Proef ingediend met het oog op het behalen

van de graad van Master in de Geneeskunde

COMPARISON OF THE EFFECT BETWEEN THE

VITAMIN K ANTAGONIST WARFARIN AND THE

NEW ORAL ANTICOAGULANTS ON AORTIC VALVE

CALCIFICATION AND ATHEROSCLEROTIC

PLAQUE PROGRESSION IN APOE-DEFICIENT

MICE

JOLIEN GEERS

2017 - 2018

Promotor: Prof. dr. B. Cosyns

Co-promotor: Isabel Remory

Geneeskunde en Farmacie

Table of contents

Dankwoord ............................................................................................................ i

Abbreviations .........................................................................................................ii

Abstract ............................................................................................................... iv

I. Introduction .................................................................................................... 1

1. Aortic valve stenosis: background .................................................................. 1

2. Anatomy of the aortic valve ........................................................................... 2

3. Pathophysiology of aortic valve stenosis .......................................................... 3

3.1 Initiation phase ...................................................................................... 5

3.2 Propagation phase .................................................................................. 5

4. The role of calcification in aortic valve stenosis ................................................ 6

4.1 Calcification as an independent predictor of outcome .................................. 6

4.2 Quantification of calcium amount and evaluation of aortic stenosis ............... 6

4.3 Factors that influence calcification – the role of vitamin K antagonists and new

oral anticoagulants ........................................................................................... 7

5. Aortic valve stenosis: consequences and side-effects ........................................ 9

II. Purpose of the study ....................................................................................11

III. Materials and Methods .................................................................................12

1. Study design ..............................................................................................12

1.1 Experiment 1 ........................................................................................12

1.2 Experiment 2 ........................................................................................15

2. Animal handling ..........................................................................................18

3. Drug preparation and administration ..............................................................19

4. Echocardiography ........................................................................................19

4.1 Imaging ...............................................................................................19

4.2 Calibrated integrated backscatter analysis ................................................20

5. SPECT/CT VCAM-1 imaging ...........................................................................21

5.1 Radionuclide labelling and injection .........................................................21

5.2 Imaging ...............................................................................................21

6. In vivo CT imaging ......................................................................................22

7. Semiquantitative plaque scoring ....................................................................22

8. Ex vivo VCAM-1 biodistribution......................................................................22

9. Ex vivo CT calcium quantification ..................................................................23

10. Tissue preservation and Von Kossa staining .................................................23

11. Statistical analysis ....................................................................................24

IV. Results .......................................................................................................25

1. Experiment 1 ..............................................................................................25

1.1 Calibrated integrated backscatter analysis ................................................25

1.2 SPECT/CT VCAM-1 imaging ....................................................................27

1.3 In vivo CT imaging ................................................................................28

1.4 Semiquantitative plaque scoring ..............................................................29

1.5 Ex vivo VCAM-1 biodistribution ...............................................................29

1.6 Ex vivo CT calcium quantification ............................................................31

1.7 Von Kossa staining ................................................................................31

2. Experiment 2 ..............................................................................................33

2.1 Body weight .........................................................................................33

2.2 Calibrated integrated backscatter analysis ................................................33

2.3 In vivo CT imaging ................................................................................35

2.4 Semiquantitative plaque scoring ..............................................................35

2.5 Ex vivo CT calcium quantification ............................................................36

2.6 Von Kossa staining ................................................................................36

V. Discussion ..................................................................................................38

VI. Conclusions ................................................................................................44

References ..........................................................................................................45

Addendum ...........................................................................................................54

i

Dankwoord

Vooraleerst wil ik mijn promotor Prof. dr. B. Cosyns bedanken voor zijn zeer goede

begeleiding bij het schrijven van deze thesis. Als diensthoofd van de afdeling Cardiologie

en als wetenschapper vol ideeën, was hij zeer geplaatst om niet enkel mijn kennis binnen

de cardiologie te verruimen, maar eveneens om mij te introduceren in het

wetenschappelijk onderzoek. Bovendien was hij altijd zeer toegankelijk en stond hij steeds

klaar om mijn vragen te beantwoorden.

Hartelijk dank ook aan Isabel Remory, die mij zeer goed begeleid heeft tijdens mijn 4

maanden in het ICMI labo. Zij heeft mij de praktische handelingen van het preklinisch

onderzoek aangeleerd, zorgde voor een zeer toffe tijd in het labo en heeft mij bovendien

het wetenschappelijk werk leren appreciëren. Daarnaast was zij een belangrijke schakel in

het tot stand komen van deze thesis en in het bijzonder dank ik haar voor het feit dat ik

steeds bij haar terecht kon.

Verder wil ik al de mensen in het ICMI labo bedanken voor de aangename samenwerking.

Ik heb mij steeds welkom gevoeld en heb met plezier deelgenomen aan de verschillende

teambuildings.

Graag wil ik mijn ouders bedanken voor hun steun en in het bijzonder mijn papa voor het

herhaaldelijk nalezen van deze thesis. Tot slot een groot woord van dank aan mijn vriend

Baptist Declerck voor het telkens opnieuw nalezen van mijn thesis, voor zijn kritische blik

en vooral voor zijn enorme steun.

ii

Abbreviations

AF Atrial fibrillation

ANGII Angiotensin II

ApoE -/- mice Apolipoprotein E deficient mice

AS Aortic valve stenosis

AV Aortic valve

AVR Aortic valve replacement

BP Blood pool

BSA Body surface area

BMP-2 Bone morphogenic protein 2

CAD Coronary artery disease

CAVD Calcific aortic valve disease

CT Computed tomography

CVD Cardiovascular disease

dB Decibel

DOACs Direct oral anticoagulants

ECG Electrocardiogram

FOV Field-of-view

GAGs Glycosaminoglycans

Gla ϒ-carboxyglutamic acid

HF Heart failure

IB Integrated backscatter

LDL Low-density lipoproteins

LV Left ventricle

LVOT Left ventricular outflow tract

MGP Matrix Gla protein

MSCT Multislice spiral computed tomography

NOACs New oral anticoagulants

OPG Osteoprotegerin

PARs Protease-activated receptors

PLAx Parasternal long axis

PSAx Parasternal short axis

PV Pixel value

RANK Receptor activator of nuclear factor kappa B

iii

RANKL Receptor activator of nuclear factor kappa B ligand

RAS Renin-angiotensin system

ROI Region of interest

S.c. Subcutaneous

SdAb Single domain antibodies

SEC Size-exclusion chromatography

SPECT Single photon emission computed tomography

TLC Thin layer chromatography

TTE Transthoracic echocardiography

VCAM-1 Vascular cell adhesion molecule-1

VECM Valvular extracellular matrix

VECs Valvular endothelial cells

VICs Valvular interstitial cells

VKA Vitamin K antagonists

iv

Abstract

Background: Aortic valve stenosis (AS) is the most common valvular heart disease and

carries an 80% 5-year risk of progression to heart failure, valve replacement or death.

Because atrium fibrillation (AF) and severe calcified AS are frequently coexisting diseases,

vitamin K antagonists (VKA) are often prescribed drugs to control blood coagulation in this

population. However, VKA can have the undesired side-effect of accelerating aortic valve

(AV) and vascular calcification. On the other hand, new oral anticoagulants (NOACs) have

no pro-calcifying properties and are a preferred alternative to the VKA. Moreover, NOACs

can have anti-inflammatory and protective activities in atherosclerosis. The question rises

if they could also interfere with AV calcification. The principal aim of this pre-clinical trial

was to demonstrate a difference in the effect between the VKA warfarin and the NOACs on

AV calcification. Their effect on atherosclerotic plaque progression, namely vascular

calcification and inflammation status, was also determined.

Methods: Two different experiments were done to investigate the effect of warfarin

treatment at different time points. During experiment 1, warfarin was subcutaneously

injected from 8 till 10 weeks and in another group from 22 till 24 weeks. In experiment 2,

treatment was given from 7 till 14 weeks. Apolipoprotein E-deficient (apoE -/-) mice on a

high-fat diet and C57BL/6J control mice were used. Calcifications of the AV were analysed

using calibrated integrated backscatter (cIB) analysis. Vascular calcifications were

determined by CT imaging, Amide analysis and by Von Kossa calcium stain. SPECT imaging,

using 99mTc-radiolabeled single domain antibodies (sdAb) against vascular cell adhesion

molecule-1 (VCAM-1) was used to look at the inflammatory stage of the plaques.

Results: A significant VCAM-1 uptake was seen in the aorta’s of the apoE -/- mice,

compared to the C57BL/6J mice. No significant difference between the different apoE -/-

groups could be detected. In both experiments, no differences in AV calcifications between

groups could be detected. During the first experiment, calcification of the aortic wall was

significantly increased in all apoE -/- mice, compared to C57BL/6J mice, but no differences

between the different apoE -/- treatment groups could be detected. During experiment 2,

Von Kossa showed significantly more calcium deposition in the proximal aorta of warfarin

treated apoE -/- mice.

Conclusions: No accelerating effect of warfarin on AV calcification could be demonstrated.

Inflammation was significantly increased in the aorta’s of all apoE -/- mice and warfarin

did not change this level of inflammation. During experiment 2, significantly more vascular

calcifications in the proximal aorta of warfarin treated apoE -/- mice were detected.

Keywords: aortic valve stenosis · vitamin K antagonists · new oral anticoagulants ·

calcification · apolipoprotein E-deficient mice

1

I. Introduction

1. Aortic valve stenosis: background

The normal aortic valve (AV) is comprised of three thin leaflets with an AV opening of 3 to

5 cm2 [1]. In aortic valve stenosis (AS), the valve opening narrows, leading to left ventricle

(LV) outflow obstruction and compromising of the cardiac function (see figure 1 below) [2].

Fig. 1 AVS. Left: localisation of the AV in the heart. Right: difference in

opening between a normal valve and a severely calcified and stenotic AV [3].

In the Western world, AS is the most common form of valvular heart disease and the third

most common cardiovascular disease (CVD) after hypertension and coronary artery

disease [4, 5]. The prevalence of AS is only about 0.2% among adults younger than 59

years, but is almost 10% in those 80 years or older, with an overall prevalence of 2.8% in

adults over 75 years of age [6]. The population at risk rises in proportion to the

improvement in life expectancy and rapidly aging society [7].

The most frequent aetiology of AS is currently the degenerative origin, referred to as the

disease spectrum calcific aortic valve disease (CAVD, 82%), followed by rheumatic (11%)

and congenital disease (aortic bicuspid valve, 5%). In developed countries, AS caused by

rheumatic heart disease is nowadays infrequent, but it is still the most common cause of

AS in developing countries [4, 7].

2

The disease spectrum of CAVD is characterised by a slowly progressive disease continuum,

with AS as the end-stage. The disease progression starts with a long unremarkable period

and once symptoms (exertional shortness of breath, angina, dizziness or syncope) develop,

there is a poor prognosis, with survival rates of only 15 – 50% at 5 years. In this stage of

severe AS, the only effective and definitive therapy is surgical or transcatheter aortic valve

replacement (AVR), to which not all patients are suited [8]. Pharmacological interventions

have thus far failed to alter the course of the disease [9]. Severe symptomatic AS in elderly

individuals presents an increasing medical and health economic dilemma. There is thus an

urgent priority to understand the pathophysiological processes leading to AS, to improve

preventive and therapeutic strategies [5].

2. Anatomy of the aortic valve

The AV is composed of three semilunar cusps or leaflets that passively open and close to

maintain unidirectional blood flow from the LV to the systemic and coronary circulation [2].

Each leaflet is organized into three layers, which is shown in figure 2: the fibrosa, the

spongiosa and the ventricularis [4].

Fig. 2 The cellular architecture of the AV. Each leaflet is organised into three

layers: the fibrosa (a collagenous layer), the spongiosa (loose connective

tissue of glycosaminoglycans (GAGs)) and the ventricularis (rich in elastin).

The valvular endothelial cells (VECs) and the valvular interstitial cells (VICs)

are the cellular components of these layers [4].

The fibrosa is located on the aortic side and is a dense collagenous layer, which provides

strength. The ventricularis is located on the ventricular side, below the inflow surface, and

is a layer rich in elastin. The spongiosa is the central layer and contains loose connective

3

tissue made up of glycosaminoglycans (GAGs). This layer permits the necessary

rearrangement of the collagen and elastin layers throughout the cardiac cycle [4, 10].

The cellular components of the AV contain a specialized set of cells, namely valvular

endothelial cells (VECs) and valvular interstitial cells (VICs). VECs are located at valvular

blood-contacting surfaces, constituting a barrier that regulates valve permeability, the

adhesion of inflammatory cells and paracrine signalling. The VICs are the major cell type,

present in all the layers of the AV. They are key in valve remodelling, regulating both the

synthesis and degradation of extracellular matrix components. They physiologically exist

in a quiescent state, with similar characteristics to fibroblasts. In response to stimulation

by molecular and mechanical triggers (including high blood pressure, altered shear stress,

cytokines and growth factors), the VECs and VICs contribute to AS pathophysiology,

altering the locals valve environment and making it calcification prone [4, 10, 11].

3. Pathophysiology of aortic valve stenosis

In the past, the disease progression of CAVD leading to AS was thought to be a mechanical

and passive condition of aging whereby “wear and tear” leads to the calcium deposition

within the valve. In de past decade, our comprehension has changed dramatically,

emphasizing the concept of an active disease process involving cellular and molecular

pathways [4, 7].

The pathophysiology of CAVD can be divided in two distinct phases: the initiation and

propagation phase, each dominated by different mechanisms. The initiating phase show

similarities with atherosclerosis, both ignited by endothelial activation/damage and an

inflammatory response and sharing common risk factors including older age, male sex,

body mass index, elevated serum lipoprotein(a) and low-density lipoproteins (LDL),

hypertension and smoking [10, 11]. However, they have a different mechanism of

evolution at tissue levels. Calcification pathways appear to predominate early in valve

disease, as opposed to vascular disease where calcification develops much later. Moreover,

only 40% of patients with CAVD have significant coronary artery disease (CAD), and most

patients with CAD do not have AS. This suggests a difference in the pathophysiological

mechanism between CAVD and atherosclerosis [9, 12].

The pathophysiology of CAVD is illustrated in figure 3 and explained below.

4

Fig. 3 Pathophysiology of CAVD. Progressive CAVD ranges from non-stenotic to sever stenosis, resulting in

valvular dysfunction, altered haemodynamic and stress (left-right, upper panel). The middle panel illustrates

the cellular involvement. Endothelial damage triggers lipid infiltration and subsequent oxidation, initiating an

inflammatory response involving macrophages, T-lymphocytes and mast cells. VICs are triggered by

inflammation to switch from phenotype, resulting in increased extracellular vesicle release, providing a nidus

for calcification. This microcalcification provokes an inflammatory response, resulting in increased apoptosis

and delayed phagocytosis thereby expanding calcium deposition. Upon propagation, pro-fibrotic and pro-

calcific mechanisms dominate. The pro-fibrotic changes, leading to collagen deposition and progressive

calcification, are mediated by reduced nitric oxide and up-regulation of renin-angiotensin system. Calcification

is the dominant process driving disease progression. The changing of the VICs to an osteoblast phenotype is

thought to play a role in the progression phase by multiple regulatory pathways including Notch, receptor

activator of nuclear factor kappa B (RANK)/receptor activator of nuclear factor kappa B ligand

(RANKL)/osteoprotegerin (OPG), Wnt/b-catenin and bone morphogenic protein-2 (BMP-2). Matrix Gla-protein

is an inhibiting factor of BMP-2 [11].

5

3.1 Initiation phase

The initiation phase is triggered by mechanical stress or decreased shear stress in the

valve, causing endothelial damage and activation. Mechanical stress is highest on the aortic

side of the leaflet, near the attachment to the aortic root and can be increased in the

context of arterial hypertension. Shear stress is highest in the cusps adjacent to the

coronary ostia because of the influence of coronary artery flow. This explains why the

noncoronary cusp is often the first cusp affected [2, 13].

The endothelial injury allows lipids, mainly LDL, to penetrate the valvular endothelium and

accumulate in the subendothelial space and in the deep layer of the fibrosa. These lipids

undergo oxidation and become highly cytotoxic and together with the endothelial damage,

they are capable to trigger inflammation within the valve. This inflammatory response

involves macrophages, T-lymphocytes and mast cells. [2, 4].

Within affected regions, microcalcifications accumulate together with lipids. Formation of

microcalcifications is mediated by the release of apoptotic bodies and extracellular vesicles,

comparable to vesicle-induced calcification in bone and the vasculature. These calcification-

prone extracellular vesicles function as nucleating sites for calcium crystal deposition and

facilitate formation of hydroxyapatite, which in turn leads to progression of the disease

[11].

3.2 Propagation phase

Compared to the initiation phase, the role of inflammation and lipid deposition is less

prominent in the propagation phase. Instead, it is characterized by fibrosis and accelerated

calcification, leading to valvular dysfunction and changes in mechanical stress and flow.

Pro-fibrotic processes are mediated by reduced nitric oxide following endothelial injury and

up-regulation of renin-angiotensin system (RAS) and formation of angiotensin II (ANGII).

A subpopulation of VICs become activated by the inflammatory activity within the valve

and differentiate into an osteoblast-like phenotype. This process is thought to be the

fundamental step in accelerating valvular calcification. Once calcification becomes

abundant, pro-osteogenic mechanisms become overwhelming, ultimately leading to severe

calcification and valvular dysfunction. The calcific regulatory pathways include Notch,

receptor activator of nuclear factor kappa B (RANK)/receptor activator of nuclear factor

kappa B ligand (RANKL)/osteoprotegerin (OPG), Wnt/b-catenin and bone morphogenic

protein-2 (BMP-2). BMP-2 is an important osteogenic differentiation factor, because it is a

key protein in phenotypic switching of VICs and hence in the development of AV

calcification. Physiologically, BMP is inhibited by matrix Gla protein (MGP). MGP exerts its

effect also through another mechanisms, like direct inhibition of the growth of

hydroxyapatite crystals [2, 11, 13, 14].

6

4. The role of calcification in aortic valve stenosis

4.1 Calcification as an independent predictor of outcome

Calcification of the AV is found to be a strong independent predictor of outcome in AS and

is associated with an increased disease progression. Therefore, severe AV calcification

correlates with an adverse clinical outcome, including an increased risk of death and the

need for AVR [15–17]. The accurate quantification of AV calcification and identifying the

factors that influence this calcification burden is therefore mandatory.

4.2 Quantification of calcium amount and evaluation of aortic stenosis

The key diagnostic tool to evaluate AS is transthoracic echocardiography (TTE). It confirms

the presence of aortic stenosis, assesses the degree of valve calcification, LV function and

wall thickness, detects the presence of other associated valve disease or aortic pathology,

provides prognostic information and can be used for follow up of progression of the disease.

The echocardiographic criteria for the definition of severe AS are described in table 1 [18].

Table 1 The echocardiographic criteria for the definition of severe AS [18, 19]. BSA = body surface area.

Valve area (cm2) <1.0

Indexed valve area (cm2/m2 BSA) <0.6

Mean gradient (mmHg) >40

Maximum jet velocity (m/s) >4.0

However, to assess the calcification burden with echocardiography, this is performed using

a subjective and visual semi-quantitative scoring system, lacking objective quantification

of the amount of calcium [20]. This is thus not suited for accurate follow up of the

calcification burden in an individual patient. Currently, Computed tomography (CT) is

considered as the reference technique to quantify and follow up AV calcifications in patients

with high cardiovascular risk. However, cardiac CT is a technique which uses irradiation

and is thus less suitable for repetitive evaluation [21, 22].

Calibrated integrated backscatter (cIB) of echocardiography has recently been validated

as a valuable, non-invasive and non-ionizing tool for the detection and quantification of

aortic valve calcifications in an in vivo rat model. cIB describes the efficiency at which

ultrasound is backscattered from a region of interest (ROI) at a certain frequency,

calibrated with a reference object to compensate for the attenuation of the backscatter

signal due to intervening structures between the skin surface and the tissue of interest

[23–26]. This method for the quantitative detection of calcification will also be used in this

study.

7

4.3 Factors that influence calcification – the role of vitamin K antagonists

and new oral anticoagulants

The cardiovascular risk factors associated with the presence and progression of CAVD and

thus with an increase of AV calcification include: older age, male sex, body mass index,

serum lipoprotein(a) and low-density lipoproteins (LDL), hypertension and smoking [10,

11]. The presence of metabolic syndrome, but also diabetes mellitus are associated with

an increased and graded likelihood of AV calcification [28]. Patients with end-stage renal

disease and secondary hyper-parathyroidism (sHPT) are known to have increased AV

calcification and stenosis compared to the general population [29].

Although vitamin K antagonists (VKA) are the most widely used anti-thrombotic drugs,

chronic treatment has been associated with undesired AV and vascular calcifications. After

more than 30 years of research, the new oral anticoagulants (NOACs) have been

introduced and may provide an alternative for numerous well documented side-effects

associated with the use of VKA [30, 31]. This thesis focuses on the effects of VKA and the

NOACs on AV calcifications.

4.3.1 Vitamin K antagonists

The VKA are widely used for the treatment and prophylaxis of thromboembolic diseases.

VKA block the vitamin K epoxide reductase complex that drives conversion of certain

glutamate residues of vitamin K-dependent coagulation factors into ϒ-carboxyglutamic acid

(Gla)-residues. The vitamin K-dependent coagulation factors include factor II, VII, IX and

X. However, VKA therapy may have undesired side-effects, because a number of proteins

outside the coagulation system also require ϒ-glutamylcarboxylation to become biologically

active, like MGP [32].

MGP is a 10-kDa protein that contains 5 residues of the vitamin K-dependent calcium

binding amino acid, namely Gla. The protein is secreted by vascular smooth muscle cells

and macrophages in the artery. MGP is a vitamin K-dependent protein not related to blood

coagulation, but also affected by VKA [30, 33]. Animal models showed that MGP is a strong

inhibitor of calcification of arterial vessel wall and cartilage. Transgenic MGP-deficient mice

died within 6 to 8 weeks after birth due to massive calcification and rupture of the arteries

[34]. The inhibitory mechanism of MGP involves inhibition of BMP-2, suppression of

osteochondrogenic transdifferentiation of vascular smooth muscle cells and direct inhibition

of hydroxyapatite crystal growth [32].

To become active and to perform his function, MGP requires vitamin K-dependent ϒ-

carboxylation. Concordantly, VKA treatment is associated with upregulation of

uncarboxylated MGP and arterial and valvular calcification [32]. Already in 1998, Price et

8

al., noticed that feeding rats with a diet containing warfarin and vitamin K11 induced focal

calcification in the media of major arteries and in aortic heart valves [33]. This process

was accelerated by vitamin D and reversed by high intake of vitamin K after discontinuation

of warfarin [35, 36]. Warfarin initiated calcium accumulation in the arteries leads to

increased accumulation of MGP in these areas. This can be explained by assuming a

feedback mechanism by which increased local calcium stimulates MGP expression in an

attempt to prevent calcification. It has been shown in cell cultures and vascular tissue that

warfarin up-regulates the mRNA expression of MGP. However, this increase in MGP is not

able to prevent calcium, because during warfarin treatment, most of the MGP is synthesized

as uncarboxylated, inactive species [33, 35]. Administration of warfarin in an

apolipoprotein E deficient (apoE -/-) mouse2 model showed no change in plaque expansion,

neither number nor size distribution of plaques were affected. On the other hand, calcium

levels were already elevated after one week of warfarin treatment and increased further

after 4 weeks of therapy. Von Kossa analysis of serial sections showed that warfarin

accelerated both medial and intimal calcification of atherosclerotic plaque [32].

The detrimental effect of warfarin was also identified in humans, where patients using VKA

demonstrated more vascular and valvular calcification [11]. Histopathologic examination

of excised AV in patients who underwent AVR for AS revealed that the amount of

calcification was significantly (more than 2-fold) greater in patients receiving preoperative

VKA treatment than in patients without VKA treatment [37]. These results were confirmed

one year later in a study from Koos et al [38]. Patients with CAVD underwent multislice

spiral computed tomography (MSCT) imaging to quantitate the amount of calcification in

the AV and coronary arteries. Long VKA treatment was associated with a significant

increase in both coronary artery and valvular calcium scores [38].

4.3.2 New oral anticoagulants

The NOACs are also called direct oral anticoagulants (DOACs) because of their direct

inhibiting action on either thrombin (FIIa) (dabigatran) or factor Xa (apixaban, edoxaban

and rivaroxaban). They do not interfere with the vitamin K cycle and are thus a preferred

alternative to the VKA. NOACs have an improved efficacy/safety ratio, a predictable

anticoagulant effect without need for routine coagulation monitoring, and fewer food and

drug interactions compared with VKA [39]. Moreover, in contrast to the VKA, there are no

pro-calcifying properties ascribed to these NOACs, these could even have an anti-

atherogenic or atheroprotective effect [40–43].

1 Co-administration of vitamin K1 is able to counteract the effect of warfarin on the carboxylation of blood coagulation factors by the liver, but not in bone and arterial vessel wall. Counteracting the effect of warfarin at the level of the liver is necessary to keep the prothrombin times and hematocrits normal and to prevent bleeding of the animals [33]. 2 ApoE -/- mouse is a widely used mouse model for atherosclerosis research because it develops atherosclerotic

plaques resembling those of human [43].

9

Accumulating evidence suggests that thrombin and factor Xa are important modulators in

atherosclerotic disease in ways that do not directly involve thrombus formation, such as

signalling through protease-activated receptors (PARs). Activation of PARs induce

promotion of inflammation and leukocyte migration, which results in the initiation of

atherosclerosis [40, 41]. Preclinical studies have provided evidence for the effects of direct

Xa or thrombin inhibition beyond anticoagulation, including anti-inflammatory and

protective activities in atherosclerotic plaque development [40]. Oral administration of the

direct thrombin inhibitor dabigatran leads to a reduced atherosclerotic lesion size with

enhanced plaque stability and improves endothelial function by reducing oxidative stress

in apoE -/- mice [42, 44–46]. Even in an apoE -/- mouse model with hypercoagulable

phenotype (TMpro/pro), dabigatran could counteract the pro-inflammatory and pro-

atherogenic state [47]. Similar effects are shown for factor Xa inhibitors. Administration of

rivaroxaban in apoE -/- mice reduced the development of atherosclerosis and stabilized

atherosclerotic plaques by inhibiting lipid deposition, collagen loss and macrophage

accumulation [43].

The NOACs are very recent drugs and their long term effects in humans are still unknown.

Preclinical evidence shows that NOACs are effective in reducing inflammatory cytokines

and atherosclerotic cascades; however, clinical data are lacking. Because atherosclerosis

and CAVD share principal risk factors and histological features, the question rises if the

NOACs could also interfere with the degenerative process leading to calcium deposition in

AS.

5. Aortic valve stenosis: consequences and side-effects

AS is the second most common indication for cardiac surgery and carries an 80% 5-year

risk of progression to heart failure, valve replacement or death [14]. It is a common

condition associated with major morbidity and mortality, due to progressive valve

narrowing that increases the pressure afterload on the left ventricle. This increased

ventricular wall stress stimulates concentric remodelling and hypertrophy of the LV

myocardium, a compensatory mechanism that initially restores wall stress and preserves

LV function. However, in the long term, if the systolic load is not relieved, this

compensatory mechanism fails, hypertrophy switches towards eccentric remodelling, filling

pressures rises and heart failure develops. There is thus evidence that increasing levels of

hypertrophy may in fact be maladaptive [2, 48, 49].

Furthermore, the LV hypertrophy can result in diastolic dysfunction and increased left atrial

pressures, which may precipitate atrial fibrillation (AF). AF consists of rapid, disorganized

and asynchronous contraction of atrial muscle with consequent deterioration of atrial

mechanical function. It is characterized by the absence of atrial complexes (p-waves) on

the electrocardiogram (ECG) and a fast and irregular ventricular rate. AF is the most

10

common arrhythmia in the general population and is increasingly frequent as population

ages: 0.02% among those 18 – 39 years of age, 11.6% among those >75 years of age.

Some of the risk factors for AF are male gender, obesity, hypertension, diabetes, coronary

artery disease, valvular heart disease and cardiomyopathy [50, 51].

Because of the high prevalence of AF in the elderly population as well as similar risk factors

for AF en AS, both conditions frequently coexist. The prevalence of AF ranges from 9% in

patients with moderate AS to 27% in patients with severe AS. In some patients with severe

AS, AF is the first sign of clinical decompensation and may be a marker of chronic pressure

or volume overload [52, 53]. The prevalence of AF is the highest in patients with severe

calcified AS. A reason for the high prevalence of severe valve calcifications in this

population might be the preventive anticoagulant treatment with a VKA of these patients

with concomitant AF.

11

II. Purpose of the study

The disease AS is the most common form of valvular heart disease and carries an 80% 5-

year risk of progression to heart failure, valve replacement or death [4, 5, 14]. Because

AF and severe calcified AS are frequently coexisting diseases, VKA are often prescribed

drugs to control blood coagulation in this population. However, VKA therapy can have the

undesired side-effect of accelerating AV and vascular calcification [30, 32, 33]. After more

than 30 years of research, the NOACs have been introduced and because they do not

interfere with the vitamin K cycle, they are a preferred alternative to the VKA. There are

no pro-calcifying properties ascribed to these NOACs, moreover, they can even have anti-

inflammatory and protective activities in atherosclerotic plaque development [40–43].

Because atherosclerosis and AS share principal risk factors and histological features, the

question rises if the NOACs could also interfere with the degenerative process leading to

calcium deposition in AS.

The principal aim of this pre-clinical trial is to demonstrate a difference in the effect of the

most frequently used VKA warfarin and the NOACs on AV calcification. This could offer new

insights in the pathophysiology of the frequent disease AS and allow transposition in the

clinical management of these patients.

Secondly, it is attributed to warfarin to increase vascular calcification burden and on the

other hand, it is shown for the NOACs to decrease inflammation status at the vascular

level. Therefore, we also want to demonstrate the effect of these two drugs on

atherosclerotic plaque progression.

The VKA and the NOACs will be compared in vivo in an apoE -/- mouse model. This common

genetically manipulated model lacks apolipoprotein E, which is a ligand for lipoprotein

receptors and mediates the cellular uptake of several different lipoproteins, like cholesterol,

triglycerides and very low-density lipoproteins (VLDL). Consequently, this model shows

impaired clearing of plasma lipoproteins and develops atherosclerotic lesions, resembling

those observed in humans. These atherosclerotic lesions are exacerbated when mice are

fed a high-cholesterol, high-fat, Western-type diet [43, 54].

The technique of cIB is already validated as a non-invasive and non-ionizing technique for

the quantification and detection of progression and regression of AV calcifications in rats

[23–26]. In this study, this technique will be validated in an in vivo mouse model.

This thesis includes the first part of the study, namely the evaluation of warfarin treatment

in the apoE -/- mouse model.

12

III. Materials and Methods

1. Study design

Two studies were done to investigate the effect of warfarin treatment at different time

points. The two study protocols were approved by the ethical committee for animal studies

of the Vrije Universiteit Brussel (project number: 16-272-1 and 17-272-7 for the first and

second experiment respectively). Guidelines of the National Institute of Health principles

of laboratory animal care (NIH publication 86-23, revised 1985) were followed during the

studies. The study designs of the first and second experiment are explained below.

1.1 Experiment 1

30 female mice of four weeks old were delivered from Charles River (France): 3 C57BL/6J

and 27 apoE -/- mice. All the mice were prospectively divided in four groups (table 2).

Table 2 Overview of the different study groups of experiment 1.

Group Number Mouse type Diet Medication

1 3 C57BL/6J Normal /

2 9 ApoE -/- High-fat diet /

3 9 ApoE -/- High-fat diet Vitamin K + warfarin earlya

4 9 ApoE -/- High-fat diet Vitamin K + warfarin lateb

a Treated from the age of 8 till the age of 10 weeks

b Treated from the age of 22 till the age of 24 weeks

After arrival, apoE -/- mice were put on a high-fat diet and one week of acclimatization

took place. At the age of 6 weeks, all mice underwent a baseline echocardiography. When

the mice were 8 weeks old, treatment with warfarin + vitamin K was started in group 3.

This group of mice was daily injected for 3 weeks. When the mice were 21 weeks old,

echocardiography was performed and inflammation status was determined by SPECT/CT

scan with radionuclide injection. After the in vivo imaging, at the age of 22 weeks,

treatment was started for group 4. This group was also daily injected with warfarin +

vitamin K for 3 weeks. At the age of 24 and 26 weeks, all the mice underwent an

echocardiography again. When the mice were 27 weeks old, a SPECT/CT scan with

radionuclide injection was done for the last time. 170 minutes after tracer injection, the

mice were killed and all the major organs and the aorta with the AV were dissected. During

dissection, the aorta’s were examined for semiquantitative plaque scoring. All organs were

weighted and radioactivity levels in each tissue were calculated. Afterwards, the aorta’s

were embedded in paraffin for ex vivo CT imaging and calcium scoring. At last, slices of 5

μm thick were generated of one paraffin embedded aorta from every group and stained

with Von Kossa calcium stain. A schematic overview of the study design for the different

groups of experiment 1 is shown in figure 4a-4d below.

13

Fig

. 4

a S

tudy d

esig

n g

roup 1

.

Fig

. 4

b S

tudy d

esig

n g

roup 2

.

14

Fig

. 4

c S

tudy d

esig

n g

roup 3

.

Fig

. 4

d S

tudy d

esig

n g

roup 4

.

15

1.2 Experiment 2

15 female wild type C57BL/6J and 40 female apoE -/- mice were delivered from Charles

River (France) at the age of 4 weeks. All the mice were prospectively divided in four groups

(table 3).

Table 3 Overview of the different study groups with the number of mice, the mouse type, their diet and therapy.

Group Number Mouse type Diet Medication

1 15 C57BL/6J Normal /

2 15 ApoE -/- High-fat diet /

3 12 ApoE -/- High-fat diet Vitamin K

4 13 ApoE -/- High-fat diet Vitamin K + warfarin

The first week after arrival, no interventions were done, so the mice could acclimatize. At

week 2, when the mice were 6 weeks, a baseline echocardiography was performed on the

four groups. After the imaging, three mice from both group 1 and 2 were killed with sodium

pentobarbital and the hearts with the aorta’s were dissected and embedded in paraffin. Ex

vivo high resolution CT of these paraffin blocks was performed. From week 3 till the end

of the study, all the apoE -/- mice were put on a high-fat diet. At week 4, treatment with

vitamin K alone (group 3) and vitamin K + warfarin (group 4) was started. The first 5 days

of treatment, injections were given daily. From day 6 till the end of the treatment,

injections were given every two days. At the age of 8, 11 and 15 weeks, echocardiography

and in vivo CT was performed on all animals in each group. At the age of 8 and 11 weeks,

3 mice from each group were killed with sodium pentobarbital and the hearts with the

aorta’s were dissected and embedded in paraffin for ex vivo CT imaging. At the age of 15

weeks, all the animals left were killed and the hearts with the aorta’s were at their turn

dissected. Now, the aorta’s were scored semiquantitatively under a microscope according

to the relative plaque size. Subsequently, the hearts with the aorta’s were embedded in

paraffin and underwent ex vivo CT imaging. At last slices of 5 μm thick were generated

from tissue of one C57BL/6J mouse and from all apoE -/- mice and stained with Von Kossa

calcium stain. A schematic overview of the study design for the different groups of

experiment 2 is shown in figure 5a-5d.

16

Fig

. 5

a S

tudy d

esig

n g

roup 1

.

Fig

. 5

b S

tudy d

esig

n g

roup 2

.

17

Fig

. 5

c S

tudy d

esig

n g

roup 3

.

Fig

. 5

d S

tudy d

esig

n g

roup 4

.

18

2. Animal handling

All mice were housed in stainless steel cages with sawdust bedding. They were kept at a

12h day/night cycle, an average room temperature of 24°C, and a relative humidity of

50%. Food and water were provided at libitum. A standard rodent diet (SAFE, France) was

given to C57BL/6J mice throughout the study. All apoE -/- mice were put on a high-fat

diet. The diet consist a Clinton/Cybulsky High Fat Rodent diet with 1.25% cholesterol

without menadione (D16030101 from Research Diets Inc.). Menadione is a product of

vitamin K catabolism and the likely intermediate in the synthesis of vitamin K2. Schurgers

et al. has demonstrated that high vitamin K2 supplementation is able to inhibit warfarin-

induced arterial calcification [55]. Therefore, a diet without menadione was chosen. The

high-fat diet is used to accelerate the atherosclerotic lesion formation but for the rest, the

diet meets all of the animals nutritional needs.

The animal welfare was guaranteed during the experiment. Physical condition of the

animals was assessed daily. All procedures were performed under inhalation anesthesia

(isoflurane gas). During the in vivo imaging, the body temperature of the animals was

maintained by the use of a heating pad and heating lamp. The contact gel, used during the

echocardiographic examination, was warmed up before applying it to the chest of the

animals.

ApoE -/- mice have a thickened skin and they can show scratch wounds due to scratching.

We tried to prevent these wounds by putting a wooden cube in the cage. By scratching the

cube, the mouse will keep its nails short. If necessary, the nails can be cut when the animal

is a sleep for an investigation. If these scratch wounds developed, they were checked daily

and treated with aluminium wound spray. This spray is an astringent and antibacterial

spray, which promotes wound healing. It acts as a wound plaster, which allows the wound

to breathe and at the same time keeps the wound clean. If the scratch wounds covered

more than 30% of the body surface, a human end-point was reached and the animal was

killed.

The medication was administered by subcutaneous (s.c.) injections. Vitamin K was always

co-injected with warfarin. Vitamin K acts as an antidote at the liver and prevents bleeding.

However, mice were closely looked at until 30 minutes after the injections were given, to

check that no bleeding was occurring. The s.c. injection sites were rotated to keep the skin

healthy. This rotation will prevent scarring and hardening of the fatty tissue. If bruising,

bleeding, blood in urine was yet noticed, a human end-point was reached and the animal

was killed. Other human end-points were rapid weight loss of 15-20% within a few days,

gradual weight loss of 20% over an extended period of time, leading to emaciation, no

production of faeces, no more explorative behaviour. Signs of thrombus formation, like

paralysis, was also a human end-point.

19

Animals were killed with an intravenous injection in the tail vein of 120 mg/kg sodium

pentobarbital. Prior to injection, animals were sedated with isoflurane inhalation

anesthesia.

3. Drug preparation and administration

Animals were s.c. injected with 150 mg/kg warfarin (4-Hydroxy-3-(3-oxo-1-

phenylbutyl)coumarin (A4571), purchased from Sigma-Aldrich BVBA/SPRL, Diegem,

Belgium) and 15 mg/kg konakion (fytomenadion, Roche, purchased from the farmacy).

Konakion is the synthetic form of vitamin K1 and was used as a source of vitamin K antidote

at the liver. Warfarin and konakion were diluted with sodium chloride 0.9% to a

concentration of 15 mg/mL and 1.5 mg/mL respectively. A solution of 100 μL per 10 gram

body weight was injected in the back of the animals.

4. Echocardiography

4.1 Imaging

The mice were anesthetized with 5% isoflurane gas (S.A. Abbott N.V., Ottignies, Belgium)

with 1.5L O2/min as a carrier gas. Subsequently, the mice were put under continuous

isoflurane gas (2.5%) anaesthesia with 1.5L O2/min as carrier gas. The anterior chest wall

was shaved and the mice were placed in left lateral decubitus on a wooden bench in order

to obtain optimal image quality and views. Electrocardiogram electrodes were fixed on the

paws (see figure 6). A Vivid 7 Pro system (GE Medical Systems, Milwaukee, WI, USA) with

a 10 MHz neonatal probe was used.

Fig. 6 Preparation of the mouse before starting an echocardiographic

examination.

20

4.2 Calibrated integrated backscatter analysis

For the quantitative detection of AV calcification, cIB was used. Two-dimensional

transthoracic parasternal short- and long axis (PSAx and PLAx) ultrasound images were

obtained in all mice. Control settings such as gain compensation, focus position and frame

rate (PLAx: 186.1 fps, PSAx: 375 fps) were maintained constant for every measurement

to obtain uniform brightness and to facilitate reproducible sampling of backscatter values.

Integrated backscatter (IB) analysis was performed offline with EchoPac PC software

(version 110.0.0, GE Medical Systems, Milwaukee, Wi, USA). IB values were acquired and

averaged over 3 cardiac cycles in quantitative analysis mode by placing fifteen 0.4 x 0.4

mm circle-shaped sample volumes (regions of interest, ROIs) on the AV in all images.

Global ROI I and II were placed around the whole AV in respectively the PSAx and PLAx.

Two blood pools (BP) were placed: BP1 of the atrial cavity in the PSAx and BP2 of the left

ventricular outflow tract (LVOT) in PLAx. The positions of the ROIs were within the same

region for all animals and during all cardiac cycles (see figure 7).

Fig. 7 Acquisition and calculation of cIB values. Fifteen ROIs were placed on the short

axis of the AV (A) and on the parasternal long axis of the AV (B). Global ROIs I and II

were placed around the whole AV. Two blood pools were placed: BP1 of the atrial cavity

and BP2 of the ventricular outflow tract.

21

To compensate for the attenuation of the backscatter signal due to intervening structures

between the skin surface and the ROIs, cIB was determined by subtracting IB values from

the tissue with the blood pool IB values adjacent to the tissue, as in the following formula:

cIB value (ROI X) = [IB value (ROI X) – IB value (ROI BPX)].

cIB values give a more precise estimation of the increase of echogenicity, independent

from the background signal. Mean cIB values for each structure were calculated as follow:

cIB value (ROIAVSA (aortic valve, short axis)) = mean cIB value (ROI 1 to 10);

cIB value (ROIAVLA (aortic valve, long axis)) = mean cIB value (ROI 11 to 15);

cIB value (ROIAVTOT (aortic valve, total)) = mean cIB value (ROIAVSA, ROIAVLA);

cIB value (ROIAVGLOB (aortic valve, global)) = mean cIB value (ROI I, II);

5. SPECT/CT VCAM-1 imaging

Vascular cell adhesion molecule-1 (VCAM-1) was used to look at the inflammatory stage

of the plaques. VCAM-1 is expressed on endothelial cells and plays an important role in

disease progression by attracting inflammatory cells to the developing lesions. In advanced

lesions, VCAM-1 is also expressed at the level of neovessels and may reflect the ongoing

inflammation within the plaque. These properties make VCAM-1 an attractive target for the

molecular imaging of inflammation in atherosclerosis [56]. The in vivo VCAM-1 imaging

will be performed through SPECT imaging using 99mTc-radiolabeled single-domain

antibodies (sdAb3).

5.1 Radionuclide labelling and injection

First, the VCAM-1 sdAb were radiolabeled with 99mTc following standard procedures and

tracer quality was evaluated via retention on size-exclusion chromatography (SEC)

columns and filters, and instant thin layer chromatography (iTLC) analysis. Five minutes

prior to tracer administration, the mice were anesthetized using inhaled isoflurane gas.

The purified radiotracer was intravenously injected through the tail vein with 200 μL tracer

(10 microgram 3R15b-Tc99m).

5.2 Imaging

Two hours post-injection, animals were scanned on a 75-pinhole stationary detector SPECT

system (U-SPECT-II, Milabs BV, The Netherlands) under continuous inhalation anesthesia.

At the age of 21 weeks, a 5 minutes total body SPECT scan was taken, followed by a 30

minutes focused thorax SPECT scan with field-of-view (FOV) including only the thorax (no

liver). The FOV was extended to include the liver, spleen and kidneys on the second time

3 sdAb are small antigen-binding fragments, derived from camelids heavy-chain-only antibodies and are proven

to be an effective molecular imaging probe [68].

22

point imaging, namely when the mice were 27 weeks old. With this FOV, the scan time

was increased to 45 minutes. The SPECT imaging settings were kept constant: scan mode

was set to spiral with a normal step mode. Subsequently, the functional SPECT scan was

each time followed by a 1:45 minutes CT scan (U-CT, Milabs BV, The Netherlands) for

anatomical information and image reconstruction in order to obtain high precision data

analysis. The CT settings were also kept constant during all scans: normal scan mode and

partial scan angle. The 2D fused SPECT/CT images were analysed using Amide software

(Amide1.0.4 64-bit, copr. Andreas Loening) and in 3D by Osirix software (OsiriX 3.6.1,

64-bit, Pixmeo, Geneva, Switzeland).

6. In vivo CT imaging

During the first experiment, a fast CT scan was performed after the SPECT scan, for

anatomical information and image reconstruction.

During the second experiment, no SPECT scans were taken, but a full-body CT (U-CT,

Milabs BV, The Netherlands) was performed under continuous inhalation anesthesia. The

tube voltage was set to 60 kV with maximum power (615 μA). To lower the ionizing

radiation doses, but to maintain a good image quality, normal scan mode with full scan

angle was chosen. These settings corresponded to a scan time of 4:14 minutes and a total

radiation dose of 464,2 mGy (determined with dose wire). All mice were scanned under

equal conditions.

7. Semiquantitative plaque scoring

After the mice were killed, the hearts and the aorta’s were dissected. The aorta was scored

semiquantitatively under a microscope according to the relative plaque size: (-) no plaque

visible, (+) a small plaque covering less than 50% of the surface of the aorta and (++)

large plaques covering more than 50% [57].

8. Ex vivo VCAM-1 biodistribution

For the first experiment, the mice were killed 170 minutes after tracer injection and all the

major organs were dissected for ex vivo validation of tracer biodistribution. All tissues

(thymus, hearts, aorta, blood, lung, liver, spleen, pancreas, stomach, small intestine,

colon, left and right kidneys, lymph nodes, muscles, bones and tail) were weighted and

radioactivity in each tissue was counted with a gamma-counter. Counts were converted to

radioactivity levels based on a standard of known activity and adjusted for radioactive

decay. Radioactivity levels in each tissue were calculated as percentage of injected activity,

divided by the weight of the tissue (% IA/g). Lesion-to-blood and lesion-to-muscle were

determined from the biodistribution data.

23

9. Ex vivo CT calcium quantification

Ex vivo high-resolution micro-CT (U-CT, Milabs BV, The Netherlands) of the paraffin-fixed

tissues was performed. The tube voltage was set to 60 kV with maximum power (615 μA),

an accurate scan mode with full scan angle and a scan time of 15:03 minutes was chosen.

All samples were scanned under equal conditions.

The ex vivo micro-CT paraffin-fixed tissue images were analysed in the medical image data

analysis tool Amide (Amide1.0.4 64-bit, copr. Andreas Loening). All images were visually

screened for the presence of calcification of the aorta. If a calcium spot was seen in the

tissue, this was indicated as (+), if no calcium was observed, it was indicated as (-). The

plaques were quantified by selecting the tissue above a calcium threshold. This calcium

threshold was determined in each image individually as the mean grayscale of the myocard

+ 2 standard deviations. The maximum and minimum intensity and the size of every plaque

were calculated automatically.

10. Tissue preservation and Von Kossa staining

The tissues were washed 5 times in phosphate buffered saline (PBS) and subsequently

transferred to 70% ethanol. For experiment 1, only the aorta’s with the aortic valves were

embedded, for experiment 2, the hearts with the aorta’s were embedded. The hearts

and/or aorta’s underwent a histokinette overnight program for fixation, dehydration and

infiltration of the histological tissue samples with paraffin wax. The generated paraffin

embedded tissues were stored at room temperature.

Slices of 5 µm thick were stained with Von Kossa calcium stain. First, sections were

deparaffinised and hydrated to distilled water. Slides were incubated in Silver Nitrate

Solution (5%) for 30 – 60 minutes with exposure to ultraviolet light. Then, they were

incubated in Sodium Thiosulfate Solution (5%) for 2 – 3 minutes and subsequently in

Nuclear Fast Red Solution for 5 minutes. Between every step, a washing step was

implemented. At last, slides were dehydrated very quickly in 3 changes of fresh 100%

alcohol and coated with mounting medium and a coverslip. Tissue analysis was performed

with a PC digital image camera (Digital Sight DS-5M, Nikon Corp, Japan) mounted on an

Axiolab Zeiss light microscope (Carl Zeiss Corp, Germany).

24

11. Statistical analysis

Data are expressed as mean ± standard deviation. Comparisons in cIB, VCAM-1

biodistribution and calcium area between groups were performed by using one-way ANOVA

tests. Bonferroni correction for multiple comparisons was applied for the cIB, when

comparing the different ROIs AVSA, AVLA, AVTOT and AVGLOB. For the categorical

variables, identification of calcium spot and semiquantitative plaque scoring, data are

presented in frequency distribution tables (absolute values and percentages) and analysed

by chi-square tests. When comparing the apoE -/- mice only, data are reduced to 2x2

tables and analysed by Fisher exact tests. A value of P < 0.05 was considered significant.

Statistical analysis was performed in Graphpad Prism 6 software and SPSS version 24.

25

IV. Results

1. Experiment 1

1.1 Calibrated integrated backscatter analysis

At baseline, the C57BL/6J mice had a significant lower cIB value for the ROI AVTOT

compared to all the apoE -/- groups (p = 0.0053, see figure 8). No significant difference

was found between the different apoE -/- groups (p = 0.3581). At 21, 24 and 26 weeks,

no significant differences between groups were observed (ROI AVTOT at 21 weeks: p =

0.0350, at 24 weeks: p = 0.2050 and at 26 weeks: p = 0.5959). For each group, no

significant difference in cIB over time was observed (ROI AVTOT over time: p = 0.1569, p

= 0.0891, p = 0.0950 and p = 0.0296 for group 1, 2, 3 and 4 respectively). The mean cIB

values of the different groups at the different time points are described in addendum 1.

26

Fig

. 8

Com

parison o

f cIB

valu

es for

the d

iffe

rent

RO

Is b

etw

een t

he d

iffe

rent gro

ups a

t baseline, 21 w

eeks, 24 w

eeks a

nd 2

6 w

eeks. RO

I =

regio

n o

f in

tere

st,

AVSA =

aort

ic valv

e short

axis

, AVLA =

aort

ic valv

e lo

ng axis

, AVTO

T =

aort

ic valv

e to

tal, AVG

LO

B =

aort

ic valv

e glo

bal, cIB

=

calibra

ted in

tegra

ted

backscatt

er,

dB =

decib

el. T

hre

shold

for

sig

nific

ance p

< 0

.0125 (

corr

ection for

multip

le c

om

parisons).

* p

= 0

.0053.

*

27

1.2 SPECT/CT VCAM-1 imaging

On the scans at 21 and 27 weeks, 99mTc-labeled VCAM-1 uptake was observed in lymphoid

tissues (bone marrow, axillary and brachial lymph nodes and thymus) of the apoE -/- mice.

Lower uptake in bone marrow and in the axillary and brachial lymph nodes and no uptake

in thymus was shown for the C57BL/6J mice. At both time points, uptake in the aorta of

the apoE -/- mice could be detected. There was no significant difference in VCAM-1 uptake

at the level of the aorta between the different apoE -/- groups at each time point (p =

0.1960 and p = 0.7610 for 21 and 27 weeks respectively) and between the different time

points (p = 0.1043, p = 0.1579 and p = 0.1160 for group 2, 3 and 4 respectively). For the

C57BL/6J control mice, no uptake in the aorta could be detected. At 27 weeks, a bigger

field of view was used and the clearance route of the low molecular weight sdAb via kidney

and urine is shown on these images. SPECT/CT VCAM-1 maximum intensity projection

images of one mouse of each group at the two time points are shown in figure 9 and 10.

Fig. 9 SPECT/CT focused thorax maximum intensity projection (MIP) images of 21 weeks

old mice 2 hours after i.v. injection of 99mTc-labeled VCAM-1 sdAb. (A) group 1, (B) group

2, (C) group 3, (D) group 4.

28

Fig. 10 SPECT/CT maximum intensity projection (MIP) images of 27 weeks old mice 2

hours after i.v. injection of 99mTc-labeled VCAM-1 sdAb. At this time point, a bigger field

of view is used, including the liver and kidneys. (A) group 1, (B) group 2, (C) group 3,

(D) group 4. BM = bone marrow, Ao = aorta, Tm = thymus, LN = lymph nodes.

1.3 In vivo CT imaging

For anatomic information and image reconstruction, the functional SPECT scan was each

time followed by a 1:45 minutes in vivo micro-CT scan and the obtained images were

visually analysed in the medical image data analysis tool Amide. At both time points,

namely 21 and 27 weeks, calcium spots were seen in the aorta’s of apoE -/- mice. For

C57BL/6J mice, calcium spots were never seen. Figure 11 represents in vivo micro-CT

images of one animal of each group at 27 weeks. Calcium spots, visible in the aorta, are

designated with red arrows.

Fig. 11 In vivo micro-CT images of one mouse of each group at 27 weeks. (A)

group 1, (B) group 2, (C) group 3, (D) group 4. Red arrows are pointing to

calcium spots, visible in the aorta.

29

1.4 Semiquantitative plaque scoring

During dissection at the end of the study, the aorta’s were scored semiquantitatively under

a microscope according to the relative plaque size. For each group, this score (absolute

values and percentages) is represented in table 4 below. There was a significant difference

in plaque scoring between the apoE -/- groups and the C57BL/6J group (p < 0.0001).

Between the apoE -/- groups mutually, no significant difference was found (p = 0.0550).

Table 4 Semiquantitative plaque scoring of the aorta for the different study groups. Results are shown as

absolute values and percentages. (-) no plaque visible, (+) a small plaque covering < 50% of the surface of

the aorta, (++) large plaques covering > 50%.

Group 1: Group 2: Group 3: Group 4:

C57BL/6J ApoE -/- no

injections

ApoE -/- vit K +

warfarin early

ApoE -/- vit K

+ warfarin late

(-) 3 (100%) 0 0 0

(+) 0 6 (75%) 3 (42.9%) 7 (100%)

(++) 0 2 (25%) 4 (57.1%) 0

1.5 Ex vivo VCAM-1 biodistribution

The 99mTc-labeled VCAM-1 sdAb showed normal kidney retention in both C57BL/6J and

apoE -/- mice (137.09 ± 6.75 %IA/g; 138.00 ± 22.15 %IA/g) and low uptake in non-

VCAM-1 expression organs like muscle and heart (<0.5 %IA/g). VCAM-1 expression was

observed in lymphoid tissues (i.e. bone marrow, lymph nodes, spleen and thymus) of both

C57BL/6J control and apoE -/- mice. Although, this uptake was lower for the C57BL/6J

mice compared to all apoE -/- groups. A significant higher VCAM-1 uptake was found in

the aorta’s of all apoE -/- mice (1.45 ± 0.41 %IA/g), compared to the C57BL/6J mice (0.22

± 0.11 %IA/g) (p = 0.0015). There was no significant difference in uptake in the aorta

between the different apoE -/- groups (p = 0.8633). The 99mTc-labeled VCAM-1 sdAb

exhibited a lesion-to-blood ratio of 0.69 ± 0.22 %IA/g, 1.92 ± 0.45 %IA/g, 1.59 ± 0.30

%IA/g and 1.53 ± 0.34 %IA/g for group 1, 2, 3 and 4 respectively. Lesion-to-blood ratio

is significantly different between group 1 and group 2 (p = 0.0222) and all the apoE -/-

mice have a lesion-to-blood ratio > 1. Small decrease in lesion-to-blood ratio for the

warfarin treated mice (group 3 and 4) is observed compared to the control apoE -/- mice

(group 2) due to higher blood retention in the warfarin treated apoE -/- mice. The lesion-

to-muscle ratio for the VCAM-1 sdAb was 5.36 ± 2.48 %IA/g, 11.48 ± 5.93 %IA/g, 8.29

± 2.42 %IA/g and 11.67 ± 2.54 %IA/g for group 1, 2, 3 and 4 respectively.

Ex vivo biodistribution of the 99mTc-labeled VCAM-1 sdAb, expressed as %IA/g tissue, for

the different study groups are shown in figure 12. An overview of these values for each

tissue individually are summarized in addendum 2.

30

Fig

. 1

2 E

x v

ivo b

iodis

trib

ution d

ata

of

99mTc-l

abele

d V

CAM

-1 s

dAb 2

h a

fter

inje

ction,

expre

ssed a

s %

IA/g

of dis

secte

d t

issue.

31

1.6 Ex vivo CT calcium quantification

The paraffin-fixed aorta’s underwent a high resolution micro-CT and the obtained images

were analysed in the medical image data analysis tool Amide as described in materials and

methods. The results of these analysis are shown in table 5. There was no calcium spot

visualised in the embedded aorta of the C57BL/6J mouse, while in all the aorta’s of the

apoE -/- mice, calcium spots were detected (p=0.0030). There was no significant difference

in size and maximal intensity of the detected calcium spots between the different apoE -/-

groups (p=0.9754 and p=0.2484 respectively) .

Table 5 Results of the ex vivo micro-CT paraffin-fixed image analysis in the medical image data analysis tool Amide.

For each group, the number of mice where a calcium spot was detected, is described. Of these calcium spots, the

size and maximal intensity and the corresponding p-value is described. Data are presented as means ± standard

deviations.

Group 1: Group 2: Group 3: Group 4: P-value

C57BL/6J ApoE -/- no

injections

ApoE -/- vit K +

warfarin early

ApoE -/- vit K +

warfarin late

Visualisation of

calcium spot (n)

(+) 0 5 4 4 0.0030

(-) 1 0 0 0

Quantification of

calcium spot

Size (mm3) 4.20 ± 0.77 4.41 ± 1.30 4.47 ± 3.20 0.9754

Maximal

intensity (AU)

1892 ± 175.2 1717 ± 147.7 1608 ± 361.3 0.2484

1.7 Von Kossa staining

Slices of 5 µm thick were generated of one paraffin embedded aorta from every group and

stained with Von Kossa calcium stain (see figure 13). In all the apoE -/- groups, plaques

with calcium deposits in the aorta were visualised. No difference between the different

apoE -/- groups could be detected. In the C57BL/6J mouse, no plaques, nor calcium

deposits were detected.

32

Fig. 13 Slices of 5 µm thick of the paraffin embedded aorta’s, stained with Von Kossa calcium stain. Calcium is represented

in brown, nuclei in red and cytoplasm in pink. (A) group 1, (B) group 2, (C) group 3, (D) group 4. AV = aortic valve.

Aortic wall

Myocard Plaques

AV

Plaques

Brachiocephalic

artery Left common

carotid artery

A

B

C D

Aortic wall

AV

Brachiocephalic

artery

Plaques

33

2. Experiment 2

2.1 Body weight

Mean body weight of the mice and the corresponding p-value at different time points during

the study are shown in table 6 below. There was no significant difference in body weight

between groups at baseline (p = 0.8883). The mice in group 1, the C57BL/6J strain on

normal diet, weighed significantly more than the apoE -/- mice on high-fat diet (group 2,

3 and 4) at 8 and 11 weeks (p < 0.0001). At 15 weeks, there was only a significant

difference between group 1 and 3 (p = 0.0487). During the study, mean body weight

significantly increased in all the groups (p < 0.0001 for group 1, 3 and 4; p = 0.0002 for

group 2).

Table 6 Means of body weight (g) of the different groups during the study course and corresponding p-value

at the different time points. Data are presented as means ± standard deviations.

Baseline 8 weeks 11 weeks 15 weeks

Body weight (g)

Group 1 17.23 ± 0.91 18.75 ± 1.07 20.54 ± 0.78 22.32 ± 1.58

Group 2 17.23 ± 1.36 17.57 ± 0.97 18.98 ± 0.92 20.70 ± 1.48

Group 3 17.35 ± 1.00 16.63 ± 0.74 17.77 ± 0.96 20.07 ± 1.05

Group 4 17.01 ± 0.94 17.42 ± 1.13 18.65 ± 1.02 21.36 ± 1.20

P-value 0.8883 <0.0001 <0.0001 0.0487

2.2 Calibrated integrated backscatter analysis

At baseline, 11 weeks and 15 weeks, there was no significant difference in cIB values for

the different ROIs between groups (see figure 14). At 8 weeks, there was a significant

difference in cIB values between group 2 and group 4 for the ROIs AVSA and AVTOT and

between group 2 and group 3 for the ROI AVSA (p = 0.0110, p = 0.0102 and p = 0.0110

respectively). For each group, no significant difference in cIB over time was observed (ROI

AVTOT over time: p = 0.1233, p = 0.0699, p = 0.3253 and p = 0.0765 for group 1, 2, 3

and 4 respectively). The mean cIB values of the different groups at the different time points

are summarized in addendum 3.

34

Fig

. 1

4 Com

parison of

cIB

valu

es fo

r th

e diffe

rent

RO

Is betw

een th

e diffe

rent

gro

ups at

baseline,

8 w

eeks,

11 w

eeks and 15 w

eeks.

RO

I =

regio

n o

f in

tere

st,

AVSA =

aort

ic v

alv

e s

hort

axis

, AVLA =

aort

ic v

alv

e long a

xis

, AVTO

T =

aort

ic v

alv

e t

ota

l, A

VG

LO

B =

aort

ic v

alv

e

glo

bal, c

IB =

calibra

ted inte

gra

ted b

ackscatt

er,

dB =

decib

el. T

hre

shold

for

sig

nific

ance p

< 0

.0125 (

corr

ection f

or

multip

le c

om

parisons).

* p

= 0

.0110, ** p

= 0

.0102.

*

**

35

2.3 In vivo CT imaging

All animals underwent in vivo micro-CT at 8, 11 and 15 weeks. Fig 15 represents these in

vivo images of one animal of each group at 15 weeks. Using the analysis tool Amide for

visual evaluation of these images, no calcium spots were detected at any time points.

Fig. 15 In vivo micro-CT images of one mouse of each group at 15 weeks. (A) group 1, (B)

group 2, (C) group 3, (D) group 4.

2.4 Semiquantitative plaque scoring

At baseline, 8 weeks and 11 weeks, a cohort of 3 mice, and at the end of the study, the

remaining mice were killed and the hearts with the aorta’s were dissected. Already at 11

weeks, calcifications on the aorta’s of the apoE -/- mice were visible macroscopically.

During dissection, the aorta’s were scored semiquantitatively under a microscope according

to the relative plaque size. The results of these plaque scoring of the aorta’s at the end of

the study (15 weeks) are represented in table 7 below. There were no plaques detected in

C57BL/6J mice, while all the apoE -/- mice had certain amount of plaques. No major

differences in plaque extent between the apoE -/- groups were seen (p = 0.3010).

Table 7 Semiquantitative plaque scoring of the aorta for the different groups. Results are shown as absolute

values and percentages. (-) no plaque visible, (+) a small plaque covering less than 50% of the surface of the

aorta, (++) large plaques covering more than 50%.

Group 1: Group 2: Group 3: Group 4:

C57BL/6J ApoE -/- no

injections

ApoE -/- vit K

alone

ApoE -/- vit K

+ warfarin

(-) 6 (100%) 0 0 0

(+) 0 6 (100%) 4 (66.7%) 5 (83.3%)

(++) 0 0 2 (33.3%) 1 (16.7%)

36

2.5 Ex vivo CT calcium quantification

The paraffin-fixed hearts with the aorta’s underwent an ex vivo high resolution micro-CT

and the obtained images were analysed in the medical image data analysis tool Amide as

described in materials and methods. The results of these analysis are described below in

table 8. There was a significant difference in the number of mice with a calcium spot

between the four groups (p = 0.0410). When the mice who received warfarin (group 4)

were compared to the other apoE -/- mice (group 2 and 3), this difference was non-

significant (p = 0.1730). The size and maximal intensity of the detected calcium spots were

greater for the warfarin treated apoE -/- mice, although this difference was also non-

significant (p = 0.3593 and p = 0.6840 respectively).

Table 8 Results of the ex vivo micro-CT paraffin-fixed image analysis in the medical image data analysis tool Amide.

For each group, the number of mice where a calcium spot was detected, is described (absolute values and

percentages). Of these calcium spots, the size and maximal intensity and the corresponding p-value is described.

Data are presented as means ± standard deviations.

Group 1: Group 2: Group 3: Group 4: P-value

C57BL/6J ApoE -/- no

injections

ApoE -/-

vit K alone

ApoE -/- vit K

+ warfarin

Visualisation of

calcium spot (n)

(+) 3 (50%) 1 (16.7%) 5 (83.3%) 6 (85.7%) 0.0410

(-) 3 (50%) 5 (83.3%) 1 (16.7%) 1 (14.3%)

Quantification of

calcium spot

Size (mm3) 0.11 ± 0.08 0.16 0.16 ± 0.08 0.19 ± 0.08 0.3593

Maximal

intensity (AU)

809.90 ±

19.06

724.01 891.50 ±

259.00

916.70 ±

104.70

0.6840

2.6 Von Kossa staining

Slices of 5 µm thick were generated of the paraffin embedded myocard and aorta from one

mouse of group 1 and from all mice of group 2, 3 and 4, killed at 15 weeks. These slices

were stained with Von Kossa and are shown in figure 16 below. For group 1, no plaques

neither calcium was present in the aorta. For group 2 and 3, plaques at the level of the

proximal aorta were present, but only in one out of six for group 2 and in none out of six

for group 3, calcium accumulation was detected. For group 4, all mice had plaques in their

proximal aorta and in 5 out of 7 mice, calcium was present in these plaques. The apoE -/-

mice of group 4 had thus significantly more calcium accumulation in their proximal aorta.

(p = 0.0140). Little calcium deposition in the AV of mice of all groups was seen.

37

Fig. 16 5 µm slices of paraffin embedded myocard and aorta, stained with Von Kossa. Calcium is represented in brown, nuclei

in red and cytoplasm in pink. A = group 1, B = group 2, C = group 3, D = group 4, AV = aortic valve.

A

C

B

D

Myocard

Aortic wall

Myocard Myocard

Myocard

AV

AV

AV

Aortic wall

AV

Plaques

Aortic wall

Plaques

Plaques

38

V. Discussion

Warfarin treatment was successfully applied in this animal model. The antidote vitamin K,

at a dose of 15 mg/kg, was sufficient to prevent bleeding. Based on the cIB analysis of

both experiments, no differences in valve calcification between different treatment groups

were detected. For experiment 1, calcium accumulation in the aortic wall of the apoE -/-

mice was seen on CT images and with Von Kossa staining, although no differences between

the different treatment groups could be detected. For experiment 2, no significant calcium

deposition in the aorta’s of the apoE -/- mice could be detected on CT images. However,

with the Von Kossa stain, the apoE -/- mice who received warfarin had significant more

calcium accumulation in their plaques at the level of the proximal aorta. VCAM-1 uptake

was significantly increased in the aorta’s of all apoE -/- mice, compared to the C57BL/6J

mice, and warfarin treatment did not change this level of VCAM-1 expression.

Study design

During the first experiment, warfarin was administered both early (from 8 till 11 weeks)

and late (from 22 till 25 weeks) in the disease progress. The late time point was based on

Schurgers et al., who investigated the effect of warfarin in apoE -/- mice who were already

for 12 weeks on a high-fat diet [32]. Administration early in the disease progress was

applied because at this stage, the NOACs are hypothesized to have the best anti-

atherogenic effect [58]. Because a lot of calcium spots were present in the apoE -/- mice,

but no difference between the warfarin treated group and the other apoE -/- mice could be

determined, we postulated that the atherosclerotic process was already advanced too far.

Therefore, during the second experiment, warfarin was administered from 7 till 14 weeks

and calcification burden was evaluated immediately after warfarin treatment was stopped.

A dosage of warfarin of 150 mg/kg was administered. This 150 mg/kg was determined by

Price et al., to be the minimum dosage needed to see an effect on arterial calcifications

[33]. The antidote vitamin K, at a dose of 15 mg/kg, was sufficient to neutralize the effect

of warfarin on the coagulation pathway. A loading dose of the antidote was shown to be

expendable. The biological half-life of vitamin K is too short to have an effect after 24 hours

pre-treatment.

39

Animal model

Genetically manipulated apoE -/- mice show impaired clearing of plasma lipoproteins and

are a well-known model for atherosclerosis research. Most studies who investigated the

effect of warfarin or the NOACs in apoE -/- mice only focussed on vascular calcification and

plaque burden, but didn’t include AV calcifications [32, 42, 43]. Regarding AS research,

mouse models do suffer from several important limitations. Wild-type mice on standard

diet do not exhibit spontaneous calcification and consequently the study of AS in mouse

models requires dietary and/or genetic or other interventions to induce a valvular calcium

burden [59]. Second, the anatomical structure of mouse aortic valves differs drastically

from that of humans. Their leaflets are usually only 5 – 10 cells thick and they do not have

a trilayer structure, making translation to clinical practice challenging [60, 61]. Another

important limitation is their size, which is too small to study hemodynamic on

echocardiography with good quality. On the other hand, rats are very cost-effective, easy

to manipulate and have a good size to perform echocardiography, but only few rat models

of CAVD are available to date. And like in mice, heart valves of rats also consist of less

layers than in humans. Both rabbits and pigs are valuable models to investigate aortic

valve disease, due to their trilayer valve anatomy and similar cholesterol metabolism to

humans. However, for easy housing and manipulation, their size is too big [60–62].

Calibrated integrated backscatter analysis

To quantify AV calcifications, the technique of cIB was used. In the first experiment, a

significant difference in cIB at baseline was found between the C57BL/6J mice and the

apoE -/- mice. However, the number of animals in the C57BL/6J group was too low (n=3)

to allow accurate comparison of the cIB values. To assure that this difference was not a

real feature, perhaps due to genetic strain difference, a higher amount of animals, namely

15 animals, were included in the control C57BL/6J group in experiment 2. Now, no

difference in cIB between the different strains was found.

We were not able to demonstrate a difference in cIB values and thus valve calcification

between the warfarin treated group and the other animals. A possible explanation could

be an unreliable use of cIB in our mouse model. The technique of cIB has recently been

validated as a valuable tool for the detection and quantification of AV calcifications in an in

vivo rat model [23–26]. However, it has not yet been validated in an in vivo mouse model,

as used in this study. A Vivid 7 Pro system with a 10 MHz neonatal probe was used to

obtain the images. Because a mouse is much smaller than a rat and the influence of the

sternum is more troublesome, it could be that, using this probe, we were not able to obtain

good quality echocardiographic images. If true, this could interfere with the reliability of

40

the cIB values. With this in mind, during the first experiment, echocardiographic images

were also made with a Vevo 2100 system at the Infinity lab of the UZ Gent. The Vevo 2100

has a 40 MHz probe allowing a better view of the heart chambers in mice. We have not yet

been able to analyse the images obtained with the Vevo 2100. In the future, a comparison

between the Vivid 7 Pro and the Vevo 2100 images can be done to critically analyse the

reliability of our cIB values and the use of this technique in an in vivo mouse model.

There is also an influence of angle-dependent scattering of ultrasound, which could limit

quantitative diagnosis [23, 26]. However, we eliminated such interference by keeping any

variables similar in each specimen. We used the same transducer and the same settings

for each animal and calculated cIB values by subtracting the background signal of the

adjacent blood pool. Also, by considering the mean cIB value of different ROIs for each

structure, as described in the methods, more diffracted ROIs get cancelled out.

Subcutaneous tissues cause diffraction of the ultrasound waves, leading to background

reverberation, so animal weight could be important when comparing cIB values [23, 26].

Therefore, the weight of the animals was carefully recorded throughout the study course

of experiment two. At baseline, no difference in weight was observed between the

C57BL/6J mice and the apoE -/- mice (17.23 ± 0.91 and 17.20 ± 1.10 gram respectively).

This is in line with the literature; apoE -/- mice are born healthy, with a similar body weight

as wild-type C57BL/6J mice [63]. All mice increased in weight over time. However, the

weight gain of the C57BL/6J group, on average 5.09 gram, was higher compared to the

three apoE -/- groups (3.47, 2.72 and 4.35 gram for group 2, 3 and 4 respectively). In

general, the lower weight gain of apoE -/- mice, compared to apoE +/+ mice (like C57BL/6J

mice), is mostly due to a decrease in fat mass [64]. Moreover, female apoE -/- mice were

used in this study and it is known for apoE -/- mice that females have lower body weights

than their male counterparts [63, 64]. The weight of the C57BL/6J mice significantly

increased during the study (p < 0.0001), but on the other hand, no significant increase in

cIB over time was observed (group 1 ROIAVTOT over time: p = 0.1233). In this study, no

influence of weight on cIB could be observed.

Semiquantitative plaque scoring

During dissection, the plaque burden at the level of the aorta was scored, both in

experiment 1 and 2. No major differences were seen in the scoring results between the

two experiments. This semiquantitative scoring renders ordinal values, going from no, to

moderate and severe plaque burden. This scoring is based on the surface of the aorta

covered by plaques, this is not an indication of calcium. It is a subjective score and,

although performed by the same person, intra-observer variation exists. Intra-observer

41

variation can definitely play a role, because the two experiments were performed with

more than a year in between. However, if the plaque burden is indeed not different between

experiment 1 and 2, this means that the amount of plaques is not changing much over the

period from 15 weeks (end of experiment 2) to 28 weeks (end of experiment 1). However,

the content of the plaques present will probably change as the atherosclerotic process

evolves.

CT calcium quantification with Amide software

On the ex vivo CT images in experiment 1, no calcium spots were detected in the C57BL/6J

mice and on the other hand, in every apoE -/- mouse, at least one calcium spot was

detected. On ex vivo CT images in experiment 2, calcium spots were not abundantly

present throughout the aorta’s. In the apoE -/- group who received no injections (group

2), only 1 mouse out of 6 was positive for calcium spot. In the apoE -/- group who only

received vitamin K (group 3), 5 mice out of 6 were positive for calcium spot. These findings

were against our expectations, we did not expect to find a difference between group 2 and

3, because vitamin K is not known to influence calcification burden. In the C57BL/6J group,

3 mice out of 6 were positive for calcium spot, which was also against our expectations.

C57BL/6J mice are wild-type mouse and are not prone to develop atherosclerosis and

calcium plaques, like the apoE -/- strain. Because calcium spots were also detected in the

C57BL/6J group, we can conclude that the chosen calcium threshold was probably too low.

Afterwards, using a more stringent threshold was not possible, because then, almost no

calcium spots were detected in the apoE -/- mice. We can postulate that calcium was not

abundantly present to allow this analysis with Amide.

When the mean size and mean maximal intensity of the calcium spots in experiment 1

were compared with those found in experiment 2, much higher values were found (mean

size experiment 1: 4.36 ± 1.76 mm3 versus experiment 2: 0.16 ± 0.08 mm3, mean

maximal intensity experiment 1: 1739 ± 228.07 AU versus experiment 2: 835.53 ± 127.59

AU). Although, no difference in plaque scoring during dissection was found, ex vivo CT

analysis showed that the plaque composition and calcium amount indeed differs between

the two experiments. Between the three apoE -/- groups themselves, no significant

difference in size or maximal intensity of the calcium spots were detected with Amide

software.

42

Von Kossa calcium stain

One C57BL/6J mouse of each experiment was stained and no plaques, nor calcium

deposition in the aorta’s were present. On the other hand, in the aorta’s of all apoE -/-

mice who were stained during the first experiment, plaques with calcium deposition were

present, but no differences between groups were seen. During the second experiment, the

apoE -/- mice who received warfarin had significant more calcium accumulation in the

plaques, present in the proximal aorta (p = 0.0140). So, although no difference in

calcification by Amide analysis could be quantified, Von Kossa staining showed that, during

the second experiment, there was indeed more calcium accumulation in the warfarin

treated apoE -/- mice. These calcium depositions were present in the proximal aorta, near

the AV and can be described as calcifications of the AV apparatus.

Calcifications of the AV leaflets were seen, but to see if differences between the different

treatment groups are present, we still need to look closer to the valves. In the C57BL/6J

mouse of experiment 2, little calcium deposition in the AV was detected. We didn’t expect

this result, as wild-type mice on normal diet are not prone to develop AV calcifications

[59]. However, tissue of only one C57BL/6J mouse from the whole group was examined.

To further investigate this result, more tissue stainings from C57BL/6J mice need to be

performed.

SPECT VCAM-1 imaging and biodistribution

Because the NOACs could have an anti-atherogenic and anti-inflammatory effect [40–43],

SPECT VCAM-1 imaging was used in the first experiment, to look at the inflammatory stage

of the plaques. VCAM-1 uptake was seen in all the aorta’s of the apoE -/- mice and only

very little uptake could be detected in the aorta’s of the C57BL/6J mice (1.45 ± 0.41 %IA/g

and 0.22 ± 0.11 %IA/g respectively, p = 0.0015). This is in line with the genetic

characteristics of the apoE -/- strain, namely to develop atherosclerosis in a short time

[43, 54]. No differences in VCAM-1 uptake in the aorta’s of the different apoE -/- groups

were seen (1.53 ± 0.27 %IA/g, 1.42 ± 0.54 %IA/g and 1.40 ± 0.43 %IA/g for group 2, 3

and 4 respectively, p = 0.8633). Higher amount of plaques would lead to an increase in

VCAM-1 and subsequently in an increase of sdAb uptake. This means that warfarin does

not have an influence on the inflammatory status of the plaques nor on the number of

plaques present. This is in line with the study of Schurgers et al., who demonstrated that

administration of warfarin in apoE -/- mice showed no change in plaque expansion, neither

number nor size distribution of plaques [32].

43

Future perspectives

The techniques used to evaluate AV and vascular calcifications can be further optimised.

For the detection and quantification of AV calcification in the future, cIB needs to be

improved and validated in an in vivo mouse model or other image modalities for

visualisation of valve calcification needs to be found. The micro-CT, used in this study, had

a pixel value of 0,166 mm. In the future, a CT scan with higher resolution can improve

the accuracy of our images and maybe influence our results. Also the technique of calcium

quantification with Amide need to be analysed critically and other data analysis tools need

to be considered.

A possible alternative for the study design in the future is to give warfarin for a longer

period, namely starting at the early time point and administering till the late time point. At

different time points during this treatment period, a cohort of mice could be killed and

calcification burden could be investigated over a longer period. Another possibility is to

increase the dose of warfarin administration. However, a further increase of the warfarin

dosage seems hard since we already used a 75 fold higher dosage then what is used to

have an effect on the coagulation pathway [65]. In analogy with the study of Schurgers et

al., warfarin and vitamin K1 can also be administered oral [32]. During this study, we didn’t

chose this method of administration because to produce this supplemented diet, it was

very expensive. On the other hand, the use of apoE -/- mice, regarding AS research, do

have several limitations, for example they have a different AV anatomy then humans [60–

62]. We can conclude that new and improved animal models of valve disease are still

desirable.

44

VI. Conclusions

Based on the cIB analysis of both experiments, no differences in AV calcification between

groups were detected. We were thus not able to demonstrate that warfarin treatment

accelerates the process of AV calcification. This study showed that inflammation status was

increased in the aorta’s of all apoE -/- mice and moreover, that warfarin does not have an

influence on the amount of plaques and does not change the level of inflammation. No

difference in vascular calcification between the warfarin treated apoE -/- mice and the

other apoE -/- mice were detected by Amide analysis of the CT images. However, Von

Kossa staining during experiment 2 showed a significant higher amount of calcium

deposition in the proximal aorta of the apoE -/- mice, treated with warfarin. This is in line

with the previously demonstrated increasing effect of warfarin on calcification burden.

However, this study wasn’t able to answer all the predefined questions, further research

with refinement of the study design, animal model and technical aspects to quantify the

calcification burden are needed.

45

References

[1] J. J. Thaden, V. T. Nkomo, and M. Enriquez-Sarano, “The Global Burden of Aortic

Stenosis,” Prog. Cardiovasc. Dis., vol. 56, no. 6, pp. 565–571, May 2014.

[2] M. R. Dweck, N. A. Boon, and D. E. Newby, “Calcific aortic stenosis: A disease of the

valve and the myocardium,” J. Am. Coll. Cardiol., 2012.

[3] A. Patel and A. J. Kirtane, “Aortic Valve Stenosis,” JAMA Cardiol., vol. 1, no. 5, p.

623, Aug. 2016.

[4] F. Akerström, M. G. Barderas, and L. Rodríguez-Padial, “Aortic stenosis: a general

overview of clinical, pathophysiological and therapeutic aspects.,” Expert Rev.

Cardiovasc. Ther., vol. 11, no. 2, pp. 239–50, Feb. 2013.

[5] A. L. Sverdlov, D. T. Ngo, M. J. Chapman, O. A. Ali, Y. Y. Chirkov, and J. D. Horowitz,

“Pathogenesis of aortic stenosis: not just a matter of wear and tear.,” Am. J.

Cardiovasc. Dis., vol. 1, no. 2, pp. 185–99, 2011.

[6] N. Rashedi and C. M. Otto, “Aortic Stenosis: Changing Disease Concepts.,” J.

Cardiovasc. Ultrasound, vol. 23, no. 2, pp. 59–69, Jun. 2015.

[7] P. Carità, G. Coppola, G. Novo, G. Caccamo, M. Guglielmo, F. Balasus, S. Novo, S.

Castrovinci, M. Moscarelli, K. Fattouch, and E. Corrado, “Aortic stenosis: insights on

pathogenesis and clinical implications.,” J. Geriatr. Cardiol., vol. 13, no. 6, pp. 489–

98, Sep. 2016.

[8] A. Vahanian, O. Alfieri, F. Andreotti, M. J. Antunes, G. Barón-Esquivias, H.

Baumgartner, M. A. Borger, T. P. Carrel, M. De Bonis, A. Evangelista, V. Falk, B.

Iung, P. Lancellotti, L. Pierard, S. Price, H. J. Schäfers, G. Schuler, J. Stepinska, K.

Swedberg, J. Takkenberg, U. O. Von Oppell, S. Windecker, J. L. Zamorano, M.

Zembala, J. J. Bax, C. Ceconi, V. Dean, C. Deaton, R. Fagard, C. Funck-Brentano,

D. Hasdai, A. Hoes, P. Kirchhof, J. Knuuti, P. Kolh, T. McDonagh, C. Moulin, B. A.

Popescu, Ž. Reiner, U. Sechtem, P. A. Sirnes, M. Tendera, A. Torbicki, L. Von

Segesser, L. P. Badano, M. Bunc, M. J. Claeys, N. Drinkovic, G. Filippatos, G. Habib,

A. Pieter Kappetein, R. Kassab, G. Y. H. Lip, N. Moat, G. Nickenig, C. M. Otto, J.

Pepper, N. Piazza, P. G. Pieper, R. Rosenhek, N. Shuka, E. Schwammenthal, J.

Schwitter, P. T. Mas, P. T. Trindade, and T. Walther, “Guidelines on the management

of valvular heart disease (version 2012),” Eur. Heart J., 2012.

[9] G. Thanassoulis, “Lipoprotein (a) in calcific aortic valve disease: from genomics to

novel drug target for aortic stenosis.,” J. Lipid Res., vol. 57, no. 6, pp. 917–24, Jun.

2016.

46

[10] N. M. Rajamannan, F. J. Evans, E. Aikawa, K. J. Grande-Allen, L. L. Demer, D. D.

Heistad, C. A. Simmons, K. S. Masters, P. Mathieu, K. D. O’Brien, F. J. Schoen, D.

A. Towler, A. P. Yoganathan, and C. M. Otto, “Calcific aortic valve disease: Not simply

a degenerative process: A review and agenda for research from the national heart

and lung and blood institute aortic stenosis working group,” Circulation, 2011.

[11] F. E. C. M. Peeters, S. J. R. Meex, M. R. Dweck, E. Aikawa, H. J. G. M. Crijns, L. J.

Schurgers, and B. L. J. H. Kietselaer, “Calcific aortic valve stenosis: hard disease in

the heart,” Eur. Heart J., Nov. 2017.

[12] P. Carità, G. Coppola, G. Novo, G. Caccamo, M. Guglielmo, F. Balasus, S. Novo, S.

Castrovinci, M. Moscarelli, K. Fattouch, and E. Corrado, “Symposium: Transcatheter

aortic valve implantation Aortic stenosis: insights on pathogenesis and clinical

implications,” J. Geriatr. Cardiol. 489 498 J Geriatr Cardiol, vol. 13, no. 13, pp.

489–498, 2016.

[13] R. V. Freeman, “Spectrum of Calcific Aortic Valve Disease: Pathogenesis, Disease

Progression, and Treatment Strategies,” Circulation, vol. 111, no. 24, pp. 3316–

3326, Jun. 2005.

[14] K. D. O’Brien, “Pathogenesis of Calcific Aortic Valve Disease: A Disease Process

Comes of Age (and a Good Deal More),” Arterioscler. Thromb. Vasc. Biol., vol. 26,

no. 8, pp. 1721–1728, Aug. 2006.

[15] R. Rosenhek, T. Binder, G. Porenta, I. Lang, G. Christ, M. Schemper, G. Maurer, and

H. Baumgartner, “Predictors of Outcome in Severe, Asymptomatic Aortic Stenosis,”

N. Engl. J. Med., vol. 343, no. 9, pp. 611–617, Aug. 2000.

[16] E. Nagy, P. Eriksson, M. Yousry, K. Caidahl, E. Ingelsson, G. K. Hansson, A. Franco-

Cereceda, and M. Bäck, “Valvular osteoclasts in calcification and aortic valve stenosis

severity,” Int. J. Cardiol., vol. 168, no. 3, pp. 2264–2271, Oct. 2013.

[17] R. Koos, A. H. Mahnken, A. M. Sinha, J. E. Wildberger, R. Hoffmann, and H. P. Kühl,

“Aortic Valve Calcification as a Marker for Aortic Stenosis Severity: Assessment on

16-MDCT,” Am. J. Roentgenol., vol. 183, no. 6, pp. 1813–1818, Dec. 2004.

[18] H. Baumgartner, V. Falk, J. J. Bax, M. De Bonis, C. Hamm, P. J. Holm, B. Iung, P.

Lancellotti, E. Lansac, D. R. Muñoz, R. Rosenhek, J. Sjögren, P. Tornos Mas, A.

Vahanian, T. Walther, O. Wendler, S. Windecker, and J. L. Zamorano, “2017

ESC/EACTS Guidelines for the management of valvular heart disease: The Task

Force for the Management of Valvular Heart Disease of the European Society of

Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery

(EACTS).,” Eur. Heart J., Aug. 2017.

47

[19] A. Vahanian, O. Alfieri, F. Andreotti, M. J. Antunes, G. Baron-Esquivias, H.

Baumgartner, M. A. Borger, T. P. Carrel, M. De Bonis, A. Evangelista, V. Falk, B.

Iung, P. Lancellotti, L. Pierard, S. Price, H.-J. Schafers, G. Schuler, J. Stepinska, K.

Swedberg, J. Takkenberg, U. O. Von Oppell, S. Windecker, J. L. Zamorano, M.

Zembala, J. J. Bax, H. Baumgartner, C. Ceconi, V. Dean, C. Deaton, R. Fagard, C.

Funck-Brentano, D. Hasdai, A. Hoes, P. Kirchhof, J. Knuuti, P. Kolh, T. McDonagh,

C. Moulin, B. A. Popescu, Z. Reiner, U. Sechtem, P. A. Sirnes, M. Tendera, A.

Torbicki, A. Vahanian, S. Windecker, B. A. Popescu, L. Von Segesser, L. P. Badano,

M. Bunc, M. J. Claeys, N. Drinkovic, G. Filippatos, G. Habib, A. P. Kappetein, R.

Kassab, G. Y. H. Lip, N. Moat, G. Nickenig, C. M. Otto, J. Pepper, N. Piazza, P. G.

Pieper, R. Rosenhek, N. Shuka, E. Schwammenthal, J. Schwitter, P. T. Mas, P. T.

Trindade, and T. Walther, “Guidelines on the management of valvular heart disease

(version 2012): The Joint Task Force on the Management of Valvular Heart Disease

of the European Society of Cardiology (ESC) and the European Association for

Cardio-Thoracic Surgery (EACTS),” Eur. Heart J., vol. 33, no. 19, pp. 2451–2496,

Oct. 2012.

[20] H. Baumgartner, J. Hung, J. Bermejo, J. B. Chambers, A. Evangelista, B. P. Griffin,

B. Lung, C. M. Otto, P. A. Pellikka, and M. Quinones, “Echocardiographic Assessment

of Valve Stenosis: EAE/ASE Recommendations for Clinical Practice,” J. Am. Soc.

Echocardiogr., vol. 22, no. 1, pp. 1–23, 2009.

[21] D. Messika-Zeitoun, M.-C. Aubry, D. Detaint, L. F. Bielak, P. A. Peyser, P. F. Sheedy,

S. T. Turner, J. F. Breen, C. Scott, A. J. Tajik, and M. Enriquez-Sarano, “Evaluation

and Clinical Implications of Aortic Valve Calcification Measured by Electron-Beam

Computed Tomography,” Circulation, vol. 110, no. 3, pp. 356–362, Jun. 2004.

[22] K. Maganti, V. H. Rigolin, M. E. Sarano, and R. O. Bonow, “Valvular heart disease:

diagnosis and management.,” Mayo Clin. Proc., vol. 85, no. 5, pp. 483–500, May

2010.

[23] K. Gillis, G. Bala, B. Roosens, I. Remory, H. De Raeve, S. Tierens, S. Hernot, G. Van

Camp, S. Droogmans, and B. Cosyns, “Echocardiographic integrated backscatter for

the differentiation between aortic valve calcification and valvular myxoid

degeneration in rats,” Eur. Hear. J. - Cardiovasc. Imaging, vol. 15, no. 9, pp. 1042–

1047, Sep. 2014.

[24] B. Roosens, G. Bala, K. Gillis, I. Remory, S. Droogmans, J. Somja, E. Delvenne, J.

De Nayer, J. Schiettecatte, P. Delvenne, P. Lancellotti, G. Van Camp, and B. Cosyns,

“Echocardiographic integrated backscatter for detecting progression and regression

of aortic valve calcifications in rats,” Cardiovasc. Ultrasound, vol. 11, no. 1, p. 4,

48

Jan. 2013.

[25] B. Roosens, G. Bala, S. Droogmans, J. Hostens, J. Somja, E. Delvenne, J.

Schiettecatte, P. Delvenne, T. Lahoutte, G. Van Camp, and B. Cosyns, “Occurrence

of cardiovascular calcifications in normal, aging rats.,” Exp. Gerontol., vol. 47, no.

8, pp. 614–9, Aug. 2012.

[26] B. Roosens, S. Droogmans, J. Hostens, J. Somja, E. Delvenne, S. Hernot, G. Bala,

C. Degaillier, V. Caveliers, P. Delvenne, T. Lahoutte, G. Van Camp, and B. Cosyns,

“Integrated backscatter for the in vivo quantification of supraphysiological vitamin

D(3)-induced cardiovascular calcifications in rats.,” Cardiovasc. Toxicol., vol. 11, no.

3, pp. 244–52, Sep. 2011.

[27] D. S. Owens, R. Katz, J. Takasu, R. Kronmal, M. J. Budoff, and K. D. O’Brien,

“Incidence and progression of aortic valve calcium in the Multi-ethnic Study of

Atherosclerosis (MESA).,” Am. J. Cardiol., vol. 105, no. 5, pp. 701–8, Mar. 2010.

[28] R. Katz, N. D. Wong, R. Kronmal, J. Takasu, D. M. Shavelle, J. L. Probstfield, A. G.

Bertoni, M. J. Budoff, and K. D. O’Brien, “Features of the metabolic syndrome and

diabetes mellitus as predictors of aortic valve calcification in the multi-ethnic study

of atherosclerosis,” Circulation, vol. 113, no. 17, pp. 2113–2119, 2006.

[29] B. Roosens, G. Bala, S. Droogmans, J. Hostens, J. Somja, E. Delvenne, J.

Schiettecatte, P. Delvenne, V. Caveliers, T. Lahoutte, G. Van Camp, and B. Cosyns,

“Echocardiographic Integrated Backscatter for Assessing Reduction of Aortic Valve

Calcifications by R-568 in a Rat Model of Chronic Kidney Disease,” Ultrasound Med.

Biol., vol. 39, no. 11, pp. 2075–2083, Nov. 2013.

[30] M. L. L. Chatrou, K. Winckers, T. M. Hackeng, C. P. Reutelingsperger, and L. J.

Schurgers, “Vascular calcification: The price to pay for anticoagulation therapy with

vitamin K-antagonists,” Blood Rev., vol. 26, no. 4, pp. 155–166, Jul. 2012.

[31] R. J. M. W. Rennenberg, B. J. Van Varik, L. J. Schurgers, K. Hamulyak, H. Ten Cate,

T. Leiner, C. Vermeer, P. W. De Leeuw, and A. A. Kroon, “Chronic coumarin

treatment is associated with increased extracoronary arterial calcification in

humans.”

[32] L. J. Schurgers, I. A. Joosen, E. M. Laufer, M. L. L. Chatrou, M. Herfs, M. H. M.

Winkens, R. Westenfeld, V. Veulemans, T. Krueger, C. M. Shanahan, W. Jahnen-

Dechent, E. Biessen, J. Narula, C. Vermeer, L. Hofstra, and C. P. Reutelingsperger,

“Vitamin K-Antagonists Accelerate Atherosclerotic Calcification and Induce a

Vulnerable Plaque Phenotype,” PLoS One, vol. 7, no. 8, p. e43229, Aug. 2012.

49

[33] P. A. Price, S. A. Faus, and M. K. Williamson, “Warfarin Causes Rapid Calcification of

the Elastic Lamellae in Rat Arteries and Heart Valves,” Arterioscler. Thromb. Vasc.

Biol., vol. 18, no. 9, 1998.

[34] G. Luo, P. Ducy, M. D. McKee, G. J. Pinero, E. Loyer, R. R. Behringer, and G.

Karsenty, “Spontaneous calcification of arteries and cartilage in mice lacking matrix

GLA protein,” Nature, vol. 386, no. 6620, pp. 78–81, Mar. 1997.

[35] L. J. Schurgers, H. M. H. Spronk, B. A. M. Soute, P. M. Schiffers, J. G. R. DeMey,

and C. Vermeer, “Regression of warfarin-induced medial elastocalcinosis by high

intake of vitamin K in rats.,” Blood, vol. 109, no. 7, pp. 2823–31, Apr. 2007.

[36] P. A. Price, S. A. Faus, and M. K. Williamson, “Warfarin-induced artery calcification

is accelerated by growth and vitamin D.,” Arterioscler. Thromb. Vasc. Biol., vol. 20,

no. 2, pp. 317–27, Feb. 2000.

[37] L. J. Schurgers, H. Aebert, C. Vermeer, B. Bultmann, and J. Janzen, “Oral

anticoagulant treatment: friend or foe in cardiovascular disease?,” Blood, vol. 104,

no. 10, pp. 3231–3232, Nov. 2004.

[38] R. Koos, A. H. Mahnken, G. Mühlenbruch, V. Brandenburg, B. Pflueger, J. E.

Wildberger, and H. P. Kühl, “Relation of Oral Anticoagulation to Cardiac Valvular and

Coronary Calcium Assessed by Multislice Spiral Computed Tomography,” Am. J.

Cardiol., vol. 96, no. 6, pp. 747–749, Sep. 2005.

[39] J. Steffel, P. Verhamme, T. S. Potpara, P. Albaladejo, M. Antz, L. Desteghe, K. G.

Haeusler, J. Oldgren, H. Reinecke, V. Roldan-Schilling, N. Rowell, P. Sinnaeve, R.

Collins, A. J. Camm, H. Heidbüchel, G. YH Lip, J. Weitz, L. Fauchier, D. Lane, G.

Boriani, A. Goette, R. Keegan, R. MacFadyen, C.-E. Chiang, B. Joung, and W.

Shimizu, “The 2018 European Heart Rhythm Association Practical Guide on the use

of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation,”

Eur. Heart J., Mar. 2018.

[40] J.-B. Kim, H. J. Joung, J. M. Lee, J. S. Woo, W. Kim, K. S. Kim, K. H. Lee, and W.

Kim, “Evaluation of the vascular protective effects of new oral anticoagulants in high-

risk patients with atrial fibrillation (PREFER-AF): study protocol for a randomized

controlled trial,” Trials, vol. 17, no. 1, p. 422, 2016.

[41] E. Blessing, Q. Zhou, F. Bea, M. Preusch, H. Wang, B. Isermann, K. Shahzad, and

H. A. Katus, “Evaluation of plaque stability of advanced atherosclerotic lesions in Apo

E-deficient mice after treatment with the oral factor Xa inhibitor rivaroxaban,”

Mediators Inflamm., vol. 2011, 2011.

50

[42] I.-O. Lee, M. T. Kratz, S. H. Schirmer, M. Baumhakel, and M. Bohm, “The Effects of

Direct Thrombin Inhibition with Dabigatran on Plaque Formation and Endothelial

Function in Apolipoprotein E-Deficient Mice,” J. Pharmacol. Exp. Ther., vol. 343, no.

2, pp. 253–257, 2012.

[43] T. Hara, D. Fukuda, K. Tanaka, Y. Higashikuni, Y. Hirata, S. Nishimoto, S. Yagi, H.

Yamada, T. Soeki, T. Wakatsuki, M. Shimabukuro, and M. Sata, “Rivaroxaban, a

novel oral anticoagulant, attenuates atherosclerotic plaque progression and

destabilization in ApoE-deficient mice,” Atherosclerosis, vol. 242, no. 2, pp. 639–

646, 2015.

[44] M. R. Preusch, N. Ieronimakis, E. S. Wijelath, S. Cabbage, J. Ricks, F. Bea, M. Reyes,

J. van Ryn, and M. E. Rosenfeld, “Dabigatran etexilate retards the initiation and

progression of atherosclerotic lesions and inhibits the expression of oncostatin M in

apolipoprotein E-deficient mice.,” Drug Des. Devel. Ther., vol. 9, pp. 5203–11, 2015.

[45] S. Pingel, V. Tiyerili, J. Mueller, N. Werner, G. Nickenig, and C. Mueller, “Thrombin

inhibition by dabigatran attenuates atherosclerosis in ApoE deficient mice.,” Arch.

Med. Sci., vol. 10, no. 1, pp. 154–60, Feb. 2014.

[46] N. P. E. Kadoglou, P. Moustardas, M. Katsimpoulas, A. Kapelouzou, N.

Kostomitsopoulos, K. Schafer, A. Kostakis, and C. D. Liapis, “The Beneficial Effects

of a Direct Thrombin Inhibitor, Dabigatran Etexilate, on the Development and

Stability of Atherosclerotic Lesions in Apolipoprotein E-deficient Mice,” Cardiovasc.

Drugs Ther., vol. 26, no. 5, pp. 367–374, Oct. 2012.

[47] J. I. Borissoff, J. J. T. Otten, S. Heeneman, P. Leenders, R. van Oerle, O. Soehnlein,

S. T. B. G. Loubele, K. Hamulyák, T. M. Hackeng, M. J. A. P. Daemen, J. L. Degen,

H. Weiler, C. T. Esmon, J. van Ryn, E. A. L. Biessen, H. M. H. Spronk, and H. ten

Cate, “Genetic and Pharmacological Modifications of Thrombin Formation in

Apolipoprotein E-deficient Mice Determine Atherosclerosis Severity and

Atherothrombosis Onset in a Neutrophil-Dependent Manner,” PLoS One, vol. 8, no.

2, p. e55784, Feb. 2013.

[48] M. Kupari, H. Turto, and J. Lommi, “Left ventricular hypertrophy in aortic valve

stenosis: preventive or promotive of systolic dysfunction and heart failure?”

[49] R. Yotti and J. Bermejo, “Left ventricular hypertrophy in aortic valve stenosis: friend

or foe?”

[50] P. Kirchhof, S. Benussi, D. Kotecha, A. Ahlsson, D. Atar, B. Casadei, M. Castella, H.-

C. Diener, H. Heidbuchel, J. Hendriks, G. Hindricks, A. S. Manolis, J. Oldgren, B. A.

Popescu, U. Schotten, B. Van Putte, P. Vardas, S. Agewall, J. Camm, G. Baron

51

Esquivias, W. Budts, S. Carerj, F. Casselman, A. Coca, R. De Caterina, S. Deftereos,

D. Dobrev, J. M. Ferro, G. Filippatos, D. Fitzsimons, B. Gorenek, M. Guenoun, S. H.

Hohnloser, P. Kolh, G. Y. H. Lip, A. Manolis, J. McMurray, P. Ponikowski, R. Rosenhek,

F. Ruschitzka, I. Savelieva, S. Sharma, P. Suwalski, J. L. Tamargo, C. J. Taylor, I.

C. Van Gelder, A. A. Voors, S. Windecker, J. L. Zamorano, K. Zeppenfeld, J. L.

Zamorano, V. Aboyans, S. Achenbach, S. Agewall, L. Badimon, G. Barón-Esquivias,

H. Baumgartner, J. J. Bax, H. Bueno, S. Carerj, V. Dean, Ç. Erol, D. Fitzsimons, O.

Gaemperli, P. Kirchhof, P. Kolh, P. Lancellotti, G. Y. H. Lip, P. Nihoyannopoulos, M.

F. Piepoli, P. Ponikowski, M. Roffi, A. Torbicki, A. Vaz Carneiro, S. Windecker, H. G.

Hayrapetyan, F. X. Roithinger, F. Aliyev, A. Chasnoits, G. H. Mairesse, D. L.

Matičević, T. Shalganov, B. Skorić, L. Antoniades, M. Taborsky, S. Pehrson, S.

Khaled, P. Kampus, A. Hedman, L. Poposka, J.-Y. Le Heuzey, K. Estadashvili, D.

Bänsch, Z. Csanádi, D. Keane, R. Beinart, F. Romeo, K. Koshumbayeva, G.

Bajraktari, A. Mirrakhimov, O. Kalejs, S. Nasr, G. Marinskis, C. Dimmer, M. Sammut,

A. Grosu, S. Abdelali, M. E. W. Hemels, O.-G. Anfinsen, B. Średniawa, P. Adragao,

G.-A. Dan, E. N. Mikhaylov, M. Zavatta, T. Potpara, P. H. Slovenia, I. Zupan, A.

Arenal, F. Braunschweig, D. Shah, A. S. Ouali, M. Demir, O. Sychov, and E. Duncan,

“2016 ESC Guidelines for the management of atrial fibrillation developed in

collaboration with EACTS,” Eur. Heart J., vol. 37, no. 38, pp. 2893–2962, Oct. 2016.

[51] J. V. (Ian) Nixon, The AHA Clinical Cardiac Consult, 3rd etitio. Wolters Kluwer

Lippincott Williams & Wilkins, 2011.

[52] F. Levy, D. Rusinaru, S. Maréchaux, V. Charles, M. Peltier, and C. Tribouilloy,

“Determinants and Prognosis of Atrial Fibrillation in Patients With Aortic Stenosis,”

Am. J. Cardiol., vol. 116, no. 10, pp. 1541–1546, Nov. 2015.

[53] S. Stortecky, L. Buellesfeld, P. Wenaweser, D. Heg, T. Pilgrim, A. A. Khattab, S.

Gloekler, C. Huber, F. Nietlispach, B. Meier, P. Juni, and S. Windecker, “Atrial

Fibrillation and Aortic Stenosis: Impact on Clinical Outcomes Among Patients

Undergoing Transcatheter Aortic Valve Implantation,” Circ. Cardiovasc. Interv., vol.

6, no. 1, pp. 77–84, Feb. 2013.

[54] G. Kolovou, K. Anagnostopoulou, D. P. Mikhailidis, and D. V Cokkinos,

“Apolipoprotein E knockout models.,” Curr. Pharm. Des., vol. 14, no. 4, pp. 338–51,

2008.

[55] L. J. Schurgers, H. M. H. Spronk, B. A. M. Soute, P. M. Schiffers, J. G. R. DeMey,

and C. Vermeer, “Regression of warfarin-induced medial elastocalcinosis by high

intake of vitamin K in rats,” Blood, vol. 109, no. 7, pp. 2823–2831, 2007.

52

[56] G. Bala, A. Blykers, C. Xavier, B. Descamps, A. Broisat, C. Ghezzi, D. Fagret, G. Van

Camp, V. Caveliers, C. Vanhove, T. Lahoutte, S. Droogmans, B. Cosyns, N.

Devoogdt, and S. Hernot, “Targeting of vascular cell adhesion molecule-1 by 18F-

labelled nanobodies for PET/CT imaging of inflamed atherosclerotic plaques,” Eur.

Hear. J. – Cardiovasc. Imaging, vol. 17, no. 9, pp. 1001–1008, Sep. 2016.

[57] J. De Vos, I. Mathijs, C. Xavier, S. Massa, U. Wernery, L. Bouwens, T. Lahoutte, S.

Muyldermans, and N. Devoogdt, “Specific Targeting of Atherosclerotic Plaques in

ApoE Mice Using a New Camelid sdAb Binding the Vulnerable Plaque Marker LOX-1,”

Mol Imaging Biol, vol. 16, pp. 690–698, 2014.

[58] J. I. Borissoff, J. J. T. Otten, S. Heeneman, P. Leenders, R. van Oerle, O. Soehnlein,

S. T. B. G. Loubele, K. Hamulyák, T. M. Hackeng, M. J. A. P. Daemen, J. L. Degen,

H. Weiler, C. T. Esmon, J. van Ryn, E. A. L. Biessen, H. M. H. Spronk, and H. ten

Cate, “Genetic and pharmacological modifications of thrombin formation in

apolipoprotein e-deficient mice determine atherosclerosis severity and

atherothrombosis onset in a neutrophil-dependent manner.,” PLoS One, vol. 8, no.

2, p. e55784, 2013.

[59] M. Guerraty and E. R. Mohler, “Models of Aortic Valve Calcification,” J. Investig.

Med., vol. 55, no. 6, pp. 278–283, Sep. 2007.

[60] Y. Wang, H. Lu, Z. Huang, H. Lin, Z. Lei, X. Chen, M. Tang, F. Gao, M. Dong, R. Li,

and L. Lin, “Apolipoprotein E-knockout mice on high-fat diet show autoimmune injury

on kidney and aorta,” Biochem. Biophys. Res. Commun., vol. 450, no. 1, pp. 788–

793, Jul. 2014.

[61] The Jackson Laboratory, “Body Weight Information for B6.129P2-Apoetm1Un/J

(002052).” [Online]. Available: https://www.jax.org/jax-mice-and-services/strain-

data-sheet-pages/body-weight-chart-002052.

[62] B. Roosens, G. Bala, S. Droogmans, G. Van Camp, J. Breyne, and B. Cosyns, “Animal

models of organic heart valve disease,” Int. J. Cardiol., vol. 165, no. 3, pp. 398–

409, May 2013.

[63] B. Emini Veseli, P. Perrotta, G. R. A. De Meyer, L. Roth, C. Van der Donckt, W.

Martinet, and G. R. Y. De Meyer, “Animal models of atherosclerosis,” Eur. J.

Pharmacol., vol. 816, pp. 3–13, Dec. 2017.

[64] S. M. Hofmann, D. Perez-Tilve, T. M. Greer, B. A. Coburn, E. Grant, J. E. Basford, M.

H. Tschöp, and D. Y. Hui, “Defective lipid delivery modulates glucose tolerance and

metabolic response to diet in apolipoprotein E-deficient mice.,” Diabetes, vol. 57,

no. 1, pp. 5–12, Jan. 2008.

53

[65] C. Foerch, K. Arai, G. Jin, K.-P. Park, S. Pallast, K. van Leyen, and E. H. Lo,

“Experimental Model of Warfarin-Associated Intracerebral Hemorrhage,” Stroke, vol.

39, no. 12, pp. 3397–3404, Dec. 2008.

[66] K. L. Sider, M. C. Blaser, and C. A. Simmons, “Animal models of calcific aortic valve

disease.,” Int. J. Inflam., vol. 2011, p. 364310, 2011.

[67] G. S. Getz and C. A. Reardon, “Animal models of atherosclerosis.,” Arterioscler.

Thromb. Vasc. Biol., vol. 32, no. 5, pp. 1104–15, May 2012.

[68] A. Broisat, S. Hernot, J. Toczek, J. De Vos, L. M. Riou, S. Martin, M. Ahmadi, N.

Thielens, U. Wernery, V. Caveliers, S. Muyldermans, T. Lahoutte, D. Fagret, C.

Ghezzi, and N. Devoogdt, “Nanobodies Targeting Mouse/Human VCAM1 for the

Nuclear Imaging of Atherosclerotic Lesions,” Circ. Res., vol. 110, no. 7, pp. 927–

937, Mar. 2012.

54

Addendum

Addendum 1: Means cIB values of the AVSA, AVLA, AVTOT and AVGLOB for each group

during the study course of experiment one. Data are presented as means ± standard

deviations. dB: decibel.

Baseline 21 weeks 24 weeks 26 weeks

cIB AVSA (dB)

Group 1 3.49 ± 1.65 5.32 ± 4.72 6.89 ± 1.92 10.43 ± 2.73

Group 2 7.09 ± 2.96 9.97 ± 3.35 11.45 ± 3.21 11.91 ± 3.94

Group 3 5.34 ± 2.09 10.23 ± 2.62 8.94 ± 1.43 11.56 ± 4.16

Group 4 8.90 ± 3.06 10.04 ± 3.45 11.25 ± 1.74 14.42 ± 1.20

cIB AVLA (dB)

Group 1 7.27 ± 2.73 10.16 ± 0.86 13.81 ± 4.56 14.68 ± 4.19

Group 2 12.66 ± 1.75 15.68 ± 2.75 14.65 ± 2.76 15.04 ± 2.97

Group 3 13.32 ± 3.44 13.09 ± 1.50 16.85 ± 4.89 14.64 ± 5.86

Group 4 12.81 ± 3.54 14.90 ± 3.91 16.05 ± 2.00 15.53 ± 2.11

cIB AVTOT (dB)

Group 1 5.38 ± 1.99 7.74 ± 2.68 10.35 ± 3.08 12.56 ± 2.94

Group 2 9.88 ± 1.95 12.82 ± 2.62 13.05 ± 2.10 13.47 ± 2.95

Group 3 9.33 ± 1.89 11.66 ± 1.73 12.90 ± 2.32 13.10 ± 4.33

Group 4 10.86 ± 2.07 12.47 ± 2.77 13.65 ± 1.69 14.98 ± 1.14

cIB AVGLOB (dB)

Group 1 9.01 ± 3.13 11.64 ± 2.44 15.75 ± 4.87 15.24 ± 2.55

Group 2 12.70 ± 2.33 15.82 ± 4.43 14.68 ± 3.11 14.87 ± 3.10

Group 3 12.48 ± 2.46 13.26 ± 0.99 15.05 ± 3.38 14.37 ± 4.20

Group 4 12.86 ± 1.38 15.45 ± 2.93 16.74 ± 3.51 17.08 ± 1.96

55

Addendum 2: Ex vivo biodistribution data of 99mTc-labeled VCAM-1 sdAb, expressed as

%IA/g tissue, 2h after injection in the different study groups. Data are expressed as means

± standard deviations.

Group 1: Group 2: Group 3: Group 4:

C57BL/6J ApoE -/- no

injections

ApoE -/- vit K +

warfarin early

ApoE -/- vit K +

warfarin late

Thymus 0.86 ± 0.15 2.21 ± 0.47 2.21 ± 1.07 2.34 ± 0.58

Heart 0.14 ± 0.06 0.36 ± 0.12 0.39 ± 0.12 0.39 ± 0.10

Lungs 0.82 ± 0.25 1.81 ± 0.32 1.74 ± 0.57 1.79 ± 0.45

Aorta 0.22 ± 0.11 1.53 ± 0.27 1.42 ± 0.54 1.40 ± 0.43

Liver 0.63 ± 0.16 1.63 ± 0.16 1.58 ± 0.37 2.16 ± 0.47

Spleen 3.72 ± 0.76 10.44 ± 1.59 8.98 ± 2.10 11.26 ± 2.21

Pancreas 0.11 ± 0.03 0.24 ± 0.03 0.28 ± 0.09 0.26 ± 0.04

Kidney L 135.10 ± 10.10 142.94 ± 13.41 118.78 ± 13.60 155.29 ± 29.17

Kidney R 139.08 ± 3.40 134.42 ± 25.42 120.37 ± 15.31 156.22 ± 36.01

Stomach 0.38 ± 0.09 0.81 ± 0.28 0.74 ± 0.26 0.87 ± 0.26

Small intestine 0.28 ± 0.09 0.58 ± 0.06 0.60 ± 0.15 0.61 ± 0.09

Large intestine 0.22 ± 0.05 0.43 ± 0.06 0.44 ± 0.10 0.48 ± 0.06

Muscle 0.04 ± 0.01 0.16 ± 0.06 0.18 ± 0.09 0.12 ± 0.02

Bone 0.47 ± 0.01 1.20 ± 0.34 0.85 ± 0.30 1.30 ± 0.41

Lymph nodes 0.29 ± 0.02 1.25 ± 0.33 1.23 ± 0.52 1.52 ± 0.37

Blood 0.31 ± 0.13 0.82 ± 0.13 0.91 ± 0.33 0.92 ± 0.22

56

Addendum 3: Means cIB values of the AVSA, AVLA, AVTOT and AVGLOB for each group

during the study course of experiment two. Data are presented as means ± standard

deviations. dB: decibel.

Baseline 8 weeks 11 weeks 15 weeks

cIB AVSA (dB)

Group 1 7.61 ± 3.35 8.01 ± 2.54 8.92 ± 1.98 8.56 ± 1.85

Group 2 8.75 ± 4.62 6.60 ± 2.49 6.69 ± 2.17 9.70 ± 1.61

Group 3 8.72 ± 3.90 9.78 ± 2.58 9.17 ± 2.33 8.69 ± 2.72

Group 4 8.26 ± 2.73 9.71 ± 2.70 7.09 ± 4.30 10.48 ± 2.11

cIB AVLA (dB)

Group 1 14.20 ± 2.51 14.14 ± 2.25 16.26 ± 2.37 15.54 ± 3.67

Group 2 14.13 ± 2.41 13.33 ± 2.58 16.65 ± 2.78 15.78 ± 3.51

Group 3 14.43 ± 3.34 13.65 ± 2.47 15.44 ± 1.88 15.39 ± 2.48

Group 4 16.08 ± 3.37 15.81 ± 3.80 15.45 ± 2.72 16.58 ± 3.13

cIB AVTOT (dB)

Group 1 10.91 ± 2.11 11.08 ± 1.92 12.59 ± 1.73 12.05 ± 1.43

Group 2 11.44 ± 2.55 9.96 ± 1.74 11.67 ± 1.86 12.74 ± 2.47

Group 3 11.57 ± 2.11 11.71 ± 1.45 12.30 ± 1.71 12.04 ± 1.89

Group 4 12.17 ± 1.66 12.76 ± 2.59 11.27 ± 2.92 13.53 ± 2.17

cIB AVGLOB (dB)

Group 1 14.55 ± 2.44 14.14 ± 2.46 15.01 ± 1.65 15.28 ± 1.87

Group 2 14.80 ± 2.75 12.30 ± 1.92 15.21 ± 2.51 15.32 ± 2.89

Group 3 14.81 ± 2.40 14.53 ± 1.98 14.20 ± 4.03 15.29 ± 2.26

Group 4 16.07 ± 2.09 16.08 ± 3.92 15.43 ± 3.12 16.26 ± 2.25