Neurohumoral response and pathophysiologic changes during progression of mitral regurgitation in the Cavalier King Charles spaniel

(1)

!"#$%&#'%$()$"+,%-+"(-.,(/&%,&0+1%)%2133&(-2"+.#$1-2

,$%2$"++1%-%4'1/$()$"2#$21/(/1%-1-/&"5(6()1"$71-25&($)"+*

+,(-1")*

!

8-."$+9$1:++%-;<()%=>$61

!

"#"$%&'#!$'((%)"'+,!

!

Department of Equine and Small Animal Medicine Faculty of Veterinary Medicine

University of Helsinki Helsinki, Finland

and

&-./012!345.627-4.!'.87-757/!940!&/6-:2;!)/8/20:<!

=/;8-.>-?!3-.;2.6!

!

To be presented for public examination, with the permission of the Faculty of Veterinary Medicine, University of Helsinki, in Auditorium XIV, Unioninkatu 34, Helsinki,

29.11.2013, at 12 o’clock noon.

!

! Helsinki 2013

(2)

Supervised by

Professor Emeritus Anna-Kaisa Järvinen, DVM, PhD,

Department of Equine and Small Animal Medicine, University of Helsinki, Finland

Professor Emeritus Peter Lord, BVSc, FRCVS,

Department of Clinical Sciences, Swedish University of Agricultural Sciences, Uppsala, Sweden

Professor Frej Fyhrquist, MD, PhD

Minerva Institute for Medical Research, Biomedicum Helsinki 2U, Helsinki, Finland

Reviewed by

Professor mso Lisbeth Høier Olsen, dr.med.vet., Department of Veterinary Disease Biology, Faculty of Life Sciences, University of Copenhagen, Denmark

Associate Professor (Cardiology) Karsten E. Schober, DVM, PhD, Dipl ECVIM- CA (cardiology), Department of Veterinary Clinical Sciences, !The Ohio State University! Veterinary Hospital, U.S.A.

Opponent

MD, PhD, docent Jarkko Magga, Department of Medicine, Oulu University Hospital, Finland

ISBN 978-952-10-9340-1 (paperback)

ISBN 978-952-10-9341-8 (pdf) http://ethesis.helsinki.fi Unigrafia

Helsinki 2013

(3)

''T To o m myy ffaam miillyy''

(4)

*

(6)

?@ABCDEF8GGH9IB8DBE!C

2D two-dimensional

ACEI angiotensin converting enzyme inhibitor AIC Akaike's information criterion

ANP pro–A–type (or atrial) natriuretic peptide ARJ regurgitant jet area

AUC area under the curve

BMI body mass index

BNP pro–B–type (or brain) natriuretic peptide BSA body surface area

cGMP cyclic guanosine monophosphate CHF congestive heart failure

CI 95% confidence interval CKCS Cavalier King Charles spaniel CMR compensated mitral regurgitation

CO cardiac output

CSA cross sectional area

CT computed tomography

CV coefficient of variation

DMR decompensated mitral regurgitation (i.e. heart failure) ECG electrocardiogram

EDD end-diastolic diameter EDRF endothelium-relaxing factor EDTA ethylenediaminetetraacetic acid EDV end-diastolic volume

eNOS endothelial (i.e. constitutional) nitric oxide synthase ESV end-systolic volume

ET-1 endothelin-1

FPRNA first pass radionuclide angiocardiography FSV forward stroke volume

HPLC high-performance liquid chromotography HR

Hr

hazard ratio heart rate

IQR interquartile range

LA/Ao left atrium to aortic root ratio LAA left atrial area

LAP left atrial pressure

LH left heart

LVEDD left ventricular end-diastolic diameter

(7)

LVESD left ventricular end-systolic diameter LVP left ventricular pressure

MMVD myxomatous mitral valve disease MR mitral regurgitation

MRI magnetic resonance imaging MRV mitral regurgitant volume NEP neutral endopeptidase

NO nitric oxide

NOS nitric oxide synthase

NOx nitrite (NO2-) and nitrate (NO3-) NP natriuretic peptide

NPR natriuretic peptide receptor (types A-C) nPTT heart rate normalized pulmonary transit time

NT–proANP N–terminal pro–A–type (or atrial) natriuretic peptide NT–proBNP N–terminal pro–B–type (or brain) natriuretic peptide NYHA New York Heart Association

PAP pulmonary arterial pressure PBV pulmonary blood volume PDE phosphodiesterase PH pulmonary hypertension

PISA Proximal Isovelocity Surface Area PVR pulmonary vascular resistance

QUEST Quality of life and Extension of Survival Time

RH right heart

ROA regurgitant orifice area

ROC receiver operating characteristic SD standard deviation

ß-AR beta adrenoreceptor(s1-3)

SVEP Scandinavian Enalapril Prevention study VHS vertebral heart system/score/scale/size VTI velocity time integral

(8)

J@ABCDEFEHBKB!8A<LGAB58DBE!C*

This thesis is based on the following original articles referred to in the text by their Roman numbers:

I Eriksson, A.S., Järvinen, A.-K., Eklund, K.K., Vuolteenaho, O.J., Toivari, M.H., Nieminen, M.S., 2001. Effect of age and body weight on neurohumoral variables in healthy Cavalier King Charles Spaniels. American Journal of Veterinary Research 62, 1818-1824.

II Lord, P., Eriksson, A., Häggström, J., Järvinen, A.-K., Kvart, C., Hansson, K., Maripuu, E., Mäkelä, O., 2003. Increased Pulmonary Transit Times in Asymptomatic Dogs with Mitral Regurgitation. Journal of Veterinary Internal Medicine 17, 824-829.

III Carlsson, C., Häggström, J., Eriksson, A., Järvinen, A.-K., Kvart, C., Lord, P., 2009. Size and shape of right heart chambers in mitral valve regurgitation in small-breed dogs. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 23, 1007- 1013.

IV Eriksson, A., Hansson, K., Häggström, J., Järvinen, A.-K., Lord, P., 2010. Pulmonary Blood Volume in Mitral Regurgitation in Cavalier King Charles Spaniels. Journal of Veterinary Internal Medicine 24, 1393-1399.

V Eriksson, A.S., Häggström, J., Duelund Pedersen, H., Hansson, K., Järvinen, A.-K., Haukka, J., Kvart, C., 2013. Increased NT-proANP predicts risk of congestive heart failure in Cavalier King Charles spaniels with mitral regurgitation caused by myxomatous valve disease.

Submitted to Journal of Veterinary Cardiology.

The copyright holders of these original publications and original figures reused in this thesis have kindly granted their permission to reproduce the articles and figures, respectively. Efforts have been made to receive permissions from all original copyright holders.

(9)

M@8GCDH85D

Myxomatous mitral valve disease (MMVD) causing mitral regurgitation (MR) is a common cause of heart failure in dogs. To date there is no treatment to slow the progression of the disease. The course of MVMD is highly variable but many dogs develop congestive heart failure (CHF). The progression of disease and prognosis has been assessed based on patient parameters such as heart rate, murmur grade, radiographic and echocardiographic heart size and recently measurement of blood natriuretic peptide (NP) levels. However, many aspects of pathophysiology affecting these measurements used for risk evaluation are poorly defined.

In humans N-terminal pro A-type natriuretic peptide (NT-proANP) levels increase with age. Whether this happens in dogs is not known. There are several reports on NT- proANP and NT-pro B-type natriuretic peptide (NT-proBNP) values in different disease stages of MR in dogs using traditional cross-sectional statistical methods, i.e.

one group is compared to another. However, in cross-sectional methods there is no way to control for individual change over time. Novel longitudinal statistical methods would control for random individual variation over time.

Aims. The objective of this study was to add to the knowledge of different pathophysiological processes affecting measures used in the diagnosis of dogs with mitral insufficiency. The specific questions were as follows:

Paper I – Does plasma concentration of NT-proANP and nitric oxide (NO) increase with age in healthy dogs, as it does in humans?

Paper II – Is there evidence of decreased heart pumping function before onset of CHF, specifically, does blood pulmonary transit time normalized to heart rate (nPTT) increase in heart failure, as reported in humans?

Paper III – Does right-sided heart enlargement contribute to heart enlargement seen on thoracic radiographs in dogs with mitral regurgitation?

Paper IV – Does pulmonary blood volume (PBV) increase in mitral regurgitation and is it associated with increased nPTT?

Paper V – Can CHF be predicted by increased plasma concentrations of NT- proANP or vital signs, such as heart rate and severity of heart murmur?

Methods. The material consisted of dogs with natural mitral regurgitation on placebo or enalapril, an angiotensin converting enzyme inhibitor (ACEI), as well as from separately recruited normal control dogs. Focus was put on measurement of plasma neurohumoral parameters, such as the NT-proANP and nitric oxide (NO), and first pass radionuclide angiocardiography (FPRNA) to evaluate overall heart pump function and possible right sided heart enlargement associated with pulmonary hypertension. Echocardiography and thoracic radiographs were used as reference methods.

(10)

Results. (Paper I) In normal dogs both NT-proANP and NO, measured as its plasma end-products nitrate and nitrite (NOx), increased with age (r= 0.69; P= 0.002 and 0.61; P= 0.01, respectively). Median plasma NT-proANP concentrations for dogs under 5.9 years of age was 230 pmol/l (inter quartile range [IQR] 190– 270) and for dogs older than 5.9 years 405 pmol/l (307– 530), with no overlap between groups. The corresponding discrimination was not seen for NOx. Paper II) Heart rate normalized pulmonary transit times (nPTT) increased with degree of MR. In healthy dogs without MR nPTT was 4.4± 0.6 (SD), dogs with MR not in failure 6.3± 1.6, and dogs with CHF 11.8± 3.4 (P< 0.001). (Paper III) The size of the right chamber increased late in severe MR. (Paper IV) Pulmonary blood volume (PBV) was strongly associated with nPTT (R²= 0.85, P< 0.0001), whereas there was no association to forward stroke volume (FSV; R²= 0.0004, P= 0.89). Increase in PBV appeared late in the preclinical phase before (CHF). (Paper V) NT-proANP was identified as an independent risk factor of CHF. The hazard ratio (HR) for every increase in 100 pmol/l of NT-proANP was 1.21 (P< 0.0002). The Kaplan-Meier median survival time estimate was 11 months (9-14 [95% Confidence Interval (CI)]) for dogs with plasma NT-proANP concentrations >1000 pmol/l, whereas it was 54 months (46 -> infinity) for dogs with concentrations " 1000 pmol/l (P< 0.0001). Dogs that developed CHF had a lower mean plasma level of NOx than dogs not reaching CHF (censored) during the follow- up period (23 vs. 28 µmol/l, P= 0.016). Heart rate (>130 vs. <115 beats/min, HR= 6.7, P<0.001) and heart murmur (HR= 9.5 for grade 3-4(/6) vs. 1-2, and HR= 17 for grade 5-6 vs. 1-2, respectively; both P<0.001) increased risk of heart failure.

Conclusions

Specific normal values for natriuretic peptides should be established for different age groups of dogs. Heart rate, heart murmur and plasma measurements of NT- proANP can be used to predict risk of and time to congestive heart failure (CHF) in dogs with MR. Heart rate normalized PTT (nPTT) is a robust measure of heart pump function and pulmonary blood circulation in MR. Both nPTT and pulmonary blood volume increase before onset of CHF. Apparent right-sided enlargement on radiographs is due to them being displaced by very large left heart chambers as they enlarge only in severe MR after onset of CHF.

(11)

N@H9IB9OEFDP9ABD9H8DLH9*

N@?!(/#$()&1+/%$0(-.",1."'1%)%20%4."2"-"$(/16"'1/$()6()6".1+"(+"

The mitral valve is situated between the left atrium (LA) and the left ventricle (LV), (Figure 1). The canine mitral valve complex consists of the annulus, two leaflets (anterior and posterior) of different shape and size, numerous chordae tendineae, and the papillary muscles (Fox 2012). The leaflets open in diastole to direct the blood flow from the LA to the left ventricle. When the heart is contracting, the mitral valves close to prevent backflow (regurgitation) of blood into the left atrium (Frater and Ellis 1961).

Under normal conditions blood entering the right ventricle (RV) via the right atrium (RA) is pumped through pulmonary arterial system at a mean pressure about one seventh that in the systemic arteries. The blood then passes through the lung capillaries, where carbon dioxide is released and oxygen is taken up. The oxygen-rich blood returns via the pulmonary veins to the LA, where it is pumped from the left ventricle to the periphery, thus completing the cycle. Pressure within peripheral vessels progressively decreases because the total cross-sectional area of smaller blood vessels increase. As a result, blood flow becomes slow in the capillaries, whereas flow (and pressure) is high near the heart. The total blood volume pumped from both ventricles is equal, but the volume in one area of the body may increase providing the blood in another region decreases. In exercise pulmonary and muscle blood volume increase, whereas blood volume in the intestinal system decreases. However, regulatory mechanisms synchronize the amount of blood pumped from right and left chambers and enable pulmonary blood pressure to remain unchanged as pulmonary blood volume (PBV) increase (Berne and Levy 1998). Proper closure of the mitral and

Figure 1 Cross-section of a heart showing the valves (left panel) and an echocardiography (right) of an enlarged heart due to mitral valve degeneration. The mitral valves are irregularly thickened and there is flail of the anterior leaflet (arrow) LA= left atrium, LV=

left ventricle, RV= right ventricle

(12)

tricuspid valves is of course essential for this synchrony to maintain.

In addition to the mitral valve leaflets, the chordae tendinae are a critical component of the functioning of the mitral valve apparatus (Fox 2012). These chordae connect the underside of the valve leaflets to the papillary muscles along the LV wall.

The cordae tendineae generally branch and are of variable thickness (Levine, Handschumacher et al. 1989). Each mitral leaflet receives chordae from both the anterior and posterior papillary muscles (Fox 2012). In order to withstand the high tensile stress applied to the chordae, they are comprised of highly organized collagen fibbers that are orientated in the direction of load (Connell, Han et al. 2012). As the left ventricle contracts during systole, the papillary muscles contract as well and the chordae in turn provide mechanical support to the mitral valve leaflets, preventing prolapse of valves into the LA during the high pressure of left ventricular systole (Silbiger and Bazaz 2009).

Descriptions of myxomatous mitral valve disease (MMVD) date back to early 20^th century (Jarcho 1975). In the early days the nodular mitral valve degeneration was considered equivalent to endocarditis (Munich 1935, Detweiler 1956). Since then several names have been proposed to describe the disease, among others, chronic valvular disease (CVD)(Detweiler, Patterson et al. 1961), endocardiosis (Jubb 1963), chronic mitral valve fibrosis (Das and Tashjian 1965), chronic myxomatous valve disease(CMVD) (Häggström, Hansson et al. 2000), degenerative mitral valve disease (DMVD) (Häggström, Duelund Pedersen et al. 2004), and chronic mitral valve insufficiency (CMVI) (Borgarelli, Savarino et al. 2008).

In chronic mitral valve disease, the valves gradually start leaking causing regurgitation of blood from the LV back into the LA. The lesions develop slowly over years, causing progressive mitral valve regurgitation, which overloads both the ventricle and the left atrium (called volume overload), and in many dogs results in CHF (Fox 2012). This is a critical event, requiring immediate treatment to save the life of the dog.

N@JP("'%.0-('133&(-2"+(-.3)1-13()+12-+%4'1/$()$"2#$21/(/1%-*

Abnormal coaptation of the mitral valve leaflets during systole creates a regurgitant orifice. The systolic pressure gradient between the left ventricle (LV) and the left atrium (LA) acts as the driving force for regurgitant flow, which results in a regurgitant volume. The regurgitant volume creates a volume overload state by entering the LA in systole and the LV in diastole. This ejection into the LA induces a low-pressure form of volume overload because the excess LV volume is ejected into the low-pressure left atrium, creating a unique form of haemodynamic stress of pure LV stretch in the absence of high pressure systolic load (Ahmed, McGiffin et al. 2009).

Chronic MR leads to dilatation of both LA and LV. Slowly progressing LA dilatation accommodates the regurgitant volume. This prevents LA pressure (LAP) and the pulmonary capillary pressure from rising rapidly.

(13)

In chronically instrumentated dogs with experimentally induced MR a diurnal rhythm of LAP was detected (Ishikawa, Tanaka et al. 2009). Left atrial pressures where slightly higher during daytime. Moreover, feeding caused an abrupt increase in LAP (average of 17 mmHg). In another experimental study, dogs with LAP> 30mmHg created signs of CHF (Miller, Eyster et al. 1986). Increased LAP was associated with decreased arterial oxygen tension, reduced lung volumes, and increased pulmonary vascular resistance.

Dogs running on a treadmill had, compared with the resting state, an increase in cardiac output (25–73%), a rise in mean pulmonary arterial pressure (PAP) from 18 to 28 mmHg, but no consistent change in mean left atrial pressure (LAP). Pulmonary vascular resistance (PVR) decreased during exercise while pressure difference between LA and PAP increased. At rest peak and mean flow velocities in the pulmonary artery averaged 136 and 35 cm/sec, respectively, and increased to 176 and 71 cm/sec, respectively, during exercise (Elkins and Milnor 1971).

The experimental data underlines the dynamic nature of cardiopulmonary blood circulation and suggests further studies to better understand the interaction of pump function, pulmonary blood flow and pressure (measured as nPTT) and mechanisms leading development of cardiogenic pulmonary edema.

N@J@?*Q1/$()*$"2#$21/(/1%-*(-.*$"'%."))1-2*.#"*/%*&("'%.0-('13*3&(-2"+*

The determinants of mitral regurgitant volume (MRV) are represented in the orifice equation, which is based on the fluid mechanical principle of Torricelli:

[1] !"# ! !"#!!!!!!!!"#"$!!! !!"# ! !"#!,

where LVP is left ventricular and LAP left atrial pressure, respectively. Thus, MRV is dependent upon the square root of the systolic pressure gradient between the LV and the LA, duration of systole (Tsyst), and regurgitant orifice area (ROA), where C is a constant (Ahmed, McGiffin et al. 2009).

In mitral regurgitation the LV is unloaded in both early and late systole via ejection into the low-pressure left atrium and this is reflected by a low LV mass and mass-to-volume ratio in patients with MR (Wisenbaugh, Spann et al. 1984). Dilatation of the LV occurs due to the addition of new sarcomeres in series (eccentric hypertrophy), enabling maintenance of forward stroke volume (FSV), despite severe MR (Adelmann 2011). Increased LV volume allows total stroke volume to increase, in turn increasing FSV, compensating for the volume lost to regurgitation. In addition the relatively thin LV wall enhances diastolic filling (Carabello 2008). Indeed MR is one of the very few cardiac diseases in which diastolic function is supernormal (Zile, Tomita et al. 1991). However, even though the diastolic enlargement maintain FSV (Kittleson 1998) and prevents a rise in resting heart rate (Häggström, Hamlin et al.

1996, Rasmussen, Falk et al. 2011), it has been shown in human patients, that both

(14)

inotropic and chronotropic responses failed to rise normally on exercise (Tan, Wenzelburger et al. 2009, Adelmann 2011).

In severe MR (Figure 2) volume adapting mechanisms fail to maintain forward stroke volume and end-systolic volume index increase (Borgarelli, Savarino et al.

2008). As a response, resting heart rate is increased and respiratory sinus arrhythmia is lost (Häggström, Hamlin et al. 1996, Kittleson 1998, Rasmussen, Falk et al. 2011). As regurgitation further continues to increase, the valves become incapable of preventing pulmonary venous pressure from exceeding the threshold for pulmonary edema or of maintaining forward cardiac output, a condition called ‘‘symptomatic’’ or decompensated MR or congestive heart failure (CHF) (Kittleson 1998, Häggström, Duelund Pedersen et al. 2004, Olsen, Häggström et al. 2010).

Until recently, much of the research efforts in MMVD have concentrated on detrimental left ventricular (LV) remodelling and its pathophysiology (Urabe, Mann et al. 1992, Dillon, Dell'Italia et al. 2012) in an expectation to find ways to inhibit the process. The gradual disease progression leads to mitral annulus stretch and enlargement of the regurgitant orifice, further increasing the regurgitant volume (Dillon, Dell'Italia et al. 2012). Consequently, LV remodelling is secondary to gradual increase in MR and a compensation to maintain cardiac output (CO). To a certain point remodelling is therefore considered beneficial for the patient but extreme or maladaptive remodelling, associated with decreased oxygen supply, contractile dysfunction, oxidative stress and myocardial damage, is characterized by poor prognosis.

In experimental models of canine MR ventricular cardiomyocytes demonstrate dysfunction based on decreased cell shortening and reduced intracellular calcium transients before chamber enlargement or decreases in contractility in the whole heart can be clinically appreciated (Dillon, Dell'Italia et al. 2012).

Figure 2. Colour Doppler flow showing retrograde turbulent regurgitation through leaking mitral valves from the left ventricle to the atrium

(15)

In dogs with natural MMVD, extensive remodelling has been associated with worse outcome (Borgarelli, Savarino et al. 2008). Dogs with a left ventricular end- diastolic diameter (LVEDD) to Ao ratio >3 have an increased risk of CHF (Reynolds, Brown et al. 2012). It has also been demonstrated that LVEDD normalized for body size (nLVEDD) (Cornell, Kittleson et al. 2004) change more rapidly in dogs that die due to MMVD than in controls (Hezzell, Boswood, Moonarmart et al. 2012). When MR progresses the apex of the LV becomes more rounded (Dillon, Dell'Italia et al.

2012) as measured by a decreased sphericity index (Serres, Chetboul et al. 2008).

Severe LV dilation with reduced eccentricity in experimental MR caused the interventricular septum to bulge and compress and displace the right ventricle (Dillon, Dell'Italia et al. 2012). If the RV is similarly displaced in natural MR, it might cause the apparent bulging of the right side of the heart seen on radiographs.

Efforts to attenuate remodelling and thereby improve outcome in natural MR have been unsuccessful. Preventive angiotensin converting enzyme inhibition (ACEI) by enalapril (the SVEP trial) had no effect on time to onset of CHF of dogs with preclinical (asymptomatic) MMVD (Kvart, Häggström et al. 2002) although the addition of enalapril to standard treatment of dogs with moderate to severe MR or dilated cardiomyopathy delayed onset of CHF (Ettinger, Benitz et al. 1998). However, a similar larger study showed only a modest increase in time to CHF in dogs with moderate to severe MR (Atkins, Keene et al. 2007). In an experimental model of subacute MR, an angiotensin II blocker decreased systemic vascular resistance but did not prevent adverse LV chamber and cardiomyocyte remodelling (Perry, Wei et al.

2002).

Altogether it is often assumed that most cases of chronic heart failure are associated with remodelling causing low cardiac output and reduced stroke volume – however, "the inconvenient truth is that stroke volume and cardiac output are often relatively normal" (MacIver and Dayer 2012). Accumulated data therefore suggests that remodelling is a consequence of MR not the primary cause of heart failure.

Preventive treatment should primarily target the degenerating valves. Moreover, when treatment of MMVD due to CHF is needed, the critical signs (pulmonary congestion and edema) are caused by increased pulmonary blood volume and pressure, not myocardial failure per se (Chin, Channick et al. 2009).

Development of secondary pulmonary hypertension (PH) in association with valvular dysfunction has been shown to be a marker of advanced disease (Kellihan and Stepien 2012). In humans with severe degenerative MR the development of PH, and particularly the presence of symptoms, was correlated to worst outcome (Rosenhek, Rader et al. 2006). To prevent secondary PH and detrimental myocardial remodelling the consensus treatment recommendation for humans with severe MR is mitral valve repair (Bonow, Carabello et al. 2008). Results are favourable in cases not complicated by secondary heart muscle cell injuries especially due to coronary artery disease (Bonow, Carabello et al. 2008, Warnes, Williams et al. 2008, Ahmed, McGiffin et al.

2009, De Meyer, De Keulenaer et al. 2010).

(16)

N@J@J*F(3/%$+*(44"3/1-2*%-+"/*%4*,#)'%-($0*"."'(*

Pulmonary vascular fluid leaves the capillary and enters the interstitial space in proportion to the net capillary hydrostatic pressure minus the net osmotic pressure across the vessel wall according to the fluid filtration equation (Starling 1896).

Interstitial fluid influx may be enhanced by an increase in the filtration pressure or a decrease in the oncotic pressure gradient (Sartori, Allemann et al. 2007). Until recently, the pathogenesis (and hence treatment) of pulmonary edema was based on this pathological finding of congested lungs leading to the concept that pulmonary edema is caused by fluid accumulation that because of severe heart failure is directed backward into the lungs (Cotter, Kaluski et al. 2001).

Accordingly, when the mitral valve fails, left atrial pressure (LAP) may increase substantially. The resulting increase in pulmonary capillary pressure forces excess fluid filtration through the pulmonary capillary walls and into the lung tissue. At first, the fluid collects within the lung interstitial space. If LAP exceeds a critical level of around 25 mmHg, the volume of edema fluid will overwhelm the capacity of the interstitial spaces and fluid will flood the airways and alveoli. This airway edema directly interferes with gas exchange, and it can kill the patient. However, although this concept is still acknowledged, survival of chronic cases indicates more complex adaptive mechanisms (Drake and Doursout 2002).

First, entry of diluted fluid into the interstitium lowers the interstitial osmotic pressure, and, according to Starling’s Law, this results in an increase of the absorption force (Sartori, Allemann et al. 2007). Second, interstitial matrix glycosaminoglycan content and connective tissue able to absorb fluid increase its volume by as much a 40% (Drake and Doursout 2002). Third, the lymphatic system which is responsible for the interstitial fluid removal, has at least an order of magnitude of reserve flow capacity (Sartori, Allemann et al. 2007). Forth, permeability studies in monolayers of rat pulmonary microvascular endothelial cells suggest that the hydraulic conductance of the alveolar capillaries may be much lower than previously thought, and may represent an additional safety factor against alveolar fluid flooding (Parker, Stevens et al. 2006). Finally, lung specimens obtained from patients with severe chronic heart failure presented alveolar fibrosis, while electron micrographs showed thickening of the capillary endothelial and alveolar epithelial cell basement membranes. These changes are thought to reduce the permeability of the alveolocapillary membrane (Sartori, Allemann et al. 2007). Despite the dramatical consequences of cardiac pulmonary edema little is known on pulmonary haemodynamics and possible changes related to severity of MR in dogs.

The pulmonary blood flow is regulated by a myriad of mediators, including bradykinin, prostacyclins, thromboxane A, angiotensin coverting enxyme (ACE and angiotensin, nitric oxide (NO) and endothelin-1 (ET-1) (Chin, Channick et al. 2009).!

One explanation for increased pulmonary vascular resistance in HF patients is that pulmonary circulation is a major source of ET-1, and ET-1 levels correlate with pulmonary vascular resistance (PVR)" Selective ET-1 antagonists could therefore be

(17)

useful in the treatment of PH (Ooi, Colucci et al. 2002) but clinical trials with ET-1 receptor antagonists in humans have been disappointing, probably because the beneficial haemodynamic effects of ET-1 antagonists are counterbalanced by a negative effect on myocardial cell survival and because of adverse effects (Schorlemmer, Matter et al. 2008).

Pulmonary edema and congestion, as a consequence of secondary PH due to severe MR, are readily diagnosed on thoracic radiographs. The presence or absence of pulmonary edema is often used as a gold standard or cut-off point in clinical studies on heart failure (Hansson, Häggström et al. 2009). However, few studies have assessed associations of pulmonary blood flow, pressure (assessed as nPTT) and volume (PBV) with severity of volume overload, such as in MMVD. To date no such clinical studies have been done in dogs with natural MMVD.

Normal pulmonary blood volume has been estimated in humans (Dock, Kraus et al. 1961, Lewis, Gnoj et al. 1970, Goto, Hirano et al. 1981, van der Walt, Van Rooyen et al. 1981) and animals (van der Walt, Van Rooyen et al. 1981, van Aarde, Littlejohn et al. 1984, van den Brom and Stokhof 1989, van Rooyen and van der Walt 1989). In normal subjects PBV increased in exercise up to 100% (Falch and Stromme 1979, Kuikka, Timisjarvi et al. 1979, Iskandrian, Hakki et al. 1981, Hopkins, Kleinsasser et al. 2007) while the blood volume of the spleen decreased 40% (Froelich, Strauss et al.

1988) but pulmonary transit time normalized to heart rate (nPTT) remained almost unaltered (Falch and Stromme 1979, Kuikka, Timisjarvi et al. 1979, Iskandrian, Hakki et al. 1981, Kuikka 2000). This means that the increased heart rate is matched by forward stroke volume (FSV) and increased PBV to keep nPTT stable, although PTT decreases [Formula 2]. Exercise induced maximal increase of cardiac output 220%, stroke volume 30%, and pulmonary blood volume 30% (Kuikka, Timisjarvi et al.

1979).!Thus regulation of PBV is dynamic both with regards to total volume and distribution and varies with posture and activity (Choi, Choe et al. 2003, Chon, Beck et al. 2006).

In athletes resting PBV and PTT was found to be increased compared to untrained humans (Falch and Stromme 1979). The increased volume of flowing blood and increased stroke volumes in athletes probably allows for a reduction in flow velocity causing increased PTT. However, humans with heart failure have a reduced capacity to tolerate activity induced increase in PBV and are prone to pulmonary edema (Hannan, Vojacek et al. 1981, Slutsky, Carey et al. 1982, Hirakawa, Suzuki et al. 1995, Butler, Chomsky et al. 1999, Agostoni, Cattadori et al. 2003).

A higher exercise to resting ratio of PBV has been reported in humans with coronary artery disease (Okada, Pohost et al. 1979). This may relate to generation of pulmonary hypertension (PH) during exercise (Magne, Lancellotti et al. 2010) and increase in pulmonary vascular resistance (PVR), as reported in asymptomatic humans with moderate mitral regurgitation due to MMVD (Magne, Lancellotti et al. 2010).

Normally there is a compensatory decrease in PVR as arterial pressure increase but approximately one-fourth of human patients with moderate resting MR developed severe MR during exercise. Exercise-induced changes in degenerative MR correlated

(18)

with changes in systolic PAP. In addition, marked exercise-induced increase in MR severity was associated with reduced symptom-free survival (Magne, Lancellotti et al.

2010).

Moreover, gravity affects regional pulmonary blood flow (lower regions hold more blood) (Friedman and Braunwald 1966). The ratio of blood between the upper (U) and the lower (L) parts of the lungs were higher in humans with MR than controls (mean U/L # 1 vs. 0.4, respectively in erect position) (Friedman and Braunwald 1966, Krishnamurthy, Srinivasan et al. 1972). The U/L ratio changed with position due to gravitation in controls, whereas this change was blunted in patients with MR (Friedman and Braunwald 1966).

Although pulmonary transit time increase in normal subjects with exercise, the heart rate normalized PTT (nPTT) does not. Because nPTT (and cardiopulmonary transit time) increase in humans with heart failure (Hannan, Vojacek et al. 1981, van der Walt, Van Rooyen et al. 1981, Liu, Kiess et al. 1986) nPTT could potentially be a useful measure of to what extent MR causes disturbances in pulmonary blood flow.

However, studies in dogs with natural MMVD have so far not been performed.

N@J@M*R"/"$'1-(/1%-*%4*,#)'%-($0*S)%%.*6%)#'"*

The transit time of blood (t) through an organ is determined by the flow (F) and the blood volume (V) of the organ: t = V/F. Pulmonary transit time (PTT) can be expressed standardized to heart rate by means of dividing PTT by mean R-R interval time (mRR) from subsequent ECG recordings. Then heart-rate normalized PTT (nPTT) is the number of stroke volumes that the pulmonary vascular bed holds at a given moment, described by the relationship (Dock, Kraus et al. 1961, Lewis, Gnoj et al. 1970, van der Walt, Van Rooyen et al. 1981):

[2] !"## ! !"# !"# !!" !"" ! !"# !"#!

which gives !"#! !!!"##!!!!"#,

where PBV= pulmonary blood volume, Hr= heart rate (= 1/mRR) and FSV by definition refers to forward SV, not total SV. Cardiac output (CO) in turn equals FSV times HR. Therefore, PBV equals PTT times CO. Pulmonary blood volume index (PBVI) is calculated by dividing PBV by body surface area (BSA).

N@J@N*5)1-13()*+12-+*1-*'1/$()*$"2#$21/(/1%-*

The first detectable sign of MMVD in dogs is a heart murmur. The severity of the regurgitation increases over years as the primary lesion progresses. Dogs that have MMVD and MR, but do not show clinical signs caused by heart failure or signs of left sided cardiomegaly (Table 1) (Atkins, Bonagura et al. 2009), have no problems to compensate for the small valve leakage by increasing the total stroke volume (SV) and possibly heart rate (Kittleson 1998). The regurgitant fraction at this stage has been estimated to be <30-50% in both dogs and humans (Kittleson and Brown 2003,

(19)

Ahmed, McGiffin et al. 2009). Compensatory mechanisms may balance the haemodynamics for a long time keeping even dogs with severe MR asymptomatic (Borgarelli, Crosara et al. 2012). However, overt clinical signs of congestive heart failure may develop rapidly. Common owner complaints include difficulty breathing (dyspnoea), breathing too rapidly (tachypnea), coughing, abdominal swelling (ascites), fainting (syncope), not eating well or not eating at all (anorexia), losing weight, or inability to exercise normally (exercise intolerance) (Kittleson 1998).

With time valve regurgitation progress to haemodynamically significant, as evidenced by radiographic or echocardiographic findings of left-sided heart enlargement, but dogs may stay asymptomatic (Atkins, Bonagura et al. 2009). If CHF develops suddenly dogs often require hospitalization, whereas more moderate and clinically stable can be managed at home (Atkins, Bonagura et al. 2009, Atkins and Häggström 2012).

When the regurgitant volume exceeds the capability of the heart to compensate with further increase in stroke volume, increase in heart rate help maintain forward stroke volume. Heart rate increase and heart rate variability decrease with degree of MR (Rasmussen, Falk et al. 2011). In humans, resting heart rate was a better predictor of heart failure than measures of LV function (Castagno, Skali et al. 2012). A substudy (Castagno, Skali et al. 2012) to the human CHARM trial (Pfeffer, Swedberg et al.

2003) assessed the association of heart rate and outcomes in a broad spectrum of patients with chronic heart failure. A significant risk of worse outcome in patients with high resting heart rate was found. The relationship of heart rate and outcomes was similar across LV ejection fraction category (Castagno, Skali et al. 2012). The study group resemble dogs with MMVD in which resting heart rate may be normal but rise significantly with severe worsening of MMVD (Häggström, Hamlin et al. 1996, Borgarelli, Savarino et al. 2008, Rasmussen, Falk et al. 2011).

To describe signs and progression of MMVD several classification schemes have been proposed. The classification schemes differ particularly in how they classify dogs with heart enlargement caused by significant haemodynamic changes but before dogs show overt signs of congestive heart failure. The ISACHC scheme (see Table 1 for details) was proposed in 1994 to serve the veterinary profession better than the human New York Heart Association (NYHA) heart failure classes scheme to classify dogs with heart disease. Yet, the ISACHC classification scheme rely much on owner provided information. The Scandinavian modified NYHA classification scheme was developed to provide better objective tools to classify dogs in the SVEP-study (see paper V for details) (Kvart, Häggström et al. 2002). The newest ACVIM scheme (American College of Veterinary Internal Medicine) is a consensus statement (Atkins, Bonagura et al. 2009) aiding more precise classification of dogs related to their need of treatment.

Nearly half of human patients with symptoms of heart failure are found to have a normal LV ejection fraction, called heart failure with normal (or preserved) ejection fraction (HFNEF or HFpEF, respectively). Yet they may have abnormal diastolic and/or systolic function (Sanderson 2007). This underlines the problems in

(20)

categorizing patients according to LV function. In a paper by Bonagura and Schober problems related to assessments of ventricular function in dogs with MMVD are reviewed (Bonagura and Schober 2009). Obviously, due to the nature of compensatory mechanism in MR, other than traditional LV measures may provide better risk stratification. Increased respiratory rate has been identified as one clinical sign with good sensitivity and specificity to detect CHF caused by MR in dogs (Schober, Hart et al. 2010). Pulmonary blood volume and nPTT are potential indices of the compliance of the pulmonary blood flow. As long as the pulmonary blood flow remain compliant, clinical signs of congestion are unlikely to develop. However, no clinical studies are available on the value of such measures for risk assessment of CHF in dogs with MR.

N@M<(/&%2"-"+1+%4."2"-"$(/16"'1/$()6()6".1+"(+"*

During the course of aging, the mitral valve remodels in generally predictable ways. Myxomatous (degenerative) mitral valve disease (MMVD) is associated with aging in both dogs and humans (Pedersen and Häggström 2000, Orton, Lacerda et al.

2012). However, the connection between these aging changes and the morbidity and mortality that accompany pathologic conditions has not been clarified (Connell, Han et al. 2012).

Table 1. The difference and overlap of three commonly used heart failure classification systems are described; simplified and combined from sources: ACVIM, American College of Veterinary Internal Medicine (Atkins, Bonagura et al. 2009); Scandinavian modified NYHA, New York Heart Association (Kvart, Häggström et al. 2002) and (The Criteria Committee of the New York Heart Association 1994); ISACHC, International Small Animal Cardiac Health Council (ISACHC 1994)

(21)

Myxomatous valvular remodelling is associated with characteristic histopathologic features. Changes include expansion of extracellular matrix with glycosaminoglycans and proteoglycans; valvular interstitial cell alteration; and attenuation or loss of the collagen-laden fibrosa layer (Olsen, Mortensen et al. 2003, Disatian, Ehrhart et al.

2008, Fox 2012). The degeneration is most pronounced at the free margins of the leaflets (Disatian, Ehrhart et al. 2008, Aupperle, Marz et al. 2009). The elastic fibers become fragmented and split and collagen bundles disorientated by glycosaminoglycan infiltration. The collagen fibrils also appear disrupted. Similar changes occur in the chordae tendineae (Akhtar, Meek et al. 1999, Aupperle and Disatian 2012).

In studies of experimental MR in dogs the gradual disease progression leads to mitral annulus stretch and enlargement of the regurgitant orifice, further increasing the regurgitant volume (Dillon, Dell'Italia et al. 2012). This feedback mechanism is thought to be the reason for accelerating heart and chamber size in the natural disease before onset of CHF (Lord, Hansson et al. 2010, Lord, Hansson et al. 2011).

Elongation and rupture of chordae tendineae is a common complication of MMVD that is considered a significant event leading to acute worsening of mitral regurgitation and sudden pulmonary edema resulting in a poor prognosis (Ettinger and Buergelt 1969, Olsen, Mortensen et al. 2003). In a study comprising 706 dogs with MMVD 114 (16%) was diagnosed with ruptured chordae tendinae. In 96% of dogs the ruptures were detected in the anterior leaflet. The median survival time was 425 days (Serres, Chetboul, Tissier et al. 2007). The survival time was longer than previously reported, though it is likely that dogs included had rupture of minor chordae. Dogs lost from follow-up (50%) may also have biased the result (Borgarelli and Buchanan 2012).

Severe MMVD can generate high velocity jets of mitral regurgitation that strike and injure left atrial endocardium. Such jets may traumatize the atrial endocardium and result in focal, fibrotic, endocardial contact lesions (known as ‘jet’ lesions). Severe lesions can induce endocardial tears (Liu 1999). Dogs may survive small and partial thickness tears of the atrial wall and affected animals may have several tears (Buchanan 1964, Fox 2012). Left atrial rupture with acute cardiac tamponade is an uncommon but catastrophic sequela (Buchanan 1977, Olsen, Häggström et al. 2010).

There are close similarities between human mitral valve prolapse (MVP), i.e.

abnormal systolic protrusion of mitral valve leaflets into the left atrium, and canine MMVD (Pomerance and Whitney , Pedersen and Häggström 2000). In both species, the principal macroscopic findings are enlarged, thickened leaflets, interchordal hooding and elongated chordae tendineae (Pedersen and Häggström 2000).

The progression of MMVD is slow and characterized by a long preclinical period with no or very slow increase in heart size (Borgarelli, Savarino et al. 2008). Affected geriatric dogs may therefore die for other reasons before progressing to CHF.

However, at some point the rate of increase in heart size escalates (Lord, Hansson et al.

2010). Typically this will lead to congestive heart failure (CHF) within approximately one year after the beginning of escalation before symptoms of CHF appear (Lord, Hansson et al. 2011, Reynolds, Brown et al. 2012).

(22)

N@N<$"6()"-3"%4'1/$()6()6".1+"(+"

Myxomatous mitral valve disease (MMVD) (Pedersen and Häggström 2000) is encountered in all breeds but is most prevalent in small to medium sized dog breeds, such as the Dachshund, Cavalier King Charles and Cocker spaniel, Poodle, Schnauzer, Chihuahua and some Terriers (Buchanan 1999).

Depending on country and breed distribution MMVD is estimated to represent around 75% of all cardiovascular disease in dogs (Häggström, Hansson et al. 1992, Buchanan 1999, Egenvall, Bonnett et al. 2006, Olsen, Häggström et al. 2010). In England 50% of Cavalier King Charles (CKC) spaniels have a murmur due to MR by the age of 5–6 years and at 10 years of age, the prevalence of murmurs approaches 100% (Darke 1987). Examinations of 160 Cavaliers in Australia revealed mitral murmurs in only 25% of dogs over 4 years of age which led the authors to conclude that MMVD prevalence in this breed was lower in Australia than in other countries (Malik, Hunt et al. 1992).

Mildly affected valves can function adequately and the lesions have no haemodynamic effect. Yet mitral valves from dogs with normal cardiac physical and echocardiographic examinations may present mild myxomatous changes on necropsy (Borgarelli, Tursi et al. 2011). The diagnosis of MMVD is therefore dependent on diagnostic methods used (Pedersen, Lorentzen et al. 1999) affecting the detected prevalence.

Because the incidence (risk of developing a new condition, or the amount of new disease cases, within a specified period of time) of MMVD increases with age (Whitney 1974, Häggström, Hansson et al. 1992), the reported prevalence (proportion of a population with a condition at a point (or period) of time) of MMVD also increases with age. The prevalence of MMVD therefore varies with whether a study is based on clinical or necropsy material (Borgarelli and Buchanan 2012).

N@TG$"".(-.+"U.1+,%+1/1%-*

The majority of the breeds at elevated risk for MMVD are small or toy breeds with average adult weights under 9 kg. However, for reasons not known, e.g. the Brussels griffon seems to be an exception to this rule (Parker and Kilroy-Glynn 2012). Another breed at low risk is the West Highland White terrier in Finland (Anna-Kaisa Järvinen, personal communication). Although the disease is more commonly diagnosed in small- breed dogs, it can also occur in large-breed dogs, such as the German Shepherd and the Great Dane (Thrusfield, Aitken et al. 1985, Borgarelli, Zini et al. 2004). In some breeds, such as the Cavalier King Charles spaniel (CKCS), approximately half of all dogs have developed a cardiac murmur by 5–6 years of age, and by the age of 10 years, most dogs are affected (Häggström, Hansson et al. 1992, Beardow and Buchanan 1993).

(23)

Dachshunds are less predisposed to the condition, but by about 10 years of age, half of the population has developed a regurgitant murmur (Olsen, Fredholm et al.

1999). Because CKCS may develop heart murmurs at younger age than other breeds they may have a higher rate of death caused by MMVD than other breeds (Häggström, Hansson et al. 1992), albeit later research has shown that, after progression to CHF, Cavaliers in fact have a better outcome than other breeds (Häggström, Boswood et al.

2008).

In conclusion, the reports on prevalence of MMVD in different breeds are inconclusive, probably much because samples are either regional, biased by insufficient sample size or biased selection of, e.g. at dog shows or referral hospitals.

Male dogs are reported to have a higher risk than females of developing MMVD (Thrusfield, Aitken et al. 1985). Males may also develop the disease at a younger age than females (Sisson, Kvart et al. 1999), although this has not been confirmed in all studies (Beardow and Buchanan 1993, Pedersen, Lorentzen et al. 1999). MMVD is also reported to progress faster in male than in female Dachshunds (Olsen, Fredholm et al. 1999). However, in a study including several breeds, males did not have an increased risk of mortality (Borgarelli, Savarino et al. 2008). Possibly unrelated to canine MR but of interest is that women had more severe regurgitation, when indexed for body size, and higher mortality than men (Avierinos, Inamo et al. 2008). The conflicting results on the effect of gender imply the need for further long-term studies in dogs.

N@VP"$1/(S1)1/0

A polygenic mode of inheritance has been proposed in Dachshunds (Olsen, Fredholm et al. 1999) and CKCS (Swenson, Häggström et al. 1996). Heritability estimates of 0.67 (± 0.071, standard error) for the grade of murmur and 0.33 (± 0.072) for the presence/absence of murmur in CKCS have been estimated (Lewis, Swift et al.

2011). Recently, two loci associated with development of MMVD in CKCS were identified (Madsen, Olsen et al. 2011). Some of the genes in these regions are good candidates for the development of MMVD because they play a role in the composition of connective tissue, collagen formation, and deposition of proteoglycans and hyaluronates (Madsen, Olsen et al. 2011). These are thought to be some of the causes of normal degeneration of mitral valves that are accentuated in dogs with MMVD (Connell, Han et al. 2012).

The choice of parents with no or low intensity murmurs can decrease the morbidity in offspring (Swenson, Häggström et al. 1996, Olsen, Fredholm et al. 1999).

Obviously though, parents have to be old enough to manifest their predisposition, otherwise the result of a breeding programme will be incomplete, as was found by impaired breeding results in Sweden (Lundin and Kvart 2010).

(24)

N@WD1'"3%#$+"1-/&",$%2$"++1%-%4'1/$()6()6".1+"(+"

Several studies have described the time course of MMVD. In a study including 558 dogs counting 36 breeds, more than 70% of asymptomatic dogs were alive at the end of the follow-up period of 6.6 years (Borgarelli, Savarino et al. 2008). In another study comprising 55 asymptomatic dogs, 82% of dogs were still asymptomatic at 12 months from inclusion in the study (Chetboul, Serres et al. 2009). A study aimed at evaluating the efficacy of treatment with enalapril (an ACEI) in delaying the onset of heart failure (VETPROOF) showed a median time before onset of CHF of 2.3 years for the treated group and 2.1 for the placebo group (Atkins, Keene et al. 2007). The SVEP study, with the same aim but including only Cavalier King Charles spaniels, reached similar results (Kvart, Häggström et al. 2002). Data from a study evaluating natural history and risk factors for progression of MMVD in the preclinical period, revealed a median survival time of 28 months in 256 dogs with normal heart size and 27 months in dogs with heart enlargement. The same study showed that 13% of dogs progressed to a more advanced stage of MMVD during the observation period (Borgarelli, Crosara et al. 2012).

The aforementioned reports provide evidence that asymptomatic MMVD is a relatively benign condition similar to what has been reported in people with asymptomatic MR due to MMVD (Freed, Levy et al. 1999, Singh, Evans et al. 1999, Enriquez-Sarano, Akins et al. 2009). However, the calculated time from onset of disease until CHF will naturally be affected by the severity of MR in the study material recruited. To date we have few methods to objectively assess haemodynamics, activation of compensatory mechanisms and overall pump function.

LA size is associated with worse outcome has been documented both in humans (Cleland, Coletta et al. 2005, Bonow, Carabello et al. 2008, Borg, Pearce et al. 2009, Hunt, Abraham et al. 2009, Le Tourneau, Messika-Zeitoun et al. 2010) and dogs (Häggström, Hansson et al. 2000, Hansson, Häggström et al. 2002, Bernay, Bland et al.

2010, Hezzell, Boswood, Chang et al. 2012, Reynolds, Brown et al. 2012). Even though the regurgitant fraction correlates with the regurgitant orifice area (ROA), which correlates with the degree of MR, the compliance of LA (denoted as LAP - LVP, Formula 2, page 23) and the pulmonary vascular bed may affect individual adaptation to increased regurgitation volumes. In support of this, LA volume indexed to body surface area (BSA) was associated with survival in human patients with MR independent of regurgitant volume (Le Tourneau, Messika-Zeitoun et al. 2010). The authors’ interpretation was that LA volume integrates not only MR degree but also other measures of the valve disease severity as well as compliance of LA. Therefore, although an association is expected, the degree of regurgitation and size of LA are probably not directly related to pulmonary blood pressures and volumes. Further research is needed to explore these relationships as well as their relationship to clinical signs of MR.!

The survival time has been studied to determine the effects of different treatments (for a comparison, see Atkins and Häggström (2012). From a clinical research point of

(25)

view, different classification schemes (Table 1) make conclusions difficult to draw when comparing treatment studies. Within all schemes the period with an enlarged heart due to MR, but no clinical signs, may in fact last for years (Kvart, Häggström et al. 2002, Borgarelli, Savarino et al. 2008). Because we cannot know when during that time period dogs have entered a study, this will heavily influence estimated survival times when CHF is the chosen outcome. However, the results of these studies suggest that the survival of dogs with moderate to severe CHF caused by MMVD typically is below one year but may reach 2 years.

N@X* Q"/&%.+* /%* (++"++* ,(/&%,&0+1%)%20* %4* ."2"-"$(/16"* '1/$()* 6()6"*

.1+"(+"*

N@X@?*8#+3#)/(/1%-*

Since early days auscultation has been the way to diagnose heart murmurs related to heart failure. The earliest description of the heart sounds comes from William Harvey’s "De Motu Cordis" in 1628, which first described human blood circulation correctly. The first stethoscope, a cylinder of wood perforated in its entre longitudinally, was invented by Laennec in 1816 (Hanna and Silverman 2002). Efforts to refine diagnostics with phonocardiography has been made since the 50's (Mannheimer 1946, Gardiner 1957), electronic stethoscopes (Grenier, Gagnon et al.

1998) and computer assisted analysis (Watrous, Thompson et al.

2008) without substantial benefit over ordinary auscultation (Grenier, Gagnon et al. 1998, Watrous, Thompson et al. 2008). Significant associations have been detected between severity of MR and signal analysis sound variables (e.g.

frequency, first frequency peak, energy ratio) (Ljungvall, Ahlström et al. 2009). Future signal analysis could therefore add to normal auscultation in assessment of the severity of MR.

In MMVD a systolic regurgitant murmur is best heard over the left cardiac apex. In the early phases the murmur may be intermittent. The intensity of the murmur has been correlated with the severity of MMVD (Häggström, Figure 3. Drawing of the first stethoscope

(1819 by Laennec) made of perforated wooden

(26)

Kvart et al. 1995, Ljungvall, Ahlström et al. 2009). In the very early stage a midsystolic click may be the only auscultatory finding, although the low intensity may facilitate detection only on phonocardiography. This click is considered a reliable indicator of mitral valve prolapse in humans (Pedersen and Häggström 2000).

In many mildly affected dogs, physical manoeuvres may increase murmur intensity (Pedersen, Häggström et al. 1999). In more severe cases of MMVD the regurgitant murmur becomes holosystolic and is frequently heard on the right side of the thorax as well (Häggström, Kvart et al. 1995, Pedersen, Häggström et al. 1999).

Large-breed dogs affected by MMVD are more often present with atrial fibrillation and myocardial failure. The intensity of murmurs may therefore not correlate to the severity of the heart failure (Borgarelli, Zini et al. 2004).

Experienced cardiologists perform better than inexperienced in classifying degrees of murmurs as well as distinguishing regurgitant murmurs from ejection flow murmurs (Pedersen, Häggström et al. 1999). Ejection flow murmurs or mild regurgitant murmurs have been described in athletes (Mukerji, Alpert et al. 1989, Barrett, Ayub et al. 2012), racing dogs (Marin, Brown et al. 2007, Bavegems, Duchateau et al. 2011) and cats (Nakamura, Rishniw et al. 2011), without apparent echocardiographic evidence of heart disease. Although innocent physiologic flow murmurs may be distinguished from regurgitant murmurs as low grade high frequency murmurs near the base of the heart (Pedersen, Häggström et al. 1999). The ACVIM consensus statement (Atkins, Bonagura et al. 2009) recommends that echocardiography should be performed to answer questions regarding the cause of the murmur and presence of cardiac enlargement in dogs with suspected MMVD. Not surprisingly, presence of murmur has been identified as a risk factor for cardiac death (Borgarelli, Crosara et al.

2012). However, so far there are no papers published on increase in risk with change in murmur intensity.

N@X@J*H(.1%)%20*

Although echocardiography in many ways is more specific than radiography in the diagnosis of heart disease, thoracic radiography is still a key diagnostic tool in dogs presented with signs of cardiac or respiratory disease. Cough and dyspnoea are signs common to both categories of disease. The value of radiography is in the assessment of the haemodynamic consequences of chronic MR (global heart size and presence of pulmonary congestion and edema). It also helps to exclude other possible causes of respiratory signs (Häggstrom, Kvart et al. 2005, Hansson, Häggström et al. 2009), because respiratory signs are unlikely the reason for respiratory signs in a dog with normal heart size (Lamb, Tyler et al. 2000, Guglielmini, Diana et al. 2009). Thoracic radiographs are also often used in clinical studies to objectively verify subjective clinical signs of CHF. However, pulmonary findings may be inconclusive, since early radiographic changes of pulmonary interstitial edema and bronchial patterns resemble the radiographic appearance of chronic airway disease (Häggstrom, Kvart et al. 2005).

(27)

Therefore the determination of onset of CHF is made by both clinical and radiographic evidence (Hansson, Häggström et al. 2002).

The vertebral heart system (VHS, also called VH score, scale or size) was introduced as an index of heart size (Figure 4) independent of the size of the dog (Buchanan and Bucheler 1995). The lengths of the long and short axes of the heart are scaled against the length of vertebrae dorsal to the heart beginning with the fourth thoracic vertebra. The sum of the long and short axes of the heart (scaled on the vertebrae) is the VHS (Buchanan 2000). The origin of the left main stem bronchus, seen in cross section, is suggested as the reference point at the heart base (Hansson, Häggström et al. 2005). An increased VHS is consistent with heart size enlargement as seen in advanced chronic MMVD (Lamb, Tyler et al. 2000, Nakayama, Nakayama et al. 2001, Saida, Tanaka et al. 2006, Häggström, Boswood et al. 2008, Borgarelli and Buchanan 2012, Reynolds, Brown et al. 2012). VHS has been used as an index in studies measuring the rate of increase heart size!(Lord, Hansson et al. 2010, Lord, Hansson et al. 2011).

The VHS method for heart size is reported to be independent of observer experience, but dependent of individual observers selection of reference points and transformation of long and short axis dimensions into VHS units (Nakayama, Nakayama et al. 2001, Hansson, Häggström et al. 2009). Values significantly greater than the original published reference VHS of 9.7±0.5 have been reported in whippets (11.3±0.5) (Bavegems, Caelenberg et al. 2005), greyhounds (10.1±0.2) (Marin, Brown et al. 2007), American Pittbull Terrier (10.9±0.4) (Cardoso, Caludino et al. 2011), as well as Pomeranian, Bulldog, and Boston Terrier breed groups (Jepsen-Grant, Pollard et al. 2013). This is because in the original paper VHS was a mean derived from several breeds. Also in contrast to the Buchanan publications, later work has reported a

Figure 4. The long and the short axis of the heart is measured and measurements are superimposed on the vertebrae starting at the 4^th vertebra. The sum of the number of vertebrae is the vertebral heart score (VHS)

(28)

significant effect on calculated VHS between radiographs taken in right vs. left recumbency (9.8±0.6 vs. 9.5±0.8; P<0.0004) (Greco, Meomartino et al. 2008).

Whether and to what extent aforementioned breed related differences are affected by altered measurement methods (Kosic, Krstic et al. 2007, Greco, Meomartino et al.

2008) remains unsolved.

An increase in VHS above normal reference values and signs of venous congestion and pulmonary infiltrates indicate CHF (Kvart, Häggström et al. 2002) but, as pointed out before, radiographic changes are sometimes inconclusive. The dog could potentially benefit from medication before clinical signs of CHF recognizable on radiographs develop. Recognition of early changes in pulmonary haemodynamics and blood volume in dogs with natural MMVD warrants other clinical diagnostic methods so far not described.

N@X@M*93&%3($.1%2$(,&0*

With the introduction of high-resolution colour Doppler ultrasound equipment, echocardiography has become the most frequently utilized procedure to provide information on mitral valve anatomy, MR severity, LV size and function, as well as cardiac and vascular pressures in MMVD dogs (Chetboul and Tissier 2012). However, owing to volume overload and complex haemodynamic changes associated with disease progression, the detection of myocardial dysfunction in the setting of chronic MR still remains challenging (Serres, Chetboul et al. 2008, Bonagura and Schober 2009).

One of the methods commonly used to assess MR severity in dogs with MMVD consists of calculating the maximal ratio of the regurgitant jet area (ARJ) to LA area (ARJ/LAA) using colour flow Doppler mode (Muzzi, de Araujo et al. 2003, Gouni, Serres et al. 2007). MR is usually considered as mild if the ARJ/LAA ratio is <20- 30%, moderate if it is $30% but "70%, or severe if it is >70%. The major advantage of this colour Doppler mapping method is the rapidity and ease of data acquisition, and its good repeatability and reproducibility for a trained observer (Muzzi, de Araujo et al.

2003, Zoghbi, Enriquez-Sarano et al. 2003). However, this method is only semiquantitative and affected by gain and pulse repetitive frequency settings of the ultrasound machine (Chetboul and Tissier 2012). Eccentric jets tend to cause underestimation and central jets to overestimation of the regurgitant jet (O'Gara, Sugeng et al. 2008).

The PISA (Proximal Isovelocity Surface Area) method, also called the flow convergence method, is routinely used in human medicine to quantify MR (Zoghbi, Enriquez-Sarano et al. 2003, O'Gara, Sugeng et al. 2008). This technique has been shown to be repeatable and reproducible in the awake-dog for a trained observer (Gouni, Serres et al. 2007). However, there are several technical pitfalls with the PISA method, such as a non-circular regurgitant orifice or the presence of multiple regurgitant jets (Chetboul and Tissier 2012). A high heart rate combined with lateral

(29)

movement and twisting of the heart, as well as respiratory movements, make PISA estimations challenging and time consuming in the clinical setting in dogs.

Echocardiography provides an easy and fast method to assess remodelling of the heart. Left atrial (LA) enlargement, as assessed by the LA to aortic root ratio (LA/Ao) (Figure 5), has been proven to correlate to disease severity (Häggström, Hansson et al. 1997,

Hansson, Häggström et al. 2002, Chetboul, Serres et al. 2009, Borgarelli, Crosara et al.

2012, Hezzell, Boswood, Moonarmart et al. 2012). A common way of assessing LA/Ao is in a short axis view in early ventricular diastole (Hansson, Häggström et al.

2002). However, inter-observer variabilities of approximately 10-20% has been reported (Chetboul, Athanassiadis et al. 2004). Therefore assessments between different observers should be interpreted with caution. In humans the preferred method to assess the size of both LA and LV is the volumetric method in a 2D long axis view (Lang, Bierig et al. 2005). In dogs M-mode is still often preferred in the clinical setting due to the ease and speed.

Ejection fraction (EF) and fractional shortening (FS) are commonly used to assess myocardial function. The EF is calculated as left ventricular end-diastolic volume (EDV)– LV end-systolic volume (ESV)/ EDV (Feigenbaum 1993). A simple approximation used commonly in echocardiography is the Teichholz formula (Teichholz, Kreulen et al. 1976) by use of measurements derived from M-mode:

[3] !"# !! !!!!!""!!!!!!!!!! !!""!,

Teichholz formula does not correct for spherical remodelling in the late phase of progressive MMVD. Therefore LV volumes tend to be overestimated (Serres, Chetboul et al. 2008, Tidholm, Bodegard-Westling et al. 2011) and planimetric methods should be preferred (Lang, Bierig et al. 2006, O'Gara, Sugeng et al. 2008, Bonagura and Schober 2009, Chetboul and Tissier 2012).

The ESV indexed to body surface area (ESVI) is increased in systolic dysfunction (Chetboul and Tissier 2012) and is one of the criteria used in the diagnosis of decreased myocardial function (Dukes-McEwan, Borgarelli et al. 2003, Borgarelli, Tarducci et al. 2007). Unfortunately, echocardiographic indices of the LV are more or less load dependent. Especially EF (and FS) measurements are suspect in patients with MR, because both afterload and preload are altered (Wisenbaugh, Spann et al. 1984).

A substantial LV dysfunction may exist even though EF is normal (Enriquez-Sarano, Akins et al. 2009). In humans the lower limit for normal EF is usually set to 0.55

Figure 5. A transverse echocardographic still frame of the left atrium (LA) and aorta (Ao) showing an extremely high left atrium to aortic root ratio (LA/Ao)

Neurohumoral response and pathophysiologic changes during progression of mitral regurgitation in the Cavalier King Charles spaniel

!"#$%&#'%$()*$"+,%-+"*(-.*,(/&%,&0+1%)%213*3&(-2"+*.#$1-2*

,$%2$"++1%-*%4*'1/$()*$"2#$21/(/1%-*1-*/&"*5(6()1"$*71-2*5&($)"+*

+,(-1")*

!

!

8-."$+*9$1:++%-;<()%=>$61*

!

!

"#"$%&'#!$'((%)*"*'+,!

''T To o m myy ffaam miillyy''

Table of Contents

*

?@*ABCD*EF*8GGH9IB8DBE!C*

J@*ABCD*EF*EHBKB!8A*<LGAB58DBE!C*

M@*8GCDH85D*

N@*H9IB9O*EF*DP9*ABD9H8DLH9*

N@?*!(/#$()*&1+/%$0*(-.*",1."'1%)%20*%4*."2"-"$(/16"*'1/$()*6()6"*.1+"(+"*

N@J*P("'%.0-('13*3&(-2"+*(-.*3)1-13()*+12-+*%4*'1/$()*$"2#$21/(/1%-*

N@M*<(/&%2"-"+1+*%4*."2"-"$(/16"*'1/$()*6()6"*.1+"(+"*

N@N*<$"6()"-3"*%4*'1/$()*6()6"*.1+"(+"*

N@T*G$"".*(-.*+"U*.1+,%+1/1%-*

N@V*P"$1/(S1)1/0*

N@W*D1'"*3%#$+"*1-*/&"*,$%2$"++1%-*%4*'1/$()*6()6"*.1+"(+"*