Contribution to the Diagnosis and Prognosis of Canine

PAUL FRIDTJOF MÕTSKÜLA CONTRIBUTION TO THE DIAGNOSIS AND PROGNOSIS OF CANINE CARDIAC DISEASE THROUGH CONTRIBUTION TO THE DIAGNOSIS AND PROGNOSIS OF CANINE CARDIAC DISEASE THROUGH ELECTRODIAGNOSTICS AND DIAGNOSTIC IMAGING ELECTRODIAGNOSTICS AND DIAGNOSTIC IMAGING ELEKTRO- JA PILTDIAGNOSTIKA TÄIENDAVAD RAKENDUSED KOERTE SÜDAMEHAIGUSTE DIAGNOOSIMISEL NING PROGNOOSIMISEL

PAUL FRIDTJOF MÕTSKÜLA

A Thesis for applying for the degree of Doctor of Philosophy in Veterinary Science

Väitekiri filosoofiadoktori kraadi taotlemiseks loomaarstiteaduse erialal

Tartu 2020

Eesti Maaülikooli doktoritööd

Doctoral Theses of the Estonian University of Life Sciences

CONTRIBUTION TO THE DIAGNOSIS AND PROGNOSIS OF CANINE CARDIAC DISEASE THROUGH ELECTRODIAGNOSTICS AND DIAGNOSTIC IMAGING

ELEKTRO- JA PILTDIAGNOSTIKA TÄIENDAVAD RAKENDUSED KOERTE SÜDAMEHAIGUSTE DIAGNOOSIMISEL NING PROGNOOSIMISEL

PAUL FRIDTJOF MÕTSKÜLA

A Thesis for applying for the degree of Doctor of Philosophy in Veterinary Science

Väitekiri filosoofiadoktori kraadi taotlemiseks loomaarstiteaduse erialal

Tartu 2020 Institute of Veterinary Medicine and Animal Sciences Estonian University of Life Sciences According to Decision N° 4 of October 01, 2020, the Doctoral Committee of Veterinary and Food Science of Estonian University of Life Sciences has accepted the thesis as ready for defence of the degree of Doctor of Philosophy in Veterinary Science.

Opponent: Prof. Dr Gerhard Wess, Dip ACVIM Dip ECVIM-CA (cardiology) Dip ECVIM-CA (internal medicine) Small Animal Clinic Faculty of Veterinary Medicine Ludwig Maximilian University, Munich, Germany

Supervisors: Prof. Toomas Orro, DVM PhD Chair of Clinical Veterinary Medicine Institute of Veterinary Medicine and Animal Sciences Estonian University of Life Sciences

Prof. Virginia Luis Fuentes, MA VetMB PhD CertVR DVC MRCVS DipACVIM DipECVIM Department of Clinical Science and Services Queen Mother Hospital for Animals The Royal Veterinary College, UK

Prof. David Connolly, BSc BVetMed PhD CertVC CertSAM DipECVIM (Cardiology) MRCVS Department of Clinical Science and Services Queen Mother Hospital for Animals The Royal Veterinary College, UK

Dr. Ranno Viitmaa, DVM PhD Chair of Clinical Veterinary Medicine Institute of Veterinary Medicine and Animal Sciences Estonian University of Life Sciences

Defence of the thesis: International video conference in the BigBlueButton web room at 10:00, March 8, 2021.

The publication of this dissertation is supported by the Estonian University of Life Sciences.

CONTENTS

LIST OF ORIGINAL PUBLICATIONS...... 10 ABBREVIATIONS...... 12 1. INTRODUCTION...... 14 2. REVIEW OF THE LITERATURE...... 17 2.1. Classification of cardiomyopathies...... 17 2.1.1. Classification of human cardiomyopathies...... 17 2.1.2. Classification of canine cardiomyopathies...... 18 2.1.2.1. Boxer cardiomyopathy...... 18 2.1.2.2. Cardiomyopathies and other cardiac pathologies in Dogue de Bordeaux...... 20 2.2. Methods used for diagnostics of cardiomyopathies...... 21 2.2.1. Clinical features...... 21 2.2.2. Electrodiagnostics...... 22 2.2.2.1. 12-lead ECG...... 22 2.2.2.2. Ambulatory ECG (Holter monitoring)...... 23 2.2.2.3. Event monitoring...... 24 2.2.3. Diagnostic imaging...... 24 2.2.3.1. Radiography...... 24 2.2.3.2. Echocardiography...... 24 2.2.3.3. cMRI...... 25 2.2.4. Genetics...... 26 2.3. Prognostic factors...... 27 2.3.1. Dilated cardiomyopathy...... 28 2.3.2. Arrhythmogenic right ventricular cardiomyopathy...... 29 3. AIMS OF THE STUDY...... 31 4. MATERIALS AND METHODS...... 32 4.1. Study population...... 32 4.1.1. Inclusion and exclusion criteria...... 32 4.1.2. Baseline clinical data (I–III)...... 33 4.1.3. Survival information (I, II)...... 33 4.2. Echocardiography data (I–IV)...... 33 4.3. Electrocardiographic data (I–III)...... 35 4.4. Cardiac magnetic resonance imaging data (IV)...... 35 4.5. Ethical considerations...... 36 4.6. Statistical analysis...... 37 5. RESULTS...... 39 5.1. Study population (I–IV)...... 39 5.2. Electrocardiographic data (I–III)...... 42 5.3. Echocardiographic data (I–IV)...... 43 5.4. Follow-up and survival information (I, II)...... 46 5.5. Cardiac magnetic resonance imaging data (IV)...... 47 6. DISCUSSION...... 49 6.1. General considerations...... 49 6.2. Prognostic factors of arrhythmogenic right ventricular cardiomyopathy in Boxers (I, II)...... 50 6.2.1. Syncope...... 50 6.2.2. Systolic dysfunction...... 50 6.2.3. Signalment: sex and age...... 51 6.2.4. Tricuspid annular plane systolic excursion measurement (II)...... 52 6.2.5. Ventricular arrhythmias (I)...... 53 6.3. Cardiac findings in a population of Dogue de Bordeaux (III)..54 6.3.1. Population overview...... 54 6.3.2. Supraventricular arrhythmias and left ventricular systolic dysfunction...... 55 6.3.3. Supraventricular arrhythmias and right atrial and ventricular dilation...... 56 6.3.4. Ventricular arrhythmias and right ventricular dilation...... 56 6.3.5. Supraventricular tachycardia and age...... 57 6.4. Comparison of echocardiography with cardiac magnetic resonance imaging (IV)...... 58 7. CONCLUSIONS...... 61 8. REFERENCES...... 62 9. SUMMARY IN ESTONIAN...... 83 10. ACKNOWLEDGEMENTS...... 91 ORIGINAL PUBLICATIONS...... 93 CURRICULUM VITAE...... 137 ELULOOKIRJELDUS...... 140 LIST OF PUBLICATIONS...... 143 LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the original studies reported in the following four original publications (I–IV), referred to in the text by their respective Roman numerals:

I Mõtsküla P.F., Linney C., Palermo V., Connolly D.J., French A., Dukes McEwan J., Luis Fuentes V., 2013. Prognostic value of 24-hour ambulatory ECG (Holter) monitoring in Boxer dogs. Journal of Veterinary Internal Medicine 27, 904–912.

II Kaye B.M., Borgeat K., Mõtsküla P.F., Luis Fuentes V., Connolly D.J., 2015. Association of tricuspid annular plane systolic excursion with survival time in Boxer dogs with ventricular arrhythmias. Journal of Veterinary Internal Medicine 29, 582– 588.

III McAulay G., Borgeat K., Sargent J., Mõtsküla P.F., Neves J., Dukes-McEwan J., Luis Fuentes V., 2018. Phenotypic description of cardiac findings in a population of Dogue de Bordeaux with an emphasis on atrial fibrillation. Veterinary Journal 234, 111– 118.

IV Sargent J., Connolly D.J., Watts V., Mõtsküla P.F., Volk H.A., Lamb C.R., Luis Fuentes V., 2015. Assessment of mitral regurgitation in dogs: comparison of results of echocardiography with magnetic resonance imaging. Journal of Small Animal Practice 56, 641–650.

10 The contribution of the authors to the research papers

Original idea, Data Preparation Paper Data analysis study design collection of manuscript PM VLF DC PM DC VLF I PM CL VP PM VLF VP CL JDM VP CL AF

BK KB DC II DC KB KB BK PM KB BK DC PM VLF

GM VLF KB GM JS KB GM VLF JS III GM VLF JS PM PM JN JDM PM JN

JS DC VLF JS VW VLF JS VLF DC CL IV JS VLF VW CL PM HV PM VW PM – Paul Mõtsküla, DC – David Connolly, VLF – Virginia Luis Fuentes, JD – Jo Dukes McEwan, AF – Anne French, VP – Valentina Palermo, CL – Chris Linney, BK – Ben Kaye, KB – Kieran Borgeat, JS – Julia Sargent, VW – Victoria Watts, HV – Holger Volk, CL – Chris Lamb, GM – Gavin McAulay, JN – João Neves

I would like to regret that the Estonian University of Life Sciences was omitted as the institution of my employment affiliation and PhD studies in three of the publication of the original articles that this thesis is based upon.

11 ABBREVIATIONS

2D two-dimensional AF atrial fibrillation AIC arrhythmia induced cardiomyopathy ARVC arrhythmogenic right ventricular cardiomyopathy CHF congestive heart failure CI confidence interval cMRI cardiac magnetic resonance imaging cMRI-RF regurgitant fraction determined by cardiac magnetic resonance imaging DCM dilated cardiomyopathy DdB Dogue de Bordeaux ECG electrocardiography

Evel peak early transmitral filling velocity FS% fractional shortening HCM hypertrophic cardiomyopathy ICC intraclass correlation coefficient IQR interquartile range ISFC International Society and Federation of Cardiology JAR jet area ratio LA left atrium; left atrial LA/Ao left atrial to aortic ratio LV left ventricle, left ventricular LVDdn normalised left ventricular diastolic diameter LVIDd left ventricular internal diameter in diastole LVIDs left ventricular internal diameter in systole

LVIDs>35mm/CHF group with systolic dysfunction or presence of congestive heart failure

LVIDs≤35mm/no CHF group without systolic dysfunction or congestive heart failure M-mode motion mode ultrasound imaging MMVD myxomatous mitral valve degeneration

12 MR mitral regurgitation MRT magnetresonantstomograafia non-CHDN dogs with non-cardiac neoplasia, echocardiographically detected cardiac soft tissue masses, or congenital heart disease PISA proximal isovelocity surface area RCM restrictive cardiomyopathy RV right ventricle, right ventricular S’ pulsed wave tissue Doppler imaging derived systolic myocardial velocity of the lateral tricuspid annulus SVT supraventricular tachycardia TAPSE tricuspid annular plane systolic excursion TNM tumour, (lymph)node involvement, metastasis UCM unclassified cardiomyopathy VC vena contracta VPC ventricular premature contraction vs versus VT ventricular tachycardia WHO World Health Organisation

13 1. INTRODUCTION

The original 1980 World Health Organisation (WHO)/International Society and Federation of Cardiology (ISFC) classification defined cardiomyopathies as the heart muscle diseases of unknown cause and included dilated (DCM), hypertrophic (HCM) and restrictive (RCM) cardiomyopathies (WHO/ISFC task force, 1980). The updated WHO/ ISFC classification, published in 1995, clarified the definition by stating that cardiomyopathies are diseases of myocardium associated with cardiac dysfunction. Also, the arrhythmogenic right ventricular cardiomyopathy (ARVC) and unclassified cardiomyopathy (UCM) were added to the new classification. Furthermore, as the new classification introduced the term “specific cardiomyopathies” to describe the heart muscle diseases that are associated with specific cardiac or systemic disorders based on the dominant pathophysiology or aetiological/pathogenetic factors (Richardson et al., 1996).

Arrhythmia induced cardiomyopathy (AIC), also referred to as tachycardia induced cardiomyopathy (Shinbane et al., 1997; Ellis and Josephson, 2013), is a condition affecting the myocardium as a result of rapid atrial or ventricular arrhythmias, resulting in myocardial remodelling as well as systolic and diastolic dysfunction (Simantirakis et al., 2012; Gopinathannair et al., 2015; Van Gelder et al., 2016). AIC is a generally reversible form of cardiomyopathy if the causative tachyarrhythmia is successfully treated (Shinbane et al., 1997; Simantirakis et al., 2012; Ellis and Josephson, 2013; Sossalla and Vollmann, 2018). Whilst the exact pathophysiological mechanism for AIC remains unclear, metabolic dysfunction, neurohormonal mechanisms and abnormal cellular calcium- handling are commonly implicated (Raymond-Paquin et al., 2018). The most common arrhythmias associated with AIC in humans are atrial fibrillation (AF), focal atrial tachycardia and frequent ventricular ectopy (Kusano, 2018; Raymond-Paquin et al., 2018).

Cardiac mortality accounts for a large proportion of overall mortality in the canine population. The most common causes of cardiac mortality in this species are myxomatous mitral valve degeneration (MMVD) and cardiomyopathies, accounting for approximately 1/3 of deaths in dogs in heartworm non-endemic areas (Egenvall et al., 2006). The estimated prevalence of cardiomyopathies varies largely depending on the type of

14 cardiomyopathy and the dog breed. A study in the first opinion practices estimated the prevalence of DCM to be 0.05% (Gane et al., 2017). In some breeds the prevalence of DCM is estimated to be as high as 24% in Irish Wolfhounds (Vollmar, 2000), 36–47% in Great Danes (Stephenson et al., 2012) and 58% in Doberman Pinchers (Wess et al., 2010). The exact prevalence of ARVC in Boxer dogs or other breeds is not known, mainly due to the lack of definitive diagnostic criteria, difficulties in ante-mortem diagnostics and variable clinical presentation (Hariu and Carpenter, 2010). However, the prevalence appears to be higher in certain families of dogs, suggesting a significant genetic predisposition (Meurs, 2017).

The study of cardiomyopathies in animals serves not only a purpose for veterinary medicine, but also contributes to a better understanding of the underlying pathophysiological processes, thus providing a valuable model to the mechanisms of human cardiomyopathies (McCauley and Wehrens, 2009). Feline (Fox et al., 2000) and canine (Basso et al., 2004; Vischer et al., 2017) ARVC as well as the AIC (Zupan et al., 1996; Shinbane et al., 1997) have been described as a model for the human counterpart.

This thesis focuses on the various diagnostic and prognostic aspects of canine cardiomyopathies. The first and second articles focus on the prognostic information of an electrodiagnostic (I) and echocardiographic (II) method used for diagnosis of ARVC. Despite being commonly used, these methods have not been previously validated in animal cohorts for prognostic purposes.

Despite a large overlap between the clinical signs, progression and diagnostic findings between different canine cardiomyopathies, it is commonly accepted that significant breed differences exist. It is generally preferable to avoid common terms such as DCM or ARVC when describing a similar disease in different breeds, and it would be more appropriate to use the terminology referring to a specific breed (e.g. Boxer cardiomyopathy). To further fill the gaps in the current understanding of the pathogenesis, and clinical course of canine cardiomyopathies, article III is focused on the description of cardiac pathologies and their prevalence in another large dog breed commonly known to have a high prevalence of cardiomyopathies and cardiac arrhythmias, the Dogue de Bordeaux (DdB).

15 The definite ante-mortem diagnosis of cardiomyopathies is challenging as the definitive diagnosis should be established histologically. Although myocardial biopsies can be obtained, it is not a feasible method in the clinical setting due to the requirement for general anaesthesia, very specialised equipment and skills as well as the possible risk of adverse, potentially life-threatening complications. Therefore, minimally- or non-invasive diagnostic methods would be preferred. Cardiac magnetic resonance imaging (cMRI) has emerged as a second-line diagnostic modality in diagnostic workup of human cardiomyopathies, especially for the ARVC. To date, only very limited information has been published on the use of cMRI in veterinary cardiology. Echocardiography, although being safer and technically easier to perform, has many limitations due to geometrical assumptions. The cMRI provides more accurate and reproducible measurement of cardiac chamber dimensions, volumes and function (Sugeng et al., 2010; Oomen et al., 2018). Article IV seeks to fill this gap in the veterinary literature. Although the article describes this method in the context of valvular heart disease rather than a myocardial disease, it contributes to the application of cMRI in veterinary cardiology. Furthermore, through comparison of the volumetric echocardiographic and cMRI methods, the findings of this article help to validate the safer and more readily available echocardiographic method used in veterinary cardiology.

16 2. REVIEW OF THE LITERATURE

2.1. Classification of cardiomyopathies

2.1.1. Classification of human cardiomyopathies

The term ‘cardiomyopathy’ was originally used to describe heart muscle disease based on morphological and pathophysiological characteristics (Elliott et al., 2008). The original 1980 WHO/ISFC classification defined cardiomyopathies as the heart muscle diseases of unknown cause, differentiating them from the heart muscle diseases of known cause (WHO/ISFC task force, 1980) (Richardson et al., 1996). With the increasing knowledge and understanding of the diseases, the difference between these two groups became less clear, leading to the need for an update of this classification. This updated WHO/ISFC classification was published in 1995, clarifying the definition by stating that cardiomyopathies are diseases of the myocardium associated with cardiac dysfunction (Richardson et al., 1996). The original classification, consisting of three types of cardiomyopathies – DCM, HCM, and RCM – was supplemented with ARVC and UCM. This classification was based on the dominant pathophysiological mechanism or known aetiological/ pathogenetic factors (Richardson et al., 1996). Also the term ‘specific cardiomyopathies’ was introduced describing the heart muscle diseases associated with specific cardiac or systemic disorders, e.g. hypertensive, inflammatory, or metabolic cardiomyopathies (Richardson et al., 1996). Further updates to the classification were made to reflect the increase in understanding of the aetiological/physiological mechanisms and the genetic background of the disease, and to ensure the continued practical value in clinical practice (Maron et al., 2006). The European Society of Cardiology classification defined cardiomyopathy as ‘A myocardial disorder in which the heart muscle is structurally and functionally abnormal, in the absence of coronary artery disease, hypertension, valvular disease and congenital heart disease sufficient to cause the observed myocardial abnormality’ (Elliott et al., 2008). This morpho- functional classification also considered the familial/genetic background of the cardiomyopathies, sub-classifying each phenotype to familial and non-familial forms. The non-familial forms were further subclassified into idiopathic (no known cause) and acquired cardiomyopathies (Elliott

17 et al., 2008). A more detailed phenotype-genotype based classification was proposed in 2013, similar to the TNM (tumour, lymph node involvement, metastasis) staging used in oncology (Arbustini et al., 2013). This classification was aimed towards an increased clinical by incorporating the morpho-functional, organ/system involvement, genetic, aetiological characteristics as well as the clinical stage (Westphal et al., 2017). Although such level of detail of classification may be useful in human medicine for improved diagnosis and treatment outcomes, to the author’s knowledge it has neither been implemented nor validated in veterinary medicine.

2.1.2. Classification of canine cardiomyopathies

In the veterinary literature, cardiomyopathies are commonly described based on morphological appearance, similarly to the human WHO/ ISFC classification. The two most prevalent types of cardiomyopathies in dogs are DCM and ARVC; HCM has also been described but is relatively uncommon in this species.

However, there is a significant inter-breed variation in the prevalence of the different forms of cardiomyopathy and mortality rates as well as clinical course and prognosis (Calvert et al., 1997; Dambach et al., 1999; Vollmar, 2000; Meurs, 2004; Martin et al., 2009; Wess et al., 2010; Stephenson et al., 2012; Aguilar and Nattel, 2018). Therefore, it may be more appropriate to use the terminology based on the breed (e.g. Boxer cardiomyopathy, Doberman cardiomyopathy, Portuguese Water Dog Cardiomyopathy). Even with such breed-specific classification, it is possible that multiple phenotypically similar, but genetically different diseases are grouped under one single category. Or due to variable penetrance, one gene defect may lead to differences in the phenotypic expression.

2.1.2.1. Boxer cardiomyopathy

Boxer cardiomyopathy was first described by Harpster in 1983 (Harpster, 1983). His description included three clinical categories – concealed, overt and myocardial dysfunction forms (Meurs, 2004). The disease is commonly referred to as ARVC due to the similarities between Boxer cardiomyopathy and human ARVC (Meurs et al., 1999; Hariu and Carpenter, 2010; Protonotarios et al., 2011). Despite the advancement 18 in knowledge, the ante-mortem diagnostic criteria and natural history of ARVC in Boxer dogs remain poorly defined and the definitive diagnosis is based on the necropsy finding of fibro-fatty replacement of the right ventricular (RV) myocardium (Meurs et al., 1999; Basso et al., 2004). Episodic weakness and syncope are the most common clinical signs, although this presentation is also common in other conditions such as aortic stenosis or neurally mediated syncope (Fisher, 1971; Thomason et al., 2008; Bussadori et al., 2009). The most common diagnostic tool used for diagnosing ARVC in Boxers is 24-hour ambulatory (Holter) monitoring, with various cut-off values for the number of ventricular premature contractions (VPC)/24h being proposed to distinguish between affected and nonaffected individuals (e.g. 50 (Meurs et al., 1999), 100 (Meurs, 2004; O’Grady and O’Sullivan, 2004; Smith et al., 2008; Scansen et al., 2009; Visser et al., 2013), 1000 (Spier and Meurs, 2004a; Baumwart and Meurs, 2005; Hariu and Carpenter, 2010; Palermo et al., 2011). The criteria for the diagnosis of human ARVC include the number of VPC/24h as a minor criterion, with 500 VPC/24h being used as a cut-off value (Marcus et al., 2010). The presence of sustained or non-sustained ventricular tachycardia (VT), a major criterion in human ARVC diagnostics, has not been used in dogs, although a modification of the Lown (Lown and Wolf, 1971) grading system of severity of the ventricular escape beats has been recommended by Spier and Meurs (2004b).

In the absence of systolic dysfunction, the echocardiographic examination findings in Boxer ARVC may be completely unremarkable (Baumwart et al., 2005). The quantitative assessment of the RV dimensions and function using echocardiography is hindered by the complex geometry, significant loading dependence and ventricular interdependence (Visser et al., 2015). Borgquist et al. (2014) found that in comparison with cMRI, echocardiographic examination was unreliable in the diagnosis of ARVC in people with subtle structural changes. The major echocardiographic criteria for the diagnosis of human ARVC include RV wall motion abnormalities (akinesia, dyskinesia) or aneurysm as well as reduced RV fractional area change (Marcus et al., 2010). In humans the RV function is commonly assessed by measurement of tricuspid annular plane systolic excursion (TAPSE) (Dini et al., 2007; Gupta et al., 2008; Haddad et al., 2008; Mercer-Rosa et al., 2013; Saguner et al., 2014). Reduced TAPSE values have been associated with a worse clinical outcome in people with ARVC (Saguner et al., 2014). Multiple studies have assessed TAPSE in

19 healthy dogs, and breed specific reference intervals have been provided (Pariaut et al., 2012; Visser et al., 2015). TAPSE measurement is relatively easy to acquire and the results are reproducible. cMRI allows for the detection of, and quantification of, fibro-fatty replacement in the myocardium as well as quantification of the accurate and reproducible measurement of chamber dimensions, volume and function. Also, in comparison to echocardiography, cMRI allows the better detection of regional RV myocardial function (Oomen et al., 2018). The use of cMRI is also reported in a canine ARVC model, although due to the technical difficulties and availability of suitable equipment, this modality is not commonplace in veterinary medicine (Basso et al., 2004; Baumwart et al., 2009).

Histopathological changes in ARVC have been typically described as affecting only the RV, although left ventricular (LV) involvement is also recognised in people and in dogs (Norman et al., 2005; Basso et al., 2009; Sen-Chowdhry et al., 2010). Atrial myocardial involvement has also been shown in dogs with ARVC (Vila et al., 2017).

The family history of ARVC in a close relative is considered a major criterion for diagnosis in people (Marcus et al., 2010). An autosomal dominant mode of inheritance with incomplete penetrance has been proposed for ARVC in dogs (Meurs et al., 1999). Over 20 gene defects leading to ARVC development have been identified in humans, but only one mutation has been identified in the dog (Meurset al., 2010; Cattanach et al., 2015; Ohno, 2016; Meurs, 2017; Oomen et al., 2018).

2.1.2.2. Cardiomyopathies and other cardiac pathologies in Dogue de Bordeaux

Canine cardiomyopathies are mainly reported in large breed dogs (Martin et al., 2009). The DdB is a large brachycephalic breed with a high reported prevalence of various cardiac conditions including congenital defects, such as subaortic stenosis and tricuspid valve dysplasia as well as acquired DCM (Borgarelli et al., 2006; Höllmer et al., 2008; Martin et al., 2009; Oliveira et al., 2011; Ohad et al., 2013). Locatelli et al. (2011) have suggested that a breed predisposition to supraventricular tachycardias (SVT) exists in this breed, and Ohad et al. (2013) reported that all DdBs diagnosed with tricuspid valve dysplasia had AF at the time of

20 diagnosis. A recent questionnaire-based study revealed that almost 5% of DdBs had been diagnosed with a cardiac disease as diagnosed by a veterinarian, and 11% of dogs were classified as having a possible or probable cardiac disease (McAulay et al., 2018). Furthermore, the same study reported a high rate of sudden death (31% of all reported deaths) and of the sudden deaths, 35% were attributed to a cardiac cause. The DdB breed has been anecdotally reported as having a high prevalence of AF and other SVT. The presence of AF in the absence of detectable structural cardiac changes is described as ‘lone’ AF (Menaut et al., 2005). Any prolonged or sustained tachyarrhythmia can cause the development of tachycardia-induced cardiomyopathy (Zupan et al., 1996; Shinbane et al., 1997; Ellis and Josephson, 2013; Gopinathannair et al., 2015). It is often difficult, if not impossible to distinguish, whether the systolic dysfunction is a result of the tachycardia or a primary cardiomyopathy has led to development of tachyarrhythmias (Simantirakis et al., 2012). It is plausible that cardiomyopathies affecting the atrial myocardium could lead to the development of supraventricular arrhythmias (Vila et al., 2017).

2.2. Methods used for diagnostics of cardiomyopathies

2.2.1. Clinical features

The spectrum of clinical signs of cardiomyopathies depends on the type and severity of the disease as well as the haemodynamic consequences. The patients may be completely asymptomatic, with the changes detected only on tests performed as a part of screening (e.g. systolic dysfunction or frequent arrhythmias). They may present with clinical signs suggestive of persistent or intermittent reduction of cardiac output (e.g. exercise intolerance, intermittent weakness, syncope), signs of left- or right-sided congestion and fluid retention (e.g. pulmonary oedema, ascites) or, in extreme cases, with sudden cardiac death (Harpster, 1983). Significant variation of the prevalence and severity of clinical signs has also been documented to depend on the breed (Freeman et al., 1996; Tidholm and Jönsson, 1996; Calvert et al., 1997; Brownlie and Cobb, 1999; Dambach et al., 1999; Meurs et al., 2001a; Harmon et al., 2017).

The majority of cardiomyopathies are diagnosed in middle-aged or older dogs with the exceptions of the Portuguese Water Dog (Dambach et al., 1999) and Toy Manchester Terriers (Legge et al., 2013). An inheritable 21 arrhythmic disease leading to sudden cardiac death has been reported in young German Shepherd dogs (Moise et al., 1994). However, these dogs did not have any detectable structural cardiac changes (Ribas and Pariaut, 2018) and only very minimal changes in 12-lead electrocardiography (ECG) (Mérot et al., 2000; Protas et al., 2005).

The clinical characteristics of ARVC in Boxer dogs are similar to those of human ARVC. Typically, three forms have been described – the concealed (asymptomatic), characterised by ventricular arrhythmias detectable on electrocardiographic examination; the overt, characterised by tachyarrhythmias and clinical signs of syncope or episodic weakness; and the myocardial systolic dysfunction (DCM-phenotype) (Harpster, 1983; Harpster, 1991; Meurs, 2004). The last form is diagnosed least frequently, and these dogs may present with clinical signs consistent with congestive heart failure (CHF) (Meurs, 2017). Whether these three forms represent a continuum of a single disease, or different cardiomyopathies, remains under debate. The physical examination in the asymptomatic ARVC may be completely normal, or arrhythmias may be detected on auscultation and pulse deficits noted on simultaneous palpation of peripheral pulses (Hariu and Carpenter, 2010). With myocardial dysfunction form, a soft-to-moderate left basilar or left apical systolic murmur or a gallop sound may be auscultated (Baumwart et al., 2005).

The DdB is a breed predisposed to both congenital and acquired cardiac diseases (Borgarelli et al., 2006; Höllmer et al., 2008; Martin et al., 2009; Oliveira et al., 2011; Ohad et al., 2013). Additionally, a breed predisposition to SVT has been suspected (Locatelli et al., 2011) and Ohad et al. (2013) documented AF in all DdBs diagnosed with tricuspid valve dysplasia. Whether such predisposition to supraventricular arrhythmias is caused by an underlying cardiomyopathy or results from the dilation of the cardiac chambers caused by structural heart disease warrants further investigation.

2.2.2. Electrodiagnostics

2.2.2.1. 12-lead ECG

Supraventricular and ventricular arrhythmias are a common feature in cardiomyopathies, especially ARVC, and therefore ECG is an essential part of the diagnostic evaluation. The characteristic ECG finding in 22 ARVC is the presence of VPC with left bundle branch morphology with the positive deflection of the QRS complex in leads II, III, and aVF (Harpster, 1991; Kraus et al., 2002). In human cardiology, various depolarisation, conduction and repolarisation abnormalities, detected by 12-lead ECG, are used as major and minor criteria for ARVC diagnosis (Marcus et al., 2010). These criteria have not been validated for the use in Boxers or other canine cardiomyopathies.

The main limitation for the use of a brief in-house diagnostic method for ARVC, is its insensitivity for the detection of intermittent arrhythmias. Furthermore, a large day-to-day variability (> 80%) (Spier and Meurs, 2004a) in the number of VPC has been documented in Boxer dogs, limiting the usefulness of this tool for screening purposes. However, in-hospital ECG is highly specific for the detection of > 50 VPC/24h (Meurs et al., 2001b).

2.2.2.2. Ambulatory ECG (Holter monitoring)

A 24-hour ambulatory ECG (Holter monitor) allows significantly better detection of intermittent arrhythmias (Meurs et al., 2001b) and is therefore the main diagnostic tool for screening and diagnosis of ARVC in dogs (Hariu and Carpenter, 2010; Meurs, 2017). The large day-to-day variability of the number of VPC in affected Boxers should be considered when interpreting results (Spier and Meurs, 2004a). Although healthy mature dogs can have a small number of VPC (Meurs et al., 2001c), detection of a large number (> 100 VPC/24h) or morphologically complex ventricular arrhythmias can be considered highly suggestive of ARVC (Meurs, 2017). Various cut-off values have been proposed to differentiate affected and unaffected individuals (Meurs et al., 1999; Meurs, 2004; O’Grady and O’Sullivan, 2004; Baumwart and Meurs, 2005; Smith et al., 2008; Scansen et al., 2009; Hariu and Carpenter, 2010; Palermo et al., 2011). In human medicine, the presence of non-sustained or sustained VT with left bundle branch morphology with superior axis is considered a major criterion and the presence of sustained or non- sustained VT with left bundle branch morphology with inferior axis or the presence of > 500 VPC/24h is considered a minor criterion for the diagnosis of ARVC (Oomen et al., 2018).

23 2.2.2.3. Event monitoring

Due to the intermittent nature of arrhythmias and clinical signs, the 24-hour ambulatory ECG may prove insufficient for detection of the underlying arrhythmic mechanism. The usefulness of implantable loop recorders (Bright and Cali, 2000; MacKie et al., 2010), adhesive patch monitors (Lichtenberger et al., 2018) and smartphone-based ECG monitors (Vezzosi et al., 2019) have been described in the veterinary literature.

2.2.3. Diagnostic imaging

2.2.3.1. Radiography

The diagnostic yield of thoracic radiography for canine cardiomyopathies is variable. DCM is a progressive disease and the early signs of cardiomegaly can be subtle. In more advanced stages, the atrial and ventricular dilation becomes more obvious; in some cases, the right atrial or ventricular enlargement can also be present. The main indication for thoracic radiography is assessment of the presence of left-sided congestive signs or presence of pleural effusion (McEwan, 2000; Meurs, 2017).

2.2.3.2. Echocardiography

Echocardiography allows for the visualisation and measurement of the cardiac chamber dimensions, estimation of global and regional myocardial function and the measurement of blood flow velocity through the cardiac structures, making it an essential tool for the screening, diagnosis and monitoring of the progression of cardiomyopathies. The European Society of Veterinary Cardiology has proposed guidelines for the diagnosis of canine DCM (Dukes-McEwan et al., 2003), and more recently the screening guidelines for Doberman Pinchers to aid with the establishment of early diagnosis of DCM in this breed (Wess et al., 2017). More breed-specific information to distinguish normal and affected individuals in different breeds have been recently published (Stephenson et al., 2012; Dickson et al., 2016; Brungs et al., 2018).

24 Echocardiography has an invaluable role in the diagnosis of AVRC with ventricular systolic dysfunction (Baumwart et al., 2005). Although the RV size and function are more difficult to quantify by echocardiography, careful evaluation may also detect right atrial and ventricular enlargement or RV dysfunction (Meurs, 2017). More recently Visser et al. (2017) have assessed the feasibility of the echocardiographic assessment of the RV function and published reference intervals for healthy dogs. In human ARVC diagnostics, the presence of echocardiographic evidence of regional RV wall motion abnormalities (akinesia, dyskinesia, aneurysm) together with other RV functional indices are considered major or minor criteria (Marcus et al., 2010). Cunningham et al. (2018) showed that two echocardiographic indicators of RV function, the TAPSE and pulsed wave tissue Doppler imaging-derived systolic myocardial velocity of the lateral tricuspid annulus (S’) are reduced in Boxer dogs with ARVC compared to healthy controls (Smith et al., 2007). Advanced echocardiographic techniques, such as tissue doppler and strain echocardiography, have been used in the assessment of cardiomyopathies in both people and dogs, but these methods require further research for sufficient validation (Chetboul et al., 2007; Prakasa et al., 2007).

2.2.3.3. cMRI cMRI technology has improved rapidly over the last decades and has become a tool commonly used in human cardiology. The benefit of cMRI is the ability of image acquisition in virtually any plane as well as superior soft tissue differentiation. The main limitations of this technology are the cost of the equipment, the relatively long time required for image acquisition and, for the assessment of cardiac structures, respiratory and cardiac motion artefacts reducing the image quality (Labruyère and Schwarz, 2013). cMRI provides a more accurate and reproducible measurement of cardiac chamber dimensions, volumes and function compared to echocardiography, making it an excellent tool for the assessment of regional RV wall akinesia or dyskinesia, volumetric measurements of the RV and the assessment of RV ejection fraction (Sugeng et al., 2010; Oomen et al., 2018). cMRI measurements of RV wall function and RV volumes have been included as major and minor criteria for the diagnosis of human ARVC (Marcus et al., 2010). Baumwart et al. (2009) have described the cMRI findings of the RV morphology and function in Boxer dogs with ARVC.

25 Another area of interest in human cardiology research where cMRI is used, is the assessment of the volume of valvular regurgitation (Thavendiranathan et al., 2012). Although cMRI equipment has become more readily available in veterinary medicine, its use has been limited mainly to research settings (Gilbert et al., 2010; Labruyère and Schwarz, 2013). One of the main limitations for the practical use of cMRI in cardiac patients is the requirement for prolonged general anaesthesia. MMVD is the most common acquired cardiac disease in dogs and the third most common cause of death in this species (Häggström et al., 2004). Despite the limitations cMRI, as an accurate diagnostic tool, could improve the diagnosis and prognostication of MMVD as well as canine cardiomyopathies (Gilbert et al., 2010).

2.2.4. Genetics

Cardiomyopathies have historically been defined by their phenotypic expression (hypertrophic, dilated, arrhythmogenic, restrictive) (Elliott et al., 2008). The first cardiac disease-causing gene mutation in people was identified in 1990 (Geisterfer-Lowrance et al., 1990). Since then a large number of gene mutations have been detected in association with various forms of human cardiomyopathies (Jacoby and McKenna, 2012; Maron et al., 2014; Oomen et al., 2018). The large number of associated gene defects, as well as the fact that the majority of such mutations appear to be limited to one single family, has limited the usefulness of mutational analysis in predicting prognosis or phenotypic expression of the disease (Maron et al., 2014). To date, only a limited number of mutations have been detected in canine cardiomyopathies. This may partially be explained by the slower progression of the research in this field in veterinary medicine, but is more likely to be associated with the significant genetic homogeneity in veterinary species within breeds (Stern and Ueda, 2019). As many inherited cardiomyopathies in humans are similar to canine equivalent diseases, the research into cardiomyopathies in veterinary species is likely to serve as a useful model for human research (Basso et al., 2004; Mausberg et al., 2011; Fox et al., 2014; Simpson et al., 2015).

DCM is highly prevalent in Doberman Pinchers, Great Danes, Irish Wolfhounds, Newfoundland Dogs and many other large breed dogs. The prevalence of DCM in Dobermans in Europe is 58% and the autosomal dominant trait of inheritance with variable penetrance is

26 similar to the majority of the human DCM forms (Meurs et al., 2007; Wess et al., 2010). In Dobermans, two mutations have been associated with increased risk of development of the DCM – one affecting the pyruvate dehydrogenase kinase 4 gene (Meurs et al., 2012) and another single nucleotide polymorphism affecting chromosome 5 (Mausberg et al., 2011). In Dobermans DCM appears to be inherited in autosomal dominant pattern with incomplete penetrance (Meurs et al., 2012). In Great Dane dogs, a breed with the second most prevalent diagnosis with DCM with reported prevalence varying from 11.8% (Tarducci et al., 2003) to 35.6% (Stephenson et al., 2012), the genetic mutations leading to development of DCM have not yet been identified. However, the heritability has been reported to be either x-linked (Meurs et al., 2001a) or autosomal dominant (Stephenson et al., 2012). In Irish Wolfhounds, the prevalence of DCM has been reported to vary between 24.2% (Vollmar, 2000) and 25.5% (33.5% in male dogs) (Distl et al., 2007). The mode of inheritance in this breed is not completely understood, but an oligogenic inheritance is suspected (Philipp et al., 2007; Simpson et al., 2016). In Newfoundland dogs, the estimated prevalence of DCM is 1.3–17.6% (Dukes-McEwan et al., 2003) and an autosomal dominant trait has been suggested. In Portuguese Water Dogs, the DCM presents as a juvenile form, causing mortality in dogs between 2 and 32 weeks of age; the causal mutation has not been identified (Werner et al., 2008). Although Boxer dogs have an adult-onset cardiomyopathy that can present with the typical features of DCM, it would be more appropriate to use the term ‘Boxer cardiomyopathy’ as a majority of the cases present with the features typical of human ARVC (Dukes-McEwan et al., 2003). It appears to be inherited in a autosomal dominant pattern with age- dependent penetrance (Meurs et al., 1999). A genome-wide association study has revealed a 8 base-pair deletion that is significantly associated with ARVC (Meurs et al., 2010). However, the same mutation did not explain the occurrence of ARVC in a United Kingdom study (Cattanach et al., 2015).

2.3. Prognostic factors

Cardiomyopathies represent a diverse group of diseases affecting the myocardium. Although the classification is mainly based on the predominant pathophysiology or phenotypic expression (e.g. DCM, HCM, RCM, ARVC), it is both generally accepted and genetically and morphologically demonstrated that a similar phenotype can be caused

27 by significantly different pathological mechanisms. Furthermore, gene- environment interactions may also alter the phenotypic expression of the particular disease (Pinto et al., 2016). Such large diversity and multifactorial pathogenesis also has a significant prognostic impact and very few universal prognostic factors have been described to cover all aspects of either human or canine cardiomyopathies (Hazebroek et al., 2015).

2.3.1. Dilated cardiomyopathy

DCM is characterised by a long progression with a high risk of fatal arrhythmias (Stolfo et al., 2018). The condition is best regarded not as a single disease entity, but rather as a nonspecific phenotype resulting from a variety of genetic and environmental factors (Japp et al., 2016). The clinical features and prognosis of DCM have been extensively studied and described in the literature. Early reports were mainly based on clinical, pathological, electrocardiographic and echocardiographic data, and also compared the effect of treatment on the long term outcome (Alvarez et al., 2007; Kubo et al., 2008). The severity of systolic dysfunction, as well as the presence of CHF appear to be more universally associated with a worse clinical outcome (Japp et al., 2016; Halliday et al., 2017; Miles et al., 2019; Schultheiss et al., 2019). However, the clinical and echocardiographic parameters have been shown to be not very accurate in predicting the long term outcome (Stolfo et al., 2018; Merlo et al., 2018; Schultheiss et al., 2019). Myocardial fibrosis has been associated with increased risk of arrhythmogenesis and sudden cardiac death (Assomull et al., 2006). Advanced imaging modalities, such as cMRI, are promising for the acquisition of diagnostic information as well as risk stratification (Japp et al., 2016; Becker et al., 2018).

Early studies into canine DCM reported poor overall prognosis and were not able to identify universal prognostic indicators (Tidholm et al., 2001). The presence of heart failure, ascites, reduced ejection fraction and increased end-systolic LV volume as well as restrictive trans-mitral flow pattern have been considered useful prognostic indicators (Borgarelli et al., 2006). The presence of sustained ventricular tachycardia on 24-h ambulatory ECG has also been shown to be a negative prognostic indicator in occult DCM for Doberman Pinchers (Calvert et al., 1997). The overt phase of DCM is consistent with an advanced stage and presence of CHF and is generally associated with a poor prognosis,

28 whilst the prognosis of dogs with pre-clinical DCM is better (Monnet et al., 1995; Tidholm and Jönsson, 1997; Harmon et al., 2017). Early detection of occult DCM can improve survival by removing the affected dogs from the breeding programmes (Wess et al., 2017).

2.3.2. Arrhythmogenic right ventricular cardiomyopathy

ARVC is a disease causing fibrofatty replacement of the myocardium that may lead to ventricular dysfunction and ventricular arrhythmias and is associated with an increased risk of SCD (Marcus et al., 2010; Calkins et al., 2017). Although SCD is a relatively rare event in ARVC patients, Naneix et al. (2015) showed that ARVC was the predominant cause of cardiomyopathy related SCD. Multiple genetic defects have been identified, but the disease expression is highly variable, even in the families with the same genetic defect, which is likely due to the presence of modifier genes and environmental influences (Sen-Chowdhry et al., 2007). Electrical abnormalities often precede the structural myocardial changes, increasing the risk of development of dangerous clinical signs such as syncope or SCD. The prevention of SCD involves both antiarrhythmic medication and implantation of cardioverter devices. However, implantable cardioverter devices can have severe side effects such as inappropriate shocks or lead failures, risk stratification for patients with ARVC is of paramount importance (Calkins et al., 2017; Maupain et al., 2018; Bosman et al., 2018). Most of the information on prognostic factors is based on single-centre reports or small multicentre registries (Lemola et al., 2005; Calkins et al., 2017). Recently multiple excellent review articles on diagnostic and prognostic criteria and risk stratification have been published (Bosman et al., 2018; Cadrin- Tourigny et al., 2019). Most of the reports identify male sex, clinical signs of syncope, RV dysfunction and the presence of sustained or non-sustained VT or ventricular fibrillation as consistently predictive of worse clinical outcomes (Bosman et al., 2018). Other negative prognostic indicators include reduced LV systolic function, higher physical activity, extent of structural disease and electrical instability (frequency of VPC an CT) (Brun et al., 2016; Calkins et al., 2017; Maupain et al., 2018). The prognosis appears to be better in asymptomatic patients and patients without sustained or non-sustained VT (Brun et al., 2016; Calkins et al., 2017; Maupain et al., 2018). In humans, ARVC has an age-dependent presentation with approximately 1/5 of patients presenting after the age of 50 years (Bhonsale et al., 2017). In the late presentation cohort:

29 male sex, RV structural disease, lack of family history and inducibility by electrophysiologic study were associated with increased risk (Bhonsale et al., 2017). An excellent consensus statement with recommendations on the diagnosis and management of human ARVC was recently published (Towbin et al., 2019).

Human and canine ARVC have significant similarities in clinical presentation, pathological and diagnostic findings as well as long term prognosis (Basso et al., 2004). Since its first description in 1983 (Harpster, 1983), extensive research has been conducted on it. However, due to the absence of uniform diagnostic criteria, the results of these studies are difficult to compare (Vischer et al., 2017). The LV involvement, and presence of CHF, appear to be associated with a shorter survival in the majority of the studies on Boxer ARVC (Hariu and Carpenter, 2010; Palermo et al., 2011; Meurs et al., 2014). Similar findings were described in English Bulldogs with presumed ARVC (Cunningham et al., 2018). The presence of syncopal events, although a factor relevant to the diagnosis of ARVC in Boxers, remains controversial as a prognostic indicator (Palermo et al., 2011; Caro-Vadillo et al., 2013; Chamas et al., 2016). More recently, echocardiographic indices of RV function, (Cunningham et al., 2018) and various ambulatory ECG variables, have been assessed for their prognostic value (Palermo et al., 2011; Chamas et al., 2016; Carvalho et al., 2019). In these studies indicators of TAPSE, as well as the presence of VT and higher complexity of the ventricular arrhythmias, appear to be associated with symptomatic status and indicate negative survival. Heart rate variability indices, appear to be poor prognostic indicators for mortality (Chamas et al., 2016). However, it is important to note that most of these studies were often limited to small populations and have used different diagnostic criteria, hindering reliable comparison of their results (Vischer et al., 2017).

30 3. AIMS OF THE STUDY

The overall aim of this thesis was to contribute to the diagnostic and prognostic yield of electrodiagnostic and diagnostic imaging modalities on canine cardiac disease.

The more specific aims of this study were as follows:

• To assess the prognostic value of Holter monitoring in Boxer dogs with suspected ARVC (I). • To evaluate the clinical utility of TAPSE in Boxer dogs with syncope or suspected ARVC (II). • To describe and analyse the echocardiographic and electrocardiographic findings in a population of DdBs presented for second opinion cardiology referral (III). • To evaluate the feasibility of cMRI for the quantification of mitral regurgitation (MR) in small breed dogs and to identify the echocardiographic variables that best correlate with the severity of MR as determined by cMRI (IV).

31 4. MATERIALS AND METHODS

4.1. Study population

4.1.1. Inclusion and exclusion criteria

The medical records of the Queen Mother Hospital for Animals, The Royal Veterinary College (I–III), the Small Animal Teaching Hospital, University of Liverpool (I, III) and Hospital for Small Animals, Royal (Dick) School Veterinary Studies (I) were retrospectively reviewed. The main inclusion criteria were Boxer dogs undergoing Holter monitoring and echocardiographic examination (I, II) or DdBs undergoing Holter monitoring (III). The electrocardiographic data for the DdB study (III) were collected from original ECG, Holter monitoring, echocardiographic timing ECG or attending clinicians’ records. The Holter monitoring for studies I and II must have contained at least 19 hours of valid data and an echocardiographic examination within one month of Holter available for review. All dogs were referred to a cardiology service for investigation of clinical signs, or physical examination findings suggestive of a cardiac disease (i.e. not for the screening for a breed club programme) (I–III).

For the cMRI study (IV) patients with body weight ≤ 20 kg presenting for magnetic resonance imaging for neurological disease that had concurrent MMVD were prospectively recruited after acquisition of the informed consent from the owner and a full physical and neurological examination. The basis for the diagnosis of MMVD was presence of a systolic heart murmur and characteristic findings from two-dimensional (2D) or colour-flow Doppler echocardiography.

Dogs with documented congenital heart disease, including those with aortic blood flow velocities ≥ 2.25 m/s, or if they had a concurrent diagnosis of a potentially life limiting disease (I, II) or insufficient echocardiographic or electrocardiographic/Holter data (I–III) were excluded from the study. Dogs with a high risk of CHF were excluded from the cMRI study (IV).

32 4.1.2. Baseline clinical data (I–III)

Baseline clinical data included signalment (breed, age, sex, body weight), history of episodic weakness or collapse that was believed to be consistent with syncope, the presence of CHF before presentation, and the presence of concurrent diseases. CHF was considered to be present based on the radiographic evidence of cardiogenic pulmonary oedema, or if the dog had clinical evidence of ascites with either concurrent jugular distension/pulsation or hepatic congestion, and a clinical response to diuretic treatment.

4.1.3. Survival information (I, II)

Information on survival was obtained from medical records. If insufficient information was available, a questionnaire was mailed to the owners. The owners that did not respond to the questionnaire were contacted by phone, and if the survival information was still incomplete, the referring veterinarians were contacted. The outcome data included patient status (dead/alive), the date of death or euthanasia, and whether the death was caused by deterioration of the cardiac condition (uncontrolled CHF, worsening of CHF, unacceptable increase in the frequency of syncope). Sudden cardiac death was defined as an unexpected death without apparent clinical signs during the preceding 24-hour period.

4.2. Echocardiography data (I–IV)

Echocardiography was performed by a board-certified cardiologist or a cardiology resident under the direct supervision of a board-certified cardiologist in unsedated dogs (I–IV); echocardiographic examination after cMRI data acquisition was performed when the patient was still under the influence of general anaesthesiaIV ( ). Echocardiography was performed with appropriate echocardiographic equipment and softwareabcde with simultaneous ECG recording.

a Philips Sonos 5500; Philips Healthcare, Surrey, UK b GE Vivid 5, GE Healthcare, Hatfield, UK c GE Vivid 7, GE Healthcare, Hatfield, UK d GE Vivid I, GE Healthcare, Hatfield, UK e GE EchoPAC, GE Healthcare, Hatfield, UK 33 Standard echocardiographic examination was performed as described by Thomas et al. (1993). For calculation of the left atrial to aortic ratio (LA/ Ao) diameters of the left atrium (LA) and aorta were measured using 2D echocardiography in right parasternal short axis view of the heart base from the first frame after aortic valve closure (Hanssonet al., 2002).

Motion mode ultrasound imaging (M-mode) echocardiography using the leading-edge-to-leading-edge method was used to measure LV internal diameters in systole (LVIDs) and diastole (LVIDd) and to calculate fractional shortening (FS%) (Sahn et al., 1978; Thomas et al., 1993). For the Boxer studies systolic dysfunction was defined as LVIDs > 35mm (Palermo et al., 2011) (I, II) or FS% < 25%, or ejection fraction (EF%) < 50% measured using a modified Simpson’s method II( ) (Boon, 2011). For the DdB study (III), the LV systolic and diastolic measurements were normalised to body weight. The normalised left ventricular systolic diameter measurements exceeding the 95% confidence interval (CI) of the predicted values were defined as a marker of systolic dysfunction, and the normalised left ventricular diastolic diameter (LVDdn) measurement exceeding the 95% CI of the predicted values were defined as a marker of LV dilation (LVDdn > 1.85) (Cornell et al., 2004).

Spectral Doppler echocardiography using a subcostal window was used to measure aortic flow velocity I( –III) (Lehmkuhl and Bonagura, 1994; Bussadori et al., 2000; Höllmer et al., 2008) and pre-ejection period to ejection time ratio (de Madron, 2015).

Right-sided dilation was defined as RV diastolic dimensions exceeding 50% of the diameter of the LV or right atrial dimensions that were subjectively larger than the LA (Menaut et al., 2005; Palermo et al., 2011). In the presence of LV dilation, the assessment of the RV size was based on the clinical experience of the echocardiographer. The presence of systolic pulmonary hypertension (III) was defined as tricuspid regurgitation velocity > 3.1 m/s (Stepien, 2009).

TAPSE measurement (II) was performed by a single observer using 2D anatomic M-mode on a left apical 4-chamber image (Pariaut et al., 2012). Intra-observer variability of TAPSE measurement was based on 5 repeated measurements on randomly selected images from 10 dogs.

34 For the cMRI study (IV), a combination of M-mode, 2D, colour flow Doppler and spectral Doppler images were used to record LA diameter, LA area, MR area, aortic diameter, amplitude of mitral valve prolapse,

LVIDd, peak early transmitral filling velocity (Evel), vena contracta (VC) width of the MR jet, proximal isovelocity surface area of the MR jet (PISA), MR velocity and aortic velocity. LA area and MR area were used to calculate the jet area ratio (JAR). All measurements were obtained retrospectively from pre- and post cMRI studies. To check inter- observer repeatability and agreement, a second observer repeated the echocardiographic measurements in all dogs.

4.3. Electrocardiographic data (I–III)

In the Boxer studies (I, II) only Holter data were used for analysis and, if multiple recordings were available from the same dog, only the first recording was used. Two centres used the same Holter analysis servicef and in the third centre the raw Holter data was reviewedg. For study I, several variables were recorded to be used for the univariable analysis of the prognostic value. For study II, the dogs were dichotomized into two groups based on Holter analysis: ≥ 50 VPC/24h and < 50VPC/24h. VT was recorded as present or absent.

For the DdB study (III), the electrocardiographic data were collected from the 12-lead ECG, Holter recording or echocardiographic timing ECG. AF, SVT, ventricular arrhythmias were documented. A single VPC was listed as ‘isolated’. Multiple VPC (couplets, triplets, bigeminy, trigeminy, VT) were listed as ‘complex’. Dogs, in which only isolated VPC occurred, were further dichotomised using a cut-off value of 100 VPC/24h.

4.4. Cardiac magnetic resonance imaging data (IV) cMRI examinations were performed using a 1.5 T Intera Pulsar MRIh using a flexible, sensitivity-encoding, phased array surface coili. Animals were under general anaesthesia throughout the procedure. Dogs were premedicated using diazepam and either alphaxalone or propofol and f Laboratory Corporation of America, Ambulatory Monitoring Services, Burlington, NC g Novacor Holter Soft, Novacor UK Ltd. Kent, UK h Intera Pulsar System, Philips Medical Systems, Castlefield Road, Reigate, Surrey. i SENSE Flex M or SENSE Body coil 35 maintained with sevoflurane. Intermittent positive pressure ventilation was maintained throughout the procedure using a mechanical ventilator. Breath-holding to avoid respiratory motion during periods of image acquisition was achieved through induction of transient apnoea following a brief period of hyperventilation. The maximum additional anaesthesia time allowed for cMRI was limited to 50 minutes, after which the dogs were recovered even if the cMRI acquisition was incomplete.

Survey gradient echo imaging was performed to acquire transverse, sagittal and dorsal standard localising scans from which two- and four-chamber LV long axis cine images were prescribed. For image acquisition an cMRI compatible vectorcardiogram gating device was used for retrospective gating. A series of short axis images of the LV was also obtained using two- and four-chamber LV long axis sequences.

T1 weighted Turbo field echo, a series of images of the LV outflow tract, were obtained to identify a site in the proximal portion of the aorta and perpendicular to aortic flow. Phase contrast angiography imaging at this site was used to obtain velocity-encoded maps enabling quantitative measurements of aortic blood flow. cMRI scans were analysed offline independently by a single observer using a dedicated software packagej. The LV volumes were calculated by a modified Simpson’s rule using a semi-automated contour tracing software programme as previously described (Alfakih et al., 2003). The total LV stroke volume was calculated by subtracting the minimum LV volume from the maximum LV volume. MR volume was calculated by subtracting aortic stroke volume from LV stroke volume. MR fraction (cMRI-RF) was defined as the MR volume as a proportion of the LV stroke volume. Five complete cMRI studies were measured independently by two cardiology residents.

4.5. Ethical considerations

All studies were approved by the Ethics and Welfare Committees of each participating institution.

j Extended MR Workspace Software Package ver. R.6.2.1. Philips Medical Systems, Castlefield Road, Reigate, Surrey 36 4.6. Statistical analysis

Data were analysed using SPSS Statistics version 18.0 (I), version 21 (II), version 20 (IV) and GraphPad Prism version 5.0 (I), version 6.0 (II–IV). The distributions of data for patient population characteristics, echocardiographic and electrocardiographic variables and image acquisition time was assessed by visual assessment of the data distribution graphs as well as using the Shapiro-Wilk test (I, II, IV) and the D’Agostino and Pearson omnibus normality test (III). The average values of the normally distributed data were expressed as mean ± standard deviation. Non-normally distributed data were expressed as median and interquartile range (IQR) (I, III, IV) or median and range (II). Binomial and categorical data were expressed as numbers and percentages (I).

The Fisher’s exact test was used to determine the differences between the study groups for the categorical data (I, III) and one-way analysis of variance for the normally distributed continuous data (III). In studies I and II, the Mann-Whitney U-test was used to assess the association between the study groups and the non-normally distributed variables. For study III the Kruskal-Wallis and Dunn’s multiple comparison tests were used for the assessment of the associations between groups and non-normally distributed variables.

Survival analysis (I, II) was performed using an endpoint of death or euthanasia related to heart disease (cardiac mortality). Dogs alive at the end of the study period or died of noncardiac disease or lost to follow-up were right-censored. Kaplan-Meier survival estimates and log-rank tests were used to identify the variables that significantly influenced cardiac mortality. Cox proportional hazards analysis was used to construct univariable time-to-event models and the variables for which the association between the variable and mortality resulted in a P < 0.2 were included in the multivariable analysis. Multivariable Cox proportional hazards analysis was performed using a manual forward stepwise approach. The assumptions of Cox proportional hazards model were tested using log cumulative hazard plots and distribution of Schoenfeld residuals was used to assess the proportional hazards assumption, the overall model fit was evaluated according to the distribution of Cox- Snell residuals.

37 In the cMRI study (IV), the inter-observer agreement was assessed using the intraclass correlation coefficient (ICC) and two-way mixed single measures model, with the results expressed as ICC (95% CI). The lower and upper thresholds of 95% CI were truncated at zero and at one if the estimated CI exceeded these limits. The inter-observer agreement for cMRI-derived volumes was assessed graphically using Bland-Altman plots. Differences between the measurements from the two echocardiographic examinations were assessed using a paired Student’s t-test for the parametric data and the Wilcoxon matched-pairs signed rank test for the non-parametric data. Correlations between the cMRI-RF and selected echocardiographic variables were tested using Pearson’s correlation coefficient (r).

The significance was set at the 5% level for all tests.

38 5. RESULTS

5.1. Study population (I–IV)

Studies I and II were based on Boxer dogs only. For the Holter study (I), a total of 163 dogs were identified from 3 participating centres. After exclusion of the dogs with elevated aortic velocity and insufficient Holter recording (Fig. 1a), 122 dogs were included in the analysis. The sex distribution in the whole population was almost equal (52% male), but in the systolic dysfunction/CHF group (LVIDs>35mm) there was a significantly larger proportion (80% vs 42%, P < 0.001) of male dogs than in the group with echocardiographically normal systolic function

(LVIDs≤35mm). The median age at presentation for the whole population was 6.9 years (IQR 4.1–9.0), and there was no significant difference in age between the two groups (P = 0.54). However, a bimodal age distribution with two peaks at approximately 1 year and 8 years of age was noted (Fig. 2). The young dogs were more likely to be presented for investigation of syncope (P = 0.040).

For study II 81 Boxer dogs were identified (Fig. 1b), but 31 had to be excluded from the analysis due to the inadequate echocardiographic images available, leaving 50 dogs for analysis. No sex difference was seen in the study population with 50% of the dogs being male. The median age was 6.0 years (0.5–12.5). Syncope was the primary presenting sign in 78% (39/50) and CHF was a historical or presenting sign in 14% (7/50) dogs.

39 D E

EƵŵďĞƌŽĨ ŝĚĞŶƚŝĨŝĞĚĚŽŐƐ ϭϲϯ EƵŵďĞƌŽĨ ŝĚĞŶƚŝĨŝĞĚ ϴϭ ǆůĐƵƐŝŽŶ͗ŽƌƚŝĐ ϰϬ ĚŽŐƐ ǀĞůŽĐŝƚǇхϮ͘ϮϱŵͬƐ ;ϭϮϯͿ ǆĐůĐƵƐŝŽŶ͗ ǆĐůƵƐŝŽŶ͗ ϭ ;ϭϮϮͿ /ŶĂĚĞƋƵĂƚĞ ŝŶƐƵĨĨŝĐŝĞŶƚ,ŽůƚĞƌ ĞĐŚŽ ϯϭ ŝŵĂŐĞƐ ^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶ zĞƐ͗Ϯϴ ;>s/ƐхϯϱŵŵͿ EŽ͗ϵϰ /ŶĐůƵĚĞĚŝŶ ƚŚĞ ϱϬ ĂŶĂůǇƐŝƐ WƌĞƐĞŶĐĞŽĨ,& EŽ͗ϵϭ zĞƐ͗ϯ

/ŶĐůƵĚĞĚŝŶƚŚĞ ĂŶĂůǇƐŝƐ ϵϭ ϯϭ

>s/Ɛ ͬ,& 'ZKhW >s/ƐPPQR&+) хϯϱŵŵ

Figure 1. Summary of the population included in the Boxer studies I (a) and II (b). The arrow-shaped boxes indicate the number of excluded dogs. The dogs with normal systolic function and presence of congestive heart failure (CHF) (pentagon arrow) were combined with the systolic dysfunction group for the analysis. LVIDs≤35mm/no

CHF – group without systolic dysfunction or CHF; LVIDs>35mm/CHF – group with systolic dysfunction or presence of CHF

Figure 2. The age distribution of the study population was suggestive of a bimodal age distribution with young dogs being more likely to be presented for the investigation of a) - syncope episodes (P = 0.040) and b) - ventricular tachycardia (VT) occurring more frequently in older dogs (P = 0.015)

Sixty-four DdB dogs were identified for studyIII (Fig. 3). The mean age at presentation was 4.3 years (± 2.1) with a male predominance (male/ female = 49/15). The predominant presenting complaint was detection of an arrhythmia on auscultation (51.6%) and exercise intolerance (41%). Other common complaints suggestive of a cardiac disease were systolic murmur (32.8%), collapse (31.3%) and abdominal distension (25%). An audible gallop sound was detected in 1 dog (1.6%).

41 EƵŵďĞƌŽĨŝĚĞŶƚŝĨŝĞĚĚŽŐƐ ϲϰ

WƌŝŽƌĚŝĂŐŶŽƐŝƐŽĨŶŽŶͲ ;ϱϴͿ ĐĂƌĚŝĂĐŶĞŽƉůĂƐŝĂ ϲ

ĂƌĚŝĂĐƐŽĨƚƚŝƐƐƵĞŵĂƐƐŽŶ ϳ ;ϱϭͿ ĞĐŚŽĐĂƌĚŝŽŐƌĂƉŚǇ

ŽŶŐĞŶŝƚĂůĐĂƌĚŝĂĐĚĞĨĞĐƚƐ ϭϳ ;ϯϰͿ

ϭϮ Ϯ W^ s^ ^ ^^ W

/ŶĐůƵĚĞĚŝŶƚŚĞĂŶĂůǇƐŝƐ͗ ŶŽĐŽŶŐĞŶŝƚĂůŚĞĂƌƚ ĚŝƐĞĂƐĞ͕ĐĂƌĚŝĂĐŵĂƐƐŽƌ ϯϰ ŶŽŶͲĐĂƌĚŝĂĐŶĞŽƉůĂƐŝĂ ;ŶŽŶͲ,EŐƌŽƵƉͿ Figure 3. Summary of non-cardiac neoplasia, cardiac soft tissue mass and congenital heart disease in the study III population. The numbers in ovals were excluded from the further assessment; only the data from dogs in the non-CHDN group were included in the analysis. SAS – subaortic stenosis, PDA – patent ductus arteriosus, PS – pulmonic stenosis, VSD – ventricular septal defect, ASD – atrial septal defect

The recruited population of the cMRI study (IV) can be considered typical for MMVD. There were 15 Cavalier King Charles Spaniels and one each of the following breeds: whippet, Border collie, beagle, Shih Tzu, Bichon Frise and toy poodle. The mean age was 7.67 (±2.73 years) and there was an almost equal sex distribution (11 males, 10 females). The systolic murmur intensity varied from grade I/VI to V/VI.

5.2. Electrocardiographic data (I–III)

The distribution of the collected Holter data (I) between the centres was not significantly different apart from the maximum P( < 0.001) and minimum (P < 0.001) one-minute heart rate. When comparing the Holter variables between the groups of dogs without LV systolic dysfunction or

CHF (LVIDs≤35 mm/no CHF) and the dogs with LV systolic dysfunction or CHF (LVIDs>35mm/CHF), the latter group was significantly more likely to have an increased number of ventricular [mean VPC/24h (P < 0.001), > 50 VPC/24h (P < 0.001), > 100 VPC/24h (P < 0.001), > 1000 VPC/24h (P = 0.016)] and supraventricular [number of atrial premature complexes in 24 hours (P = 0.004), number of SVT episodes in 24 hours (P = 0.011)] arrhythmias; increased complexity of ventricular

42 arrhythmias [presence of VT (P = 0.027), presence of couplets or triplets (P = 0.013)], as well as increased mean one minute heart rate (P = 0.026).

The dogs in the study II were divided into two groups based on the frequency of the VPC using the cut-off value of < 50 VPC/24h (‘non- ARVC’, n = 22) or ≥ 50 VPC/24h (‘ARVC’, n = 28). Of the dogs presenting for investigation of syncope, 56% (22/39) were classified as ARVC; six dogs in ARVC group did not have a history of syncope. The dogs in the ARVC group were significantly older (7.7 years, range 0.5–12.5) than those in the non-ARVC group (4.9 years, range 0.9–10, P = 0.046).

Of all DdBs in study III (n = 64), supraventricular arrhythmia was detected in 30 (46.9%) – persistent AF in 25 (39.1%) and SVT in 5 (7.8%). The dogs with supraventricular arrhythmias also had higher heart rates at presentation (P < 0.001). Holter data were available for 16/64 (25%) DdBs, of which 9/16 (56%) had AF.

Excluding the dogs with non-cardiac neoplasia, echocardiographically detected cardiac soft tissue masses, or congenital heart disease (non- CHDN group, n = 34), AF was present in 19 (56%) and 3 (9%) had SVT. The dogs with SVT were significantly younger than the other dogs (P = 0.027). The median heart rate of the non-CHDN dogs with AF (n = 19) at presentation was 200 beats per minute, with 17 (89%) having a heart rate > 160 beats per minute. Only one of the three dogs with no evidence of structural heart disease and AF had a heart rate of < 160 beats per minute. Four non-CHDN dogs with sinus rhythm had systolic dysfunction, with 2 of these dogs having > 1000 VPC/24h.

Ventricular arrhythmias were detected in 15/64 (23.4%) of all dogs; 8 dogs had isolated VPC and 7 complex VPC. Holter data were available for 13/15 (87%) dogs with ventricular arrhythmia, with a median of 105 (IQR 5–543) VPC/24h. In dogs with isolated VPC on Holter, 6/7 (86%) had less than 100 VPC/24h.

5.3. Echocardiographic data (I–IV)

Of the 122 dogs in the Holter study (I), 28 dogs had LV systolic dysfunction. Furthermore, three dogs with echocardiographically normal ventricular systolic dimensions (LVIDs ≤ 35mm) had CHF

43 at presentation, leaving 91 dogs in the LVIDs≤35mm group available for survival analysis (Fig. 1).

The dogs with LV dilation (LVIDs > 35mm) were significantly more likely to have LA dilation, reduced FS% and presence of CHF (P < 0.01 for each variable). There were strong associations between LV systolic dimension (LVIDs > 35mm) and reduced FS% (FS% < 25%); LA dilation (LA/Ao > 1.5) and the presence of CHF; LA dilation and increased number of VPC over 24 hours (> 50VPC/24h); and between reduced FS% and > 50VPC/24h (all pairwise P-values < 0.001) in these dogs. The presence of syncope was not significantly associated with any of these indices. The echocardiographic findings for the LVIDs≤35mm group are presented below under survival information.

TAPSE measurements (II) were recorded from the stored echocardiographic studies for 50 dogs; a minimum of 2 consecutive measurements were obtained in 45 (90%) of dogs and 3 consecutive measurements in 27 dogs (54%). Intra-observer variability of TAPSE measurements was 8%. Mean TAPSE for the entire population was 15.2 mm (±3.9 mm). A weak negative correlation was detected between TAPSE and both absolute weight (r = -0.324, P = 0.022) and the weight scaled to the 1/3 power (r = -0.295, P = 0.038). As this correlation was weak and unlikely to be clinically significant, for the analysis TAPSE was used as an absolute value.

TAPSE was significantly lower in the ARVC group (13.9 ± 4.04mm vs 16.8 ± 3.21 mm, P = 0.008). Weak to moderate correlations were found between TAPSE and the following variables: ejection fraction (r = 0.354, P = 0.017), FS% (r = 0.424, P = 0.002), and a negative correlation for LA/Ao (r = -0.448, P = 0.002). TAPSE was significantly lower in dogs with LV systolic dysfunction compared to that of the dogs with normal LV function (14.1 ± 3.78 mm vs 16.8 ± 3.73 mm, P = 0.02); TAPSE was also lower in dogs with VT recorded on Holter (12.9 ± 3.21 mm vs 16.1 ± 3.87 mm, P = 0.008).

Of the 34 dogs in the non-CHDN group in study III, 18 (52.9%) had echocardiographic evidence of LV systolic dysfunction and 18 (52.9%) had right atrial and/or ventricular dilation; 13 (38%) had both right- sided dilation and LV systolic dysfunction. The echocardiographic

44 and electrocardiographic findings in the non-CHDN group dogs are depicted in more detail in Fig. 4.

Pulmonary hypertension was detected in 9/64 (14%) of dogs, 4 of which also had right-sided atrial or ventricular dilation. Of the 6 non- CHDN dogs: in 4/6 pulmonary hypertension was mild, in the remaining 2 it was moderate. The dogs with moderate pulmonary hypertension had atrial fibrillation and dilation of the right heart. Collapse was reported in 20/64 (31.3%) of all dogs in the study and was not significantly associated with the presence of pulmonary hypertension (P = 0.252).

ĚǁŝƚŚŶŽĐŽŶŐĞŶŝƚĂůŚĞĂƌƚĚŝƐĞĂƐĞ͕ĐĂƌĚŝĂĐŵĂƐƐŽƌŶŽŶͲ ĐĂƌĚŝĂĐŶĞŽƉůĂƐŝĂ;ŶŽŶͲ,EͿ Ŷсϯϰ

^ŝŶƵƐƌŚǇƚŚŵ ƚƌŝĂůĨŝďƌŝůůĂƚŝŽŶ ^ƵƉƌĂǀĞŶƚƌŝĐƵůĂƌƚĂĐŚǇĐĂƌĚŝĂ ŶсϭϮ Ŷсϭϵ Ŷсϯ

^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶ ^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶ ^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶ ŶсϮ Ŷсϯ ŶсϬ

^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶĂŶĚ ^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶĂŶĚ ^ǇƐƚŽůŝĐĚǇƐĨƵŶĐƚŝŽŶĂŶĚ ƌŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ƌŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ƌŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ŶсϮ ŶсϭϬ Ŷсϭ

ZŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ZŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ZŝŐŚƚͲƐŝĚĞĚĚŝůĂƚŝŽŶ ŶсϬ Ŷсϰ Ŷсϭ Figure 4. The association between systolic dysfunction or right-sided dilation and the arrhythmia in the non-CHDN dogs

The echocardiographic assessment in study IV focused on the difference between the conscious and anaesthetised subjects as well as inter-observer repeatability of various measurements. No significant differences were detected between the studies performed on conscious and anaesthetised animals. The inter-observer variability was least for

LVIDd and Evel measurements (ICC 0.98, 95% CI 0.91–0.99 and ICC 0.96, 95% CI 0.88–0.99, respectively), followed by LA to aortic ratio (LA/Ao) (ICC 0.86, 95% CI 0.52 –0.96). The least reproducible 2D measurement was mitral prolapse (ICC 0.75, 95% CI 0.31–0.92). Of the colour flow Doppler echocardiographic variables, the inter-observer variability was less for VC than for PISA and JAR (ICC 0.89, 95% CI

45 0.61–0.98, ICC 0.52, 95% CI 0–0.88, and ICC 0.34, 95% CI 0–0.74, respectively).

5.4. Follow-up and survival information (I, II)

Median follow-up period for all dogs in the Holter study (I) was 500 days (IQR 186–917) with the follow-up period being longer for the dogs that were alive at the end of the study in all dogs (P < 0.001) as well as in the LVIDs≤35mm group (P = 0.005).

Kaplan-Meier survival estimates of the whole population showed that the median survival time was significantly shorter for the systolic dysfunction group compared to the dogs in the LVIDs≤35mm group (242 vs 2083 days, P < 0.001). The presence of syncope before presentation was not associated with decreased survival time.

In the LVIDs≤35mm group, male sex, > 50VPC/24h, the presence of polymorphic VPC, and the presence of VT were all significantly associated with shorter survival (P = 0.035, P = 0.003, P = 0.019, P = 0.002, respectively). Increased LA size (LA/Ao > 1.5), decreased FS% (FS% < 25%) the presence of syncope, atrial premature complexes or SVT were not predictive of survival outcome in this group. All three commonly used cut-off values of VPC/24h (50, 100, and 1000 VPC/24h) to differentiate Boxers affected and not affected by ARVC were significantly associated with shorter survival P( = 0.002, 0.018 and 0.019, respectively).

A bimodal age distribution was detected, and the young dogs were more likely to have < 50 VPC/24h (P < 0.001); the dogs presenting with more complex ventricular arrhythmias [polymorphic VPC (P = 0.001), bigeminy or trigeminy (P = 0.008), couplets or triplets (P < 0.001) and VT (P = 0.015)] were older. The dogs in the youngest age tertile had significantly longer survival than the other two tertiles P( < 0.001) and as there was no significant difference in survival between the two higher age tertiles (P = 0.574), an age of 4.5 years was used as a cut-off value in the multivariable statistical analysis.

Based on the multivariable Cox proportional hazards analysis, only age > 4.5 years (HR, 28.82, 95% CI, 3.03–247.31, P = 0.003), the presence of VT (HR, 3.407, 95% CI, 1.37–8.45, P = 0.008), and male sex (HR,

46 2.73, 95% CI, 1.04–7.12, P = 0.009) were independent predictors of survival in the LVIDs≤35mm group.

By the end of the TAPSE study (II), 26/50 (52%) of Boxer dogs were dead – 20 dogs (40%) for cardiac and 6 dogs (12%) for non-cardiac causes; 4 dogs (8%) were lost to follow up and 20 (40%) were alive. The median survival time to cardiac death for all dogs was 440 days (6–2083 days). For the multivariable Cox proportional hazards analysis four different models were used, to evaluate the association of TAPSE on time to cardiac death, to control for the effect of other measured variables. All models suggested that TAPSE < 15.1 mm was associated with a shorter time to cardiac death, independent of the presence of CHF (P = 0.02), systolic dysfunction (P = 0.023), detectable VT on Holter (P = 0.023), or diagnosis of ARVC (P = 0.029).

5.5. Cardiac magnetic resonance imaging data (IV)

The initial cMRI examinations were limited due to the difficulties of obtaining suitable images due to technical issues (e.g. usage of a suboptimal surface coil in the first seven dogs) and inability to complete the image acquisition within the timeframe specified in the study protocol (first nine dogs). After the initial learning curve, the complete data sets with optimal image quality were obtained for the final 12 consecutive dogs for a mean time of 34.5 minutes (±7.5 minutes).

Measurement of the first 5 complete datasets by two independent observers had excellent agreement for LV stroke volume and aortic stroke volume determinations with an ICC ≥ 0.99. The 95% inter- observer limits of agreement assessed using the Bland-Altmann method were narrower for LV stroke volume than aortic stroke volume (-1.15 to 0.55 ml, bias -0.30 and -1.89 to 0.07 ml, bias -0.56, respectively).

Complete cMRI data were available for 10 dogs to test correlations between cMRI-RF and echocardiographic variables. VC/Ao and Evel had the strongest correlations with MR severity (r = 0.89, 95% CI, 0.60–0.97, P = 0.001, and r = 0.86, 95% CI, 0.52–0.97, P = 0.001, respectively). The presence of multiple MR jets precluded VC measurements in one dog.

Correlations for other echocardiographic variables were: LA/Ao (r = 0.78, 95% CI, 0.30–0.94, P = 0.008), Prolapse/Ao (r = 0.72, 95% CI,

47 0.16–0.93, P = 0.02). The PISA method had a moderate correlation with cMRI-RF (r = 0.75, 95% CI, 0.09–0.95, P = 0.03), but was limited in two dogs due to multiple jets or a small jet size. Correlation of JAR with cMRI-RF was not significant (r = 0.48, 95% CI, -0.21–0.85,P = 0.16).

48 6. DISCUSSION

6.1. General considerations

The studies included in this thesis were set up to contribute to the diagnostics and prognostics of canine cardiomyopathies. Despite the high prevalence of cardiomyopathies in this species, only limited prognostic information is validated in clinical cohorts of patients. A major limitation to the availability of such information is the difficulty in early diagnosis due to the long pre-clinical (asymptomatic) phase. Furthermore, the ante-mortem diagnosis of cardiomyopathies presents a challenge as the findings of electrodiagnostics or diagnostic imaging modalities may be subtle or even non-existent (Tidholm et al., 2001; Meurs, 2004).

More specifically studiesI and II were aimed to improve the prognostic value of the information gained from commonly used diagnostic tests used in the diagnosis of Boxer cardiomyopathy – Holter monitoring (I) and two-dimensional echocardiography (II). Study III aimed to describe the cardiac pathologies in a population of another large breed dog, the DdB, known to have a high prevalence of congenital and acquired cardiac diseases and frequent heart rhythm disturbances.

Although cMRI is considered an essential tool in human cardiology, especially for the diagnostics of cardiomyopathies, to date this method is not widely used in animals. The main limitation for the usage of cMRI in the veterinary clinical setting is the relatively long image acquisition time and requirement for general anaesthesia. Although study IV focused on a valvular heart disease rather than cardiomyopathies, it was aimed to contribute to the diagnostic methods available in veterinary cardiology.

Results of studies I–III may have been influenced by their retrospective design. However, due to the limitations of the ante-mortem diagnostics and classification of cardiomyopathies, a prospective study could have introduced significant selection bias, making the results less applicable to the general population seen in clinical practice. Nevertheless, the retrospective design allowed us to additionally detect factors affecting mortality (calculation of the estimated survival times in studies I and II).

49 6.2. Prognostic factors of arrhythmogenic right ventricular cardiomyopathy in Boxers (I, II)

The ante mortem diagnostic criteria for ARVC in dogs are poorly defined. Some of the previous studies looking at the survival in Boxer dogs have attempted to limit their sample population to dogs presumptively affected with ARVC (Meurs, 2011; Palermo et al., 2011). In our study we attempted to overcome this limitation by looking for the prognostic factors in a broader population of Boxer dogs. This would be more useful in a typical clinical situation where the actual ARVC-status is uncertain. However, an extensive investigation to rule out potential alternative causes of ventricular arrhythmias was not attempted in each case, and therefore other conditions leading to ventricular arrhythmias could have been undetected in some dogs.

6.2.1. Syncope

The association of the presence of syncope with cardiac morbidity and mortality in humans with ARVC remains a subject of debate (Corrado et al., 2000; Hulot et al., 2004; Ciaramella et al., 2009; Pinamonti et al., 2011). In our studies, the most common reason for seeking a referral was a history of syncope (76% in study I and 78% in study II). Unlike the results of the study by Palermo et al. (2011), our findings suggest that the presence of syncope is not associated with a worse outcome. However, it is important to note that we also included Boxer dogs that did not meet the most common diagnostic criteria for ARVC, and thus a larger proportion of young dogs with neurocardiogenic syncope may have been included. The patients with neurocardiogenic syncope are generally considered to have a better prognosis. This may also be supported by the finding of a bimodal age distribution seen in our population. Fewer dogs in our study had LV systolic dysfunction than in Palermo’s study and only a small proportion of dogs in Palermo’s study had a Holter recording available, making it difficult to draw conclusions on the association between syncope and severity or frequency of arrhythmias in their study.

6.2.2. Systolic dysfunction

Previous studies have shown that LV systolic dysfunction was associated with a worse outcome in dogs with suspected ARVC (Meurs, 50 2011; Palermo et al., 2011). The results of study I also supported that observation. It is therefore understandable, that indicators of an advanced cardiac disease such as reduced FS%, dilation of the LA, presence of CHF, and the presence of > 50 VPC/24 h were predictors of a negative outcome in our overall study population. Furthermore, all these factors were strongly correlated with each other.

The long-term prognosis of Boxer dogs with CHF and systolic dysfunction is poor, and the clinical management of these cases differs significantly from that of Boxer dogs with normal echocardiographic findings and a suspected diagnosis of ARVC. As there is only limited information on the prognosis of in Boxers without systolic dysfunction (defined in our study as LVIDs ≤ 35 mm group) we aimed to identify prognostic factors for this sub-population. Therefore, dogs with LVIDs > 35 mm or the presence of CHF were excluded from further statistical analysis in study I.

6.2.3. Signalment: sex and age

Previous reports on sex distribution in ARVC Boxers has varied from male predominance (Harpster, 1991; Palermo et al., 2011) to female predominance (Baumwart et al., 2005). No sex difference was present in the sample population included in studies I and II (52% and 50% males, respectively). However, a male predominance (80%) was seen in the systolic dysfunction/CHF (LVIDs>35 mm/CHF) group, whilst a female predominance (42% male) was present in the normal systolic function/ no CHF (LVIDs>35 mm/no CHF) group. The Kaplan-Meier survival estimates showed that male dogs in the latter group had a significantly worse prognosis, and this correlation remained also significant in the final multivariable Cox proportional hazards model. In study II, no significant effect of sex was seen when comparing the ARVC and non- ARVC groups even when other variables such as presence of systolic dysfunction, presence of CHF and detectable VT were included (multivariable Cox proportional hazards analysis).

Age distribution in study I between systolic dysfunction group and normal systolic function group was not significantly different. The explanation to the bimodal age distribution seen in study I with two peaks at approximately 1 and 8 years of age might be that young dogs are more likely to be referred for isolated fainting episodes, whilst older

51 dogs are more likely to be presented for the investigation of more severe clinical signs such as CHF, arrhythmias detected on the routine physical examination or syncope associated with severe arrhythmias attributable to ARVC. A very mild age difference seen in study II, where the dogs in ARVC groups were older (4.9 vs 7.7 years), may further support this explanation.

Other authors have also suggested a bimodal age distribution in syncopal Boxers (Harpster, 1983; Meurs, 2004; Martin et al., 2009; Palermo et al., 2011). In our population, older dogs were more likely to have an increased complexity of ventricular arrhythmias (polymorphism, bigeminy, trigeminy, couplets, triplets, and VT) compared to young dogs. Furthermore, young dogs were more likely to have < 50 VPC/24 h. Whether this finding is caused by a more advanced disease state in the older population or the young dogs having a less malignant condition like neurocardiogenic syncope (Fisher, 1971), cannot be determined from our results.

The non-linearity of the age distribution as well as the association of multiple variables with age suggest that the long-term prognosis is not linearly associated with the age. This is further supported by the survival analysis of the age textiles, as the youngest tertile had the best outcome. This finding is very important in the clinical setting, as it suggests that young Boxers have a markedly better prognosis whilst the older population is more likely to be presented for a less benign condition such as ARVC or systemic disease. Interestingly in study II, TAPSE measurement was not associated with age, suggesting that it is a valuable prognostic factor when assessing dogs with suspected ARVC independent of age.

6.2.4. Tricuspid annular plane systolic excursion measurement (II)

Our results show that TAPSE can be used as a useful indirect estimate of RV function that was significantly associated with outcome in our study. Even when the presence of CHF, detection of VT on the Holter recording as well as LV systolic dysfunction were included in the multivariable analysis, TAPSE < 15.1 mm was still associated with a shorter time to an endpoint of cardiac death. This finding is in accordance with human studies, that have shown that TAPSE measurement < 18.1

52 mm is associated with a greater incidence of cardiac mortality in people with heart failure (Vizzardi et al., 2012). Although our results showed a weak correlation between TAPSE and LV systolic dysfunction, the measurement of TAPSE < 15.1 mm was still associated with shorter survival time even after the dogs with reduced systolic function were excluded from the statistical analysis.

Measurement of TAPSE is technically easy and the repeatability is good. However, it is important to notice that acquisition of appropriate images for the accurate measurement may be moderately challenging for an inexperienced operator or in a poorly compliant dog. Malalignment of the image may lead to false underestimation of TAPSE. In our retrospective analysis of previously recorded echocardiographic images, data from 31/81 (38%) otherwise suitable patients had to be excluded due to inadequate alignment or lack of recordings of appropriate imaging views. This was mainly because previously such images were not included in the routine echocardiographic protocol. Pariaut et al. (2012) have shown that suitable images for TAPSE measurement can be successfully acquired in dogs over a wide range of body weights and breeds.

The main practical implication of our findings is that TAPSE, a measurement that is simple to acquire with standard echocardiographic equipment, provides additional useful prognostic information in Boxer dogs with suspected ARVC.

6.2.5. Ventricular arrhythmias (I)

Presence of VT is considered one of the major criteria in the diagnosis of human ARVC (Marcus et al., 2010). Our study was the first to show that the presence of VT in Boxer dogs without echocardiographic evidence of structural heart disease is also associated with an increased risk of cardiac death: the median survival time in the LVIDs≤35 mm group was 1,244 days in the dogs with VT and > 2,083 days in the absence of VT.

The complexity and frequency of ventricular arrhythmias is associated with a worse prognosis in humans with ARVC (Corrado et al., 2000). Veterinary studies have failed to show similar association. However, the Cox univariable model showed that the presence of polymorphic VPC

53 and > 50 VPC/24 h were both significantly associated with a worse outcome in our study population. Perhaps due to the limited size of the study population, the presence of bigeminy or trigeminy, and couplets or triplets was not predictive of a worse outcome.

In human cardiology, a grading scheme of the frequency and complexity of ventricular arrhythmias has been recommended (Lown and Wolf, 1971). A modification of the Lown grading has been used in veterinary studies, although this scheme or its modification has not been adequately validated (Meurs et al., 1999; Meurs, 2011; Santilli et al., 2011). The data presented in our study could contribute to development of a new arrhythmia-based grading scheme. However, any of such schemes should be validated in a prospective cohort study.

6.3. Cardiac findings in a population of Dogue de Bordeaux (III)

6.3.1. Population overview

This study was aimed to describe the phenotype of DdB within a cardiac referral population in the United Kingdom, specifically with reference to the prevalence and associated features of supraventricular arrhythmias, especially AF.

The majority (58/64) of dogs presented for investigation of a suspected cardiac disease; the remaining six dogs had a previous diagnosis of non- cardiac neoplasia were examined to evaluate myocardial function prior to chemotherapy. Similarly to previous reports in the literature (Höllmer et al., 2008; Oliveira et al., 2011), subaortic stenosis was the most prevalent congenital heart disease in this population. Other congenital defects included patent ductus arteriosus, pulmonic stenosis, ventricular septal defect, atrial septal defect. Cardiac soft tissue masses were detected from echocardiographic examination of seven dogs, with only one confirmed histopathologically as hemangiosarcoma. For phenotypic echocardiographic and electrocardiographic descriptions, the dogs with congenital heart disease, cardiac soft tissue masses, and non-cardiac neoplasia were omitted from the analysis (non-CHDN group).

54 6.3.2. Supraventricular arrhythmias and left ventricular systolic dysfunction

Supraventricular arrhythmias were commonly detected in the study population with AF documented in 39% of all dogs. After exclusion of the dogs with congenital heart disease or neoplasia, the prevalence of AF was present in 56%. The prevalence in our study was higher than the 7.7% previously reported by Locatelli et al. (2011). This discrepancy was likely due to the differences in the inclusion criteria: the patients included in our study were selected from a referral population. In these patients a frequent cause for seeking a referral was auscultation of an arrhythmia or history of exercise intolerance. Despite the selection bias, our findings suggest that in the cardiac referral population of DdB in the United Kingdom, AF is the predominant arrhythmia.

Sixteen percent (3/19) of dogs with AF in the non-CHDN group, had a structurally normal heart on echocardiographic examination. The term ‘lone’ AF has been used to describe the AF in the absence of evidence of cardiac remodelling (Takemura et al., 2002; Menaut et al., 2005). More recently, however, the use of ‘lone’ AF has been questioned as the ability to diagnose the underlying structural changes has improved (Wyse et al., 2014). Menaut et al. (2005) have reported slow ventricular rates in dogs with AF and no evidence of structural heart disease. In our study, only one non-CHDN dog with AF and without structural heart disease, had a heart rate < 160 beats per minute on presentation.

A DCM-phenotype, consisting of increased LV dimensions, reduced systolic and diastolic function (Tidholm et al., 2001), was documented in 26% (9/34) of the non-CHDN DdBs. AF was present in more than half of these dogs (5/9). DCM phenotype leads to increased filling pressures and atrial dilation which, in turn may precipitate arrhythmias, particularly AF (Dukes-McEwan et al., 2003). Previous studies in other large dog breeds have shown a prevalence of AF in dogs with DCM as high as 45% and the most common breeds affected were the Irish wolfhound, Great Dane and Newfoundland dog (Tidholm and Jönsson, 1996; Vollmar, 2000; Meurs et al., 2001a; Martin et al., 2009). AF is commonly reported in the Irish wolfhound, with and without DCM. In one report up to 87.6% of Irish wolfhounds with a diagnosis of DCM had AF (Vollmar, 2000). The fact, that AF is so commonly found in large breeds, leaves the debate open whether supraventricular arrhythmias occur secondary to a

55 primary cardiomyopathy or the DCM phenotype in these dogs results from a prolonged or sustained tachycardia (Brownlie and Cobb, 1999).

6.3.3. Supraventricular arrhythmias and right atrial and ventricular dilation

Right sided atrial or ventricular dilatation was a common finding in this study. This was documented in 41% (27/64) of all dogs and over half of non-CHDN dogs without the evidence of significant pulmonary hypertension. Ohad et al. (2013) described the finding of AF in DdBs with tricuspid dysplasia. Tricuspid dysplasia is defined as a congenital malformation of the tricuspid valve apparatus, including the valve leaflets, chordae tendineae and papillary muscles that commonly results in annular distortion, tethering of the leaflets and valvular insufficiency commonly leading to right atrial dilatation (Oyama and Sisson, 2001; Ohad et al., 2013). Functional tricuspid insufficiency in humans has been reported secondary to annular distortion and alteration in papillary muscle geometry secondary to RV and LV dilatation, systolic dysfunction and respiratory phase (Rogers and Bolling, 2009; Topilsky et al., 2014; Mas et al., 2015). Due to the complex geometry, the accurate measurement of the severity of tricuspid insufficiency is challenging (Lancellottiet al., 2010). In humans without structural heart disease, chronic AF can lead to development of tricuspid regurgitation dependent of the AF type and duration (Najib et al., 2012; Park et al., 2015). As the influence of these factors in dogs is unknown, it was impossible to define correctly the normal or dysplastic tricuspid morphology in our study population and therefore only the right sided chamber dilation was reported.

Many dogs with AF in our study population had right atrial or ventricular dilatation and systolic dysfunction. The observation that the dogs in the non-CHDN group did not have right sided dilatation in the absence supraventricular arrhythmia or systolic dysfunction suggests that right sided dilatation might result from myocardial dysfunction rather than from a primary valvular disease in some DdBs.

6.3.4. Ventricular arrhythmias and right ventricular dilation

RV dilatation is a common finding in Boxer dogs with ARVC. Retrospective studies have reported that 21.5–43% of Boxer dogs with cardiomyopathy had RV dilatation (Basso et al., 2004; Palermo et al., 56 2011). Frequent and complex ventricular arrhythmias are a hallmark of ARVC (Harpster, 1991; Meurs et al., 1999; Basso et al., 2004; Palermo et al., 2011). Ventricular arrhythmias were also commonly found in our study, with almost a quarter (15/64) of the whole study population having ventricular arrhythmias documented. It is important to note, however, that Holter recordings were not performed in all cases, thus under-diagnosis of ventricular arrhythmias was likely. A small number of ventricular arrhythmias can be found in apparently normal dogs (Meurs et al., 2001c), with normal Dobermans and Boxers reportedly having < 50 and < 91 VPC/24, respectively (Stern et al., 2010; Wess et al., 2010). Using a dichotomised cut-off of 100 VPC/24h, approximately half of the dogs with ventricular arrhythmia in our study were classified as having a low frequency of ventricular arrhythmias.

6.3.5. Supraventricular tachycardia and age

SVT was documented in five dogs in our study. Despite the small number of dogs with SVT, the finding that the three non-CHDN dogs with SVT were younger than dogs with sinus rhythm and AF is noteworthy. SVT in dogs is commonly occurring with structural heart disease and systemic disease (Finster et al., 2008). Atrial pathology as a predisposing factor for SVT development has been described in dogs with ARVC (Fox et al., 2000; Basso et al., 2004; Vila et al., 2017), and SVT has also been reported in Boxer cardiomyopathy (Baumwart et al., 2005). It is likely, that paroxysmal SVT might predispose to development of AF as a result of atrial stretch, reduced atrial refractoriness and vagal activation during tachycardia (Chang et al., 2008). Santilli et al. (2010) documented focal atrial tachycardia by electrophysiological study in two DdBs in a series of 16 dogs diagnosed with focal atrial tachycardia. Out of those 16 dogs, seven had a surface ECG diagnosis of AF, focal atrial tachycardia occurred in younger dogs, and the it was mainly originating from the right atrium. Due to the retrospective nature of our study and the absence of electrophysiological mapping, it was impossible to detect the origin of the SVT and any temporal relationship between arrhythmia types and progression. However, it is possible that in some DdBs SVT might precede the development of AF.

57 6.4. Comparison of echocardiography with cardiac magnetic resonance imaging (IV) cMRI is a non-invasive diagnostic imaging modality that lacks the risk of radiation exposure and has excellent detail for imaging soft tissues. The main limitations for its use in veterinary medicine are the long acquisition time and the resulting requirement for chemical restraint, neither of which are ideal in patients with an advanced cardiac disease (Labruyère and Schwarz, 2013). Moreover, cMRI requires special expertise and very advanced scanners, limiting its use in veterinary medicine (Mai et al., 2010). In humans, cMRI is widely used for the assessment of cardiac and great vessel morphology as well as LV function, myocardial perfusion and viability (Constantine et al., 2004). Detailed discussions on recent advances on the common applications for cMRI in human medicine can be found elsewhere (Saeed et al., 2015; Mojibian and Pouraliakbar, 2018). cMRI also plays a key role in the diagnosis of ARVC, with cMRI findings included as both minor and major criteria in the revised task force criteria for the diagnosis of ARVC. Additionally it is also extensively used for risk stratification in ARVC patients (Femia et al., 2017; Bennett et al., 2019; Zaman et al., 2019). Another important application of cMRI in human medicine is the calculation of ventricular volumes and accurate calculation of the mitral and aortic regurgitation fraction (Mojibian and Pouraliakbar, 2018). Only a limited number of studies using cMRI are available in the veterinary literature, mainly being used in experimental studies on dogs (Gilbert et al., 2010; Mai et al., 2010; Labruyère and Schwarz, 2013; Sieslack et al., 2014).

In experimental models, cMRI has been shown to be accurate in LV volume quantification in dogs (Dell’italia et al., 1994) and has been used as a gold standard for comparison of echocardiographic measurements of stroke volume and ejection fraction (Kim et al., 2013). Our study was designed to compare the echocardiographic variables of the severity of MMVD with an indirect method to quantify cMRI-RF. cMRI in both humans and dogs is challenging and prone to imaging artefacts. The major sources of artefacts in cMRI are respiratory motion, cardiac motion and extremes of blood flow velocity (MacDonald et al., 2005). Using the image acquisition guidelines developed for the use in human medicine often results in sub-optimal image quality veterinary species. Significant differences in the size and anatomical orientation

58 of canine hearts resulted in poor image alignment in both short and long axis images in the early phases of our study. Furthermore, the ECG-gating (R-wave recognition) for the image acquisition during the initial attempts was also contribution to inadequate image quality. With increasing operator experience and effort put into correct positioning, gating and sequencing, as well as using different type of coils and breath- holding, a significant reduction of acquisition time and improvement of image quality was achieved. After the first nine dogs, the image acquisition protocol was considered adequate.

Whilst image acquisition was challenging, the image analysis of our technique was excellent, with a good inter-observer reproducibility. This is encouraging, supporting the use of cMRI as a useful quantitative research tool. As we could not justify extending the anaesthetic time, the repeatability of the repeatability of the image acquisition was not assessed in the current study.

Multiple studies have shown that mitral inflow E-wave velocity is an independent echocardiographic predictor of cardiac mortality in canine MMVD (Borgarelli et al., 2008; Borgarelli et al., 2012; Hezzell et al., 2012;

Sargent et al., 2015b). The correlation of Evel with cMRI-RF reflects the increased preload and LA to LV pressure gradient with increasing MR volume (Thomas et al., 1998). We found that Evel and VC/Ao had the strongest correlations with cMRI-RF severity in dogs with relatively mild MMVD. As cMRI-RFis a simple and highly repeatable measurement, our results support the inclusion of this in the assessment of canine MMVD.

VC diameter is another predictor of survival in canine MMVD (Sargent et al., 2015). However, its acquisition is limited by technical difficulty and further affected by the presence of multiple regurgitant orifices and other factors dictated by the laws of physics, such as the temporal variation in VC diameter throughout systole (Zoghbi et al., 2003; Moraldo et al., 2013). The correlation VC/Ao with cMRI-RF severity is encouraging, but further investigation is required for its validation.

Despite a moderate correlation of PISA-RF and cMRI-RF, the inter-observer reproducibility was poor. Furthermore, the PISA-RF measurement shares many limitations with VC measurement, the

59 PISA-RF method is likely to introduce errors, limiting its use in clinical setting (Moraldo et al., 2013).

Although the chamber enlargement in dogs with MMVD is well- established (Hansson et al., 2002; Borgarelli et al., 2008; Borgarelli et al., 2012; Hezzell et al., 2012), these changes occur relatively late in dogs with asymptomatic MMVD. Therefore, the indicators of LA or LV dilation are less useful in the early stages of this disease for the identification of patients at greatest risk of progression. Furthermore, the reproducibility of LA/Ao in this study was inferior to Evel and LVIDd. Measurement of LA/Ao is further hindered when the measurement is preformed from views that have the potential of inclusion of the entrance of the pulmonary veins.

Even though the dogs with MMVD and a larger volume MR would be expected to have a more rapid disease progression, it has not been proven by clinical studies. The variables identifying the severity of MR in asymptomatic dogs (such as VC/Ao and Evel) could prove useful in identification of dogs at greatest risk of progression. Certainly Evel, which correlated very well with cMRI-RF in this study, is already widely established as an important prognostic indicator in MMVD (Borgarelli et al., 2008; Borgarelli et al., 2012; Hezzell et al., 2012; Sargent et al., 2015). Whether a combination of different approaches of assessing the severity of MR in dogs would improve the prognostic yield requires further investigations.

60 7. CONCLUSIONS

The major conclusions of the work presented in this thesis are:

• Valuable prognostic information can be gained from Holter monitoring in Boxer dogs with a suspicion of ARVC. The presence of VT, older age, and male sex are associated with shorter survival. Boxer dogs < 4.5 years of age with normal echocardiographic examination and < 50 VPC/24h have a good long-term prognosis. • TAPSE adds valuable prognostic information for the management of Boxer dogs with syncope or suspected ARVC. More specifically, TAPSE < 15.1 mm is associated with shorter cardiac survival in Boxer dogs with ≥ 50 VPC/24h, irrespective of presence of LV systolic dysfunction, CHF or VT. • Congenital heart diseases and cardiac soft tissue masses are prevalent in DdB dogs. The most common congenital disease in this population was subaortic stenosis. Supraventricular arrhythmias, specifically fast AF, are prevalent in this breed. AF was often associated with cardiac remodelling, especially dilation of the right atrium or ventricle and reduced systolic function. • Despite being technically challenging, cMRI is a feasible method for

quantification of MR in small dogs. Mitral Evel and VC/Ao were the echocardiographic measures which had the best correlation to cMRI measurements allowing the comparison of the results obtained by the two diagnostic methods.

61 8. REFERENCES

Aguilar, M., Nattel, S., 2018. Taking the pulse of atrial fibrillation: a practical approach to rate control. Can. J. Cardiol. 34, 1526–1530. Alfakih, K., Plein, S., Thiele, H., Jones, T., Ridgway, J.P., Sivananthan, M.U., 2003. Normal human left and right ventricular dimensions for MRI as assessed by turbo gradient echo and steady-state free precession imaging sequences. J. Magn. Reson. Imaging 17, 323–329. Alvarez, J.A., Wilkinson, J.D., Lipshultz, S.E., 2007. Outcome predictors for pediatric dilated cardiomyopathy: a systematic review. Prog. Pediatr. Cardiol. 23, 25–32. Arbustini, E., Narula, N., Dec, G.W., Reddy, K.S., Greenberg, B., Kushwaha, S., Marwick, T., Pinney, S., Bellazzi, R., Favalli, V., Kramer, C., Roberts, R., Zoghbi, W.A., Bonow, R., Tavazzi, L., Fuster, V., Narula, J., 2013. The MOGE(S) classification for a phenotype- genotype nomenclature of cardiomyopathy. J. Am. Coll. Cardiol. 62, 2046–2072. Assomull, R.G., Prasad, S.K., Lyne, J., Smith, G., Burman, E.D., Khan, M., Sheppard, M.N., Poole-Wilson, P.A., Pennell, D.J., 2006. Cardiovascular magnetic resonance, fibrosis, and prognosis in dilated cardiomyopathy. J. Am. Coll. Cardiol. 48, 1977–1985. Basso, C., Corrado, D., Marcus, F.I., Nava, A., Thiene, G., 2009. Arrhythmogenic right ventricular cardiomyopathy. Lancet 373, 1289– 1300. Basso, C., Fox, P.R., Meurs, K.M., Towbin, J.A., Spier, A.W., Calabrese, F., Maron, B.J., Thiene, G., 2004. Arrhythmogenic right ventricular cardiomyopathy causing sudden cardiac death in boxer dogs: a new animal model of human disease. Circulation 109, 1180–1185. Baumwart, R.D., Meurs, K.M., 2005. Assessment of plasma brain natriuretic peptide concentration in Boxers with arrhythmogenic right ventricular cardiomyopathy. Am. J. Vet. Res. 66, 2086–2089. Baumwart, R.D., Meurs, K.M., Atkins, C.E., Bonagura, J.D., DeFrancesco, T.C., Keene, B.W., Koplitz, S., Fuentes, V.L., Miller, M.W., Rausch, W., Spier, A.W., 2005. Clinical, echocardiographic, and electrocardiographic abnormalities in Boxers with cardiomyopathy

62 and left ventricular systolic dysfunction: 48 cases (1985–2003). J. Am. Vet. Med. Assoc. 226, 1102–1104. Baumwart, R.D., Meurs, K.M., Raman, S. V, 2009. Magnetic resonance imaging of right ventricular morphology and function in boxer dogs with arrhythmogenic right ventricular cardiomyopathy. J. Vet. Intern. Med. 23, 271–274. Becker, M.A.J., Cornel, J.H., van de Ven, P.M., van Rossum, A.C., Allaart, C.P., Germans, T., 2018. The prognostic value of late gadolinium- enhanced cardiac magnetic resonance imaging in nonischemic dilated cardiomyopathy. JACC Cardiovasc. Imaging 11, 1274–1284. Bennett, R.G., Haqqani, H.M., Berruezo, A., Della Bella, P., Marchlinski, F.E., Hsu, C.J., Kumar, S., 2019. Arrhythmogenic cardiomyopathy in 2018–2019: ARVC/ALVC or both? Heart Lung Circ. 28, 164–177. Bhonsale, A., Te Riele, A.S.J.M., Sawant, A.C., Groeneweg, J.A., James, C.A., Murray, B., Tichnell, C., Mast, T.P., van der Pols, M.J., Cramer, M.J.M., Dooijes, D., van der Heijden, J.F., Tandri, H., van Tintelen, J.P., Judge, D.P., Hauer, R.N.W., Calkins, H., 2017. Cardiac phenotype and long-term prognosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia patients with late presentation. Heart Rhythm 14, 883–891. Boon, J., 2011. Evaluation of size, function and hemodynamics, in: Veterinary Echocardiography. pp. 206–207. Borgarelli, M., Crosara, S., Lamb, K., Savarino, P., La Rosa, G., Tarducci, A., Haggstrom, J., 2012. Survival characteristics and prognostic variables of dogs with preclinical chronic degenerative mitral valve disease attributable to myxomatous degeneration. J. Vet. Intern. Med. 26, 69–75. Borgarelli, M., Santilli, R.A., Chiavegato, D., D’Agnolo, G., Zanatta, R., Mannelli, A., Tarducci, A., 2006. Prognostic indicators for dogs with dilated cardiomyopathy. J. Vet. Intern. Med. 20, 104–110. Borgarelli, M., Savarino, P., Crosara, S., Santilli, R.A., Chiavegato, D., Poggi, M., Bellino, C., La Rosa, G., Zanatta, R., Haggstrom, J., Tarducci, A., 2008. Survival characteristics and prognostic variables of dogs with mitral regurgitation attributable to myxomatous valve disease. J. Vet. Intern. Med. 22, 120–128. Borgquist, R., Haugaa, K.H., Gilljam, T., Bundgaard, H., Hansen, J., Eschen, O., Jensen, H.K., Holst, A.G., Edvardsen, T., Svendsen,

63 J.H., Platonov, P.G., 2014. The diagnostic performance of imaging methods in ARVC using the 2010 task force criteria. Eur. Heart J. Cardiovasc. Imaging 15, 1219–1225. Bosman, L.P., Sammani, A., James, C.A., Cadrin-Tourigny, J., Calkins, H., van Tintelen, J.P., Hauer, R.N.W., Asselbergs, F.W., te Riele, A.S.J.M., 2018. Predicting arrhythmic risk in arrhythmogenic right ventricular cardiomyopathy: A systematic review and meta-analysis. Heart Rhythm 15, 1097–1107. Bright, J.M., Cali, J. V, 2000. Clinical usefulness of cardiac event recording in dogs and cats examined because of syncope, episodic collapse, or intermittent weakness: 60 cases (1997–1999). J. Am. Vet. Med. Assoc. 216, 1110–1114. Brownlie, S.E., Cobb, M.A., 1999. Observations on the development of congestive heart failure in Irish wolfhounds with dilated cardiomyopathy. J. Small Anim. Pract. 40, 371–377. Brun, F., Groeneweg, J.A., Gear, K., Sinagra, G., van der Heijden, J., Mestroni, L., Hauer, R.N., Borgstrom, M., Marcus, F.I., Hughes, T., 2016. Risk stratification in arrhythmic right ventricular cardiomyopathy without implantable cardioverter-defibrillators. JACC Clin. Electrophysiol. 2, 558–564. Brungs, A., Vollmar, A., Reese, S., Poulsen Nautrup, C., 2018. Echocardiographic indices of age- and gender-dependent cardiac remodeling over the adult lifespan in Irish Wolfhounds. J. Vet. Cardiol. 20, 307–318. Bussadori, C., Amberger, C., Le Bobinnec, G., Lombard, C.W., 2000. Guidelines for the echocardiographic studies of suspected subaortic and pulmonic stenosis. J. Vet. Cardiol. 2, 15–22. Bussadori, C., Pradelli, D., Borgarelli, M., Chiavegato, D., D’Agnolo, G., Menegazzo, L., Migliorini, F., Santilli, R., Zani, A., Quintavalla, C., 2009. Congenital heart disease in boxer dogs: Results of 6 years of breed screening. Vet. J. 181, 187–192. Cadrin-Tourigny, J., Bosman, L.P., Tadros, R., Talajic, M., Rivard, L., James, C.A., Khairy, P., 2019. Risk stratification for ventricular arrhythmias and sudden cardiac death in arrhythmogenic right ventricular cardiomyopathy: an update. Expert Rev. Cardiovasc. Ther. 17, 645–651.

64 Calkins, H., Corrado, D., Marcus, F., 2017. Risk stratification in arrhythmogenic right ventricular cardiomyopathy. Circulation 136, 2068–2082. Calvert, C.A., Pickus, C.W., Jacobs, G.J., Brown, J., 1997. Signalment, survival, and prognostic factors in Doberman pinschers with end- stage cardiomyopathy. J. Vet. Intern. Med. 11, 323–326. Caro-Vadillo, A., García-Guasch, L., Carretón, E., Montoya-Alonso, J.A., Manubens, J., 2013. Arrhythmogenic right ventricular cardiomyopathy in boxer dogs: a retrospective study of survival. Vet. Rec. 172, 268. Carvalho, E.R., Zacché, E., Fenerich, M., Camacho, A.A., Sousa, M.G., 2019. Prognostic value of coupling interval, prematurity index and heart rate variability in Boxer dogs. Pesqui. Veterinária Bras. 39, 635– 642. Cattanach, B.M., Dukes-McEwan, J., Wotton, P.R., Stephenson, H.M., Hamilton, R.M., 2015. A pedigree-based genetic appraisal of Boxer ARVC and the role of the Striatin mutation. Vet. Rec. 176, 492. Chamas, P.P.C., Oliveira, V.M.C., Yamaki, F.L., Goldfeder, G.T., Larsson, M.H.M.A., 2016. Valor prognóstico da variabilidade da frequência cardíaca e da eletrocardiografia ambulatorial em cães Boxer com cardiomiopatia arritmogênica do ventrículo direito . Arq. Bras. Med. Veterinária e Zootec. 68, 1219–1227. Chang, S.L., Tai, C.T., Lin, Y.J., Lo, L.W., Tuan, T.C., Udyavar, A.R., Tsao, H.M., Hsieh, M.H., Hu, Y.F., Chiang, S.J., Chen, Y.J., Wongcharoen, W., Ueng, K.C., Chen, S.A., 2008. Electrophysiological characteristics and catheter ablation in patients with paroxysmal supraventricular tachycardia and paroxysmal atrial fibrillation. J. Cardiovasc. Electrophysiol. 19, 367–373. Chetboul, V., Gouni, V., Sampedrano, C.C., Tissier, R., Serres, F., Pouchelon, J.-L., 2007. Assessment of regional systolic and diastolic myocardial function using tissue Doppler and strain imaging in dogs with dilated cardiomyopathy. J. Vet. Intern. Med. 21, 719–730. Ciaramella, P., Basso, C., Loria, A. Di, Piantedosi, D., 2009. Arrhythmogenic right ventricular cardiomyopathy associated with severe left ventricular involvement in a cat. J. Vet. Cardiol. 11, 41–45. Constantine, G., Shan, K., Flamm, S.D., Sivananthan, M.U., 2004. Role of MRI in clinical cardiology. Lancet 363, 2162–2171.

65 Cornell, C.C., Kittleson, M.D., Della Torre, P., Häggström, J., Lombard, C.W., Pedersen, H.D., Vollmar, A., Wey, A., 2004. Allometric scaling of M-mode cardiac measurements in normal adult dogs. J. Vet. Intern. Med. 18, 311–321. Corrado, D., Basso, C., Thiene, G., 2000. Arrhythmogenic right ventricular cardiomyopathy: diagnosis, prognosis, and treatment. Heart 83, 588–595. Cunningham, S., Sweeney, J.T., MacGregor, J., Barton, B.A., Rush, J.E., 2018. Clinical features of English bulldogs with presumed arrhythmogenic right ventricular cardiomyopathy: 31 cases (2001- 2013). J. Am. Anim. Hosp. Assoc. 54, 95–102. Cunningham, S.M., Aona, B.D., Antoon, K., Rush, J.E., Barton, B.A., 2018. Echocardiographic assessment of right ventricular systolic function in Boxers with arrhythmogenic right ventricular cardiomyopathy. J. Vet. Cardiol. 20, 343–353. Dambach, D.M., Lannon, A., Sleeper, M.M., Buchanan, J., 1999. Familial dilated cardiomyopathy of young Portuguese water dogs. J. Vet. Intern. Med. 13, 65–71. de Madron, E., 2015. Global left ventricular systolic function assessment, in: Clinical Echocardiography of the Dog and Cat. Elsevier, pp. 111– 125. Dell’italia, L.J., Blackwell, G.G., Pearce, D.J., Thorn, B., Pohost, G.M., 1994. Assessment of ventricular volumes using cine magnetic resonance in the intact dog: a comparison of measurement methods. Invest. Radiol. 29, 162–167. Dickson, D., Shave, R., Rishniw, M., Harris, J., Patteson, M., 2016. Reference intervals for transthoracic echocardiography in the English springer spaniel: a prospective, longitudinal study. J. Small Anim. Pract. 57, 520–528. Dini, F.L., Conti, U., Fontanive, P., Andreini, D., Banti, S., Braccini, L., De Tommasi, S.M., 2007. Right ventricular dysfunction is a major predictor of outcome in patients with moderate to severe mitral regurgitation and left ventricular dysfunction. Am. Heart J. 154, 172– 179. Distl, O., Vollmar, A.C., Broschk, C., Hamann, H., Fox, P.R., 2007. Complex segregation analysis of dilated cardiomyopathy (DCM) in Irish wolfhounds. Heredity 99, 460–465.

66 Dukes-McEwan, J., Borgarelli, M., Tidholm, A., Vollmar, A.C., Häggström, J., 2003. Proposed guidelines for the diagnosis of canine idiopathic dilated cardiomyopathy. J. Vet. Cardiol. 5, 7–19. Egenvall, A., Bonnett, B.N., Häggström, J., 2006. Heart disease as a cause of death in insured Swedish dogs younger than 10 years of age. J. Vet. Intern. Med. 20, 894–903. Elliott, P., Andersson, B., Arbustini, E., Bilinska, Z., Cecchi, F., Charron, P., Dubourg, O., Kühl, U., Maisch, B., McKenna, W.J., Monserrat, L., Pankuweit, S., Rapezzi, C., Seferovic, P., Tavazzi, L., Keren, A., 2008. Classification of the cardiomyopathies: A position statement from the european society of cardiology working group on myocardial and pericardial diseases. Eur. Heart J. 29, 270–276. Ellis, E.R., Josephson, M.E., 2013. What About Tachycardia-induced Cardiomyopathy? Arrhythm. Electrophysiol. Rev. 2, 82–90. Femia, G., Sy, R.W., Puranik, R., 2017. Systematic review: impact of the new task force criteria in the diagnosis of arrhythmogenic right ventricular cardiomyopathy. Int. J. Cardiol. 241, 311–317. Finster, S.T., Defrancesco, T.C., Atkins, C.E., Hansen, B.D., Keene, B.W., 2008. Supraventricular tachycardia in dogs: 65 cases (1990–2007). J. Vet. Emerg. Crit. Care 18, 503–510. Fisher, E.W., 1971. Fainting in Boxers-the possibility of vaso-vagal syncope (Adams-Stokes attacks). J. Small Anim. Pract. 12, 347–349. Fox, P.R., Basso, C., Thiene, G., Maron, B.J., 2014. Spontaneously occurring restrictive nonhypertrophied cardiomyopathy in domestic cats: a new animal model of human disease. Cardiovasc. Pathol. 23, 28–34. Fox, P.R., Maron, B.J., Basso, C., Liu, S.K., Thiene, G., 2000. Spontaneously occurring arrhythmogenic right ventricular cardiomyopathy in the domestic cat: a new animal model similar to the human disease. Circulation 102, 1863–1870. Freeman, L.M., Michel, K.E., Brown, D.J., Kaplan, P.M., Stamoulis, M.E., Rosenthal, S.L., Keene, B.W., Rush, J.E., 1996. Idiopathic dilated cardiomyopathy in Dalmatians: nine cases (1990–1995). J. Am. Vet. Med. Assoc. 209, 1592–1596. Gane, S., O’Neill, D., Boswood, A., Church, D., Brodbelt, D., Mattin, M., 2017. Prevalence and clinical outcomes of dilated cardiomyopathy

67 (DCM) in dogs under first opinion veterinary care in the UK, in: BSAVA Congress Proceedings 2017. British Small Animal Veterinary Association, p. 471. Geisterfer-Lowrance, A.A.T., Kass, S., Tanigawa, G., Vosberg, H.-P., McKenna, W., Seidman, C.E., Seidman, J.G., 1990. A molecular basis for familial hypertrophic cardiomyopathy: a β cardiac myosin heavy chain gene missense mutation. Cell 62, 999–1006. Gilbert, S.H., McConnell, F.J., Holden, A. V, Sivananthan, M.U., Dukes- McEwan, J., 2010. The potential role of MRI in veterinary clinical cardiology. Vet. J. 183, 124–134. Gopinathannair, R., Etheridge, S.P., Marchlinski, F.E., Spinale, F.G., Lakkireddy, D., Olshansky, B., 2015. Arrhythmia-induced cardiomyopathies mechanisms, recognition, and management. J. Am. Coll. Cardiol. 66, 1714–1728. Gupta, S., Khan, F., Shapiro, M., Weeks, S.G., Litwin, S.E., Michaels, A.D., 2008. The associations between tricuspid annular plane systolic excursion (TAPSE), ventricular dyssynchrony, and ventricular interaction in heart failure patients. Eur. J. Echocardiogr. 9, 766–771. Haddad, F., Hunt, S.A., Rosenthal, D.N., Murphy, D.J., 2008. Right ventricular function in cardiovascular disease, part I: anatomy, physiology, aging, and functional assessment of the right ventricle. Circulation 117, 1436–1448. Häggström, J., Duelund Pedersen, H., Vet Sci, D., Kvart, C., 2004. New insights into degenerative mitral valve disease in dogs. Vet. Clin. North Am. Small Anim. Pract. 34, 1209–1226. Halliday, B.P., Cleland, J.G.F., Goldberger, J.J., Prasad, S.K., 2017. Personalizing risk stratification for sudden death in dilated cardiomyopathy. Circulation 136, 215–231. Hansson, K., Häggström, J., Kvart, C., Lord, P., 2002. Left atrial to aortic root indices using two-dimensional and M-mode echocardiography in cavalier King Charles spaniels with and without left atrial enlargement. Vet. Radiol. Ultrasound 43, 568–575. Hariu, C.D., Carpenter, D.H., 2010. Arrhythmogenic right ventricular cardiomyopathy in boxers. Compend Contin Educ Vet. 32, E1–E7.

68 Harmon, M.W., Leach, S.B., Lamb, K.E., 2017. Dilated Cardiomyopathy in Standard Schnauzers: Retrospective Study of 15 Cases. J. Am. Anim. Hosp. Assoc. 53, 38–44. Harpster, N.K., 1991. Boxer cardiomyopathy. A review of the long-term benefits of antiarrhythmic therapy. Vet. Clin. North Am. Small Anim. Pract. 21, 989–1004. Harpster, N.K., 1983. Boxer Cardiomyopathy, in: Kirk, R. (Ed.), Current Veterinary Therapy VIII. Philadelphia, PA, pp. 329–337. Hazebroek, M.R., Moors, S., Dennert, R., van den Wijngaard, A., Krapels, I., Hoos, M., Verdonschot, J., Merken, J.J., de Vries, B., Wolffs, P.F., Crijns, H.J.G.M., Brunner-La Rocca, H.-P., Heymans, S., 2015. Prognostic relevance of gene-environment interactions in patients with dilated cardiomyopathy. J. Am. Coll. Cardiol. 66, 1313–1323. Hezzell, M.J., Boswood, A., Moonarmart, W., Elliott, J., 2012. Selected echocardiographic variables change more rapidly in dogs that die from myxomatous mitral valve disease. J. Vet. Cardiol. 14, 269–279. Höllmer, M., Willesen, J.L., Jensen, A.T., Koch, J., 2008. Aortic stenosis in the Dogue de Bordeaux. J. Small Anim. Pract. 49, 432–437. Hulot, J.-S., Jouven, X., Empana, J.-P., Frank, R., Fontaine, G., 2004. Natural history and risk stratification of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Circulation 110, 1879–1884. Jacoby, D., McKenna, W.J., 2012. Genetics of inherited cardiomyopathy. Eur. Heart J. 33, 296–304. Japp, A.G., Gulati, A., Cook, S.A., Cowie, M.R., Prasad, S.K., 2016. The diagnosis and evaluation of dilated cardiomyopathy. J. Am. Coll. Cardiol. 67, 2996–3010. Kim, J.H., Lee, M.S., Lee, S.Y., Kim, S.Y., Lee, S.Y., Lee, S.J., Park, Y.W., Yeo, J.H., Song, S.H., Park, N.W., Hong, S.W., Choi, S.I., Eom, K.D., 2013. Contrast echocardiography to assess left ventricular volume and function in Beagle dogs: comparison with 3-Tesla dual source parallel cardiac magnetic resonance imaging. Vet. J. 198, 450–456. Kraus, M.S., Moïse, N.S., Rishniw, M., Dykes, N., Erb, H.N., 2002. Morphology of ventricular arrhythmias in the boxer as measured by 12-lead electrocardiography with pace-mapping comparison. J. Vet. Intern. Med. 16, 153–158.

69 Kubo, T., Matsumura, Y., Kitaoka, H., Okawa, M., Hirota, T., Hamada, T., Hitomi, N., Hoshikawa, E., Hayato, K., Shimizu, Y., Yamasaki, N., Yabe, T., Nishinaga, M., Takata, J., Doi, Y., 2008. Improvement in prognosis of dilated cardiomyopathy in the elderly over the past 20 years. J. Cardiol. 52, 111–117. Kusano, K., 2018. Ventricular arrhythmias in nonischemic cardiomyopathy. J. Arrhythmia 34, 336–346. Labruyère, J., Schwarz, T., 2013. CT and MRI in veterinary patients: an update on recent advances. In Pract. 35, 546–563. Lancellotti, P., Moura, L., Pierard, L.A., Agricola, E., Popescu, B.A., Tribouilloy, C., Hagendorff, A., Monin, J.L., Badano, L., Zamorano, J.L., Sicari, R., Vahanian, A., Roelandt, J.R.T.C., 2010. European association of echocardiography recommendations for the assessment of valvular regurgitation. Part 2: mitral and tricuspid regurgitation (native valve disease). Eur. J. Echocardiogr. 11, 307–332. Legge, C.H., López, A., Hanna, P., Côté, E., Hare, E., Martinson, S.A., 2013. Histological characterization of dilated cardiomyopathy in the juvenile Toy Manchester Terrier. Vet. Pathol. 50, 1043–1052. Lehmkuhl, L.B., Bonagura, J.D., 1994. Comparison of transducer placement sites for Doppler echocardiography in dogs with subaortic stenosis. Am. J. Vet. Res. 55, 192–198. Lemola, K., Brunckhorst, C., Helfenstein, U., Oechslin, E., Jenni, R., Duru, F., 2005. Predictors of adverse outcome in patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy: long term experience of a tertiary care centre. Heart 91, 1167–1172. Lichtenberger, J., Meurs, K.M., Côté, E., 2018. Preliminary assessment of a novel 14-day electrocardiographic adhesive patch monitor in dogs. J. Am. Anim. Hosp. Assoc. 54, 138–143. Locatelli, C., Santini, A., Bonometti, G.A., Palermo, V., Scarpa, P., Sala, E., Brambilla, P.G., 2011. Echocardiographic values in clinically healthy adult Dogue de Bordeaux dogs. J. Small Anim. Pract. 52, 246–253. Lown, B., Wolf, M., 1971. Approaches to sudden death from coronary heart disease. Circulation 44, 130–142. MacDonald, K.A., Kittleson, M.D., Reed, T., Larson, R., Kass, P., Wisner, E.R., 2005. Quantification of left ventricular mass using cardiac

70 magnetic resonance imaging compared with echocardiography in domestic cats. Vet. Radiol. Ultrasound 46, 192–199. MacKie, B.A., Stepien, R.L., Kellihan, H.B., 2010. Retrospective analysis of an implantable loop recorder for evaluation of syncope, collapse, or intermittent weakness in 23 dogs (2004–2008). J. Vet. Cardiol. 12, 25–33. Mai, W., Weisse, C., Sleeper, M.M., 2010. Cardiac magnetic resonance imaging in normal dogs and two dogs with heart base tumor. Vet. Radiol. Ultrasound 51, 428–435. Marcus, F.I., McKenna, W.J., Sherrill, D., Basso, C., Bauce, B., Bluemke, D.A., Calkins, H., Corrado, D., Cox, M.G.P.J., Daubert, J.P., Fontaine, G., Gear, K., Hauer, R., Nava, A., Picard, M.H., Protonotarios, N., Saffitz, J.E., Yoerger Sanborn, D.M., Steinberg, J.S., Tandri, H., Thiene, G., Towbin, J.A., Tsatsopoulou, A., Wichter, T., Zareba, W., 2010. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/ dysplasia: proposed modification of the task force criteria. Circulation 121, 1533–1541. Maron, B.J., Ommen, S.R., Semsarian, C., Spirito, P., Olivotto, I., Maron, M.S., 2014. Hypertrophic cardiomyopathy: present and future, with translation into contemporary cardiovascular medicine. J. Am. Coll. Cardiol. 64, 83–99. Maron, B.J., Towbin, J.A., Thiene, G., Antzelevitch, C., Corrado, D., Arnett, D., Moss, A.J., Seidman, C.E., Young, J.B., 2006. Contemporary definitions and classification of the cardiomyopathies. Circulation 113, 1807–1816. Martin, M.W.S., Stafford Johnson, M.J., Celona, B., 2009. Canine dilated cardiomyopathy: a retrospective study of signalment, presentation and clinical findings in 369 cases. J. Small Anim. Pract. 50, 23–29. Mas, P.T., Rodríguez-Palomares, J.F., Antunes, M.J., 2015. Secondary tricuspid valve regurgitation: a forgotten entity. Heart 101, 1840– 1848. Maupain, C., Badenco, N., Pousset, F., Waintraub, X., Duthoit, G., Chastre, T., Himbert, C., Hébert, J.-L., Frank, R., Hidden-Lucet, F., Gandjbakhch, E., 2018. Risk atratification in arrhythmogenic right ventricular cardiomyopathy/dysplasia without an implantable cardioverter-defibrillator. JACC. Clin. Electrophysiol. 4, 757–768.

71 Mausberg, T.-B., Wess, G., Simak, J., Keller, L., Drögemüller, M., Drögemüller, C., Webster, M.T., Stephenson, H., Dukes-McEwan, J., Leeb, T., 2011. A locus on chromosome 5 is associated with dilated cardiomyopathy in Doberman Pinschers. PLoS One 6, e20042. McAulay, G., Borgeat, K., Luis Fuentes, V., 2018. An online health survey of Dogue de Bordeaux owners and breeders with special emphasis on cardiac disease. Vet. J. 232, 78–82. McCauley, M.D., Wehrens, X.H.T., 2009. Animal models of arrhythmogenic cardiomyopathy. Dis. Model. Mech. 2, 563–570. McEwan, J.D., 2000. Canine dilated cardiomyopathy 1. Breed manifestations and diagnosis. In Pract. 22, 520–530. Menaut, P., Bélanger, M.C., Beauchamp, G., Ponzio, N.M., Moïse, N.S., 2005. Atrial fibrillation in dogs with and without structural or functional cardiac disease: a retrospective study of 109 cases. J. Vet. Cardiol. 7, 75–83. Mercer-Rosa, L., Parnell, A., Forfia, P.R., Yang, W., Goldmuntz, E., Kawut, S.M., 2013. Tricuspid annular plane systolic excursion in the assessment of right ventricular function in children and adolescents after repair of tetralogy of fallot. J. Am. Soc. Echocardiogr. 26, 1322– 1329. Merlo, M., Cannatà, A., Gobbo, M., Stolfo, D., Elliott, P.M., Sinagra, G., 2018. Evolving concepts in dilated cardiomyopathy. Eur. J. Heart Fail. 20, 228–239. Mérot, J., Probst, V., Debailleul, M., Gerlach, U., Moise, N.S., Le Marec, H., Charpentier, F., 2000. Electropharmacological characterization of cardiac repolarization in German shepherd dogs with an inherited syndrome of sudden death: Abnormal response to potassium channel blockers. J. Am. Coll. Cardiol. 36, 939–947. Meurs, K.M., 2017. Arrhythmogenic right ventricular cardiomyopathy in the Boxer dog: an update. Vet. Clin. North Am. Small Anim. Pract. 47, 1103–1111. Meurs, K.M., 2011. Research abstract program of the 2011 ACVIM Forum Denver, Colorado, June 15–18, 2011. J. Vet. Intern. Med. 25, 632–767. Meurs, K.M., 2004. Boxer dog cardiomyopathy: an update. Vet. Clin. North Am. Small Anim. Pract. 34, 1235–1244.

72 Meurs, K.M., Fox, P.R., Norgard, M., Spier, A.W., Lamb, A., Koplitz, S.L., Baumwart, R.D., 2007. A prospective genetic evaluation of familial dilated cardiomyopathy in the Doberman pinscher. J. Vet. Intern. Med. 21, 1016–1020. Meurs, K.M., Lahmers, S., Keene, B.W., White, S.N., Oyama, M.A., Mauceli, E., Lindblad-Toh, K., 2012. A splice site mutation in a gene encoding for PDK4, a mitochondrial protein, is associated with the development of dilated cardiomyopathy in the Doberman pinscher. Hum. Genet. 131, 1319–1325. Meurs, K.M., Mauceli, E., Lahmers, S., Acland, G.M., White, S.N., Lindblad-Toh, K., 2010. Genome-wide association identifies a deletion in the 3′ untranslated region of Striatin in a canine model of arrhythmogenic right ventricular cardiomyopathy. Hum. Genet. 128, 315–324. Meurs, K.M., Miller, M.W., Wright, N.A., 2001a. Clinical features of dilated cardiomyopathy in Great Danes and results of a pedigree analysis: 17 cases (1990–2000). J. Am. Vet. Med. Assoc. 218, 729–32. Meurs, K.M., Spier, A.W., Miller, M.W., Lehmkuhl, L., Towbin, J.A., 1999. Familial ventricular arrhythmias in Boxers. J. Vet. Intern. Med. 13, 437–439. Meurs, K.M., Spier, A.W., Wright, N.A., Hamlin, R.L., 2001b. Comparison of in-hospital versus 24-hour ambulatory electrocardiography for detection of ventricular premature complexes in mature Boxers. J. Am. Vet. Med. Assoc. 218, 222–224. Meurs, K.M., Spier, A.W., Wright, N.A., Hamlin, R.L., 2001c. Use of ambulatory electrocardiography for detection of ventricular premature complexes in healthy dogs. J. Am. Vet. Med. Assoc. 218, 1291–1292. Meurs, K.M., Stern, J.A., Reina-Doreste, Y., Spier, A.W., Koplitz, S.L., Baumwart, R.D., 2014. Natural history of arrhythmogenic right ventricular cardiomyopathy in the boxer dog: a prospective study. J. Vet. Intern. Med. 28, 1214–1220. Miles, C., Fanton, Z., Tome, M., Behr, E.R., 2019. Inherited cardiomyopathies. BMJ 365, 1–8. Moise, N.S., Meyers-Wallen, V., Flahive, W.J., Valentine, B.A., Scarlett, J.M., Brown, C.A., Chavkin, M.J., Dugger, D.A., Renaud-Farrell, S., Kornreich, B., Schoenborn, W.C., Sparks, J.R., Gilmour, R.F., 1994.

73 Inherited ventricular arrhythmias and sudden death in German shepherd dogs. J. Am. Coll. Cardiol. 24, 233–243. Mojibian, H., Pouraliakbar, H., 2018. Cardiac Magnetic Resonance Imaging, in: Practical Cardiology. Elsevier, pp. 159–166. Monnet, E., Orton, E.C., Salman, M., Boon, J., 1995. Idiopathic Dilated Cardiomyopathy in Dogs: Survival and Prognostic Indicators. J. Vet. Intern. Med. 9, 12–17. Moraldo, M., Cecaro, F., Shun-Shin, M., Pabari, P.A., Davies, J.E., Xu, X.Y., Hughes, A.D., Manisty, C., Francis, D.P., 2013. Evidence-based recommendations for PISA measurements in mitral regurgitation: systematic review, clinical and in-vitro study. Int. J. Cardiol. 168, 1220–1228. Najib, M.Q., Vinales, K.L., Vittala, S.S., Challa, S., Lee, H.R., Chaliki, H.P., 2012. Predictors for the development of severe tricuspid regurgitation with anatomically normal valve in patients with atrial fibrillation. Echocardiography 29, 140–146. Naneix, A.L., Périer, M.C., Beganton, F., Jouven, X., Lorin De La Grandmaison, G., 2015. Sudden adult death: an autopsy series of 534 cases with gender and control comparison. J. Forensic Leg. Med. 32, 10–15. Norman, M., Simpson, M., Mogensen, J., Shaw, A., Hughes, S., Syrris, P., Sen-Chowdhry, S., Rowland, E., Crosby, A., McKenna, W.J., 2005. Novel mutation in desmoplakin causes arrhythmogenic left ventricular cardiomyopathy. Circulation 112, 636–642. O’Grady, M.R., O’Sullivan, M.L., 2004. Dilated cardiomyopathy: an update. Vet. Clin. North Am. Small Anim. Pract. 34, 1187–1207. Ohad, D.G., Avrahami, A., Waner, T., David, L., 2013. The occurrence and suspected mode of inheritance of congenital subaortic stenosis and tricuspid valve dysplasia in Dogue de Bordeaux dogs. Vet. J. 197, 351–357. Ohno, S., 2016. The genetic background of arrhythmogenic right ventricular cardiomyopathy. J. Arrhythmia 32, 398–403. Oliveira, P., Domenech, O., Silva, J., Vannini, S., Bussadori, R., Bussadori, C., 2011. Retrospective review of congenital heart disease in 976 dogs. J. Vet. Intern. Med. 25, 477–483.

74 Oomen, A.W.G.J., Semsarian, C., Puranik, R., Sy, R.W., 2018. Diagnosis of arrhythmogenic right ventricular cardiomyopathy: progress and pitfalls. Heart Lung Circ. 27, 1310–1317. Oyama, M.A., Sisson, D.D., 2001. Evaluation of canine congenital heart disease using an echocardiographic algorithm. J. Am. Anim. Hosp. Assoc. 37, 519–535. Palermo, V., Stafford Johnson, M.J., Sala, E., Brambilla, P.G., Martin, M.W.S., 2011. Cardiomyopathy in Boxer dogs: a retrospective study of the clinical presentation, diagnostic findings and survival. J. Vet. Cardiol. 13, 45–55. Pariaut, R., Saelinger, C., Strickland, K.N., Beaufrère, H., Reynolds, C.A., Vila, J., 2012. Tricuspid annular plane systolic excursion (TAPSE) in dogs: reference values and impact of pulmonary hypertension. J. Vet. Intern. Med. 26, 1148–1154. Park, J.H., Shin, S.H., Lee, M.J., Lee, M.D., Shim, H.I., Yoon, J., Oh, S., Kim, D.H., Park, S.D., Kwon, S.W., Woo, S.I., Park, K.S., Kwan, J., 2015. Clinical and echocardiographic factors affecting tricuspid regurgitation severity in the patients with lone atrial fibrillation. J. Cardiovasc. Ultrasound 23, 136–142. Philipp, U., Broschk, C., Vollmar, A., Distl, O., 2007. Evaluation of tafazzin as candidate for dilated cardiomyopathy in Irish wolfhounds. J. Hered. 98, 506–509. Pinamonti, B., Dragos, A.M., Pyxaras, S.A., Merlo, M., Pivetta, A., Barbati, G., Di Lenarda, A., Morgera, T., Mestroni, L., Sinagra, G., 2011. Prognostic predictors in arrhythmogenic right ventricular cardiomyopathy: results from a 10-year registry. Eur. Heart J. 32, 1105–1113. Pinto, Y.M., Elliott, P.M., Arbustini, E., Adler, Y., Anastasakis, A., Böhm, M., Duboc, D., Gimeno, J., De Groote, P., Imazio, M., Heymans, S., Klingel, K., Komajda, M., Limongelli, G., Linhart, A., Mogensen, J., Moon, J., Pieper, P.G., Seferovic, P.M., Schueler, S., Zamorano, J.L., Caforio, A.L.P., Charron, P., 2016. Proposal for a revised definition of dilated cardiomyopathy, hypokinetic non-dilated cardiomyopathy, and its implications for clinical practice: A position statement of the ESC working group on myocardial and pericardial diseases. Eur. Heart J. 37, 1850–1858.

75 Prakasa, K.R., Wang, J., Tandri, H., Dalal, D., Bomma, C., Chojnowski, R., James, C., Tichnell, C., Russell, S., Judge, D., Corretti, M., Bluemke, D., Calkins, H., Abraham, T.P., 2007. Utility of tissue Doppler and strain echocardiography in arrhythmogenic right ventricular dysplasia/ cardiomyopathy. Am. J. Cardiol. 100, 507–512. Protas, L., Sosunov, E.A., Anyukhovsky, E.P., Moïse, N.S., Rosen, M.R., Robinson, R.B., 2005. Regional dispersion of L-type calcium current in ventricular myocytes of German shepherd dogs with lethal cardiac arrhythmias. Heart Rhythm 2, 172–176. Protonotarios, N., Anastasakis, A., Antoniades, L., Chlouverakis, G., Syrris, P., Basso, C., Asimaki, A., Theopistou, A., Stefanadis, C., Thiene, G., McKenna, W.J., Tsatsopoulou, A., 2011. Arrhythmogenic right ventricular cardiomyopathy/dysplasia on the basis of the revised diagnostic criteria in affected families with desmosomal mutations. Eur. Heart J. 32, 1097–1104. Raymond-Paquin, A., Nattel, S., Wakili, R., Tadros, R., 2018. Mechanisms and clinical significance of arrhythmia-induced cardiomyopathy. Can. J. Cardiol. 34, 1449–1460. Ribas, T., Pariaut, R., 2018. Inherited ventricular arrhythmias in German Shepherd dogs, in: Guide to Canine and Feline Electrocardiography. John Wiley & Sons, Ltd, Chichester, UK, pp. 315–318. Richardson, P., McKenna, R.W., Bristow, M., Maisch, B., Mautner, B., O’Connell, J., Olsen, E., Thiene, G., Goodwin, J., Gyarfas, I., Martin, I., Nordet, P., 1996. Report of the 1995 World Health Organization/ International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies. Circulation 93, 841–842. Rogers, J.H., Bolling, S.F., 2009. The tricuspid valve: current perspective and evolving management of tricuspid regurgitation. Circulation 119, 2718–2725. Saeed, M., Van, T.A., Krug, R., Hetts, S.W., Wilson, M.W., 2015. Cardiac MR imaging: current status and future direction. Cardiovasc. Diagn. Ther. 5, 290–310. Saguner, A.M., Vecchiati, A., Baldinger, S.H., Rüeger, S., Medeiros- Domingo, A., Mueller-Burri, A.S., Haegeli, L.M., Biaggi, P., Manka, R., Lüscher, T.F., Fontaine, G., Delacŕtaz, E., Jenni, R., Held, L., Brunckhorst, C., Duru, F., Tanner, F.C., Delacrétaz, E., Jenni, R.,

76 Held, L., Brunckhorst, C., Duru, F., Tanner, F.C., 2014. Different prognostic value of functional right ventricular parameters in arrhythmogenic right ventricular cardiomyopathy/dysplasia. Circ. Cardiovasc. Imaging 7, 230–239. Sahn, D.J., DeMaria, A., Kisslo, J., Weyman, A., 1978. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 58, 1072– 1083. Santilli, R.A., Bontempi, L. V., Perego, M., 2011. Ventricular tachycardia in English bulldogs with localised right ventricular outflow tract enlargement. J. Small Anim. Pract. 52, 574–580. Santilli, R.A., Perego, M., Perini, A., Moretti, P., Spadacini, G., 2010. Electrophysiologic characteristics and topographic distribution of focal atrial tachycardias in dogs. J. Vet. Intern. Med. 24, 539–545. Sargent, J., Connolly, D.J., Watts, V., Mõtsküla, P., Volk, H.A., Lamb, C.R., Luis Fuentes, V., 2015a. Assessment of mitral regurgitation in dogs: comparison of results of echocardiography with magnetic resonance imaging. J. Small Anim. Pract. 56, 641–650. Sargent, J., Muzzi, R., Mukherjee, R., Somarathne, S., Schranz, K., Stephenson, H., Connolly, D., Brodbelt, D., Fuentes, V.L., 2015b. Echocardiographic predictors of survival in dogs with myxomatous mitral valve disease. J. Vet. Cardiol. 17, 1–12. Scansen, B.A., Meurs, K.M., Spier, A.W., Koplitz, S., Baumwart, R.D., 2009. Temporal variability of ventricular arrhythmias in Boxer dogs with arrhythmogenic right ventricular cardiomyopathy. J. Vet. Intern. Med. 23, 1020–1024. Schultheiss, H.P., Fairweather, D.L., Caforio, A.L.P., Escher, F., Hershberger, R.E., Lipshultz, S.E., Liu, P.P., Matsumori, A., Mazzanti, A., McMurray, J., Priori, S.G., 2019. Dilated cardiomyopathy. Nat. Rev. Dis. Prim. 5, 1–19. Sen-Chowdhry, S., Morgan, R.D., Chambers, J.C., McKenna, W.J., 2010. Arrhythmogenic cardiomyopathy: etiology, diagnosis, and treatment. Annu. Rev. Med. 61, 233–253. Sen-Chowdhry, S., Syrris, P., McKenna, W.J., 2007. Role of genetic analysis in the management of patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy. J. Am. Coll. Cardiol. 50, 1813– 1821.

77 Shinbane, J.S., Wood, M.A., Jensen, D.N., Ellenbogen, K.A., Fitzpatrick, A.P., Scheinman, M.M., 1997. Tachycardia-induced cardiomyopathy: a review of animal models and clinical studies. J. Am. Coll. Cardiol. 29, 709–715. Sieslack, A.K., Dziallas, P., Nolte, I., Wefstaedt, P., Hungerbuhler, S.O., Hungerbühler, S.O., Hungerbuhler, S.O., 2014. Quantification of right ventricular volume in dogs: a comparative study between three- dimensional echocardiography and computed tomography with the reference method magnetic resonance imaging. BMC Vet. Res. 10, 242. Simantirakis, E.N., Koutalas, E.P., Vardas, P.E., 2012. Arrhythmia- induced cardiomyopathies: the riddle of the chicken and the egg still unanswered? Europace 14, 466–473. Simpson, S., Dunning, M.D., Brownlie, S., Patel, J., Godden, M., Cobb, M., Mongan, N.P., Rutland, C.S., 2016. Multiple genetic associations with irish wolfhound dilated cardiomyopathy. Biomed Res. Int. 2016, 1–14. Simpson, S., Edwards, J., Ferguson-Mignan, T.F.N., Cobb, M., Mongan, N.P., Rutland, C.S., 2015. Genetics of human and canine dilated cardiomyopathy. Int. J. Genomics 2015, 1–13. Smith, C.E., Freeman, L.M., Meurs, K.M., Rush, J.E., Lamb, A., 2008. Plasma fatty acid concentrations in Boxers and Doberman Pinschers. Am. J. Vet. Res. 69, 195–198. Smith, C.E., Freeman, L.M., Rush, J.E., Cunningham, S.M., Biourge, V., 2007. Omega-3 fatty acids in Boxer dogs with arrhythmogenic right ventricular cardiomyopathy. J. Vet. Intern. Med. 21, 265–273. Sossalla, S., Vollmann, D., 2018. Arrhythmia-induced cardiomyopathy. Dtsch. Aerzteblatt 115, 335–341. Spier, A.W., Meurs, K.M., 2004a. Assessment of heart rate variability in Boxers with arrhythmogenic right ventricular cardiomyopathy. J. Am. Vet. Med. Assoc. 224, 534–537. Spier, A.W., Meurs, K.M., 2004b. Evaluation of spontaneous variability in the frequency of ventricular arrhythmias in Boxers with arrhythmogenic right ventricular cardiomyopathy. J. Am. Vet. Med. Assoc. 224, 538–541.

78 Stephenson, H.M., Fonfara, S., López-Alvarez, J., Cripps, P., Dukes- McEwan, J., 2012. Screening for dilated cardiomyopathy in Great Danes in the United Kingdom. J. Vet. Intern. Med. 26, 1140–1147. Stepien, R.L., 2009. Pulmonary arterial hypertension secondary to chronic left-sided cardiac dysfunction in dogs. J. Small Anim. Pract. 50, 34–43. Stern, J.A., Meurs, K.M., Spier, A.W., Koplitz, S.L., Baumwart, R.D., 2010. Ambulatory electrocardiographic evaluation of clinically normal adult Boxers. J. Am. Vet. Med. Assoc. 236, 430–433. Stern, J.A., Ueda, Y., 2019. Inherited cardiomyopathies in veterinary medicine. Pflugers Arch. Eur. J. Physiol. 471, 745–753. Stolfo, D., Ceschia, N., Zecchin, M., De Luca, A., Gobbo, M., Barbati, G., Gigli, M., Masè, M., Pinamonti, B., Pivetta, A., Merlo, M., Sinagra, G., 2018. Arrhythmic Risk Stratification in Patients With Idiopathic Dilated Cardiomyopathy. Am. J. Cardiol. 121, 1601–1609. Sugeng, L., Mor-Avi, V., Weinert, L., Niel, J., Ebner, C., Steringer- Mascherbauer, R., Bartolles, R., Baumann, R., Schummers, G., Lang, R.M., Nesser, H.J., 2010. Multimodality comparison of quantitative volumetric analysis of the right ventricle. JACC Cardiovasc. Imaging 3, 10–18. Takemura, N., Nakagawa, K., Hirose, H., 2002. Lone atrial fibrillation in a dog. J. Vet. Med. Sci. 64, 1057–1059. Tarducci, A., Borgarelli, M., Zanatta, R., Cagnasso, A., 2003. Asymptomatic dilated cardiomyopathy in Great Danes: clinical, electrocardiographic, echocardiographic and echo-Doppler features. Vet. Res. Commun. 27, 799–802. Thavendiranathan, P., Phelan, D., Thomas, J.D., Flamm, S.D., Marwick, T.H., 2012. Quantitative assessment of mitral regurgitation: Validation of new methods. J. Am. Coll. Cardiol. 60, 1470–1483. Thomas, L., Foster, E., Schiller, N.B., 1998. Peak mitral inflow velocity predicts mitral regurgitation severity. J. Am. Coll. Cardiol. 31, 174– 179. Thomas, W.P., Gaber, C.E., Jacobs, G.J., Kaplan, P.M., Lombard, C.W., Vet, M., Moise, N.S., Moses, B.L., 1993. Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. J. Vet. Intern. Med. 7, 247–252.

79 Thomason, J.D., Kraus, M.S., Surdyk, K.K., Fallaw, T., Calvert, C.A., 2008. Bradycardia-associated syncope in 7 boxers with ventricular tachycardia (2002–2005). J. Vet. Intern. Med. 22, 931–936. Tidholm, A., Häggström, J., Borgarelli, M., Tarducci, A., 2001. Canine idiopathic dilated cardiomyopathy. Part I: aetiology, clinical characteristics, epidemiology and pathology. Vet. J. 162, 92–107. Tidholm, A., Jönsson, L., 1997. A retrospective study of canine dilated cardiomyopathy (189 cases). J. Am. Anim. Hosp. Assoc. 33, 544–550. Tidholm, A., Jönsson, L., 1996. Dilated cardiomyopathy in the Newfoundland: a study of 37 cases (1983–1994). J. Am. Anim. Hosp. Assoc. 32, 465–470. Topilsky, Y., Nkomo, V.T., Vatury, O., Michelena, H.I., Letourneau, T., Suri, R.M., Pislaru, S., Park, S., Mahoney, D.W., Biner, S., Enriquez- Sarano, M., 2014. Clinical outcome of isolated tricuspid regurgitation. JACC Cardiovasc. Imaging 7, 1185–1194. Towbin, J.A., McKenna, W.J., Abrams, D.J., Ackerman, M.J., Calkins, H., Darrieux, F.C.C., Daubert, J.P., de Chillou, C., DePasquale, E.C., Desai, M.Y., Estes, N.A.M., Hua, W., Indik, J.H., Ingles, J., James, C.A., John, R.M., Judge, D.P., Keegan, R., Krahn, A.D., Link, M.S., Marcus, F.I., McLeod, C.J., Mestroni, L., Priori, S.G., Saffitz, J.E., Sanatani, S., Shimizu, W., van Tintelen, J.P., Wilde, A.A.M., Zareba, W., 2019. 2019 HRS expert consensus statement on evaluation, risk stratification, and management of arrhythmogenic cardiomyopathy: Executive summary. Heart Rhythm 16, e373–e407. Van Gelder, I.C., Rienstra, M., Crijns, H.J.G.M., Olshansky, B., 2016. Rate control in atrial fibrillation. Lancet 388, 818–828. Vezzosi, T., Tognetti, R., Buralli, C., Marchesotti, F., Patata, V., Zini, E., Domenech, O., 2019. Home monitoring of heart rate and heart rhythm with a smartphone-based ECG in dogs. Vet. Rec. 184, 96. Vila, J., Pariaut, R., Moïse, N.S., Oxford, E.M., Fox, P.R., Reynolds, C.A., Saelinger, C., 2017. Structural and molecular pathology of the atrium in boxer arrhythmogenic right ventricular cardiomyopathy. J. Vet. Cardiol. 19, 57–67. Vischer, A.S., Connolly, D.J., Coats, C.J., Fuentes, V.L., McKenna, W.J., Castelletti, S., Pantazis, A.A., 2017. Arrhythmogenic right ventricular cardiomyopathy in Boxer dogs: the diagnosis as a link to the human disease. Acta Myol. 36, 135–150.

80 Visser, L.C., 2017. Right Ventricular Function. Vet. Clin. North Am. Small Anim. Pract. 47, 989–1003. Visser, L.C., Scansen, B.A., Schober, K.E., 2013. Single left coronary ostium and an anomalous prepulmonic right coronary artery in 2 dogs with congenital pulmonary valve stenosis. J. Vet. Cardiol. 15, 161–169. Visser, L.C., Scansen, B.A., Schober, K.E., Bonagura, J.D., 2015. Echocardiographic assessment of right ventricular systolic function in conscious healthy dogs: repeatability and reference intervals. J. Vet. Cardiol. 17, 83–96. Vizzardi, E., D’Aloia, A., Bordonali, T., Bugatti, S., Piovanelli, B., Bonadei, I., Quinzani, F., Rovetta, R., Vaccari, A., Curnis, A., Dei Cas, L., 2012. Long-term prognostic value of the right ventricular myocardial performance index compared to other indexes of right ventricular function in patients with moderate chronic heart failure. Echocardiography 29, 773–778. Vollmar, A.C., 2000. The prevalence of cardiomyopathy in the Irish Wolfhound: a clinical study of 500 dogs. J. Am. Anim. Hosp. Assoc. 36, 125–132. Werner, P., Raducha, M.G., Prociuk, U., Sleeper, M.M., Van Winkle, T.J., Henthorn, P.S., 2008. A novel locus for dilated cardiomyopathy maps to canine chromosome 8. Genomics 91, 517–521. Wess, G., Domenech, O., Dukes-McEwan, J., Häggström, J., Gordon, S., 2017. European Society of Veterinary Cardiology screening guidelines for dilated cardiomyopathy in Doberman Pinschers. J. Vet. Cardiol. 19, 405–415. Wess, G., Schulze, A., Butz, V., Simak, J., Killich, M., Keller, L.J.M., Maeurer, J., Hartmann, K., 2010. Prevalence of dilated cardiomyopathy in Doberman Pinschers in various age groups. J. Vet. Intern. Med. 24, 533–538. Westphal, J.G., Rigopoulos, A.G., Bakogiannis, C., Ludwig, S.E., Mavrogeni, S., Bigalke, B., Doenst, T., Pauschinger, M., Tschöpe, C., Schulze, P.C., Noutsias, M., 2017. The MOGE(S) classification for cardiomyopathies: current status and future outlook. Heart Fail. Rev. 22, 743–752.

81 WHO/ISFC task force, 1980. Report of the WHO/ISFC task force on the definition and classification of cardiomyopathies. Br. Heart J. 44, 672–673. Wyse, D.G., Van Gelder, I.C., Ellinor, P.T., Go, A.S., Kalman, J.M., Narayan, S.M., Nattel, S., Schotten, U., Rienstra, M., 2014. Lone atrial fibrillation: does it exist? J. Am. Coll. Cardiol. 63, 1715–1723. Zaman, S., Goldberger, J.J., Kovoor, P., 2019. Sudden death risk- stratification in 2018–2019: the old and the new. Heart. Lung Circ. 28, 57–64. Zoghbi, W.A., Enriquez-Sarano, M., Foster, E., Grayburn, P.A., Kraft, C.D., Levine, R.A., Nihoyannopoulos, P., Otto, C.M., Quinones, M.A., Rakowski, H., Stewart, W.J., Waggoner, A., Weissman, N.J., 2003. Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J. Am. Soc. Echocardiogr. 16, 777–802. Zupan, I., Rakovec, P., Budihna, N., Brecelj, A., Koželj, M., 1996. Tachycardia induced cardiomyopathy in dogs; relation between chronic supraventricular and chronic ventricular tachycardia. Int. J. Cardiol. 56, 75–81.

82 9. SUMMARY IN ESTONIAN

Elektro- ja piltdiagnostika täiendavad rakendused koerte südamehaiguste diagnoosimisel ning prognoosimisel

Kokkuvõte

Kardiomüopaatia ehk südamelihase mittepõletikuline haigus on talitlushäiretega kulgevate südame lihaskesta haiguste üldnimetus.

Maailma Terviseorganisatsiooni (World Health Organisation, WHO) 1980. aasta klassifikatsiooni kohaselt jaotatakse kardiomüopaatia morfoloogiliste tunnuste alusel dilateeruvaks (dilated cardiomyopathy, DCM), hüpertroofiliseks hypertrophic ( cardiomyopathy, HCM) ja restriktiivseks (restrictive cardiomyopathy, RCM) kardiomüopaatiaks. Seda klassifikatsiooni täiendati 1995. aastal, kui lisandusid arütmogeenne parema vatsakese kardiomüopaatia (arrhythmogenic right ventricular cardiomyopathy, ARVC) ja klassifitseerimata kardiomüopaatia (unclassified cardiomyopathy, UCM). Täiendatud klassifikatsiooni lisati ka mõiste „spetsiifilised kardiomüopaatiad“, et kirjeldada südamelihase muutusi, mis tekivad seoses spetsiifiliste südame- või süsteemsete haigustega.

Arütmiast tingitud kardiomüopaatia (arrhythmia induced cardiomyopathy, AIC) on patoloogia, mille korral tekivad tahhüarütmiate e südamerutluse tagajärjel südame remodelleerumise (südame suuruse, kuju ning funktsiooni muutuse) ja süstoolse ning diastoolse talitluse häirumisega kulgevad muutused. AIC kohta kasutatakse kirjanduses tihti ka terminit tahhükardiast tingitud kardiomüopaatia. AIC täpne patofüsioloogiline tekkemehhanism on ebaselge, kuid südamelihases tekkinud muutused on neid esile kutsunud tahhüarütmia eduka ravi korral enamasti pöörduvad. Inimestel on kõige sagedamad AIC-d põhjustavad rütmihäired kodade virvendusarütmia, kodade fokaalne tahhükardia ja sagedad vatsakeste ekstrasüstolid (ventricular premature contractions, VPC).

Kardiomüopaatia on üks sagedasemaid koerte haigestumuse ja suremuse põhjuseid. Selle levimus sõltub suurel määral kardiomüopaatia tüübist ja koera tõust. DCM-i levimus võib esmatasandi loomakliiniku patsientide hulgas olla hinnanguliselt 0,05%. Erinevate uuringute alusel on levimus iiri hundikoertel 24%, taani dogidel 36–47% ning dobermannidel 58%.

83 ARVC-d on kirjeldatud eeskätt bokseritel ning üksikute haigusjuhtumite või väikeste haiguslugude seeriatena ka teistel tõugudel. Selle kardiomüopaatia liigi levimust nii bokseritel kui ka teistel tõugudel on raske täpselt hinnata, sest puuduvad ühesed diagnostilised kriteeriumid, haiguse diagnoosimine looma eluajal on keeruline ja haiguse kliiniline kulg on väga muutlik.

Loomade ja inimeste kardiomüopaatia tunnused ja muutused (etioloogia, histoloogia, diagnostika, patofüsioloogia, prognostika jne) on väga sarnased, mistõttu on võimalik loomade haigusi kasutada mudelina inimeste vastavate haiguste käsitlemisel. Kasside ja koerte kardiomüopaatiat on kirjeldatud inimeste vastava haiguse mudelina.

Käesoleva uurimistöö raames uuriti koerte kardiomüopaatia mitmesuguseid diagnostilisi ja prognostilisi aspekte. I ja II artiklis keskenduti prognostilisele teabele, mida on võimalik saada üldlevinud elektrokardiograafilistest I( ) ja ehhokardiograafilistest II( ) uuringutest, mida kasutatakse bokserite ARVC diagnostikas. III artikkel kirjeldas südamepatoloogiaid ühel teisel suure südamehaiguste levimusega tõul bordoo dogil. ARVC elupuhune diagnostika on raskendatud, sest see eeldab südamelihase histoloogilist uurimist. Olgugi et südamelihase biopsia võtmine on tehniliselt võimalik, ei ole see tehniliste probleemide ja potentsiaalselt eluohtlike tüsistuste tõttu tavapraktikas rakendatav meetod. Humaanmeditsiinis on kardiomüopaatia, eeskätt ARVC diagnostikas väga tähtis osa südame magnetresonantstomograafial (MRT). MRT-d on veterinaarkardioloogias suhteliselt vähe uuritud, sest see eeldab nii kalli aparatuuri olemasolu kui ka patsiendi anesteseerimist uuringu tegemiseks. IV artikli eesmärk oli anda täiendavaid teadmisi MRT rakendamisel veterinaarkardioloogias. Täpsemalt võrreldi selles artiklis ultraheliuuringul ja MRT-uuringul saadud südame alaosade mahtude mõõtmistulemusi.

Töö eesmärgid

Käesoleva töö peamine eesmärk oli anda täiendavat teavet kardiomüopaatia, eeskätt ARVC diagnostika ja prognostika kohta.

Väitekirja kitsamad eesmärgid olid alljärgnevad.

84 • Hinnata Holteri uuringu prognostilist väärtust ARVC kahtlusega bokseri tõugu koertel (I). • Selgitada TAPSE (tricuspid annular plane systolic excursion, eesti keeles parema koja-vatsakese klapi fibroosrõnga liikumise ulatus süstoli ajal) kliinilist rakendust minestamishoogude e sünkoopide või ARVC kahtlusega bokseri tõugu koertel (II). • Kirjeldada ehho- ja elektrokardiograafilisi leide kardioloogilisele eriarsti vastuvõtule suunatud bordoo dogidel (III). • Uurida südame MRT-uuringu rakendatavust mitraalklapi e vasaku koja-vatsakese klapi regurgitatsiooni e tagasivoolu kvantitatiivseks hindamiseks väikest tõugu koertel ja tuvastada südame ultraheliuuringu käigus mõõdetavad näitajad, mis korreleeruvad kõige paremini südame MRT-uuringul mõõdetud mitraalklapi regurgitatsiooni ulatusega (IV).

Uurimismetoodika ja -tulemused

Väitekirja esimeses uuringus (I) analüüsiti retrospektiivselt Holteri ja südame ultraheliuuringu tulemusi bokseri tõugu koertel, kes olid saadetud täiendavale kardioloogilisele eriarsti vastuvõtule kolme erinevasse teise tasandi kliinikusse seoses esmatasandi arsti poolt täheldatud südamehaigusele viitavate kliiniliste tunnuste või kliinilise läbivaatuse leiuga (n = 163). Kaasasündinud südamehaigustega või muude potentsiaalselt elu lühendavate mitte-südamehaigustega patsiendid, samuti ebapiisava andmehulgaga patsiendid jäeti analüüsist välja (n = 41). Analüüsi kaasati 122 koera andmed. Uuringupopulatsioonis oli isaseid ja emaseid koeri enam-vähem võrdselt (52% isased) ja kogu uuringupopulatsiooni mediaanvanus oli 6,9 aastat. Populatsioonis täheldati vanusejaotuse kaht sagedustõusu, mille tipud olid ühe ja kaheksa eluaasta juures. Noori koeri saadeti märkimisväärselt sagedamini (P = 0,040) minestushoogude uurimiseks edasi. Kuna südamepuudulikkuse tunnustega ja oluliselt halvenenud süstoolse talitlusega patsiente on teiste uuringute, eeskätt ultraheliuuringuga võimalik lihtsalt tuvastada, keskenduti elumust mõjutavate tegurite analüüsil normaalse süstoolse talitlusega ja ilma südamepuudulikkuse tunnusteta patsientide [LVIDs≤35 mm rühm; n = 91; LVIDs (left ventricular internal diameter in systole, eesti keeles vasaku vatsakese süstoolne siseläbimõõt)] andmetele. Selles rühmas oli keskmine elumus oluliselt pikem kogu uuringupopulatsiooni elumusest (> 2083 vs 242 päeva, P < 0,001). Sünkoobi esinemine ei olnud seotud lühema elumusega. Statistiliselt oluliselt lühem elumus seondus selles

85 rühmas sooga (isasloomadel lühem), > 50 VPC esinemisega ööpäevas, polümorfsete VPC-de esinemisega ja vatsakeste tahhükardia esinemisega (vastavalt P = 0,035, P = 0,003, P = 0,019, P = 0,002). Analüüsides elumusandmeid vanuserühmade kaupa, esines noortel loomadel harvem > 50 VPC-d ööpäevas (P < 0,001). Vanematel koertel esines hoopis sagedamini vatsakeste tahhükardiat (P = 0,015) ja teisi kompleksseid vatsakeste rütmihäireid. Kuna elumus oli noorimal vanusekolmandikul pikem kui kahel vanemal kolmandikul (P < 0,001), kasutati vanusepiiri 4,5 aastat vanuserühmade eristamiseks mitme muutujaga statistilises analüüsis. Tuginedes mitme muutujaga Coxi regressiooni riskitegurite analüüsi tulemustele, olid üksnes vanus > 4,5 aasta (P = 0,003), vatsakeste tahhükardia esinemine (P = 0,008) ja isane sugu (P = 0,009) lühema elumuse sõltumatud ennustajad.

Väitekirja teise uuringusse (II) kaasati patsiendid samasugustel alustel – bokseri tõugu koerad, kelle puhul kardioloogilisele eriarsti vastuvõtule suunamise põhjus oli südamehaigusele viitavate anamneesiandmete või kliiniliste leidude esinemine. Esialgu tuvastatud 81 koerast arvati 31 analüüsist välja südame ultraheliuuringu andmete vähesuse tõttu. Seega analüüsiti 50 koera andmeid, kellest 78% oli varem täheldatud sünkoopi ning 14% südame paispuudulikkust. Koerte mediaanvanus oli 6 aastat ja neist pooled olid isased. Analüüsiks jaotati koerad kahte rühma – ARVC kahtlusega, kellel Holteri EKG-l oli ≥ 50 VPC / 24h, n = 28; ja ARVC kahtluseta (< 50 VPC / 24 h, n = 22). ARVC rühmas olevad koerad olid tunduvalt vanemad kui mitte-ARVC rühmas (7,7 vs. 4,9 aastat, P = 0,046). TAPSE oli märgatavalt väiksem ARVC rühmas (13,9 ± 4,04 mm vs. 16,8 ± 3,21 mm, P = 0,008). Veel oli TAPSE kergelt kuni mõõdukalt seotud vasaku vatsakese väljutusfraktsiooni (r = 0,354, P = 0,017) ja lühenemisfraktsiooniga (r = 0,424, P = 0,002) ning negatiivselt seotud vasaku koja suurusega (r = -0,448, P = 0,002). Mitme muutujaga analüüsiks koostati neli mudelit, hindamaks TAPSE < 15,1 mm seost lühema elumusega, kaasates mudelitesse paispuudulikkuse esinemise, süstoolse talitlushäire, Holteri uuringul tuvastatud vatsakeste tahhükardia esinemise, ja ARVC diagnoosi. Kõikides mudelites oli TAPSE < 15,1 mm seotud oluliselt lühema elumusega (vastavalt P = 0,02, P = 0,023, P = 0,023, P = 0,029).

Väitekirja kolmandasse uuringusse (III) kaasati bordoo dogid, kes olid suunatud kardioloogilisele eriarsti vastuvõtule seoses südamehaigusele viitavate anamneesiandmete või kliinilise läbivaatuse leiuga. Neile oli

86 tehtud südame ultraheliuuring ja salvestatud oli ka 12-lülituseline EKG või 24-tunnine Holteri EKG (n = 64). Koerte mediaanvanus oli 4,3 aastat ja 77% koertest olid isased. Peamised kliinilised leiud, mille uurimiseks patsiendid edasi suunati, olid arütmia (52%), koormustaluvuse vähenemine (41%), süstoolne südamekahin (33%), kollaps (31%) ja kõhu suurenemine (25%). Südamega mitteseonduv ebanormaalne koemass oli diagnoositud kuuel koeral. Kliinilisel läbivaatusel täheldati galoppheli ühel koeral. Ultraheliuuringul täheldati südamega seonduvat koemassi seitsmel ja kaasasündinud südamehaigusi 17 koeral. Kõige sagedam kaasasündinud südamehaigus oli aordiklapialune kitsenemus (n = 12). Avatud arterioosjuha diagnoositi kahel koeral ja lisaks diagnoositi kopsutüve suistiku kitsenemus, vatsakeste vaheseina defekt ning kodade vaheseina defekt. Kogu uuringupopulatsioonis esines supraventrikulaarseid rütmihäireid 47% patsientidest – püsiv kodade virvendusarütmia 39%, supraventrikulaarne tahhükardia 8%. Holteri EKG oli salvestatud 25% (16/64) patsientidest, kellest üheksal (56%) esines kodade virvendusarütmia. Rütmihäirete esinemissageduse statistiline analüüs viidi läbi koertel, kellel ei tuvastatud südamega mitteseonduvaid uudismoodustisi, ultraheliuuringul täheldatavaid südamega seonduvaid koemasse ega kaasasündinud südamehaigusi (mitte-CHDN-i rühm, n = 34; CHDN – congenital heart disease and neoplasia). Neist loomadest 56% esines kodade virvendust ja 9% supraventrikulaarset tahhükardiat. Kodade virvendusarütmia esinemisel oli südamelöökide mediaansagedus 200 korda minutis ja 89% kodade virvendusega patsientidest oli südame löögisagedus > 160 korda minutis. Ühel kodade virvendusega koeral ei täheldatud südames struktuurseid muutusi ja tema südame löögisagedus oli < 160 korda minutis. Ventrikulaarseid rütmihäireid täheldati 23% kõigist (15/64) koertest. Kaheksal koeral esines üksikuid VPC-sid ning seitsmel kompleksseid VPC-sid. 87% vatsakeste rütmihäiretega patsientidest (13/15) oli salvestatud ka Holteri EKG ja mediaanväärtusena esines neil 105 VPC-d 24 tunni jooksul. Isoleeritud VPC-dega koertest 89% esines < 100 VPC-d 24 tunni jooksul. Südame ultraheliuuringul täheldati mitte-CHDN-i rühma koertest (n = 34) 53% vasaku vatsakese süstoolset talitlushäiret ja 53% parema koja või vatsakese laienemist; 13 koeral (38%) esines korraga nii parempoolne südame laienemine kui ka vasaku vatsakese talitlushäire.

Väitekirja neljandasse uuringusse (IV) kaasati tüüpilised mitraalklapi müksomatoosse degeneratsiooniga patsiendid kehamassiga ≤ 20 kg, kelle peamised kliinikusse suunamise põhjused olid närvisüsteemi

87 haigused, mille uuringuks kasutati MRT-uuringut. Uuringusse kaasati 15 Cavalier King Charlesi spanjelit ning üks koer igast järgnevast tõust: whippet, borderkolli, beagle, shih tzu, bichon frise ja kääbuspuudel. Patsientide mediaanvanus oli 7,7 aastat ja sooline jaotus oli praktiliselt võrdne (52% isasloomi). Kõigil loomadel täheldati süstoolset südamekahinat tugevusega I / VI–V / VI. Mitraalklapi müksomatoosse degeneratsiooni diagnoosi alus oli süstoolse südamekahina olemasolu ning tüüpiliste kahedimensionaalsete või Doppleri värviuuringu ehhokardiograafiliste leidude esinemine. Uuringusse kaasati 21 koera, kellest üheksal ei õnnestunud protokollis sätestatud aja jooksul (50 min) täieliku ja piisava kvaliteediga südame MRT-uuring.

Uuringu käigus hinnati südame MRT käigus mõõdetud mitraalklapi regurgitatsioonifraktsiooni seost järgmiste ehhokardiograafiliste mõõtmiste tulemustega: vena contracta (kitsaim tagasivoolujoa piirkond vahetult peale lekkiva klapi suuet) läbimõõdu suhe aordi läbimõõtu, varajase transmitraalvoolu kiirus, mitraalklapi prolapsi suhe aordi läbimõõtu, vasaku vatsakese diastoolse läbimõõdu suhe aordi läbimõõtu, vasaku koja läbimõõdu suhe aordi läbimõõtu, mitraalklapi regurgitatsioonijoa pindala suhe vasaku koja pindalasse ning proksimaalse samakiirus ala (proximal isovelocity surface area, PISA) pindala alusel arvutatud regurgitatsioonifraktsioon. Südame ultraheliuuring sooritati kõigil patsientidel esmalt ilma anesteesiata/sedatsioonita ning seda korrati pärast südame MRT-uuringu tegemist anesteseeritud patsiendil. Südame MRT-uuringuks vajalik maksimaalne lubatav lisaanesteesia aeg oli 50 minutit. Südame MRT-uuringu detailse tehnilise kirjelduse võib leida IV artiklist. Südame ultraheliuuringul tehtud mõõtmiste tulemustes anesteseeritud ja mitteanesteseeritud patsientidel olnud olulist erinevust ei leitud. Kahe erineva uurija mõõtmiste tulemuste erinevus väljendatuna rühmade-sisese korrelatsioonikordaja ning 95% usaldusvahemikuna (95% UV) oli kõige väiksem vasaku vatsakese diastoolse siseläbimõõdu (0,98, 95% UV 0,91–0,99) ning varajase transmitraalvoolu kiiruse korral (0,96, 95% UV 0,88–0,99), veidi suurem vasaku koja läbimõõdu ja aordi läbimõõdu suhte korral (0,86, 95% UV 0,52–0,96) ning kõige suurem mitraalklapi prolapsi (0,75, 95% UV 0,31–0,92) hindamisel. Doppleri värviuuringu puhul oli kahe uurija tulemuste erinevus kõige väiksem vena contracta läbimõõdu korral (0,89, 95% UV 0,61–0,98), mõnevõrra suurem PISA (0,52, 95% UV 0,00–0,88) ja kõige suurem mitraalklapi regurgitatsioonijoa pindala ja vasaku koja pindala suhte (0,34, 95% UV 0,00–0,74) puhul. Südame MRT-uuringu käigus salvestatud kujutisi

88 hindasid kaks erinevat uurijat ning tunnistasid vasaku vatsakese löögimahu ja aordi löögimahu korratavuse väga heaks (≥ 0,99).

Täielik südame MRT- ja ehhokardiograafia uuringu andmehulk saadi kümne koera kohta, kelle andmed kaasati lõplikku kahe meetodi võrdlevasse analüüsi. Seoseid südame MRT käigus mõõdetud regurgitatsioonifraktsiooni ja ultraheliuuringu käigus mõõdetud näitajatega hinnati Pearsoni korrelatsioonikoefitsiendi (r) abil. Kõige olulisem seos südame MRT käigus mõõdetud regurgitatsioonifraktsiooniga leiti vena contracta ja aordi läbimõõdu suhte (r = 0,89, 95% UV 0,60–0,97, P = 0,001) ning varajase transmitraalvoolu kiiruse korral (r = 0,86, 95% UV 0,52–0,97, P = 0,001). Mitme mitraalregurgitatsioonijoa samaaegne esinemine muutis vena contracta mõõtmise ühel koeral võimatuks. Vasaku koja läbimõõdu ja aordi läbimõõdu suhe (r = 0,78, 95% UV 0,30–0,94, P = 0,008) ning mitraalklapi prolapsi ja aordi läbimõõdu suhe (r = 0,72, 95% UV 0,16–0,93, P = 0,02) MRT käigus mõõdetud regurgitatsioonifraktsiooni oli nõrgem. PISA meetodil arvutatud ja südame MRT abil mõõdetud regurgitatsioonifraktsiooni seos oli mõõdukas (r = 0,75, 95% UV 0,09–0,95, P = 0,03), kuid see meetod ei olnud rakendatav kahel koeral seoses mitme regurgitatsioonijoa esinemise või selle väiksusega.

Järeldused

• Holteri monitooring annab väärtuslikku prognostilist teavet ARVC kahtlusega bokseri tõugu koertel. Vatsakeste tahhükardia esinemine, kõrgem iga ja isassugu on seotud lühema elumusega. Alla 4,5 aasta vanustel bokseri tõugu koertel, kelle südames ei täheldata ultraheliuuringul patoloogilisi muutusi ja kellel esineb vähem kui 50 VPC-d 24 tunni jooksul, on pikaajaline prognoos hea. • TAPSE annab väärtuslikku prognostilist teavet minestamisepisoodidega või ARVC kahtlusega bokseri tõugu koerte käsitlemisel. TAPSE < 15,1 mm seostub suurema südamehaigustest põhjustatud suremusega bokseri tõugu koertel, kellel on ööpäevas ≥ 50 VPC-d sõltumata sellest, kas neil esineb süstoolse talitluse häire, südame paispuudulikkus või vatsakeste tahhükardia. • Bordoo dogidel on levinud südamehaigused ja südamega seonduvad koemassid. Kõige sagedamaks kaasasündinud südamehaiguseks uuritud patsientidel oli aordiklapi(alune) kitsenemus. Uuritud tõul on levinud supraventrikulaarsed rütmihäired, eeskätt kodade

89 virvendusarütmia. Kodade virvendusarütmia oli seotud südame remodelleerumise, eeskätt parema koja või vatsakese laienemise või halvenenud süstoolse talitlusega. • Vaatamata sellele, et südame MRT on tehniliselt keeruline, on see siiski meetod, mida saab mitraalklapi regurgitatsiooni hindamisel rakendada. Varajase transmitraalvoolu ja vena contracta läbimõõdu ja aordi läbimõõdu suhe korreleerusid kõige paremini südame MRT- uuringul saadud mõõtmistulemustega, mis annab võimaluse nende uurimismeetodite kasutamisel saadavate tulemuste võrdlevaks rakendamiseks.

90 10. ACKNOWLEDGEMENTS

I would like to thank my supervisors Professor Toomas Orro, Professor Virginia Luis Fuentes, Professor David Connolly and Doctor Ranno Viitmaa for their support, input and encouragement for completion of my thesis.

I would like to thank my senior clinical training scholarship supervisors from the Royal Veterinary College, University of London: Professor David John Connolly, Professor Virginia Luis Fuentes and Professor Adrian Boswood, for all the teaching, advice and support leading to my board certification in veterinary cardiology as well as their guidance and invaluable input to my research publications.

I would like to thank all the other co-authors of the publications that are the basis of this thesis: Dr. Chris Linney, Dr. Valentina Palermo, Prof. Anne French, Prof. Joanna Dukes McEwan, Dr. Benjamin Kaye, Dr. Kieran Borgeat, Dr. Gavin McAulay, Dr. Julia Sargent, Dr. João Neves, Ms. Victoria Watts, Prof. Holger Volk and Prof. Christopher Lamb. Also, I would like to thank Dr. Yu-Mei Chang for her invaluable help with the statistical analysis.

I would like to thank Prof. Christian Burvenich from the Department of Nutrition, Genetics and Ethology, University of Ghent and Prof. David Richard Arney from the Chair of Animal Nutrition, Estonian University of Life Sciences for their constructive criticism and invaluable feedback during the final stages of the preparation of this thesis.

I would like to thank my colleagues on clinics for their support with the clinical work and taking over many of my tasks, thus helping me to make time for writing. Also, I would like to thank all my students for their inspiration and constant challenging that has helped me to become a better teacher.

Lastly, I would like to express my sincere gratitude and appreciation to my family and friends for their kindness and support during this interesting albeit at times difficult journey. It would not have been possible to accomplish this thesis without your pat on the back!

ORIGINAL PUBLICATIONS Mõtsküla PF, Linney C, Palermo V, Connolly DJ, French A, Dukes McEwan J, Luis Fuentes V., 2013. PROGNOSTIC VALUE OF 24-HOUR AMBULATORY ECG (HOLTER) MONITORING IN BOXER DOGS. Journal of Veterinary Internal Medicine 27, 904–912. J Vet Intern Med 2013;27:904–912

Prognostic Value of 24-Hour Ambulatory ECG (Holter) Monitoring in Boxer Dogs

P.F. Mõtskula,€ C. Linney, V. Palermo, D.J. Connolly, A. French, J. Dukes McEwan, and V. Luis Fuentes

Background: Boxer dogs are reported to be predisposed to arrhythmogenic right ventricular cardiomyopathy (ARVC), but the natural history has not been well characterized and inconsistent diagnostic criteria have been applied to identify affected dogs. Echocardiographic examination findings are unremarkable in many affected Boxer dogs, and in these dogs, 24-hour ambulatory ECG (Holter) monitoring often is used for diagnostic and prognostic purposes, despite limited information available relating Holter findings to outcome. Hypothesis/Objectives: Boxers with complex ventricular arrhythmias at initial presentation will have shorter survival times. The objective was to investigate the prognostic value of Holter monitoring in Boxer dogs. Animals: One hundred and twenty-two Boxer dogs seen at 3 university referral hospitals. Methods: Retrospective study. Survival times were obtained for Boxer dogs evaluated by echocardiography and a 24-hour Holter ECG. Kaplan-Meier survival analysis was used to estimate the median survival time and Cox proportional hazards analysis was used to identify variables independently associated with cardiac mortality. Results: Outcome data were obtained for 122/163 dogs meeting the inclusion criteria. Of the 70 dogs that had died, 45 were considered to have suffered cardiac-related deaths. Median survival was significantly longer in dogs with a left ventricular systolic diameter (LVIDs) 35 mm compared with those with LVIDs > 35 mm (P < .001). Multivariable analysis in dogs with LVIDs 35 mm showed that the presence of ventricular tachycardia, age >4.5 years, and male sex were independent predictors of cardiac mortality. Conclusions and Clinical Importance: Holter monitoring in Boxer dogs provides valuable prognostic information. Key words: Electrocardiography; VPC; Dog; ARVC.

rrhythmogenic right ventricular cardiomyopathy A(ARVC) has been reported in Boxer dogs, but the Abbreviations: diagnostic criteria and natural history of ARVC in this 2D 2-dimensional 1–5 breed remain ill-defined. A definitive diagnosis usu- Ao aortic root dimension ally is based on necropsy findings of fibrofatty replace- Aovel aortic flow velocity ment of the right ventricular myocardium.1–6 The APC atrial premature contraction antemortem diagnosis commonly is based on the pres- ARVC arrhythmogenic right ventricular cardiomyopathy ence of an arbitrary number of ventricular premature CHF congestive heart failure complexes (VPC) in a 24-hour Holter recording. Many CI confidence interval cutoff values have been proposed for the number of ECG electrocardiography VPC/24 h to differentiate between affected and nonaf- FS% fractional shortening fected individuals (eg, 50,4 100,6–9 1,0003,10–12). Syn- HR hazard ratio cope is one of the most common clinical signs in IQR interquartile range affected Boxers, but syncope also can be associated LA left atrial dimension with other conditions such as aortic stenosis13 or neu- LA/Ao left atrial to aortic root ratio rally mediated syncope.14,15 LV left ventricle LVIDd left ventricular internal diameter in diastole Echocardiographic abnormalities are present in LVIDs left ventricular internal diameter in systole some affected dogs, and the association with these and SVT supraventricular tachycardia prognosis in Boxer dogs diagnosed with ARVC is well VPC ventricular premature contraction recognized. Palermo et al showed that Boxers diag- VT ventricular tachycardia nosed with ARVC and systolic dysfunction (defined as

left ventricular internal diameter in systole >35 mm) From the The Royal Veterinary College, London, UK (Mõtskula,€ had shorter survival times than Boxers with ARVC Connolly, Luis Fuentes); the School of Veterinary Sciences, and normal systolic function.3 In the same study, the University of Liverpool, Neston, UK (Linney, Dukes McEwan); and and the Royal (Dick) School of Veterinary Studies, University of presence of syncope and increased frequency and com- Edinburgh, Edinburgh, UK (Palermo, French). The results of part of plexity of ventricular arrhythmias also were associated this study were presented in ACVIM forum 2012, New Orleans, LA. with shorter survival times. In another study of 64 Corresponding author: P.F. Mõtskula,€ DVM, MSc, MRCVS, Boxers with ARVC, Meurs et al showed that the pres- The Royal Veterinary College, Hawkshead Lane, North Mymms, ence of congestive heart failure (CHF) or poor systolic Hertfordshire AL9 7TA, UK; e-mail: [email protected]. function, but not the presence of syncope, was associ- Submitted December 4, 2012; Revised March 27, 2013; ated with shorter survival times.1 These studies elected Accepted April 11, 2013. Copyright © 2013 by the American College of Veterinary Internal to limit their analysis to dogs diagnosed with ARVC Medicine based on Holter-based diagnostic criteria. Similar stud- 10.1111/jvim.12107 ies in a wider population of Boxers would be more

95 Holter Monitoring in Boxer Dogs 905 applicable to Boxers seen in general practice, but have the definitions presented in the Table 1. The number of arrhyth- not been reported. mia episodes was normalized to a 24-hour period. We hypothesized that in Boxers with normal systolic Survival information was obtained from a range of sources as function, cardiac mortality would be increased in those necessary, starting with the medical records, a questionnaire with complex ventricular arrhythmias at initial presen- mailed to owners, and finally telephone follow-up with owners who did not respond to the mailed questionnaire. If survival data tation compared with those without. The aim of this were still incomplete, referring veterinarians were contacted. Out- study was to investigate the prognostic value of Holter come data included patient status (dead/alive), date of death or monitoring in Boxers presenting to a referral institu- euthanasia, and whether the death was caused by deterioration tion and in the subpopulation of dogs with normal of a cardiac condition. Sudden cardiac death was defined as an systolic function. unexpected death without apparent clinical signs during the preceding 24-hour period.1 Dogs that were still alive were right- Materials and Methods censored at the time of last veterinary contact or the date of the interview with the owner. The Ethics and Welfare Committees of the participating insti- Statistical analysis was performed by commercially available tutions approved the protocol used in this study. statistical software.g-i Significance was set at the 5% level for all For this multicenter study, we reviewed the medical records tests, apart from the univariable Cox proportional hazards analy- from 2004 to 2011 of all Boxer dogs undergoing Holter monitor- sis where variables with a P value of <.2 were included in the ing and echocardiographic examination at 3 referral centers: the multivariable model. Continuous data were expressed as median Queen Mother Hospital for Animals, The Royal Veterinary and interquartile range (IQR). Binomial and categorical data College; the Small Animal Teaching Hospital, University of were expressed as number and percentage. Categorical data were Liverpool; and the Hospital for Small Animals, Royal (Dick) assessed by the Fisher’s exact test. The Mann-Whitney U-test School of Veterinary Studies. All dogs were assessed either by a was used to compare nonnormally distributed data. Hartigan’s board-certified cardiologist or a resident in training under direct dip test statistic was used for testing unimodality. supervision of a board-certified cardiologist. Dogs were included For survival analysis, the endpoint was defined as death or in the study only if they were referred for investigation of clinical euthanasia related to heart disease (cardiac mortality). Dogs that signs (such as congestive heart failure, syncope, exercise intoler- died or were euthanized for other reasons, were alive at the end ance, episodic weakness, or arrhythmias noted on routine physical of the study, or were lost to follow-up were right-censored. Kap- examination), and the study population did not include dogs lan-Meier survival curves were constructed and analyzed by the presented for routine screening for breeding purposes. Data were obtained from medical records, echocardiographic images, and Table 1. List of variables recorded from the Holter Holter reports. recordings and the respective definitions. Boxer dogs with at least 1 Holter recording with a minimum Holter variables of 19 hours of valid data and an echocardiographic examination Presence and total number of atrial premature within 1 month of Holter recording were included in the study. contractions (APC) Dogs with documented congenital heart disease were excluded Presence and total number of episodes of supraventricular from the study, including those with aortic blood flow velocities tachycardia (SVT) (Ao ) 2.25 m/s.16 vel Presence of episodes of bradycardia Baseline clinical details obtained from medical records at the Maximal duration of bradycardia episode time of first Holter recording included age, sex, body weight, a Minimal average ventricular rate during bradycardia episode history of syncope or episodic weakness or collapse that was Presence and total number of ventricular arrhythmias believed to be consistent with syncope, the presence of congestive Total number of ventricular premature contractions (VPCs) heart failure before presentation, and concurrent diseases. The Presence of polymorphic VPCs diagnosis of CHF was based on radiographic evidence of pulmo- Presence of bigeminy or trigeminy nary venous congestion or edema, or when thoracic radiographs Presence of couplets or triplets were not available, the attending clinician’s judgment was based Presence and total number of episodes of ventricular on a combination of history, physical examination (the presence tachycardia (VT) of tachypnea and crackles, or jugular distension and ascites), and Longest duration of VT episode ultrasound findings (atrial enlargement with pleural effusion, Average ventricular rate of fastest VT episode abdominal effusion, or both). The presence of syncope before Minimal, mean, and maximal 1-minute average heart rate Holter recording was determined based on the history recorded Definitions in the patient records. Pause >2.5 seconds All echocardiographic measurements were made off-line by 1 Bradycardia—minimum of 4 consecutive normally conducted observer in each center by proprietary software.a,b,c,d None of the sinus-beats at the rate of <60 beats per minute dogs was sedated for the echocardiographic examination. For cal- SVT—minimum of 3 consecutive normally conducted culation of LA/Ao, left atrial diameter (LA) and aortic root diame- sinus-beats at >160 beats per minute ter (Ao) were measured by 2-dimensional (2D) echocardiography APC—normally conducted sinus complex with RR interval from the first frame after aortic valve closure.17,18 M-mode echo- of <45% of the preceding RR interval cardiography was used to measure left ventricular (LV) LV internal Bigeminy—minimum of 2 VPCs alternated with diameters in systole (LVIDs) and diastole (LVIDd) and to calcu- 1 sinus complex late fractional shortening (FS%) using the leading-edge-to-leading- Trigeminy—minimum of 2 VPCs alternated with edge dimensions.19,20 Spectral Doppler echocardiography was used 2 sinus complexes to measure subcostal aortic velocity (Aovel).21 Dogs were classified Couplet—2 consecutive VPCs as having systolic dysfunction based on LVIDs > 35 mm.3 Triplet—3 consecutive VPCs Two centers used the same commercial Holter analysis servicee VT—more than 3 consecutive VPCs at the rate and in the 3rd center, raw Holter data were reviewed.f One inves- of >100 beats per minute tigator in each center reviewed available Holter recordings using

96 906 Mõtskula€ et al log-rank test to identify factors that significantly influenced car- leaving 122 dogs included in the study. Myocardial histo- diac mortality, and to estimate median survival time. Survival pathology or necropsy information was not available for analysis was carried out for the whole population, and separately any of the dogs. Signalment, clinical findings (the pres- for dogs with normal left ventricular systolic function (defined as 3 ence of syncope and CHF), survival data, echocardio- LVIDs 35 mm) and no evidence of CHF (LVIDs 35 mm graphic findings, and results of Holter monitoring are group). The Cox proportional hazards analysis was used to detect the summarized in Table 2. Data are presented separately for the whole population, the LVIDs 35 mm group, and association between each variable and time to event in the LVIDs 35 mm group. Results were expressed as the hazard ratio the LVIDs>35 mm group. Age distribution, the presence (HR) with 95% confidence intervals (95% CI). A multivariable of syncope, and CHF are shown in Figure 1. model was constructed using a manual forward stepwise approach The distribution of Holter variables among the 3 to identify variables that independently predicted outcome. Pair- centers was not statistically significant, apart from wise interactions of the variables in the final model were evaluated, maximum (P < .001) and minimum (P < .001) 1-min- the proportional hazard assumption was assessed by log cumulative ute heart rate. No syncope episodes were documented hazard plots and Schoenfeld residuals, and overall model fit by 22 during Holter recording in any dog. graphical assessment of Cox-Snell residuals. Twenty-eight dogs had LV systolic dysfunction (defined as LVIDs > 35 mm). Three dogs with Results LVIDs 35 mm had CHF at presentation, leaving A total of 163 dogs were identified based on a search 91 dogs in the LVIDs 35 mm group available for sta- of the medical records of the 3 participating centers. tistical analysis. The 3 excluded dogs all had marked Forty dogs were excluded for Aovel 2.25 m/s and 1 biatrial and right ventricular enlargement on 2D echo- dog had insufficient duration of the Holter recording, cardiographic examination.

Table 2. Signalment, clinical ﬁndings (presence of syncope and congestive heart failure), echocardiographic ﬁnd- ings, and Holter results for dogs without left ventricular systolic dysfunction or congestive heart failure (LVIDs 35 mm/no CHF); dogs with left ventricular systolic dysfunction, congestive heart failure (LVIDs>35 mm/ CHF), or both; and the whole population of dogs (All dogs).

LVIDs 35 mm/no CHF LVIDs>35 mm/CHF All Dogs Variable Median (IQR) n (%) Median (IQR) n (%) P Median (IQR) n (%) Male 38 (42%) 28 (80%) <.001 63 (52%) Age (years) 6.5 (2.8–9.0) 91 (100%) 7.7 (5.9–9.0) 31 (100%) .054 6.9 (4.1–9.0) 122 (100%) Body weight (kg) 28.2 (25.0–33.6) 91 (100%) 33.8 (29.0–37.9) 31 (100%) <.001 30.0 (25.6–34.8) 122 (100%) Syncope on presentation 69 (76%) 24 (77%) .857 93 (76%) Aortic velocity (m/s) 1.81 (1.50–1.97) 91 (100%) 1.59 (1.28–1.87) 31 (100%) .014 1.73 (1.42–1.95) 122 (100%) LA/Ao 1.42 (1.31–1.55) 84a (92%) 2.21 (1.70–2.51) 30 (97%) <.001 1.49 (1.35–1.73) 114a (93%) LVIDs (mm) 26.41 (23.3–29.6) 91 (100%) 41.9 (36.1–45.7) 31 (100%) <.001 28.2 (24.7–34.1) 122 (100%) FS% 29 (24–32) 89a (98%) 16 (10–25) 30a (97%) <.001 26 (21–31) 119a (98%) Presence of APCs 43 (47%) 21 (68%) .049 64 (52%) Number of APC/24 h 4 (2–32) 512 (5–2,152) .004 8.5 (2–472) Presence of SVT 18 (20%) 11 (35%) .038 27 (22%) Number of SVT episodes 2 (1–19) 53 (5–204) .011 5 (2–62) Presence of VPCs 86 (95%) 31 (100%) .183 117 (96%) Average VPC/24 h 36 (2.9–1,001) 967 (208–3,376) <.001 199 (5.5–1,410) >50 VPC/24 h 44 (48%) 27 (87%) <.001 71 (58%) >100 VPC/24 h 40 (44%) 26 (83%) <.001 66 (54%) >1,000 VPC/24 h 23 (25%) 15 (48%) .016 38 (31%) Presence of polymorphic 32 (35%) 24 (77%) <.001 56 (46%) VPCs Presence of bigeminy 36 (40%) 18 (58%) .073 54 (44%) or trigeminy Presence of couplets 54 (59%) 26 (84%) .013 80 (66%) or triplets Presence of VT 22 (24%) 14 (45%) .027 36 (30%) Number of VT episodes 8.5 (3–24.3) 8 (1.75–69.25) .761 8 (3.0–32.5) Minimum heart rate (bpm) 50 (43–59) 91 (100%) 55 (46–73) 31 (100%) .068 51 (44–61) 122 (100%) Mean heart rate (bpm) 87 (77–100) 91 (100%) 97 (84–110) 31 (100%) .026 90 (78–104) 122 (100%) Maximum heart rate (bpm) 186 (171–206) 91 (100%) 192 (166–215) 31 (100%) .520 190 (171–206) 122 (100%)

IQR, interquartile range; CHF, congestive heart failure; LA, left atrial diameter; Ao, aortic diameter; FS%, fractional shortening; APC, atrial premature complex; SVT, supraventricular tachycardia; VPC, ventricular premature complex; VT, ventricular tachycardia; HR, heart rate.

The P value refers to a comparison between the [LVIDs 35 mm/no CHF] and [LVIDs>35 mm/CHF] subpopulations. aThe number of the dogs in which the data were available.

97 Holter Monitoring in Boxer Dogs 907

A All dogs B All dogs

Syncope 15 15 VT Yes

10 10 Number of dogs Number of dogs

5 5

0 0 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 Age range Age range

Fig 1. Age distribution of the whole population suggests a bimodal age distribution. Young dogs were more likely presented for investigation of syncope (P = .040) (A), but VT occurred more frequently in older dogs (P = .015) (B).

1.0 1.0 A LVIDs<35mm B Syncope - No

LVIDs>35mm Syncope - Yes

0.8 0.8

p<0.001 p=0.385

0.6 0.6

Fraction survival 0.4 Fraction survival 0.4

0.2 0.2

0.0 0.0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500

Days Days

Fig 2. Kaplan-Meier survival curves in overall population. Dogs in group (A) had signiﬁcantly shorter median survival time (MST)

(242 days, 95% CI 0–512 days) than dogs in LVIDs 35 mm group (2,083 days, 95% CI 1,336–2,830 days), P < .001. The presence of syncope (B) did not signiﬁcantly inﬂuence survival (P = .385). MST for dogs with syncope was 1,244 days (95% CI 737–1,750 days) versus >2,107 days in dogs without syncope.

Dogs with LVIDs > 35 mm were more likely to Kaplan-Meier analysis of the whole population, median have increased left atrial size, decreased FS%, and survival time was significantly shorter for dogs with CHF. There were significant associations between LVIDs > 35 mm compared with the LVIDs 35 mm LVIDs > 35 mm and FS < 25% (P < .001), LA/ group (242 versus 2,083 days, P < .001, Fig 2). The Ao > 1.5 (P < .001), the presence of CHF (P < .001), presence of syncope before presentation was not associ- and >50 VPC/24 h (P < .001) as well as between ated with decreased survival time. FS < 25% and CHF (P = .002), and >50 VPC/24 h In the LVIDs 35 mm group, male sex, the presence of (P = .002). The presence of syncope, however, was not >50 VPC/24 h, the presence of polymorphic VPCs, and significantly associated with any of these indices. the presence of VT all were significantly associated with Median follow-up period for the whole population shorter survival. The presence of couplets or triplets, was 500 days (IQR, 186–917 days). The follow-up per- bigeminy or trigeminy, the presence of bradycardia epi- iod for dogs that were alive at the end of the study sodes, and the presence of pauses did not reach statistical period was significantly longer than the follow-up per- significance. Increased left atrial size (LA/Ao > 1.5), iod for dogs that had died, both in the whole popula- decreased fractional shortening (FS < 25%), the pres- tion (P < .001) as well as in the LVIDs 35 mm group ence of syncope, APC, or SVT were not predictive of (P = .005). outcome in the LVIDs 35 mm group (Table 3, Fig 3). Kaplan-Meier survival curves were constructed The Kaplan-Meier analysis was used to assess the prog- separately to evaluate cardiac mortality in the whole nostic value of the 3 cutoff values commonly used in the population and the LVIDs 35 mm group. Based on the literature to differentiate between Boxers affected and

98 908 Mõtskula€ et al

Table 3. Median survival time and 95% CI in the predictors of survival times in the LVIDs 35 mm group LVIDs 35 mm group based on the Kaplan-Meier anal- (Fig 6). ysis. Variables in the table were significantly associated with median survival times. The presence of bigeminy Discussion or trigeminy, and couplets or triplets did not reach statistical significance. Two previous studies describing outcome in Boxer dogs limited the population to dogs affected with Median Survival ARVC.1,3 Because the criteria for diagnosis of ARVC Variable Time (days; 95% CI) P (log-rank) are still poorly defined, we assessed the prognostic fac- Sex tors in a broader population of Boxer dogs that would Male 1,244; 852–1,635 .035 better reflect a typical practice situation. Extensive Female >2,083 investigations to rule out other causes of ventricular VT arrhythmias were not conducted in most dogs, and Yes 1,244; 351–2,137 .002 conditions other than ARVC may have been responsi- No >2,083 ble for ventricular arrhythmias in some dogs. In our Polymorphic VPC Yes 1,660; 681–2,638 .019 study, the most common reason for seeking a referral No 2,083; 893–3,272 was a history of syncope. There are conflicting results >50 VPC/24 h for the association between the presence of syncope Yes 1,393; 895–1,890 .003 and cardiac morbidity or mortality in humans with No >2,083 ARVC.23–26 Our results suggest that the presence of Age tertiles syncope is not associated with a worse prognosis, in >4.67 years >2,083 .001 contrast with Palermo’s findings.3 However, our popu- 1,393; 703–2,083 lation included Boxer dogs that did not meet the most 4.67–8.07 years common diagnostic criteria for ARVC, and may have >8.07 years 1,086; 493–1,679 included a larger proportion of young dogs with neurocardiogenic syncope; these cases would be expected to have a better prognosis. In addition, fewer dogs had not affected by ARVC. Dogs with increased numbers of LV systolic dysfunction than in Palermo’s study. Only VPCs using any of the 3 previously suggested cutoff val- a small proportion of dogs in Palermo’s study had a ues (50, 100, and 1,000 VPC/24 h) had significantly Holter recording available, making it difficult to draw shorter survival times (P = .002, .018, and .019, respec- conclusions on the association between syncope and tively, Fig 4). severity or frequency of arrhythmias. Furthermore, the Age distribution was bimodal with 2 peaks (P = .04) at arrhythmias responsible for syncope are not necessarily approximately 1 and 8 years of age (Fig 1). There was no the same as those responsible for sudden death. correlation between age at diagnosis and the presence of The bimodal age distribution seen in our population LVIDs > 35 mm (P = .612), sex (P = .953), the presence may be explained by the hypothesis that young dogs of APCs (P = .133), the presence of SVT (P = .772), or are more likely to be referred for isolated fainting epi- LA/Ao (P = .362). Younger dogs were more likely to be sodes associated with neurocardiogenic syncope, presented for investigation of syncope (P = .040), and whereas older dogs are more likely to be presented for were more likely to have <50 VPC/24 h (P = .001). Dogs other signs of congestive heart failure, or syncope presenting with polymorphic VPCs (P < .001), bigeminy associated with arrhythmias attributable to ARVC. or trigeminy (P = .008), couplets or triplets (P < .001), The effect of treatment was not assessed in our study, and VT (P = .015) were older. because treatment protocols were not standardized. Dogs in the youngest age tertile had significantly As with the studies by Palermo et al3 and Meurs longer survival than the other 2 tertiles (P < .001). et al,1 our results showed that left ventricular systolic There was no significant difference in survival between dysfunction was associated with increased cardiac the 2 higher age tertiles (P = .574; Fig 5). An age of mortality. As expected, other factors that were signifi- 4.5 years consequently was used as a cutoff value in cantly associated with worse outcome in our overall the multivariable statistical analysis between young population were FS < 25%, LA/Ao > 1.5, the presence and older dogs. of congestive heart failure, and the presence of Variables included in the multivariable Cox propor- >50 VPC/24 h, because all of these factors were tional hazards model for the LVIDs 35 mm group strongly correlated with one another. Boxer dogs with included sex, age >4.5 years, LA/Ao, FS%, the pres- CHF and systolic dysfunction are well recognized to ence of SVT, the presence of >50 VPC/24 h, the pres- have a poor prognosis, and their management usually ence of polymorphic VPCs, the presence of bigeminy is very different from the management of Boxer dogs or trigeminy, the presence of couplets or triplets, and with normal echocardiographic findings diagnosed the presence of VT. Based on the multivariable Cox with ARVC, and thus, dogs with LVIDs > 35 mm or proportional hazard analysis, only age >4.5 years (HR, the presence of CHF were excluded from further statis- 28.82; 95% CI, 3.03–274.31; P = .003), the presence of tical analysis. The Boxers with LVIDs 35 mm were a VT (HR, 3.407; 95% CI, 1.37–8.45; P = .008), and male diverse group with varied ages and etiologies. Referral sex (HR, 2.73; 95% CI, 1.04–7.12) were independent to a second-opinion hospital often was sought because

99 Holter Monitoring in Boxer Dogs 909

1.0 1.0 A Female B VT - No

Male VT - Yes

0.8 0.8

0.6 0.6

p=0.035 p=0.002 Fraction survival Fraction 0.4 Fraction survival 0.4

0.2 0.2

0.0 0.0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500

Days Days

1.0 C 1.0 D <50 VPC/24h Polymorphic VPC - No

>50 VPC/24h Polymorphic VPC - Yes

0.8 0.8

0.6 0.6

p=0.003 p=0.019

Fraction survival 0.4

0.2 0.2

0.0 0.0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Days Days

Fig 3. Kaplan-Meier survival curves (cardiac mortality) in the LVIDs 35 mm group. Male dogs (A), dogs with ventricular tachycardia (B), >50 VPC/24 h (C), and polymorphic VPCs (D) had shorter median survival times.

1.0 A 1.0 B 1.0 C <50 VPC/24h < 100 VPC < 1000 VPC

>50 VPC/24h > 100 VPC > 1000 VPC

0.8 0.8 0.8

0.6 0.6 0.6

p=0.003 p=0.018 p=0.019 Fraction survival Fraction Fraction survival 0.4 Fraction survival 0.4 0.4

0.2 0.2 0.2

0.0 0.0 0.0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500

Days Days Days

Fig 4. Kaplan-Meier survival curves (cardiac mortality) in the LVIDs 35 mm group. Median survival was shorter for Boxers exceeding any of the cutoff values used for distinguishing between Boxers affected and not affected by ARVC (>50 VPC/24 h (A), >100 VPC/24 h (B), or >1,000 VPC/24 h (C)). of syncope, detection of an arrhythmia during routine A bimodal age distribution in syncopal Boxers has physical examination, episodic weakness, or exercise been suggested by other authors.2,3,6,27 This finding intolerance. Because the prognosis in Boxers with also was noted in our population, and the age distribu- normal echocardiography is very variable, it is impor- tion pattern was similar to Palermo’s findings.3 In our tant to identify prognostic factors, because information population, there was no association between age at on outcome in these dogs is scant in the current presentation and sex, the presence of APC, the literature. presence of SVT, or the proportion of dogs with LA/

100 910 Mõtskula€ et al

1.0 polymorphic VPCs and >50 VPC/24 h were signiﬁ-

<4.67 years cantly associated with worse outcome in the Cox uni-

4.67-8.07 years variable model, whereas the presence of bigeminy or 0.8 >8.07 years trigeminy, and couplets or triplets failed to reach statistical signiﬁcance in predicting a worse outcome in our

0.6 population. Classiﬁcation of ventricular arrhythmias as “polymorphic ventricular ectopy” may not have been accurate in all cases, because changes in QRS morphol-

Fraction survival 0.4 ogy also can be the result of changes in body position; p (all groups)=0.001 however, the association with worse survival is other- p (T2, T3)=0.574 wise diﬃcult to explain. The complexity of ventricular 0.2 arrhythmias has not been clearly associated with increased mortality in veterinary studies, but is considered a negative prognostic factor in the human medical 0.0 25 0 500 1000 1500 2000 2500 literature. Neither the presence of APCs or SVT nor

Days the presence of bradycardia episodes or pauses was associated with outcome in the LVIDs<35 mm group.

Fig 5. Kaplan-Meier survival curves in LVIDs 35 mm group The reported sex distribution in ARVC Boxers in pre- comparing effect of age by tertiles. The youngest tertile vious publications has varied from male predomi- (<4.67 years) had significantly longer median survival than the rest nance3,28 to female predominance.29 In our total of the population (T2, 4.67–8.07 years; T3, >8.07 years). No signif- population, a very mild male predominance was docu- icant difference was present between the 2 oldest age groups. mented, with a smaller proportion of male dogs in the LVIDs 35 mm group. The Kaplan-Meier survival analysis of the LVIDs 35 mm group showed that male dogs Male sex HR 2.73, 95%CI 1.04-7.12 in this group have a significantly worse prognosis. The bimodal age distribution and significant association between many variables and age suggest that the Presence of VT HR 3.41, 95%CI 1.37-8.45 correlation between age and prognosis is not likely to be linear. The Kaplan-Meier survival analysis of age HR 28.82, 95%CI 3.03-274.31 Age > 4.5 y tertiles supported this conclusion, because the first tertile did significantly better than the older age groups,

1 0 whereas there was no significant difference between the 10 0 1 2 older tertiles (Fig 5). This information is clinically Hazard ratio +/- 95%CI (log10) important, as it suggests that young syncopal Boxers Fig 6. Multivariate Cox proportional hazard analysis indicates are more likely to present with benign disease, such as that male sex, the presence of ventricular tachycardia, and age vagally mediated (neurocardiogenic) syncope and over 4.5 years are independent predictors of worse long-term out- therefore carry a markedly better prognosis, whereas come. older dogs could present with ARVC or systemic diseases that carry a poorer long-term prognosis. The definitive diagnostic criteria for ARVC have not Ao > 1.5. With increasing age, the proportion of dogs been established. This makes it difficult to provide a cut- having polymorphic VPC, bigeminy or trigeminy, cou- off age to distinguish dogs with benign syncope from plets or triplets, or VT increased. This, in combination dogs with ARVC. The number of VPC/24 h alone can- with a larger proportion of young dogs having not be used to differentiate between dogs affected and <50 VPC/24 h, suggests that older dogs have increased not affected with ARVC, because exceeding any of the complexity and frequency of ventricular arrhythmias. previously suggested cutoff values was associated with Whether this finding is associated with more advanced shorter survival times. A grading scheme based on fre- disease in the older population or caused by the pres- quency and severity of ventricular arrhythmias on ence of a less malignant condition in younger dogs, ambulatory ECG monitoring has been suggested by such as vasovagal syncope,14 was not determined. Lown30 for use in human cardiology. A modification in However, the fact that only 2 dogs in the 1st age tertile Lown grading has been used in veterinary studies, but reached the primary endpoint later in life may support we lack data on the association with such schemes the latter hypothesis and is the most likely explanation and cardiac mortality.1,4,31 The data presented in our of the bimodal age distribution. study could form the basis for constructing a new It has long been suspected that the presence of VT in arrhythmia-grading scheme that would ideally be tested an echocardiographically normal Boxer is associated in a validation cohort of Boxers before being used as a with an increased risk of cardiac death, and our find- classification scheme for severity of arrhythmias in this ings supported this suspicion. Dogs in LVIDs 35 mm breed. group with VT had a median survival time of On the basis of our results, we suggest a better prog- 1,244 days (95% CI, 351–2,136 days) versus >2,083 nosis in Boxer dogs under 4.5 years of age, with no days in dogs without VT. In addition, the presence of echocardiographic abnormalities and <50 monomorphic

101 Holter Monitoring in Boxer Dogs 911

VPC/24 h. The multivariable Cox model identified male Acknowledgments sex, age >4.5 years, and the presence of ventricular tachycardia as independent predictors of a worse outcome. The authors thank Yu-Mei Chang for her help with statistical analyses. The study was done in London, UK. The study was not supported by a grant. Limitations Conflict of Interest Declaration: Authors disclose no Several limitations may have affected the results of conflict of interest. this study. The retrospective nature of the study may have influenced the results, such as the accurate docu- References mentation of syncope and congestive heart failure. The 1. Meurs K, Lahmers S, Keene B. ACVIM Forum Research data were collected from 3 different centers, with 2 Abstracts Program. J Vet Intern Med 2011;25:649 (abstract centers using the same commercial diagnostic labora- C.11). tory for analyzing the Holter recordings, which may 2. Harpster NK. Boxer cardiomyopathy. In: Kirk R, ed. Cur- have resulted in differences in analysis, and errors of rent Veterinary Therapy VIII. Philadelphia, PA: WB Saunders; interpretation. Periods of Holter recording artifacts 1983:329–337. also may have led to false-negative and false-positive 3. Palermo V, Stafford Johnson MJ, Sala E, et al. Cardiomy- opathy in Boxer dogs: A retrospective study of the clinical pre- classification of arrhythmias. The use of an arbitrary sentation, diagnostic findings and survival. J Vet Cardiol cutoff value for the classification of dogs as having 2011;13:45–55. normal or abnormal systolic function may have led to 4. Meurs KM, Spier AW, Miller MW, et al. Familial ventric- misclassification of some patients. However, the chosen ular arrhythmias in Boxers. J Vet Intern Med 1999;13:437–439. cutoff value allows comparison of our study with pre- 5. Basso C, Fox PR, Meurs KM, et al. Arrhythmogenic right vious studies.3 The effect of treatment on outcome was ventricular cardiomyopathy causing sudden cardiac death in not assessed. Lastly, the small number of patients was Boxer dogs: A new animal model of human disease. Circulation also a limitation in our study. 2004;109:1180–1185. 6. Meurs KM. Boxer dog cardiomyopathy: An update. Vet Clin North Am Small Anim Pract 2004;34:1235. Conclusions 7. Smith CE, Freeman LM, Meurs KM, et al. Plasma fatty Despite these limitations, the results of this study acid concentrations in Boxers and Doberman Pinschers. Am J Vet Res 2008;69:195–198. confirm that valuable prognostic information can be 8. Scansen B, Meurs K, Spier A, et al. Temporal variability of gained from Holter monitoring in Boxer dogs that are ventricular arrhythmias in Boxer dogs with arrhythmogenic right presented for investigation of suspected ARVC. The ventricular cardiomyopathy. J Vet Intern Med 2009;23:1020–1024. presence of systolic dysfunction or congestive heart 9. O’Grady MR, O’Sullivan ML. Dilated cardiomyopathy: An failure also is associated with a poor prognosis and update. Vet Clin North Am Small Anim Pract 2004;34:1187. shorter life expectancy. Syncope was not associated 10. Spier AW, Meurs KM. Assessment of heart rate variabil- with outcome in our population. Boxer dogs ity in Boxers with arrhythmogenic right ventricular cardiomyopa- <4.5 years of age, with normal echocardiographic thy. J Am Vet Med Assoc 2004;224:534–537. examination findings and fewer than 50 monomorphic 11. Baumwart RD, Meurs KM. Assessment of plasma brain VPC/24 h, have a good long-term prognosis. The pres- natriuretic peptide concentration in Boxers with arrhythmogenic right ventricular cardiomyopathy. Am J Vet Res 2005;66:2086–2089. ence of ventricular tachycardia, older age, and male 12. Hariu CD, Carpenter DH. Arrhythmogenic right ventricu- sex are associated with shorter survival times. Addi- lar cardiomyopathy in Boxers. Compend Contin Educ Vet tional studies with a larger number of patients and 2010;32:E1–E7. including dogs presented for screening are required to 13. Bussadori C, Pradelli D, Borgarelli M, et al. Congenital formulate and validate an arrhythmia-grading scheme heart disease in Boxer dogs: Results of 6 years of breed screen- that could be used for screening purposes. ing. Vet J 2009;181:187–192. 14. Fisher EW. Fainting in Boxers–the possibility of vasovagal syncope (Adams-Stokes attacks). J Small Anim Pract 1971;12:347–349. Footnotes 15. Thomason J, Kraus M, Surdyk K, et al. Bradycardia- associated syncope in 7 Boxers with ventricular tachycardia a Philips Sonos 5500; Philips Healthcare, Surrey, UK (2002–2005). J Vet Intern Med 2008;22:931–936. b GE Vivid 5; GE Healthcare, Milwaukee, WI 16. Bussadori C, Amberger C, Le Bobinnec G, Lombard CW. c GE Vivid 7; GE Healthcare Guidelines for the echocardiographic studies of suspected subaor- d GE EchoPAC PC; GE Healthcare tic and pulmonic stenosis. J Vet Cardiol 2000;2:15–22. e Laboratory Corporation of America, Ambulatory Monitoring 17. Hansson K, H€aggstrom€ J, Kvart C, Lord P. Left atrial to Services, Burlington, NC aortic root indices using two-dimensional and M-mode echocar- f Novacor Holter Soft, Novacor UK Ltd, Kent, UK diography in Cavalier King Charles Spaniels with and without g PASW Statistics 18.0 for Windows; SPSS Inc, Chicago, IL left atrial enlargement. Vet Radiol Ultrasound 2005;43:568–575. h Prism 5 for Mac OS X ver 5.0a, GraphPad Software, Inc. La 18. Rishniw M, Erb HN. Evaluation of four 2-dimensional Jolla, CA echocardiographic methods of assessing left atrial size in dogs. i package in R 2.15.2 software, R Foundation, Vienna, J Vet Intern Med 2000;14:429–435. Austria 19. Thomas WP, Gaber CE, Jacobs GJ, et al. Recommenda- tions for standards in transthoracic two-dimensional echocardiography in the dog and cat. J Vet Intern Med 1993;7:247–252.

102 912 Mõtskula€ et al

20. Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommen- 26. Basso C, Corrado D, Marcus FI, et al. Arrhythmogenic dations regarding quantitation in M-mode echocardiography: right ventricular cardiomyopathy. Lancet 2009;373:1289–1300. Results of a survey of echocardiographic measurements. Circula- 27. Martin MWS, Johnson MJS, Celona B. Canine dilated tion 1978;58:1072–1083. cardiomyopathy: A retrospective study of signalment, presenta- 21. Lehmkuhl LB, Bonagura JD. Comparison of transducer tion and clinical findings in 369 cases. J Small Anim Pract placement sites for Doppler echocardiography in dogs with sub- 2009;50:23–29. aortic stenosis. Am J Vet Res 1994;55:192–198. 28. Harpster NK. Boxer cardiomyopathy. A review of the 22. Dohoo I, Martin W, Stryhn H. Modelling survival data. long-term benefits of antiarrhythmic therapy. Vet Clin North Am In: Dohoo I, Martin W, Stryhn H, eds. Veterinary Epidemiologi- Small Anim Pract 1991;21:989–1004. cal Research, 2nd ed. Charlottetown: VER Inc; 2010:467–528. 29. Baumwart RD, Meurs KM, Atkins CE, et al. Clinical, 23. Hulot J-S, Jouven X, Empana J-P, et al. Natural history echocardiographic, and electrocardiographic abnormalities in and risk stratification of arrhythmogenic right ventricular dyspla- Boxers with cardiomyopathy and left ventricular systolic dysfunc- sia/cardiomyopathy. Circulation 2004;110:1879–1884. tion: 48 cases (1985–2003). J Am Vet Med Assoc 2005;226:1102– 24. Pinamonti B, Dragos AM, Pyxaras SA, et al. Prognostic 1104. predictors in arrhythmogenic right ventricular cardiomyopathy: 30. Lown B, Calvert AF, Armington R, Ryan M. Monitoring Results from a 10-year registry. Eur Heart J 2011;32: for serious arrhythmias and high risk of sudden death. Circula- 1105–1113. tion 1975;52(6 Suppl):III189–III198. 25. Corrado D, Basso C, Thiene G. Arrhythmogenic right 31. Santilli RA, Bontempi LV, Perego M. Ventricular tachy- ventricular cardiomyopathy: Diagnosis, prognosis, and treatment. cardia in English Bulldogs with localised right ventricular outflow Heart 2000;83:588–595. tract enlargement. J Small Anim Pract 2011;52:574–580.

103

II Kaye BM, Borgeat K, Mõtsküla PF, Luis Fuentes V, Connolly DJ., 2015. ASSOCIATION OF TRICUSPID ANNULAR PLANE SYSTOLIC EXCURSION WITH SURVIVAL TIME IN BOXER DOGS WITH VENTRICULAR ARRHYTHMIAS. Journal of Veterinary Internal Medicine 29, 582–588. J Vet Intern Med 2015;29:582–588

Association of Tricuspid Annular Plane Systolic Excursion with Survival Time in Boxer Dogs with Ventricular Arrhythmias

B.M. Kaye, K. Borgeat, P.F. Motsk~ ula,€ V. Luis Fuentes, and D.J. Connolly

Background: Tricuspid annular plane systolic excursion (TAPSE) is a useful estimate of right ventricular function in humans. Reference intervals for dogs have been generated, but the value of measuring TAPSE in other diseases, or investigat- ing the association between TAPSE and outcome, is unknown. Hypothesis: TAPSE is lower in Boxer dogs with ≥50 VPCs/24 h on Holter than in dogs with fewer ventricular ectopics, and lower TAPSE is associated with a shorter survival time. Animals: Fifty Boxer dogs that presented for investigation of syncope or suspected arrhythmogenic right ventricular cardiomyopathy (ARVC) at a veterinary teaching hospital (2004–2011). Methods: Retrospective study. Clinical records, Holter, and echocardiographic data were reviewed. TAPSE was measured in a blinded manner on stored echocardiographic cine-loops using anatomic M-mode. Outcome information was obtained and death was classiﬁed as cardiac or noncardiac. Survival analysis was performed using Kaplan-Meier curves and Cox proportional hazards models. Results: TAPSE was lower in Boxers with ≥50 VPCs/24 h (13.9 4.04 mm) than Boxers with <50 VPCs/24 h (16.8 3.21 mm; P < .001). TAPSE <15.1 mm was associated with shorterÆ cardiac survival time in all dogs (P = .004) and Æ also in dogs without left ventricular dysfunction (P = .035). When controlling for other variables, including ventricular tachycardia on Holter and left ventricular systolic dysfunction, multivariable analysis showed that TAPSE remained an independent predictor of time to cardiac death (HR >4.09, 95%CI 1.15–16.9, P < .029). Conclusions and Clinical Importance: TAPSE oﬀers prognostic value for Boxer dogs, including those with apparently normal systolic function and ≥50 VPCs/24 h on Holter analysis. Key words: Arrhythmia; Canine; Echocardiography; Holter; M-mode.

Introduction Abbreviations: rrhythmogenic right ventricular cardiomyopathy ARVC arrhythmogenic right ventricular cardiomyopathy (ARVC) is a primary myocardial disease character- A cMRI cardiac magnetic resonance imaging ized by fibro-fatty infiltration of the right, and occasion- VPC ventricular premature complex ally left, ventricle. LV left ventricular The diagnosis of ARVC in humans is based upon a VT ventricular tachycardia combination of genetic screening, family history, ambu- RV right ventricular latory electrocardiography (Holter), and the detection TAPSE tricuspid annular plane systolic excursion of right ventricular adipose tissue on cardiac magnetic CHF congestive heart failure 1 resonance imaging (cMRI). There is an extensive litera- LA left atrium ture describing the genetic, clinical, echocardiographic, Ao aorta and electrophysiologic characteristics of ARVC in LVIDs left ventricular internal diameter in systole 2–6 Boxer dogs. FS% fractional shortening There are several prognostic indicators for Boxers SD systolic dysfunction dogs with ARVC. Shorter survival times are associated EF% ejection fraction with echocardiographic evidence of left ventricular (LV) systolic dysfunction,5,7 increased frequency and complexity of VPCs, including ventricular tachycardia (VT), and polymorphic VPCs.7 In human medicine, right ventricular (RV) function is From the Royal Veterinary College, Hatfield, UK; Highcroft commonly assessed by measurement of tricuspid annu- Veterinary Referrals, Bristol, UK; Estonian University of Life lar plane systolic excursion (TAPSE) using M-mode Sciences, Kreutzwaldi, Tartu, Estonia. echocardiography.8–11 TAPSE is a quantitative estimate Where the work was done: Royal Veterinary College. of longitudinal RV shortening. It does not account for Corresponding author: K. Borgeat, Highcroft Veterinary Refer- the radial and circumferential motion of the RV myo- rals, Bristol, BS14 9BE, UK; e-mail: kieran.borgeat@highcroft- vet.co.uk. cardial wall, but physiologic studies in humans have Submitted September 4, 2014; Revised December 21, 2014; shown that RV shortening is greater longitudinally than Accepted February 10, 2015 radially and torsion contributes less to RV than LV 12 Copyright © 2015 The Authors. Journal of Veterinary Internal output. TAPSE was significantly correlated with RV Medicine published by Wiley Periodicals, Inc. on behalf of the ejection fraction in validation studies against the gold American College of Veterinary Internal Medicine. standard of radio-nuclide angiography,13 and reduced This is an open access article under the terms of the Creative TAPSE is associated with a worse outcome in people Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original with congenital heart disease, pulmonary hypertension, work is properly cited and is not used for commercial purposes. acute symptomatic pulmonary embolism, hypertrophic 8,14–16 DOI: 10.1111/jvim.12572 cardiomyopathy, and ARVC. Several studies have

107 Cardiac Death in Boxer Dogs 583 also reported a correlation in humans between TAPSE Information on survival was obtained via several means; com- and echocardiographic measures of LV systolic func- plete medical records, questionnaires mailed to owners, and tele- tion,17 illustrating the interdependence between the left phone contact with clients and referring veterinarians. and right heart. Documented outcome data included whether the dog was alive or Assessing RV function has received recent attention dead, the date of death or euthanasia, and whether the death was cardiac related. Cardiac death was defined as death or euthanasia in animals. The index of myocardial performance (or associated with clinical signs attributable to cardiac disease TEI index) has been evaluated in Boxer dogs with (uncontrolled CHF, worsening CHF, or unacceptable increase in ARVC, and measurements suggested worse global myo- the frequency of syncope) or an unexpected death without appar- cardial performance in affected dogs, independent of ent clinical signs during the preceding 24-hour period. Necropsy body weight and heart rate variability.18 Reference data were not available. intervals for TAPSE in normal dogs of various breeds 19 and body weights have been derived. In the same Echocardiography study, dogs with pulmonary hypertension had lower TAPSE than dogs without pulmonary hypertension. In Echocardiographya was performed in unsedated dogs and all addition, TAPSE is easily obtained and repeatable.1 echocardiographic images were acquired by a cardiology diplomate This suggests that TAPSE could be a practical and clin- or cardiology Resident under direct supervision. For calculation of ically useful measurement in RV dysfunction in dogs, as left atrial to aortic root ratio (LA/Ao), diameters of the left atrium has already been reported in humans. To the authors’ (LA) and aorta were measured using 2D echocardiography in a right parasternal short axis view of the heart base, from the first knowledge, the association between TAPSE and out- frame after aortic valve closure (end-systolic).21 M-mode echocar- come has not been reported in dogs for any disease pro- diography was used to measure (leading-edge to leading-edge) LV cess. internal diameters in systole (LVIDs) and diastole, and to calculate The aim of this study was to evaluate the clinical util- fractional shortening (FS%).22,23 Spectral Doppler echocardiogra- ity of TAPSE in a cohort of Boxer dogs presented for phy was used to measure aortic outflow velocity from a subcostal evaluation of syncope or suspected ARVC. We hypoth- window. Left ventricular systolic dysfunction (SD) was deemed esized that TAPSE would be lower in dogs with ≥50 present where: LVIDs>35 mm,5,7 FS%<25%, or ejection fraction (EF%) <50%.7 EF% was measured by the Modified Simpson’s VPCs/24 h on Holter analysis compared to dogs with 24 <50 VPCs/24 h. We also hypothesized that lower method. TAPSE would be significantly associated with a shorter All TAPSE measurements were performed by a single observer (BK), blinded to the dog outcome. TAPSE was measured using time to an end-point of cardiac death. 2D anatomic M-mode on a suitable left apical 4-chamber image as previously described in dogs.19 The initial measurement was taken Materials and Methods at the onset of ventricular systole, defined as the onset of the QRS complex on simultaneous ECG. Cardiac cycles within 1 beat of an Dogs and Clinical Data atrial or ventricular premature complex were not considered suitable for measurement. Where possible, 3 consecutive measure- Records of client-owned Boxer dogs that presented to a veteri- ments of TAPSE were performed and the mean value recorded. nary teaching hospital for investigation of suspected ARVC or Intraobserver variability in TAPSE measurement was assessed other clinical signs of cardiac disease (2004–2011) were retrospec- using images from a randomly selected subset of 10 dogs, each tively reviewed. All dogs were assessed under the supervision of a measured 5 times. Coefficient of variation (CV) was calculated and Board-certified cardiologist. Dogs were eligible for inclusion if they defined as acceptable if CV<10%. had at least 1 Holter recording with a minimum of 19 hours of valid data and an echocardiographic examination within 1 month Holter ECG of Holter available for review. Dogs with documented congenital heart disease were excluded from the study, including those with Holter records were reviewed by a single operator (PM), 20 aortic blood flow velocities ≥2.25 m/s. Dogs were also excluded blinded to outcome. Dogs were dichotomized into 2 groups based if stored echocardiographic images were not suitable for measure- on Holter analysis: ≥50 VPCs/24 h and <50 VPCs/24 h. Ventricu- ment of TAPSE by off-line measurement software, or if they had a lar tachycardia, defined as >3 consecutive VPCs with a sustained concurrent diagnosis of a potentially life-limiting disease, such as heart rate >100 beats/min, was also recorded as present/absent.7 malignant neoplasia. All Boxer dogs in this study were referred to Where multiple Holter ECGs were performed, the first Holter, the cardiology service of the Royal Veterinary College for the fur- which was placed within 1 month of echocardiography, was used ther assessment of clinical signs or physical exam findings includ- for analysis. ing collapse or arrhythmia and were not screened for ARVC as part of a breed club program. Baseline clinical details obtained from medical records at the Statistical Analysis time of the first Holter recording included age, sex, body weight, and presenting clinical signs, defined as: absent; a history of syn- Statistical analyses were performed by commercially available software.bc Normality was determined graphically and using the cope, episodic weakness/collapse that was believed to be consistent Shapiro-Wilk test. Normally distributed data were represented as with syncope; or the presence/absence of congestive heart failure mean ( standard deviation) and compared using an independent (CHF) at or before initial presentation. CHF was considered pres- Æ t ent if radiographic evidence of cardiogenic pulmonary edema was samples -test. Nonnormally distributed data were represented as median (range) and compared using a Mann-Whitney U-test. Con- detected or if the dog had clinical evidence of ascites with either tinuous variables were assessed for linear association using a Pear- concurrent jugular distension/pulsation or hepatic congestion, with a clinical response to diuretic treatment. Medication that dogs son’s correlation test. Survival analysis was performed using an end-point of cardiac were receiving at the time of presentation was also documented. death. Animals dying of noncardiac disease or alive at the end of

108 584 Kaye et al the study period were right-censored. Continuous variables were evidence of systolic dysfunction. Of the 28 dogs with explored by division into groups based on quartiles, and the effect ≥50 VPCs, 20 (71%) had systolic dysfunction (Table 2). of these variables on survival was evaluated using Kaplan-Meier Among dogs presenting with syncope, ≥50 VPCs were survival analysis and Log-rank tests. Where 1 or more groups had detected in 22/39 (56%) cases. Six dogs that had ≥50 disproportionately different hazards to the others, these variables VPCs, had no history of syncope. Dogs with ≥50 VPCs were presumed to exhibit a threshold effect and were included in Cox models as categorical variables using a cut-off established by were significantly older (7.7 years, range 0.5–12.5 years) the exceptional quartile/s. For example, LA:Ao was included in than those with <50VPCs (4.9 years, range 0.9–10 years; the analysis dichotomized either side of the median value. P = .046). There was no significant difference in weight Univariable time-to-event models were constructed using Cox or sex between groups (Table 2). proportional hazards analysis. Factors where P < .2 in the univariable analysis were included in the multivariable analysis. Multivar- Tricuspid Annular Plane Systolic Excursion iable Cox proportional hazards analysis was performed in a forward stepwise manner. The assumptions of Cox proportional A minimum of 2 consecutive TAPSE measurements hazards were tested using log cumulative hazard plots and Schoen- was obtained in 45/50 (90%) of dogs, and 3 consecutive feld residuals. Overall model fit was evaluated using Cox-Snell TAPSE measurements were achievable in 27/50 (54%) residuals. Statistical significance was set at the 5% level. of stored echocardiographic studies. Intraobserver variability of TAPSE measurement was 8%. Mean TAPSE Results for the entire population was 15.2 mm ( 3.9 mm). Æ Population TAPSE exhibited a weak negative correlation with both absolute weight (r = 0.324, P = .022) and weight Eighty-one eligible Boxer dogs were evaluated for scaled to the 1/3 powerÀ (r = 0.295, P = .038; Fig 1).22 ARVC over the period (2004–2011). Thirty-one were There was no significant differenceÀ in TAPSE between excluded because of inadequate quality, poor alignment genders (P = .39). Because correlation between TAPSE of echocardiographic images for measurement of and weight was weak and unlikely to be clinically rele- TAPSE, or both; leaving 50 dogs with complete data vant (Fig 1), TAPSE was evaluated as an absolute mea- on echocardiography, 24 hours Holter and outcome surement, rather than an indexed value. Dogs where TAPSE could be appropriately measured. diagnosed with ≥50 VPCs/24 h had a significantly lower Fifty percent of the population studied was male. TAPSE (13.9 4.04 mm) than the <50 VPCs/24 h Median age was 6.0 years (0.5–12.5 years) and mean group (16.8 Æ3.21 mm; P = .008, Table 2). TAPSE weight was 30.2 kg ( 6.2 kg). Males were significantly demonstratedÆ weak to moderate correlation at the 5% Æ heavier (34.2 4.7 kg) than female dogs significance level with the following variables: ejection Æ (26.3 4.7 kg; P < .001). Syncope was the primary fraction (r = 0.354, P = .017), fractional shortening Æ presenting sign in 39/50 (78%) dogs. Exactly, 7/50 (r = 0.424, P = .002), and LA:Ao (r = 0.448, (14%) dogs presented with or had a history of CHF. P = .002). Dogs with LV systolic dysfunction hadÀ sig- Medication histories were available for all Boxer dogs nificantly lower TAPSE (14.1 3.78 mm) than those (Table 1). Thirty-five dogs were not receiving any medi- with normal LV systolic functionÆ (16.8 3.73 mm; cation, 7 dogs were receiving antiarrhythmic medica- P = .02). Also, dogs with detectable VT onÆ Holter had tion, either alone or in combination with other lower TAPSE (12.9 3.21 mm) than those without VT therapeutics, and 10 dogs were receiving a diuretic, (16.1 3.87 mm; P Æ= .008; Fig 2). either alone or in combination with other therapeutics. Æ Fifty-six percentage (28/50) of Boxer dogs had ≥50 Survival Analysis VPCs on 24 hour Holter analysis. Of the remaining 22 dogs with <50 VPCs, 11 (50%) had echocardiographic At the end of the study period, 26/50 (52%) Boxer dogs were dead; 20 dogs (40%) had suffered a cardiac Table 1. Medical treatment received by Boxer dogs at death, 6 (12%) dogs died because of noncardiac causes, the time of examination. 20 (40%) remained alive, and 4 (8%) dogs were lost to Medication Number Receiving follow-up. Median survival time to cardiac death for all dogs was 440 days (6–2,083 days). None 35 The presence of ≥50 VPCs on 24 hour Holter (yes Furosemide 10 Pimobendan 8 versus no) and the presence of systolic dysfunction Benazapril 5 (present versus absent) were included in the Cox pro- Spironolactone 4 portional hazards analysis as separate variables. The Enalapril 4 results of univariable Cox proportional hazards analysis Digoxin 4 are shown in Table 3. Lower TAPSE (<15.1 mm; cut- Sotalol 3 off based upon a threshold effect seen when dichotomiz- Amiloride 1 ing either side of median TAPSE value for the whole Amiodarone 1 population, as described above) was significantly associ- Hydrochlorothiazide 1 ated with a shorter time to cardiac death (hazard ratio Mexiletine 1 6.1, 95% confidence intervals 1.8–21; P = .004; Fig 3). Phenobarbitone 1 Ramipril 1 In the 19 dogs without echocardiographic evidence of LV systolic dysfunction, TAPSE remained significantly

109 Cardiac Death in Boxer Dogs 585

Table 2. Characteristics of the 50 Boxer dogs included in survival analysis: tricuspid annular plane systolic excursion (TAPSE) and age at diagnosis were signiﬁcantly diﬀerent between groups. Normally distributed data are represented mean ( standard deviation) and nonnormally distributed data are represented median (range). Æ Factor Non-ARVC: <50 VPCs on Holter) ARVC: ≥50 VPCs on Holter P Value Number 22 28 – Age (years) 4.9 (0.9–10.0) 7.7 (0.5–12.5) .046 Male: number (%) 9 (41%) 16 (57%) .39 Weight (kg) 28.8 ( 5.57) 31.4 ( 6.48) .14 Æ Æ Systolic dysfunction: number (%) 11 (50%) 20 (71%) .079 TAPSE (mm) 16.8 ( 3.21) 13.9 ( 4.04) .008 Æ Æ

Fig 2. Scatter plots to illustrate the association of tricuspid annular plane systolic excursion (TAPSE) with Holter variables. TAPSE was signiﬁcantly lower in dogs with ventricular tachycardia (VT) detected on Holter ECG (P = .008).

death, independent of the presence of CHF, systolic dysfunction, and detectable VT. Weight and age each lost statistical signiﬁcance when included with TAPSE <15.1 mm, which remained signiﬁcantly associated with a shorter survival time.

Discussion Our results show that measurement of TAPSE in Fig 1. Scatter plots to show correlation between tricuspid annular Boxer dogs with suspected ARVC provides useful prog- plane systolic excursion (TAPSE) and weight. Both absolute nostic information, in addition to a 24-hour Holter weight (r = 0.324, P = .022) and weight scaled to the 1/3 power recording. Multivariable analysis showed that TAPSE À 22 (r = 0.295, P = .038) exhibited a weak, negative correlation <15.1 mm remained significantly associated with a À with TAPSE. shorter time to an end-point of cardiac death, even when controlling for the presence of CHF, Holter detection of associated with a shorter time to cardiac death VT, and echocardiographic evidence of LV systolic dys- (P = .035; Fig 4). function. This is similar to findings in humans, where an Variables carried forward to the multivariable Cox absolute TAPSE measurement <18.1 mm is associated analysis were: TAPSE (<15.1 mm), weight (continuous with a greater incidence of cardiac mortality in dogs with variable), presence of ≥50VPCs on 24 hour Holter, LV heart failure.25 Although our data suggested that TAPSE systolic dysfunction, LA:Ao ratio (≥1.76), CHF (yes) exhibited weak to moderate correlation with echocardio- and VT (yes). Because 20 cardiac deaths were recorded, graphic estimates of LV systolic function, when dogs multivariable models were limited to a maximum of 2 with identifiable LV systolic dysfunction were excluded factors. To avoid violation of the assumptions of Cox from survival analysis, dogs with TAPSE <15.1 mm still proportional hazards analysis, we chose not to include had a significantly shorter survival time than those with the combination of 2 echocardiographic variables (spe- higher TAPSE. Despite a small sample size and low cifically LA size and LV systolic dysfunction) in the event rate in this group of dogs, our analysis did show a same model. significant difference, suggesting that this is a genuine Multivariable analysis (Table 4) suggested that finding which merits further prospective study and vali- TAPSE was associated with a shorter time to cardiac dation in a larger cohort of Boxer dogs.

110 586 Kaye et al

Table 3. Results of univariable Cox proportional hazards analysis to evaluate factors associated with a shorter time to cardiac death.

Hazard 95% Confidence Factor Ratio Intervals P Value TAPSE (≥15.1 mm) Ref. TAPSE (<15.1 mm) 6.1 1.8–21.0 .004 Weight: continuous 1.1 1.0–1.2 .028 per +1 kg Sex: Male = No Ref. Sex: Male = Yes 1.8 0.7–4.5 .23 ARVC diagnosis = no Ref. ARVC diagnosis = yes 7.3 2.1–25.5 .002 LV systolic Ref. dysfunction = No LV systolic 5.9 1.3–26.1 .019 Fig 4. Kaplan-Meier curve to show the effect of tricuspid annular dysfunction = Yes plane systolic excursion (TAPSE) on survival time to cardiac death LA:Ao (<1.76) Ref. in dogs without echocardiographic evidence of left ventricular sys- LA:Ao (≥1.76) 2.4 0.9–6.8 .088 tolic dysfunction. Dogs with TAPSE below the population median Presence of CHF = No Ref. (<15.1 mm) had a significantly shorter survival time to cardiac Presence of CHF = Yes 5.7 2.1–16.0 .001 death than those with higher TAPSE measurements (P = .035). Presence of VT = No Ref. Presence of VT = Yes 5.6 2.2–14.2 <.001 and breeds in 1 recent study.1 That said, the acquisition TAPSE, tricuspid annular plane systolic excursion; VPCs, of appropriate images for measurement of TAPSE ventricular premature complexes; LV, left ventricle; LA, left might be prone to error if performed by an inexperi- atrium; Ao, aorta; CHF, congestive heart failure; VT, ventricular enced operator or in a poorly compliant dog. tachycardia. Our results showed a weak negative correlation between TAPSE and weight (absolute and scaled to the 1/3 power, Fig 1). Interestingly, a previous study reported a positive correlation in a population of normal dogs of different breeds and a greater weight range.22 This discrepancy between our data and theirs is difficult to explain but might reflect the different populations used and in particular the fact that our population consisted of dogs with myocardial disease. Our study design does not allow us to explain why heavier dogs in this population might have had lower TAPSE values. Despite this correlation reaching statistical significance, we elected to analyze TAPSE as an absolute measurement rather than indexing for body weight as previously described.22 This decision was made because the correlation was weak and the scatter plots of this data suggested that it was unlikely to be clinically relevant. We also felt that an absolute measurement would have greater clinical utility than a calculated indexed Fig 3. Kaplan-Meier curve to show the effect of tricuspid annular value. We believe that this is justified in Boxer dogs, plane systolic excursion (TAPSE) on survival time to cardiac death. Boxer dogs with TAPSE below the population median because the range of body sizes among dogs of 1 breed (<15.1 mm) had a significantly shorter survival time to cardiac is more limited than that evaluated by a previous study 22 death than those with higher TAPSE measurements (P < .001). across different breeds. Another study used a single cut-off value of LVIDs >35 mm to quantify systolic This study indicates that measurement of TAPSE is dysfunction in Boxers,5 and similar single measurements straightforward and repeatable, with acceptable intraob- are commonly used among veterinary cardiologists server variability. However, during our inclusion of without indexing to a measure of body size.5,8,10,26 Body dogs for this study, we discovered that 31/81 echocar- weight was also not significantly different between the diographic studies (38%) from suitable Boxer dogs were ≥50 and <50 VPCs/24 h groups in our study, and excluded from analysis because of poor alignment or although body weight was significantly associated with image quality. This is at least in part a consequence of outcome at the univariable level, this association did the retrospective nature of this study using previously not remain present at the multivariable level. acquired images. Prospective acquisition of suitable There is currently no definitive consensus among vet- images for TAPSE measurement was successfully per- erinary cardiologists as to how ARVC in Boxer dogs is formed in 80 dogs over a wide range of body weights best diagnosed.3,7,27–29 We acknowledge that our cut-off

111 Cardiac Death in Boxer Dogs 587

Table 4. Results of multivariable Cox proportional hazards analysis, performed to evaluate the association of tricuspid annular plane systolic excursion (TAPSE) on time to cardiac death when controlling for the effect of other measured variables. Reduced TAPSE lost statistical significance when patients had R-on-T phenomenon detected on Holter, but did demonstrate a significant association with shorter survival time when controlling for the presence of congestive heart failure (CHF) and detection of ventricular tachycardia (VT).

Parameter Hazard Ratio 95% Conﬁdence Intervals P Value Model 1 TAPSE <15.1 mm 4.09 1.15–14.5 .029 ARVC diagnosis 5.18 1.43–18.7 .012 Model 2 TAPSE <15.1 mm 4.64 1.27–16.9 .02 CHF (yes) 3.21 1.11–9.29 .031 Model 3 TAPSE <15.1 mm 4.37 1.23–15.5 .023 LV systolic dysfunction 3.69 0.81–16.9 .093 Model 4 TAPSE <15.1 mm 4.36 1.22–15.6 .023 VT detected 4.05 1.54–10.7 .005

value of 50 VPCs on 24-hour Holter analysis is a low measure accurately using echocardiography.12 Despite threshold for ARVC diagnosis. However, it is more this, TAPSE appears to be a reasonable, indirect estimate likely to include Boxer dogs that are in the early stage of RV function and was significantly associated with out- of the disease. Evidence suggests that arrhythmias may come in our study. As further imaging techniques are precede the development morphologic abnormalities in studied to evaluate the right heart, including evaluating dogs with ARVC.30 Therefore, in Boxer dogs that pres- RV myocardial strain or RV geometrical changes by 3D ent for investigation of clinical signs, our TAPSE find- ultrasound technology and MRI techniques, more accu- ings should provide clinicians with a more powerful rate measures of RV function may be validated for use in prognostic indicator of survival, irrespective of VPC dogs. However, TAPSE has the benefit of accessibility to number. Our results suggest that TAPSE is a clinically a greater number of practitioners, being simple to per- useful measurement, giving additional relevant informa- form with standard echocardiographic equipment and an tion whichever cut-off value the clinician should choose appropriate level of operator experience. to diagnose ARVC. In conclusion, our results indicate that TAPSE Day-to-day variation of VPC number on Holter can be <15.1 mm in Boxer dogs with ≥50 VPCs/24 h on Holter >80% in dogs with ARVC,31 and there can be significant analysis is associated with a shorter time to cardiac annual variation in the number of VPCs detected.3 Some death, irrespective of the presence of CHF, echocardio- dogs with a diagnosis of ARVC have fewer than 50 VPCs graphic evidence of LV systolic dysfunction, and Holter the year after a Holter showing ≥300 VPCs, despite detection of VT. This easy to obtain and repeatable receiving no antiarrhythmic medication whatsoever.3 It measurement therefore gives the clinician valuable prog- would therefore be interesting to evaluate TAPSE in a nostic information in addition to that provided by Hol- longitudinal fashion in a similar cohort of dogs. ter analysis alone. Other limitations of this study are typical for any retrospective data analysis. Boxer dogs had diagnostic tests and treatment provided based upon individual decision making of the primary clinician managing each case at Footnotes the time it is presented. Differences in treatment and diagnostic protocol could not be accounted for in our 1 Visser LC, Scansen BA, Brown NV, et al. ACVIM Forum analysis. An inability to obtain 3 consecutive TAPSE Research Abstracts Program. J Vet Intern Med. 2014;28:1000. measurements in 46% of stored images was the result (abstract C.2). a of frequent ventricular arrhythmias in many of the Vivid 7 with EchoPac off-line measurement software, GE Boxer dogs and could have impacted on the accuracy systems, Hatfield, United Kingdom. b IBMÒ SPSSÒ Statistics Version 21; for Windows 7, IBM (UK) of the TAPSE measurement. However despite this, Ltd, Portsmouth, UK. TAPSE remained significant following multivariable c GraphPad Prism Version 6.0; for Windows 7, GraphPad Soft- analysis when applied to affected dogs in a working ware Inc. Sandiego, CA, USA. clinical environment. Misclassification of clinical status, such as the presence/absence of CHF, and outcome (cardiac versus noncardiac death) are possible, despite robust criteria. A further limitation is that these dogs Acknowledgments were placed on multiple different treatment protocols, which may have altered over time as their disease pro- The authors thank Yu-Mei Chang for her statistical gressed. The study was not designed to assess the effect advice. of medication on TAPSE. Funding: This research was not supported by a grant Because of its complex geometry with dense trabecula- or any other source of funding. This data have not been tion and load dependence, the RV is challenging to presented at any congress or meeting.

112 588 Kaye et al

Conflict of Interest Declaration: Drs. Connolly, Luis acute symptomatic pulmonary embolism. J Thromb Haemost Fuentes, and Borgeat have performed consultancy work 2014;12:1020–1027. for Boehringer Ingelheim, Novartis Animal Health, and 16. Finocchiaro G, Knowles JW, Pavlovic A, et al. Prevalence Royal Canin. and clinical correlates of right ventricular dysfunction in patients Off-label Antimicrobial Declaration: Authors declare with hypertrophic cardiomyopathy. Am J Cardiol 2014;113:361– 367. no off-label use of antimicrobials. 17. Lopez-Candales� A, Rajagopalan N, Saxena N, et al. Right ventricular systolic function is not the sole determinant of tricus- References pid annular motion. Am J Cardiol 2006;98:973–977. 18. Baumwart RD, Meurs KM, Bonagura JD. Tei index of 1. Marcus FI, McKenna WJ, Sherrill D, et al. Diagnosis of ar- myocardial performance applied to the right ventricle in normal rhythmogenic right ventricular cardiomyopathy/dysplasia: Pro- dogs. J Vet Intern Med 2005;19:828–832. posed modification of the task force criteria. Circulation 19. Pariaut R, Saelinger C, Strickland KN, et al. Tricuspid 2010;121:1533–1541. annular plane systolic excursion (TAPSE) in dogs: Reference val- 2. Meurs KM, Stern JA, Sisson DD, et al. Association of ues and impact of pulmonary hypertension. J Vet Intern Med dilated cardiomyopathy with the striatin mutation genotype in 2012;26:1148–1154. Boxer dogs. J Vet Intern Med 2013;27:1437–1440. 20. Bussadori C, Amberger C, Le Bobinnec G, et al. Guidelines 3. Meurs KM, Stern JA, Reina-Doreste Y, et al. Natural history for the echocardiographic studies of suspected subaortic and pul- of arrhythmogenic right ventricular cardiomyopathy in the Boxer monic stenosis. J Vet Cardiol 2000;2:15–22. dog: A prospective study. J Vet Intern Med 2014;28:1214–1220. 21. Hansson K, H€aggstrom€ J, Kvart C, et al. Left atrial to 4. Basso C, Fox PR, Meurs KM, et al. Arrhythmogenic right aortic root indices using two-dimensional and M-mode echocar- ventricular cardiomyopathy causing sudden cardiac death in Boxer diography in Cavalier King Charles Spaniels with and without dogs: A new animal model of human disease. Circulation left atrial enlargement. Vet Radiol Ultrasound 2002;43:568– 2004;109:1180–1185. 575. 5. Palermo V, Johnson MJS, Sala E, et al. Cardiomyopathy in 22. Thomas WP, Gaber CE, Jacobs GJ, et al. Recommenda- Boxer dogs: A retrospective study of the clinical presentation, tions for standards in transthoracic two-dimensional echocardiog- diagnostic findings and survival. J Vet Cardiol 2011;13:45–55. raphy in the dog and cat. J Vet Intern Med 1993;7:247–252. 6. Harpster NK. Boxer Cardiomyopathy. In: Kirk R, ed. Cur- 23. Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations rent Veterinary Therapy VIII. Philadelphia, PA: WB Saunders; regarding quantitation in M-mode echocardiography: Results of a 1983:329–337. survey of echocardiographic measurements. Circulation 7. Motsk~ ula€ PF, Linney C, Palermo V, et al. Prognostic value 1978;58:1072–1083. of 24-hour ambulatory ECG (Holter) monitoring in Boxer dogs. J 24. Boon JA. Veterinary Echocardiography, 2nd ed. West Sus- Vet Intern Med 2013;27:904–912. sex, UK: John Wiley & Sons; 2011:206–207. 8. Saguner AM, Vecchiati A, Baldinger SH, et al. Different 25. Vizzardi E, D’Aloia A, Bordonali T, et al. Long-term prog- prognostic value of functional right ventricular parameters in ar- nostic value of the right ventricular myocardial performance index rhythmogenic right ventricular cardiomyopathy/dysplasia. Circ compared to other indexes of right ventricular function in patients Cardiovasc Imaging 2014;7:230–239. with moderate chronic heart failure. Echocardiography 9. Dini FL, Conti U, Fontanive P, et al. Right ventricular dys- 2012;29:773–778. function is a major predictor of outcome in patients with moderate 26. Smets P, Daminet S, Wess G. Simpson’s Method of to severe mitral regurgitation and left ventricular dysfunction. Am Discs for Measurement of Echocardiographic End-Diastolic Heart J 2007;154:172–179. and End-Systolic Left Ventricular Volumes: Breed-Specific Refer- 10. Gupta S, Khan F, Shapiro M, et al. The associations ence Ranges in Boxer Dogs. J Vet Intern Med 2014;28:116– between tricuspid annular plane systolic excursion (TAPSE), ven- 122. tricular dyssynchrony, and ventricular interaction in heart failure 27. Stern JA, Meurs KM, Spier AW, et al. Ambulatory electro- patients. Eur J Echocardiogr 2008;9:766–771. cardiographic evaluation of clinically normal adult Boxers. J Am 11. Mercer-Rosa L, Parnell A, Forfia PR, et al. Tricuspid Vet Med Assoc 2010;236:430–433. annular plane systolic excursion in the assessment of right ventric- 28. Spier AW, Meurs KM. Assessment of heart rate variability ular function in children and adolescents after repair of tetralogy in Boxers with arrhythmogenic right ventricular cardiomyopathy. of Fallot. J Am Soc Echocardiogr 2013;26:1322–1329. J Am Vet Med Assoc 2004;224:534–537. 12. Haddad F, Hunt SA, Rosenthal DN, et al. Right ventricu- 29. Oxford EM, Danko CG, Fox PR, et al. Change in b-cate- lar function in cardiovascular disease, part I: Anatomy, physiol- nin localization suggests involvement of the canonical Wnt path- ogy, aging, and functional assessment of the right ventricle. way in Boxer dogs with arrhythmogenic right ventricular Circulation 2008;117:1436–1448. cardiomyopathy. J Vet Intern Med 2013;28:92–101. 13. Ueti OM, Camargo EE, de A Ueti A, et al. Assessment of 30. Baumwart RD, Meurs KM, Raman SV. Magnetic reso- right ventricular function with doppler echocardiographic indices nance imaging of right ventricular morphology and function in derived from tricuspid annular motion: Comparison with radionu- Boxer dogs with arrhythmogenic right ventricular cardiomyopathy. clide angiography. Heart 2002;88:244–248. J Vet Intern Med 2009;23:271–274. 14. Forfia PR, Fisher MR, Mathai SC, et al. Tricuspid annular 31. Spier AW, Meurs KM. Evaluation of spontaneous variabil- displacement predicts survival in pulmonary hypertension. Am J ity in the frequency of ventricular arrhythmias in Boxers with ar- Respir Crit Care Med 2006;174:1034–1041. rhythmogenic right ventricular cardiomyopathy. J Am Vet Med 15. Lobo JL, Holley A, Tapson V, et al. Prognostic significance Assoc 2004;22:538–541. of tricuspid annular displacement in normotensive patients with

113

III McAulay G, Borgeat K, Sargent J, Mõtsküla PF, Neves J, Dukes- McEwan J, Luis Fuentes V., 2018. PHENOTYPIC DESCRIPTION OF CARDIAC FINDINGS IN A POPULATION OF DOGUE DE BORDEAUX WITH AN EMPHASIS ON ATRIAL FIBRILLATION. Veterinary Journal 234, 111–118.

The Veterinary Journal 234 (2018) 111–118

Contents lists available at ScienceDirect

The Veterinary Journal

journal homepage: www.elsevier.com/locate/tvjl

Original Article

Phenotypic description of cardiac ﬁndings in a population of Dogue de

Bordeaux with an emphasis on atrial ﬁbrillation

a, b,1 b,2 b,3 c c

G. McAulay *, K. Borgeat , J. Sargent , P. Mõtsküla , J. Neves , J. Dukes-McEwan ,

V. Luis Fuentes

Cardio-respiratory Referrals, New Priory Vets Brighton, BN1 8QR, UK

Royal Veterinary College, Department of Clinical Science and Services, Hatﬁeld AL9 7TA, UK

Small Animal Teaching Hospital, Institute of Veterinary Science, University of Liverpool, Chester High Road, Neston CH64 7TE, UK

A R T I C L E I N F O A B S T R A C T

Article history:

Accepted 21 February 2018 The aim of this study was to describe the clinical phenotype of Dogue de Bordeaux (DdB) referred for

cardiac investigation, with particular reference to the prevalence of atrial ﬁbrillation and associated

Keywords: features. Review of canine medical records of two United Kingdom veterinary referral hospitals identiﬁed

Canine 64 DdB with available echocardiographic and electrocardiographic (ECG)/Holter data. Atrial ﬁbrillation

Cardiomyopathy was documented in 25 (39%) dogs and supraventricular tachycardia was recorded in ﬁve (7.8%) dogs. In a

Dogue de Bordeaux subset of 34 dogs, excluding congenital heart disease (n = 17), presence of a cardiac mass (n = 7) and non-

Supraventricular tachycardia

cardiac neoplasia (n = 6), 19 (56%) dogs had atrial ﬁbrillation, with a median heart rate of 200 beats per

Subaortic stenosis

min (bpm) on presentation. Atrial ﬁbrillation was inconsistently associated with cardiac chamber

remodelling, but was frequently associated with systolic dysfunction (13/19, 68.4%) and right sided atrial

or ventricular dilatation (14/19, 73.7%) in dogs with atrial ﬁbrillation in this subset. No dogs in this subset

had right sided atrial or ventricular dilatation in the absence of supraventricular arrhythmia or systolic

dysfunction. The absence of structural heart disease in some dogs with supraventricular arrhythmias

suggests that an underlying primary arrhythmic process might be responsible for initiating remodelling,

although a primary cardiomyopathy cannot be ruled out.

Introduction tachyarrhythmia has been suggested (Locatelli et al., 2011). In a

study by Ohad et al. (2013), all DdB diagnosed with tricuspid

The Dogue de Bordeaux (DdB) is a large brachycephalic breed of dysplasia had AF.

dog, originating in France. The population and cardiac pathology of Atrial ﬁbrillation has a complex pathophysiology, whereby

DdB in the United Kingdom (UK) has not been well described; functional and structural mechanisms promote electrical re-entry

however, anecdotal reports suggest that atrial ﬁbrillation (AF) is (Brundel et al., 2005). Atrial ﬁbrillation is frequently associated

prevalent in this breed and a recent study suggested that the breed with heart disease; increased atrial dimensions, inﬂammation,

may have a high incidence of sudden and unexpected death ﬁbrosis, alterations in autonomic tone and ion channel expression

(McAulay et al., 2018). The DdB is affected by a range of cardiac promote development of AF. Large breeds of dog might be

conditions, including subaortic stenosis (SAS), tricuspid dysplasia predisposed to AF because their atrial dimensions provide the

and dilated cardiomyopathy (DCM) (Borgarelli et al., 2006; critical atrial mass required to support wavelet re-entry (Moore

Höllmer et al., 2008; Martin et al., 2009; Oliveira et al., 2011; and Spear, 1987; Guglielmini et al., 2000). Atrial ﬁbrillation in the

Ohad et al., 2013). A breed predisposition to supraventricular absence of structural heart disease has been described as ‘lone’ AF

(Menaut et al., 2005). Frequent arrhythmia, including AF, can

promote cardiac chamber remodelling and a dilated cardiomyop-

athy (DCM) phenotype, designated tachycardia-induced cardio-

* Corresponding author.

E-mail address: [email protected] (G. McAulay). myopathy (Shinbane et al., 1997).

Present address: Langford Vets, University of Bristol, Bristol BS40 5DU, UK. Since AF can be a cause, or consequence, of structural or

Present address: Southern Counties Veterinary Specialists, Ringwood, Hamp- functional disease, clarifying the association of AF with phenotype

shire BH24 3JW, UK.

ﬁ

3 may be helpful in de ning the underlying disease process. This

Present address: Anderson Moores Veterinary Specialists, Hursley, Winchester,

study sought to describe the phenotype of DdB within a cardiac

Hampshire SO21 2LL, UK.

https://doi.org/10.1016/j.tvjl.2018.02.015

117

112 G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118

Statistical analysis was performed using Prism 6 (GraphPad Software). Data

referral population, speciﬁcally with reference to the prevalence

were assessed for normality visually and using the D’Agostino and Pearson omnibus

and associated features of AF. We hypothesised that AF is the

normality test. Normally distributed data were expressed as mean standard

predominant arrhythmia in DdBs undergoing cardiac investigation Æ

deviation (SD). Non-normally distributed data were expressed as median and inter-

ﬂ

and that AF re ects a primary arrhythmia, rather than consequence quartile range (IQR). Normally distributed, continuous data were assessed by one-

of structural heart disease. We hypothesised that AF would occur in way analysis of variance (ANOVA). Non-parametric data were compared using the

Kruskal–Wallis test and Dunns multiple comparisons test. Fisher’s exact test was

some dogs in the absence of cardiac chamber remodelling.

used to compare differences in proportions of categorical data. Statistical

signiﬁcance was set at P = 0.05.

Materials and methods Results

Study group

This study was approved by the Royal Veterinary College Ethics and Welfare The study included 64 DdB (Fig. 1). Presenting clinical signs are

Committee, (approval number 1296; date of approval 16th October 2014). The

summarised in Table 1. Six of 64 (9.4%) DdB had a prior diagnosis of

records of The Queen Mother Hospital for Animals, Royal Veterinary College,

non-cardiac neoplasia and were examined to evaluate myocardial

Hatﬁeld, UK, and Small Animal Teaching Hospital, University of Liverpool, Neston,

UK, were searched for DdB presenting to the cardiology services from March 2005 function prior to chemotherapy. Seven of 64 (10.9%) DdB had a

to July 2013. Dogs were excluded if echocardiographic or electrocardiographic (ECG) cardiac mass identiﬁed echocardiographically. No dogs with a

data were unavailable.

cardiac mass had a prior diagnosis of non-cardiac neoplasia.

Cardiac masses were in variable locations and all were suspected to

Echocardiography

be neoplastic, but only one case (with haemangiosarcoma) was

ﬁ

Echocardiographic examinations were performed as described by Thomas et al. con rmed using histopathology. Seventeen of 64 DdB were

(1993), using a Vivid 7 (GE Healthcare), with simultaneous ECG recording, and diagnosed with congenital defects. Twelve of 64 (18.8%) had SAS

reviewed and measured using EchoPAC Clinical Workstation Software (GE

(median velocity 3.47 m/s, IQR 2.6–5.4). Five dogs had other

Healthcare). Left ventricular measurements were obtained in M-mode using the

congenital defects, including two with patent ductus arteriosus,

leading edge to leading edge method (Sahn et al., 1978). The mean of three

measurements for dogs in sinus rhythm and ﬁve measurements for dogs with and one each with pulmonic stenosis, ventricular septal defect and

arrhythmia were recorded for each variable. Left ventricular systolic and diastolic atrial septal defect.

measurements were normalised to body weight (Cornell et al., 2004). Normalised

Supraventricular arrhythmia was identiﬁed in 30/64 (46.9%)

left ventricular systolic diameter (LVDsn) measurements exceeding 95% conﬁdence

DdB, persistent AF in 25/64 (39.1%) and SVT in 5/64 (7.8%). Dogs

interval (CI) predicted values were deﬁned as a marker of systolic dysfunction

with supraventricular arrhythmia had higher heart rates at

(LVDsn > 1.26). Normalised left ventricular diastolic diameter (LVDdn) measure-

ments exceeding 95% CI predicted values were deﬁned as a marker of ventricular presentation, more severe left ventricular dilatation and systolic

dilatation (LVDdn >1.85). dysfunction, reduced FS% and increased PEP:ET compared to dogs

The pre-ejection period to ejection time ratio (PEP:ET) was calculated from

in sinus rhythm (Table 2). Dogs with supraventricular arrhythmia

Doppler interrogation of aortic ﬂow obtained from the left apical ﬁve chamber or

were more likely to have left and/or right sided atrial dilatation, or

subcostal view (de Madron, 2015). Left atrial diameter was indexed to aortic

diameter (LA:Ao) (Hansson et al., 2002). Right sided dilatation was deﬁned as right left ventricular dilatation, than those in sinus rhythm ( = 0.001).

ventricular diastolic dimensions exceeding 50% of the diameter of the left ventricle Dogs with right sided atrial or ventricular dilatation had higher

or right atrial dimensions that were subjectively larger than the left atrium in any heart rates than those without atrial dilatation (190 77 bpm

view (Bright et al., 2005; Menaut et al., 2005; Palermo et al., 2011). If dogs exhibited Æ

versus 125 52 bpm; P = 0.001). There was no signiﬁcant differ-

left ventricular dilatation, right ventricular dilatation was assessed subjectively on Æ

ence in the frequency of supraventricular arrhythmia between

the basis of clinical experience.

Pulmonary hypertension (systolic) was diagnosed when Doppler-derived dogs with or without SAS (P = 0.117), other congenital disease

tricuspid regurgitation velocity exceeded 3.1 m/s (>38 mmHg) (Stepien, 2009) (P > 0.999), a cardiac mass (P = 0.433) or non-cardiac neoplasia

and was classiﬁed as mild (<50 mmHg), moderate ( 50 mmHg to <80 mmHg) or

(P = 0.1094). No dogs with a prior diagnosis of non-cardiac

severe ( 80 mmHg). The aortic valve, outﬂow tract and Doppler proﬁle were

neoplasia had supraventricular arrhythmia.

assessed; velocities > 2.5 m/s were considered to be consistent with SAS (Bussadori

Systolic dysfunction was observed in 21/64 (32.8%) DdB and left

et al., 2000; Höllmer et al., 2008). Pulmonic valve morphology was assessed

according to Bussadori et al. (2000); pulmonic stenosis was deﬁned by pulmonic ventricular dilatation in 13/64 (20%) DdB, occurring concurrently

velocities > 2.25 m/s. in 11/64 (17.2%) cases. Systolic dysfunction was evident in 5/34

(14.7%) DdB classiﬁed as having sinus rhythm (no evidence of

Electrocardiography

ECG data were collected from the original ECG, Holter, echocardiographic

timing ECG and attending clinicians’ records. Holter data were quantitatively

analysed by Laboratory Corporations of America, Ambulatory Monitoring Services.

For initial analysis, AF and supraventricular tachycardia (SVT) were described

collectively as ‘supraventricular arrhythmia’ (Santilli et al., 2008). Atrial ﬁbrillation

was diagnosed when QRS complexes occurred without periodicity in the absence of

P waves or ﬂutter waves (Miller et al., 1999). Supraventricular tachycardia was

deﬁned as three or more QRS complexes < 0.07 s duration, occurring regularly at an

instantaneous rate > 160 beats per min (bpm), paroxysmally or persistently, and

deemed to be physiologically inappropriate by the attending clinician or on review

of Holter data with reference to the Holter activity diary. Dogs in sinus rhythm with

no evidence of paroxysmal supraventricular arrhythmia on review were classed as

‘sinus rhythm’. Ventricular arrhythmia was described as ‘isolated’ if only individual

ventricular premature complexes (VPCs) were identiﬁed, or ‘complex’ if couplets,

triplets, bigeminy, trigeminy and/or ventricular tachycardia were documented.

Dogs with only isolated VPCs on Holter were dichotomised with a cut off of 100

VPCs/24 h.

Statistical analysis

After initial assessment of the whole population, dogs with congenital heart

disease, cardiac masses and non-cardiac neoplasia were excluded from further

assessment, and the clinical, echocardiographic and ECG ﬁndings of the remaining

subset (non-CHDN) were explored further. Fig. 1. Summary of 64 Dogue de Bordeaux (DdB) presented for cardiac evaluation.

RVC, Royal Veterinary College; UoL, University of Liverpool.

118

G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118 113

Table 1

2/3 dogs (median SVT 240 bpm) (Fig. 2; Table 3). The three dogs

Signalment and presenting clinical signs in study group of 64 Dogue de Bordeaux.

with SVT were younger than the 31 dogs with other arrhythmias

n (P = 0.027; Fig. 3). Non-CHDN dogs with AF (n = 19) had a median

Signalment heart rate of 200 bpm at presentation, with 17/19 (89%) of these

Age (years) 4.3 ( 2.1) dogs having a heart rate 160 bpm. Three (15.8%) non-CHDN dogs

Sex (male:female) 49:15 with AF did not have any chamber dilatation and were considered

Weight (kg) 57.1 ( 8.4)

Æ not to have structural cardiac heart disease, although one was

classiﬁed with systolic dysfunction. Only one of these three dogs

Clinical signs

Auscultated arrhythmia 33 (51.6%) had a heart rate < 160 bpm (84 bpm). Four non-CHDN dogs in sinus

Exercise intolerance 26 (41%) rhythm had systolic dysfunction, with 2/4 demonstrating >1000

Anorexia 24 (37.5%)

VPCs/24 h on Holter. Overall, 18/34 (52.9%) non-CHDN dogs

Systolic murmur 21 (32.8%)

demonstrated systolic dysfunction and 18/34 (52.9%) demonstrat-

Dyspnoea 21 (32.8%)

ed right sided dilatation (atrial or ventricular), with 38% showing

Collapse 20 (31.3%)

Abdominal distension 16 (25%) evidence of both systolic dysfunction and right sided dilatation. Of

Weight loss 16 (25%) non-CHDN dogs with AF, 13/19 (68.5%) had systolic dysfunction

Cough 7 (10.9%)

and 14/19 (73.7%) had right sided dilatation. No non-CHDN dogs in

Gallop sound 1 (1.6%)

sinus rhythm had right sided dilatation without concurrent systolic

Normally distributed data expressed as mean standard deviation (SD).

Æ dysfunction (Fig. 4). Left heart dilatation and systolic dysfunction

were present in 9/34 (26%) non-CHDN dogs and ﬁve (55%) of these

had AF. Normalised left ventricular dimensions and indexed left

paroxysmal supraventricular arrhythmia on review). Pulmonary atrial dimensions within the non-CHDN dogs are displayed in

hypertension was documented in 9/64 (14%) DdB, of which 4/9 had Fig. 5. Pulmonary hypertension was identiﬁed in six non-CHDN

right sided atrial or ventricular dilatation. Collapse was reported in dogs, four with mild and two with moderate severity; the latter

20/64 (31.3%) DdB and was not associated with pulmonary two had AF and right sided atrial/ventricular dilatation. The

hypertension (P = 0.252). Holter data were available for 16/64 prevalence of collapse did not differ between sinus, SVT or AF

(25.0%) DdB, of which 9/16 (56.3%) had AF. Ventricular arrhythmia rhythm (P = 0.958).

was detected in 15/64 (23.4%) DdB, of which eight had isolated

VPCs and seven had complex arrhythmias (Table 2). Holter data Discussion

were available for 13/15 (86.7%) dogs with ventricular arrhythmia,

with a median of 105 (IQR 5-543) VPCs/24 h. In dogs with isolated In this study supraventricular arrhythmias were detected in a

VPCs on Holter, 6/7 (85.7%) had less than 100 VPCs/24 h. high proportion of DdBs assessed by veterinary cardiologists at two

In the subset of 34 non-CHDN dogs; 19 (56%) had AF, 12 (35%) UK referral hospitals. Atrial ﬁbrillation was documented in 39% of

had sinus rhythm and three (9%) had SVT, which was paroxysmal in all DdB. In the non-CHDN subset (DdB without congenital heart

Table 2

Signalment and echocardiographic and electrocardiographic ﬁndings in 64 Dogue de Bordeaux with sinus rhythm and supraventricular arrhythmia.

Sinus rhythm Supraventricular arrhythmia P value

Presentation

Number of cases 34 30

RVC cases:UoL cases 20:14 23:7 0.183

Age (years) 4.7 2.34 3.9 2.0 0.13

Æ Æ

Sex (male:female) 25:9 24:6 0.571

Weight (kg) 56.3 7.1 58.1 9.7 0.387

Æ Æ

HR presentation 104 [84–129] 190 [120–240] <0.001

History of collapse or syncope 10/24 (29.4%) 10/20 (33.3%) 0.791

Echocardiography

LA/Ao 1.31 [1.18–1.50] 1.8 [1.48–2.06] <0.001

LVDdn 1.5 0.34 1.71 0.37 0.032

Æ Æ

LVDsn 1.00 0.30 1.34 0.34 0.002

Æ Æ

FS% 28.6 9.9 15.0 9.0 <0.001

Æ Æ

PEP:LVET 0.41 0.10 0.57 0.13 <0.001

Æ Æ

TRV (m/s) 2.8 [2.6–3.1] 2.9 [2.3–3.0] 0.912

Right sided dilatation 6/27 (18.1%) 21/8 (72%) 0.001

No remodelling (No LVD, LAD or RSD) 23 (67%) 4 (13.3%) 0.007

Rhythm

HR ECG/Holter mean 80 [68–125] 135 [98–240] 0.032

Ventricular arrhythmia 5 10 0.138

Complex ventricular arrhythmia (couplets, triplets bigeminy, trigeminy and/or ventricular tachycardia) 4 3 >0.999

Speciﬁc diagnosis

SAS 9 (26.5%) 3 (10%) 0.119

CHD (not SAS) 3 (8.8%) 2 (6.7%) >0.999

Cardiac mass 5 (14.7%) 2 (6.7%) 0.433

Non-cardiac neoplasia 6 (17.6%) 0 (0%) 0.016

Normally distributed data expressed as mean standard deviation (SD) and non-normally distributed data are expressed as median [inter-quartile range].

RVC, Royal Veterinary College; UoL, University of Liverpool; HR, heart rate; LA, left atrial diameter; Ao, aortic diameter; LVDdn, left ventricular diameter diastole normalised;

LVDsn, left ventricular diameter systole normalised; FS, fractional shortening; PEP, pre-ejection period; ET, left ventricular ejection time; TRV, tricuspid regurgitant velocity;

LVD, left ventricular dilatation; LAD, left atrial dilatation; RSD, right sided dilatation; SAS, subaortic stenosis.

119

114 G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118

heart rates have been reported in dogs with AF without structural

or functional heart disease (Menaut et al., 2005). In our study,

only one non-CHDN dog with AF without structural or functional

disease had a heart rate < 160 bpm on presentation. Whether fast

AF rates at presentation reﬂect more advanced underlying

disease, stress or a breed propensity to tachyarrhythmia remains

unclear.

Nine of 34 (26%) non-CHDN DdB had increased left heart

dimensions and systolic dysfunction, consistent with a dilated

cardiomyopathy (DCM) phenotype, and 5/9 (55%) of these dogs

had AF. DCM has been characterised by systolic and diastolic

dysfunction, and variable ventricular dilatation (Kittleson, 1998;

Tidholm et al., 2001), which can precipitate valvular incompetence,

atrial dilatation and arrhythmias, particularly AF (Dukes-McEwan

et al., 2003). The prevalence of AF in a previous retrospective study

of DCM was 45%, with a high prevalence in the Irish wolfhound,

Great Dane and Newfoundland (Tidholm and Jönsson, 1996;

Vollmar, 2000; Meurs et al., 2001a; Martin et al., 2009). DCM has

been reported previously in the DdB; however, the inconsistent

relationship of AF to the DCM phenotype in this population

suggests that left ventricular dilatation does not always precede

development of AF.

Atrial ﬁbrillation occurs frequently in the Irish wolfhound, with

Fig. 2. Summary of congenital heart disease, cardiac mass and non-cardiac

and without DCM, and is reported in up to 87.6% of Irish

neoplasia (CHDN) in the study group of Dogue de Bordeaux (DdB) and development

wolfhounds with a diagnosis of DCM (Vollmar, 2000). Studies in

of subset of dogs without congenital heart disease, cardiac mass and non-cardiac

neoplasia (non-CHDN). SAS, subaortic stenosis. Irish wolfhounds support an association between AF and DCM,

which develops in the majority of dogs with AF (Simpson et al.,

ﬂ

disease or neoplasia), the prevalence of atrial ﬁbrillation was 56%. 2016). Whether DCM in the Irish wolfhound re ects a primary

A previous Italian study reported AF in 3/39 (7.7%) apparently cardiomyopathy, atrial dysfunction or sustained tachycardia is

healthy DdB (Locatelli et al., 2011 ). Since dogs in our study were unclear (Brownlie and Cobb, 1999). Familial DCM associated with

drawn from a cardiology referral population with a high prevalence early AF has been reported in human beings (Sébillon et al., 2003),

of auscultated arrhythmia and exercise intolerance, it is unsur- but sustained tachyarrhythmia, including AF, may result in

prising that the prevalence of AF in our study was higher than in a biventricular systolic dysfunction and eccentric dilatation (Wilson

general population; however, our ﬁndings support the hypothesis et al., 1987; Shinbane et al., 1997; Schoonderwoerd et al., 2001;

that AF is the predominant arrhythmia in DdB undergoing cardiac Dukes-McEwan et al., 2003).

investigation in the UK. A feature of DdBs in this study was the high frequency of right

In the non-CHDN subset, 3/19 DdB with AF had a structurally sided atrial or ventricular dilatation, which were evident in 27/64

normal heart. Atrial ﬁbrillation in the absence of structural (42.1%) of all dogs and over half of non-CHDN dogs in the absence

cardiac disease has been described as ‘lone’ AF (Takemura et al., of severe pulmonary hypertension. In a retrospective study of DCM

2002; Menaut et al., 2005). However, the term ‘lone’ AF has been in the Irish wolfhound, predominantly right sided dilatation was

questioned because of inconsistent deﬁnition and improved reported in 5.4% dogs and right sided changes only in 4.2% of dogs

ability to detect predispositions to AF (Wyse et al., 2014). Low (Vollmar, 2000).

Table 3

Signalment, echocardiographic and electrocardiographic ﬁndings in subset of Dogue de Bordeaux excluding congenital heart disease, cardiac mass and non-cardiac neoplasia.

Sinus AF SVT P value

Signalment

n 12 19 3

Age (years) 4.9 [3.1–6.2] 4.1 [3.1–5.6] 1.5 [1–2.3] 0.027

Sex (male:female) 10:2 15:4 2:1 (66.6% male) 0.813

Weight (kg) 53.5 [48.4–54.7] 61.9 [52.9–67.6] 50.2 [45.1–66.0] 0.153

HR presentation 128 [120–170] 200 [160–240] 120 [84–360] 0.054

History of collapse or syncope 5 (41%) 8 (42.1%) 1 (33.3%) 0.959

Echocardiography

LA/Ao 1.39 [1.21–1.56] 1.83 [1.59–2.16] 1.31 [1.27–1.33] 0.004

LVDdn 1.42 [1.25–2.05] 1.37 [1.27–1.57] 1.47 [1.4–1.55] 0.324

LVDsn 1.04 [0.83–1.51] 1.37 [1.27–1.57] 1.14 [1.02–1.26] 0.298

FS (%) 25.3 [16.4–28.0] 11.7 [7.81–21.1] 16.5 [11.9–21.0] 0.017

PEP:LVET 0.48 [0.36–0.57] 0.57 [0.54–0.69] 0.41 [0.4–0.57] 0.017

TRV (m/s) 2.58 [2.58–3.00] 3 [2.4–3.3] 2.9 [2.8–3.0] 0.79

Left-sided dilatation 4 (33.3%) 5 (26.3%) 0 (0%) 0.504

Right sided dilatation 2 (16.7%) 14 (73.7%) 2 (66.6%) 0.007

Data are expressed as median [interquartile range]. AF, atrial ﬁbrillation; SVT, supraventricular tachycardia; HR, heart rate; LA, left atrial diameter; Ao, aortic diameter; LVDdn,

left ventricular diameter diastole normalised; LVDsn, left ventricular diameter systole normalised; FS, fractional shortening; PEP, pre-ejection period; ET, left ventricular

ejection time; TRV, tricuspid regurgitant velocity. Groups were compared with a Kruskal–Wallis test and the P value is reported with pair-wise comparisons using Dunn’s multiple comparisons test.

Within each row, data that were not signiﬁcantly different between the two indicated groups.

120

G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118 115

arrhythmia has been documented in 32% of apparently normal

dogs (Meurs et al., 2001b), with normal Dobermans and Boxers

reportedly having a ventricular burden < 50 and < 91VPCs/24 h,

respectively (Stern et al., 2010; Wess et al., 2010). On the basis of

a dichotomised cut-off of 100 VPCs/24 h, approximately half of

the DdB with ventricular arrhythmia in our study had isolated

arrhythmia of a low frequency.

Atrial ﬁbrillation has been described previously in DdBs with

tricuspid dysplasia (Ohad et al., 2013). Tricuspid dysplasia in dogs

has been deﬁned by echocardiographic characteristics including

leaﬂet thickening, tethering and valvular insufﬁciency, as assessed

by right atrial dilatation and regurgitant jet area (Bonagura et al.,

1999; Oyama and Sisson, 2001; Ohad et al., 2013). In human beings,

functional tricuspid insufﬁciency due to annular distortion and

Fig. 3. Scatter plot of age associated with rhythm diagnosis in subset of 34 Dogue de alterations in papillary muscle geometry has been reported to be

Bordeaux (DdB) without congenital heart disease, cardiac mass or non-cardiac secondary to right and left ventricular dilatation, systolic

neoplasia (non-CHDN). AF, atrial ﬁbrillation; SVT, supraventricular tachycardia.

dysfunction and respiratory phase (Rogers and Bolling, 2009;

Topilsky et al., 2014; Tornos Mas et al., 2015). The assessment of

Right ventricular dilatation has been associated with

tricuspid insufﬁciency by regurgitant jet area has been discouraged

arrhythmogenic right ventricular cardiomyopathy (ARVC) in

because of technical limitations (Lancellotti et al., 2010). Alter-

dogs. In retrospective studies of Boxer cardiomyopathy, 21.5–

ations in papillary muscle geometry result in the echocardio-

43% of dogs had right ventricular dilatation (Basso et al., 2004;

graphic appearance of leaﬂet tethering and reduced valvular

Palermo et al., 2011). It is possible DdB in this study were

coaptation (Spinner et al., 2012). In human beings with chronic AF

affected by a cardiomyopathy with a predisposition for right

in the absence of organic heart disease, the development and

sided pathology. A hallmark of ARVC is frequent ventricular

severity of tricuspid regurgitation has been associated with AF type

arrhythmia, often with complexity, including couplets, triplets

and duration (Najib et al., 2011; Park et al., 2015). In view of the

and ventricular tachycardia (Harpster, 1991; Meurs et al., 1999;

unknown inﬂuence of these factors in dogs, we could not

Basso et al., 2004; Palermo et al., 2011). In our study, 15/64

conﬁdently deﬁne normal or dysplastic tricuspid morphology

(23.4%) DdBs had documented ventricular arrhythmia; however,

and only described right sided dilatation. Although many DdBs

since Holter was inconsistently performed, ventricular arrhyth-

with AF displayed right sided atrial or ventricular dilatation and

mia might have been under-diagnosed. Alternatively, given the

systolic dysfunction, no non-CHDN dogs had right sided dilatation

high frequency of DdB with AF in our study, some supraventric-

without supraventricular arrhythmia or systolic dysfunction

ular QRS complexes may have been erroneously classiﬁed as

(Fig. 4). This observation suggests that right sided dilatation might

ventricular because of intermittent aberrant conduction with

reﬂect cardiac dysfunction rather than primary valvular disease in

bundle branch block, despite manual Holter review. Ventricular

some DdBs.

DdB with no congenital heart disease, cardiac mass or non-cardiac neoplasia (non-CHDN)

n = 64

Sinus rhythm Supraventricular tachycardia

n = 12 n = 3 Atrial fibrillation

n = 19 Systolic Systolic Right-sided Systolic Systolic Right-sided dysfunction dysfunction dilatation dysfunction dysfunction dilatation and right- and right- sided sided dilatation dilatation

n = 2 n = 2 n = 0 n = 0 n = 1 n = 1

Systolic Systolic Right-sided dysfunction dysfunction dilatation and right- sided dilatation

n = 3 n = 10 n = 4

Fig. 4. Systolic dysfunction and right sided dilatation associated with rhythm diagnosis in subset of 34 Dogue de Bordeaux (DdB) without congenital heart disease, cardiac

mass or non-cardiac neoplasia (non-CHDN).

121

116 G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118

tachycardia (Chang et al., 2008). Focal atrial tachycardia (FAT) was

reported in two DdB in a series of 16 dogs diagnosed with FAT by

electrophysiological study (Santilli et al., 2010). Seven of these

dogs had a surface ECG diagnosis of AF. The authors noted that FAT

occurred in younger dogs, as reported in this study, and that FAT

predominantly originated within the right atrium. Of the three

non-CHDN dogs with SVT in our study, 2/3 had right sided

dilatation potentially supportive of right sided pathology (Figs. 4

and 5). Although our study cannot clarify any temporal relation-

ship between arrhythmia types, it is possible that SVT might

precede and degenerate into AF in some DdB.

Additional ﬁndings in this retrospective study included the

identiﬁcation of cardiac masses in 7/64 (10.9%) DdB and increased

aortic velocities, consistent with SAS, in 12/64 (18.8%) DdB. The

high prevalence of increased aortic velocities is consistent with

Danish and Israeli studies, which reported subaortic stenosis in 17%

and 24% of DdBs in screened and referral populations, respectively

(Höllmer et al., 2008; Ohad et al., 2013).

Our study has several limitations, inherent in any retrospective

studies. The inclusion of dogs with concurrent disease might

inﬂuence heart rate and rhythm, and no account was taken of the

potential inﬂuence of medications on heart rate or cardiac

dimensions. We elected to use normalised left ventricular M-

mode measurements and deﬁned values exceeding 95% CI

predicted values to categorise left ventricular dimensions and

systolic function because of the availability of these images. There

are acknowledged limitations of M-mode measurements and it

would have been preferable to have assessed left ventricular

dimensions and function by calculation of end systolic volume

index using Simpson’s method of discs (Lang et al., 2005; Tidholm

et al., 2010). The use of 95% CI predicted values determined by the

Cornell method may be insensitive because of the wide prediction

intervals. Quantitative evaluation of right sided dimensions would

have been preferable; however, the retrospective nature of this

study meant that images optimised for right sided evaluation were

rarely available. Our inability to more clearly deﬁne the presence

and magnitude of right sided dilatation is a substantial limitation.

Holter ECG was inconsistently available and this may have

limited our ability to obtain representative heart rates. Some cases

presenting with sinus rhythm may have had unidentiﬁed

paroxysmal supraventricular arrhythmia, predisposing to systolic

dysfunction or chamber dilatation; our deﬁnition of SVT may

include some dogs with sinus tachycardia. No cases had an

electrophysiological study to clarify the mechanism of arrhythmia,

so some rhythms may have been misclassiﬁed on surface ECG.

Conclusions

Fig. 5. Box and whisker plots of (a) left atrial-to-aortic ratio (LA:Ao); (b) normalised

left ventricular systolic dimension (LVIDsn); (c) normalised left ventricular diastolic

In this UK referral population of DdB, cardiac masses and

dimension (LVIDsn) according to rhythm diagnosis in subset of 34 Dogue de

subaortic stenosis were relatively frequent. The prevalence of

Bordeaux (DdB) without congenital heart disease, cardiac mass or non-cardiac

ﬁ

neoplasia (non-CHDN). supraventricular arrhythmia, speci cally fast AF with ventricular

response rate exceeding 160 bpm, was high. Atrial ﬁbrillation was

Five dogs in our study had supraventricular tachycardia (one associated with cardiac chamber dilatation in the majority of dogs.

ﬁ

persistent and four paroxysmal). Although the number of dogs Most dogs with atrial brillation had right sided dilatation and

with SVT was small, the three non-CHDN dogs with SVT were systolic dysfunction; however, this was inconstantly associated

younger than dogs with sinus rhythm and AF. In dogs, SVT has been with left ventricular dilatation or a DCM phenotype. After

associated with structural heart disease and systemic disease excluding dogs with congenital heart disease and neoplasia, no

(Kittleson, 1998; Finster et al., 2008). Atrial pathology predisposing remaining dogs had right sided dilatation in the absence of

to supraventricular arrhythmia has been described in association supraventricular arrhythmia or systolic dysfunction. In view of the

ﬁ

with ARVC (Fox et al., 2000; Basso et al., 2004; Mano et al., 2013; subjective classi cation of right sided dilatation in this retrospec-

Vila et al., 2017), and SVT has been reported in Boxers with ARVC tive study, any inferences drawn from these observations should be

(Baumwart et al., 2005). In human beings, paroxysmal SVT is viewed cautiously. Potential aetiologies include primary cardio-

closely associated with paroxysmal AF and it has been suggested myopathy with right sided predilection and/or predisposition to

that SVT might predispose to AF development because of atrial supraventricular arrhythmia with cardiac remodelling secondary

stretch, reduced atrial refractoriness and vagal activation during to tachycardia-induced cardiomyopathy. Since some cardiomyop-

athies may be associated with a propensity for supraventricular

122

G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118 117

arrhythmia, these mechanisms may not be mutually exclusive. Kittleson, M.D., 1998. Primary myocardial disease leading to chronic myocardial

failure. In: Kittleson, M.D., Kienle, R.D. (Eds.), Small Animal Cardiovascular

Serial echocardiographic and Holter studies may help to clarify if

Medicine. Mosby, St Louis, MO, USA, pp. 319–346.

supraventricular arrhythmia typically develops as a consequence

Lancellotti, P., Moura, L., Pierard, L.A., Agricola, E., Popescu, B.A., Tribouilloy, C.,

of chamber dilatation, or if it precedes cardiac chamber Hagendorff, A., Monin, J.-L., Badano, L., Zamorano, J.L., 2010. European

Association of Echocardiography recommendations for the assessment of

remodelling in the DdB, and whether the ﬁndings from this small

valvular regurgitation. Part 2: mitral and tricuspid regurgitation (native valve

study are representative of the wider DdB population.

disease). European Journal of Echocardiography 11, 307–332.

Lang, R.M., Bierig, M., Devereux, R.B., Flachskampf, F.A., Foster, E., Pellikka, P.A.,

Conﬂict of interest statement Picard, M.H., Roman, M.J., Seward, J., Shanewise, J.S., et al., 2005.

Recommendations for chamber quantiﬁcation: a report from the American

Society of Echocardiography’s Guidelines and Standards committee and the

ﬁ

None of the authors of this paper have a nancial or personal Chamber Quantiﬁcation Writing Group, developed in conjunction with the

relationship with other people or organisations that could European Association of Echocardiography, a branch of the European Society of

Cardiology. Journal of the American Society of Echocardiography 18, 1440–1463.

inappropriately inﬂuence or bias the content of the paper.

Locatelli, C., Santini, A., Bonometti, G.A., Palermo, V., Scarpa, P., Sala, E., Brambilla, P.

G., 2011. Echocardiographic values in clinically healthy adult Dogue de Bordeaux

Acknowledgements dogs. Journal of Small Animal Practice 52, 246–253.

McAulay, G., Borgeat, K., Fuentes, V.L., 2018. An online health survey of Dogue de

Bordeaux owners and breeders with special emphasis on cardiac disease. The

Preliminary results were presented at the Veterinary Cardio-

Veterinary Journal 232, 78–82.

vascular Society Meeting at the British Small Animal Veterinary Mano, H., Watanabe, I., Okumura, Y., Sonoda, K., Nagashima, K., Nakai, T., Ohkubo, K.,

Kofune, T., Haruta, H., Kofune, M., et al., 2013. Atrial tachycardia in a patient with

Association Congress, Birmingham, UK, 5 April 2017.

arrhythmogenic right ventricular cardiomyopathy/dysplasia. Journal of

Arrhythmia 29, 238–241.

References Martin, M.W.S., Stafford Johnson, M.J., Celona, B., 2009. Canine dilated

cardiomyopathy: a retrospective study of signalment, presentation and clinical

ﬁndings in 369 cases. Journal of Small Animal Practice 50, 23–29.

Basso, C., Fox, P.R., Meurs, K.M., Towbin, J.A., Spier, A.W., Calabrese, F., Maron, B.J.,

Menaut, P., Bélanger, M.C., Beauchamp, G., Ponzio, N.M., Moïse, N.S., 2005. Atrial

Thiene, G., 2004. Arrhythmogenic right ventricular cardiomyopathy causing

ﬁbrillation in dogs with and without structural or functional cardiac disease: a

sudden cardiac death in boxer dogs: a new animal model of human disease.

retrospective study of 109 cases. Journal of Veterinary Cardiology 7, 75–83.

Circulation 109, 1180–1185.

Meurs, K.M., Miller, M.W., Wright, N.A., 2001a. Clinical features of dilated

Baumwart, R.D., Meurs, K.M., Atkins, C.E., Bonagura, J.D., DeFrancesco, T.C., Keene, B.

cardiomyopathy in Great Danes and results of a pedigree analysis: 17 cases

W., Koplitz, S., Fuentes, V.L., Miller, M.W., Rausch, W., et al., 2005. Clinical,

(1990-2000). Journal of the American Veterinary Medical Association 218, 729–

echocardiographic, and electrocardiographic abnormalities in Boxers with

732.

cardiomyopathy and left ventricular systolic dysfunction: 48 cases (1985-2003).

Meurs, K.M., Spier, A.W., Wright, N.A., Hamlin, R.L., 2001b. Use of ambulatory

Journal of the American Veterinary Medical Association 226, 1–3.

electrocardiography for detection of ventricular premature complexes in

Bonagura, J., Lehmkuhl, L.B., Lehmkuhl, L., 1999. Congenital heart disease. In: Fox, P.

healthy dogs. Journal of the American Veterinary Medical Association 218,

R., Sisson, D.D., Moise, S. (Eds.), Textbook of Canine and Feline Cardiology.

1291–1292.

Saunders, St Louis, MO, USA, pp. 471–535.

Meurs, K.M., Spier, A.W., Miller, M.W., Lehmkuhl, L., Towbin, J.A., 1999. Familial

Borgarelli, M., Santilli, R.A., Chiavegato, D., D’Agnolo, G., Zanatta, R., Mannelli, A.,

ventricular arrhythmias in Boxers. Journal of Veterinary Internal Medicine 13,

Tarducci, A., 2006. Prognostic indicators for dogs with dilated cardiomyopathy.

437–439.

Journal of Veterinary Internal Medicine 20, 104–110.

Miller, M.W., Tilley, L.P., Smith, K.F.W., Fox, P.,1999. Electrocardiography. In: Fox, P.R.,

Bright, J.M., Martin, J.M., Mama, K., 2005. A retrospective evaluation of transthoracic

Sisson, D.D., Moise, S. (Eds.), Textbook of Canine and Feline Cardiology.

biphasic electrical cardioversion for atrial ﬁbrillation in dogs. Journal of

Saunders, St Louis, MO, USA, pp. 67–105.

Veterinary Cardiology 7, 85–96.

Moore, E.N., Spear, J.F., 1987. Electrophysiological studies on atrial ﬁbrillation. Heart

Brownlie, S.E., Cobb, M.A., 1999. Observations on the development of congestive

Vessels Supplement 2, 32–39.

heart failure in Irish wolfhounds with dilated cardiomyopathy. Journal of Small

Najib, M.Q., Vinales, K.L., Vittala, S.S., Challa, S., Lee, H.R., Chaliki, H.P., 2011.

Animal Practice 40, 371–377.

Predictors for the development of severe tricuspid regurgitation with

Brundel, B.J.J.M., Melnyk, P., Rivard, L., Nattel, S., 2005. The pathology of atrial

anatomically normal valve in patients with atrial ﬁbrillation. Echocardiography

ﬁbrillation in dogs. Journal of Veterinary Cardiology 7, 121–129.

29, 140–146.

Bussadori, C., Amberger, C., Le Bobinnec, G., Lombard, C.W., 2000. Guidelines for the

Ohad, D.G., Avrahami, A., Waner, T., David, L., 2013. The occurrence and suspected

echocardiographic studies of suspected subaortic and pulmonic stenosis.

mode of inheritance of congenital subaortic stenosis and tricuspid valve

Journal of Veterinary Cardiology 2, 15–22.

dysplasia in Dogue de Bordeaux dogs. The Veterinary Journal 197, 351–357.

Chang, S., Tai, C., Lin, Y., Lo, L., Tuan, T., Udyavar, A.R., Tsao, H., Hsieh, M., Hu, Y.,

Oliveira, P., Domenech, O., Silva, J., Vannini, S., Bussadori, R., Bussadori, C., 2011.

Chiang, S., et al., 2008. Electrophysiological characteristics and catheter ablation

Retrospective review of congenital heart disease in 976 dogs. Journal of

in patients with paroxysmal supraventricular tachycardia and paroxysmal atrial

Veterinary Internal Medicine 25, 477–483.

ﬁbrillation. Journal of Cardiovascular Electrophysiology 19, 367–373.

Oyama, M.A., Sisson, D.D., 2001. Evaluation of canine congenital heart disease using

Cornell, C.C., Kittleson, M.D., Torre Della, P., Haggstrom, J., Lombard, C.W., Pedersen,

an echocardiographic algorithm. Journal of the American Animal Hospital

H.D., Vollmar, A., Wey, A., 2004. Allometric scaling of M-mode cardiac

Association 37, 519–535.

measurements in normal adult dogs. Journal of Veterinary Internal Medicine 18,

311–321. Palermo, V., Stafford Johnson, M.J., Sala, E., Brambilla, P.G., Martin, M.W.S., 2011.

Cardiomyopathy in Boxer dogs: a retrospective study of the clinical

de Madron, É., 2015. Global left ventricular systolic function assessment. In: de

presentation, diagnostic ﬁndings and survival. Journal of Veterinary Cardiology

Madron, E., Chetboul, V., Bussadori, C. (Eds.), Clinical Echocardiography of the

13, 45–55.

Dog and Cat. Elsevier, St Louis, MO, USA, pp. 111–125.

Park, J.-H., Shin, S.-H., Lee, M.-J., Lee, M.-D., Shim, H.-I., Yoon, J., Oh, S., Kim, D.-H.,

Dukes-McEwan, J., Borgarelli, M., Tidholm, A., 2003. Proposed guidelines for the

Park, S.-D., Kwon, S.-W., et al., 2015. Clinical and echocardiographic factors

diagnosis of canine idiopathic dilated cardiomyopathy. Journal of Veterinary

affecting tricuspid regurgitation severity in the patients with lone atrial

Cardiology 5, 7–19.

ﬁbrillation. Journal Cardiovascular Ultrasound 23, 136–137.

Finster, S.T., DeFrancesco, T.C., Atkins, C.E., Hansen, B.D., Keene, B.W., 2008.

Rogers, J.H., Bolling, S.F., 2009. The tricuspid valve: current perspective and evolving

Supraventricular tachycardia in dogs: 65 cases (1990-2007). Journal of

management of tricuspid regurgitation. Circulation 119, 2718–2725.

Veterinary Emergency and Critical Care 18, 503–510.

Sahn, D.J., DeMaria, A., Kisslo, J., Weyman, A., 1978. Recommendations regarding

Fox, P.R., Maron, B.J., Basso, C., Liu, S.K., Thiene, G., 2000. Spontaneously occurring

quantitation in M-mode echocardiography: results of a survey of

arrhythmogenic right ventricular cardiomyopathy in the domestic cat: a new

echocardiographic measurements. Circulation 58, 1072–1083.

animal model similar to the human disease. Circulation 102, 1863–1870.

Santilli, R.A., Perego, M., Crosara, S., Gardini, F., Bellino, C., Moretti, P., Spadacini, G.,

Guglielmini, C., Chetboul, V., Pietra, M., Pouchelon, J.L., Capucci, A., Cipone, M., 2000.

2008. Utility of 12-lead electrocardiogram for differentiating paroxysmal

Inﬂuence of left atrial enlargement and body weight on the development of

supraventricular tachycardias in dogs. Journal of Veterinary Internal Medicine

atrial ﬁbrillation: retrospective study on 205 dogs. The Veterinary Journal 160,

235–241. 22, 915–923.

Santilli, R.A., Perego, M., Perini, A., Moretti, P., Spadacini, G., 2010. Electrophysiologic

Hansson, K., Haggstrom, J., Kvart, C., Lord, P., 2002. Left atrial to aortic root indices

characteristics and topographic distribution of focal atrial tachycardias in dogs.

using two-dimensional and M-mode echocardiography in Cavalier King Charles

Journal of Veterinary Internal Medicine 24, 539–545.

spaniels with and without left atrial enlargement. Veterinary Radiology 43,

568–575. Schoonderwoerd, B.A., Gelder, I.C., Veldhuisen, D.J., Tieleman, R.G., Grandjean, J.G.,

Bel, K.J., Allessie, M.A., Crijns, H.J.G.M., 2001. Electrical remodeling and atrial

Harpster, N.K., 1991. Boxer cardiomyopathy. A review of the long-term beneﬁts of

dilation during atrial tachycardia are inﬂuenced by ventricular rate: role of

antiarrhythmic therapy. Veterinary Clinics of North America: Small Animal

developing tachycardiomyopathy. Journal of Cardiovascular Electrophysiology

Practice 21, 989–1004.

12, 1404–1410.

Höllmer, M., Willesen, J.L., Jensen, A.T., Koch, J., 2008. Aortic stenosis in the Dogue de

Bordeaux. Journal of Small Animal Practice 49, 432–437.

123

118 G. McAulay et al. / The Veterinary Journal 234 (2018) 111–118

Sébillon, P., Bouchier, C., Bidot, L.D., Bonne, G., Ahamed, K., Charron, P., Drouin- Tidholm, A., Jönsson, L., 1996. Dilated cardiomyopathy in the Newfoundland: a

Garraud, V., Millaire, A., Desrumeaux, G., et al., 2003. Expanding the phenotype study of 37 cases (1983-1994). Journal of the American Animal Hospital

of LMNA mutations in dilated cardiomyopathy and functional consequences of Association 32, 465–470.

these mutations. Journal of Medical Genetics 40, 560–567. Tidholm, A., Westling, A.B., Höglund, K., Ljungvall, I., Haggstrom, J., 2010.

Shinbane, J.S., Wood, M.A., Jensen, D.N., Ellenbogen, K.A., Fitzpatrick, A.P., Comparisons of 3-, 2-dimensional, and M-mode echocardiographical methods

Scheinman, M.M., 1997. Tachycardia-induced cardiomyopathy: a review of for estimation of left chamber volumes in dogs with and without acquired heart

animal models and clinical studies. Journal American College of Cardiology 29, disease. Journal of Veterinary Internal Medicine 24, 1414–1420.

709–715. Topilsky, Y., Nkomo, V.T., Vatury, O., Michelena, H.I., Letourneau, T., Suri, R.M.,

Simpson, S., Dunning, M.D., Brownlie, S., Patel, J., Godden, M., Cobb, M., Mongan, N. Pislaru, S., Park, S., Mahoney, D.W., Biner, S., et al., 2014. Clinical outcome of

P., Rutland, C.S., 2016. Multiple genetic associations with Irish wolfhound isolated tricuspid regurgitation. JACC Cardiovascular Imaging 7, 1185–1194.

dilated cardiomyopathy. BioMed Research International 2016, 6374082. Tornos Mas, P., Rodríguez-Palomares, J.F., Antunes, M.J., 2015. Secondary tricuspid

Spinner, E.M., Lerakis, S., Higginson, J., 2012. Correlates of tricuspid regurgitation as valve regurgitation: A forgotten entity. Heart 101, 1840–1848.

determined by 3D echocardiography: pulmonary arterial pressure, ventricle Vila, J., Pariaut, R., Moïse, N.S., Oxford, E.M., Fox, P.R., Reynolds, C.A., Saelinger, C.,

geometry, annular dilatation, and papillary muscle displacement. Circulation 5, 2017. Structural and molecular pathology of the atrium in boxer

43–50. arrhythmogenic right ventricular cardiomyopathy. Journal of Veterinary

Stepien, R.L., 2009. Pulmonary arterial hypertension secondary to chronic left-sided Cardiology 19, 57–67.

cardiac dysfunction in dogs. Journal of Small Animal Practice 50, 34–43. Vollmar, A.C., 2000. The prevalence of cardiomyopathy in the Irish wolfhound: a

Stern, J.A., Meurs, K.M., Spier, A.W., Koplitz, S.L., Baumwart, R.D., 2010. Ambulatory clinical study of 500 dogs. Journal of the American Animal Hospital Association

electrocardiographic evaluation of clinically normal adult Boxers. Journal of the 36, 125–132.

American Veterinary Medical Association 236, 430–433. Wess, G., Schulze, A., Butz, V., Simak, J., Killich, M., Keller, L.J.M., Maeurer, J.,

Takemura, N., Nakagawa, K., Hirose, H., 2002. Lone atrial ﬁbrillation in a dog. Journal Hartmann, K., 2010. Prevalence of dilated cardiomyopathy in Doberman

of Veterinary Medical Science 64, 1057–1059. Pinschers in various age groups. Journal of Veterinary Internal Medicine 24,

Thomas, W.P., Gaber, C.E., Jacobs, G.J., Kaplan, P.M., Lombard, C.W., Moïse, N.S., 533–538.

Moses, B.L., 1993. Recommendations for standards in transthoracic two- Wilson, J.R., Douglas, P., Hickey, W.F., Lanoce, V., Ferraro, N., Muhammad, A., Reichek,

dimensional echocardiography in the dog and cat. Echocardiography N., 1987. Experimental congestive heart failure produced by rapid ventricular

Committee of the Specialty of Cardiology, American College of Veterinary pacing in the dog: cardiac effects. Circulation 75, 857–867.

Internal Medicine. Journal of Veterinary Internal Medicine 7, 247–252. Wyse, D.G., van Gelder, I.C., Ellinor, P.T., Go, A.S., Kalman, J.M., Narayan, S.M., Nattel,

Tidholm, A., Haggstrom, J., Borgarelli, M., Tarducci, A., 2001. Canine idiopathic S., Schotten, U., Rienstra, M., 2014. Lone atrial ﬁbrillation: does it exist? Journal

dilated cardiomyopathy. Part I: Aetiology, clinical characteristics, epidemiology of the American College of Cardiology 63, 1715–1723.

and pathology. The Veterinary Journal 162, 92–107.

124 IV Sargent J, Connolly DJ, Watts V, Mõtsküla PF, Volk HA, Lamb CR, Luis Fuentes V., 2015. ASSESSMENT OF MITRAL REGURGITATION IN DOGS: COMPARISON OF RESULTS OF ECHOCARDIOGRAPHY WITH MAGNETIC RESONANCE IMAGING. Journal of Small Animal Practice 56, 641–650. / PETSAVERS PAPER Assessment of mitral regurgitation in dogs: comparison of results of echocardiography with magnetic resonance imaging

J. Sargent, D. J. Connolly, V. Watts, P. Mõtsküla, H. A. Volk, C. R. Lamb and V. Luis Fuentes

The Royal Veterinary College, University of London, Herts AL9 7TA

OBJECTIVES: Echocardiography is used routinely to assess mitral regurgitation severity, but echocardiographic measures of mitral regurgitation in dogs have not been compared with other quantitative methods. The study aim was to compare echocardiographic measures of mitral regurgitation with cardiac magnetic resonance imaging-derived mitral regurgitant fraction in small-breed dogs.

METHODS: Dogs with myxomatous mitral valve disease scheduled for magnetic resonance imaging

sava.com assessment of neurological disease were recruited. Correlations were tested between cardiac magnetic resonance imaging-derived mitral regurgitant fraction and the following echocardiographic measures: vena contracta/aortic diameter, transmitral E-wave velocity, amplitude of mitral prolapse/ b aortic diameter, diastolic left ventricular diameter:aortic diameter, left atrium:aortic diameter, mitral regurgitation jet area ratio and regurgitant fraction calculated using the proximal isovelocity surface area method.

RESULTS: Measurement of cardiac magnetic resonance imaging-derived mitral regurgitant fraction was attempted in 21 dogs. Twelve consecutive, complete studies were obtained and 10 dogs were included in the final analysis: vena contracta/aortic diameter (r=0·89, p=0·001) and E-wave velocity (r=0·86, p=0·001) had the strongest correlations with cardiac magnetic resonance imaging-derived mitral regurgitant fraction. E velocity had superior repeatability and could be measured in all dogs. The presence of multiple jets precluded vena contracta/aortic diameter measurement in one dog.

CLINICAL SIGNIFICANCE: Measurement of cardiac magnetic resonance imaging-derived mitral regurgitant fraction is feasible but technically demanding. The echocardiographic measures that correlated most closely with cardiac magnetic resonance imaging-derived mitral regurgitant fraction were vena www. contracta/aortic diameter and E-wave velocity.

Journal of Small Animal Practice (2015) 56, 641–650 DOI: 10.1111/jsap.12410 // Accepted: 3 September 2015; Published online: 16 October 2015

INTRODUCTION will have a poorer outcome with increased risk of congestive cardiac failure and premature death, but this has never Mitral regurgitation (MR) associated with myxomatous mitral been clearly demonstrated. This is in part due to lack of a p: valve disease (MMVD) is highly prevalent in older small breed validated gold standard technique with which to accurately dogs (Whitney 1974, Thrusfield et al. 1985) and is the third quantify MR. most common cause of death in this species (Haggstrom et al. Several studies have investigated the association of echocar- 2004). It is reasonable to suspect that dogs with worse MR diographic measures on survival or onset of congestive heart

127 J. Sargent et al.

failure. Chronic left heart volume overload is known to result in ( Thavendiranathan et al. 2012b) and therefore cMRI could be left atrial (LA) enlargement and increased left ventricular (LV) useful as a reference method with which to compare echocardio- diastolic diameter, and tends to occur late in the clinical course graphic measurements. of MMVD, with the most rapid changes in the 6 months imme- The aims of the present study were to: 1) investigate the fea- diately preceding onset of congestive heart failure (Lord et al. sibility of MR quantification using cMRI in small breed dogs; 2010, Hezzell et al. 2012, Reynolds et al. 2012). Early transmi- and, 2) identify the echocardiographic variables that best cor- tral (E-wave) peak velocity (Evel) is directly influenced by MR relate with severity of MR as determined by cMRI. volume (Thomas et al. 1998) and velocities that are greater than 1·2 m/s are associated with disease progression and increased cardiac mortality (Borgarelli et al. 2008, 2012, Hezzell et al. 2012, MATERIALS AND METHODS Sargent et al. 2015). Quantitative or semi-quantitative colour flow Doppler (CFD) The study protocol was approved by the Ethics and Welfare based methods to evaluate MR in dogs have been extrapolated committee of The Royal Veterinary College (URN 2010:1078). from techniques used in people (Zoghbi et al. 2003, Grayburn Dogs were recruited with informed owner consent, including et al. 2012, Messika-Zeitoun et al. 2013). The most commonly discussion of the additional time required for acquisition of the reported methods to semi-quantify MR in dogs are the ratio of cMRI images. A complete medical history and physical exami- MR jet area to LA area (the jet area ratio, or JAR), to reflect mitral nation (including neurological examination) was collected prior regurgitant volume (Kittleson & Brown 2003, Muzzi et al. 2003, to recruitment. The inclusion criteria for recruitment were dogs Chetboul & Tissier 2012) and the proximal isovolumetric sur- presenting for magnetic resonance imaging (MRI) to investigate face area (PISA) methods (Schwammenthal et al. 1996, Kittleson suspected neurological disease; concurrent MMVD; and body & Brown 2003, Choi et al. 2004, Gouni et al. 2007). The PISA weight ≤20kg. The diagnosis of MMVD was based on detec- method is useful to estimate effective MR orifice area (EROA) tion of a systolic heart murmur and echocardiographic find- and regurgitant fraction (PISA-RF) but is inaccurate for evalua- ings characteristics of MMVD, including MR confirmed using tion of the eccentric or multiple mitral regurgitant jets that are colour-flow Doppler (CFD), and two-dimensional (2D) echo- frequently encountered in canine MMVD (Chetboul & Tissier cardiographic evidence of mitral valve prolapse and thicken- 2012). In people, measurement of the vena contracta (minimum ing. Dogs believed to be at high risk of congestive heart failure diameter) of the MR jet (VC) is a widely accepted method to (advanced MMVD or suggestive history) were excluded. approximate the regurgitant orifice area (Grayburn et al. 2012) that is not influenced by machine gain settings or jet eccentricity Echocardiography protocol (Hall et al. 1997). The application of this technique has recently Echocardiography was performed prospectively prior to MRI been described for dogs with MMVD (Di Marcello et al. 2014, studies and repeated after cMRI data acquisition while the Sargent et al. 2015). patient was still under the influence of general anaesthesia. Cardiac magnetic resonance imaging (cMRI) is considered an All studies were recorded on one of two echocardiographic accurate technique for quantifying MR in people (Thavendira- machines (Vivid7 and Vivid i, General Electric Medical Systems nathan et al. 2012a,b). To date, experience of cMRI in veterinary Ultrasound, 71 Great North Road, Hatfield) by the same cardi- patients is limited. The need for general anaesthesia to provide ology resident. Standard echocardiographic views were obtained sufficient restraint for MRI inhibits its use in cardiac patients from the right parasternal, subcostal, left apical and left cranial but there has been recent interest in the use of cMRI to fur- parasternal positions using 7·5 MHz or 3·5 MHz phased-array ther characterise cardiac and/or pericardial neoplasia (Mai et al. transducers with concurrent ECG recording. A combination of 2010, Boddy et al. 2011, Gallach & Mai 2013, to image complex M-mode, two-dimensional echocardiography, CFD and spec- canine congenital cardiac defects (Garcia-Rodriguez et al. 2009, tral Doppler echocardiography was used to record LA diameter, Lopez-Alvarez et al. 2011, Schwarz et al. 2009) and to study LA area, MR jet area, aortic diameter (Ao), amplitude of mitral animal models of human cardiac disease (Gilbert et al. 2010). valve prolapse (Prolapse), LV diastolic diameter (LVIDd), Evel, A recent study used cMRI as a gold standard reference method vena contracta of the MR jet (VC), proximal isovelocity sur- with which to compare measurements of LV volume and func- face area of the MR jet (PISA), MR velocity, and aortic velocity tion made using contrast echocardiography (Kim et al. 2013). (Table 1). Echocardiographic measurements were performed There have been other studies of remodelling of the right and off-line using a dedicated workstation and proprietary software left ventricles in response to chronic mitral valve disease, but no (Echopac, General Electric Medical Systems Ultrasound, 71 quantification of MR was attempted (Gilbert et al. 2010, Young Great North Road, Hatfield, AL9 5EN). Measurements were et al. 1996). obtained retrospectively from pre- and post cMRI studies by As facilities for cMRI are limited in veterinary practice, echo- the observer who had acquired the echocardiographic data. cardiography is likely to remain the standard clinical tool for A second observer (American College of Veterinary Internal assessing severity of MMVD in dog. Although indirect mea- Medicine Diplomate in Cardiology) repeated echocardio- surement of MR volumes using cMRI has not been validated graphic measurements in all dogs from studies acquired at the in dogs, there are fewer geometric assumptions for assessment of end of cMRI data acquisition in order to evaluate inter-observer ventricular volumes with cMRI than with 2D echocardiography reproducibility.

128 Assessment of mitral regurgitation in dogs

Table 1. Echocardiographic parameters of interest Echo variable Technique Measurement LVIDd:Ao LVIDd: M-mode from RPxSA at papillary muscle LVIDd: leading edge to leading edge technique at QRS onset Left ventricular diastolic level Ao: aortic diameter measured in systole between open aortic valve diameter: aortic diameter Ao: 2D Left parasternal cranial view leaflets LA:Ao 2D RPxSA view at the level of the heart base LA and Ao diameter: measured at the first frame of aortic valve Left atrial diameter: aortic closure at the beginning of diastole (Hansson et al. 2002) diameter Prolapse:Ao Prol: 2D RPxLA optimised for left ventricular inlet Prol: Maximum excursion of either mitral leaflet towards the left Amplitude of mitral pro- Ao: 2D Left parasternal cranial view atrium beyond a line drawn at the level of the mitral annulus lapse: aortic diameter Ao: aortic diameter measured in systole between open aortic valve leaflets

Evel PW spectral Doppler from LAp4Ch view with The maximal value of the spectral signal of the mitral E wave Peak early transmitral filling sample volume at tips of open leaflets velocity JAR Colour flow Doppler in LAp4Ch view MR jet area: Planimetry of the largest area of blue or aliased signal, Mitral regurgitant jet area: without including any flow rebounding from the furthest LA wall. left atrial area LA area: Planimetry of the endocardial border of the LA, obtained from the same frame as the jet area, following digital subtraction of the colour signal. VC Colour flow Doppler without variance in a LAp4Ch Width of the narrowest point of the colour MR jet beyond the region Vena contracta width view in ‘zoom’ mode, from the frame in systole of proximal flow convergence. Where multiple VC jets were pres- where VC appeared smallest and where the ent, the VC of the largest jet was recorded. region of proximal flow convergence, VC, and jet were all clearly visible. 2 PISA MR volume MR flow rate: Colour flow Doppler in LAp4Ch view MR flow rate: 2x π r x Alv(where r= the radius of the isovelocity Mitral regurgitant volume in “zoom” with adjustment of Nyquist limit to semicircle of aliasing in the proximal flow convergence region and

calculated by proximal achieve a symmetrical semicircle of aliasing in Alv is the aliasing velocity ) isovelocity surface area the proximal flow convergence region. method Peak MR velocity and MR velocity time integral EROA=MR peak velocity×Mitral flow rate (VTI) from CW spectral Doppler of MR jet MR volume=EROA×mitral VTI 2 Aortic stroke volume Aortic velocity time integral: high pulse repetition Aortic stroke volume = π x r x VTIAo (where r= the radius of the aorta

frequency spectral Doppler from subcostal view. and VTIAo is the aortic VTI. Aortic radius from LCrPX view PISA-RF LAp4Ch The proximal isovelocity surface area derived value for MR=the MR Regurgitant fraction calcu- Subcostal volume/(Ao SV+MR volume)×100 lated by proximal isoveloc- LCrPX view ity surface method

2D Two-dimensional echocardiography, Ao Aortic diameter, LA Left atrial diameter, LVIDd Left ventricular diastolic diameter, Prolapse Amplitude of prolapse, RPxSA Right parasternal short axis view, RPxLA Right parasternal long axis four chamber view cMRI protocol A series of short axis images of the LV was obtained using the Dogs received premedication with butorphanol or methadone. 2- and 4-chamber VLA sequences. Care was taken to ensure that General anaesthesia was induced using diazepam and either alphax- the image plane was parallel to the plane of the mitral annu- alone or propofol, and maintained with sevoflurane. Cardiac MRI lus and that acquired images encompassed the entire LV and LA was performed after acquisition of clinically-indicated magnetic from the apex of the LV to the dorsal aspect of the LA. For acqui- resonance images of the brain and/or spinal cord by the same sition of the LV short axis stack a flip angle of 60°, median repeti- cardiology resident and certified MRI technologist. A 1·5 Tesla tion time of 3·9 seconds (range 3·6 to 4·0) and median echo time MRI (Intera Pulsar System, Philips Medical Systems, Castlefield of 1·9 seconds (range 1·8 to 2·0) was used. Road, Reigate, Surrey) and a flexible, sensitivity-encoding, phased Depending on the size of dog, 9 to 10 images of 7 to 8 mm array surface coil (SENSE™ Flex M or SENSE™ Body coil) thickness were obtained with an interslice gap of 0 mm and 30 were used. Intermittent positive pressure ventilation was main- frames per cardiac cycle (Fig. 2). Image slice thickness was chosen tained throughout using a mechanical ventilator. Breath-holding to achieve the greatest number of slices with acceptable signal- to avoid respiratory motion during periods of image acquisition to-noise ratio. Acquisition of the short axis images required 3 was achieved though induction of transient apnoea following a breath-hold sequences for the required acquisition time. brief period of controlled hyperventilation. Survey gradient echo Using T1-weighted Turbo field echo, a series of images of the imaging was performed to acquire transverse, sagittal and dorsal LV outflow tract (LVOT) in a plane orthogonal to the short axis standard localising scans (scouts) from which 2- and 4-chamber images were then obtained to identify a site in the proximal por- LV long axis (VLA) cine images were prescribed (Fig. 1). Cine tion of the aorta that was cranial to the sinus of Valsalva (at the image acquisition utilised a gradient echo pulse sequence with sinotubular junction), and perpendicular to aortic flow (Fig. 3). balanced steady-state free precession (bright-blood imaging). For Phase contrast angiography imaging at this site, with the image these sequences, an MRI compatible vectocardiogram (VCG) gat- plane aligned parallel to the aortic valve plane, was used to obtain ing device was used to apply retrospective gating. velocity-encoded maps that enabled quantitative measurements

129 J. Sargent et al.

of aortic blood flow. For quantitative aortic flow measurements a flip angle of 15°, median repetition time of 5·3 seconds (range 5·2 to 5·6) and median echo time of 3·9 seconds (range 3·6 to 4·0) was used. Velocity phase sensitivity was adjusted manually to avoid aliasing of aortic flow (Fig. 4). The maximum time allowed for additional general anaesthesia for cMRI acquisition was limited to 50 minutes, after which dogs were recovered from anaesthesia even if the cMRI acquisition was incomplete. Analysis of cardiac images was performed offline by a single observer using a dedicated software package (Extended MR Workspace Software Package Version R.6.2.1. Philips Medical Systems, Castlefield Road, Reigate, Surrey) at a separate time point, subsequent to measurement of echocardiographic data.

LV volumes were calculated by a modified Simpson’s rule, tracing the ventricular endocardial borders in all frames from apex to base through the whole cardiac cycle using a semi-automated, contour-tracing software programme. Papillary muscles were excluded but LV trabeculae were included in the ventricular lumen volume. The most basal slice included was the slice where the endocardial contours were interrupted by the LV outflow (even at end diastole) but greater than 50% of the ventricular lumen circumference was surrounded by a wall thickness consistent with ventricular myocardium (Alfakih et al. 2003). Total LV stroke volume was calculated by subtracting the minimum LV volume from maximum LV volume. Active contouring was applied to the aortic vessel walls in the aortic velocity-encoded FIG 1. Example of a 4-chamber left ventricular long axis image of the heart demonstrating mitral regurgitation (arrow) characterised by signal images for calculation of the aortic stroke volume using the quan- void in the left atrium. LA Left atrium, LV Left ventricle titative flow analysis software programme. Mitral regurgitant

FIG 2. Images from a short axis stack through the heart, encompassing the apex of the left ventricle to the roof of the left atrium. Each frame represents the same point of the cardiac cycle. Ao Aorta, LA Left atrium, LV Left ventricle

130 Assessment of mitral regurgitation in dogs

FIG 3. Example of orthogonal acquisitions of the left ventricular outflow tract used to prescribe velocity-encoded imaging of aortic flow at the sinotubular junction. Ao Aorta, LA Left atrium, LV Left ventricle

FIG 4. Velocity-encoded phase contrast angiography image of aortic flow at the level of the sinotubular junction as shown in Fig 3. Aliased flow is depicted as a black “core” in image (A) that is absent in image (B), following manual adjustment of velocity phase sensitivity volume was calculated by subtracting aortic stroke volume from Bland-Altman plots to calculate the limits of agreement for con- LV stroke volume. Mitral regurgitant fraction (cMRI-RF) was tinuous variables. The difference between echocardiographic defined as the MR volume as a proportion of the LV stroke vol- measurements between the first and second examination was ume. Five complete cMRI studies were measured independently assessed using a paired Student’s t-test for parametric data and by two cardiology residents. the Wilcoxon matched-pairs signed rank test for non-parametric Statistical analysis was carried out using either IBM SPSS Sta- data. Correlations between cMRI-RF and the following echo- tistics 20.0, or GraphPad Prism 6 for Windows. Significance was cardiographic variables were tested using Pearson’s correlation set at p<0·05. The distributions of data for patient population coefficient (r): vena contracta/aortic diameter (VC:Ao), Evel, Pro- characteristics, duration of image acquisition and echocardio- lapse/aortic diameter (Prolapse:ao), LVIDd/aortic diameter, left graphic variables were assessed for normality using the Shap- atrium/aortic diameter (LA:Ao), JAR and PISA-RF. iro-Wilk test and observations on graphs. Results for normally distributed data are presented as mean (±standard deviation) and median (interquartile range) for non-normally distributed RESULTS data. Inter-observer agreement was tested using intraclass correlation coefficient (ICC) and two-way mixed, single measures Twenty-one dogs were recruited between February 2011 and model. Results are expressed as ICC (95% confidence interval). February 2014. There were 15 Cavalier King Charles spaniels The lower and upper bounds of 95% confidence interval for and one each of the following breeds; whippet, Border collie, ICC were truncated at zero and at one, if the estimated 95% beagle, Shih Tzu, Bichon Frise and toy poodle. Mean age was CI for ICC exceeded the limits. The inter-observer agreement 7·67 years (± 2·73 years), with 11 males and 10 females. Murmur for cMRI-derived volumes were also assessed graphically using intensities ranged from grade 1 to 5/6.

131 J. Sargent et al.

Significant differences in echocardiographic measures between DISCUSSION the conscious and anaesthetised studies were not detected. Based on ICC result, inter-observer repeatability for LVIDd and E vel Our study compared echocardiographic variables associated with [0·98 (0·91 to 0·99) and 0·96 (0·88 to 0·99) was superior to that increasing MMVD severity with an indirect method to quan- of LA:Ao (0·86 (0·52 to 0·96)]. Of the 2D echocardiographic tify cMRI-RF. In experimental dog models, cMRI imaging has measurements, Prolapse was the least reproducible with an ICC been shown to be highly accurate in quantifying LV volumes of 0·75 (0·31 to 0·92). Of the CFD echocardiographic variables, (Dell’Italia et al. 1994) and has been used as a gold standard com- inter-observer repeatability was better for VC [0·89 (0·61 to parison technique for echocardiographic measurements of SV 0·98) than PISA and JAR (0·52 (0 to 0·88) and 0·34 (0 to 0·74)], and ejection fraction (Kim et al. 2013). To the authors’ knowl- respectively. edge, no studies have validated phase-velocity encoded imaging Several of the initial cMRI studies were curtailed because of for measurement of aortic stroke volume or to calculate cMRI-RF difficulties obtaining suitable images. For example, image qual- in dogs, but this method has been used to compare effects of two ity was subjectively poor for the first seven dogs that utilised the anesthetic protocols on aortic flow (Drees 2015). Direct, SENSE™ Flex-M surface coil, primarily because of low signal et al. phase-contrast imaging and indirect CMRI methods have been to noise ratio. Changing to the SENSE™ Body coil resulted in superior image quality and improved signal-to-noise ratio. described to quantitate MR in people (Thavendiranathan et al. Difficulties obtaining suitable images resulted in incomplete 2012a,b). cMRI examinations for the first nine dogs because a study was The main obstacle to validate any technique for quantifica- terminated when study time approached 50 minutes, according tion of MR has been the lack of a practical and accurate refer- to protocol. The last part of the study protocol was the phase ence method. Invasive catheterisation techniques are subject to contrast velocity-encoded mapping of aortic outflow, hence no inherent limitations associated with assumptions of LV geometry, aortic stroke volume data were obtained in studies terminated validity of calculations in the presence of multiple valve insuffi- before completion. Improvements in operator expertise enabled ciencies, and inaccuracies in the quantification of cardiac output, more rapid completion of the study protocol. Complete stud- and have now been widely superseded by cMRI as an alterna- ies with optimal quality allowing estimation of LV and aortic tive reference method to quantitate MR in people. (Shanks et al. stroke volume were obtained for the final 12 consecutive dogs 2010, Cawley et al. 2013, Son et al. 2013) in a mean time of 34·5 min (±7·1 min). Mean cMRI-RF was cMRI in dogs is feasible, but technically challenging. During 17·6% (±17%). the present study, increasing operator experience led to a reduc-

Measurement of the first 5 complete studies by two indepen- tion in acquisition time and improvement in study quality. Ini- dent observers had excellent agreement for LV stroke volume tial cMRI attempts were compromised by difficulties applying and aortic stroke volume determinations, with an ICC≥0·99. acquisition guidelines for humans to the smaller, relatively more The 95% inter-observer limits of agreement assessed using the rounded hearts of dogs. Sub-optimal images occurred because Bland-Altman method were narrower for LV stroke volume than of incorrect image alignment, particularly in attempted short aortic stroke volume (−1·15 to 0·55 mL, bias −0·30 and −1·89 axis images. Poor image alignment was recognised by apparent to 0·07 mL, bias −0·56, respectively). The first of these five dogs foreshortening of the LV on VLA images, and inadvertent inclu- produced inconceivably low values for aortic stroke volume; the sion of the LVOT in four chamber studies or the right cardiac result was considered erroneous and the dog was excluded from chambers in two chamber views, resulting in oblique slice acqui- further analysis. An error in data storage precluded collection of sition for the short axis image series. Critical attention to cardiac data for the aortic stroke volume for another dog that therefore anatomy subsequently enabled increased accuracy with fewer was also excluded. attempts. Dogs were initially scanned in dorsal recumbency Complete cMRI data was available for 10 dogs to test cor- because this was the position used for neurological imaging; how- relations between cMRI-RF and echocardiographic variables ever, after changing to lateral recumbency it was found that alignment of image planes was easier, which also enabled increased (Fig. 5). VC:Ao (n=9) and Evel (n=10) had the strongest correlation with MR severity (r=0·89, 95% CI 0·60 to 0·97, p=0·001 accuracy of image acquisition. Finally, a marked improvement in and r=0·86, 95% CI 0·52 to 0·97, p=0·001, respectively). One R-wave recognition to trigger acquisition sequences was achieved dog had multiple MR jets that precluded VC measurement. by repositioning the ECG electrodes from the feet to the thorax. Correlations for other echocardiographic variables were LA:Ao Three major sources of artefact in cMRI are respiratory motion, (r = 0·78, 95% CI 0·30 to 0·94, p=0·008), Prolapse:Ao (r=0·72; cardiac motion and extremes of blood flow velocity (MacDonald 95% CI 0·16 to 0·93, p=0·02) and LVIDd:Ao (r=0·67, 95% CI et al. 2005). These were largely overcome by the use of breath- 0·08 to 0·92, p=0·032). The PISA method (n=8) had moder- holding, retrospective VCG gating and the SENSE™ body coil. ate correlation with cMRI-RF (r=0·75; 95% CI 0·09 to 0·95, Successful completion of the acquisition protocol was achieved p=0·03), but could not be measured in two dogs because of consistently after the first nine dogs. small jet size and the presence of multiple jets, respectively. Cal- Inter-observer reproducibility for the image analysis compo- culation of JAR was possible for all dogs but correlation with nent of this technique was excellent, supporting use of cMRI cMRI-RF was not significant (r=0·48, 95% CI −0·21 to 0·85, as a quantitative research tool. It would have been desirable to p=0·16). evaluate repeatability of the image acquisition component but

132 Assessment of mitral regurgitation in dogs

FIG 5. The associations between various echocardiographic variables and the cMRI-RF. Evel Peak early transmitral filling velocity, JAR Mitral regurgitant jet area indexed to left atrial area, LVIDd:Ao Left ventricular diastolic diameter indexed to aortic diameter, LA:Ao Left atrial diameter indexed to aortic diameter, PISA-RF Mitral regurgitant volume calculated by proximal isovelocity surface area method, Prolapse:Ao Amplitude of mitral prolapse indexed to aortic diameter, r Pearson’s correlation coefficient, VC:Ao Vena contracta indexed to aortic diameter. we could not justify further prolongation of anaesthetic time in related to poor alignment of velocity-encoded maps with aortic this group of patients. The dog excluded due to perceived inac- flow and inappropriate adjustment of velocity phase sensitivity. curate values for aortic stroke volume was the first dog for which Anaesthetic time constraints precluded detailed manual adjust- quantitative flow-analysis was attempted. Inaccuracies were likely ments for this dog.

133 J. Sargent et al.

Our study found that Evel and VC:Ao had the strongest cor- in all dogs in the present study, both the correlation of JAR with relations with cMRI-RF severity in dogs with relatively mild MR severity and its reproducibility were poor. Measuring the MMVD, although the confidence intervals for all correla- same echocardiographic studies limited the impact of operator, tions were large and overlapping. Mitral inflow E-wave veloc- patient and machine factors on inter-observer repeatability but ity independently predicts cardiac mortality in canine MMVD frame selection and subtle differences in measurement technique (Borgarelli et al. 2008, 2012, , Hezzell et al. 2012, Sargent et al. between observers likely contributed to the low ICC values. JAR

2015). The strong correlation between Evel with cMRI-RF reflects may still have a role in the assessment of multiple MR jets, but the increase in the LA to LV pressure gradient with increasing it does not appear to be suitable as a stand-alone measure of MR MR volume (Thomas et al. 1998). This is a simple, highly repeat- severity in dogs. able measurement and these results strongly support its inclusion The prognostic importance of chamber enlargement in dogs in the assessment of canine MMVD. with MMVD is well-established (Borgarelli et al. 2008, 2012, Vena contracta diameter is predictive of survival in canine Lord et al. 2010, Haggstrom et al. 2008, Hezzell et al. 2012) but MMVD (Sargent et al. 2015), but it is technially challenging these changes occur relatively late in dogs with asymptomatic to measure. Meticulous image acquisition is essential to obtain MMVD, so are less useful in early stages of this disease to identify repeatable results (Biner et al. 2010, Moraldo et al. 2013). Inter- patients at greatest risk of progression. Reproducibility of LA:Ao observer reproducibility for VC measurement was good, but infe- in this study was inferior to Evel and LVIDd. Measurement of rior to Evel. In people, reproducibility is significantly improved LA:Ao can be challenging, especially when measured from views when the effective regurgitant orifice is readily identifiable. We that include the entrance of the pulmonary vein into the LA as selected a frame in which the mitral valve was closed and where this can introduce uncertainty as to the limits of LA free wall. The the proximal surface area, VC and jet could be seen simultane- correlation between LA:Ao and cMRI-RF was similar to LVIDd. ously. As there is temporal variation in VC diameter during sys- Mitral valve prolapse is a common and early change in dogs tole (Zoghbi et al. 2003, Moraldo et al. 2013) the selection of with MMVD that may precede the development of MR or a heart the frame for VC measurement is likely to contribute to inter- murmur (Nakayama et al. 1996, Pedersen et al. 1999, Olsen et al. observer variability. Values for VC:Ao could not be obtained in 2003). Prolapse severity is likely to directly influence MR volume two dogs (whereas Evel could be measured in all dogs) because and is an independent predictor of survival in MMVD (Sargent of multiple MR jets. Summation of VC diameter measurements et al. 2015). Accurate assessment of mitral prolapse using 3DE in people with multiple MR jets does not accurately reflect MR has shown promise in people to identify patients at increased risk severity (Zoghbi et al. 2003) and we anticipated that the same of chordae tendineae rupture and disease progression (Lee et al. would be true in dogs, but there were insufficient numbers to 2013). We found a moderate correlation between prolapse and draw conclusions in this study. Summation of multiple VC MR severity in agreement with the results of previous studies areas measured using 3D echocardiography has been validated (Pedersen et al. 1999, Olsen et al. 2003, Terzo et al. 2009). Inter- in people (Hyodo et al. 2012), and the reproducibility of VC observer reproducibility for prolapse in this study was moderate. area measurement using 3D echocardiography is superior to 2D The MV has a saddle-shaped anatomy and exhibits translational methods for eccentric and multiple MR jets (Little et al. 2008, systolic motion (Komeda et al. 1996) and, depending on the Yosefy et al. 2009). view, it is easy to over- or underestimate the degree of prolapse. Although PISA-RF showed moderate correlation with cMRI- Previous studies have demonstrated good repeatability in mea- RF, inter-observer reproducibility was poor. In addition to shar- surements of mitral prolapse when categorized into absent, mild ing the limitations previously listed for VC measurement, the and severe prolapse rather than when considered as a continuous multiple components of the PISA-RF method provide many variable (Komeda et al. 1996, Pedersen et al. 1996). opportunities for introducing error (Enriquez-Sarano et al. 1995, It would be logical to assume that dogs with MMVD and Biner et al. 2010, Moraldo et al. 2013). The shape of the regur- worse MR would have more rapid disease progression but this gitant orifice area will alter the shape of the proximal flow con- has not been proven in dogs. Variables that identify severe MR vergence region and affect both the accuracy and reproducibility in asymptomatic dogs (such as VC:Ao and Evel) might be valu- of the calculated EROA value (Simpson et al. 1996, Moraldo able to identify dogs at greatest risk of progression. Certainly et al. 2013) and, in people, reproducibility is improved for cen- Evel , which correlated very well with cMRI-RF in this study, is tral regurgitant jets (Biner et al. 2010, Moraldo et al. 2013). This already widely established as an important prognostic indicator technique may be a more appropriate in dogs with “functional in MMVD (Sargent et al. 2015, , Borgarelli et al. 2008, 2012, MR”, associated with MV annulus dilation. Hezzell et al. 2012). The American Society of Echocardiography Measuring JAR provides a relatively rapid, qualitative evalu- and American Heart Association currently recommend a com- ation of MR jet size, shape and number, but values are affected posite approach to assess MR severity in people (Zoghbi et al. by patient, operator and machine factors, including alteration 2003) and further study is required to demonstrate whether a in haemodynamic loading conditions, LA function, transducer similar approach would be useful in dogs. frequency and gain settings (Enriquez-Sarano et al. 1993, 2005, The present study had several limitations. Importantly, the Chetboul & Tissier 2012). JAR likely underestimates the volume results of cMRI were not compared to a second reference tech- of eccentric jets, commonly encountered in MMVD (Chen et al. nique. This reflects the lack of a true “gold-standard” to quantify 1991, Grayburn et al. 2012). Although JAR could be measured MR. The small number of dogs included represents a relatively

134 Assessment of mitral regurgitation in dogs

narrow spectrum of disease and consequently may limit how the B orgarelli, M., Savarino, P., Crosara, S., et al. (2008) Survival characteristics and prognostic variables of dogs with mitral regurgitation attributable to myxoma- results might be applied to the wider population. This limita- tous valve disease. Journal of Veterinary Internal Medicine 22, 120-128 tion primarily reflects our reluctance to impose extended anaes- Borgarelli, M., Crosara, S., Lamb, K., et al. (2012) Survival characteristics and prognostic variables of dogs with preclinical chronic degenerative mitral thetic time on patients with more advanced cardiac disease, as valve disease attributable to myxomatous degeneration. Journal of Veterinary well as possible pre-selection bias if fewer dogs with severe car- Internal Medicine 26, 69-75 C awley, P. J., Hamilton-Craig, C., Owens, D. S., et al. (2013) Prospective compari- diac disease were referred for extensive neurological examina- son of valve regurgitation quantitation by cardiac magnetic resonance imaging tion. Measurement of cMRI-RF was not possible for the first and transthoracic echocardiography. Circulation: Cardiovasc Imaging 6, 48-57 C hen, C. G., Thomas, J. D., Anconina, J., et al. (1991) Impact of impinging wall jet nine dogs, however, the protocol was consistently achieved in on color Doppler quantification of mitral regurgitation. Circulation 84, 712-720 all remaining dogs. It is therefore plausible that after an initial C hetboul, V. & Tissier, R. (2012) Echocardiographic assessment of canine degenerative mitral valve disease. Journal of Veterinary Cardiology 14, 127-148 training period, this protocol should be feasible across the wider C hoi, H., Lee, K., Lee, H., et al. (2004) Quantification of mitral regurgitation using population of dogs. A further limitation was the requirement for proximal isovelocity surface area method in dogs. Journal of Veterinary Science 5, 163-171 general anaesthesia for cMRI protocols. No significant influence D ell’Italia, L. J., Blackwell, G. G., Pearce, D. J., et al. (1994) Assessment of ven- of anaesthesia on echocardiographic variables was demonstrated. tricular volumes using cine magnetic resonance in the intact dog. A comparison of measurement methods. Investigative Radiology 29, 162-167 This is likely, in part, due to the selection of anaesthetic protocols D i Marcello, M., Terzo, E., Locatelli, C., et al. (2014) Assessment of mitral regur- that had minimal cardiovascular effects, however, the study may gitation severity by Doppler color flow mapping of the vena contracta in dogs. Journal of Veterinary Internal Medicine 28, 1206-1213 have had insufficient power to detect changes in cardiac haemo- Drees, R., Johnson, R. A., Stepien, R. L., et al. (2015) Effects of two different anes- dynamics under anaesthesia. The decision was made in this study thetic protocols on cardiac flow measured by two dimensional phase contrast magnetic resonance imaging. Veterinary Radiology & Ultrasound 56, 168-175 to include LV trabeculae and exclude LV papillary muscles when E nriquez-Sarano, M., Tajik, A. J., Bailey, K. R. et al. (1993) Color flow imaging com- measuring LV stroke volume and this decision is likely to have pared with quantitative Doppler assessment of severity of mitral regurgitation: influence of eccentricity of jet and mechanism of regurgitation. Journal of the affected absolute values. Although superior repeatability is sug- American College of Cardiology 21, 1211-1219 gested when both are excluded in people this has not been evalu- E nriquez-Sarano, M., Miller, F. A., Jr., Hayes, S. N., et al. (1995) Effective mitral regurgitant orifice area: clinical use and pitfalls of the proximal isovelocity sur- ated in dogs (Papavassiliu et al. 2005). face area method. Journal of the American College of Cardiology 25, 703-709 This study represents pilot data for a small number of canine E nriquez-Sarano, M., Avierinos, J. F., Messika-Zeitoun, D., et al. (2005) Quantitative determinants of the outcome of asymptomatic mitral regurgita- patients with relatively mild MR. Despite these limitations, we tion. New England Journal of Medicine 352, 875-883 conclude that although technically challenging, cMRI is a feasi- G allach, R. G. & Mai, W. (2013) Cardiac MRI findings in a dog with a diffuse pericardial mesothelioma and pericardial effusion. Journal of the American Animal ble method to quantify mitral regurgitant fraction in small dogs. Hospital Association 49, 398-402 Although cMRI is not a validated “gold-standard” in dogs, these Garcia-Rodriguez, M. B., Granja, M. A., Garcia, C. C., et al. (2009) Complex cardiac congenital defects in an adult dog: an ultrasonographic and magnetic reso- results provide further support for E as an important echocar- nance imaging study. Canadian Veterinary Journal 50, 933-935

vel diographic indicator of MR severity in dogs with asymptomatic G ilbert, S. H., McConnell, F. J., Holden, A. V., et al. (2010) The potential role of MRI in veterinary clinical cardiology. Veterinary Journal 183, 124-134 MMVD. Vena contracta width has a similarly strong association G ouni, V., Serres, F. J., Pouchelon, J. L., et al. (2007) Quantification of mitral with MR severity but is technically challenging to measure and valve regurgitation in dogs with degenerative mitral valve disease by use of the proximal isovelocity surface area method. Journal of the American Veterinary is unsuitable for use in dogs with multiple MR jets. A composite Medical Association 231, 399-406 approach to assessing severity of MR in dogs may be useful and G rayburn, P. A., Weissman, N. J. & Zamorano, J. L. (2012) Quantitation of mitral regurgitation. Circulation 126, 2005-2017 this hypothesis warrants further investigation. H aggstrom, J., Duelund Pedersen, H. & Kvart, C. (2004) New insights into degenerative mitral valve disease in dogs. Veterinary Clinics of North America: Small Animal Practice 34, 1209-1226, vii-viii A cknowledgements Haggstrom, J., Boswood, A., O’Grady, M., et al. (2008) Effect of Pimobendan or Benazepril Hydrochloride on survival times in dogs with congestive heart fail- We gratefully acknowledge the financial support for this proj- ure caused by naturally occuring myxomatous mitral valve disease: the QUEST ect received from Novartis, PetSavers and the Veterinary study. Journal of Veterinary Internal Medicine 22, 1124-1135 H all, S. A., Brickner, M. E., Willett, D. L., et al. (1997) Assessment of mitral Cardiovascular Society, technical support from Mrs Peveril regurgitation severity by Doppler color flow mapping of the vena contracta. Johnstone of Philips HealthCare and advice from colleagues at Circulation 95, 636-642 Hansson, K., Haggstrom, J., Kvart, C., et al. (2002) Left atrial to aortic root the Royal Brompton and Harefield NHS Foundation Trust, par- indices using two-dimensional and M-mode echocardiography in cavalier King ticularly Consultant Cardiologist Dr Sonya V. Babu-Narayan. Charles spaniels with and without left atrial enlargement. Veterinary Radiology & Ultrasound, 43, 568-575 H ezzell, M. J., Boswood, A., Moonarmart, W., et al. (2012) Selected echocardio- Conflict of interest graphic variables change more rapidly in dogs that die from myxomatous mitral valve disease. Journal of Veterinary Cardiology 14, 269-279 None of the authors of this article has a financial or personal H yodo, E., Iwata, S., Tugcu, A., et al. (2012) Direct measurement of multiple vena relationship with other people or organisations that could inap- contracta areas for assessing the severity of mitral regurgitation using 3D TEE. JACC Cardiovasc Imaging 5, 669-676 propriately influence or bias the content of the paper. K im, J. H., Lee, M. S., Lee, S. Y., et al. (2013) Contrast echocardiography to assess left ventricular volume and function in Beagle dogs: comparison with 3-Tesla dual source parallel cardiac magnetic resonance imaging. Veterinary References Journal 198, 450-456 Alfakih, K., Plein, S., Thiele, H., et al. (2003) Normal human left and right ventricu- K ittleson, M. D. & Brown, W. A. (2003) Regurgitant fraction measured by using lar dimensions for MRI as assessed by turbo gradient echo and steady-state the proximal isovelocity surface area method in dogs with chronic myxomatous free precession imaging sequences. Journal of Magnetic Resonance Imaging mitral valve disease. Journal of Veterinary Internal Medicine 17, 84-88 17, 323-329 K omeda, M., Glasson, J. R., Bolger, A. F., et al. (1996) Three-dimensional B iner, S., Rafique, A., Rafii, F., et al. (2010) Reproducibility of proximal isovelocity dynamic geometry of the normal canine mitral annulus and papillary muscles. surface area, vena contracta, and regurgitant jet area for assessment of mitral Circulation 94, II159-163 regurgitation severity. JACC Cardiovascular Imaging 3, 235-243 L ee, A. P., Hsiung, M. C., Salgo, I. S., et al. (2013) Quantitative analysis of mitral B oddy, K. N., Sleeper, M. M., Sammarco, C. D., et al. (2011) Cardiac magnetic valve morphology in mitral valve prolapse with real-time 3-dimensional echocar- resonance in the differentiation of neoplastic and nonneoplastic pericardial diography: importance of annular saddle shape in the pathogenesis of mitral effusion. Journal of Veterinary Internal Medicine 25, 1003-1009 regurgitation. Circulation 127, 832-841

135 J. Sargent et al.

L ittle, S. H., Pirat, B., Kumar, R., et al. (2008) Three-dimensional color Doppler Schwammenthal, E., Chen, C., Giesler, M., et al. (1996) New method for accurate echocardiography for direct measurement of vena contracta area in mitral calculation of regurgitant flow rate based on analysis of Doppler color flow regurgitation: in vitro validation and clinical experience. JACC Cardiovasc maps of the proximal flow field. Validation in a canine model of mitral regur- Imaging 1, 695-704 gitation with initial application in patients. Journal of the American College of L opez-Alvarez, J., Dukes-McEwan, J., Martin, M. W., et al. (2011) Balloon dilation Cardiology 27, 161-172 of an imperforate cor triatriatum dexter in a Golden Retriever with concurrent Schwarz, T., Rossi, F., Wray, J. D., et al. (2009) Computed tomographic and mag- double-chambered right ventricle and subsequent evaluation by cardiac magnetic resonance imaging features of canine segmental caudal vena cava apla- netic resonance imaging. Journal of Veterinary Cardiology 13, 211-218 sia. Journal of Small Animal Practice 50, 341-349 L ord, P., Hansson, K., Kvart, C., et al. (2010) Rate of change of heart size before Shanks, M., Siebelink, H. M., Delgado, V., et al. (2010) Quantitative assessment congestive heart failure in dogs with mitral regurgitation. Journal of Small of mitral regurgitation: comparison between three-dimensional transesopha- Animal Practice 51, 210-218 geal echocardiography and magnetic resonance imaging. Circ Cardiovasc MacDonald, K. A., Kittleson, M. D., Reed, T., et al. (2005) Quantification of left ven- Imaging 3, 694-700 tricular mass using cardiac magnetic resonance imaging compared with echo- Simpson, I. A., Shiota, T., Gharib, M., et al. (1996) Current status of flow conver- cardiography in domestic cats. Veterinary Radiology & Ultrasound 46, 192-199 gence for clinical applications: is it a leaning tower of “PISA”? Journal of the M ai, W., Weisse, C. & Sleeper, M. M. (2010) Cardiac magnetic resonance imaging American College of Cardiology 27, 504-509 in normal dogs and two dogs with heart base tumor. Veterinary Radiology & Son, J. W., Chang, H. J., Lee, J. K., Chung, H. J., et al. (2013) Automated quanti- Ultrasound 51, 428-435 fication of mitral regurgitation by three dimensional real time full volume color M essika-Zeitoun, D., Topilsky, Y. & Enriquez-Sarano, M. (2013) The role of Doppler transthoracic echocardiography: a validation with cardiac magnetic res- echocardiography in the management of patients with myxomatous disease. onance imaging and comparison with two dimensional quantitative methods. Cardiology Clinics 31, 217-229 Journal of Cardiovascular Ultrasound 21, 81-89 Moraldo, M., Cecaro, F., Shun-Shin, M., et al. (2013) Evidence-based recommen- Terzo, E., Di Marcello, M., McAllister, H., et al. (2009) Echocardiographic assess- dations for PISA measurements in mitral regurgitation: systematic review, clini- ment of 537 dogs with mitral valve prolapse and leaflet involvement. Veterinary cal and in-vitro study. International Journal of Cardiology 168, 1220-1228 Radiology & Ultrasound 50, 416-422

Muzzi, R. A., de Araujo, R. B., Muzzi, L. A., et al. (2003) Regurgitant jet area by Thavendiranathan, P., Phelan, D., Collier, P., et al. (2012a) Quantitative assess- Doppler color flow mapping: quantitative assessment of mitral regurgitation ment of mitral regurgitation: how best to do it. JACC. Cardiovascular imaging severity in dogs. Journal of Veterinary Cardiology 5, 33-38 5, 1161-1175 Nakayama, T., Wakao, Y., Nemoto, H., et al. (1996) Mitral valve protrusion Thavendiranathan, P., Phelan, D., Thomas, J. D., et al. (2012b) Quantitative assessed by use of B-mode echocardiography in dogs with mitral regurgitation. assessment of mitral regurgitation: validation of new methods. Journal of the American Journal of Veterinary Research 57, 791-797 American College of Cardiology 60, 1470-1483 Olsen, L. H., Martinussen, T. & Pedersen, H. D. (2003) Early echocardiographic Thomas, L., Foster, E. & Schiller, N. B. (1998) Peak mitral inflow velocity predicts predictors of myxomatous mitral valve disease in dachshunds. Veterinary mitral regurgitation severity. Journal of the American College of Cardiology 31, Record 152, 293-297 174-179 Papavassiliu, T., Kuhl, H. P., Schroder, M., et al. (2005) Effect of endocardial Thrusfield, M. V., Aitken, C. G. G. & Darker, P. G. G. (1985) Observations on breed trabeculae on left ventricular measurements and measurement reproducibility and sex in relation to canine heart valve incompetence. Journal of Small Animal at cardiovascular MR imaging. Radiology 236, 57-64 Practice 26, 709-717 Pedersen, H. D., Lorentzen, K. A. & Kristensen, B. O. (1996) Observer variation Whitney, J. C. (1974) Observations on the effect of age on the severity of heart in the two-dimensional echocardiographic evaluation of mitral valve prolapse in valve lesions in the dog. Journal of Small Animal Practice 15, 511-522 dogs. Veterinary Radiology & Ultrasound 37, 367-372 Yosefy, C., Hung, J., Chua, S., et al. (2009) Direct measurement of vena contracta Pedersen, H. D., Lorentzen, K. A. & Kristensen, B. O. (1999) Echocardiographic area by real-time 3-dimensional echocardiography for assessing severity of mitral valve prolapse in cavalier King Charles spaniels: epidemiology and prog- mitral regurgitation. American Journal of Cardiology 104, 978-983 nostic significance for regurgitation. Veterinary Record 144, 315-320 Young, A. A., Orr, R., Smaill, B. H., et al. (1996) Three-dimensional changes in left Reynolds, C. A., Brown, D. C., Rush, J. E., et al. (2012) Prediction of first onset and right ventricular geometry in chronic mitral regurgitation. American Journal of congestive heart failure in dogs with degenerative mitral valve disease: the of Physiology 271, H2689-2700 PREDICT cohort study. Journal of Veterinary Cardiology 14, 193-202 Zoghbi, W. A., Enriquez-Sarano, M., Foster, E., et al. (2003) Recommendations Sargent, J., Muzzi, R., Mukherjee, R., et al. (2015) Echocardiographic predictors for evaluation of the severity of native valvular regurgitation with two-dimen- of survival in dogs with myxomatous mitral valve disease. Journal of Veterinary sional and Doppler echocardiography. Journal of the American Society of Cardiology 17, 1-12 Echocardiography 16, 777-802

136 CURRICULUM VITAE

Name: Paul Fridtjof Mõtsküla Date of birth: 15.02.1974 E-mail: [email protected] Institution: Estonian University of Life Sciences, Institute of Veterinary Medicine and Animal Sciences, Chair of Clinical Veterinary Medicine Kreutzwaldi 62, 51006 Tartu, Estonia Position: Lecturer

Professional employment: 2019–… Estonian University of Life Sciences, Institute of Veterinary Medicine and Animal Sciences – Lecturer in cardiorespiratory medicine (0.3) 2013–2019 Estonian University of Life Sciences, Institute of Veterinary Medicine and Animal Sciences – Lecturer in cardiorespiratory medicine (0.2) 2019–… Vetekar OÜ (Cardio-respiratory consultancy), Estonia – Veterinary cardiologist, founder, CEO 2015–2018 Anderson Moores Veterinary Specialists, UK – Veterinary cardiologist (1.0) 2010–2013 Royal Veterinary College, University of London, UK – Senior clinical training scholar (resident) in veterinary cardiology (1.0) 2004–2010 Estonian University of Life Sciences, Estonia – staff clinician (1.0) 1997–2004 Private veterinary practice

Education: 2014–2020 Estonian University of Life Sciences, Institute of Veterinary Medicine and Animal Sciences, Doctoral studies

137 2017 Royal College of Veterinary Surgeons (RCVS), UK Awarded: RCVS Recognised Veterinary Specialist in Small Animal Cardiology 2016 European College of Veterinary Internal Medicine- Companion Animals: Awarded: Dip ECVIM-CA (Cardiology), EBVS® European Veterinary Specialist in Small Animal Cardiology 2010–2013 Royal Veterinary College, University of London, UK. Senior clinical training scholarship (residency) in veterinary cardiology. Awarded degree: MVetMed (Master of Veterinary Medicine) 1998–2002 Estonian University of Life Sciences, Estonia. Awarded degree: MSc (Master of Veterinary Science) 1992–1997 Estonian University of Life Sciences, Estonia. Veterinary medicine. Awarded degree: DVM (Degree in Veterinary Medicine), cum laude 1981–1992 Gustav Adolf Grammar School, Estonia. Graduated with honours (“silver medal”)

Professional membership and administrative duties: 2019–… European College of Veterinary Internal Medicine (ECVIM-CA). Member of Cardiology Education Committee 2018–… European Board of Veterinary Specialisation (EBVS). National representative for Estonia 2017–… Royal College of Veterinary Surgeons, UK. Member; Specialist in Veterinary Cardiology 2016–… European College of Veterinary Internal Medicine - Companion Animals. EBVS® European Specialist in Small Animal Cardiology 2004–… European Society of Veterinary Cardiology. member

138 1999–… Estonian Veterinary Association. Member; Past president (2002–2006) 1998–… Estonian Small Animal Veterinary Association. Member

Participation in research projects: 2009–2012 Research project founded by the Estonian Research Council “Measuring of host inflammatory response as a research tool in clinical veterinary science” (8–2/ T9001VLVL)

Teaching activities: VL.1285 Väikeloomade sisehaigused I Small animal internal medicine I VL.1296 Animal physiology II; Loomafüsioloogia II VL.1298 Väikeloomade meditsiini erikursus Special course in small animal medicine VL.1292 Väikeloomade meditsiini õppepraktika Practical training in small animal medicine

139 ELULOOKIRJELDUS

Nimi: Paul Fridtjof Mõtsküla Sünniaeg: 15.02.1974 E-mail: [email protected] Töökoht: Eesti Maaülikool Veterinaarmeditsiini ja loomakasvatuse instituut Kliinilise veterinaarmeditsiini õppetool Kreutzwaldi 62, 51006 Tartu, Estonia Ametikoht: Lektor

Teenistuskäik: 2019–... Eesti Maaülikool, veterinaarmeditsiini ja loomakasvatuse instituut. Kliinilise veterinaarmeditsiini õppetool. Kardiopulmonoloogia lektor (0,30) 2013–2019 Eesti Maaülikool, veterinaarmeditsiini ja loomakasvatuse instituut. Kliinilise veterinaarmeditsiini õppetool. Kardiopulmonoloogia lektor (0,20) 2019–... Vetekar OÜ. Veterinaarkardioloog, tegevjuht. (1,00) 2015–2018 Anderson Moores Veterinary Specialists, Hursley, UK. Veterinaarkardioloog (1,00) 2010–2013 Kuninglik Veterinaarkolledž, Londoni Ülikool, Ühendkuningriik. Veterinaarkardioloogia resident (1,00) 2004–2010 Eesti Maaülikool, Veterinaarmeditsiini ja loomakasvatuse instituut, teraapia osakond. Loomaarst (1,00) 1997–2004 Veterinaar-erapraksis. Loomaarst (1,00)

Haridus: 2014–2020 Eesti Maaülikool, veterinaarmeditsiini ja loomakasvatuse instituut. Doktorantuur

140 2017 Kuninglik loomaarstide kolleegium (Royal College of Veterinary Surgeons - RCVS), Ühendkuningriik. RCVS tunnustatud veterinaarspetsialist väikeloomade kardioloogia alal 2016 Euroopa veterinaar-sisehaiguste kolleegium. Dip ECVIM-CA (kardioloogia) EBVS® Euroopa veterinaarspetsialist väikeloomade kardioloogia alal 2010–2013 Kuninglik Veterinaarkolledž, Londoni Ülikool, Ühendkuningriik. Vanem kliinilise koolituse stipendiaat (resident) veterinaarkardioloogia erialal Omistatud kraad: MVetMed (loomaarstiteaduse magister) 1998–2002 Eesti Maaülikool. Magistrantuur Omistatud kraad: MSc (veterinaarmeditsiini magistri teaduskraad) 1992–1997 Eesti Maaülikool. Loomaarstiõpe Omistatud kraad: DVM (loomaarstikraad), kiitusega (cum laude) 1981–1992 Gustav Adolfi Gümnaasium. Keskharidus, hõbemedaliga

Erialaorganisatsioonide liikmelisus, administratiivne tegevus: 2019–… Euroopa veterinaar-sisehaiguste kolleegium (ECVIM- CA). Kardioloogia hariduskomitee liige 2018–… Euroopa veterinaar-spetsialiseerumise nõukogu (European Board of Veterinary Specialisation – EBVS). Eesti esindaja 2017–… Kuninglik loomaarstide kolleegium (Royal College of Veterinary Surgeons – RCVS), Ühendkuningriik. Liige, veterinaarkardioloogia spetsialist 2016–… Euroopa veterinaar-sisehaiguste kolleegium. EBVS® Euroopa spetsialist väikeloomade kardioloogia alal. 2004–… Euroopa veterinaarkardioloogia ühing. Liige 1999–… Eesti loomaarstide ühing. Liige. President aastatel 2002– 2006

141 1998–… Eesti väikeloomaarstide selts. Liige

Osalemine uurimisprojektides: 2009–2012 Eesti Teadusagentuuri rahastatud projekt „Põletikuvastase mõõtmise kasutamine teadusuuringutes kliinilises veterinaarmeditsiinis“

Ülikoolis läbiviidav õppetöö: VL.1285 Väikeloomade sisehaigused I VL.1296 Loomafüsioloogia II VL.1298 Väikeloomade meditsiini erikursus VL.1292 Väikeloomade meditsiini õppepraktika

142 LIST OF PUBLICATIONS

1.1. Scholarly articles indexed by Web of Science, Science Citation Index Expanded, Social Sciences Citation Index, Arts & Humanities Citation Index and/or indexed by Scopus (excluding chapters in books)

Van Hoek, I., Hodgkiss-Geere, H., Bode, E., Hamilton-Elliott, J., Motskula, P., Palermo, V., Pereira, Y.M., Gulshaw, G.J., Ivanova, A., Dukes-McEwan, J., 2020. Associations among Echocardiography, cardiac biomarkers, insulin metabolism, morphology, and inflammation in cats with asymptomatic hypertrophic cardiomyopathy. J. Vet. Intern. Med. 34, 591–599.

McAulay, G., Borgeat, K., Sargent, J., Mõtsküla, P., Neves, J., Dukes- McEwan, J., Luis Fuentes, V., 2018. Phenotypic description of cardiac findings in a population of Dogue de Bordeaux with an emphasis on atrial fibrillation. Vet. J. 234, 111–118.

Wustefeld-Janssens, B.G., Burrow, R., Mõtsküla, P., Martin, M., Dukes- McEwan, J., 2016. Clinical findings and treatment outcomes for cats diagnosed with patent ductus arteriosus in the UK: a retrospective study of 19 cases (2004–2012). Vet. Rec. 179.

Jokelainen, P., Mõtsküla, P.F., Heikkinen, P., Ülevaino, E., Oksanen, A., Lassen, B., 2016. Dirofilaria repens microfilaremia in three dogs in Estonia. Vector Borne Zoonotic Dis. 16, 136–8.

Sargent, J., Connolly, D.J., Watts, V., Mõtsküla, P., Volk, H.A., Lamb, C.R., Luis Fuentes, V., 2015. Assessment of mitral regurgitation in dogs: comparison of results of echocardiography with magnetic resonance imaging. J. Small Anim. Pract. 56, 641–650.

Kaye, B.M., Borgeat, K., Mõtsküla, P.F., Luis Fuentes, V., Connolly, D.J., 2015. Association of tricuspid annular plane systolic excursion with survival time in Boxer dogs with ventricular arrhythmias. J. Vet. Intern. Med. 29, 582–588.

143 Mõtsküla, P.F., Linney, C., Palermo, V., Connolly, D.J., French, A., Dukes McEwan, J., Luis Fuentes, V., 2013. Prognostic value of 24-hour ambulatory ECG (Holter) monitoring in Boxer dogs. J. Vet. Intern. Med. 27, 904–912.

Procoli, F., Mõtsküla, P.F., Keyte, S. V., Priestnall, S., Allenspach, K., 2013. Comparison of histopathologic findings in duodenal and ileal endoscopic biopsies in dogs with chronic small intestinal enteropathies. J. Vet. Intern. Med. 27, 268–274.

3.4. Articles/presentations published in conference proceedings van Hoek, I., Hodgkiss-Geere, H., Bode, E., Mõtsküla, P., Palermo, V., Martinez-Pereira, Y., Lazxalde, J., Dukes McEwan, J., 2019. Diet-induced reduction of cardiac wall thickness, Troponin-I and IGF-1 in cats with asymptomatic hypertrophic cardiomyopathy. ECVIM-CA congress, Milan, Italy. van Hoek, I., Hodgkiss-Geere, H., Bode, E., Mõtsküla, P., Palermo, V., Martinez-Pereira, Y., Dukes McEwan, J., 2018. Associations between echocardiography, cardiac biomarkers, insulin metabolism, morphology and inflammation in feline asymptomatic hypertrophic cardiomyopathy. ACVIM Forum, Seattle, WA, USA.

Mõtsküla, P., 2018. Approach to dogs with incidental murmurs. FECAVA Eurocongress, Tallinn, Estonia.

Jokelainen, P., Mõtsküla, P., Heikkinen, P., Ülevaino, E., Oksanen, A., Lassen, B., 2015. Dirofilaria repens in three dogs in Estonia. 6th Conference of the Scandinavian-Baltic Society for Parasitology, Uppsala, Sweden.

Sargent, J., Watts, V., Volk, H., Mõtsküla, P., Lamb, C., Luis Fuentes, V., 2013. Quantitative magnetic resonance imaging of mitral regurgitation in dogs: development of a method. BSAVA Congress, Birmingham, UK.

Mõtsküla, P., Linney, C., Connolly, DJ., Chang, YM., Pace, C., Dukes McEwan, J., Luis Fuentes, V., 2012. Prognostic value of 24-hour

144 ambulatory ECG (Holter) monitoring in Boxer dogs. ACVIM Forum, New Orleans, LA, USA.

3.5. Articles/presentations published in local conference proceedings

Jokelainen, P., Mõtsküla, P., Heikkinen, P., Ülevaino, E., Oksanen, A., Lassen, B., 2016. Dirofilaria repens in three dogs in Estonia. Abstract in proceedings and poster session presented at: Annual Veterinary Congress, Helsinki, Finland.

Mõtsküla, P., 2019. Small animal cardiology. Autumn seminar of Estonian Small Animal Veterinary Association, Pärnu, Estonia.

5.1. Conference abstracts

Jokelainen, P., Mõtsküla, P., Heikkinen, P., Ülevaino, E., Oksanen, A., Lassen, B., 2015. Dirofilaria repens in three dogs in Estonia. Poster presented at: Conference “Veterinaarmeditsiin 2015”, Tartu, Estonia.

6.3. Popular science articles

Mõtsküla, P., 2018 Juhusliku kliinilise leiuna täheldatud südamekahina käsitlus koertel. Eesti Loomaarstlik Ringvaade, 3, 2–5.

Mõtsküla, P. 2016 Veterinaarmeditsiin, see on imelihtne... Eesti Loomaarstlik Ringvaade, 3, 25–27.

Russak, Õ.-M., Bogdanova, K., Mõtsküla, P., 2014. Kasvajate käsitlemine veterinaarias: Lümfoom. Eesti Loomaarstlik Ringvaade, 2, 3–9.

Ritson, K., Mõtsküla, P., 2013. Eesnäärmehaigused. Eesti Loomaarstlik Ringvaade, 3, 8–12.

145 Toom, M., Mõtsküla, P., 2009. Granulomatoosne hepatiit Kesk-Aasia lambakoeral - kliiniline haiguslugu. Eesti Loomaarstlik Ringvaade, 1, 9–12.

Mõtsküla, P., 2002. Ultraheliuuringu käigus mõõdetavate lineaarmõõtmete usaldusväärsus. Eesti Loomaarstlik Ringvaade, 2, 2–8.

Mõtsküla, P., Viitmaa, R., Kuks, A., 2002. Persisteeriv arterioosjuha koeral Kliiniline haiguslugu. Eesti Loomaarstlik Ringvaade, 1, 3–6.

Mõtsküla, P., 2001. Ultraheliuuringu käigus tekkivad artefaktid. Eesti Loomaarstlik Ringvaade, 3–4, 107–114.

146

VIIS VIIMAST KAITSMIST

JANAR KALDER SOLAR AND WIND ENERGY SEASONAL HEAT STORAGE SYSTEMS FOR RESIDENTIAL BUILDINGS IN NORDIC CLIMATE PÄIKESE- JA TUULEENERGIA SESOONNE SOOJUSSALVESTUS ELAMUTE KÜTTEKS PÕHJAMAISES KLIIMAS Professor Andres Annuk, dotsent Eugen Kokin 13. november 2020

OTTAR TAMM THE EFFECTS OF LAND USE AND CLIMATE CHANGE ON THE HYDROPOWER POTENTIAL IN ESTONIAN RIVERS MAAKASUTUSE JA KLIIMAMUUTUSE MÕJU EESTI JÕGEDE HÜDROENERGEETILISELE POTENTSIAALILE Dotsent Toomas Tamm 11. detsember 2020

ALEKSANDR SEMJONOV EVALUATION OF A FIXED-DOSE COMBINATION OF BUTORPHANOL- AZAPERONE-MEDETOMIDINE (BAM) FOR CHEMICAL IMMOBILISATION OF AFRICAN LION, BLESBOK, AND CHEETAH BUTORFANOOLI-ASAPEROONI-MEDETOMIDIINI (BAM) FIKSEERITUD ANNUSTE KOMBINATSIOONI KASUTAMINE AAFRIKA LÕVI, BLESBOKI JA GEPARDI KEEMILISEKS IMMOBILISEERIMISEKS Doktor Vladimir Andrianov, konsultant doktor Jacobus Phillippus Raath (Wildlife Pharmaceuticals South Africa) 17. detsember 2020

HELIS LUIK-LINDSAAR EESTI PÕLLUMAJANDUSETTEVÕTETE JA MAAPIIRKONNA KOHALIKE OMAVALITSUSTE TEHNILISE EFEKTIIVSUSE HINDAMINE ANDMERAJA ANALÜÜSI MEETODIGA EVALUATION OF TECHNICAL EFFICIENCY IN FARMS AND RURAL MUNICIPALITIES IN ESTONIA USING DATA ENVELOPMENT ANALYSIS Professor Rando Värnik, vanemteadur Ants-Hannes Viira 21. detsember 2020

KAIRI MAILEHT FÜTOPLANKTON JÄRVEDE SEISUNDI INDIKAATORINA PHYTOPLANKTON AS ECOLOGICAL QUALITY INDICATOR OF LAKES Professor Emeritus Ingmar Ott, juhtivteadur Peeter Nõges 19. veebruar 2021

ISSN 2382-7076 ISBN 978-9949-698-63-9 (printed) ISBN 978-9949-698-64-6 (pdf)