treatment of DCM, the improvement of outcome remains problematic. Stem cell therapy has been extensively studied in preclinical and clinical models of ischemic heart disease showing potential benefit. Dilated cardiomyopathy is associated with a major health burden and few studies have been performed on cell therapy for DCM. In this systematic review we aim to provide an overview of performed preclinical and clinical studies on cell therapy for DCM. Methods and Results A systematic search, critical appraisal and summarized outcomes are presented. In total 29 preclinical and 15 clinical studies were included. Methodological quality of reported studies in general was low based on the Centre for Evidence Based Medicine (CEBM) criteria, Oxford University. A large heterogeneity in inclusion criteria, procedural characteristics and used outcome measures was noted. The majority of studies showed a significant increase in left ventricular ejection fraction after cell therapy during follow-up. Conclusions Stem cell therapy has shown moderate but significant effects in clinical trials for ischemic heart disease, but it still remains to be determined if we can extrapolate these results to DCM patients. There is a need for methodological sound studies to elucidate underlying mechanisms and translate these into improved therapy for clinical practice. To validate safety and efficacy of cell therapy for DCM, adequate randomized (placebo) controlled trials using different strategies are mandatory.

176

CELL THERAPY FOR DILATED CARDIOMYOPATHY

INTRODUCTION Dilated cardiomyopathy (DCM) is the most common form of non-ischemic cardiomyopathy worldwide.1 Estimated prevalence in adults is 1/2,500 individuals, with an incidence of 7/100,000 per year. It is defined by dilatation and impaired contraction of heart muscle, in the absence of an obvious etiology such as ischemic heart disease or valve disease. Causes are multifactorial including genetics as major contributor. Dilated cardiomyopathy can lead to sudden cardiac death and heart failure, and carries a heavy burden on health care resources because of a high rate of hospital admission and occasional need for ventricular assist device and/or heart transplantation. Over the last years progress has been made in treating DCM, but improving outcome remains a difficult goal to achieve. Stem cell therapy for cardiac repair has been extensively studied in acute myocardial infarction and chronic ischemic heart disease showing potential benefit.2-4 Proposed working mechanisms include for example transdifferentiation of stem cells to cardiomyocytes, activation of intrinsic cardiac stem cells, paracrine effects and angiogenesis. Only a few studies have been reported on cell therapy for DCM, with first-in-man trials since 2006. To the best of our knowledge, no complete overview has been reported so far. In this systematic review, preclinical and clinical studies on cell therapy for DCM are reported including validity, procedural characteristics, safety and outcome measures.

METHODS Data collection and analysis We searched PubMed and Embase for reports of preclinical (all years) and clinical (2006-2012) therapeutic studies on various types of cell therapy for non-ischemic DCM (search syntaxes in Supplementary data, Table S1-2). In order to prevent selection and retrieval bias, outcome was not included in the search syntaxes. Only English and German articles were included, no

restrictions were imposed on timing of assessment and follow-up. Excluded were abstracts, reviews, reports from scientific sessions and discussions, case reports, in vitro studies, studies only on ischemic heart disease, pediatric studies and articles describing study design. Retrieved studies were used for cross-reference checking and carefully examined to exclude potentially duplicate or overlapping data. Reviews and case reports were available for background information. A separate search was performed on ClinicalTrials.gov for ongoing trials (Supplementary Table S3). Authors were contacted for missing data. Assessment of quality and risk of bias For assessment of quality, risk of bias (e.g., selection or information bias) and heterogeneity, remaining articles were considered for critical appraisal. Quality of studies were assessed in duplicate (by G.K. and J.G.) using standard forms for relevance and validity according to the Centre for Evidence Based Medicine, Oxford University (CEBM) criteria.5 Study design was determined, adequate randomization (concealed allocation) or comparable groups when not randomized, losses to follow-up and blinding of physician/treatment officer or researcher were reviewed.

177

PART THREE CHAPTER 9

Data extraction and management Study characteristics and used outcome measures were summarized. If necessary, data were estimated graphically or calculated with standard deviations from available data using SPSS 17.0 (SPSS, Chicago, IL, USA). Two-tailed p-values <0.05 were considered significant.

RESULTS Preclinical studies Literature search Our preclinical search yielded 1455 unique articles (Figure 1). After screening of title and abstract using in- and exclusion criteria 29 relevant articles remained, reference checking retrieved no new articles. Study characteristics and quality assessment Quality assessment of included studies is depicted in Supplementary data (Table S4). In summary, around half (15/29) was conducted in a randomized fashion, one study (Jin et al.) reported allocation concealment.6 One study reported blinding of the treatment officer.7 Twelve studies reported blinding of the effect investigators for treatment. All relevant data of the 29 studies are depicted in Table 1. Two studies used a large animal model (Hata et al.: dogs, Psaltis et al.: sheep), all other studies used rodent or rabbit models. Delivery route varied between

Figure 1. Flowchart of search strategy preclinical studies. ENG = English GER = German DCM = dilated cardiomyopathy.

178

CELL THERAPY FOR DILATED CARDIOMYOPATHY

intracoronary or intravenous infusion, intramuscular injection, sheet-graft or surgical injections and in total 10 different stem cell types were used. Fractional shortening, fractional area change and ejection fraction were determined by (2D)-echocardiography in almost all studies, only Psaltis et al. used MRI to determine ejection fraction. Delta pressure/delta time (dP/dt) and peak systolic pressure were measured using hemodynamic measurements. Five studies reported baseline data, showing no significant differences in the study groups. Twenty-one out of 29 studies had sufficient and comparable proportions of follow-up and length of follow-up varied between 2-21 weeks. Lin et al., Sun et al. and Kondoh et al. (2006 and 2007) reported interim results on ejection fraction or fractional shortening. Most studies (24/29) used comparable treatment methods between control and therapy groups. Eighteen out of 29 studies (62%) were published in a journal with impact factor ≥ 3.00. Efficacy and safety in animal models All animal studies demonstrated a positive effect of stem cell therapy on left ventricular function, of which 24 studies showed significant improvement, ranging from 3% to 24% on echocardiographic evaluation (Table 1). In general, no safety issues were found in these studies. Of 18 studies that assessed mortality, 3 studies found improvement in survival.8-10 Ten studies found equal mortality between treatment and control groups, other studies did not mention significance and no cardiac tamponade was reported. Four studies detected no arrhythmias, other studies did not mention any occurrence of arrhythmias.10-13 Clinical studies Literature search Our clinical search strategy yielded 339 unique hits (Figure 2). Screening of title and abstract using in- and exclusion criteria resulted in 16 relevant articles. One relevant peer-reviewed article was found not indexed on PubMed or Embase.14 All principle investigators from ongoing trials were contacted for up-to-date trial results (see ongoing clinical trials section), this resulted in 2 additional complete manuscripts.15, 16 Reference checking retrieved no new articles. Four articles

were excluded because of overlapping data. In total 15 articles were considered for critical appraisal (4 RCTs, 2 non-randomized controlled trials, 9 cohort studies). Authors were contacted for articles not full-text available and missing data (8 authors responded). Study characteristics and quality assessment Four out of 15 studies were randomized, 2 non-randomized controlled trials used matched control groups (Table 2). Data in Supplementary material shows further methodological quality of the included studies (Table S5). Allocation concealment was not reported in the RCTs. There was no patient or physician blinding performed in these clinical studies, nor placebo used in the controlled trials. Two of the controlled trials described equally treated groups aside from the allocated treatment.17, 18 Losses to follow-up were reported and adequate (≤20%) in 13 out of 15 studies. All articles were published in peer-reviewed journals, 5 of the 15 studies (33%) in a journal with impact factor ≥ 3.00. Sample size ranged from 3-131 persons and overall follow-up duration from 3-60 months.

179

180 67/20

35/18 120/40

Lu (2006) 46

41/11 26/14

Pouly (2004) 52

Ohno (2002)

25/16

3 ?/15

30/10

Nomura (2006) 50

Nakajima (2008)

Nagaya (2005) 48

Mu (2011)

40/8

Lin (2010)

100/24

85/57

Kondoh (2007) 44

20/10

Kondoh (2006) 9

52/18

38/14

Jin (2010) 6

Ishida (2004)

Garbade (2009) 42

31/15

Dhein (2006) 7

19/6 160/?

De Angelis (2010) 8

24/8

Chen (2010) 41

15/7

12/5

n= (total/cell treated)

Chen (2006) 13

Model

Baba (2007) 11

Small animals

Psaltis (2010)

Hata (2006) 29

Large animals

Author

Table 1. Preclinical study characteristics and outcomes

Yes

Randomized

Follow-up (weeks)

MDPC

BMMC

MSC

BM-MSC

BMMC

ADMSC

BMMC

BM-MSC

CPC

BM-MSC

BMMC

MPC

Cell type

FAC

PSP

dP/dT

dP/dt

dP/dT

Outcome

BF, rel

CT, abs

BF, abs

CT, abs

CT, rel

CT, abs

CT, rel

CT, abs

CT, rel

CT, abs

CT, rel

CT, abs

CT, rel

CT, abs

Measuring method

+ 24%

≈ + 68%

≈ + 5%

+ 3.1%

≈ + 11%

+ 32%/- 18%

+ 11%

+ 4,1%

+ 31%

≈ + 5%

+ 35%/- 32%

+ 4.1%

+ 3%

+ 62%/- 88%

≈ + 24%

+ 6.4%

+ 11.1%

+ 27%/- 20%

+ 10%

+ 6.5%

Measured effect

0.019

<0.001

<0.05

<0.0001

N.S.

<0.05

N.S.

<0.01

<0.05

“significant”

<0.05

P-value

PART THREE CHAPTER 9

25/12 118/8

Werner(2005) 57

Zhou (2007)

42/18 20/10

28/10

13 4 15 4 21 4 12

Yes No No No Yes Yes Yes

Determined using MRI. bAverage difference in PSP c Singla used 6-8 animals per group, 1 group underwent cell therapy.

MSC

BMMC

EPC

BMMC

CM or ES

dP/dT

Cell type ADMSC: adipose-derived stem cells BM(M)C: bone marrow (mononuclear) cells CM: conditioned medium CPC: cardiac progenitor cells EPC: endothelial progenitor cells ES: embryonic stem cells MB: myoblasts MDPC: myosphere-derived progenitor cell MPC: mesenchymal precursor cells MSC: mesenchymal stem cells

4 2

No No

Abbreviations: N.S. = non-significant EF = ejection fraction dP/dt = delta pressure/delta time FS = fractional shortening FAC = fractional area change PSP = peak systolic pressure.

Disease model 1 porcine myosin + Freund’s adjuvant 2 genetic δ-sarcoglycan deficiency 3 doxorubicin or adriamycin 4 pacing-induced DCM (230/min) 5 BIO TO-2 hamsters 6 Kir 6.2 knockout + aortic constriction

Zhang (2005) 58

Yamada (2008)

?/?

Suzuki (2001) 56

Tezuka (2008)

16/8

Sun (2009) 55

?/? c

Singla (2012) 54

33/15

Shabbir (2009) 53 CT, abs

≈ + 13%

+ 0.1%

+ 3.8%

+ 12.5%

+ 15%

+ 15%/- 37%

+ 20%/- 20%

+ 5.5%

+ 5%

<0.001

N.S.

<0.05

unknown

<0.05

0.01

0.0002

<0.05

≈ italic = estimated effect based on graphics. dP/dt: measured effect (maximum / minimum).

abs = absolute difference rel = relative difference

Measuring method BF = baseline vs follow-up CT = control vs therapy

CT, abs

BF, abs

BF, rel

CT, rel

CT, abs

CELL THERAPY FOR DILATED CARDIOMYOPATHY

181

PART THREE CHAPTER 9

Figure 2. Flowchart of search strategy clinical trials. ENG = English GER = German DCM = dilated cardiomyopathy.

Definitions used for DCM were different between studies. Seven studies reported excluding patients with coronary artery disease on coronary angiography.14, 16, 18-22 Only Schannwell et al. performed endomyocardial biopsy in all patients to exclude myocarditis, Seth et al. performed biopsies in 8/85 patients. Three studies included patients with ischemic cardiomyopathy in a separate group, these subgroups were excluded for this review.14, 23, 24 For Fatkhutdinov et al., the groups undergoing concomitant (valve) surgery were also excluded. For cell delivery, the majority used an intracoronary route versus a surgical procedure with intra-myocardial injections (11 vs. 4). Cell type was different between studies, mostly bone marrow stem cells were used, two studies used fetal material. Kirillov et al. had one group with prenatal allogenic skeletal myoblasts and one with prenatal mesenchymal stem cells.25 Efficacy outcomes in clinical trials Main outcomes reported in clinical trials were left ventricular ejection fraction (LVEF), New York Heart Association (NYHA) class, Quality of Life (QoL) and mortality. All but one article mainly used (2D)-echocardiography for determining ejection fraction, Fischer-Rasokat et al. used angiography.26 Cardiac magnetic resonance imaging (MRI) was only used in certain subgroups. Mean difference in LVEF between baseline and final follow-up values is shown in Table 2 (baseline and final LVEF values are shown in Supplementary material, Table S6). LVEF values were calculated for Arg端ero et al. from available patient data, LVEF was estimated in the therapy groups of Fatkhutdinov et al. and the control groups of Vrtovec et al. (2011 and 2013) as these were only graphically depicted. The majority of studies (10/15) showed a significant increase in LVEF in the therapy group (range 3.2% to 11%). Four studies found no significant difference between baseline and follow-up in the therapy groups.17, 20, 24, 25 Kalil et al. showed a significant

182

0.03 unknown

+ 5.9 + 0.4 + 4.6 ≈-2 + 5.7 ≈-3 ≈0 ≈-5 +9 0 + 8.7

12 60

unknown 40 12 12 12 6 6 ± 13b 12

41 28(61)c 27 55(131)c 55 6(14) 8 10(20) 10 3(5)d 10 3 33 9 9

52 27

controls therapy controls therapy controls therapy controls therapy controls

group 1 group 2

Argüero (2006) 23

Benetti (2010) 19

Chin (2011) 24

Fischer-Rasokat (2009) 26

Kalil (2008) 28

Kirillov (2007) 25

Martino (2010) 20

Ruengsakulrach (2011) 14

Suarez de Lezo (2013) 16

BMC-MN BMMC

Cell type ACP = angiogenic cell precursors ASM = allogenic skeletal myoblasts BM(M)C = bone marrow (mononuclear) cells HFDSC = human fetal-derived stem cells MSC = mesenchymal stem cell MN = mononuclear

ACP

Surgical

0.001

0.03

N.S.

0.023

<0.001

N.S.

0.005

0.0059

N.S.

<0.01

N.S.

unknown

0.02

N.S.

<0.05

0.04

Other abbreviations EF = ejection fraction G-CSF = granulocyte colony-stimulating factor IC = intracoronary N.S. = non-significant ≈ italic = estimated difference based on graphics.

+ 11

+ 4.4

+ 1.2

+ 1.6

prenatal MSC IC

- 0.5

+ 3.2

+ 15.1

+ 8.2

BMC

BM-MSC

HFDSC

autologous CD34+

BMC-MN

allogenic MSC

autologous CD34+

BMC

Surgical

prenatal ASM

Design RCT = randomized controlled trial Non-RCT = non-randomized controlled trial Cohort = cohort study a Randomized to G-CSF administration (14) or G-CSF associated to BMSC intracoronary infusion (8). b Bocchi et al. mean 468±374 days (up to 1190 days), Ruengsakulrach et al. mean 409.7±352.4 days. c All patients received G-CSF before randomization, n = 6 (2011) and n = 21 (2013) patients were excluded in this phase. d 3/5 patients included in preliminary report, outcomes calculated.

Cohort

Schannwell (2008) 18

Fatkhutdinov (2010) 17

Non-RCTs

Vrtovec (2013) 15

Vrtovec (2011) 22

0.016

44(85)

+ 8.8

therapy

Seth (2010) 21

BMC + 5.6

± 15b

Bocchi (2010) 27

RCTs 8

P-value

14(23)a

EF mean difference (%)

controls

Cell type

therapy

Delivery route

Follow up (months)

Group

n(total)

Author (year)

Table 2. Clinical study characteristics and outcomes

CELL THERAPY FOR DILATED CARDIOMYOPATHY

183

PART THREE CHAPTER 9

decrease of 0.5% (p = 0.023) in mean ejection fraction at final follow up (12 months), in this study LVEF did increase in the first two months after surgical cell therapy but decreased afterwards. In the control groups increase of mean LVEF from baseline was less, or LVEF decreased (range ≈ -5% to 5.6%). The highest increase was found in the control group of Bocchi et al., which underwent G-CSF administration. Seth et al. showed a significant difference (p = <0.05) between therapy and controls (LVEF: 5.9% vs. 0.4%) and Vrtovec et al. found a significant change in LVEF mean difference in both studies.15, 21, 22 NYHA class improved in all 9 studies which reported functional class, with significant improvements in more than half of them.14, 17, 19, 20, 27 Quality of life was determined by using questionnaires and was reported to be significantly improved in six studies.14, 19-21, 27, 28 Fatkhutdinov et al. found an improvement in QoL at three months during follow up, but at final follow up a non-significant difference. Safety outcomes in clinical trials Mortality was reported in all studies except for Bocchi et al. Comparing therapy and control groups, Vrtovec et al. (2011) found a significantly lower incidence of heart transplantation or mortality in the therapy group (7% vs. 30%, p = 0.03). At 5-year follow-up, Vrtovec et al. (2013) found a lower total mortality in patients receiving stem cell therapy compared to controls (14% vs. 35%, p = 0.01), occurrence of heart transplantation did not differ between the 2 groups. The other controlled studies found similar mortality rates. Chin et al., Fatkhutdinov et al. and Schannwell et al. found no mortality in both groups and of the 85 patients included by Seth et al. 12 treated patients vs. 14 controls had died after 3 years of follow-up. In the study by Argüero et al., which mainly included patients with ischemic cardiomyopathy, two patients died because of intractable arrhythmias, but there was no information given on to which group these patients belonged. Benetti et al., Kalil et al. and Martino et al. reported some cases with de novo (supra) ventricular arrhythmias, quickly terminated with chemical or electrical cardioversion. The other studies did not show an increase in arrhythmias (only Seth et al. gave no information on occurrence of arrhythmias).

DISCUSSION In total 29 preclinical and 15 clinical studies were included in this systematic review on cell therapy for DCM. Main findings were: (i) a low methodological quality in general; (ii) a large heterogeneity in inclusion criteria, procedural characteristics and used outcome measures; (iii) only 2 preclinical studies involving large animals have been reported; (iv) the majority of clinical studies (from in total 513 enrolled patients) showed a significant increase in LVEF, improvement in NYHA class and QoL at follow-up after cell therapy; (v) several reports on potentially lifethreatening arrhythmias; (vi) there is no increased mortality after cell therapy. From “bench-to-bedside” Study validity in preclinical studies showed a lack in methodological quality, around half was randomized and only one study used concealed allocation. No blinding was performed in the

184

CELL THERAPY FOR DILATED CARDIOMYOPATHY

majority of included articles. We found a large heterogeneity between studies in disease models, therapy and used endpoints. It is therefore difficult to combine the results in order to infer a stronger conclusion on stem cell therapy. However, all studies did show a positive effect on left ventricular function, mostly by attenuation of deterioration, also shown in the studies reporting interim results (Figure 3). Largest improvement was found by De Angelis et al. using CPCâ&#x20AC;&#x2122;s on doxorubicine-induced cardiomyopathy in rats. Contractility was also significantly and positively influenced. Comparing stem cell treated animals with controls, 10 out of 29 studies reported equal mortality or sometimes even improvement of survival was reported. No conclusion can be drawn regarding arrhythmias, while it was reported in only 4 studies. Main proposed mechanisms of stem cell effects on DCM include reduction of fibrosis and angiogenesis by direct (cell engraftment, regenerative transdifferentiation) and paracrine support.29, 30 No additional fibrosis is seen following therapy, sometimes even attenuation is witnessed. It remains possible that publication bias and methods of measurement attributed to an overestimation and positive image of stem cell therapy for DCM. Left ventricular (LV) hemodynamics and MRI are more objective than 2D-echocardiography in measuring LV-function. An optimal animal model for DCM is still lacking. Overall, stem cell therapy seems feasible and effective in preclinical models, however large animal studies are scarce in the transition from bench-to-bedside. Back from â&#x20AC;&#x153;bedside-to-benchâ&#x20AC;? Quality assessment, safety and efficacy in clinical trials In 2006 consensus of the task force of the European Society of Cardiology (ESC) addressed the possible need for autologous cell therapy in DCM.31 In the same year first-in-man clinical trials were reported and in summary these studies were predominantly pioneering trials assessing safety and efficacy. There is a lack in methodological quality, as mainly non-randomized trials were performed. Lack of allocation concealment, placebo effect or procedural characteristics such as delivery techniques and growth factor administration could have influenced the outcome. Especially studies with a short follow-up period should be interpreted with caution whether the positive outcome would be sustained over a longer period. We noticed a large heterogeneity in

inclusion criteria, procedural characteristics and used outcome measures.

Figure 3. Preclinical studies reporting interim results on fractional shortening (A) or left ventricular ejection fraction (B). FS = fractional shortening LVEF = left ventricular ejection fraction.

185

PART THREE CHAPTER 9

These studies showed no increase in mortality in the cell therapy groups. There were several reports on potentially life-threatening arrhythmias.19, 20, 23, 28 Furthermore, a case report (n=2) describing a pilot trial on cell therapy in DCM did show severe complications after intracoronary balloon inflation, with hypotension and ventricular arrhythmias in one patient and severe spasm with hypotension requiring bailout stenting in another, therefore this trial was prematurely terminated.32 So these procedures do carry an arrhythmic risk, which should be assessed in future studies. There seems to be a positive effect of cell therapy for DCM on LVEF, NYHA class and QoL, although this should be verified in large, randomized trials. Also more studies are needed for determining optimal delivery route and cell type for cardiac repair. Outcomes should be given as absolute differences, in order to more accurately estimate actual treatment effect. This would result in clear indications for cell therapy in non-ischemic cardiomyopathy, preferable above random patient selection. Using multiple “rising” imaging modalities such as MRI, (3D-) echocardiography and single-photon emission computed tomography (SPECT) would provide more accuracy for determining outcomes.33 Although we would prefer MRI to determine outcome measures, certain heart failure patients would be ineligible because of implanted devices such as implantable cardioverter-defibrillator (ICD) and cardiac resynchronization therapy (CRT-D). Besides clinically relevant LVEF measurements, detailed measuring of local cardiac function can provide information about local changes induced by cardiac regenerative therapy.34 Combined global and regional myocardial function measurements are recommended to assess the effects of cardiac regenerative therapy. Elucidating underlying mechanisms Current evidence shows controversy on stem cell transdifferentiation to cardiomyocytes contributing to cardiac functional gain, which led to investigating paracrine mechanisms for repair.35, 36 In contrast to cell therapy for ischemic cardiomyopathy where angiogenesis plays an important role, in DCM heart muscle disease and myocardial dysfunction poses a different background. However, marked vascular derangements and impaired vasculogenic and angiogenic responses have also been reported in idiopathic DCM.37 Proposed underlying working mechanisms should be differentiated between ischemic cardiomyopathy and DCM, for example by elucidating these issues in preclinical studies. Findings from clinical studies should also be clarified in collaboration with preclinical research. As progress is being made for DCM in unraveling underlying genetic mutations and molecular mechanisms (e.g., calcium metabolism), these could be used in future for tailored cell and/or gene therapy.38, 39 Ongoing clinical trials Several clinical trials on cell therapy for DCM are underway (Table 3). Nine trials were found on ClinicalTrials.gov, see Supplementary data for search syntax (Table S3). All principle investigators were contacted for up-to-date trial results, this resulted in 2 complete manuscripts which have been added to the clinical studies section. Most studies are randomized, only one study is performed double-blind and placebo controlled with another control group receiving G-CSF. Delivery routes are intracoronary, intra-myocardial or transendocardial (mainly using the NOGA® XP/MyoStar™ injection catheter system) and different cell types are used. Primary outcome measures are assessing safety or LVEF and results are expected.

186

RCT

NCT01302171 (REGENERATE-DCM)

NCT01350310 (NOGA-DCM)

NCT00765518 (IMPACT-DCM)

NCT01392625 (POSEIDON-DCM)

NCT01020968

NCT00333827 (MiHeart)

NCT00721045 (Revascor) single-blind

double-blind

open-label

single-blind

double-blind

Blinding

Cell type

BM-MNC vs. Placebo/G-CSF CD34+ CRCs autologous vs. allogenic MSC CRCs BMC MPCs

Delivery route IC IC vs. IM IM TE TE IC TE

safety

LVEF

safety

LVEF (echocardiography)

LVEF (MRI)

Primary outcome measure (modality)

Jul 2011

Feb 2009

Feb 2012

Jul 2015

Feb 2012

Mar 2013

Dec 2012

(Estimated) Completion date

unknown

active, not recruiting

recruiting

completed

recruiting

Status

Abbreviations: RCT = randomized controlled trial IC = intracoronary IM = intra-myocardial TE = transendocardial BM-MNC = bone marrow mononuclear cells G-CSF = granulocyte colony-stimulating factor CRCs = cardiac repair cells BMC = bone marrow cells MPCs = mesynchymal precursor cells MSC = mesenchymal stem cells LVEF = left ventricular ejection fraction MRI = magnetic resonance imaging.

Design

Identifier (short name)

Table 3. Ongoing clinical trials on cell therapy for dilated cardiomyopathy

CELL THERAPY FOR DILATED CARDIOMYOPATHY

187

PART THREE CHAPTER 9

Limitations The large heterogeneity in study characteristics of included reports could have influenced outcome, therefore we examined reasons for heterogeneity using standardized criteria. Differences between animal DCM models and human disease carries limitations when translating from bench-to-bedside. Publication bias cannot be ruled out as smaller studies are prone to publication bias. Estimates of effect are vulnerable to confounding by lack of randomization. Only reported outcomes were included as we did not have access to complete individual data and, accordingly, mean values are provided; authors were contacted in case of missing data. Future implications Only a few studies on cell therapy for DCM have been performed until now and methodological quality in general was low. The majority of studies show a significant positive effect on LVEF. There is a need for methodological sound (pre)clinical studies to elucidate underlying mechanisms and translate these into improved therapy for clinical practice. Clearly described uniform definitions for inclusion should be used in future studies, for example derived from the ESC classification.40 Adequately randomized (placebo) controlled trials are mandatory to validate safety and efficacy of cell therapy for DCM, preferably with long term follow-up. These studies should focus on occurrence of arrhythmias, mortality, optimal cell delivery techniques and cell type using different modern imaging modalities. Stem cell therapy has shown moderate but significant effects in clinical trials for ischemic heart disease, compromising nearly 2000 patients.4 It still remains to be determined if we can extrapolate these results to DCM patients. Nevertheless, this exciting field of research has shown promising results so far, and warrants continuation of ongoing study programs.

188

CELL THERAPY FOR DILATED CARDIOMYOPATHY

REFERENCES 1. 2.

5. 6. 7.

10. 11.

12. 13.

14. 15.

16.

17.

18.

19.

Jefferies JL, Towbin JA. Dilated cardiomyopathy. Lancet. 2010;375:752-62. van der Spoel TI, Jansen of Lorkeers SJ, Agostoni P, van Belle E, Gyongyosi M, Sluijter JP, et al. Human relevance of pre-clinical studies in stem cell therapy: systematic review and metaanalysis of large animal models of ischaemic heart disease. Cardiovasc Res. 2011;91:649-58. Strauer BE, Steinhoff G. 10 years of intracoronary and intramyocardial bone marrow stem cell therapy of the heart: from the methodological origin to clinical practice. J Am Coll Cardiol. 2011;58:1095-104. Zimmet H, Porapakkham P, Porapakkham P, Sata Y, Haas SJ, Itescu S, et al. Short- and longterm outcomes of intracoronary and endogenously mobilized bone marrow stem cells in the treatment of ST-segment elevation myocardial infarction: a meta-analysis of randomized control trials. Eur J Heart Fail. 2012;14:91-105. Centre for Evidence Based Medicine, Oxford University. EBM Tools. Critical appraisal sheets. [14 Feb 2012]; Available from: http://www.cebm.net/index.aspx?o=1157 (14 February 2012). Jin B, Luo XP, Ni HC, Li Y, Shi HM. Cardiac matrix remodeling following intracoronary cell transplantation in dilated cardiomyopathic rabbits. Mol Biol Rep. 2010;37:3037-42. Dhein S, Garbade J, Rouabah D, Abraham G, Ungemach FR, Schneider K, et al. Effects of autologous bone marrow stem cell transplantation on beta-adrenoceptor density and electrical activation pattern in a rabbit model of non-ischemic heart failure. J Cardiothorac Surg. 2006;1:17. De Angelis A, Piegari E, Cappetta D, Marino L, Filippelli A, Berrino L, et al. Anthracycline cardiomyopathy is mediated by depletion of the cardiac stem cell pool and is rescued by restoration of progenitor cell function. Circulation. 2010;121:276-92. Kondoh H, Sawa Y, Miyagawa S, Sakakida-Kitagawa S, Memon IA, Kawaguchi N, et al. Longer preservation of cardiac performance by sheet-shaped myoblast implantation in dilated cardiomyopathic hamsters. Cardiovasc Res. 2006;69:466-75. Yamada S, Nelson TJ, Crespo-Diaz RJ, Perez-Terzic C, Liu XK, Miki T, et al. Embryonic stem cell therapy of heart failure in genetic cardiomyopathy. Stem Cells. 2008;26:2644-53. Baba S, Heike T, Yoshimoto M, Umeda K, Doi H, Iwasa T, et al. Flk1(+) cardiac stem/progenitor cells derived from embryonic stem cells improve cardiac function in a dilated cardiomyopathy mouse model. Cardiovasc Res. 2007;76:119-31. Tezuka A, Kawada T, Nakazawa M, Masui F, Konno S, Nitta S, et al. Which skeletal myoblasts and how to be transplanted for cardiac repair? Biochem Biophys Res Commun. 2008;369:270-6. Chen M, Fan ZC, Liu XJ, Deng JL, Yang Q, Huang DJ. Cell transplantation with a catheter-based approach: an efficient method for the treatment of heart failure with multiple lesions. Cell Prolif. 2006;39:471-7. Ruengsakulrach P, Visudharom K, Chaothawee L, Belkin M. Peripheral Blood Stem Cell for Cardiomyopathy. Bangkok Med J. 2011;2:19-29. Vrtovec B, Poglajen G, Lezaic L, Sever M, Domanovic D, Cernelc P, et al. Effects of Intracoronary CD34+ Stem Cell Transplantation in Nonischemic Dilated Cardiomyopathy Patients: 5-Year Follow-Up. Circ Res. 2013;112:165-73. Suárez de Lezo J, Herrera C, Romero M, Pan M, Suárez de Lezo J, Carmona MD, et al. Functional improvement in patients with dilated cardiomyopathy after the intracoronary infusion of autologous bone marrow mononuclear cells. Rev Esp Cardiol (Engl Ed). 2013;66:450-7. Fatkhutdinov T, D’Yachkov A V, Koroteyev AV, Goldstein DV, Bochkov NP. Safety and efficiency of transplantation of allogenic multipotent stromal cells in surgical treatment of dilatated cardiomyopathy. Bull Exp Biol Med. 2010;149:119-24. Schannwell CM, Köstering M, Zeus T, Brehm M, Erdmann G, Fleissner T, et al. [Humane autologe intrakoronare stammzelltransplantation zur myokardregeneration bei dilatativer cardiomyopathie (NYHA stadium II-IV)]. The Düsseldorf Autologous Bone Marrow Cells in Dilated Cardiomyopathy Trial. ABCD Trial. J Kardiologie. 2008;15:23-30. Benetti F, Penherrera E, Maldonado T, Vera YD, Subramanian V, Geffner L. Direct myocardial implantation of human fetal stem cells in heart failure patients: long-term results. Heart Surg Forum. 2010;13:E31-5.

189

PART THREE CHAPTER 9

20. Martino HF, Oliveira PS, Souza FC, Costa PC, Assuncao ESE, Villela R, et al. A safety and feasibility study of cell therapy in dilated cardiomyopathy. Braz J Med Biol Res. 2010;43:989-95. 21. Seth S, Bhargava B, Narang R, Ray R, Mohanty S, Gulati G, et al. The ABCD (Autologous Bone Marrow Cells in Dilated Cardiomyopathy) trial a long-term follow-up study. J Am Coll Cardiol. 2010;55:1643-4. 22. Vrtovec B, Poglajen G, Sever M, Lezaic L, Domanovic D, Cernelc P, et al. Effects of intracoronary stem cell transplantation in patients with dilated cardiomyopathy. J Card Fail. 2011;17:272-81. 23. Arguero R, Careaga-Reyna G, Castano-Guerra R, Garrido-Garduno MH, Magana-Serrano JA, de Jesus Nambo-Lucio M. Cellular autotransplantation for ischemic and idiopathic dilated cardiomyopathy. Preliminary report. Arch Med Res. 2006;37:1010-4. 24. Chin SP, Poey AC, Wong CY, Chang SK, Tan CS, Ng MT, et al. Intramyocardial and intracoronary autologous bone marrow-derived mesenchymal stromal cell treatment in chronic severe dilated cardiomyopathy. Cytotherapy. 2011;13:814-21. 25. Kirillov AM, Fatkhudinov T, Dyachkov AV, Koroteev AV, Goldshtein DV, Bochkov NP. Transplantation of allogenic cells in the therapy of patients with dilated cardiomyopathy. Bull Exp Biol Med. 2007;144:635-9. 26. Fischer-Rasokat U, Assmus B, Seeger FH, Honold J, Leistner D, Fichtlscherer S, et al. A pilot trial to assess potential effects of selective intracoronary bone marrow-derived progenitor cell infusion in patients with nonischemic dilated cardiomyopathy: final 1-year results of the transplantation of progenitor cells and functional regeneration enhancement pilot trial in patients with nonischemic dilated cardiomyopathy. Circ Heart Fail. 2009;2:417-23. 27. Bocchi EA, Bacal F, Guimaraes G, Mendroni A, Mocelin A, Filho AE, et al. Granulocyte-colony stimulating factor or granulocyte-colony stimulating factor associated to stem cell intracoronary infusion effects in non ischemic refractory heart failure. Int J Cardiol. 2010;138:94-7. 28. Kalil RA, Ott D, Santâ&#x20AC;&#x2122;Anna R, Dias E, Marques-Pereira JP, Delgado-Canedo A, et al. Autologous transplantation of bone marrow mononuclear stem cells by mini-thoracotomy in dilated cardiomyopathy: technique and early results. Sao Paulo Med J. 2008;126:75-81. 29. Hata H, Matsumiya G, Miyagawa S, Kondoh H, Kawaguchi N, Matsuura N, et al. Grafted skeletal myoblast sheets attenuate myocardial remodeling in pacing-induced canine heart failure model. J Thorac Cardiovasc Surg. 2006;132:918-24. 30. Psaltis PJ, Carbone A, Nelson AJ, Lau DH, Jantzen T, Manavis J, et al. Reparative effects of allogeneic mesenchymal precursor cells delivered transendocardially in experimental nonischemic cardiomyopathy. JACC Cardiovasc Interv. 2010;3:974-83. 31. Bartunek J, Dimmeler S, Drexler H, Fernandez-Aviles F, Galinanes M, Janssens S, et al. The consensus of the task force of the European Society of Cardiology concerning the clinical investigation of the use of autologous adult stem cells for repair of the heart. Eur Heart J. 2006;27:1338-40. 32. Widimsky P, Penicka M. Complications after intracoronary stem cell transplantation in idiopathic dilated cardiomyopathy. Int J Cardiol. 2006;111:178-9. 33. Marwick TH, Raman SV, Carrio I, Bax JJ. Recent developments in heart failure imaging. JACC Cardiovasc Imaging. 2010;3:429-39. 34. van Slochteren FJ, Teske AJ, van der Spoel TI, Koudstaal S, Doevendans PA, Sluijter JP, et al. Advanced measurement techniques of regional myocardial function to assess the effects of cardiac regenerative therapy in different models of ischaemic cardiomyopathy. Eur Heart J Cardiovasc Imaging. 2012;13:808-18. 35. Segers VF, Lee RT. Stem-cell therapy for cardiac disease. Nature. 2008;451:937-42. 36. Maxeiner H, Krehbiehl N, Muller A, Woitasky N, Akinturk H, Muller M, et al. New insights into paracrine mechanisms of human cardiac progenitor cells. Eur J Heart Fail. 2010;12:730-7. 37. Roura S, Bayes-Genis A. Vascular dysfunction in idiopathic dilated cardiomyopathy. Nat Rev Cardiol. 2009;6:590-8. 38. Jessup M, Greenberg B, Mancini D, Cappola T, Pauly DF, Jaski B, et al. Calcium Upregulation by Percutaneous Administration of Gene Therapy in Cardiac Disease (CUPID): a phase 2 trial of intracoronary gene therapy of sarcoplasmic reticulum Ca2+-ATPase in patients with advanced heart failure. Circulation. 2011;124:304-13.

190

CELL THERAPY FOR DILATED CARDIOMYOPATHY

39. Sun N, Yazawa M, Liu J, Han L, Sanchez-Freire V, Abilez OJ, et al. Patient-specific induced pluripotent stem cells as a model for familial dilated cardiomyopathy. Sci Transl Med. 2012;4:130ra47. 40. Elliott P, Andersson B, Arbustini E, Bilinska Z, Cecchi F, Charron P, et al. Classification of the cardiomyopathies: a position statement from the European Society Of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J. 2008;29:270-6. 41. Chen Y, Liu W, Li W, Gao C. Autologous bone marrow mesenchymal cell transplantation improves left ventricular function in a rabbit model of dilated cardiomyopathy. Exp Mol Pathol. 2010;88:311-5. 42. Garbade J, Dhein S, Lipinski C, Aupperle H, Arsalan M, Borger MA, et al. Bone marrow-derived stem cells attenuate impaired contractility and enhance capillary density in a rabbit model of Doxorubicin-induced failing hearts. J Card Surg. 2009;24:591-9. 43. Ishida M, Tomita S, Nakatani T, Fukuhara S, Hamamoto M, Nagaya N, et al. Bone marrow mononuclear cell transplantation had beneficial effects on doxorubicin-induced cardiomyopathy. J Heart Lung Transplant. 2004;23:436-45. 44. Kondoh H, Sawa Y, Fukushima N, Matsumiya G, Miyagawa S, Kitagawa-Sakakida S, et al. Combined strategy using myoblasts and hepatocyte growth factor in dilated cardiomyopathic hamsters. Ann Thorac Surg. 2007;84:134-41. 45. Lin YC, Leu S, Sun CK, Yen CH, Kao YH, Chang LT, et al. Early combined treatment with sildenafil and adipose-derived mesenchymal stem cells preserves heart function in rat dilated cardiomyopathy. J Transl Med. 2010;8:88. 46. Lu C, Arai M, Misao Y, Chen X, Wang N, Onogi H, et al. Autologous bone marrow cell transplantation improves left ventricular function in rabbit hearts with cardiomyopathy via myocardial regeneration-unrelated mechanisms. Heart Vessels. 2006;21:180-7. 47. Mu Y, Cao G, Zeng Q, Li Y. Transplantation of induced bone marrow mesenchymal stem cells improves the cardiac function of rabbits with dilated cardiomyopathy via upregulation of vascular endothelial growth factor and its receptors. Exp Biol Med (Maywood). 2011;236:1100-7. 48. Nagaya N, Kangawa K, Itoh T, Iwase T, Murakami S, Miyahara Y, et al. Transplantation of mesenchymal stem cells improves cardiac function in a rat model of dilated cardiomyopathy. Circulation. 2005;112:1128-35. 49. Nakajima H, Sakakibara Y, Tambara K, Marui A, Yoshimoto M, Premaratne GU, et al. Delivery route in bone marrow cell transplantation should be optimized according to the etiology of heart disease. Circ J. 2008;72:1528-35. 50. Nomura T, Ashihara E, Tateishi K, Asada S, Ueyama T, Takahashi T, et al. Skeletal myospherederived progenitor cell transplantation promotes neovascularization in delta-sarcoglycan knockdown cardiomyopathy. Biochem Biophys Res Commun. 2007;352:668-74. 51. Ohno N, Fedak PW, Weisel RD, Mickle DA, Fujii T, Li RK. Transplantation of cryopreserved muscle cells in dilated cardiomyopathy: effects on left ventricular geometry and function. J Thorac Cardiovasc Surg. 2003;126:1537-48. 52. Pouly J, Hagege AA, Vilquin JT, Bissery A, Rouche A, Bruneval P, et al. Does the functional efficacy of skeletal myoblast transplantation extend to nonischemic cardiomyopathy? Circulation. 2004;110:1626-31. 53. Shabbir A, Zisa D, Suzuki G, Lee T. Heart failure therapy mediated by the trophic activities of bone marrow mesenchymal stem cells: a noninvasive therapeutic regimen. Am J Physiol Heart Circ Physiol. 2009;296:H1888-97. 54. Singla DK, Ahmed A, Singla R, Yan B. Embryonic Stem Cells Improve Cardiac Function in Doxorubicin-Induced Cardiomyopathy Mediated through Multiple Mechanisms. Cell Transplant. 2012;21:1919-30. 55. Sun CK, Chang LT, Sheu JJ, Chiang CH, Lee FY, Wu CJ, et al. Bone marrow-derived mononuclear cell therapy alleviates left ventricular remodeling and improves heart function in rat-dilated cardiomyopathy. Crit Care Med. 2009;37:1197-205. 56. Suzuki K, Murtuza B, Suzuki N, Smolenski RT, Yacoub MH. Intracoronary infusion of skeletal myoblasts improves cardiac function in doxorubicin-induced heart failure. Circulation. 2001;104:I213-7.

191

PART THREE CHAPTER 9

57. Werner L, Deutsch V, Barshack I, Miller H, Keren G, George J. Transfer of endothelial progenitor cells improves myocardial performance in rats with dilated cardiomyopathy induced following experimental myocarditis. J Mol Cell Cardiol. 2005;39:691-7. 58. Zhang J, Li GS, Li GC, Zhou Q, Li WQ, Xu HX. Autologous mesenchymal stem cells transplantation in adriamycin-induced cardiomyopathy. Chin Med J (Engl). 2005;118:73-6. 59. Zhou C, Yang C, Xiao S, Fei H. Feasibility of bone marrow stromal cells autologous transplantation for dilated cardiomyopathy. J Huazhong Univ Sci Technolog Med Sci. 2007;27:75-8

192

CELL THERAPY FOR DILATED CARDIOMYOPATHY

SUPPLEMENTAL MATERIAL Table S1. Search syntax preclinical trials (May 1st, 2012) PubMed (DCM OR IDCM OR cardiomyopathy) AND (dog OR canine OR sheep OR ovine OR pig OR pigs OR porcine OR swine OR bovine OR rabbit OR rabbits OR rodents OR rat OR rats OR mouse OR mice OR hamster OR hamsters) AND (“stem cells” OR “stem cell” OR “progenitor cells” OR “progenitor cell” OR “bone marrow” OR transplantation) 897 hits Embase (DCM OR IDCM OR cardiomyopathy) AND (dog OR canine OR sheep OR ovine OR pig OR pigs OR porcine OR swine OR bovine OR rabbit OR rabbits OR rodents OR rat OR rats OR mouse OR mice OR hamster OR hamsters) AND (‘stem cells’ OR ‘stem cell’ OR ‘progenitor cells’ OR ‘progenitor cell’ OR ‘bone marrow’ OR transplantation) 821 hits Total: 1718 hits After removal of duplicates  1455 hits

Table S2. Search syntax clinical trials (February 16th, 2012) PubMed (dilated cardiomyopathy OR DCM) AND (stem cells OR stem cell OR progenitor OR bone marrow) AND (“2006/01/01”[PDat] : “2012/02/16”[PDat]) 158 hits Embase (‘dilated cardiomyopathy’ OR ‘DCM’) AND (‘stem cells’ OR ‘stem cell’ OR ‘progenitor’ OR ‘bone marrow’) AND [2006-2012]/py 291 hits Total: 449 hits After removal of duplicates  339 hits

Table S3. Search ongoing clinical trials (November 16th, 2012) ClinicalTrials.gov Search Terms: ((cardiomyopathy AND (dilated OR idiopathic)) OR DCM) AND (stem OR cells OR cell OR progenitor OR bone marrow) | Adult, Senior

56 hits 49 excluded: - 1 terminated (NCT00612911) - 2 pediatric (NCT01219452, NCT01504594) - 4 ischemic cardiomyopathy (NCT01720888, NCT00545610, NCT00690209, NCT01670981) - 5 completed and published (NCT00284713, NCT00615394, NCT00743639, NCT00629096, NCT00629018) - 1 observational study (NCT00962364) - 1 no cell therapy (NCT00585546) - 1 no DCM (NCT00349271) - 1 cardiac resynchronization therapy (NCT00800657) - 2 gene therapy (NCT01643330, NCT00454818) - 31 others (NCT00958087, NCT00815217, NCT01080781, NCT00943059, NCT00193843, NCT00011648, NCT00785525, NCT01152918, NCT00006301, NCT00368225, NCT00353639, NCT00009568, NCT00575211, NCT00078390, NCT00193765, NCT01480479, NCT00923884, NCT00340717, NCT01675102, NCT00342706, NCT00341497, NCT00483379, NCT00867048, NCT00745953, NCT00923962, NCT00342433, NCT01165515, NCT01661283, NCT01247922, NCT01032070, NCT00433862)  7 ongoing clinical trials

193

PART THREE CHAPTER 9

Table S4. Critical appraisal preclinical studies using CEBM criteria Author

Randomized Comparable Comparable Adequate groups treatment follow-up

Blind investigator

Baba

Chen ‘06

Chen ‘10

+/-

De Angelis

Dhein

+/-

Garbade

Hata

Ishida

+/-

Jin

Kondoh ‘06

+/-

Kondoh ‘07

+/-

Lin

+/-

Nagaya

Nakajima

Nomura

Ohno

Pouly

+/-

Psaltis

+/-

Shabbir

Singla

Sun

+/-

Suzuki

Tezuka

Werner

+/-

Yamada

Zhang

+/-

Zhou

+/-

randomized with concealed allocation +

similar groups Y/N

similar treatment groups in controlled trials besides intervention Y/N

<20% lost to follow up Y/N

blind treatment officer Y/N

blind investigator Y/N

randomized, unclear concealed allocation +/nonrandomized Y = yes N = no

194

Blind treatment officer

CELL THERAPY FOR DILATED CARDIOMYOPATHY

Table S5. Critical appraisal clinical studies using CEBM criteria Author

Randomized

Similarity groups

Similar treatment

Adequate follow-up

Blinding

Arguero

n/a

unknown

Benetti

n/a

Bocchi

+/-

unknown

Chin

n/a

Fatkhutdinov

+/-

Fischer-Rasokat

n/a

Kalil

n/a

unknown

Kirillov

n/a

Martino

n/a

Ruengsakulrach

n/a

Schannwell

+/-

Seth

+/-

unknown

Suarez de Lezo

n/a

Vrtovec (2011)

+/-

unknown

Vrtovec (2013)

+/-

unknown

randomized with concealed allocation +

similar groups Y/N

similar treatment groups in controlled trials besides intervention Y/N

<20% lost to follow up Y/N

blinding of patient and physician +

randomized, unclear concealed allocation +/-

matched controls +/-

nonrandomized -

blinding of patient or physician +/-

no blinding -

Y = yes N = no

195

196 SD

Mean difference

p-value

26.6

21.8

24.5

20.5

30.2

25.9

23.4

24.4

23.8

23.3

22.5

25.5

24.3

Bocchi

Chin

Fatkhudinov

Fischer-Rasokat

Kalil

Kirillov group 1

Kirillov group 2

Martino

Ruengsakulrach

Schannwell

Seth

Suarez de Lezo

Vrtovec (2011)

Vrtovec (2013) 6.5

7.5

8.3

5.3

10.3

6.7

8.2

10.9

1.6

5.3

3.8

4.6

30.0

30.1

28.4

27.7

25.4

33.4

≈ 21

39.6

30.6

34.8

5.1

6.7

11.8

11.3

7.7

7.6

11.1

6.8

11.5

7.7

7.3

7.2

5.3

+ 5.7

+ 4.6

+ 11

+ 5.9

+ 4.4

+ 1.2

+ 1.6

- 0.5

+ 3.2

≈0

+ 15.1

+ 8.8

+ 8.2

+ 8.7

0.02

0.03

0.001

<0.05

<0.01

0.03

N.S.

0.023

<0.001

N.S.

0.016

0.005

0.0059

25.7

26.7

20.8

26.1

21.3

4.1

3.9

9.3

6.0

5.2

≈ 23

≈ 25

21.2

≈ 21

26.9

EF final

9.2

8.3

≈ - 3b

≈ - 2b

+ 0.4

≈-5

+ 5.6b

Mean difference

unknownc

N.S.c

N.S.

0.04

p-value

outcomes calculated b controls with G-CSF administration c significant difference between therapy and controls ≈ italic = estimated effect based on graphics. Abbreviations: EF = ejection fraction SD = standard deviation N.S. = non-significant.

27.3

Benetti

EF baseline

EF final

EF baseline

Controls

Therapy

Argueroa

Author

Table S6. Outcomes of ejection fraction clinical studies

PART THREE CHAPTER 9

197

PART THREE

Elucidating (Epi)genetic and Translating

Therapeutic Pathways

Chapter

Xenotransplantation of Human Cardiomyocyte Progenitor Cells Does Not Improve Cardiac Function in a Porcine Model of Chronic Ischemic Heart Failure Results from a randomized, blinded, placebo controlled trial

Submitted

Sanne J. Jansen of Lorkeers1, Johannes M.I.H. Gho1, Stefan Koudstaal1,2, Gerardus P.J. van Hout1, Peter-Paul M. Zwetsloot1, Joep W.M. van Oorschot3, Esther C.M. van Eeuwijk1, Tim Leiner3, Imo E. Hoefer1, Marie-JosĂŠ Goumans4, Pieter A. Doevendans1,2, Joost P.G. Sluijter1,2, Steven A.J. Chamuleau1,2

Department of Cardiology, Division Heart and Lungs, University Medical Center Utrecht, Utrecht, the Netherlands

ICIN - Netherlands Heart Institute, Utrecht, the Netherlands

Department of Radiology, University Medical Center Utrecht, the Netherlands

Department of Molecular Cell Biology, Leiden University Medical Center, the Netherlands

PART THREE CHAPTER 10

ABSTRACT Background Recently cardiomyocyte progenitor cells (CMPCs) were successfully isolated from fetal and adult human hearts. Direct intramyocardial injection of human CMPCs (hCMPCs) in experimental mouse models of acute myocardial infarction significantly improved cardiac function compared to controls. Aim Here, our aim was to investigate whether xenotransplantation via intracoronary infusion of fetal hCMPCs in a pig model of chronic myocardial infarction is safe and efficacious, in view of translation purposes. Methods & Results We performed a randomized, blinded, placebo controlled trial. Four weeks after ischemia/ reperfusion injury by 90 minutes of percutaneous left anterior descending artery occlusion, pigs (n=16, 68.5 Âą 5.4 kg) received intracoronary infusion of 10 million fetal hCMPCs or placebo. All animals were immunosuppressed by cyclosporin (CsA). Four weeks after infusion, endpoint analysis by MRI displayed no difference in left ventricular ejection fraction, left ventricular end diastolic and left ventricular end systolic volumes between both groups. Serial pressure volume (PV-)loop and echocardiography showed no differences in functional parameters between groups at any timepoint. Infarct size at follow-up, measured by late gadolinium enhancement MRI showed no difference between groups. Intracoronary pressure and flow measurements showed no signs of coronary obstruction 30 minutes after cell infusion. No premature death occurred in cell treated animals. Conclusion Xenotransplantation via intracoronary infusion of hCMPCs is feasible and safe, but not associated with improved left ventricular performance and infarct size compared to placebo in a porcine model of chronic myocardial infarction.

200

CMPCs IN HEART FAILURE

INTRODUCTION The heart has regenerative capacity as it harbours a pool of cardiac stem cells.1 However, this is clearly not sufficient to repair the damage caused by myocardial infarction (MI) to prevent the development of heart failure. The number of stem cells available might just be too little. Ex vivo expansion and reapplication of cardiac stem cells to the injured heart was proposed, however isolation of these cardiac stem cells remains challenging. Our lab succeeded in isolating fetal and adult cardiomyocyte progenitor cells (CMPC) from mouse and human hearts based on the stem cell antigen Sca-1.2 The Sca-1-like positive human CMPCs (hCMPCs) from fetal and adult hearts showed differentiation into spontaneously beating cardiomyocytes after stimulation with 5-azacytidine and TGFβ.3 These hCMPC-derived cardiomyocytes have functional gap junctions, enabling metabolic and electrical coupling of cells.3 Additionally, hCMPCs can also be differentiated into endothelial cells and smooth muscle cells by exposure to vascular endothelial growth factor (VEGF).2–4 Intra-myocardial injection of hCMPCs in a mouse model of acute MI led to engraftment of 3.5% of cells, differentiation towards coupled cardiomyocytes, increased vascular density, and to improved cardiac function.5 These promising results led to the current large animal study, as a next step towards potential clinical application of hCMPCs. Since the greatest burden of ischemic heart disease is based on chronic ischemic heart failure, hCMPCs were tested in a chronic disease model. In the current study, fetal hCMPCs were intracoronary administered in a porcine model of chronic ischemia/reperfusion injury (I/R). We hypothesized that this strategy is safe, improves left ventricular performance and reduces infarct size compared to placebo.

METHODS Experimental design All animal experiments were executed in conformance with the ‘Guide for the Care and Use of Laboratory Animals’. The experiment was evaluated and approved by the Animal Experiments Committee of the Utrecht University, the Netherlands (permit number 2012.II.09.145). Animals were subjected to I/R, randomized to fetal hCMPC or placebo infusion 4 weeks later and endpoint

analyses were performed 4 weeks after stem cell injection (see study protocol, Figure 1). All animals were immunosuppressed by Cyclosporin A (CsA) to facilitate xenogeneic cell treatment. Based on a power calculation (estimated effect 7.5%6, standard deviation of 5%, a power of 0.9 and alpha of 0.05) 8 pigs per group were needed. Animals that died before endpoint analysis were supplemented. Animals were randomized after surviving the initial I/R, using a computer based random order generator. Cell and placebo infusion as well as data analysis were performed in a blinded fashion (investigators, technicians and animal caretakers). Deblinding was performed after collecting and analysing all data. The primary endpoint of this study was defined as left ventricular ejection fraction (EF) at the end of follow-up, measured by magnetic resonance imaging (MRI). Secondary endpoints were left ventricular end diastolic volume and left ventricular end systolic volume (EDV and ESV)

201

PART THREE CHAPTER 10

measured by MRI, infarct size measured by ex vivo gross macroscopy after incubation with triphenyltetrazolium chloride (TTC) and late gadolinium enhancement (LGE) MRI, functional parameters serially measured by pressure volume (PV-)loop and echocardiography, coronary microvascular function by intracoronary pressure- and flow measurements and vascular density and fibrosis on histology. Cell isolation Cells were isolated from fetal human heart tissue (derived after elective abortion with informed consent) and cultured as described before.2 Shortly, tissue was minced, incubated with collagenase and grinded through a cell strainer. Cells were incubated with anti-Sca-1 microbeads and separated using a MiniMACS magnet (Miltenyi Biotec, Leiden, the Netherlands). hCMPCs were collected and dissolved in growth medium containing endothelial growth medium (EGM2, Cambrex, CC-4176), FBS, penicillin/streptomycin, non-essential amino acids and bFGF. After attachment of the cells, cells were split after +/- 3 days at 80-90% confluence in a 1:6 fashion. All pigs received hCMPCs of passage 5-7 from the same donor. Animal experiment A comprehensive description of the protocol is also available at http://www.jove.com/ video/51269.7 Female landrace specific pathogen free pigs (n=19) (van Beek SPF Varkensfokkerij B.V., Lelystad, the Netherlands), weighing 68.5 ± 5.4 kg (Upper and lower limit 60.6 – 82.0 kg) at baseline, were pre-treated with amiodaron for 10 days (1200 mg/day for 10 days, 800 mg/ day maintenance), clopidogrel for 3 days (75 mg/day) and acetylsalicylic acid for 1 day (320 mg loading dose, 80 mg/day maintenance) and a fentanyl patch (25µg/h) for 1 day. One day before cell or placebo delivery, cyclosporin (CsA) was started. Based on clinical organ transplantation protocol, pigs received a loading dose of 800 mg, then 400 mg b.i.d. for 1 week and 200 mg b.i.d for the remaining 3 weeks by oral administration (Neoral drink, 100mg/ml, Novartis Pharma bv). At the day of surgery, one dosage was i.v. infused as 200 mg in 100mL over 2 hours (Sandimmune 50mg/ml, Novartis Pharma bv). All medication, except for the fentanyl patch, was continued until the end of follow-up. Pigs received fiber rich pellets (Abdiets animal nutrition, product 2755, Woerden, The Netherlands) twice a day and water was available ad libitum. Animals were kept fasted the day of surgery (except for medication). Anaesthesia was obtained by intramuscular injection of 10 mg/kg ketamine, 0.4 mg/kg midazolam and 0.5 mg/kg atropine in the cage. Pigs were intubated and transported to the operating theatre. Maintenance anaesthesia consisted of continuous infusion of 0.5 mg/kg/h midazolam, 2.5 µg/ kg/h sufentanyl and 0.1 mg/kg/h pancuronium. Other perioperative medication consisted of 300 mg amiodaron, amoxicillin + clavulanic acid 750/75 mg and heparin (100 IE/kg after positioning the sheaths and 50 IE/kg every 2 hours). Pigs were mechanically ventilated with a positive pressure ventilator with FiO2 0.5, 10ml/kg tidal volume and a frequency of 12/minute under continuous capnography. Arterial access was achieved by cannulating the internal carotid artery with an 8F sheath. Venous access was achieved by cannulating the jugular vein with a 9F sheath. An additional arterial line was inserted in a small peripheral artery in the hind limb for continuous stable arterial pressure registration.

202

CMPCs IN HEART FAILURE

Figure 1. Study protocol

Echocardiography Pigs were positioned in the right lateral position. Parasternal short axis images were obtained at 3 levels (mitral valve, papillary muscle and apex) during 5 beats per level (iE33 ultrasound device Philips, Eindhoven, The Netherlands). Short axis images were analysed offline using Xcelera R2.L1 (Philips Healthcare, Best, The Netherlands) and fractional area shortening (FAS), fractional shortening (FS) and septal wall thickening (WTsept) at the level of the mitral valve, papillary muscle and apex (mitral, pap, apex respectively) were calculated. Because of apical dilatation of the left ventricle after infarct, the modified Simpsons rule was not applicable and no reliable volumes could be calculated. Pressure volume loop measurement Admittance based PV-loop measurements were performed as recently described.8 Cardiac output (CO) was measured three times by thermodilution and stroke volume was calculated out of three measurements. The 7F tetra-polar admittance catheter (7.0 VSL Pigtail/no lumen, Transonic Scisense, London, Canada) was inserted into the left ventricle through the arterial sheath under fluoroscopic guidance. Inferior caval vein occlusion was performed using an 8F Fogarty occlusion catheter (62080814F, Edwards Lifesciences). All measurements were performed during apnea. Data were offline analysed using Iworx analysis software (Labscribe V2.0).

Intracoronary pressure and flow velocity assessment Intracoronary pressure and flow were assessed by positioning a Combowire in the left anterior descending artery (LAD) and the left circumflex coronary artery (LCX) acting as a control subsequently. Intracoronary pressure and flow measurements, together with arterial pressure and ECG, were recorded using the ComboMap system (Volcano Corporation). An intracoronary bolus of 200 Âľg nitroglycerin was administered to prevent coronary spasms. Three successive measurements were recorded in rest and during hyperaemia, achieved by intracoronary administration of 60 Âľg adenosine. Data were analysed offline, using AMC Volcano Studymanager (versus 6.0, Borland Software Corporation and Delphi versus 2010, Embarcadero, San Francisco, CA, USA).

203

PART THREE CHAPTER 10

Ischemia/reperfusion injury After baseline measurements, the diameter of the LAD, distal to the second diagonal branch (D2) was measured in anterior-posterior and left anterior oblique 30° view. Before positioning the balloon, an intracardiac defibrillation catheter was placed in the right ventricle. A PCI-balloon catheter with a suitable diameter was positioned distal from the D2 and inflated for 90 minutes. In case of ventricular arrhythmias, chest compressions were started immediately, amiodaron 300mg was infused intravenously and pigs were defibrillated intracardiac with 50J. In case of multiple unsuccessful shocks, defibrillation was switched to transthoracic defibrillation with 150-200J. Cell infusion Ten million cells were resuspended in 10mL of phosphate buffered saline (PBS), one hour before intracoronary infusion. Cell suspension or placebo (10mL of PBS) was handed to the animal technician in a covered sterile syringe so that the personnel at the operating theatre remained blinded for the treatment allocation. An over-the-wire PCI balloon catheter (Apex, Boston Scientific), was positioned distal to the D2, according to the position of the PCI catheter during infarct creation. The balloon was inflated until pressure matched the coronary diameter. 10mL of cell suspension or placebo was infused in three sessions consisting of 30 seconds infusion of 3.3mL during 2 minutes balloon occlusion followed by 3 minutes of reperfusion. Intracoronary pressure and flow measurements were repeated, 30 minutes after the last reperfusion period, to detect possible flow restrictions caused by cell or placebo infusion. MRI At the end of follow-up, cardiac MRI and all other functional measurements were performed. All MR studies were performed on a 3T MRI scanner (Achieva TX, software release 3.2; Philips Healthcare, Best, The Netherlands). Both cine images and LGE images in short axis and twochamber long axis orientations were obtained under continuous anesthesia and mechanical ventilation. LGE images were obtained at least 15 minutes after injection of 0.2 ml/kg gadobutrol (Gadovist, Bayer Healthcare). Offline imaging analysis for functional measurements was performed in Qmass MR 7.4 enterprise solutions (Medis medical imaging systems BV, Leiden) and by Segment version v1.9R 3293 (Medviso AB, Lund, Sweden) for infarct size. Infarct size at the end of follow-up was analysed based on the 2SD method. Ex vivo Infarct size Animals were sacrificed by exsanguination under general anesthesia. The hearts were excised, washed and cooled with running tap water and the LV was then cut into 5 equal slices from apex to base. Slices were incubated in 1% TTC (Sigma-Aldrich Chemicals, Zwijndrecht, the Netherlands) in 37°C 0.9%NaCl for 15 min to discriminate infarcted tissue from viable myocardium. The infarcted area was calculated as percentage of the left ventricle.

204

CMPCs IN HEART FAILURE

Histology Sections of infarct zone, border zone and remote area of all pigs were isolated and fixated in 4% formalin for at least 1 week and then embedded in paraffin. Sections of 5 µm were stained for endothelial cells by lectin (Lectin fom Bandeiraea simplicifolia, Sigma Aldrich L5391) and for fibrosis by picrosirius red (Sirius red F3B, BDH and picrin acid, Boom. Cat. 12388). For vascular density, 5 random pictures per slide with 40x magnification were taken from the border zone area. The number of vessels was counted manually by two independent observers (S.J. and J.G.) and the absolute number of lectin positive vessels per field were averaged. For assessment of fibrosis, 3 random pictures per slide (slides for infarct zone, border zone and remote area) were taken using polarized light during one session. Pictures were analysed in Cell^p (version 5.0 Olympus) for mean grey value and percentage of fibrotic area by using the same settings for all slides. Cyclosporin assay The effect of CsA on the cells was tested by a migration assay, sprouting matrigel assay and by testing cellular growth factor release. Based on the pig serum levels of CsA, in vitro assays were executed to analyze performance of CMPCs in presence of CsA. CMPCs were maintained in 0.1% gelatin coated plates with standard CMPC culture medium. For all in vitro experiments low passages were used (passage 8-15). For all assays, standard culture medium was used as a negative control. Three different concentrations of CsA were used as experimental conditions (50 ng/mL, 150 ng/mL and 300 ng/mL). See supporting information for detailed description of the methods. Statistical analysis Data are reported as mean ± standard deviation. Comparisons of serial measurements over time within one group or for all animals combined were performed by paired t-tests. Comparisons between both groups were made by independent t-tests. Differences between conditions in cyclosporin assays were tested by one-way ANOVA. The statistical software used is IBM SPSS statistics, version 20.0 (IBM Corporation, Armonk, NY, USA).

RESULTS

A total of 19 pigs were used in this study. In 9 pigs VF occurred during occlusion, of which 8 converted successfully by defibrillation and amiodaron infusion. A total of three pigs died; one pig died of VF during LAD occlusion, one pig died by periprocedural complications before cell/ placebo infusion and one animal died one day after placebo infusion by unknown cause. All other 16 animals were included in this study. Safety Two hours after cell infusion during short coronary occlusion, high sensitive troponin I (TnI) levels were 545 ± 460 ng/L n=8 after cell infusion and 623 ± 764 ng/L n=8 after placebo infusion.

205

PART THREE CHAPTER 10

In comparison, mean TnI levels 3.5 hours after the start of 90 minutes balloon occlusion, were 1147822 ± 584559 ng/L (n=16), confirming the induction of a large MI. Intracoronary pressure and flow measurement (see below) showed no evidence for intracoronary embolization or obstruction in CMPC treated animals 30 minutes after infusion. None of the CMPC treated animals died by arrhythmias, or any other cause. Left ventricular performance MRI At the end of follow-up EF, EDV and ESV did not differ between groups. EF in CMPC treated animals was 40.6 ± 4.2 % versus 38.3 ± 8.6% in placebo animals (p=0.52). EDV was 172.7 ± 19.0 ml in CMPC treated animals and 150.0 ± 27.1 ml in placebo (p=0.07). ESV was 102.6 ± 13.3 ml in CMPC treated and 91.5 ± 14.9 ml in placebo animals (p=0.14). (Figure 2) PV-loop Serial functional measurements were performed by PV-loop. Because of technical failure, we could not complete data analysis of all animals at all time points. For EF, ESV and EDV we could include 7 out of 8 cell treated animals and all 8 placebo animals. For end systolic pressures volume relationship (ESPVR), end diastolic pressure volume relationship (EDPVR) and V0, we could include 6 animals per group. Induction of MI caused a significant decrease in EF over 4 weeks (-8.7 ± 11.6% from 54.8 ± 8.7 to 46.1 ± 9.5 p=0.01) and an increase in ESV (+40.6 ± 62.3 ml from 85.3 ± 27.1 to 125.9 ± 49.1

Figure 2. Functional measurements by MRI of cell- and placebo treated animals at follow-up. Individual data and mean with SD, n=8 per group. EF Left ventricular ejection fraction, EDV End diastolic volume, ESV End systolic volume, NS = non-significant.

206

CMPCs IN HEART FAILURE

Figure 3. Functional outcome measurements by PV-loop. Mean and SD for PV-loop parameters. EDV End diastolic volume, ESV End systolic volume, EF Left ventricular ejection fraction, ESPVR End systolic pressures volume relationship, EDPVR end diastolic pressure volume relationship. Placebo n=8, cell n=7 for ESV, EDV and EF, n=6 per group for ESPVR, EDPVR and V0.

p=0.02), and trend towards an increased EDV (+38.1 ± 73.6 ml, from 186.7 ± 37.6 ml to 224.7 ± 50.9 p=0.07), confirming successful creation of a chronic MI model. No significant differences between both groups were observed at pre-infusion time point (EF p=0.34, EDV p=0.40, ESV p=0.27). (Figure 3) For all parameters (EDV, ESV, EF, ESPVR, EDPVR and V0), no differences between both groups were found at any time point except for EDPVR at follow-up. EDPVR was 0.004 ± 0.002 for cell treated animals and 0.008 ± 0.004 for placebo animals (p=0.04). Visual appearing trends in functional outcomes are observed in Figure 3 (and Supporting table 2). However, EF did not increase after cell or placebo infusion (p=0.85 and p=0.17 respectively) EDV and ESV did not change after cell treatment (p=0.25 and p=0.58 respectively), but both significantly increased after placebo infusion (p=0.01 and p=0.004), suggesting further deterioration after placebo infusion compared to cell treatment. No statistical significant differences in ΔEDV and ΔESV were found between cell treated animals and placebo animals.

Echocardiography WTsept and FAS at the apical level significantly decreased after MI (WTsept -0.34 ± 0.23 p < 0.001, FAS -0.08 ± 0.11 p = 0.02), confirming successful apical infarct creation. A trend towards decrease of FS at the apical level was also observed (-0.50 ± 0.11 p=0.08) in both groups. For FAS, FS and WTsept, no statistical significant differences were identified between groups at any time point. A trend was observed for a higher mean FAS (mean of three levels) at follow -up in cell treated animals compared to placebo animals (cell treated 0.49 ± 0.06 placebo 0.44 ± 0.04 p=0.08). Next, no differences in decrease or increase of any other echocardiographic parameter after hCMPC or placebo infusion was observed. (Figure 4 and Supporting table 3)

207

PART THREE CHAPTER 10

Figure 4. Functional measurements by echocardiography. Mean and SD, n=8 per group. FAS Fractional area shortening, FS Fractional shortening, sWT Septal wall thickening at the level of the apex.

Intracoronary pressure and flow Hyperemic microvascular resistance is a parameter for assessing the microvascular bed and is calculated as the ratio of intracoronary pressure (Pd) and hyperemic average peak flow velocity (pAPV). It is known from previous studies that HMR increases after I/R, accompanied by decreased pAPV and a constant Pd.9 At the time point of cell infusion, measurements in one animal failed because of technical problems. HMR in the LAD region significantly increased after I/R as expected (p=0.03), Pd in the LAD significantly increased (p=0.03) and pAPV in the LAD tended towards a decrease (p=0.10). All parameters in the control region remained stable over time. Thirty minutes after cell or placebo infusion, HMR was significantly decreased in all animals (2.7 ± 1.0 mmHg/cm/s pre-infusion, 2.2 ± 0.6 mmHg/cm/s post-infusion, p = <0.001), without any differences between hCMPC and placebo treated animals (ΔHMR -0.50 ± 0.48 mmHg/cm/s for hCMPC treated, -0.56 ± 0.66 mmHg/cm/s for placebo treated animals, p=0.21). Decrease in HMR was accompanied by an increase in pAPV (41.5 ± 11.0 cm/s pre-infusion, 48.9 ± 11.7 post-infusion, p=0.01), without differences between groups (p=0.7), both advocating against intracoronary (micro-)embolization or obstruction by cell infusion. Decreased HMR and increased pAPV can be explained as a hyperemic response after short ischemia during infusion. Neither cell-, nor placebo infusion caused a change in microvascular parameters over time. Infarct size For LGE, infarct size was 17.5 ± 3.8% and 18.0 ± 4.5% of the left ventricular mass for cell treated and placebo animals respectively at follow-up. For TTC, infarct size in the cell treated group was 14.6 ± 2.5% of the left ventricle, and 13.8 ± 3.9 in the control group. There was no statistical difference in infarct size between both groups at termination of follow-up, for both modalities. (Figure 5) Histology For fibrosis, there were no differences between both groups. A significant higher vascular density was observed in the placebo group compared to the hCMPC animals. (Table 1, Figure 6)

208

CMPCs IN HEART FAILURE

Cyclosporin CsA did not affect functioning of CMPCs in vitro with regard to migration capacity, angiogenesis and growth factor secretion. See supporting information for results and figures.

Figure 5. Infarct size Representative pictures of TTC stained heart slice and LGE image of the same animal.

Table 1. Fibrosis % Fibrosis

Grey value

CMPC (n=8)

Placebo (n=8)

Infarct

39.1 ± 12.2

40.0 ± 16.0

Border

1.6 ± 1.2

2.9 ± 2.3

Remote

6.5 ± 7.8

2.8 ± 2.3

Infarct

27.0 ± 8.9

28.5 ± 12.7

Border

1.2 ± 0.7

2.0 ± 1.4

Remote

4.3 ± 4.3

2.6 ± 2.2

33.9 ± 10.1

48.3 ± 10.1 *

Vascular density * p=0.01 between CMPC and placebo.

209

PART THREE CHAPTER 10

Figure 6. Histology Representative pictures of picrosirius red staining of infarct area (A), borderzone (B) and remote area (C) and lectin staining of the borderzone (D). Table presents mean Âą SD percentage fibrosis and mean grey value of infarct area, border zone and remote area and vascular density of the borderzone of cell and placebo treated animals.

DISCUSSION In this preclinical model of chronic MI, we show that xenotransplantation via intracoronary infusion of fetal hCMPCs is feasible and safe, but we did not find a benefit in the current model on cardiac function, infarct size, fibrosis and vascular density. The non-significant difference in EF between both groups of 2.3% measured by the gold standard MRI is lower than the effect size of 7.5% as documented in a recent meta-analysis and used for power calculation. We believe that these data in the field of cardiac regeneration are of great importance. It complements the preclinical evidence needed for successful translation of cardiac regeneration by cardiac cell therapy from bench to bedside. We present a solid and sound piece of evidence since this study is conducted by clinical standards in terms of blinding, randomization and outcome assessment. Preclinical studies of cardiac cell therapy The neutral results presented in the current study, are partly in line with other large animal studies with cardiac stem cells for MI (Table 2). Over the last decade, five different cardiac stem

210

CMPCs IN HEART FAILURE

Source

Dose

Timing

Follow- up

Autologous NA

1,0E+07

4 wk

8 wk

No EF or IS improvement

Autologous NA

1,0E+07

4 wk

8 wk

No functional improvement on echo. EF improved on LV angiography

Allogeneic

1,25E+07 2-3 wk

8 wk

EF preservation, IS reduction

Allogeneic

8,5-9E+06 30 mins 2 days

No difference in EF, IS reduction

Allogeneic

1,5E+08

8 wk

No functional improvement, IS reduction

Human

Cyclosporin

1,0E+06

14 days

1 mnth Restored EF, IS reduction

Autologous

5,0E+05

3-4 mnth

1 mnth

Result

Route

Immuno

suppression

Table 2. Overview of studies of cardiac derived stem cells in pig models

CDC Johnston et al. 31 Circulation 2009

Lee et al. 32 JACC 2011

Malliaras et al. 33 Circulation 2013

Kanazawa et al.34 Circ. Heart Fail. 2015

Yee et al. *35 PLoS ONE 2014

CSC Williams et al. 22 Circulation 2013

Bolli et al. 23 Circulation 2013

Improved EF, IS reduction

IM intramyocardial injections, IC intracoronary infusion, TE transendocardial injections, EF ejection fraction, IS infarct size. * used cardiospheres instead of CDCs.

cell sources have been identified10 and two of them have been extensively tested in large animal studies. Clonogenic, multipotent and self-renewing cardiac stem cells, based on the C-kit epitope (CSCs)11,12 and cluster-forming cells, positive for C-kit and Sca-1, called cardiosphere derived cells (CDCs)13 were used in these studies. Both cell types have gone through the process of preclinical investigation and progressed from bench to bedside testing. In a recent study, expanded cells upon different cardiac progenitor cell isolation procedures were compared

(including, c-kit+, cardiospheres, and sca1+ cells), and it was suggested that these cells have highly similar transcriptional profiles.14 Therefore, it is important to closely look at particular study design aspects of the study to look for possible explanations. Study design considerations A multitude of parameters in the research field of cell therapy for cardiac repair must be considered when conducting preclinical or clinical studies, like timing of therapy, administration route, cell source, cell number, etc. In a meta-analysis including all large animal studies on cell therapy for MI, none of these parameters showed significant impact on effect size.15 Additionally, no ruling about these issues was provided based on previous CSC studies (Table 2). Therefore, we made upfront decisions about all these variables based on experience and clinical relevance.

211

PART THREE CHAPTER 10

Treatment of acute myocardial infarction is greatly improved over the last decades, but patients surviving the initial episode often suffer from chronic ischemic heart failure. The current animal model was designed to represent this group. A direct comparison of efficacy of stem cells in acute versus chronic ischemic heart disease has never been performed, but safety is at least similar in acute and chronic patients.16 Table 2 shows negative and positive results of cardiac cell therapy in both acute and chronic settings. We administered cells via intracoronary infusion, which is considered as efficient as other routes with regard to cell retention17–19 and functional improvement20,21 but is less invasive then intramyocardial injections and less time consuming compared to transendocardial delivery. We investigated efficacy of infusion of 10 million cells. Table 2 show both positive and neutral results from CDCs in comparable doses. For CSCs, promising results are found by infusion of 0.5 and 1 million cells in smaller pigs (1.5 ± 0.8kg and 35-40 kg respectively, compared to 68.5 ± 5.4 kg in the current study).22,23 Roughly, we applied 4 times higher indexed doses (37,000/kg versus 140,000/kg) of cells. These data show that relative under-dosing was certainly not anticipated for in the present study. Xenogenicity The impact of xenogenicity is important since human cells were administered in a porcine model. Alloreactivity (and xenoreactivity) depends on foreign peptide presentation by the major histocompatibility complex (MHC) on antigen presenting cells and detection by T cells.24 In clinical practice, alloreactivity is suppressed by T cell suppressors like CsA, and this strategy was applied accordingly in the present study. Next to immunosuppression, the effect of CsA includes prevention of apoptosis25 and protection of the myocardium in the setting of acute myocardial infarction.26 Since CsA was administered at 4 weeks after MI until the end of follow-up, we believe it did not affect the outcome by that cause. Adequate dosing by oral administration of CsA is safe in pigs27, but not much is known about the pharmacodynamics. We did not observe differences in serum levels of leukocytes, nor on renal function between hCMPC and placebo treated animals 30 minutes and 4 weeks after treatment, all receiving CsA. Also, we did not find a correlation between serum CsA levels at follow-up and effect size. There is no consensus about the effect of CsA on stem cells in vitro and in vivo.27 In this respect, we tested the effect of different levels of CsA on migration, angiogenesis and growth factor expression of hCMPCs in vitro (See supporting information). No significant effect of CsA was noted. Taken together, we assume that our results are not confounded by the chosen CsA regimen. Internal and external validity Internal validity has major impact on effect size, as it is known that blinded and randomized studies produce smaller effect sizes compared to non-blinded and non-randomized reports.28 Based on clinical standards, our preclinical study has been performed in a randomized, blinded (both for intervention and outcome assessment) and placebo-controlled manner. The animal model is a surrogate for human disease. We included large pigs (68.5 ± 5.4 kg), as the heart corresponds with human (coronary) anatomy and size, to increase external validity (the

212

CMPCs IN HEART FAILURE

translatability towards other models or populations). However, age and co-morbidity are not accounted for in this model. Our standardized way of infarct creation caused a decrease in EF of 8.7% based on PV-loop measurements, despite high TnI levels and a mean infarct size of 17.8%. It is known from clinical studies that patients with a low baseline EF benefit most from cell therapy.29,30 In retrospect, the anticipated effect size of 7.5% (difference in EF at follow-up between cell and placebo treated animals on MRI) might have been too high. Power calculation was based on difference in EF at follow-up, measured by MRI and not for the other functional measurement modalities (echocardiography and PV-loop), which might thereby be underpowered to report significant differences.

CONCLUSION In this randomized, blinded placebo controlled preclinical study, xenotransplantation via intracoronary infusion of fetal human CMPCs in a pig model of chronic I/R injury appeared to be feasible and safe, but showed no significant improvement with regard to cardiac function or infarct size. Acknowledgments We kindly thank Corina Metz, Lizanne Bosma, Marlijn Jansen, Joyce Visser, Martijn van Nieuwburg, Grace Croft, Gerard Marchal, Merel Schurink and Evelyn Velema for their excellent technical assistance.

213

PART THREE CHAPTER 10

REFERENCES 1. Goumans MJ, Maring JA, Smits AM. A straightforward guide to the basic science behind cardiovascular cell-based therapies. Heart. 2014;100:1153–7. 2. Smits AM, van Vliet P, Metz CH, Korfage T, Sluijter JP, Doevendans PA, et al. Human cardiomyocyte progenitor cells differentiate into functional mature cardiomyocytes: an in vitro model for studying human cardiac physiology and pathophysiology. Nat Protoc. 2009;4:232–43. 3. Goumans MJ, de Boer TP, Smits AM, van Laake LW, van Vliet P, Metz CH, et al. TGF-beta1 induces efficient differentiation of human cardiomyocyte progenitor cells into functional cardiomyocytes in vitro. Stem Cell Res. 2007;1:138–49. 4. Van Vliet P, Smits AM, de Boer TP, Korfage TH, Metz CH, Roccio M, et al. Foetal and adult cardiomyocyte progenitor cells have different developmental potential. J Cell Mol Med. 2010;14:861–70. 5. Smits AM, van Laake LW, den Ouden K, Schreurs C, Szuhai K, van Echteld CJ, et al. Human cardiomyocyte progenitor cell transplantation preserves long-term function of the infarcted mouse myocardium. Cardiovasc Res. 2009;83:527–35. 6. van der Spoel TI, Jansen of Lorkeers SJ, Agostoni P, van Belle E, Gyöngyösi M, Sluijter JPG, et al. Human relevance of pre-clinical studies in stem cell therapy: systematic review and metaanalysis of large animal models of ischaemic heart disease. Cardiovasc Res. 2011;91:649–58. 7. Koudstaal S, Jansen of Lorkeers S, Gho JM, van Hout GP, Jansen MS, Gründeman PF, et al. Myocardial infarction and functional outcome assessment in pigs. J Vis Exp. 2014. 8. Van Hout GP, de Jong R, Vrijenhoek JE, Timmers L, Duckers HJ, Hoefer IE. Admittance based pressure volume loop measurements in a porcine model of chronic myocardial infarction. Exp Physiol. 2013;98:1565–75. 9. Koudstaal S, Jansen Of Lorkeers SJ, van Slochteren FJ, van der Spoel TI, van de Hoef TP, Sluijter JP, et al. Assessment of coronary microvascular resistance in the chronic infarcted pig heart. J Cell Mol Med. 2013;17:1128–35. 10. Koudstaal S, Jansen Of Lorkeers SJ, Gaetani R, Gho JM, van Slochteren FJ, Sluijter JP, et al. Concise review: heart regeneration and the role of cardiac stem cells. Stem Cells Transl Med. 2013;2:434-43. 11. Beltrami AP, Barlucchi L, Torella D, Baker M, Limana F, Chimenti S, et al. Adult cardiac stem cells are multipotent and support myocardial regeneration. Cell. 2003;114:763–76. 12. Bearzi C, Leri A, Lo Monaco F, Rota M, Gonzalez A, Hosoda T, et al. Identification of a coronary vascular progenitor cell in the human heart. Proc Natl Acad Sci U S A. 2009;106:15885–90. 13. Messina E, De Angelis L, Frati G, Morrone S, Chimenti S, Fiordaliso F, et al. Isolation and expansion of adult cardiac stem cells from human and murine heart. Circ Res. 2004;95:911–21. 14. Gaetani R, Feyen DA, Doevendans PA, Gremmels H, Forte E, Fledderus JO, et al. Different types of cultured human adult cardiac progenitor cells have a high degree of transcriptome similarity. J Cell Mol Med. 2014;18:2147–51. 15. Jansen Of Lorkeers SJ, Eding JE, Vesterinen HM, van der Spoel TI, Sena ES, Duckers HJ, et al. Similar effect of autologous and allogeneic cell therapy for ischemic heart disease: Systematic review and meta-analysis of large animal studies. Circ Res. 2015;116:80–6. 16. De Rosa S, Seeger FH, Honold J, Fischer-Rasokat U, Lehmann R, Fichtlscherer S, et al. Procedural safety and predictors of acute outcome of intracoronary administration of progenitor cells in 775 consecutive procedures performed for acute myocardial infarction or chronic heart failure. Circ Cardiovasc Interv. 2013;6:44–51. 17. Freyman T, Polin G, Osman H, Crary J, Lu M, Cheng L, et al. A quantitative, randomized study evaluating three methods of mesenchymal stem cell delivery following myocardial infarction. Eur Heart J. 2006;27:1114–22. 18. Perin EC, Silva G V, Assad JA, Vela D, Buja LM, Sousa AL, et al. Comparison of intracoronary and transendocardial delivery of allogeneic mesenchymal cells in a canine model of acute myocardial infarction. J Mol Cell Cardiol. 2008;44:486–95. 19. van der Spoel TI, Vrijsen KR, Koudstaal S, Sluijter JPG, Nijsen JF, de Jong HW, et al. Transendocardial cell injection is not superior to intracoronary infusion in a porcine model of ischaemic cardiomyopathy: a study on delivery efficiency. J Cell Mol Med. 2012;16:2768–76.

214

CMPCs IN HEART FAILURE

20. Fukushima S, Coppen SR, Lee J, Yamahara K, Felkin LE, Terracciano CM, et al. Choice of celldelivery route for skeletal myoblast transplantation for treating post-infarction chronic heart failure in rat. PLoS One. 2008;3:e3071. 21. Ang KL, Chin D, Leyva F, Foley P, Kubal C, Chalil S, et al. Randomized, controlled trial of intramuscular or intracoronary injection of autologous bone marrow cells into scarred myocardium during CABG versus CABG alone. Nat Clin Pract Cardiovasc Med. 2008;5:663–70. 22. Williams AR, Hatzistergos KE, Addicott B, McCall F, Carvalho D, Suncion V, et al. Enhanced Effect of Human Cardiac Stem Cells and Bone Marrow Mesenchymal Stem Cells to Reduce Infarct Size and Restore Cardiac Function after Myocardial Infarction. Circulation. 2012;15:213– 23. 23. Bolli R, Tang XL, Sanganalmath SK, Rimoldi O, Mosna F, Abdel-Latif A, et al. Intracoronary delivery of autologous cardiac stem cells improves cardiac function in a porcine model of chronic ischemic cardiomyopathy. Circulation. 2013;128:122–31. 24. Felix NJ, Allen PM. Specificity of T-cell alloreactivity. Nat Rev Immunol. 2007;7:942–53. 25. Matsuda S, Koyasu S. Mechanisms of action of cyclosporine. Immunopharmacology. 2000;47:119–25. 26. Piot C, Croisille P, Staat P, Thibault H, Rioufol G, Mewton N, et al. Effect of cyclosporine on reperfusion injury in acute myocardial infarction. N Engl J Med. 2008;359:473–81. 27. Jansen Of Lorkeers SJ, Hart E, Tang XL, Chamuleau ME, Doevendans PA, Bolli R, et al. Cyclosporin in cell therapy for cardiac regeneration. J Cardiovasc Transl Res. 2014;7:475–82. 28. Sena E, van der Worp HB, Howells D, Macleod M. How can we improve the pre-clinical development of drugs for stroke? Trends Neurosci. 2007;30:433–9. 29. Schächinger V, Erbs S, Elsässer A, Haberbosch W, Hambrecht R, Hölschermann H, et al. Improved clinical outcome after intracoronary administration of bone-marrow-derived progenitor cells in acute myocardial infarction: final 1-year results of the REPAIR-AMI trial. Eur Heart J. 2006;27:2775–83. 30. Delewi R, Hirsch A, Tijssen JG, Schächinger V, Wojakowski W, Roncalli J, et al. Impact of intracoronary bone marrow cell therapy on left ventricular function in the setting of ST-segment elevation myocardial infarction: a collaborative meta-analysis. Eur Heart J. 2014;35:989–98. 31. Johnston PV, Sasano T, Mills K, Evers R, Lee ST, Smith RR, et al. Engraftment, differentiation, and functional benefits of autologous cardiosphere-derived cells in porcine ischemic cardiomyopathy. Circulation. 2009;120:1075–83. 32. Lee ST, White AJ, Matsushita S, Malliaras K, Steenbergen C, Zhang Y, et al. Intramyocardial injection of autologous cardiospheres or cardiosphere-derived cells preserves function and minimizes adverse ventricular remodeling in pigs with heart failure post-myocardial infarction. J Am Coll Cardiol. 2011;57:455–65. 33. Malliaras K, Li TS, Luthringer D, Terrovitis J, Cheng K, Chakravarty T, et al. Safety and efficacy of allogeneic cell therapy in infarcted rats transplanted with mismatched cardiosphere-derived cells. Circulation. 2012;125:100–12. 34. Kanazawa H, Tseliou E, Malliaras K, Yee K, Dawkins JF, Couto G De, et al. Cellular PostConditioning : Allogeneic Cardiosphere-Derived Cells Reduce Infarct Size and Attenuate Microvascular Obstruction When Administered After Reperfusion in Pigs With Acute Myocardial Infarction. Circ Heart Fail. 2015;8:322–32. 35. Yee K, Malliaras K, Kanazawa H, Tseliou E, Cheng K, Luthringer DJ, et al. Allogeneic Cardiospheres Delivered via Percutaneous Transendocardial Injection Increase Viable Myocardium, Decrease Scar Size, and Attenuate Cardiac Dilatation in Porcine Ischemic Cardiomyopathy. PLoS One. 2014;9:e113805.

215

PART THREE CHAPTER 10

SUPPORTING INFORMATION Supporting methods Histology For staining, slides were deparafinated by 2x10 minutes ultraclear, 2x5 minutes 99% EtOH, 2x5 minutes 96% EtOH and 5 minutes 7% EtOH. Slides were washed three times with dH2O for 5 minutes. For picrosirius staining, all slides were stained at once for 30 minutes in filtered picrosirus red. Slides were washed twice 0.2N HCl, then in dH2O for 5 minutes and dehydrated (quick 70% EtOH, 2x quick in 96% EtOH, 2x5 minutes 99% EtOH, 2x10 minutes Ultraclear), mounted with entallan and covered with coverslips. For lectin, deparaffinated slides were incubated with 3% H2O2 for 20 minutes and with 1% BSA in PBS for 15 minutes. Slides were incubated over night with Lectin (1:50 in PBS). After washing twice with PBS, slides were stained with DAB+ chromogen in DAB+ substrate buffer (Dako K3467) for a maximum of 4 minutes. Slides were washed with running tap water, dehydrated as above, mounted with entellan and covered with coverslips. Cyclosporin assays Migration assay A migration scratch assay was performed to compare migratory capacity. CMPCs were plated in 0.1% gelatin coated 6 well plates in medium with additional CsA in three concentrations. At 90% confluency a scratch was placed vertically in the middle of the well using a pipet tip, after which medium was changed and replaced with either fresh medium or medium containing CsA. Pictures of the initial scratches were immediately taken. After 3 and 5 hours new pictures were taken of the scratched area. Analysis of migration capacity was performed using the ImageJ software. Sprouting matrigel assay Angiogenesis μ-slides (Ibidi) were used to quantify matrigel tube formation. Slide chambers were coated with 10 μl ECMatrixTM (Millipore). CMPCs were seeded on the matrigel in 50 μl, with or without CsA. Tube formation was imaged after 13 hours and quantified using the Angioquant software in Matlab. Growth factor antibody array Conditioned medium was collected from the CMPCs after 9 days of culture in low serum medium without any additional growth factors, with or without CsA. Human Angiogenesis Array G1 (RayBiotech) was performed on the conditioned medium according to the provided instructions. Three conditions were tested; normal medium, 150 ng/mL and 300 ng/mL. Next, we tested culture medium from a different CMPC cell line as a positive control.

216

CMPCs IN HEART FAILURE

RESULTS Migration of CMPC was the same for all conditions (Supporting Figure 1). The migration was 51.0 ± 9.8 % for the normal medium and 42.8 ± 10.7%, 50.2 ± 10.1% and 57.5 ± 8.5% in the presence of CsA (50ng/mL, 150 ng/mL and 300 ng/mL respectively) (p=0.95) For angiogenesis, the number of segments per area, total length per area and length per segment were calculated. For all 3 measures, no difference exists between conditions (Supporting table 3, Figure 2). Conditioned medium of CMPCs cultured for 9 days contained Angiogenin (ANG), CXCL 1, 2&3 (GRO), Il-6, Il-8, CCL2 (MCP1), CCL5 (RANTES) and metallopeptidase inhibitor 1 (TIMP1). Presence of CsA did not affect growth factor secretion by CMPCs (Supporting Figure 3).

Supporting Figure 1. Quantification of cell migration.

Supporting Figure 2. Sprouting matrigel assay. A. bar plot of angiogenesis parameters. B. representative pictures of matrigel assays for all 4 conditions.

217

PART THREE CHAPTER 10

Supporting Figure 3. Array map and pictures of the array of 4 conditions at three different intensities. Positive control meaning different cell line without Cyclosporin.

Supporting table 1. Angiogenesis

218

medium

50 µg/L

150 µg/L

300 µg/L

P-value

Total number/area

1.26 ± 0.23

1.10 ± 0.09

1.05 ± 0.25

1.00 ± 0.25

0.49

Total length/area

86.0 ± 15.0

70.6 ± 12.0

63.1 ± 15.9

64.5 ± 12.8

0.25

Length/number

67.9 ± 1.8

64.1 ± 5.6

60.1 ± 1.8

64.9 ± 4.6

0.18

CMPCs IN HEART FAILURE

Supporting table 2. Functional outcome measured by PV-loop Cell Baseline

Placebo

Pre-infusion

Follow-up

Baseline

Pre-infusion

Follow-up

EDV (ml) 185.1 ± 43.7

237.1 ± 51.4

275.5 ± 64.1

188.1 ± 34.3

213.9 ± 51.3

274.9 ± 51.0

ESV (ml)

85.6 ± 33.3

141.2 ± 57.0

157.0 ± 38.3

84.9 ± 22.9

112.5 ± 39.9

154.2 ± 34.3

EF (%)

54.6 ± 9.8

43.5 ± 9.8

42.7 ± 6.0

55.0 ± 8.3

48.4 ± 9.3

43.7 ± 9.7

1.3 ± 0.4

0.9 ± 0.8

1.6 ± 0.8

1.3 ± 0.4

0.9 ±0.7

1.7 ± 0.4

ESPVR

1.3 ± 0.6

1.2 ± 0.7

1.9 ± 1.8

1.3 ± 0.6

1.5 ± 0.9

1.2 ± 0.7

EDPVR

0.022 ± 0.020

0.008 ± 0.007

0.007 ± 0.004

0.017 ± 0.013

0.006 ± 0.004

0.004 ± 0.002*

dPdT+

1559.3 ± 196.3 1483.9 ± 414.0 1667.1 ± 324.9

1544 ± 210.4

1575.9 ± 182.3 1316.9 ± 97.9

dPdT-

-1231.3 ± 195.9 -1213.1 ± 379.4 -1305.5 ± 274.7

1388.4 ± 224.7 -1223.0 ± 248.8 -1077.7 ± 269.2

-49.9 ± 71.8

-14.4 ± 71.5

-6.9 ± 139.0

-34.7 ± 49.0

-14.5 ± 57.5

-26.8 ± 54.3

Tau

45.8 ± 9.6

55.2 ± 22.4

53.4 ± 10.2

40.2 ± 8.4

48.4 ± 7.5

60.6 ± 14.8

EDV End diastolic volume, ESV End systolic volume, EF Left ventricular ejection fraction, SW Stroke work, ESPRV End systolic pressure volume relationship, EDPVR End diastolic pressure volume relationship, dPdT+ maximum pressure rise, dPdT- Minimum pressure rise, V0 Theoretical volume at zero pressure. * p=0.03 Cell treated animals compared to placebo. For EDV, ESV and EF n=7 for cell treated animals, n=8 for placebo treated animals. For SW, ESPRV, EDPVR, dPdT+, dPdT-, V0 and Tau n=6 per group.

Supporting table 3. Functional outcomes measured by echocardiography Cell Baseline t=0 (n=8)

Placebo

Pre-infusion (n=8)

Follow-up (n=8)

Baseline t=0 (n=8)

Pre-infusion (n=8)

Follow-up (n=8)

sWT mitral 0.18 ± 0.12

0.23 ± 0.12

0.27 ± 0.20

0.29 ± 0.20

0.28 ± 0.11

0.14 ± 0.16

sWT pap

0.36 ± 0.13

0.37 ± 0.15

0.35 ± 0.24

0.39 ± 0.15

0.35 ± 0.32

0.42 ± 0.41

sWT apex

0.58 ± 0.26

0.17 ±0.24

0.22 ± 0.27

0.47 ± 0.14

0.18 ± 0.21

0.23 ± 0.32

sWT mean

0.37 ± 0.13

0.26 ± 0.10

0.28 ± 0.13

0.38 ± 0.14

0.27 ± 0.16

0.26 ± 0.21

FS mitral

0.23 ± 0.06

0.29 ± 0.05

0.28 ± 0.05

0.30 ± 0.04

0.27 ± 0.06

FS pap

0.23 ± 0.05

0.28 ± 0.05

0.27 ± 0.04

0.24 ± 0.05

0.30 ± 0.04

0.25 ± 0.06

FS apex

0.24 ± 0.08

0.20 ± 0.06

0.19 ± 0.08

0.24 ± 0.05

0.19 ± 0.06

0.18 ± 0.05

FS mean

0.23 ± 0.03

0.25 ± 0.04

0.26 ± 0.05

0.24 ± 0.05

0.23 ± 0.04

FAS mitral

0.49 ± 0.09

0.55 ± 0.05

0.57 ± 0.05

0.54 ± 0.08

0.55 ± 0.07

0.53 ± 0.07

FAS pap

0.46 ± 0.08

0.47 ± 0.06

0.49 ± 0.08

0.48 ± 0.09

0.478 ± 0.12

0.44 ± 0.07

FAS apex

0.45 ± 0.07

0.40 ± 0.07

0.42 ± 0.08

0.47 ± 0.12

0.37 ± 0.08

0.36 ± 0.05

FAS mean

0.47 ± 0.03

0.42 ± 0.17

0.49 ± 0.06*

0.49 ± 0.08

0.47 ± 0.07

0.44 ± 0.04*

sWT Septal wall thickening, FS Fractional shortening, FAS Fractional area shortening at levels of the mitral valve (mitral) papillary muscle (pap), apex and mean of three levels (mean). N=8 per group. *p=0.08 for cell treated animals compared to placebo.

219

PART THREE

Elucidating (Epi)genetic and Translating

Therapeutic Pathways

Chapter

General Discussion

CHAPTER 11

GENERAL DISCUSSION The studies in this thesis tried to unravel the pathways relevant for cardiac regeneration and the basic mechanisms leading to heart failure (HF). The first part portrays the clinical course regarding the incidence of and prognostic factors associated with HF following myocardial infarction. The second part aimed to improve fibrosis detection in HF by optimizing cardiovascular magnetic resonance imaging (CMR) and histological analyses. The third and final part aimed to identify underlying (epi)genetic pathways leading to HF and focused on translation of cell therapy to the clinic. The goal of this thesis was to provide opportunities from bench-to-bedside to elucidate pathways associated with the failing heart. Opportunities in the failing heart following myocardial infarction The Global Burden of Disease 2010 Study estimated a global increased prevalence of ischaemic heart disease between 1990 and 2010, predominantly in males.1 The incidence of HF after MI could be changing due to variation in baseline population characteristics and improved treatment of MI.2 Currently, contemporary data of large cohorts regarding the incidence of HF following myocardial infarction (MI) are lacking. Information has become more widely available for clinical research with the electronic health record (EHR). While the EHR is not a direct reflection of the patients and physiology, but a reflection of the recording process inherent in healthcare with noise and feedback loops, it has been proposed to study the EHR as an object of interest in itself.3 Particularly in chronic disease, patients data is collected over time in different sources and linkage of these sources must be performed to obtain a more complete picture.4 Using linked EHR sources from primary and secondary care to identify MI patients (n = 24,745) in Chapter 2, we observed that in the current era 1) a substantial amount of patients develop HF after a first MI (24.3%, median follow-up 3.7 years) 2) increasing age, higher socioeconomic deprivation, a history of hypertension, diabetes, atrial fibrillation, peripheral arterial disease or COPD and a STEMI at presentation were important risk factors for HF following MI. The identified amendable prognostic factors can be used in future to decrease the incidence of HF following MI. The high incidence is supported by a different Swedish study (1993-2004) showing nearly one-third of MI patients aged 65-84 developed HF within 3 years. 5 On the other hand, ascertainment of cardiovascular outcomes based on clinical criteria creates a more direct reflection of the patients and physiology. In Chapter 3 we used a meticulously phenotyped cohort of patients (n = 1360) with a first ST-elevation myocardial infarction (STEMI) where followup was performed for incident HF. In this population, we found a relatively low HF incidence of 6.3% after STEMI during a median follow-up of 6.7 years. Higher peak creatine kinase (CK-MB) levels and an LAD (left anterior descending) artery culprit lesion at index STEMI were important risk factors for the outcome of HF. At the time of this research there was no EHR linkage with primary care data available in the Netherlands. With the advent of EHRs it would be important to create an infrastructure for linkage of EHR sources (primary care, secondary care and registers) for a more accurate estimation of health outcomes in the Netherlands.

222

GENERAL DISCUSSION

Improved cardiac fibrosis detection Cardiovascular Magnetic Resonance Imaging The aim of Chapter 4 was to systematically compare the relation between multiple CMR techniques and myocardial fibrosis in transverse cardiac slices of a chronic porcine infarct model. Hereby we tried to identify the most optimal technique to assess local myocardial fibrosis for translation into clinical practice. This systematic comparison can be readily adapted for use with other imaging modalities (e.g. SPECT, CT, PET and echocardiography). Myocardial strain and wall thickening showed modest explained variance for fibrosis. This might be explained by the resulting deformation caused by the contraction of cardiomyocytes in the vicinity of the infarct, as strain is not a strictly local phenomenon. For late gadolinium enhancement (LGE) imaging, we found that the full width at half maximum (FWHM) technique is preferred for the detection of myocardial fibrosis. This is in line with the patient study published by Flett et al., who studied patients with acute MI, chronic MI and hypertrophic cardiomyopathy and found that the FWHM technique for LGE quantification gave mean LGE volume results similar to manual quantification and was the most reproducible, regardless of underlying etiology.6 Myocardial fibrosis can lead to HF and act as a substrate for cardiac arrhythmias. Assessment of cardiac fibrosis is important for diagnosis, predicting prognosis, therapy guidance and monitoring. While LGE CMR provides an accurate qualitative measure of fibrosis, it has potential drawbacks as an indirect method requiring contrast agents without quantitative measurement of cardiac collagen. Assessment of extracellular volume (ECV) could be of key importance to predict prognosis in diffuse myocardial fibrosis.7, 8 Contrast-enhanced T1 mapping seems a promising method to quantify myocardial ECV in diffuse myocardial fibrosis.9 To overcome the limitations associated with contrast agent use, several endogenous contrast mechanisms for fibrosis detection are reviewed in Chapter 5. In this Chapter we also showed examples of T1rho and T2* mapping from a pilot study in a porcine chronic infarct model. Endogenous CMR fibrosis detection methods could obviate the need for exogenous contrast agents, favourable for patients with renal dysfunction, could improve reproducibility and decrease scan duration. Although T2* mapping might be a promising method for endogenous contrast, it is unclear whether postreperfusion intramyocardial hemorrhage with iron deposits can be distinguished from fibrosis.10 We recently showed the feasibility of native T1rho mapping translated from animals to patients.11 In 21 patients with chronic myocardial infarction, we found a significantly higher T1rho relaxation time in the infarct region compared to remote myocardium. Recently, a study using T1rho CMR to detect fibrosis in HCM patients has been reported.12 They observed a significant correlation between T1rho at 3SD and 4SD with visual LGE measurement. Although endogenous contrast

methods for fibrosis detection seem promising, more research is required before a large-scale application for clinical decision-making can be recommended. These study results warrant further investigation of endogenous contrast CMR in patients with replacement or diffuse fibrosis. Histological fibrosis quantification Phospholamban (PLN) plays a key role in calcium homeostasis and calcium cycling is crucial for cardiac excitation and contraction (Figure 1).13-15 While several causal PLN mutations have been described in humans, the PLN R14del mutation has been associated with dilated and

223

CHAPTER 11

arrhythmogenic cardiomyopathy.16 Previously, histological analysis of the myocardium of patients with the PLN R14del mutation has not been extensively performed. To the best of our knowledge, Chapter 6 is the first study investigating fibrosis and fatty tissue patterns in transverse heart slices of PLN R14del patients. Myocardial fibrosis was mainly observed in the posterolateral wall of the left ventricle, whilst adipose tissue was more pronounced in the outer myocardium of the right ventricle. Previous histopathological studies (not specifically PLN mutation associated) have shown similar areas of predilection in arrhythmogenic cardiomyopathy.

Figure 1. Calcium cycling in cardiomyocytes. In response to depolarization due to sodium (Na+) influx, calcium (Ca2+) enters the cytosol through the L-type calcium channels (LTCC) in the plasma membrane. This Ca2+ influx leads to calcium-induced calcium release from the sarcoplasmic reticulum mediated by the Ca2+ release channels (ryanodine receptors; RyRs). Calcium binds to troponin in the thin filaments of myofibrils to activate muscle contraction and Ca2+ is removed from the cytosol by sarcoplasmic reticulum Ca2+ ATPase (SERCA2a) and the sodium-calcium exchanger (NCX) on the plasma membrane. Phospholamban (PLN) is a reversible inhibitor of SERCA2a and in its dephosphorylated state PLN inhibits SERCA2a activity. Upon phosphorylation by protein kinase A (PKA), through the Î˛-adrenergic receptor pathway, or Ca2+/calmodulindependent protein kinase (CaMKII), PLN dissociates from SERCA2a, thereby relieving Ca2+ pump inhibition and enhancing cardiac relaxation and contractility. Phospholamban is dephosphorylated by protein phosphatase (PP1), which ends the stimulation phase and PP1 is regulated by inhibitor-1 (I-1).

224

GENERAL DISCUSSION

According to our findings the sensitivity of endomyocardial biopsies from the right ventricular septum in PLN mutation associated cardiomyopathies might be low. The systematic fibrosis quantification method presented in this chapter could be useful to assess fibrosis and fatty tissue in a broad range of human cardiomyopathies and in preclinical studies as an outcome measure. Using the systematic histological quantification method developed in Chapter 6, in Chapter 7 we found a characteristic distribution of fibrosis in genetic cardiomyopathies related to the mutation group. The posterolateral wall of the left ventricle was found to be highly discriminating for the identification of the different groups of mutations. These histological patterns may provide a roadmap for cardiac imaging in patients and could be related to genetics and molecular biology of the corresponding tissue to elucidate underlying pathophysiology. In future nomenclature of cardiomyopathy types, it might be more suitable to use the underlying genetic defect.17 Identifying genes and pathways associated with the failing heart Despite advances in genetic research, limited genetic loci associated with risk of incident HF have been identified. A genome-wide association study (GWAS) examines many common variants in individuals to determine associations of these variants with a trait.18 Typically, associations between single nucleotide polymorphisms (SNPs) and disease are studied. Previously, 1 GWAS has been reported in unrelated individuals with HF.19 This GWAS only identified a single locus in each ethnic cohort that reached genome-wide significance. This might be explained by a large heterogeneity due to inclusion of HF cases with different underlying etiologies. Furthermore, only common genetic variants with a minor allele frequency of >5% were investigated. In future, clustering of weakly associated interacting genes might be identified using a pathway-based analysis approach.20 The zebrafish embryo can be studied to investigate the role of genes associated with cardiac disease.21 Molecular and cellular techniques to characterize gene function in vivo can be used to fully elucidate the role of those found genes. While the PLN R14del mutation has been associated with both dilated cardiomyopathy and arrhythmogenic cardiomyopathy, genetic cardiomyopathies have a variable disease penetrance and phenotype severity, even in the same family, thus other regulatory factors might play a role.22, 23 Next-generation sequencing allows for profiling of DNA regulatory elements (e.g., histone marks related to active promoters and enhancers) that control the timing, localization and level of gene transcription. Changes in these epigenetic profiles could serve as an indicator of disease pathogenesis. In Chapter 8, we aimed to identify differentially regulated genes and pathways in HF by chromatin immunoprecipitation and sequencing (ChIP-seq) of human

cardiac tissue from PLN R14del cardiomyopathy patients and healthy controls. We found 208 PLN R14del mutation specific regulated genes which could be used as a biomarker for risk stratification. These genes were mostly involved with general fibrotic pathways and developmental programs. Of these genes, 55 involved in lipid metabolism can be further investigated as candidate genes associated with fibrofatty replacement. The identified differentially regulated gene regions in HF could ultimately lead to a better understanding of underlying disease mechanisms, risk stratification and novel therapeutic strategies. Previously, p300, a transcriptional activator, has been associated with hypertrophy and HF through

225

CHAPTER 11

transcription by myocyte enhancer factor-2 and GATA-4 and it has been suggested that inhibition of p300 histone acetyltransferase activity by curcumin (a polyphenol responsible for the yellow color of the spice turmeric) may provide a novel therapeutic strategy for HF in humans.24, 25 Therapeutic opportunities Using the genetic knowledge, molecular defects associated with HF can be factored in therapeutic decision making and targeted with gene therapy.23 In HF sarcoplasmic reticulum Ca2+ ATPase (SERCA2a) expression is decreased26, thereby impairing cardiac contractility and relaxation. Gene transfer of SERCA2a in preclinical models, in rodents27, pigs28 and sheep29, showed improved cardiac function and energy potential in failing hearts. This led to the initiation of human clinical trials of SERCA2a gene transfer.30 Unfortunately, the SERCA2a CUPID phase 2b, double-blind, placebo-controlled, randomised trial (ClinicalTrials.gov Identifier: NCT01643330) did not meet the primary endpoint of time to recurrent event (heart failure-related hospitalisations in the presence of terminal events), but the procedures (e.g., dosing and delivery methods) could be optimised and this might pave the way for other gene therapy trials.31 MicroRNAs implicated in HF could also be potentially important targets for HF therapy.32, 33 Using the hypothesis of enhancement of the sensitivity of cardiac myosin to calcium, another approach is directed at the activation of cardiac myosin in patients with HF and reduced ejection fraction. Omecamtiv mecarbil, a cardiac myosin activator, does not seem to increase oxygen consumption or intracellular calcium and clinical trials in chronic HF are currently underway.34 Furthermore, major advances have been made with (assist) device therapy and the described therapeutic approaches, alone or in combination, could become important weapons in the ‘war’ against HF.35 Cell therapy could be a promising technique for cardiac repair, by regeneration of heart muscle in failing hearts and thereby improving cardiac function. Currently, the ideal cell type has not yet emerged. The advantages of cell therapy in chronic cardiomyopathy include a “stable” environment and a relatively large target patient population. In Chapter 9, the results of cell therapy for dilated cardiomyopathy seem promising, but this article has recommendations for future studies, such as improved methodological quality, elucidating underlying mechanisms and validating safety and efficacy before implementation in clinical practice. This includes recommendations for clearly described uniform definitions for inclusion and use of CMR to determine outcome measures. There is a need of consensus for methodological sound studies. It might be more appropriate to use differences between study arms measured in clinical endpoints such as (recurrent) hospitalisation for HF or mortality.35 With the aim of translation to the clinic, in Chapter 10, we used the previously reported human cardiomyocyte progenitor cells (CMPCs) as cell therapy in a preclinical study.36 We performed a randomised, placebo controlled study with intracoronary infusion of CMPCs in large animal model of ischaemic HF. While this therapeutic method seems feasible and safe, we did not find significant positive effects on left ventricular performance or infarct size. Therefore translation of the current CMPCs towards a first-in-man trial is presently not justified.

226

GENERAL DISCUSSION

With the advance of induced pluripotent stem cells (iPSCs)37 functional cardiomyocytes can be generated for patient-specific disease modelling or regenerative approaches. In a recent collaboration iPSCs were generated from a patient harbouring the PLN R14del mutation and differentiated into cardiomyocytes.38 These iPSC derived cardiomyocytes recapitulated Ca2+ handling abnormalities, electrical instability, an abnormal distribution of PLN protein in the cytoplasm and increased expression of cardiac hypertrophy markers. Gene correction of the causative R14del mutation using transcription activator-like effector nucleases (TALENs) restored the phenotype. These findings may pave the way for tailored therapy of genetic mutations associated with cardiomyopathies. Conclusion The failing heart presents challenges, ranging from basic mechanisms through the translational axis to the clinical course and vice versa. In this thesis we have shed light on opportunities to elucidate underlying mechanisms, visualize pathophysiology, identify the clinical picture and pave the way for future therapeutic strategies in the failing heart.

227

CHAPTER 11

REFERENCES 1.

2. 3. 4. 5.

10.

11.

12.

13. 14. 15. 16.

17.

18.

19.

228

Moran AE, Forouzanfar MH, Roth GA, Mensah GA, Ezzati M, Flaxman A, et al. The global burden of ischemic heart disease in 1990 and 2010: the Global Burden of Disease 2010 study. Circulation. 2014;129:1493-501. Mosterd A, Hoes AW. Clinical epidemiology of heart failure. Heart. 2007;93:1137-46. Hripcsak G, Albers DJ. Next-generation phenotyping of electronic health records. J Am Med Inform Assoc. 2013;20:117-21. Stevens RJ, McManus R. Unlinked data sources underestimate risk of cardiovascular disease. BMJ. 2013;346:f3737. Shafazand M, Rosengren A, Lappas G, Swedberg K, Schaufelberger M. Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993-2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden. Eur J Heart Fail. 2011;13:135-41. Flett AS, Hasleton J, Cook C, Hausenloy D, Quarta G, Ariti C, et al. Evaluation of techniques for the quantification of myocardial scar of differing etiology using cardiac magnetic resonance. JACC Cardiovasc Imaging. 2011;4:150-6. Wong TC, Piehler K, Meier CG, Testa SM, Klock AM, Aneizi AA, et al. Association between extracellular matrix expansion quantified by cardiovascular magnetic resonance and short-term mortality. Circulation. 2012;126:1206-16. Wong TC, Piehler KM, Kang IA, Kadakkal A, Kellman P, Schwartzman DS, et al. Myocardial extracellular volume fraction quantified by cardiovascular magnetic resonance is increased in diabetes and associated with mortality and incident heart failure admission. Eur Heart J. 2014;35:657-64. Moon JC, Messroghli DR, Kellman P, Piechnik SK, Robson MD, Ugander M, et al. Myocardial T1 mapping and extracellular volume quantification: a Society for Cardiovascular Magnetic Resonance (SCMR) and CMR Working Group of the European Society of Cardiology consensus statement. J Cardiovasc Magn Reson. 2013;15:92. Gho JM, van Oorschot JW, Jansen of Lorkeers S, Froeling M, Luijten PR, Doevendans PA, et al. Endogenous contrast MRI of cardiac fibrosis: beyond late gadolinium enhancement. J Am Coll Cardiol. 2014;63(12_S):. doi:10.1016/S0735-1097(14)61008-1. van Oorschot JW, El Aidi H, Jansen of Lorkeers SJ, Gho JM, Froeling M, Visser F, et al. Endogenous assessment of chronic myocardial infarction with T(1rho)-mapping in patients. J Cardiovasc Magn Reson. 2014;16:104. Wang C, Zheng J, Sun J, Wang Y, Xia R, Yin Q, et al. Endogenous contrast T1rho cardiac magnetic resonance for myocardial fibrosis in hypertrophic cardiomyopathy patients. J Cardiol. 2015. doi: 10.1016/j.jjcc.2015.03.005. [Epub ahead of print] MacLennan DH, Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol. 2003;4:566-77. Bers DM. Cardiac excitation-contraction coupling. Nature. 2002;415:198-205. Kranias EG, Hajjar RJ. Modulation of cardiac contractility by the phospholamban/SERCA2a regulatome. Circ Res. 2012;110:1646-60. van der Zwaag PA, van Rijsingen IA, Asimaki A, Jongbloed JD, van Veldhuisen DJ, Wiesfeld AC, et al. Phospholamban R14del mutation in patients diagnosed with dilated cardiomyopathy or arrhythmogenic right ventricular cardiomyopathy: evidence supporting the concept of arrhythmogenic cardiomyopathy. Eur J Heart Fail. 2012;14:1199-207. Arbustini E, Narula N, Dec GW, Reddy KS, Greenberg B, Kushwaha S, et al. The MOGE(S) classification for a phenotype-genotype nomenclature of cardiomyopathy: endorsed by the World Heart Federation. J Am Coll Cardiol. 2013;62:2046-72. McCarthy MI, Abecasis GR, Cardon LR, Goldstein DB, Little J, Ioannidis JP, et al. Genome-wide association studies for complex traits: consensus, uncertainty and challenges. Nat Rev Genet. 2008;9:356-69. Smith NL, Felix JF, Morrison AC, Demissie S, Glazer NL, Loehr LR, et al. Association of genome-wide variation with the risk of incident heart failure in adults of European and African ancestry: a prospective meta-analysis from the cohorts for heart and aging research in genomic epidemiology (CHARGE) consortium. Circ Cardiovasc Genet. 2010;3:256-66.

GENERAL DISCUSSION

20. Wang K, Li M, Bucan M. Pathway-based approaches for analysis of genomewide association studies. Am J Hum Genet. 2007;81:1278-83. 21. Bakkers J. Zebrafish as a model to study cardiac development and human cardiac disease. Cardiovasc Res. 2011;91:279-88. 22. Cahill TJ, Ashrafian H, Watkins H. Genetic cardiomyopathies causing heart failure. Circ Res. 2013;113:660-75. 23. Fontaine GH, Zhang L. Is the phenotype-genotype relationship necessary to understand cardiomyopathies? Circ Cardiovasc Genet. 2014;7:405-6. 24. Morimoto T, Sunagawa Y, Kawamura T, Takaya T, Wada H, Nagasawa A, et al. The dietary compound curcumin inhibits p300 histone acetyltransferase activity and prevents heart failure in rats. J Clin Invest. 2008;118:868-78. 25. Yanazume T, Hasegawa K, Morimoto T, Kawamura T, Wada H, Matsumori A, et al. Cardiac p300 is involved in myocyte growth with decompensated heart failure. Mol Cell Biol. 2003;23:3593606. 26. Hasenfuss G, Reinecke H, Studer R, Meyer M, Pieske B, Holtz J, et al. Relation between myocardial function and expression of sarcoplasmic reticulum Ca(2+)-ATPase in failing and nonfailing human myocardium. Circ Res. 1994;75:434-42. 27. Sakata S, Lebeche D, Sakata N, Sakata Y, Chemaly ER, Liang LF, et al. Restoration of mechanical and energetic function in failing aortic-banded rat hearts by gene transfer of calcium cycling proteins. J Mol Cell Cardiol. 2007;42:852-61. 28. Kawase Y, Ly HQ, Prunier F, Lebeche D, Shi Y, Jin H, et al. Reversal of cardiac dysfunction after long-term expression of SERCA2a by gene transfer in a pre-clinical model of heart failure. J Am Coll Cardiol. 2008;51:1112-9. 29. Byrne MJ, Power JM, Preovolos A, Mariani JA, Hajjar RJ, Kaye DM. Recirculating cardiac delivery of AAV2/1SERCA2a improves myocardial function in an experimental model of heart failure in large animals. Gene Ther. 2008;15:1550-7. 30. Jaski BE, Jessup ML, Mancini DM, Cappola TP, Pauly DF, Greenberg B, et al. Calcium upregulation by percutaneous administration of gene therapy in cardiac disease (CUPID Trial), a first-in-human phase 1/2 clinical trial. J Card Fail. 2009;15:171-81. 31. Ratner M. Heart failure gene therapy disappoints but experts keep the faith. Nat Biotechnol. 2015;33:573-4. 32. Thum T, Gross C, Fiedler J, Fischer T, Kissler S, Bussen M, et al. MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature. 2008;456:980-4. 33. Wahlquist C, Jeong D, Rojas-Munoz A, Kho C, Lee A, Mitsuyama S, et al. Inhibition of miR-25 improves cardiac contractility in the failing heart. Nature. 2014;508:531-5. 34. Cleland JG, Teerlink JR, Senior R, Nifontov EM, Mc Murray JJ, Lang CC, et al. The effects of the cardiac myosin activator, omecamtiv mecarbil, on cardiac function in systolic heart failure: a double-blind, placebo-controlled, crossover, dose-ranging phase 2 trial. Lancet. 2011;378:676-83. 35. Braunwald E. The war against heart failure: the Lancet lecture. Lancet. 2015;385:812-24. 36. Goumans MJ, de Boer TP, Smits AM, van Laake LW, van Vliet P, Metz CH, et al. TGF-beta1 induces efficient differentiation of human cardiomyocyte progenitor cells into functional cardiomyocytes in vitro. Stem Cell Res. 2007;1:138-49. 37. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell. 2006;126:663-76. 38. Karakikes I, Stillitano F, Nonnenmacher M, Tzimas C, Sanoudou D, Termglinchan V, et al. Correction of human phospholamban R14del mutation associated with cardiomyopathy using targeted nucleases and combination therapy. Nat Commun. 2015;6:6955.

229

Samenvatting in het Nederlands List of publications Acknowledgements / Dankwoord Curriculum vitae

SAMENVATTING De studies in dit proefschrift probeerden de wegen te ontrafelen die relevant zijn voor cardiale regeneratieve geneeskunde en de basale mechanismen die leiden tot hartfalen. Het eerste deel portretteert het klinische beloop betreft de incidentie van en voorspellende factoren voor hartfalen na een hartinfarct. Het tweede deel was erop gericht om littekenweefsel in het hart beter te detecteren door optimalisatie van MRI technieken en histologische analyses. Het derde deel had ten doel om onderliggende erfelijke oorzaken te identificeren die leiden tot hartfalen en om de vertaalslag van stamceltherapie naar de kliniek te onderzoeken. Het uiteindelijke doel van dit proefschrift was mogelijkheden te creëren om resultaten uit het laboratorium naar de praktijk te vertalen en de wegen op te helderen die in verband zijn gebracht met het falende hart. Mogelijkheden in het falende hart na een hartinfarct Mogelijk verandert het vóórkomen van hartfalen na een hartinfarct vanwege verschillen in populatiekarakteristieken en een betere behandeling van het hartinfarct. Echter zijn er weinig hedendaagse studies op grote schaal verricht betreft het aantal nieuwe ziektegevallen (incidentie) van hartfalen na een hartinfarct. De ontwikkeling van elektronische patiëntendossiers (EPD’s) creëert de mogelijkheid om op grote schaal onderzoek te doen naar de epidemiologie van aandoeningen. Hoewel het EPD geen directe weerspiegeling is van de patiënt en fysiologie, met name bij chronische ziekten wordt data van patiënten verzameld gedurende de tijd in verschillende bronnen, kan door koppeling van meerdere bronnen een completer beeld worden gevormd. Door middel van een onderzoeksplatform welke meerdere EPD’s (uit Engelse huisartspraktijken, ziekenhuizen en registers) aan elkaar koppelt om patiënten met een hartinfarct te identificeren (n = 24.745) in hoofdstuk 2, zagen we dat in het huidige tijdperk 1) een substantieel deel van de patiënten hartfalen ontwikkelt na een eerste hartinfarct (24,3%, mediane follow-up 3,7 jaren) 2) een hogere leeftijd, toenemende mate van sociale deprivatie, een voorgeschiedenis met hoge bloeddruk, suikerziekte, boezemfibrilleren, perifeer vaatlijden of COPD en een ST-elevatie myocardinfarct (STEMI) bij presentatie belangrijke risicofactoren waren voor hartfalen volgend na een hartinfarct. De geïdentificeerde beïnvloedbare prognostische factoren kunnen in de toekomst gebruikt worden om de incidentie van hartfalen volgend na een hartinfarct te verminderen. Anderzijds, het vaststellen van uitkomsten betreft harten vaatziekten gebaseerd op klinische maatstaven creëert een meer directe weerspiegeling van de patiënt en fysiologie. In hoofdstuk 3 hebben we gebruik gemaakt van een zorgvuldig in kaart gebracht patiënten cohort (n = 1360) met een eerste STEMI waar vervolgonderzoek was verricht voor het optreden van hartfalen. In deze studiepopulatie vonden we een relatief lage hartfalen incidentie van 6,3% na STEMI gedurende een mediane follow-up van 6,7 jaren. Hogere piek creatine kinase (CK-MB) niveaus en een afgesloten LAD (left anterior descending) kransslagader bij index STEMI waren belangrijke risicofactoren voor de uitkomst hartfalen. Op het moment van dit onderzoek was er geen EPD koppeling met de eerste lijn beschikbaar in Nederland. Met de komst van de EPD’s is het van belang om een infrastructuur te creëren voor koppeling van verscheidene EPD bronnen (bijv. eerste lijn, tweede lijn en registers) voor een meer nauwkeurige schatting van gezondheidsuitkomsten in Nederland.

232

SAMENVATTING IN HET NEDERLANDS

Verbeterde herkenning van littekenweefsel in het hart MRI van het hart Het doel van hoofdstuk 4 was om een systematische vergelijking te maken van de relatie tussen verscheidene MRI technieken en littekenweefsel in het hart in transversale hartplakken bij een chronisch varkens infarct model. Hierbij probeerden wij de optimale techniek te identificeren om lokaal littekenweefsel in het hart vast te stellen voor translatie naar de kliniek. Deze systematische vergelijking kan eenvoudig aangepast worden voor gebruik met andere beeldvormingstechnieken (bijv. SPECT, CT, PET en echocardiografie). Lokale vervorming van de hartspier en wandverdikking op MRI lieten weinig verklaarde variantie voor littekenweefsel zien. Dit kan mogelijk verklaard worden doordat de lokale vervorming veroorzaakt wordt door samentrekking van hartspiercellen in de directe nabijheid van het infarct, daar lokale vervorming geen strikt lokaal fenomeen is. Op MRI na contrast toediening met late aankleuring (late gadolinium enhancement, LGE), bleek de halfwaardebreedte (full width at half maximum, FWHM) techniek bij voorkeur geschikt voor het herkennen van littekenweefsel in het hart. Littekenweefsel in het hart kan leiden tot hartfalen en hartritmestoornissen. Een inschatting maken van littekenweefsel in het hart is belangrijk voor het stellen van een diagnose, voorspellen van de prognose, het sturen van de behandeling en ter controle. Alhoewel de MRI met late aankleuring een nauwkeurige kwalitatieve maat van littekenweefsel in het hart verzorgt, heeft het potentiele nadelen als indirecte methode met de noodzaak om contrastmiddelen te gebruiken zonder te resulteren in een kwantitatieve maat voor littekenweefsel. Om de beperkingen gepaard gaande met het gebruik van contrastmiddelen te ondervangen, worden er een aantal MRI technieken zonder contrastmiddel (endogeen contrast) voor het herkennen van littekenweefsel besproken in hoofdstuk 5. In dit hoofdstuk worden ook voorbeelden van endogeen contrast MRI in de vorm van T1rho en T2* mapping bij een pilot studie in een chronisch varkens infarct model getoond. Endogeen contrast MRI van het hart voor herkenning van littekenweefsel kan de noodzaak van contrastmiddelen gebruik overbodig maken, wat gunstig is voor patiĂŤnten met nierfunctiestoornissen, kan de reproduceerbaarheid verbeteren en de scan tijd verkorten. Alhoewel T2* mapping een veelbelovende methode voor endogeen contrast kan zijn, is het onduidelijk of na herstel van de bloedstroom in de kransslagader bloedingen in de hartspier met ijzerafzettingen onderscheiden kunnen worden van littekenweefsel. Alhoewel endogene contrast methoden voor herkenning van littekenweefsel veelbelovend lijken, is er meer onderzoek nodig voordat het op grote schaal voor klinische besluitvorming kan worden aanbevolen. Deze studie resultaten rechtvaardigen verder onderzoek naar endogeen contrast MRI van het hart in patiĂŤnten met lokaal of diffuus littekenweefsel. Histologie Phospholamban (PLN) speelt een sleutelrol in de calciumhuishouding en calciumionen zijn cruciaal voor de elektrische activatie en het laten samentrekken van de hartspier. Alhoewel een aantal oorzakelijke mutaties in het PLN gen zijn beschreven in mensen, is de PLN R14del mutatie welke relatief vaak in Nederland voorkomt in verband gebracht met gedilateerde en aritmogene hartspierziekte. In het verleden zijn er slechts beperkte histologische (histologie = leer van de weefsels) analyses verricht van het hartweefsel van patiĂŤnten met de PLN R14del mutatie.

233

Voor zover wij weten is hoofdstuk 6 de eerste studie die onderzoek doet naar patronen van litteken- en vetweefsel in dwarsdoorsnedes van PLN R14del patiënten harten. Het littekenweefsel in het hart werd voornamelijk gezien in de achter- en zijwand van de linkerkamer, terwijl vetweefsel meer uitgesproken was in de buitenste laag hartweefsel van de rechterkamer. Naar aanleiding van onze bevindingen kan de gevoeligheid van hartbiopten uit de rechterkamer van het tussenschot in hartspierziekten met de PLN mutatie laag zijn. De gepresenteerde systematische methode om littekenweefsel te kwantificeren in dit hoofdstuk kan nuttig zijn om litteken- en vetweefsel vast te stellen in een breed scala aan hartspierziekten en als uitkomstmaat in preklinisch onderzoek. Door gebruik te maken van de methode uit hoofdstuk 6, hebben we in hoofdstuk 7 een karakteristiek verdelingspatroon van littekenweefsel in genetische hartspierziekten gevonden gerelateerd aan de mutatiegroep. De achter- en zijwand van de linkerkamer had een groot onderscheidend vermogen betreft het identificeren van de verschillende groepen mutaties. Deze patronen op histologie kunnen de weg wijzen voor beeldvorming van het hart in patiënten en kunnen gerelateerd worden aan de erfelijke informatie en processen op molecuulniveau in de cellen van het bijbehorende weefsel om onderliggende ziektemechanismen op te helderen. Bij de toekomstige naamgeving van soorten hartspierziekten is het wellicht meer gepast om het onderliggende gendefect te gebruiken. Identificeren van genen en paden die verband houden met het falende hart Ondanks vooruitgang in erfelijkheidsonderzoek, zijn er tot nu toe weinig plaatsen (loci) van genen in verband gebracht met het optreden van hartfalen. Hoewel de PLN R14del mutatie in verband gebracht is met zowel gedilateerde als aritmogene hartspierziekte, hebben genetische hartspierziekten een wisselende uiting en ernst van de ziekte, zelfs in dezelfde familie, dus mogelijk spelen andere regulerende factoren een rol. Next-generation sequencing maakt het mogelijk om regulerende elementen op DNA niveau te karakteriseren die het moment, de lokalisatie en het niveau van gen transcriptie beïnvloeden. Veranderingen in deze epigenetische karakteristieken kunnen dienen als een indicator voor het ontwikkelen van een ziekte. In hoofdstuk 8 hadden we tot doel om verschillen te identificeren in genregulatie en paden bij hartfalen tussen humaan hartweefsel van patiënten met een hartspierziekte op basis van de PLN R14del mutatie en gezonde controles met chromatine-immunoprecipitatie sequencing (ChIP-seq). We vonden 208 PLN R14del mutatie specifiek gereguleerde genen die gebruikt kunnen worden als biomarker om een risico-inschatting te maken. Deze genen waren voornamelijk betrokken bij paden die leiden tot littekenweefsel en ontwikkeling van het hart. Van deze genen zijn er 55 betrokken bij vetstofwisseling die verder bestudeerd kunnen worden als kandidaat genen voor ziekteprocessen met vervanging door litteken- en vetweefsel. De geïdentificeerde verschillend gereguleerde erfelijke gebieden in hartfalen kunnen uiteindelijk leiden tot een beter begrip van onderliggende ziektemechanismen, een betere risico-inschatting en nieuwe behandelopties. Mogelijkheden voor behandeling Door kennis van de erfelijke oorzaken te gebruiken kunnen gebreken op molecuulniveau die in verband zijn gebracht met hartfalen mogelijkheden bieden voor behandelopties, bijvoorbeeld met gentherapie. Stamceltherapie kan een veelbelovende techniek zijn voor herstel van de

234

SAMENVATTING IN HET NEDERLANDS

hartspier, door regeneratie van de hartspier bij hartfalen en daardoor een verbetering van de hartspierfunctie. Op dit moment is het ideale celtype nog niet naar voren gekomen. De voordelen van stamceltherapie bij chronische hartspierziekten behelzen een “stabiel” milieu en een relatief grote geschikte patiëntenpopulatie. In hoofdstuk 9 lijken de resultaten van stameltherapie voor gedilateerde hartspierziekte veelbelovend, maar dit hoofdstuk heeft aanbevelingen voor toekomstige studies, zoals het verbeteren van de methodologische kwaliteit, ophelderen van onderliggende mechanismen en valideren van veiligheid en effectiviteit voorafgaand aan implementatie in de klinische praktijk. Dit omvat aanbevelingen voor helder beschreven uniforme definities qua inclusiecriteria en gebruik van cardiale MRI om uitkomstmaten te bepalen. Er is een noodzaak voor overeenstemming binnen het onderzoeksveld betreft methodologisch deugdelijke studies. Met als doel om een vertaalslag naar de kliniek te maken in hoofdstuk 10, hebben we hartspier voorlopercellen (CMPCs) uit het menselijke hart als stamceltherapie gebruikt in een preklinische studie. We hebben een gerandomiseerde, placebo gecontroleerde studie met infusie van CMPCs in de kransslagader in een groot diermodel van ischemisch hartfalen uitgevoerd. Hoewel deze behandel methode uitvoerbaar en veilig lijkt, vonden we geen significante positieve effecten op de linkerkamerfunctie of grootte van het infarct. Daarom is de vertaling van de huidige CMPCs naar een first-in-man trial op dit moment nog niet gerechtvaardigd. Conclusie Het falende hart biedt uitdagingen die variëren van basale mechanismen via de translationele as naar de kliniek en omgekeerd met bevindingen uit de praktijk als aanleiding voor (laboratorium) onderzoek. In dit proefschrift hebben we mogelijkheden belicht om onderliggende mechanismen op te helderen, ziektemechanismen in beeld te brengen, het klinisch beeld te identificeren en de weg te banen voor toekomstige behandelstrategieën in het falende hart.

235

LIST OF PUBLICATIONS Ganesh SK*, Tragante V*, Guo W*, Guo Y, Lanktree MB, Smith EN, Johnson T, Castillo BA, Barnard J, Baumert J, Chang YP, Elbers CC, Farrall M, Fischer ME, Franceschini N, Gaunt TR, Gho JM, …, Doevendans PA, …, de Bakker PI*, Zhu X*, Levy D*, Keating BJ*, Asselbergs FW*. Loci influencing blood pressure identified using a cardiovascular gene-centric array. Hum Mol Genet. 2013;22:1663-78. *These authors contributed equally. Koudstaal S, Jansen of Lorkeers SJ, Gaetani R, Gho JM, van Slochteren FJ, Sluijter JP, Doevendans PA, Ellison GM, Chamuleau SA. Concise review: heart regeneration and the role of cardiac stem cells. Stem Cells Transl Med. 2013;2:434-43. Gho JM, Kummeling GJ, Koudstaal S, Jansen of Lorkeers SJ, Doevendans PA, Asselbergs FW, Chamuleau SA. Cell therapy, a novel remedy for dilated cardiomyopathy? A systematic review. J Card Fail. 2013;19:494-502. Gho JM, van Es R, Stathonikos N, Harakalova M, te Rijdt WP, Suurmeijer AJ, van der Heijden JF, de Jonge N, Chamuleau SA, de Weger RA, Asselbergs FW, Vink A. High resolution systematic digital histological quantification of cardiac fibrosis and adipose tissue in phospholamban p.Arg14del mutation associated cardiomyopathy. PloS One. 2014;9:e94820. Tragante V*, Barnes MR*, Ganesh SK*, Lanktree MB, Guo W, Franceschini N, Smith EN, Johnson T, Holmes MV, Karczewski KJ, Almoguera B, Barnard J, Baumert J, Chang YP, Elbers CC, Farrall M, Fischer ME, Gaunt TR, Gho JM, ..., Doevendans PA, ..., Levy D*, Asselbergs FW*, Munroe PB*, Keating BJ*. Gene-centric meta-analysis in 87,736 individuals of European ancestry identifies multiple blood pressure related loci. Am J Hum Genet. 2014;94:349-60. *These authors contributed equally. van Hout GP*, Jansen of Lorkeers SJ*, Gho JM, Doevendans PA, van Solinge WW, Pasterkamp GP, Chamuleau SA, Hoefer IE. Admittance based pressure volume loops versus gold standard cardiac magnetic resonance imaging in a porcine model of myocardial infarction. Physiol Rep. 2014;2:e00287. *These authors contributed equally. Koudstaal S, Jansen of Lorkeers SJ, Gho JM, van Hout GP, Jansen MS, Grundeman PF, Pasterkamp G, Doevendans PA, Hoefer IE, Chamuleau SA. Myocardial infarction and functional outcome assessment in pigs. J Vis Exp. 2014;(86). van Oorschot JW, El Aidi H, Jansen of Lorkeers SJ, Gho JM, Froeling M, Visser F, Chamuleau SA, Doevendans PA, Luijten PR, Leiner T, Zwanenburg JJ. Endogenous assessment of chronic myocardial infarction with T1ρ-mapping in patients. J Cardiovasc Magn Reson. 2014;16:104.

236

LIST OF PUBLICATIONS

Karakikes I*, Stillitano F*, Nonnenmacher M*, Tzimas C, Sanoudou D, Termglinchan V, Kong CW, Rushing S, Hansen J, Ceholski D, Kolokathis F, Kremastinos D, Katoulis A, Ren L, Cohen N, Gho JM, Tsiapras D, Vink A, Wu JC, Asselbergs FW, Li RA, Hulot JS, Kranias EG, Hajjar RJ. Correction of the human phospholamban R14del mutation associated with cardiomyopathy using targeted nucleases and combination therapy. Nat Commun. 2015;6:6955. *These authors contributed equally. van Oorschot JW*, Gho JM*, van Hout GP, Froeling M, Jansen of Lorkeers SJ, Hoefer IE, Doevendans PA, Luijten PR, Chamuleau SA, Zwanenburg JJ. Endogenous contrast MRI of cardiac fibrosis: beyond late gadolinium enhancement. J Magn Reson Imaging 2015;41:1181-9. *These authors contributed equally. Gho JM*, van Es R*, van Slochteren FJ, Jansen of Lorkeers SJ, Hauer AJ, van Oorschot JW, Doevendans PA, Leiner T, Vink A, Asselbergs FW, Chamuleau SA. A systematic comparison of cardiovascular magnetic resonance and high resolution histological fibrosis quantification in a chronic porcine infarct model. Submitted. *These authors contributed equally. Jansen of Lorkeers SJ, Gho JM, Koudstaal S, van Hout GP, Zwetsloot PP, van Oorschot JW, van Eeuwijk EC, Leiner T, Hoefer IE, Goumans MJ, Doevendans PA, Sluijter JP, Chamuleau SA. Xenotransplantation of human cardiomyocyte progenitor cells does not improve cardiac function in a porcine model of chronic ischemic heart failure. Submitted.

237

ACKNOWLEDGEMENTS / DANKWOORD In de afgelopen jaren is dit proefschrift tot stand gekomen door de samenwerking tussen uiteenlopende disciplines, in binnen- en buitenland. Ik heb volop genoten van deze leerzame en gezellige promotietijd. Hierbij wil ik iedereen bedanken die hier een bijdrage aan hebben geleverd. Geachte prof.dr. Doevendans, beste Pieter, bij ons eerste gesprek betreft een mogelijk promotietraject had je meteen een aantal mogelijkheden in gedachten betreft onderzoek en het advies was om eens met de andere begeleiders te gaan praten. Ik wist nog niet waar dit naartoe zou leiden toen je vertelde dat het met deze begeleiders een “mooi feestje” zou worden. Uiteindelijk had ik me geen ander promotietraject kunnen wensen. Dankzij je sturende rol zijn er onder andere mooie samenwerkingsverbanden ontstaan binnen SMARTCARE en met Stichting Genetische Hartspierziekte PLN die binnen dit proefschrift samenkomen en hopelijk vervolg krijgen. Beste Folkert, inmiddels prof.dr. Asselbergs, je streven om zaken bij voorkeur gisteren gedaan te krijgen en respons op e-mails met de snelheid van next-generation sequencing houdt de vaart erin. Daarnaast viel mij meteen je directheid op welke goed van pas komt om je heldere visie, die je goed weet over te brengen, waar te maken. Ik vond het leuk om relatief ongebaande paden te verkennen waarbij nieuwe onderzoekslijnen werden gecreëerd. Dank voor de mogelijkheid om aan verschillende studies in binnen- en buitenland mee te werken, als wereldreiziger schoot ook de begeleiding in Londen nooit tekort. Ik zie ernaar uit om binnenkort te gaan tennissen als je een keer in Utrecht bent. Geachte dr. Chamuleau, beste Steven, ik zal nooit de wijze les vergeten tijdens een van de eerste weekenddiensten waar ik als semi-arts meeliep in het UMC Utrecht. Zoals je van je eigen opleider had geleerd, adviseerde je me om na een paar dagen nog even contact op te nemen met de patiënt die zich op de spoedeisende hulp had gepresenteerd en weer huiswaarts was gegaan. Gelukkig voelde de patiënt zich alweer een stuk beter. Die onderwijzende en motiverende rol is je op het lijf geschreven. Een van de assistenten toentertijd raadde me aan om contact met je op te nemen als ik interesse had in wetenschappelijk onderzoek. Uiteindelijk ben ik zeer dankbaar voor dit traject in de breedste zin van het woord, met teveel opportunities om op te noemen zoals de leuke een leerzame congresbezoeken in Washington en Madrid, vooruitstrevende projecten en de nuttige besteding van het persoonlijk budget. Ik waardeer je openheid en eerlijkheid en hoop dat ik de komende tijd nog veel van je kan leren. Nu ik zo in mijn mailbox terugkijk was het de bedoeling om onderzoek te doen naar DCM (op het snijvlak imaging, cardiogenetica en stamcellen) en als ik door dit boekje heen blader is dat to the best of my knowledge aardig gelukt. Maar we zijn er nog niet, er is in de toekomst nog genoeg te onderzoeken en ik hoop dat we kunnen blijven samenwerken from bench-to-bedside.

238

ACKNOWLEDGEMENTS / DANKWOORD

Beste Aryan, dr. Vink, het is een eer voor mij dat je op het laatste moment als 2e copromotor op kan treden. Datgene wat volgens mij als zijproject is begonnen vormt nu een integraal deel van dit proefschrift. Ik zal niet vergeten dat je mij bij het eerste telefoongesprek adviseerde om in Robbins te kijken. Dank voor je intensieve begeleiding, enthousiasme en bevlogenheid. Beste leden van de leescommissie, prof.dr. R. Goldschmeding, prof.dr. M.A.A.J. van den Bosch, prof.dr. M.R. Vriens, prof.dr. A.C. van Rossum en prof.dr. E.D. de Muinck, bedankt voor de bereidheid om zitting te nemen in de beoordelingscommissie van dit proefschrift. De stafleden van de afdeling Cardiologie in het UMC Utrecht, dank voor de bijdragen aan mijn onderzoek in de afgelopen jaren en de leerzame supervisie in de kliniek. Beste dr. Kirkels, dank voor de mogelijkheid om in de kliniek aan de slag te gaan met vele onderwijs momenten. Maarten-Jan, dank voor je nimmer aflatende enthousiasme en goede raad betreft de kliniek en onderzoek. Het stafsecretariaat, Jantine, Sylvia en Tamara, dank voor al jullie hulp in de afgelopen jaren. Jonne en Ingrid, dank voor jullie gezelligheid, hulp en kunde om afspraken te maken met mijn dagelijkse begeleiders en velerlei vragen tot een goed einde te brengen. Dank aan alle mede-onderzoekers van de afgelopen jaren, Anneline (voor de discussies over PLN, tot snel in de kliniek!), Cas, Cheyenne (sightseeing Madrid), Crystel, Dirk (zie je bij de volgende BBQ), Fatih G, Freek (eindelijk gebruik maken van de Museumkaart met Mariam), Geert van H (naast onderzoek ook voor de gezamenlijke tijd in de kliniek), Gijs (voor je onderzoeksen muzikale input), Hamza, Iris, Jetske (voor de vermakelijke borrels en filmpjes), Judith (brainstormen over PLN), Manon (hopelijk tot snel in het Diak!), Mariam, Marloes, Martine, Mieke (fietsen en basketbal kijken in Washington), Mira (op zoek naar het Villa pleintje in Madrid), Peter-Paul (gecellige borrels en leerzame discussies), Remco (als kamergenoot in Oirschot tot in de Villa en organisator van menig uitje), RenĂŠ (voor het tot in de late uurtjes optimaliseren van de manuscripten, borrelen en de vele sportactiviteiten), Rosemarijn (van buur bij de IBB tot de Rotaractse activiteiten), Sanne (voor de vele en gezellige uren, Oink! momenten en hopelijk tot ziens in de kliniek!), Sjoukje, Sofieke, Stefan (van bunker tot Villa en een paar keer de Noordzee over), Thijs (nice dat je ook begonnen bent), Thomas, Tycho (van tips in de bunker, op de tennisbaan tot aan B4West), Vivan, Willemien (vanaf A0 in Amersfoort tot het ACC in Washington), Wouter G en Wouter van E, voor al jullie hulp en gezelligheid. Alle arts-assistenten Cardiologie in het UMC Utrecht, dank voor jullie collegialiteit, geduld en de leerzame en gezellige sfeer. De R&D van de Cardiologie, dank voor jullie gezelligheid, inzet en hulp bij de verschillende studies.

239

Alle betrokkenen van de Experimentele Cardiologie. Dank voor jullie bijdrage aan de projecten, met name op basaal gebied. Alle biotechnici, GDL medewerkers en dierverzorgers, dank voor jullie hulp, expertise en geduld bij de soms langdurige experimenten. Alle betrokkenen van de afdeling Radiologie betreft de vele MRI uren, in het bijzonder Joep en Jaco onder begeleiding van prof.dr. Luijten dank voor de fysieke inzet en de fysische input en prof.dr. Leiner dank voor de waardevolle scherpe klinische blik. Mooi om de vertaling van concept naar klinische studie te zien. Een aantal van de afdeling Pathologie zijn al eerder genoemd, maar hierbij wil ik ook in het bijzonder Shahrzad, Roel en de betrokken analisten bedanken voor de onmisbare hulp, mogelijkheden en expertise bij het onderzoek naar de erfelijke hartspierziekten. Nikolas, thank you for your extensive help with creating the fibrosis quantifier (FibroQuant). Magdalena, thank you for all your help and precise efforts, especially regarding epigenetics, also together with Michal. Vinicius, thanks for all your bioinformatic input and output over the last years. Alle studenten die aan de projecten bij dit proefschrift hebben meegewerkt. In het bijzonder Maartje voor de hulp bij de AGNES studie. En Einar, succes met je onderzoek. Uit het AMC Amsterdam alle betrokkenen bij de AGNES studie. Ik hoop voor de toekomst op een blijvende samenwerking, bijvoorbeeld in het GENIUS-CHD consortium. I would like to thank all the people at the Farr Institute. Thank you for all your help with epidemiological and statistical input, R-coding skills, but also the wonderful time with (International cuisine) lunches and post-work drinks in London. I am still not tired of London and hope to visit you again soon. Dr. Hajjar and Francesca, thank you for your collaboration regarding PLN and visit to your lab in New York. Hope your research will pave the way for novel therapies in genetic cardiomyopathies. Pieter Glijnis, namens Stichting Genetische Hartspierziekte PLN. Dank voor alle ondersteuning en de interesse gedurende het onderzoek, ik hoop dat er in de toekomst betere behandelmogelijkheden komen voor deze hartspierziekte. Van het Meander MC de maatschap Cardiologie en verpleegkundigen, dank voor de leerzame perioden onderzoek en in de kliniek, maar vooral ook gezellige tijd in Amersfoort. In het bijzonder dr. Senden, beste Jeff, bedankt voor de leerzame tijd in de kliniek en het advies om onderzoek te gaan doen met â&#x20AC;&#x153;stamcellenâ&#x20AC;?. Hoop jullie weer te zien bij het volgende zeiluitje!

240

ACKNOWLEDGEMENTS / DANKWOORD

William Kramer dank voor de begeleiding tijdens de studie in de congrescommissie en bij het duikgeneeskunde artikel. Wendy, dank voor je inzet en betrokkenheid bij de lay-out. Robert, dank voor je sprekende ontwerp voor de omslag. Mijn paranimfen, Ing Lok en Frebus. Beste Ing Lok, lief broertje, ik vind het mooi dat we zoveel interesses delen, we hebben de afgelopen jaren al menig uitstapje gemaakt van wintersport tot Wimbledon. Ik bewonder je sociale kant en hulpvaardigheid en vind het leuk dat je nu ook in het Meander MC aan de slag bent gegaan. Beste Frebus, mien jong, Slo is inmiddels een begrip geworden, leuk om de afgelopen jaren met je samengewerkt te hebben waarbij je unieke technische blik echt een eyeopener is geweest. Daarnaast herinner ik me ook vele mooie uitstapjes en feestjes en heb ik je zien groeien als â&#x20AC;&#x153;CEslOâ&#x20AC;? van je eigen bedrijf. Leuk dat jullie mijn paranimfen zijn. Dank aan al mijn familie en vrienden die mij de afgelopen jaren hebben gesteund en interesse hebben getoond voor wat ik ook alweer aan het doen was. Dank voor jullie gastvrijheid en gezelligheid. Mama en papa, dank voor jullie opvoeding, steun en goede raad in de afgelopen jaren waar ik altijd op kan rekenen. Mama, ik ben nog altijd dankbaar voor je advies om Biologie toch als keuzevak erbij te kiezen op de middelbare school wat ik toen toch maar heb opgevolgd.

241

CURRICULUM VITAE Johannes Michael Ing Han Gho was born on 29 September 1986 in Leusden, the Netherlands. After attending the Meridiaan College, Het Nieuwe Eemland in Amersfoort where he obtained his gymnasium diploma in 2004, he studied medicine at Utrecht University. During his medical study, he passed the propaedeutic exam cum laude (with honor), spent time abroad in Malaysia, Australia and Malawi for internships and became interested in Cardiology and scientific research. After graduation from medical school in 2010, he started working as a resident (ANIOS) at the Cardiology department of the Meander Medical Center in Amersfoort under supervision of dr. P.J. Senden. In January 2012 he started working as a PhD student at the Cardiology department of the University Medical Center Utrecht under supervision of prof.dr. P.A.F.M. Doevendans, prof.dr. F.W. Asselbergs, dr. S.A.J. Chamuleau and dr. A. Vink. During this period he worked on several research projects with a main focus on heart failure, including genetic, regenerative medicine, imaging, histological and epidemiological studies. He also conducted 10 weeks of research at the Farr Institute of Health Informatics Research, University College London (UCL), London, United Kingdom, which resulted in a manuscript included in this thesis (Chapter 2). In July 2015 he started working as a resident (ANIOS) at the Cardiology department of the University Medical Center Utrecht and he will start his formal Cardiology training under supervision of dr. J.H. Kirkels in January 2016.

242