Download CORONARY COMPUTED TOMOGRAPHY ANGIOGRAPHY IN

From the Division of Medical Imaging and Technology
Department of Clinical Science, Intervention and Technology
Karolinska Institutet, Stockholm, Sweden
CORONARY COMPUTED TOMOGRAPHY
ANGIOGRAPHY IN PATIENTS WITH
MYOCARDIAL INFARCTION AND NONOBSTRUCTED CORONARY ARTERIES
Elin Bacsovics Brolin
Stockholm 2015
All previously published papers were reproduced with permission from the publisher.
Published by Karolinska Institutet.
Printed by E-print AB 2015
© Elin Bacsovics Brolin, 2015
ISBN 978-91-7676-090-1
CORONARY COMPUTED TOMOGRAPHY
ANGIOGRAPHY IN PATIENTS WITH
MYOCARDIAL INFARCTION AND
NON-OBSTRUCTED CORONARY ARTERIES
THESIS FOR DOCTORAL DEGREE (Ph.D.)
By
Elin Bacsovics Brolin
Principal Supervisor:
Associate Professor Kerstin Cederlund
Karolinska Institutet
Department of Clinical Science,
Intervention and Technology
Division of Medical Imaging and Technology
Co-supervisors:
Associate Professor Torkel Brismar
Karolinska Institutet
Department of Clinical Science,
Intervention and Technology
Division of Medical Imaging and Technology
Doctor Andreas Rück
Karolinska Institutet
Department of Medicine Solna
Division of Cardiology
Professor Per Tornvall
Karolinska Institutet
Department of Clinical Science and Education,
Södersjukhuset
Division of Cardiology
Opponent:
Professor Göran Bergström
University of Gothenburg
Wallenberg Laboratory, Clinical Physiology
Sahlgrenska Academy, Institute of Medicine
Examination Board:
Associate Professor Sven-Göran Fransson
Linköping University
Department of Medical and Health Sciences
Division of Radiological Sciences
Associate Professor Kerstin Jensen-Urstad
Karolinska Institutet
Department of Clinical Science and Education,
Södersjukhuset
Associate Professor Elmir Omerovic
University of Gothenburg
Sahlgrenska Academy, Institute of Medicine
Department of Molecular and Clinical Medicine
ABSTRACT
Cardiovascular disease (CVD) is the number one cause of death worldwide. In Sweden,
almost 30 000 people suffer an acute myocardial infarction (AMI) each year and, despite the
greatly improved survival after AMI, CVD remains the leading cause of death among women
and men. During the last decade, there has been increasing awareness of the significant
minority of patients with acute myocardial infarction, for whom invasive coronary
angiography (ICA) does not show any coronary artery stenoses. This condition is called
myocardial infarction and non-obstructed coronary arteries (MINOCA) and is still
incompletely understood. Another condition that has gained increasing attention is Takotsubo
syndrome (TS), also known as stress-induced cardiomyopathy or the broken heart syndrome.
Patients with TS may be considered a sub-group of MINOCA. Important advances in
coronary computed tomography angiography (CTA) technology have enabled safe and
accurate non-invasive imaging of the coronary arteries. In contrast to ICA, coronary CTA
allows for detection of non-obstructive as well as obstructive coronary artery disease (CAD).
Coronary CTA is also useful for assessment of plaque characteristics and detection of
myocardial bridging (MB). In order to improve CVD risk prediction, numerous risk markers
have emerged, among them carotid artery intima-media thickness (IMT), endothelial function
determined by digital reactive hyperemia peripheral arterial tonometry (RH-PAT) and
different categories of circulating biomarkers. Coronary CTA plaque burden has a prognostic
value in CVD risk assessment, but its association with other risk markers is incompletely
studied.
There were two major aims of this thesis. The first aim was to investigate the underlying
mechanisms of MINOCA (study I and III). The second aim was to examine the association
between coronary CTA plaque burden and other risk markers of CVD (study II and IV).
In study I we compared coronary CTA plaque burden in MINOCA patients and controls,
matched by age and gender. We found that coronary CTA plaque burden was similar in the
two groups and that a large proportion of MINOCA patients (42%) had no signs of CAD at
coronary CTA. Non-obstructive CAD is most likely not a frequent cause of MINOCA.
In study II, 58 volunteers, free from clinical CVD, underwent testing for IMT and RH-PAT
as well as coronary CTA. More than half of the study group had evidence of subclinical CAD
at coronary CTA. There was no association between IMT or RH-PAT and presence or extent
of CAD. Neither evaluation of IMT nor RH-PAT can reliably be used to predict coronary
CTA plaque burden in clinically healthy subjects.
In study III the prevalence of MB, determined by coronary CTA, was compared for
MINOCA patients, including a subgroup with TS, and matched controls. The MB depiction
rate of coronary CTA and ICA was compared. MB was frequent, with a similar prevalence in
MINOCA patients, patients with TS and controls, suggesting MB is not a frequent cause of
MINOCA or TS. Coronary CTA detects significantly more MB than ICA.
In study IV, 115 subjects with predominantly low-to-intermediate CVD risk and normal or
mildly reduced kidney function, underwent coronary CTA and laboratory testing. The groups
without and with CAD differed with regard to levels of adiponectin, lipoprotein(a) and
cystatin C. However, in a multivariable logistic regression model, only male sex and levels of
cystatin C were independently associated with non-obstructive CAD at coronary CTA.
In conclusion, non-obstructive CAD is not a frequent cause of MINOCA in patients with
angiographically normal or near-normal coronary arteries. MINOCA should probably not be
considered a definitive diagnosis, but rather a working diagnosis, warranting additional
diagnostic evaluation. TS, which is one of the possible underlying causes of MINOCA, is
most likely not caused by MB. For TS, future consensus on the diagnostic criteria will
facilitate research on pathophysiological mechanisms, diagnosis, prognosis and patient
management. Circulating cystatin C was associated with non-obstructive CAD and may thus
have a potential to serve as a screening test for subclinical CAD. However, CVD risk
assessment is complex and large-scale studies are necessary to investigate which combination
of imaging parameters and other risk markers yields the most accurate individual risk
prediction.
LIST OF SCIENTIFIC PAPERS
The thesis is based on the following four papers, referred to in the text by their roman
numerals:
I. Brolin EB, Jernberg T, Brismar TB, Daniel M, Henareh L, Ripsweden J,
Tornvall P, Cederlund K. Coronary plaque burden, as determined by cardiac
computed tomography, in patients with myocardial infarction and
angiographically normal coronary arteries compared to healthy volunteers: a
prospective multicenter observational study. PloS one 2014;9(6):e99783.
doi: 10.1371/journal.pone.0099783.
II. Brolin EB, Agewall S, Brismar TB, Caidahl K, Tornvall P, Cederlund K.
Neither endothelial function nor carotid artery intima-media thickness
predicts coronary computed tomography angiography plaque burden in
clinically healthy subjects: a cross-sectional study. BMC Cardiovasc Disord
2015;15:63. doi: 10.1186/s12872-015-0061-x.
III. Brolin EB, Brismar TB, Collste O, Y-Hassan S, Henareh L, Tornvall P,
Cederlund K. Prevalence of myocardial bridging in patients with myocardial
infarction and nonobstructed coronary arteries. Am J Cardiol 2015. Advance
online publication. doi: 10.1016/j.amjcard.2015.09.017.
IV. Brolin EB, Agewall S, Cederlund K, Ekenbäck C, Henareh L, Malmqvist K,
Rück A, Svensson A, Tornvall P. Cystatin C is associated with nonobstructive coronary artery disease determined by coronary computed
tomography angiography. Submitted manuscript.
CONTENTS
1 Introduction ..................................................................................................................... 1
1.1 Coronary computed tomography angiography .................................................... 1
1.2 Other imaging modalities for coronary artery disease ......................................... 2
1.3 Myocardial infarction and non-obstructed coronary arteries............................... 3
1.3.1 Acute myocardial infarction .................................................................... 3
1.3.2 Myocardial infarction and non-obstructed coronary arteries .................. 4
1.4 Takotsubo syndrome ............................................................................................ 5
1.4.1 Imaging in Takotsubo syndrome ............................................................. 7
1.5 Myocardial bridging ............................................................................................. 8
1.6 Cardiovascular disease risk assessment ............................................................... 9
1.6.1 Carotid artery ultrasound ....................................................................... 10
1.6.2 Endothelial function............................................................................... 10
1.6.3 Circulating biomarkers .......................................................................... 11
1.6.4 Coronary computed tomography angiography plaque burden
versus other risk markers ....................................................................... 11
2 Aims of the thesis.......................................................................................................... 13
3 Materials and methods .................................................................................................. 15
3.1 Study group (I-IV) .............................................................................................. 15
3.2 Risk factor assessment (I-IV) ............................................................................. 16
3.3 Invasive coronary angiography (I, III-IV) ......................................................... 17
3.4 Coronary computed tomography angiography (I-IV)........................................ 17
3.4.1 Acquisition ............................................................................................. 17
3.4.2 Analysis.................................................................................................. 18
3.4.3 Coronary artery calcium score............................................................... 19
3.5 Endothelial function assessment (II) .................................................................. 20
3.6 Carotid artery ultrasound (II) ............................................................................. 20
3.7 Laboratory analyses (IV) .................................................................................... 21
3.8 Statistical analyses (I-IV) ................................................................................... 21
4 Main results ................................................................................................................... 23
5 Results and methodological discussion ........................................................................ 25
5.1 Study I ................................................................................................................. 26
5.1.1 Results .................................................................................................... 26
5.1.2 Methodological discussion .................................................................... 27
5.2 Study II ............................................................................................................... 27
5.2.1 Results .................................................................................................... 27
5.2.2 Unpublished results ............................................................................... 28
5.2.3 Methodological discussion .................................................................... 28
5.3 Study III .............................................................................................................. 30
5.3.1 Results .................................................................................................... 30
5.3.2 Methodological discussion .................................................................... 31
5.4 Study IV .............................................................................................................. 32
6
7
8
9
5.4.1 Results .................................................................................................... 32
5.4.2 Methodological discussion and additional analyses.............................. 34
General discussion ........................................................................................................ 37
6.1 Myocardial infarction and non-obstructed coronary arteries ............................. 37
6.2 Study group ......................................................................................................... 40
6.3 Takotsubo syndrome........................................................................................... 40
6.4 Cardiovascular disease risk assessment ............................................................. 41
Conclusions and final remarks ...................................................................................... 45
Acknowledgements ....................................................................................................... 47
References ..................................................................................................................... 48
LIST OF ABBREVIATIONS
AMI
Acute myocardial infarction
CAC
Coronary artery calcium
CAD
Coronary artery disease
CCA
Common carotid artery
CI
Confidence interval
CM
Contrast medium
CMR
Cardiovascular magnetic resonance
CT
Computed tomography
CTA
Computed tomography angiography
CVD
Cardiovascular disease
FRS
Framingham risk score
GFR
Glomerular filtration rate
HU
Hounsfield units
ICA
Invasive coronary angiography
IMT
Intima-media thickness
LAD
Left anterior descending artery
MB
Myocardial bridging
MINCA
Myocardial infarction and normal coronary arteries
MINOCA
Myocardial infarction and non-obstructed coronary arteries
NSTEMI
Non-ST elevation myocardial infarction
OR
Odds ratio
RHI
Reactive hyperemia index
RH-PAT
Digital reactive hyperemia peripheral arterial tonometry
ROI
Region of interest
SD
Standard deviation
SMINC
Stockholm Myocardial Infarction with Normal Coronaries
STEMI
ST elevation myocardial infarction
TS
Takotsubo syndrome
1 INTRODUCTION
Cardiovascular disease (CVD) is the number one cause of death globally. According to the
World Health Organization (WHO) [1] more than 17 million people died from CVD in 2012,
representing approximately one-third of all global deaths. An estimated 7.5 million of these
deaths were due to coronary heart disease [1]. In Sweden, almost 30 000 people suffer an
acute myocardial infarction (AMI) each year. Even though survival after AMI has increased
significantly during the last 20 years, CVD remains the leading cause of death among women
and men [2]. In a significant minority of patients with AMI, invasive coronary angiography
(ICA) shows no significant stenoses of the coronary arteries. The condition is called
myocardial infarction with non-obstructed coronary arteries (MINOCA) and is still
incompletely understood.
The studies of this thesis have been conceived and realized in a setting of rapid technological
advancements and a constantly expanding field of knowledge. Important advances in
computed tomography (CT) technology have enabled safe and accurate non-invasive imaging
of the heart and the coronary arteries [3,4]. There has been an increasing awareness of
MINOCA, for which there are currently no therapeutic guidelines [5]. Another condition that
has gained increasing attention during the last decade and for which underlying mechanisms
are largely unknown is the Takotsubo syndrome (TS), also referred to as stress-induced
cardiomyopathy or the “broken heart syndrome” [6-9].
The screening phase of the multicenter Stockholm Myocardial Infarction with Normal
Coronaries (SMINC) study [10] started in 2007. By then, coronary computed tomography
angiography (CTA) was an established method in the Radiology department in Huddinge,
Karolinska University Hospital, and the idea emerged to use coronary CTA to detect and
quantify coronary artery disease (CAD) in MINOCA patients and in controls (study I). It
soon became clear that a large proportion of MINOCA patients fulfilled the diagnostic
criteria for TS [10,11]. The frequent observation of myocardial bridging (MB) at coronary
CTA, in a clinical setting, together with the publication of research suggesting a causative
role of MB in ischemia and in TS [12,13] gave rise to the idea of studying MB in MINOCA
patients (study III). During the last decade, different biomarkers for improving CVD risk
prediction have attracted considerable attention and new knowledge has emerged regarding
the prognostic value of coronary CTA [14-16]. Studies II and IV investigate the association
between coronary CTA plaque burden and different biomarkers of CVD.
1.1
CORONARY COMPUTED TOMOGRAPHY ANGIOGRAPHY
Since 1998, when the first generations of multi-detector CT were introduced, remarkable
progress has been made in the field of CT technology. In order to visualize the heart and the
coronary arteries high spatial resolution is required as well as high temporal resolution. Most
64-detector CT scanners can provide spatial resolution of 0.4 mm and temporal resolution of
175 ms (i.e. the required time for data acquisition for one slice). Electrocardiographic (ECG)
synchronization makes it possible to obtain motion-free images despite the constant motion
1
of a beating heart. When 64-detector CT scanners were introduced, concerns were raised due
to high radiation doses differing significantly between CT systems and hospitals (5-30 mSv)
[17]. Since then, a number of dose-saving strategies such as dose-modulation, prospective
ECG-triggering, and low kilovoltage have been introduced and widely spread. Today, in
clinical practice, the radiation dose rarely exceeds 5 mSv and is often closer to 1 mSv [18].
Coronary CTA has thus become a clinically feasible non-invasive method for imaging the
coronary arteries.
Coronary CTA has a very good accuracy for diagnosing stenoses of the coronary arteries,
compared to ICA. In particular, an excellent negative predictive value has been demonstrated
in several meta-analyses [3,19]. Coronary CTA has proven highly sensitive in detecting
atherosclerotic plaques in the proximal segments of the coronary arteries, as compared to
intravascular ultrasound [20]. A histopathologic study [21] showed that coronary CTA
detected all advanced plaques, but missed many of the very early lesions (Stary I-II). In
contrast to ICA, coronary CTA not only shows the lumen of the coronary arteries, but also
the vessel wall and what lies beyond it. Coronary CTA thus allows for the detection and
characterization of coronary atherosclerotic plaques even when they do not give rise to
stenoses and whether they are calcified or not. Accordingly, coronary CTA makes it possible
to assess atherosclerotic plaques undetected by ICA. Figures 1(A) and 5 show different
plaque types, as seen on coronary CTA. The method has also proven suitable for anatomic
depiction of MB, since the epicardial fat and the myocardium as well as the vessel lumen, are
well visualized (Figures 3 and 6).
Recent clinical practice guidelines issued by the European Society of Cardiology (ESC) and
the American College of Cardiology Foundation (ACCF)/American Heart Association
(AHA) identify a number of appropriate clinical indications for coronary CTA, among them
ruling out CAD in symtomatic patients with intermediate pre-test probability of CAD or
before valve surgery in patients with low probability of CAD [22].
Coronary artery calcium (CAC) scoring dates back to the 1980s, when electron beam CT was
used to determine the amount of calcified plaque in the coronary arteries. The CAC score
developed by Agatston et al. (1990) [23] has been shown to improve CVD risk prediction
beyond commonly used risk assessment algorithms [23-26]. In 2007, when we started the
work eventually leading to this thesis, little was known on the prognostic value of coronary
CTA. Since then, however, several studies have demonstrated a prognostic value of coronary
CTA regarding CVD events, in asymptomatic and symptomatic subjects [14,15,27,28].
1.2
OTHER IMAGING MODALITIES FOR CORONARY ARTERY DISEASE
The atherosclerotic process is complex and involves a number of local and systemic factors,
including local hemodynamics, endothelial dysfunction, hyperlipidemia and widespread
inflammation [29]. Development of an atherosclerotic lesion is a gradual process. The earliest
imaging sign is diffuse intimal thickening, which may later progress to a localized atheroma
with a fibrous cap. If the lipid core enlarges and the fibrous cap thins, the lesion is called a
2
thin-cap fibroatheroma. This plaque phenotype is associated with an increased risk of plaque
rupture. At a later stage, increasing calcification leads to the formation of a fibrocalcific
plaque, which is considered to be more stable. Even though these plaques rarely cause
thrombosis and acute coronary syndromes they may cause chronic ischemic symptoms due to
gradual lumen narrowing [30].
For many years, ICA has been the gold standard for evaluating the coronary arteries. Spatial
and temporal resolution is very high (0.2 mm, 5 ms) compared to coronary CTA. ICA allows
for visualization of CAD with stenotic lesions, but often misses early CAD with preserved
vessel lumen and does not show the composition of atherosclerotic lesions. An advantage of
ICA is the possibility of performing physiological assessment through fractional flow reserve
measurements, which is the gold standard for identification of lesions that cause ischemia
[31]. During coronary artery catheterization, the use of coronary intravascular ultrasound
allows detection of early atherosclerotic intimal changes, due to its high spatial resolution
(100-150 µm) and is the gold standard for plaque quantification. Even higher spatial
resolution (10-40 µm) can be attained by optical coherence tomography, which is the only
method allowing direct visualization of the cap of the thin-cap fibroatheroma (<65 µm) and is
sensitive for detecting plaque rupture. The method uses near-infrared light, which does not
penetrate tissues as effectively as ultrasound and does therefore not allow visualization of the
outer vessel wall [30,32].
Hence, the earliest phases of CAD can be detected by intravascular ultrasound or by optical
coherence tomography. When atheromas have developed, they can be demonstrated and
characterized by coronary CTA and when the degree of stenosis increases they may be
visualized by ICA. If calcification ensues CAD can also be quantified by CAC scoring.
1.3
MYOCARDIAL INFARCTION AND NON-OBSTRUCTED CORONARY
ARTERIES
1.3.1 Acute myocardial infarction
In pathology, myocardial infarction is defined as myocardial cell death due to prolonged
ischemia (i.e. decreased blood flow and oxygen supply to the heart muscle). In an expert
consensus document from 2007 [33] AMI was defined as “evidence of myocardial necrosis in
a clinical setting consistent with acute myocardial ischemia”. Elevated blood levels of
sensitive and specific biomarkers such as cardiac troponins indicate myocardial injury and
cell death. Detection of a rise and/or fall in troponins is essential to the diagnosis of acute
myocardial infarction (AMI), since chronic elevation of troponins can occur in a number of
conditions, such as heart failure, renal failure or pulmonary embolism. Myocardial ischemia
typically gives rise to symptoms such as chest discomfort, discomfort of arm, mandibula or
epigastrium, shortness of breath or fatigue, but can also present with atypical symptoms or
even no symptoms. Electrocardiographic changes can be used to classify patients with AMI
into two groups: with ST elevation myocardial infarction (STEMI) or with non-ST elevation
3
myocardial infarction (NSTEMI), which has implications for recommended treatment
strategies [34].
Typically, AMI is considered to be a manifestation of pre-existing CAD, with plaque rupture
or ulceration and subsequent intraluminal thrombus formation in a coronary artery, leading to
decreased myocardial blood flow. This is classified as AMI type 1, according to the universal
classification of myocardial infarction. AMI type 2 corresponds to myocardial injury with
necrosis, where a condition other than CAD causes an imbalance in myocardial oxygen
supply and demand, e.g. tachycardia, hypertrophic cardiomyopathy or coronary spasm [33].
Early ICA studies from the 1980s [35,36] demonstrated the presence of significant
obstructive CAD in >97% of patients with STEMI or NSTEMI, stressing the importance of
obstructive CAD in AMI.
1.3.2 Myocardial infarction and non-obstructed coronary arteries
During the last two decades a large number of ICA studies have shown a higher prevalence of
non-obstructive CAD in patients with AMI than what was previously recognized. The
reported prevalence of MINOCA ranges between 3.5% and 18% of all AMIs [37-41]. The
different findings of these studies may partly be attributed to different definitions of normal
or non-obstructed coronary arteries. In women, the prevalence has been reported to be as high
as 33% [40,41]. The increasing awareness of MINOCA is likely largely due to the frequent
use of ICA [42] and the introduction of new sensitive troponin assays [43], leading to an
increasing number of patients with an unclear diagnosis: elevated troponins signaling
myocardial injury and cell death, but normal ICA. Figure 1 shows coronary CTA and ICA
findings of a patient with MINOCA.
Figure 1. The right coronary artery in a patient with MINOCA. Coronary CTA (A) shows a large atherosclerotic
plaque and more distally a small plaque, both with <20% stenosis. ICA (B) shows only minimal signs of
atherosclerosis. Abbreviations: MINOCA, myocardial infarction and non-obstructed coronary arteries; CTA,
computed tomography angiography; ICA, invasive coronary angiography.
4
The terminology for this condition has been inconsistent. Researchers have called it
myocardial infarction with normal coronary arteries (MINCA), MINOCA, myocardial
infarction with no significant CAD, acute coronary syndrome with culprit-free angiogram,
etc. For paper I we chose the term MINCA to stress that we only included patients with
normal or nearly-normal coronary arteries at ICA (<30% stenosis) in contrast to other studies
where patients with stenosis <50% were included. However, since then the term MINOCA
has become the most widely used, which is why this term was chosen for papers III and IV
and for this doctoral thesis.
Due to the limited investigation of this condition, MINOCA is still incompletely understood
and there are currently no therapeutic guidelines relating to the management of these patients.
MINOCA patients are more likely to be female and are younger compared to patients with
AMI and obstructive CAD [5,40]. Apart from that, most traditional CVD risk factors are
similar in the two patient groups. Approximately one-third of MINOCA patients were
categorized with STEMI and two-thirds with NSTEMI. A pooled meta-analysis of studies
investigating prognosis in patients with MINOCA [5] showed an in-hospital mortality of
0.9% and a 12-month all-cause mortality of 4.7%, both of which are lower than for patients
with AMI and obstructive CAD. Compared to patients with stable chest pain and nonobstructed coronary arteries, however, patients with MINOCA had a significantly poorer
prognosis [5].
Studies investigating the pathophysiological mechanisms of MINOCA have provided
evidence that the condition is most likely heterogeneous. Several potential underlying causes
have been proposed: CAD with rupture of a non-stenotic plaque, coronary embolism,
microvascular obstruction, disturbed endothelial function and vasospasm [44-46]. In a
clinical setting, patients with TS might be considered as a subgroup of MINOCA, since these
patients present with a similar clinical picture. A pooled meta-analysis of cardiovascular
magnetic resonance (CMR) imaging studies [5] showed subendocardial infarction in 24%,
myocarditis in 33%, evidence of TS in 18% and no detectable myocardial abnormalities in
26% of MINOCA patients. Inherited thrombotic disorders have been demonstrated in 14% of
patients with MINOCA [5]. Previous studies using intravascular ultrasound and coronary
CTA have suggested that CAD is an important underlying cause of MINOCA [45,47-49].
None of these studies, however, included a control group.
1.4
TAKOTSUBO SYNDROME
TS is an acute cardiac syndrome commonly triggered by an emotionally or physically
stressful event and with a clinical presentation indistinguishable from an AMI [6-9]. It is
characterized by transient left ventricular wall-motion abnormalities that are not restricted to
a single coronary distribution. Typically, the apical and mid-portions of the left ventricle are
hypokinetic, with hypercontraction of the basal left ventricle, leading to an unusual shape of
the left ventricle during systole, as shown in Figure 2. This shape resembles the ceramic pot
used by Japanese fishermen to trap octopuses, hence the term tako-tsubo (octopus pot) like
5
left ventricular dysfunction, which was introduced by Sato et al. (1990) [50]. It is estimated
that approximately 2% of patients presenting with a clinical picture of AMI have TS [7].
Figure 2. Illustration of a normal heart in systole (left) and the heart of a patient with TS (middle) in systole,
with the characteristic left ventricular apical hyopkinesia, giving the left ventricle the shape of an octopus pot
(right). Abbreviations: TS, Takotsubo syndrome.
During the last decade, a large number of case reports and small case series have been
published with an inconsistent terminology: TS, broken heart syndrome, apical ballooning
syndrome, stress-induced cardiomyopathy, etc. The typical patient is a postmenopausal
woman, presenting with chest pain or dyspnea after an event of acute emotional or physical
stress, with positive cardiac biomarkers or an abnormal electrocardiogram. The predominant
type of TS involves apical hypokinesia, but other types involving mid-ventricular, basal or
focal hypokinesia have also been described [6,51]. Right ventricular involvement has been
demonstrated for a significant proportion of patients with TS, in particular for those with
more severe impairment of left ventricular systolic function [52].
There is no consensus on the diagnostic criteria for TS. However, the Mayo clinic diagnostic
criteria (modified in 2008) [8] are widely used. These four criteria should be fulfilled:
(1) Transient hypokinesis, akinesis, or dyskinesis in the left ventricular mid segments
with or without apical involvement; the regional wall motion abnormalities extend
6
beyond a single epicardial vascular distribution; a stressful trigger is often, but not
always present.
(2) Absence of obstructive CAD or angiographic evidence of acute plaque rupture.
(3) New electrocardiographic abnormalities (either ST-segment elevation and/or T-wave
inversion) or modest elevation in cardiac troponin.
(4) Absence of pheochromocytoma and myocarditis.
These criteria have been questioned for several reasons:
(1) There are forms of TS that do not involve the mid segments of the left ventricle [51].
(2) TS may develop in patients with obstructive CAD [6,8,53]. In some cases, the stress
of an AMI may even be the trigger of TS [54].
(4) TS may coexist (or be triggered by) pheochromocytoma or myocarditis [53,55].
In a majority of patients with TS the condition is preceded by an emotionally or physically
stressful event, but in a significant minority there is no evident trigger [6,9]. Potential
physical stressors include non-cardiac surgery, critical illness with or without sepsis and
subarachnoid hemorrhage. TS is probably still underrecognized and often misdiagnosed, due
to the wide spectrum of the condition, its varied presentation and the diagnostic difficulties.
Left ventricular systolic function usually improves within weeks and total recovery
eventually occurs in most patients. However, there are complications in the acute phase of the
disease, such as fulminant heart failure and cardiogenic shock, arrhythmias and cardiac arrest
[6,56,57]. Conflicting results have been reported regarding short- and long-term prognosis.
Early studies reporting a favorable prognosis, similar to that of the general population, may
have underestimated the risk of early and late complications [6,8]. More recent studies have
reported short- and long-term mortality rates similar to those of patients suffering from AMI
and a recurrence rate of approximately 10% [6,7,58,59].
Causes of TS are still incompletely understood, even though a number of pathophysiologic
mechanisms have been proposed. A potential role of cathecholamine excess has been
suggested, as well as an involvement of the “brain-heart connection” [6,7,60]. Other proposed
mechanisms (not mutually exclusive) include coronary vasospasm, microvascular
dysfunction, left ventricular outflow tract obstruction, effects of epinephrine on the β2adrenoreceptor and catecholamine-mediated myocardial stunning [7,8,53,60,61]. A recent
publication [13] suggested MB of the left anterior descending artery (LAD) might be one
underlying cause of TS, after demonstrating an elevated prevalence of MB in patients with
TS compared to controls.
1.4.1 Imaging in Takotsubo syndrome
ICA with left ventricular angiography is central in the diagnosis of TS, to visualize the
coronary arteries (and exclude an acute coronary syndrome) as well as the function and shape
of the left ventricle. Early CMR imaging may be valuable for assessing patients with TS,
7
since functional analysis (long and short axis cine) allows for evaluation of wall-motion
abnormalities and assessment of cardiac function. In addition, T2-weighted sequences allow
for the determination of presence and extent of myocardial edema. Late gadolinium
enhancement sequences allow for detection of myocardial necrosis and/or fibrosis, as well as
differentiation between myocardial infarction and myocarditis [62,63]. Typical CMR findings
in TS are mid-apical myocardial edema with corresponding regional wall-motion
abnormalities and apical ballooning, but with no or only subtle late gadolinium enhancement
[63]. Echocardiography allows for detection of wall-motion abnormalities, assessment of
systolic function and evaluation of left ventricular outflow tract obstruction, which is
common in patients with TS [61].
1.5
MYOCARDIAL BRIDGING
Myocardial bridging (MB) refers to a condition where a segment of an epicardial coronary
artery (the “tunneled segment”) is covered by cardiac muscle fibres (the “myocardial
bridge”), thus coursing through the myocardium for a variable distance (Figure 3).
Figure 3. MB visualized by coronary CTA. Curved multiplanar reformation of the LAD of a healthy volunteer
with MB with full encasement (A). In (B) MB with partial encasement is shown, in a patient with TS.
Abbreviations: MB, myocardial bridging; CTA, computed tomography angiography; LAD, left anterior
descending artery; TS, Takotsubo syndrome.
MB is mostly found in the LAD, even though it has been reported in other coronary vessels
[64,65]. MB can be classified as MB with full encasement or MB with partial encasement,
which is further discussed in the methods section of study III. The different types of MB are
illustrated in detail in Figure 6. The reported prevalence of MB in the general population
varies greatly and its significance has been widely debated. Autopsy studies have
demonstrated a prevalence of MB ranging between 15% and 85% [66]. The detection rate at
ICA is much lower: in general between 0.5% and 4.5% [65]. Detection of MB at ICA mainly
8
relies on the presence of systolic compression of the tunneled arterial segment, also referred
to as “milking effect”, for indirect visualization of the MB. However, a short or superficial
MB might not cause any detectable systolic compression, which probably accounts for the
low prevalence reported at ICA compared to autopsy. A number of research groups have
examined the feasibility of coronary CTA for detecting MB and reported a prevalence
between 26% and 58% [13,65,67,68], which is similar to numbers reported from autopsy
studies. A very good correlation has been reported between coronary CTA and coronary
intravascular ultrasound for detecting MB of the LAD [69]. For the majority of people with
MB, the condition can be considered to be a benign congenital variant, unlikely to ever lead
to any complications [70,71]. However, serious adverse events such as myocardial ischemia,
AMI, arrhythmia and sudden death have been reported [12,66].
There are two main mechanisms by which MB is believed to cause ischemia: the first is
acceleration of the development of atherosclerosis proximal to the MB and the second is
dynamic compression of the intramural segment, resulting in hemodynamic alterations that
might affect both systolic and diastolic blood flow. Several conditions might facilitate or
trigger ischemia related to MB, among them tachycardia and increased myocardial
contractility during stress or exercise, increasing left ventricular diastolic dysfunction
associated with ageing, development of left ventricular hypertrophy or endothelial
dysfunction stimulating coronary vasospasm or thrombus formation [12,72,73].
1.6
CARDIOVASCULAR DISEASE RISK ASSESSMENT
Considering CVD worldwide, different levels of risk factors can be identified. At a
fundamental level we find determinants such as globalization, urbanization, poverty and an
ageing population [1]. These determinants have an impact on behavioural risk factors, among
them physical inactivity, an unhealthy diet and use of tobacco and alcohol. The behavioural
risk factors, in turn, manifest themselves in individuals through for instance obesity,
hypertension, hyperlipidemia or diabetes mellitus, all of which are measurable risk markers
for CVD or cerebrovascular disease events. Interventions aiming to decrease the prevalence
of CVD may thus operate at a population level (strategies to reduce the harmful use of
tobacco and alcohol, healthy school meals for children, etc.) or at an individual level [1]. In
primary prevention at the individual level, modifying behavioural risk factors (smoking
cessation, increased physical activity, etc.) and for some individuals drug treatment of for
instance hypertension or hyperlipidemia has been shown to reduce the risk of CVD.
To guide primary prevention interventions it is essential to identify individuals at increased
risk of future CVD events. CVD risk assessment is traditionally based on risk factors such as
smoking, hypertension, hyperlipidemia and diabetes mellitus or risk assessment algorithms,
such as the Framingham, the SCORE or the Reynolds risk scores [74-78]. Although these
risk factors and risk scores have proven useful on a population level, they do not identify
individuals who will eventually develop CVD [79,80]. As many as half of individuals who
suffer a CVD event have only one or none of the traditional risk factors [79]. A number of
additional risk markers that might improve the predictive power of risk scores have been
9
proposed, among them the CAC score, carotid artery ultrasound measurement of intimamedia thickness (IMT) or plaque detection, measures of peripheral endothelial function and
different categories of circulating biomarkers [25,26,79,81,82].
1.6.1 Carotid artery ultrasound
The carotid arteries are easily accessible for non-invasive ultrasound imaging that allows for
detection of plaque and measurement of IMT. IMT is measured between the luminal-intimal
and the medial-adventitial interfaces of the carotid artery wall. A good accuracy has been
demonstrated for IMT measurements of the far wall of the common carotid artery, when
compared to histologic specimens [83]. The IMT measure reflects intimal thickening, but also
medial smooth muscle hyperplasia and/or hyperthrophy. Carotid artery ultrasound findings
have implications for the risk of cerebrovascular disease. Carotid artery IMT has also proven
to be associated with prevalent CVD and to predict CVD events [81,84], motivating its use as
a surrogate marker for CAD. According to the American College of Cardiology
(ACC)/American Heart Association (AHA) guidelines from 2010 [25], measurement of IMT
can be useful for risk assessment for intermediate-risk individuals, but according to updated
guidelines from 2013 [85], routine measurement of IMT is no longer recommended for
prediction of CVD events. Recently, it has been shown that carotid plaque predicts CAD
events more accurately than IMT [82,83].
1.6.2 Endothelial function
A functioning endothelium is a fundamental element of vascular health. The endothelium is a
highly selective barrier between the circulating blood and the vascular wall and it is also
metabolically active [86]. Nitric oxide is central in the normal functioning of the
endothelium, primarily because it promotes vasodilation, but also due to other effects, such as
inhibition of platelet aggregation and leucocyte adherence [87]. In endothelial dysfunction,
there is impaired production of vasoprotective endothelial mediators and excessive
production of pro-oxidant, pro-inflammatory and pro-thrombotic mediators, thus promoting
inflammation and atherothrombosis [88]. Endothelial dysfunction is considered to be an early
marker and a precursor of atherothrombotic disease. It is closely related to CVD risk factors
and predicts CVD events in patients with or without established CVD [87,89,90]. Several
classes of CVD drugs have been shown to lead to improved endothelial function and serial
measurements of endothelial function have been proposed in order to monitor therapeutic
effects [87,88]. There are several methods for evaluating endothelial function. Quantitative
ICA with intracoronary acetylcholine is a direct method for assessing endothelium-dependent
vasodilation of the coronary arteries [88]. Since endothelial dysfunction is a systemic
disorder, measurement of peripheral endothelial function can be used as a surrogate for
coronary endothelial function [91-93]. Flow-mediated dilation of the brachial artery is the
most established method for non-invasive assessment of endothelial function. Digital reactive
hyperemia peripheral arterial tonometry (RH-PAT) is a recently introduced method for
assessing peripheral endothelial function that has the advantage of being easily accessible,
10
largely operator-independent and highly reproducible [94]. An association between RH-PAT
and several CVD risk factors has been demonstrated [95].
1.6.3 Circulating biomarkers
A large number of circulating biomarkers, representing different pathophysiologial pathways,
have been proposed to predict CAD. Markers of inflammation include C-reactive protein and
several interleukins. Lipoprotein(a) and adiponectin are markers of lipid metabolism. Cystatin
C is an accurate marker of kidney function. There are numerous other circulating biomarkers,
reflecting for instance oxidative stress, thrombosis and endothelial dysfunction [79,96-98].
Even though circulating biomarkers have received considerable attention during the last
decade, their potential role in CVD risk assessment is not yet established [79,96]. An
advantage of circulating biomarkers is that some may provide information early and some
late in the CVD process, whereas imaging biomarkers may detect subclinical disease, but not
earlier stages of CVD. Genetic markers, on the other hand, can be evaluated long before
development of CVD and provide information on an individual’s susceptibility, but fail to tell
us whether subclinical disease has developed or not. To date, studies investigating the
performance of risk prediction based on combinations of multiple circulating biomarkers,
have shown, at best, modest improvement in risk prediction [79].
1.6.4 Coronary computed tomography angiography plaque burden versus
other risk markers
A prognostic value regarding CVD events has been demonstrated for coronary CTA, as
previously discussed. IMT and RH-PAT are supposed to reflect the presence and extent of
CAD, but only a few studies have compared these measures with coronary CTA plaque
burden [99]. Even fewer have focused on clinically healthy subjects, with predominantly lowto-intermediate risk profiles [100]. An association between circulating biomarkers and CAD
determined by coronary CTA has only been studied by a small number of research groups,
with varying results [101-104]. Imai et al. [103] reported that cystatin C was associated with
the extent of non-obstructive CAD at coronary CTA. Studies by Bamberg et al. [101] and
Matsuda et al. [104] showed an association between levels of adiponectin and coronary CTA
plaque burden, whereas Caselli et al. [102] did not find an association beween circulating
adiponectin and coronary CTA findings. Lipoprotein(a) was not associated with CAD
determined by coronary CTA [101,102].
11
2 AIMS OF THE THESIS
There were two major aims of this thesis. The first aim was to investigate the underlying
mechanisms of MINOCA by studying coronary CTA plaque burden (study I) and myocardial
bridging (study III). The second aim was to examine the association between coronary CTA
plaque burden and other risk markers of CVD: carotid artery IMT and peripheral endothelial
function (study II) and circulating biomarkers (study IV).
The specific objectives of each study were:
I.
To investigate whether patients with MINOCA had a greater coronary plaque
burden determined by coronary CTA than a control group matched by age and
gender.
II.
To examine the ability of carotid artery IMT and endothelial function
determined by RH-PAT to predict coronary CTA plaque burden in clinically
healthy subjects.
III.
To investigate the prevalence of MB in patients with MINOCA and TS
compared to healthy volunteers and to compare the MB depiction rates of
coronary CTA and ICA.
IV.
To investigate whether circulating markers of lipid metabolism, inflammation
and kidney function could predict non-obstructive CAD determined by
coronary CTA.
By pursuing these aims we hope to increase the understanding of the underlying
pathophysiological mechanisms of MINOCA and to some degree TS, thereby contributing to
the development of diagnostic strategies, that may guide therapeutic management and
hopefully ultimately improve prognosis for these patients. We also hope to provide new
information in the important field of CVD risk assessment in low-to-intermediate risk groups.
13
3 MATERIALS AND METHODS
The studies of this thesis are a subset of the multicentre Stockholm Myocardial Infarction
with Normal Coronaries (SMINC) study [10,11]. The studies conform to the principles of the
Declaration of Helsinki and were approved by the Regional Ethical Review Board in
Stockholm and by the Radiation Protection Committee of the Karolinska University Hospital.
Written informed consent was obtained from all patients and healthy volunteers. Study
participants were assigned study identification numbers and records were kept confidential at
the Cardiology Unit at Karolinska University Hospital in Solna, Sweden.
3.1
STUDY GROUP (I-IV)
The study participants (MINOCA patients and controls) of the four studies of this thesis form
a subgroup of the study participants of the SMINC study [10,11].
Study I, III and IV: In the first step, patients with MINOCA were screened for the SMINC
study at five different coronary care units in the Stockholm metropolitan area between June
2007 and May 2011, as previously described [10]. Patients were eligible to take part in the
study if they were between 35 and 70 years old, fulfilled the criteria for acute myocardial
infarction (AMI) according to the universal definition of AMI [33] and underwent ICA
showing no or minimal signs of atherosclerosis (defined as the presence of plaque discernible
on ICA, but no stenosis exceeding 30% by visual estimation). All patients also underwent
cardiovascular magnetic resonance (CMR) imaging at a median of 12 days after hospital
admission. Exclusion criteria were myocarditis (based on CMR findings or a clinical
diagnosis), a clinical diagnosis of pulmonary embolism, non-sinus rhythm on admission,
pacemaker use, and a patient history of structural or coronary heart disease, chronic
obstructive lung disease or renal disease [10]. After patient inclusion, the coronary
angiograms as well as the clinical AMI diagnoses were re-evaluated by an additional
independent investigator. Patients with TS according to the Mayo clinic diagnostic criteria [8]
were considered part of the MINOCA group and thus not excluded.
A control group, matched by age and gender, was recruited using a registry comprising all
Stockholm residents. Randomly selected persons of matching age and gender to MINOCA
patients received a letter of invitation followed by a telephone call. Subjects who were willing
to participate and who had no known CVD underwent an exercise stress test. If the test was
normal they were invited to take part in the study.
In the second step, MINOCA patients and controls of the SMINC study were recruited to the
coronary CTA substudy. Additional exclusion criteria for the coronary CTA study were age
under 45, an estimated glomerular filtration rate (GFR) < 50ml/min/1.73m2 (based on serum
creatinine), previous adverse reaction to an iodine-based contrast agent and an irregular heart
rate (jeopardizing the diagnostic quality of the CT scan). Patient selection and study flow are
shown in Figure 4.
15
Figure 4. Flow chart illustrating study group selection for the SMINC coronary CTA substudy, during a 4-year
screening period in the Stockholm metropolitan area.
Study II: The study group of study II was composed of the 58 clinically healthy controls of
the group described above.
3.2
RISK FACTOR ASSESSMENT (I-IV)
Current smoking was defined as active regular use and previous smoking as a history of
regular use. Hypertension and hyperlipidemia were defined as previously diagnosed and
16
medically treated. A positive family history of CAD was defined as CVD event in a parent or
sibling before 65 years of age. Risk assessment was performed according to the Framingham
risk algorithm [74].
Study I-III: Diabetes mellitus was defined as previously diagnosed.
Study IV: An oral glucose tolerance test was performed, with measurement of the plasma
glucose concentration 2 hours after ingestion of 75 g glucose dissolved in 150 ml of water.
Impaired glucose tolerance was defined as a plasma glucose concentration ≥7.8 mmol/L and
diabetes mellitus as a plasma glucose concentration ≥11.1 mmol/L.
3.3
INVASIVE CORONARY ANGIOGRAPHY (I, III-IV)
Study I, III and IV: For patients with MINOCA, ICA was performed at the time of initial
hospital admission, according to clinical routines and using standard techniques, including a
transfemoral or transradial approach and a minimum of two different projections of each
coronary artery. ICAs were evaluated using the modified American Heart Association
17 segment classification [106]. Clinical reassessments of all ICAs were performed at a later
time by an additional independent investigator.
Study III: Additional joint readings of ICAs were performed by two experienced
interventional cardiologists focusing specifically on the presence of MB. They assessed
indirect signs of MB: systolic compression of the tunneled segment (by visual estimation) or
a localized change in direction of the vessel course towards the ventricle (the “step down-step
up” phenomenon).
3.4
CORONARY COMPUTED TOMOGRAPHY ANGIOGRAPHY (I-IV)
All study participants underwent coronary CTA at Karolinska University Hospital in
Huddinge, Sweden. For MINOCA patients, the CTA examinations were performed between
3 and 6 months after the acute event.
3.4.1 Acquisition
Examinations were performed on a 64-slice computed tomography scanner (LightSpeed VCT
XT; GE Healthcare, Milwaukee, WI, USA). A prospectively ECG-triggered scan protocol
was used: detector configuration 64 x 0.625 mm, rotation time 350 ms (temporal resolution
175 ms), tube potential 120 kVp, tube current 450-650 mA (according to patient size). The
scans were performed in diastole, in general at 70-75% of the RR interval, with 0-200 ms
padding (depending on heart rate and variability). Two patients, however, were examined
using a retrospectively ECG-gated scan protocol (dose modulation, 100 kVp, 200-450 mA),
as a slightly irregular heart rate might otherwise have compromised image quality. The
intravenous contrast medium (CM) used was the iso-osmolar iodixanol 320 mg I/mL
(Visipaque, GE Healthcare, Stockholm, Sweden), which was individually dosed based on
bodyweight (400 mg I/kg, 75-100 mL iodixanol). The CM was administered using a dualhead injector (Medrad, Stellant Dual Head Injector, Pittsburgh, PA, USA) and a triple-phase
17
protocol (1, CM bolus injection; 2, mixture of saline and CM; 3, saline bolus injection). Bolus
tracking software was used (SmartPrep, GE Healthcare, Milwaukee, WI, USA) in order to
synchronize image acquisition with optimal CM opacification of the coronary arteries. In the
absence of contraindications and depending on the initial heart rate, patients received
metoprolol (25-100 mg) per os prior to the examination. Patients also received sublingual
nitroglycerine (0.4 mg) 4 minutes before the scan.
Figure 5. Different plaque types, as seen by coronary CTA. A non-calcified plaque is shown in longitudinal and
cross section (A and B). The degree of stenosis was 20-50%. A large mixed plaque is shown in longitudinal
section (C) and in cross section at the level of non-calcified (D) and calcified (E) components. A large calcified
plaque is shown to the right. (F and G). The mixed and calcified plaques (C to G) were both eccentric in location
and the degree of stenosis was <20%. Abbreviations: CTA, computed tomography angiography.
3.4.2 Analysis
Study I-IV: The coronary CTA examinations were analyzed independently by two
experienced readers (level 2 according to ACCF/AHA levels of competence [107]) who were
blinded to all clinical information. Subsequent joint readings were performed in order to
reach a consensus. Coronary CTA analysis was performed using the CardIQ Xpress software
on the Advantage Workstation 4.4 (GE Healthcare, Milwaukee, WI, USA). Axial source
images, multiplanar and curved multiplanar reformations as well as thin-slab maximum
intensity projections were used. The optimal image display setting was chosen on an
individual basis (in general at a window width of 800-1000 Hounsfield units (HU) and a level
of 100-200 HU). Coronary arteries were subdivided into 17 segments, according to the
modified American Heart Association classification [106]. Initially, each segment was
assessed regarding image quality and evaluability. Segments were considered non-evaluable
if artifacts prevented reliable assessment of the lumen or the vessel wall (e.g. due to motion,
image noise or heavy calcification). Secondly, each segment was visually evaluated with
regard to the presence of atherosclerotic plaque (with or without stenosis) (a), plaque size (b)
18
and composition (c). A plaque was defined as any structure, discernible in at least two planes,
within or adjacent to the vessel lumen, which could be clearly separated from the lumen and
from adjacent soft tissue.
a.
Lesions were quantified for stenosis by visual estimation, comparing the
minimal lumen of the stenotic segment with the lumen of the adjacent proximal unaffected
segment, and expressed in terms of diameter stenosis: <20%, 20-50% or ≥50%. Although
arbitrary, the 50% stenosis limit is the most widely used to differentiate between nonobstructive and obstructive lesions.
b.
The size of the atherosclerotic plaque was determined by measuring the length
of the plaque on longitudinal sections and arbitrarily classified as: small (<4mm), medium
(4-8 mm) or large (≥ 8mm).
c.
Plaque composition was visually assessed based on the presence or absence of
calcified elements: non-calcified coronary artery plaque, mixed coronary artery plaque or
calcified coronary artery plaque, the latter with ≥ 50% calcified tissue [21].
In the case of more than one atherosclerotic plaque in a single segment, the greatest degree of
stenosis, the largest plaque size and the most pronounced calcification was considered.
Figure 5 shows examples of different plaque types, as seen by coronary CTA.
Study III: In addition to the coronary CTA evaluations described above, an additional joint
reading was performed, with focus on MB. The presence of MB was noted and described as
MB with partial encasement or MB with full encasement [108]. Hence, MB was defined as a
coronary artery segment in direct contact with the myocardium (partial encasement) or
completely surrounded by myocardium (full encasement), instead of being surrounded by
epicardial fat as it normally would. Figure 6 illustrates the types of MB. In the case of one
MB involving two adjacent coronary artery segments, the MB was attributed to the more
proximal segment. The length of the MB was measured on multiplanar reformations and for
full encasement, the thickness of the overlying bridge was measured in a short axis view
perpendicular to the tunneled segment.
3.4.3 Coronary artery calcium score
For the CAC score, a non-enhanced scan was performed, using a prospectively ECGtriggered scan protocol: detector configuration 64 x 0.625 mm, rotation time 350 ms, tube
potential 120 kVp, tube current 200 mA. The CAC score was calculated using semiautomatic software (SmartScore 4.0, GE Healthcare, Milwaukee, WI, USA) on the
Advantage Workstation 4.4 (GE Healthcare, Milwaukee, WI, USA). The total calcium
burden of the coronary arteries was reported in terms of AJ-130 score, based on the scoring
algorithm of Agatston et al. (1990) [23].
19
Figure 6. Classification of MB, as seen on coronary CTA. Bottom images show short axis views of the LAD,
and its relation to surrounding structures. Top images show the corresponding short axis views with annotations.
Usually, the LAD is located above the interventricular gorge, completely surrounded by epicardial fat (A). In
MB with partial encasement a vessel segment is in direct contact with left ventricular myocardium (B). In MB
with full encasement the vessel segment is completely surrounded by myocardium, either with an unmeasurable
overlying muscle bridge (C) or with measurable overlying muscle (D). The LAD is marked with a pink circle
and the myocardium-epicardium interface is marked with a blue line. RV denotes the right ventricle; LV, the left
ventricle; S, the interventricular septum; M, left ventricular myocardium; E, epicardial fat and L, lung tissue.
Abbreviations: MB, myocardial bridging; CTA, computed tomography angiography; LAD, left anterior
descending artery.
3.5
ENDOTHELIAL FUNCTION ASSESSMENT (II)
Study II: Peripheral endothelial function was assessed using the digital RH-PAT device
EndoPAT (Itamar Medical Ltd., Caesarea, Israel), according to the method described by
Hamburg et al. [109]. The method involves measuring pulse wave amplitude of the fingertip
(namely the PAT signal), using a fingertip probe, at rest and after a 5-minute occlusion of the
brachial artery, using a standard blood pressure cuff inflated to supra-systolic pressure. When
the cuff is deflated, the surge of blood causes an endothelium-dependent flow mediated
dilatation leading to reactive hyperemia and an increase in the PAT signal amplitude. A
fingertip probe is also placed on the index finger of the contralateral side for internal control.
As a measure of reactive hyperemia the reactive hyperemia index (RHI) was calculated, using
dedicated software, as the ratio between the post-occlusion and the pre-occlusion PAT signal
amplitude divided by the corresponding ratio of the contralateral side, to adjust for any
spontaneous or systemic alterations of vascular tone. The test was performed in a thermoneutral and quiet surrounding, after fasting and avoiding pre-test smoking or consumption of
caffeine.
3.6
CAROTID ARTERY ULTRASOUND (II)
Study II: Two-dimensional images of the left and right common carotid artery (CCA) were
acquired using an ultrasound scanner equipped with a 12 MHz transducer (Vivid 7, General
Electric Company, Horten, Norway). From each CCA a long-axis cine loop of three beats and
20
three diastolic images at the time of the electrocardiographic R-wave were digitally stored for
off-line analysis. The IMT of the CCA far wall was measured in three diastolic images using
semi-automatic IMT analysis software (General Electric Company). A 10 mm region of
interest (ROI) was manually placed starting 1 cm proximal to the carotid bulb. The luminalintimal and the medial-adventitial interfaces were identified automatically. In case of
suboptimal tracking, the ROI was adjusted or another diastolic frame was chosen. Manual
correction was not performed. The left and right IMT results were calculated as the mean of
three semi-automatic measurements, and finally the average of these two results was
obtained. All examinations were performed and interpreted by the same experienced
ultrasonographer, who did not have access to the coronary CTA findings.
3.7
LABORATORY ANALYSES (IV)
Study IV: For MINOCA patients, data from routine clinical chemistry analyses performed at
admission were acquired from medical records: fasting plasma triglycerides, fasting plasma
LDL cholesterol, fasting plasma HDL cholesterol [11]. Additional blood samples were
acquired at a 3-month follow-up visit. For the control group all blood sampling was
performed at a single study visit. Plasma from all study participants was stored at −80°C for
subsequent analyses. Plasma high-sensitivity C-reactive protein, lipoprotein(a) and cystatin C
were analyzed using immuno-turbidimetric assays and adiponectin using enzyme-linked
immunosorbent assay. Analyses were performed using accredited methods at an accredited
laboratory (the Karolinska University Laboratory). Estimated GFR was calculated based on
serum creatinine.
3.8
STATISTICAL ANALYSES (I-IV)
Data were characterized using descriptive statistics and assessed for normal distribution either
graphically (study II) or using the Shapiro-Wilk test (studies I, III, IV). Categorical variables
are presented as absolute value and percentage and were compared using the chi-square test
or Fisher’s exact two-sided test. Continuous variables following a normal distribution are
presented as mean ± standard deviation (SD) and non-normally distributed continuous
variables are presented as median (range or interquartile range), as specified. Student’s t-test
(for normally distributed variables) and the Mann-Whitney U-test (in the case of non-normal
distribution) were used in order to test differences between two independent groups. The
Pearson test (for normally distributed variables) or the Spearman test (in the case of nonnormal distribution) were used to test correlation between continuous variables. The 5% level
of significance was considered (two-tailed p<0.05).
Study I and II: All analyses were carried out using the SAS system (the SAS system for
Windows 9.3, SAS Institute Inc., Cary, NC, USA).
Study III: Analyses were carried out using IBM SPSS Statistics version 22.0.
Study IV: In addition to analyses described above, univariable logistic regression analyses
were performed and variables with p≤0.1 (sex, body mass index, systolic blood pressure,
21
HDL cholesterol, adiponectin, lipoprotein(a) and cystatin C) were incorporated into a
multivariable logistic regression analysis with a forward stepwise method for variable
selection (criterion for entering variable: p<0.05). Additional variables incorporated into the
multivariable regression model were: smoking, subgroup (MINOCA, control) and estimated
GFR. Levels of plasma biomarkers were dichotomized according to the median. Results of
logistic regression analyses are presented as odds ratio (OR) with a 95% confidence interval
(CI). All analyses were carried out using IBM SPSS Statistics version 22.0.
22
4 MAIN RESULTS
The principal findings were:
Study I: MINOCA patients did not have more CAD than healthy controls, matched by age
and gender. A large proportion of MINOCA patients had no signs of CAD at coronary CTA.
Study II: Neither assessment of endothelial function by the RH-PAT method nor ultrasound
measurement of carotid artery IMT predicts coronary CTA plaque burden in clinically
healthy subjects with low-to-intermediate CVD risk.
Study III: MB was frequent, with a similar prevalence in MINOCA patients, patients with
TS and controls. Coronary CTA detects MB at a significantly higher rate than ICA.
Study IV: Plasma levels of cystatin C were associated with non-obstructive CAD at coronary
CTA, independently of GFR, in subjects with low-to-intermediate CVD risk.
23
5 RESULTS AND METHODOLOGICAL DISCUSSION
Clinical characteristics of the study group (MINOCA and controls) are shown in Table 1.
Table
Clinical
characteristics
Table
1. 1.
Clinical
characteristics
MINOCA patients!!
n=57!
Age (years)
60 ± 5!
Female!
42 (74%)!
Smoking habits!
!
Never smoked
29 (52%)
Previous smoker
17 (30%)
Current smoker
10 (18%)
Family history of CAD 16 (28%)!
BMI (kg/m2)
25.8 ± 3
*
Diabetes mellitus
1 (2%)!
Treated hypertension
19 (33%)!
Treated hyperlipidemia 8 (14%)!
SBP (mm Hg)
144 (129-161)
GFR (ml/min/1.73m2) 77.9 ± 12
FRS
low
22 (39%)
intermediate
22 (39%)
high
13 (23%)
Controls!
n=58!
61 ± 6
39 (67%)!
!
33 (57%)
23 (40%)
2 (3%)
14 (24%)!
25.8 ± 3
0 (0%)!
6 (10%)!
3 (5%)!
129 (115-140)
78.1 ± 13
P
ns
ns
0.04
ns
ns
ns
0.003
ns
<0.001
ns
ns
31 (53%)
18 (31%)
9 (16%)
Abbreviations:
MINOCA,
myocardial
infarction
and non-obstructed
coronary arteries;
CAD,
coronary
artery
Abbreviations:
MINOCA,
myocardial
infarction
and non-obstructed
coronary
arteries;
CAD,
disease;
BMI,
body disease;
mass index;
SBP,
systolic
pressure;
estimated
filtration
rate based on
coronary
artery
BMI,
body
massblood
index;
SBP, GFR,
systolic
blood glomerular
pressure; GFR,
estimated
serum
creatinine;
FRS, Framingham
Score (lowRisk
risk ≤10%,
intermediate
risk 10-20%,
high risk
>20%
glomerular
filtration
rate; FRS, Risk
Framingham
Score (low
risk ≤10%,
intermediate
risk
1010-year
Data >20%
are presented
as risk).
mean!Data
± standard
deviation, as
absolute
(percentage)
or median
20%, risk).
high risk
10-year
are presented
meanvalue
± standard
deviation,
absolute
(interquartile
range). P-values
<0.05(interquartile
are given, obtained
through
Student´s
the Mann-Whitney
U-test or
value (percentage)
or median
range).
P-values
<0.05t-test,
are given,
obtained through
*
theStudent´s
chi-squaret-test,
test, as
appropriate.
Diabetes
mellitus
waschi-square
defined as previously
diagnosed.
the
Mann-Whitney
U-test
or the
test, as appropriate.
*
Diabetes mellitus was defined as previously diagnosed.
Baseline characteristics were similar in MINOCA patients and controls, apart from current
smoking and treated hypertension, which were more common in the MINOCA group. At
presentation, 56 MINOCA patients had no signs of heart failure (Killip class 1) and only one
had heart failure (Killip class 2). Signs of acute ischemia (ST-T changes or left bundle branch
block) on admission ECG were present in 31 (54%) out of whom 10 had ST elevations. The
median (interquartile range) peak troponin level was 18 (7-43) times the upper limit of
normal. Myocardial infarction was detected by CMR in 11 (19%) patients. The criteria for TS
were fulfilled in 15 (26%).
25
5.1
STUDY I
5.1.1 Results
Coronary CTA examinations of 57 MINOCA patients and 58 controls were analyzed. Fifteen
(1.9 %) individual segments in the MINOCA group and 11(1.4 %) in the control group were
non-evaluable and excluded from analyses. There were 765 evaluable segments in the
MINOCA group and 781 in the control group.
Coronary CTA plaque burden analyses are presented in Tables 2 and 3, on per patient and per
segment basis, comparing the MINOCA group with the control group.
Table
2.1.Coronary
Table
CoronaryCTA
CTAplaque
plaqueburden
burdenper
per patient
patient
Severity of CAD*
All CAD†
Calcium score (AJ-130)
No CAD
Stenosis <20%
Stenosis 20-50%
Stenosis ≥50%
0 segments
1 segments
2 segments
3 segments
4 segments
5 segments
6 segments
7 segments
8 segments
9 segments
10 segments
MINOCA patients
n=57
24 (42%)
22 (39%)
11 (19%)
0 (0%)
24 (42%)
14 (25%)
8 (14%)
4 (7%)
2 (4%)
3 (5%)
0 (0%)
0 (0%)
2 (4%)
0 (0%)
0 (0%)
6 (0-778)
Controls
n=58
25 (43%)
23 (40%)
9 (16%)
1 (2%)
25 (43%)
10 (17%)
12 (21%)
6 (10%)
0 (0%)
1 (2%)
0 (0%)
0 (0%)
1 (2%)
1 (2%)
2 (3%)
8 (0-1882)
P
ns
ns
ns
Abbreviations: CTA, computed tomography angiography; MINOCA, myocardial infarction and non-obstructed
Abbreviations:
CTA,
computed
tomography
angiography;
MINOCA,
myocardial
infarction
and non-obstructed
coronary arteries;
CAD,
coronary
artery disease;
ns, non significant.
Values
are presented
as absolute
value
†
coronary
arteries;
CAD, coronary
arterytodisease;
ns, non-significant.
Values
are presented
as absolute
(percentage)
or median
(range). *refers
the maximum
diameter stenosis;
refers
to obstructive
and non-value
obstructive or
CAD.
(percentage)
median (range). *refers to the maximum diameter stenosis; †refers to obstructive and nonobstructive CAD.
On a per patient level there were no statistically significant differences in severity or extent of
CAD. Twenty-four (42%) MINOCA patients and 25 (43%) controls had no signs of CAD.
When analyzing the data on a per segment level, however, there were statistically significant
differences regarding degree of stenosis, plaque size and plaque composition. Compared to
controls, MINOCA patients had fewer segments with stenosis ≥20%. They also had fewer
large and mixed type plaques. The CAC scores within each group were diverse, but no
significant differences were found between the groups.
26
There were no differences regarding coronary CTA plaque burden when subgroups of
MINOCA patients were compared with controls (MINOCA patients with AMI detected by
CMR, with ST elevations or with TS).
Table 3. Coronary CTA plaque burden per segment
Table 2. Coronary CTA plaque burden per segment
Severity of CAD
Plaque size
Plaque composition
No CAD
Stenosis <20%
Stenosis 20-50%
Stenosis ≥50%
No CAD
Small
Medium
Large
No CAD
Non-calcified plaque
Mixed plaque
Calcified plaque
MINOCA patients
765 segments
684 (89%)
68 (9%)
13 (2%)
0 (0%)
684 (89%)
41 (5%)
24 (3%)
16 (2%)
684 (89%)
10 (1%)
10 (1%)
61 (8%)
Controls
781 segments
687 (88%)
58 (7%)
35 (4%)
1 (0.1%)
687 (88%)
31 (4%)
29 (4%)
34 (4%)
687 (88%)
10 (1%)
28 (4%)
56 (7%)
P
<0.01*
0.04*
0.04*
Abbreviations: CTA, computed tomography angiography; MINOCA, myocardial infarction and non-obstructed
* non-obstructed
Abbreviations:
CTA,
computed
MINOCA,
myocardial
infarction and
coronary arteries;
CAD,
coronarytomography
artery disease;angiography;
Values are presented
as absolute
value (percentage).
P-values
apply toarteries;
the comparison
the four artery
categories
in theValues
two columns
to the leftas
ofabsolute
the value,value
using (percentage).
the chi-square *P-values
coronary
CAD, of
coronary
disease;
are presented
test.
apply to the comparison of the four categories in the two columns to the left of the value, using the chi-square
test.
5.1.2 Methodological discussion
For MINOCA patients, stenosis >30% at ICA was an exclusion criterion. For ethical reasons,
healthy volunteers did not undergo ICA and individuals with stenoses were thus not
excluded, which might explain the fact that there were more segments with stenosis ≥20% in
the control group than in the MINOCA patient group. Still, on per patient basis there were
roughly equal numbers of MINOCA patients (19%) and controls (18%) with stenosis ≥20%.
5.2
STUDY II
5.2.1 Results
All 58 volunteers, free from clinical CVD, underwent coronary CTA with subsequent plaque
burden analysis, as well as testing for peripheral endothelial function by the RH-PAT
method. Ultrasound of the carotid arteries and measurement of IMT were successfully
performed in 57 study participants. A non-enhanced CT scan for CAC scoring was obtained
in 51 subjects.
According to the Framingham risk algorithm [74] 31 study participants (53%) had a 10-year
risk of 10% or less, whilst 18 (31%) had a 10-20% risk and 9 (16%) had an elevated risk
exceeding 20%. Only three subjects were on lipid-lowering medication.
27
Coronary CTA demonstrated normal coronary arteries in 25 (43%) study participants. Thirtytwo (55%) subjects had non-obstructive CAD and only one (2%) had obstructive CAD. The
median (range) of the RHI was 2.2 (1.4-4.9) and of the IMT 0.70 mm (0.49 mm-0.99 mm).
The CAC score ranged from 0 to 1882, with a median of 4.
When comparing the groups with and without evidence of CAD at coronary CTA, no
statistically significant differences were found concerning RHI or IMT (Table 4). Nor was
there any correlation between the number of diseased segments on coronary CTA and RHI
(rs=0.13) or IMT (rs=0.098) (Figure 7). Not surprisingly, there was a statistically significant
difference in CAC score when comparing the groups with and without CAD demonstrated by
coronary CTA (p<0.001) (Table 4). Similarly, there was a strong correlation between the
CAC score and the number of diseased coronary segments (rs=0.86, p<0.001) (Figure 7).
Table 4. Comparison between groups with and without CAD (as determined by coronary
Table
CTA)4. Comparison between groups with and without CAD at coronary CTA (clinically healthy subjects)
RHI
IMT (mm)
CAC
No CAD
n=25
2.2 (1.4-4.9)
0.70 (0.53-0.99)
0 (0-4)
CAD
n=33
2.1 (1.4-3.6)
0.71 (0.49-0.99)
36 (0-1882)
P
ns
ns
<0.001
Abbreviations:CAD,
CAD,coronary
coronaryartery
arterydisease;
disease;CTA,
CTA,computed
computedtomography
tomographyangiography;
angiography;
Abbreviations:
RHI, reactive
RHI, reactive hyperemia index; IMT, carotid artery intima-media thickness; CAC, coronary
hyperemia index; IMT, carotid artery intima-media thickness; CAC, coronary artery calcium score; ns, non
artery calcium score; ns, non significant. Values are presented as median (range).
significant. Values are presented as median (range).
5.2.2 Unpublished results
Study I showed that MINOCA patients and controls were similar with respect to coronary
plaque burden and a previous SMINC study found no difference in RHI or IMT between the
two groups [11]. We therefore found it reasonable to conduct additional analyses with the two
groups put together into one (n=115). The median (range) of the RHI was 2.2 (1.0-4.9) and of
the IMT 0.70 mm (0.49 mm-1.0 mm). The CAC ranged from 0 to 1882, with a median of 1.5.
All results were similar to those obtained when studying only the healthy volunteers
(Table 5). Hence, there was no association between the presence or extent of CAD and RHI
or IMT, but a strong correlation between the number of diseased segments and the CAC score
(rs=0.80, p<0.001).
5.2.3 Methodological discussion
Ultrasound examinations of the carotid arteries focused only on IMT measurement of the
CCA, while additional analyses such as plaque detection might have yielded different results.
Several studies published after the initiation of this study have shown that plaque assessment
of the carotid arteries, including the carotid bulb and the internal carotid artery, predicted
CAD and outcome better than IMT alone [82,83]. IMT does not only reflect subclinical
28
atherosclerosis, but may be influenced by for instance hypertension. In addition,
atherosclerotic plaques tend to form in the carotid bulb and in the internal carotid artery,
which may also explain why IMT of the CCA has less predictive value [83]. In our study,
even though the focus of the ultrasound examination was the CCA, plaques in the carotid
bulb were evident for three subjects, all with CAD at coronary CTA.
Brolin et al. BMC Cardiovascular Disorders (2015) 15:63
Page 5 of 7
have affected the results. Since coronary CTA permits
visualization of non-stenotic lesions that might not be
visible on conventional coronary angiography, more patients with limited disease are likely to end up classified
as having CAD when coronary CTA is used instead of
conventional coronary angiography. Rohani et al., on the
contrary, reported a lack of association between endothelial function, expressed in terms of flow mediated dilation
of the brachial artery, and the angiographic extent of
coronary stenoses in patients with severe CAD [28]. A
possible implication of their findings might be that
endothelial function testing is of limited value for patients with pre-existing severe CAD. It has indeed been
B
suggested that endothelial function testing, and in particular testing involving the microvasculature as is the
case for the RH-PAT method, might be more suitable
for subjects with a low prevalence of risk factors [29].
Our findings for the present study group, however, do
not support this hypothesis.
With respect to the lack of association between IMT
and the presence and extent of CAD, our findings differ
from those of some previous studies [21, 30, 31]. These
studies, however, have mainly focused on patient groups
with a more pronounced risk profile and with more
C
severe CAD, as determined by cardiac CT or by conventional coronary angiography, compared to the present
study group. Our findings are partly supported by a recent study by Schroeder et al., who found no significant
relationship between carotid artery ultrasound findings
and coronary CTA plaque burden [22]. They did however
demonstrate a higher sensitivity of carotid ultrasound
Figure
7. Correlation
between
of diseased
compared
to CAC
for number
coronary
CTA plaque burden.
segments,
determined
by coronary
CTA, and patients
RHI
Their as
study
focused
on asymtomatic
referred
risk
a cardiovascular
(A),for
IMT
(B)stratification,
and CAC (C). Awith
significant
correlation risk profile
a only
prevalence
to that of the present
wasand
found
for CACof(rsCAD
=0.86,similar
p<0.001)
study. TheCTA,
increasing
evidence
that IMT of the common
Abbreviations:
computed
tomography
carotid artery
may not
be a good
marker
angiography;
RHI, reactive
hyperemia
index;
IMT, for coronary
atherosclerosis
also
supports
the
results
of the present
intima-media thickness; CAC, coronary artery
study [14].
calcium.
The finding of a strong correlation between CAC and
Fig. 1 Correlation between number of diseased segments, as
the presence and extent of CAD as determined by cordetermined by coronary CTA, and RHI (a), IMT (b) and CAC (c). The
onary CTA is not surprising, since the two measures
larger point (0, 0) in (c) corresponds to 23 subjects. Abbreviations:
correspond to two aspects of the atherosclerotic process
CTA, computed tomography angiography; RHI, reactive hyperemia
of the coronary vascular bed.
index; IMT, intima-media thickness; CAC, coronary artery calcium
An important limitation of the present study is the
limited sample size. Thus, lack of significant differences
and correlation may be caused by lack of power to depatients with no CAD [27]. The RHI was markedly worse tect such differences. Since no correlations were found,
in the patients with CAD of that study compared to the additional statistical analyses to determine29sensitivity, speRHI of the patients with CAD of the present study. This cificity and predictive values were not performed. The
difference might reflect a significantly greater risk factor study group mainly reflects a northern European ethnic
A
Table 5. Comparison between groups with and without CAD (as determined by coronary
Table
CTA)5. Comparison between groups with and without CAD at coronary CTA (MINOCA and controls)
RHI
IMT (mm)
CAC
No CAD
n=49
2.3 (1.3-4.9)
0.70 (0.49-0.99)
0 (0-4)
CAD
n=66
2.1 (1.0-3.6)
0.70 (0.49-1.0)
35 (0-1882)
P
ns
ns
<0.001
Abbreviations:CAD,
CAD,coronary
coronaryartery
artery
disease;
CTA,
computed
tomography
angiography;
Abbreviations:
disease;
CTA,
computed
tomography
angiography;
RHI, reactive
RHI, reactive hyperemia index; IMT, carotid artery intima-media thickness; CAC, coronary
hyperemia index; IMT, carotid artery intima-media thickness; CAC, coronary artery calcium score; ns, non
artery calcium score; ns, non significant. Values are presented as median (range).
significant. Values are presented as median (range).
!
The choice of RH-PAT, in order to ensure a highly reproducible method, instead of for
instance brachial artery flow mediated dilation might also have affected the results. It has
been shown that different risk factors affect endothelial function of the microvasculature
(measured with RH-PAT) and the conduit vessels (measured with brachial artery flow
mediated dilation) differently [110]. Another factor that may have influenced our results is
the fact that endothelial function was determined with one single test. Repeated measurement
has been proposed in order to augment accuracy, since endothelial function is a dynamic
process, which can be transiently affected by a variety of factors [111].
5.3
STUDY III
5.3.1 Results
All study participants underwent coronary CTA and all MINOCA patients also underwent
ICA, as previously described.
In total, 54 (47%) study participants had evidence of MB. MB with full encasement was
found in 33 (29%) and MB with partial encasement in 21 (18%). A majority of MBs were
located in segment 7 (mid LAD). Only three study participants had MB (detected by coronary
CTA) in vessels other than the LAD. Due to the vast majority of MBs being located in the
LAD, only these were included in further analyses.
Table 6 shows the prevalence, type and location of MB in the different study groups. There
were no statistically significant differences between MINOCA patients or the TS subgroup
and the control group regarding the prevalence or type of MB. Nor were there any differences
between the groups regarding the location, the length or the thickness of MB. Nine MINOCA
patients had no signs of CAD or MB. Out of these 9 patients, 3 were diagnosed with TS.
In the MINOCA patient group, the depiction rate of MB was compared for coronary CTA
and ICA. Coronary CTA demonstrated MB in 28 (49%) patients, while ICA showed indirect
signs of MB in 13 (23%). The difference in depiction rate between the two methods was
statistically significant (p<0.01). In four patients MB was suspected at ICA, but not
confirmed by coronary CTA. ICA thus “correctly” detected MB in 9 (16%) patients. When
comparing the cases of MB detected by both modalities with those detected by coronary CTA
30
but missed by ICA, there were no statistically significant differences regarding localization,
type, length or thickness of MB (Table 7).
Table
TS patients
patientsand
andininmatched
matchedcontrols.
controls.
Table6.6.Myocardial
Myocardialbridging
bridgingin
in MINOCA
MINOCA and
and TS
MB (coronary CTA)
Type of MB
Partial encasement
Full encasement
Location of MB
Proximal LAD
Mid LAD
Distal LAD
Length of MB (mm)
Thickness of MB (mm)*
*
MINOCA patients
n=57
28 (49%)
TS subgroup
n=15
8 (53%)
Control group
n=58
26 (45%)
11 (19%)
17 (30%)
2 (13%)
6 (40%)
10 (17%)
16 (28%)
1 (2%)
23 (40%)
4 (7%)
14.5 (2.0-51.0)
0 (0-4.8)
1 (7%)
7 (47%)
0 (0%)
7.3 (2.0-10.9)
0 (0-2.0)
0
20 (34%)
6 (10%)
13.5 (3.8-45.0)
0 (0-5.0)
P
ns
ns
ns
ns
ns
in the case of full encasement
Data are presented as absolute value (percentage) or median (range). Abbreviations: MINOCA, myocardial
Data are presented as absolute value (percentage) or median (range). Abbreviations: MINOCA, myocardial
infarction
and non-obstructed coronary arteries; TS, takotsubo syndrome; MB, myocardial bridging; CTA,
infarction and non-obstructed coronary arteries; TS, takotsubo syndrome; MB, myocardial bridging; CTA,
computed
artery;ns,
ns,non-significant.
non-significant.The
Thechi-square
chi-square
computedtomography
tomographyangiography;
angiography;LAD,
LAD, left
left anterior
anterior descending
descending artery;
test
was
used
for
comparing
the
MINOCA
group
and
the
TS
subgroup
respectively
with
the
control
group.
test was used for comparing the MINOCA group
subgroup respectively with the control group.
*
Neithertest
testdemonstrated
demonstratedany
anystatistically
statisticallysignificant
significant differences.
differences. in the case of full encasement.
Neither
!
5.3.2 Methodological discussion
The prevalence of MB according to ICA was higher than in most previous studies, but similar
to that demonstrated by Leschka et al [65]. The high depiction rate at ICA can probably be
attributed to the thorough review of the angiogram, with specific focus on detecting MB. In
the present study MB was suspected at ICA in 4 cases, where no MB was demonstrated by
coronary CTA, which occurred only once in the previous study, where automated quantitative
coronary angiography software was used to measure systolic diameter narrowing. The use of
quantitative analysis at ICA might thus pose an advantage in order to quantify systolic
compression as well as to avoid over-diagnosing MB [65].
Assessment of dynamic compression was not performed at coronary CTA, since images were
acquired only in diastole, in order to minimize radiation exposure. To adequately assess
dynamic compression it would have been necessary to perform a retrospectively gated
examination with full radiation dose throughout the cardiac cycle, resulting in significantly
higher radiation exposure [65].
Only MB of the LAD was considered, due to the facts that MB in other vessels is rare and
that dynamic compression mostly occurs in MB of the LAD [67]. The degree of dynamic
compression has been shown to correlate with the risk of ischemia [66]. Thus it seems very
unlikely that MBs in other vessels would be clinically significant.
31
Table 3. Myocardial bridging detected by coronary computed tomography angiography and unidentified or
Table
7. MB
coronary
CTA and unidentified or identified by ICA.
identified
bydetected
invasiveby
coronary
angiography.
Type of MB
Partial encasement
Full encasement
Location of MB
Proximal LAD
Mid LAD
Distal LAD
Length of MB (mm)
Thickness of MB (mm)*
MB unidentified by ICA
n=19
MB identified by ICA
n=9
7 (37%)
12 (63%)
4 (44%)
5 (56%)
1 (5%)
17 (89%)
1 (5%)
15.1 ± 13
0 (0-4.8)
1 (11%)
7 (78%)
1 (11%)
18.5 ± 8
0 (0-1.5)
P
ns
ns
ns
ns
Dataare
arepresented
presentedasasabsolute
absolutevalue
value(percentage),
(percentage), mean
mean ±
± SD
SD or
Data
or median
median (range).
(range).Abbreviations:
Abbreviations:MB,
MB,
myocardial bridging; ICA, invasive coronary angiography; LAD, left anterior descending artery; ns, nonmyocardial bridging; CTA, computed tomography angiography; ICA, invasive coronary angiography; LAD, left
significant. *in the case of full encasement
anterior descending artery; ns, non-significant. *in the case of full encasement
5.4
STUDY IV
5.4.1 Results
Clinical characteristics, including laboratory results, of study participants without and with
CAD are shown in Table 8. Approximately half of the study group had low CVD risk
according to the Framingham risk algorithm, and only one-fifth had high CVD risk. Kidney
function was in general normal or only mildly reduced.
Coronary CTA plaque burden is described in Tables 2 and 3. Briefly, there were no signs of
CAD in 49 (43%) study participants and CAD in at least one coronary segment in 66 (57%).
Among study participants with CAD, 44 (67%) had a maximum degree of stenosis <20%,
21 (32%) had a maximum degree of stenosis 20-50% and only one had a stenosis ≥50%. On a
per segment level, out of all plaques, 20 (11%) were non-calcified, 38 (22%) were mixed and
117 (67%) were calcified. Among subjects with CAD, only 12 (18%) had exclusively noncalcified plaques.
As previously described, coronary CTA plaque burden was similar in MINOCA patients and
controls (results study I) as were most baseline characteristics (Table 1). For the purpose of
the current study it was crucial to find out whether the two subsets of the study group were
also similar regarding the plasma biomarkers to be analyzed. LDL cholesterol was the only
parameter showing a statistically significant difference in concentration between MINOCA
patients and controls. The mean concentration of LDL cholesterol (±SD) was
3.2 (±0.9) mmol/L in MINOCA patients and 3.7 (±0.9) mmol/L in controls (p=0.008).
Compared to the group without CAD at coronary CTA, the study participants with CAD had
lower plasma concentrations of adiponectin and higher plasma concentrations of
lipoprotein(a) and cystatin C. They were also more likely to be male and had higher body
mass index (Table 8). Results of univariable logistic regression analyses are shown in
32
Table 9. Concentrations of circulating biomarkers adiponectin, lipoprotein(a) and cystatin C
were dichotomized, according to the median. The median was chosen as a cut-point, since
concentrations were largely within normal ranges and there were no predefined cut-off levels.
Table
Table8.8.Clinical
Clinicalcharacteristics
characteristicsofofsubjects
subjectswithout
withoutand
andwith
withCAD,
CAD,determined
determined by
by coronary
coronary CTA.
CTA.
Variable
Age (years)
Female
Smoking habits
Never smoked
Previous smoker
Current smoker
Family history of CAD
Treated hypertension
Treated hyperlipidemia
BMI (kg/m2)
OGTT
NGT
IGT
Diabetes mellitus
FRS
low
intermediate
high
SBP (mm Hg)
GFR (ml/min/1.73m2)
hs-CRP (mg/L)
HDL cholesterol (mmol/L)
LDL cholesterol (mmol/L)
Triglycerides (mmol/L)
Adiponectin (mg/L)
Lipoprotein(a) (nmol/L)
Cystatin C (mg/L)
No CAD
(n=49)
60 ± 6
40 (82%)
CAD
(n=66)
61 ± 7
41 (62%)
29 (59%)
14 (29%)
6 (12%)
10 (20%)
10 (20%)
4 (8%)
24.9 ± 3
33 (53%)
26 (40%)
6 (9%)
20 (30%)
15 (23%)
7 (11%)
26.5 ± 4
32 (67%)
12 (25%)
4 (8%)
44 (72%)
13 (21%)
4 (7%)
24 (49%)
17 (35%)
8 (16%)
131 (118-149)
80.0 ± 12
0.82 (0.39-1.85)
1.6 (1.2-2.1)
3.4 ± 1.0
0.94 (0.73-1.35)
10.5 (7.8-18.3)
16.5 (5.0-37.0)
0.79 (0.72-0.88)
29 (44%)
23 (35%)
14 (21%)
140 (125-155)
76.5 ± 13
1.1 (0.56-2.20)
1.3 (1.1-1.9)
3.5 ± 0.87
1.00 (0.88-1.30)
7.9 (6.0-13.0)
24.0 (14.0-76.0)
0.86 (0.77-0.95)
P
ns
0.023
ns
ns
ns
ns
0.026
ns
ns
ns
ns
ns
ns
ns
ns
0.048
0.027
0.012
Abbreviations:
angiography;
BMI,
body mass
Abbreviations:CAD,
CAD,coronary
coronaryartery
arterydisease;
disease;CTA,
BMI,computed
body masstomography
index; OGTT,
oral glucose
tolerance
test; NGT,
index;
OGTT,
oral
glucose
tolerance
test;
NGT,
normal
glucose
tolerance;
IGT,
impaired
glucose
tolerance;
normal glucose tolerance; IGT, impaired glucose tolerance; FRS, Framingham Risk Score (low risk
≤10%,
FRS,
Framingham
Score
(low
≤10%,
intermediate
risksystolic
10-20%,
highpressure;
risk >20%
10-year
risk); SBP,
intermediate
risk Risk
10-20%,
high
riskrisk
>20%
10-year
risk); SBP,
blood
GFR,
estimated
glomerular
hs-CRP,
high-sensitivity
C-reactive
protein;
highcreatinine;
density lipoprotein
systolic
bloodfiltration
pressure;rate;
GFR,
estimated
glomerular filtration
rate
based HDL,
on serum
hs-CRP, cholesterol;
highLDL,
low
density
lipoprotein
cholesterol
ns,
non
significant.
Data
are
presented
as
mean
±
standard
deviation,
sensitivity C-reactive protein; HDL, high density lipoprotein cholesterol; LDL, low density lipoprotein
absolute value
(percentage)
or Data
median
are given,
obtained
Student´s
cholesterol;
ns, non
significant.
are(interquartile
presented as range).
mean ±P-values
standard<0.05
deviation,
absolute
value through
(percentage)
or
t-test, the Mann-Whitney U-test or the chi-square test, as appropriate.
median (interquartile range). P-values <0.05 are given, obtained through Student´s t-test, the Mann-Whitney Utest or the chi-square test, as appropriate.
The final adjusted multivariable model incorporated all risk markers with p≤0.1 in
univariable analysis as well as smoking habits, subgroup (MINOCA, control) and estimated
GFR (Table 9). The remaining independent variables were male sex (OR 2.77, 95% CI 1.11,
6.90) and cystatin C (OR 2.50, 95% CI 1.12, 5.59).
33
When examining the correlation between the number of diseased segments at coronary CTA
and plasma biomarker levels, a statistically significant correlation was found only for cystatin
C (rs=0.25, p<0.01).
Table
9. 2.
Univariable
andand
multivariable
logistic
regression
models
for the
variable
CAD,CAD,
determined
by by
Table
Univariable
multivariable
logistic
regression
models
for outcome
the outcome
variable
determined
coronary
CTA.
coronary
CTA
Logistic regression model
Univariable
analysesa
Multivariable
adjusted modelb
Risk marker
Male sex
BMI
SBP
HDL cholesterol
Adiponectin
Lipoprotein(a)
Cystatin C
Male sex
Cystatin C
OR
2.71
1.13
1.02
0.51
0.48
1.87
2.56
2.77
2.50
95% CI
1.13, 6.52
1.01, 1.25
1.00, 1.03
0.25, 1.06
0.22, 1.03
0.87, 4.00
1.19, 5.49
1.11, 6.90
1.12, 5.59
Adiponectin,
Lipoprotein(a)
Cystatin
C were
dichotomized
equally
groups,
according
Adiponectin,
Lipoprotein(a)
andand
Cystatin
C were
dichotomized
into into
two two
equally
sizedsized
groups,
according
to theto the
a
b
a For univariable analyses results are presented for all variables with p ≤0.1.
b
median.
Stepwise
multivariable
median. For univariable analyses results are presented for all variables with p ≤0.1. Stepwise multivariable
model incorporating: sex, BMI, SBP, HDL cholesterol, Adiponectin, Lipoprotein(a), Cystatin C, subgroup
model incorporating: sex, BMI, SBP, HDL cholesterol, Adiponectin, Lipoprotein(a), Cystatin C, subgroup
(MINOCA-control), smoking and estimated glomerular filtration rate. OR, odds ratio; CI, confidence interval of
(MINOCA-control), smoking and estimated glomerular filtration rate. OR, odds ratio; CI, confidence interval of
the odds ratio; BMI, body mass index; SBP, systolic blood pressure; HDL, high-density lipoprotein cholesterol.
the odds ratio; BMI, body mass index; SBP, systolic blood pressure; HDL, high-density lipoprotein cholesterol.
!
5.4.2 Methodological discussion and additional analyses
Estimated GFR was based on serum creatinine rather than cystatin C, in order to be able to
adjust for GFR in a mutivariable regression analysis incorporating cystatin C as an
independent variable.
There was no difference between the groups without and with CAD with regard to the
Framingham risk score (FRS) [74]. However, since the FRS is primarily a tool for primary
prevention it might not be appropriate for patients suffering from MINOCA. We therefore
performed separate analyses for the control group to investigate whether the FRS was
associated with presence or extent of CAD at coronary CTA. Non-parametric testing showed
no significant difference in FRS between the groups without and with CAD (p=0.2). There
was however a statistically significant correlation between FRS and number of diseased
segments (rs=0.26, p=0.05) (Figure 8). Significant correlations were also found between FRS
and adiponectin (rs=-0.41, p<0.01) and cystatin C (rs=0.44, p<0.01) but not between FRS and
lipoprotein(a).
34
A
!
B
!
C
!
Figure 8. Correlation between the FRS and the number of diseased segments, as determined by coronary CTA
(rs=0.26) (A), plasma adiponectin (rs=-0.41) (B) and plasma cystatin C (rs=0.44) (C). Abbreviations: FRS,
Framingham risk score; CTA, computed tomography angiography.
35
36
6 GENERAL DISCUSSION
6.1
MYOCARDIAL INFARCTION AND NON-OBSTRUCTED CORONARY
ARTERIES
Study I of this thesis was the first to use coronary CTA to analyze coronary plaque burden in
patients with MINOCA, with a control group matched by age and gender for comparison. We
aimed to find out whether MINOCA patients had more CAD than controls, which might
indicate that CAD is an underlying mechanism of MINOCA. Instead, we found a similar
prevalence of CAD in patients with MINOCA and controls. In addition, a large proportion of
MINOCA patients (42%) did not have any signs of CAD at coronary CTA, which strongly
suggests there are other underlying causes for a significant number of MINOCA patients.
This finding was further supported by the fact that MINOCA patients had a lower rate of
large size and mixed type coronary artery plaques; plaque characteristics that have been
shown to imply a more vulnerable plaque type, more prone to rupture [112-114].
Coronary CTA has proven to be highly sensitive for detecting atherosclerotic plaques of the
coronary arteries [3,4,20]. Still, there are limitations to its spatial resolution, which makes it
uncertain to assess very small plaques, in particular if located distally in the coronary arteries.
Hence, early CAD, with subtle changes of the vessel wall, as well as distal lesions might
remain undetected by coronary CTA [21,115]. In this thesis, a semi-quantitative method that
relied on visual assessment was used for coronary CTA plaque analysis. It could be debated
whether an automated, non-user dependent method would have been more reliable. There
were however no robust and validated automated methods available when this study was
conducted. In contrast, semi-quantitative methods for plaque assessment have been described
in several studies, demonstrating small inter- and intraobserver variability [116,117]. In
clinical practice and in most studies, coronary CTA analysis still relies largely on visual
assessment. In a recent study [118] an automated coronary plaque quantification algorithm
was compared to expert and non-expert semi-automatic plaque assessment, with intravascular
ultrasound as the gold standard. The performance of the fully automated method was good
and comparable to non-expert readers but inferior to expert readers. There was a high
agreement between expert readings and intravascular ultrasound measurements for all
parameters analyzed [118]. Only segments with sufficient CTA and intravascular ultrasound
image quality were included in the study and performance of an automated method in the real
world scenario with motion, calcifications and other artefacts still needs to be established.
Substantial efforts have been made to search for imaging markers of plaque instability that
might allow us to identify vulnerable plaques before they rupture and thereby possibly
prevent plaque rupture and AMI or sudden death. Pathologic studies have identified the thincap fibroatheroma as a vulnerable plaque type with prognostic implications. Studies using
intravascular ultrasound and optical coherence tomography have reported that a large necrotic
core, a thin fibrous cap and positive vascular remodeling are signs of plaque instability [119].
Several coronary CTA plaque characteristics have been associated with a vulnerable plaque
type: positive vascular remodeling, low-attenuation plaque (<30 HU), large plaque area and
37
non-calcified or mixed plaque types [112-114,120]. There are several limitations when
attempting to identify vulnerable plaques by coronary CTA. The limited spatial resolution
prevents identification of thin-cap fibroatheromas and makes the measurement of positive
remodeling imprecise. Plaque HU measurements do not only depend on plaque composition,
but are also affected by technical and imaging parameters. Therefore, we chose to limit
plaque characteristic assessment to: degree of stenosis, length of plaque and non-calcified/
mixed/ calcified plaque.
An important strength of the studies in this doctoral thesis is the high diagnostic quality of the
coronary CTA scans, with very few non-evaluable coronary segments. Examinations were
performed by dedicated radiographers using a 64-detector scanner, in a centre with several
years’ experience of coronary CTA [121]. Evaluation of coronary CTA scans was performed
by two experienced readers, separately and in consensus.
Our findings partly contradict findings of a previous study [49], in which only 16% of
MINOCA patients did not have signs of CAD when examined with coronary CTA. However,
this difference can probably be explained by the difference in inclusion criteria. In the
previous study, patients with <50% angiographic diameter stenosis were included, whereas in
our study a more rigorous definition was used, including only patients with no or minimal
signs of atherosclerosis (<30%). In addition, only patients with evidence of myocardial
infarction on CMR were included in the earlier study [49], whereas we also included patients
with smaller myocardial infarctions that were proven by biochemical markers but not
detectable by CMR. In fact, a recently published meta-analysis [5] demonstrated that
subendocardial infarct can only be detected by CMR in one-fourth of patients with
MINOCA, which illustrates that the study group of the earlier study [49] might not have been
representative of all MINOCA patients. In our study, one-fifth of MINOCA patients had
myocardial infarction at CMR, which is similar to findings of the recent meta-analysis [5].
In an intravascular ultrasound study [45], plaque rupture and ulceration was found in 38% of
women with MINOCA, suggesting CAD to be an underlying cause for a large proportion of
MINOCA patients. Patients with plaque disruption had a higher degree of stenosis (median
degree of stenosis 40%) compared to patients without plaque disruption. Interestingly, none
of the patients with a normal ICA had plaque disruption [45]. Consequently, it seems likely
that the frequency of plaque disruption would be smaller in our study group, with less severe
CAD. It has been shown that culprit lesions in AMI are frequently not stenotic. Pathology
studies of acute coronary death have demonstrated that the average culprit lesion has a
diameter stenosis of approximately 50%, and rarely below 30% [122]. This further supports
our conclusion that, for the present study group, CAD is most likely not an important cause of
MINOCA.
To the best of my knowledge, to date our study is the largest study of its kind. Still, the
sample size was limited. Therefore, the lack of significant differences between the two groups
may be caused by a lack of power to detect such differences. However, the data shows a
tendency towards more severe CAD in the control group compared to the MINOCA-group,
38
rather than the other way round. In order to estimate statistical power, one could consider the
per segment analysis (n=765+781), a binomial endpoint (segment with or without CAD) and
the 5% level of significance. Anticipating a CAD prevalence of 10% in the control group and
15% in the MINOCA group would yield a power of 85%.
We found that coronary CTA detected significantly more MB than ICA. In our study MB
was detected by coronary CTA in 47% of the study participants, which is similar to the
prevalence (58%) found in a previous coronary CTA study [67], and of the same magnitude
as frequencies found in previous autopsy studies [66]. However, most previous studies using
coronary CTA showed a lower prevalence of MB [66]. This may in part reflect the superior
image quality of newer generations of CT scanners and the fact that evaluations of coronary
CTA scans were performed by experienced readers focusing specifically on the presence of
MB, and probably also the inclusion of both MB with partial encasement and MB with full
encasement. Kim et al. [67] found partial encasement in 19% and full encasement in 39% of
subjects, which is similar to our findings (partial encasement 18%; full encasement 29%).
Study III of this thesis is the first to use coronary CTA in order to compare the prevalence of
MB in MINOCA patients with a matched control group. MB was a frequent finding, with a
similar prevalence in MINOCA patients and in controls. Therefore, it is unlikely that MB is
an important cause of MINOCA. However, in order to find out whether MB might have been
an underlying mechanism of the AMI for some individuals, additional testing of the
physiological significance of the MB would be necessary, e.g. dobutamine stress coronary
angiography or intracoronary acetylcholine testing for coronary spasm and endothelial
dysfunction [72,73,123].
The studies of this thesis cannot explain the pathophysiological mechanisms of MINOCA.
However, our findings suggest that neither CAD nor MB is a frequent cause. Together with
findings of earlier and recent studies [5,10,11] our findings support that patients with
MINOCA compose a heterogeneous group, with a variety of underlying causes. For the
MINOCA patients with CAD this may, at least for some patients, have been the cause of the
myocardial infarction. However, CAD is a less likely cause in the patients with no or minimal
angiographic stenosis. Plaque disruption with transient thrombus formation might be one
plausible mechanism and vasospasm another [5,45]. There was a higher frequency of
smoking in the MINOCA group, which may increase the risk of thrombosis as well as
vasospasm. Even though MB is probably not a frequent cause of MINOCA, it should be
considered when other potential causes of the MI have been ruled out. TS was a relatively
frequent finding (26% of MINOCA patients) and the prevalence might have been
underestimated, which will be further discussed. Six out of 57 MINOCA patients showed no
signs of CAD, did not have MB and did not get a diagnosis of TS, which means that there is
still a proportion of patients where other potential causes of the MI are likely.
39
6.2
STUDY GROUP
An important strength of studies I and III is the well defined MINOCA group, where all
patients underwent CMR imaging in order to exclude those with myocarditis, which
commonly mimics AMI [124]. Limitations in the sensitivity of CMR to detect myocarditis
may have led to false-negative cases [125]. However, they were most likely few, since most
patients presenting with typical clinical features of myocarditis did not undergo ICA and were
thus not eligible for the SMINC study in the first place. During the screening period, a
majority (64%) of all MINOCA patients in the Stockholm metropolitan area were included in
the SMINC study. Of the 100 MINOCA patients included in the SMINC study, only 57
underwent coronary CTA and a possible selection bias must be considered. However, when
comparing baseline characteristics of these patients with the 43 MINOCA patients who did
not participate in the coronary CTA study, there were no significant differences. Considering
my research questions, I do not believe that there was any significant selection bias affecting
the results. The upper age limit of 70 years may have reduced the proportion of MINOCA
patients with TS, since patients with TS are typically older than other MINOCA patients [7].
The control group was composed of randomly selected volunteers matched by age and
gender, free from clinical CVD. The participation rate was approximately 50%. An important
question is whether they truly represented the population from which they were recruited. In
an ongoing large Swedish multi-centre study of cardiopulmonary disease [126], a pilot study
was performed, where randomly selected subjects (50-64 years of age) were invited to take
part in a number of examinations, among them coronary CTA. The participation rate was
lower in areas of low compared to high socio-economic status (40% vs 68%). Moreover, the
pilot study demonstrated significant differences in health related to socio-economic status,
with a higher prevalence of risk factors in areas of low socio-economic status [126]. Hence,
there seems to be a risk of recruiting controls with a lower prevalence of risk factors than the
general population. This effect would probably be even more important since controls of the
SMINC study were excluded if they had an abnormal exercise stress test. Since we found in
study I that MINOCA patients did not have more CAD than controls, these concerns
probably did not compromise our results.
6.3
TAKOTSUBO SYNDROME
For patients with TS we found a prevalence of MB similar to that of the control group. This
contradicts findings of a previous study [13] that reported a significantly higher prevalence of
MB in patients with TS compared to controls, by coronary CTA as well as by ICA. However,
consistent with the findings of our study, a recent ICA study [127] showed a similar
prevalence of MB in patients with TS and in a control group. Our findings support that MB is
not a cause of TS. In fact, it does not seem plausible that MB would cause the wall-motion
abnormality seen in TS, since it does not follow the coronary distribution.
A limitation in studies I and III might be the use of the Mayo clinic criteria [8] for diagnosis
of TS. As discussed in the introduction section there are several arguments against these
40
criteria and updated diagnostic criteria have been proposed [7, 53]. Some of these arguments
do not apply to the present study group, who by definition in the study protocol did not have
obstructive CAD or myocarditis. However, in addition to the 15 MINOCA patients who
fulfilled the diagnostic criteria for TS, there were 10 MINOCA patients for whom there was a
clinical suspicion of TS, but the left ventricular wall motion abnormality was atypical and did
not fulfil the first Mayo criterion. In some patients, a typical wall motion abnormality might
have been missed, if left ventricular angiography was not performed at the time of ICA and if
CMR was performed too late, when the wall motion abnormality had partly resolved.
Another cause for a “missed” diagnosis might be if the patient had a milder form of disease in
the TS spectrum.
6.4
CARDIOVASCULAR DISEASE RISK ASSESSMENT
When interpreting the results of studies II and IV it is important to consider the present study
group. The prevalence of CVD risk factors was relatively low, most study participants had
low-to-intermediate CVD risk by the Framingham risk algorithm [74] and a majority were
women. This contrasts to most other similar studies, where the CVD risk profile was much
more pronounced and a larger proportion of study participants were men. In our study group,
coronary CTA plaque burden was relatively mild, with almost exclusively non-obstructive
CAD. IMT and RH-PAT values were in general in the normal range, as were the levels of
circulating biomarkers. It has been shown that subjects with early stage non-obstructive CAD
are at increased risk of CVD events compared to subjects without CAD at coronary CTA
[14,15,27,28], which is why we found it important to study this predominantly female low-tointermediate risk group, with no or non-obstructive CAD.
IMT was similar in the groups with and without evidence of CAD by coronary CTA. There
was no correlation between IMT and the number of diseased coronary artery segments. Our
findings differ from findings of previous studies that focused mainly on patients with more
pronounced CVD risk and more severe CAD as determined by coronary CTA or by ICA
[99,128,129]. However, our findings are partly supported by a recent study [100] that did not
find a significant relationship between carotid artery ultrasound findings and coronary CTA
plaque burden. Their study group was composed of asymtomatic subjects referred for risk
stratification, with a CVD risk profile and a prevalence of CAD similar to that of our study
[100]. As previously mentioned, it has been shown that carotid plaque detection is more
accurate than IMT for predicting CVD events [82].
Peripheral endothelial function by the RH-PAT method was similar in subjects with and
without evidence of CAD by coronary CTA and there was no correlation between RH-PAT
and the number of diseased coronary artery segments. To the best of my knowledge, study II
is the first publication comparing endothelial function by the RH-PAT method with coronary
CTA plaque burden. A previous study [130] of 140 women with chest pain who underwent
RH-PAT testing as well as ICA, demonstrated significantly worse RHI in patients with
obstructive and non-obstructive CAD compared to patients with no CAD. In that study,
patients with CAD had markedly worse RHI than subjects with CAD of our study. This
41
difference might reflect a greater risk factor burden of the patients of the earlier study [130].
The fact that we analyzed CAD by coronary CTA instead of ICA might also have affected
the results. Since coronary CTA allows for detection of non-obstructive lesions that might not
be visible on ICA, more patients with limited disease are likely to be classified as having
CAD when coronary CTA is used. Another study [131] reported a lack of association
between endothelial function, expressed in terms of flow mediated dilation of the brachial
artery and ICA findings, in patients with severe CAD. A possible interpretation of their
findings might be that endothelial function testing is of limited value for patients with preexisting severe CAD. It has indeed been suggested that endothelial function testing, and in
particular testing involving the microvasculature as is the case for the RH-PAT method,
might be more suitable for subjects with a low prevalence of risk factors [111]. This
hypothesis is not supported by our findings. Endothelial dysfunction has been shown to
precede the appearance of CAD and to predict the progression of IMT, hence suggesting that
it represents an earlier step in the atherosclerotic process [89,132]. However, in our study, a
large number of subjects with CAD evident at coronary CTA displayed normal endothelial
function.
One explanation for the lack of association between IMT or RH-PAT and CAD might be the
fact that the different measures reflect the presence of different underlying CVD risk factors,
to which the various vascular beds are more or less susceptible. It has been shown that certain
risk factors have different impacts on IMT and CAC, with diabetes mellitus having a
relatively higher impact on IMT and blood pressure showing a stronger association with CAC
[133]. The finding of a strong correlation between CAC and the presence and extent of CAD
as determined by coronary CTA is not surprising, since the two measures correspond to two
aspects of the atherosclerotic process of the coronary vascular bed.
Among circulating biomarkers we found that plasma cystatin C was an independent predictor
of CAD, with an OR of 2.5 adjusted for confounders, including estimated GFR, based on
serum creatinine. Adiponectin levels were significantly lower and lipoprotein(a) levels were
higher in the CAD group by non-parametric analysis. However, these associations did not
remain in the multivariable logistic regression analysis.
An association between elevated plasma levels of cystatin C and CVD has been demonstrated
in patient groups with elevated CVD risk as well as in the general population [134,135],
including in subjects without chronic kidney disease (estimated GFR ≥60 ml/min/1.73m2)
[136]. Cystatin C is an accurate marker of renal function and its association with CVD might
be related to compromised renal function, possibly also within the normal GFR range. The
pathophysiological mechanisms are incompletely understood and other effects of cystatin C
on for instance atherogenesis have been suggested [137]. Two previous ICA studies
[138,139] showed associations between cystatin C levels and extent of CAD and number of
stenotic vessels (≥51%). In both studies, a majority of the study participants were men, with a
higher prevalence of CVD risk factors than in our study group. The extent and severity of
CAD was considerably more pronounced and concentrations of cystatin C were in general
42
higher than in our study. To the best of my knowledge, there is only one previous publication
[103] that has examined the association between plasma levels of cystatin C and CAD
determined by coronary CTA. That study differs fundamentally from our study since it
included only patients with evidence of CAD at coronary CTA (stenosis <50%). In addition,
a large proportion were men (59%) and the prevalence of diabetes mellitus and hypertension
was higher. Imai et al. [103] showed that cystatin C was associated with the number of
diseased segments and that increasing plasma concentrations of cystatin C had a stronger
impact on non-calcified compared to calcified plaques, suggesting a possible role for cystatin
C in the early stage of CAD. Our study provides additional information by demonstrating an
association between plasma levels of cystatin C and absence or presence of CAD at coronary
CTA, in patients with normal or mildly reduced kidney function. Our findings suggest that
there is an association between cystatin C and CAD even in subjects with lower CVD risk
than previously reported. Our study, in contrast to most previous studies, reported plasma
levels of cystatin C that were relatively low (mainly within clinical reference ranges), which
indicates that the predictive value of cystatin C applies to both low and high plasma
concentrations.
Adiponectin was lower in the group with CAD by non-parametric testing, but the difference
did not remain in the logistic regression analysis. Only a few studies [101,102,104] have
investigated the association between plasma adiponectin and coronary CTA findings, and
they have demonstrated conflicting results. One study [101] demonstrated an inverse
association between plasma levels of adiponectin and non-calcified plaque and another [104]
showed that adiponectin was an independent predictor of multivessel CAD, whereas a third
study [102] did not find a significant association between levels of adiponectin and coronary
CTA findings. In all three studies, the study participants had an elevated CVD risk and more
pronounced coronary CTA plaque burden compared to the participants of our study. Due to
the small number of subjects with exclusively non-calcified plaque of our study, analyses
stratified by plaque type were not performed.
In multivariable analysis, lipoprotein(a) was not associated with CAD, which is consistent
with two previous coronary CTA studies [101,102]. In contrast, numerous prospective studies
[140] have shown a moderately strong association of lipoprotein(a) with CVD events, largely
independent of classical CVD risk factors. This suggests that lipoprotein(a) might have
effects on CVD risk that are not mediated through an increased coronary plaque burden.
Within the study group, ranges of plasma biomarker concentrations were quite small, which
may partly account for the lack of associations between CAD and circulating biomarkers.
Due to the limited sample size, I focused on correlation and binary logistic regression
analyses and did not perform additional statistical analyses such as the C statistic or area
under the receiver operating characteristic curve.
In order for a biomarker to be clinically useful it should meet several criteria. To begin with,
measurement should be easy. In addition, the biomarker should provide new information on
top of traditional risk factors and it should have an effect on patient management [79,141]. In
43
addition to a statistically significant association with the outcome, the test needs to have an
ability to discriminate (between those who will and those who will not get the disease). It
needs to be calibrated (tested for the agreement between predicted and observed risk in
groups with different baseline risks) and tested for its ability to reclassify subjects from one
risk category to another, thus affecting treatment decisions [75,79]. Ideally, the accuracy of
the reclassification should also be explored. Hence, in order to establish the clinical
usefulness of a novel biomarker large-scale studies are necessary.
44
7 CONCLUSIONS AND FINAL REMARKS
The general aims of this thesis were to investigate the underlying mechanisms of MINOCA
and to examine the association between coronary CTA plaque burden and other risk markers
of CVD.
Our results suggest that:
o Non-obstructive CAD is not a frequent cause of MINOCA in patients with
angiographically normal or near-normal coronary arteries.
o MB is frequent and is most likely not a common cause of MINOCA or of TS.
Coronary CTA detects significantly more MBs than ICA.
o Neither IMT nor RH-PAT can reliably be used to predict subclinical CAD at coronary
CTA.
o Circulating cystatin C is an independent predictor of presence and extent of CAD at
coronary CTA.
Since the SMINC study was initiated in 2007, numerous studies have been published,
examining the prevalence, risk factors and possible pathophysiologic mechanisms of
MINOCA. There is strong evidence that the condition is heterogeneous, with a variety of
underlying causes [5,10,11]. Thus, MINOCA should not be considered a definitive diagnosis,
but rather a working diagnosis, warranting additional diagnostic evaluation. Myocarditis is
one of the conditions that may manifest itself as MINOCA. Findings of the SMINC study,
where myocarditis was excluded by CMR, suggest that TS is an important cause of
MINOCA [10,11]. Other potential causes include CAD with rupture of a non-stenotic lesion,
coronary artery spasm, thrombotic disorders and microvascular dysfunction [5,44,142].
Identifying the cause of MINOCA is crucial, since these conditions have different prognostic
implications and different treatment strategies are required in order to improve prognosis.
Additional research is needed to define optimal diagnostic strategies for patients with
MINOCA. At ICA, provocative testing may confirm coronary artery spasm and left
ventricular angiography may demonstrate the wall motion abnormality of TS [5,44]. In
addition, intravascular ultrasound may demonstrate the presence of vulnerable or ruptured
plaques [45] and if MB is suspected, testing of its physiological significance may be
performed [72,73,123]. CMR imaging will probably have a central role, since it allows for
assessment of TS as well as differentiation between myocardial infarction and myocarditis
[11,62,63]. In some cases, coronary CTA may be useful to confirm or rule out the presence of
non-obstructive CAD [143].
Regarding TS, this thesis contributed by demonstrating that MB is most likely not an
underlying cause, which highlights that the underlying pathogenic mechanisms still need to
be clarified. Establishing consensus on the diagnostic criteria will facilitate future research on
pathophysiological mechanisms, diagnosis, prognosis, and patient management [7].
45
CVD is the leading cause of death worldwide and causes considerable morbidity each year.
Improving CVD risk prediction might enable more individualized preventive measures,
which may in turn have a large impact on health and healthcare costs. However, the
relationship between different risk markers and adverse CVD events is complex and the
contribution of individual risk factors or markers is difficult to isolate. This study provides
novel insights into the association of cystatin C with non-obstructive CAD determined by
coronary CTA, in a study group with a predominantly low-to-intermediate CVD risk profile
and with normal or only mildly reduced kidney function. Cystatin C might thus have a
potential to serve as a screening test aiming to identify patients with subclinical CAD, who
are at increased risk of future CVD events. However, large-scale studies will be necessary to
establish its clinical usefulness.
A prognostic value of the total coronary CTA plaque burden (obstructive and non-obstructive
plaque) has been established [14,15,27,28] and several studies have suggested that
identification of vulnerable plaque characteristics may provide additional predictive
information [112-114,120]. However, there are several factors that may limit the usefulness
of coronary CTA to determine vulnerable plaque characteristics, a few of them previously
discussed. Keeping radiation exposure low in coronary CTA is essential for patient safety.
Even though low-dose CT protocols yield a high diagnostic accuracy for diagnosing stenoses,
evaluation of non-calcified plaque and plaque components becomes more uncertain with very
low radiation doses. One future application of coronary CTA in identification of high-risk
plaque may instead be the combined anatomic imaging of coronary CTA with imaging of
metabolism and inflammation using 18F-2-deoxyglucose (FDG) positron emission
tomography (PET) [144]. Recently, there has been increasing emphasis on identifiying “the
vulnerable patient”, rather than the vulnerable plaque, since CAD has become considered
more as a generalized disease than a focal disease. Instead of focusing on identifying
individual high-risk plaques, it may prove to be more useful to consider the extent and
inflammatory activity of CAD together with other risk markers, including markers of a
prothrombotic state [122,144]. Hence, large-scale studies with a comprehensive integrative
approach will be necessary to investigate which combination of imaging parameters and other
risk markers yields the most accurate individual risk prediction. An accurate tool for CVD
risk prediction will allow for the assignment of individuals to risk categories that can guide
management and treatment and, in the long run, lower CVD morbidity and mortality.
46
8 ACKNOWLEDGEMENTS
I wish to express my sincere gratitude to all colleagues, friends and family who have in
different ways contributed to this work.
I especially wish to thank Kerstin Cederlund, my principal supervisor, for her motivation
and support, and for sharing her immense knowledge. I cannot imagine a better PhD
supervisor or mentor!
I also wish to thank:
Per Tornvall, my co-supervisor and principal investigator of the SMINC study, without
whom this PhD project would never have been realized.
Torkel Brismar, my second co-supervisor, for sharing his energy and enthusiasm for research.
It is contagious!
Andreas Rück, my third co-supervisor, for not hesitating to ask the tough questions, that
made me think things over again.
Professor Peter Aspelin, who has created such a stimulating intellectual environment,
encouraging so many to enter research and discover its challenges and possibilities. It is a
privilege to be working in your department!
Heads of the Radiology Department at Karolinska University Hospital in Huddinge.
My former and present colleagues in the thoracic radiology section: Egon, Sören, Margareta,
Ulrika, Ming, Raquel, Andreas, Laura and Eva.
Jonaz Ripsweden, my colleague and friend. It is nice to have someone who is always a few
steps ahead, showing the way!
Anders Svensson, for helping me develop a deeper understanding of CT technology and its
possibilities and for providing us with excellent coronary CTA examinations.
Katharina Brehmer, for being such a knowledgeable and reliable colleague and friend!
My family: My husband Gergö who took care of everything while I wrote this thesis, and my
wonderful children Max and Julia, who were incredibly patient! My sisters Karin and Lisathank you for your intelligent feedback on this thesis and for always being there. My
supporting parents, Anders and Eva (who illustrated TS!).
47
9 REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
48
Cardiovascular diseases (CVDs) fact sheet n°317 [Internet]. World Health Organization
[last modified 2015 Jan; cited 2015 Oct] Available from:
http://www.who.int/mediacentre/factsheets/fs317/en/
Swedish Web-system for Enhancement and Development of Evidence-based care in
Heart disease Evaluated According to Recommended Therapies (SWEDEHEART).
Annual Report 2014. Available from: http://www.ucr.uu.se/swedeheart/
Mowatt G, Cook JA, Hillis GS, et al. 64-slice computed tomography angiography in
the diagnosis and assessment of coronary artery disease: Systematic review and metaanalysis. Heart 2008;94:1386-1393.
Fischer C, Hulten E, Belur P, et al. Coronary CT angiography versus intravascular
ultrasound for estimation of coronary stenosis and atherosclerotic plaque burden: a
meta-analysis. J Cardiovasc Comput Tomogr 2013;7(4):256-266.
Pasupathy S, Air T, Dreyer RP et al. Systematic review of patients presenting with
suspected myocardial infarction and nonobstructive coronary arteries. Circulation
2015; 131:861-870.
Templin C, Ghadri JR, Diekmann J, et al. Clinical features and outcomes of Takotsubo
(stress) cardiomyopathy. N Eng J Med 2015;373:929-938.
Akashi YJ, Nef HM, Lyon AR. Epidemiology and pathophysiology of Takotsubo
syndrome. Nat Rev Cardiol 2015;12:387-397
Prasad A, Lerman A, Rihal CS. Apical ballooning syndrome (tako-tsubo or stress
cardiomyopathy): A mimic of acute myocardial infarction. Am Heart J 2008;155:408417.
Pilgrim T, Wyss T. Takotsubo cardiomyopathy or transient left ventricular apical
ballooning syndrome: A systematic review. Int J Cardiol 2008;124:283-292.
Collste O, Sorensson P, Frick M, et al. Myocardial infarction with normal coronary
arteries is common and associated with normal findings on cardiovascular magnetic
resonance imaging: results from the Stockholm Myocardial Infarction with Normal
Coronaries study. J Intern Med 2013;273:189-196.
Daniel M, Ekenback C, Agewall S, et al. Risk factors and markers for myocardial
infarction with angiographically normal coronary arteries. Am J Cardiol 2015;116:838844.
Mohlenkamp S, Hort W, Ge J, et al. Update on myocardial bridging. Circulation
2002;106:2616-2622.
Migliore F, Maffei E, Marra MP, et al. LAD coronary artery myocardial bridging and
apical ballooning syndrome. JACC Cardiovasc Imaging 2013;6:32-41.
Andreini D, Pontone G, Mushtaq S, et al. A long-term prognostic value of coronary CT
angiography in suspected coronary artery disease. JACC Cardiovasc Imaging
2012;5:690-701.
Cho I, Chang HJ, Sung JM, et al. Coronary computed tomographic angiography and
risk of all-cause mortality and nonfatal myocardial infarction in subjects without chest
pain syndrome from the CONFIRM Registry (coronary CT angiography evaluation
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
for clinical outcomes: an international multicenter registry). Circulation 2012;126:304313.
Hadamitzky M, Achenbach S, Al-Mallah M, et al. Optimized prognostic score for
coronary computed tomographic angiography results From the CONFIRM Registry
(COronary CT Angiography EvaluatioN For Clinical Outcomes: An InteRnational
Multicenter Registry). J Am Coll Cardiol 2013;62:468-476.
Hausleiter J, Meyer T, Hermann F, et al. Estimated radiation dose associated with
cardiac CT angiography. J Am Med Assoc 2009;301(5):500-507.
Flohr T, De Cecco CN, Schmidt B, et al. Computed tomographic assessment of
coronary artery disease. Radiol Clin N Am 2015;53:271-285.
Abdulla J, Abildstrom SZ, Gotzsche O, et al. 64-multislice detector computed
tomography coronary angiography as potential alternative to conventional coronary
angiography: A systematic review and meta-analysis. Eur Heart J 2007;28:3042-3050.
Voros S, Rinehart S, Qian Z, et al. Coronary atherosclerosis imaging by coronary CT
angiography: Current status, correlation with intravascular interrogation and metaanalysis. JACC Cardiovasc Imaging 2011;4:537-548.
Leschka S, Seitun S, Dettmer M, et al. Ex vivo evaluation of coronary atherosclerotic
plaques: Characterization with dual-source CT in comparison with histopathology.
J Cardiovasc Comput Tomogr 2010;4:301-308
Al-Mallah M, Aljizeeri A, Villines T, et al. Cardiac computed tomography in current
cardiology guidelines. J Cardiovasc Comput Tomogr 2015;doi:
10.1016/j.jcct.2015.09.003
Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium
using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827-832.
Greenland P, Bonow RO, Brundage BH, et al. ACCF/AHA 2007 clinical expert
consensus document on coronary artery calcium scoring by computed tomography in
global cardiovascular risk assessment and in evaluation of patients with chest pain: A
report of the American College of Cardiology Foundation Clinical Expert Consensus
Task Force (ACCF/AHA Writing Committee to Update the 2000 Expert Consensus
Document on Electron Beam Computed Tomography) developed in collaboration with
the Society of Atherosclerosis Imaging and Prevention and the Society of
Cardiovascular Computed Tomography. J Am Coll Cardiol 2007;49:378-402.
Greenland P, Alpert JS, Beller GA, et al. 2010 ACCF/AHA guideline for assessment of
cardiovascular risk in asymptomatic adults: A report of the American College of
Cardiology Foundation/American Heart Association Task Force on Practice
Guidelines. J Am Coll Cardiol 2010;56(25):e50-103.
Yeboah J, Carr JJ, Terry JG, et al. Computed tomography-derived cardiovascular risk
markers, incident cardiovascular events, and all-cause mortality in nondiabetics: the
Multi-Ethnic Study of Atherosclerosis. Eur J Prev Cardiol 2014;21(10):1233.1241.
Pundziute G, Schuijf JD, Jukema JW, et al. Evaluation of plaque characteristics in
acute coronary syndromes: Non-invasive assessment with multi-slice computed
tomography and invasive evaluation with intravascular ultrasound radiofrequency data
analysis. Eur Heart J 2008;29:2373-2381.
49
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
50
Min JK, Dunning A, Lin FY, et al. Age- and sex-related differences in all-cause
mortality risk based on coronary computed tomography angiography findings.
J Am Coll Cardiol 2011;58(8):849-860.
Libby P, Ridker P, Hansson G. Progress and challenges in translating the biology of
atherosclerosis. Nature 2011;473:317-325.
Sandfort V, Lima J, Bluemke D. Noninvasive imaging of atherosclerotic plaque
progression- status of coronary computed tomography angiography.
Circ Cardiovasc Imaging 2015;8:e003316.
Park H, Heo R, Hartaigh B, et al. Atherosclerotic plaque characteristics by CT
angiography identify coronary lesions that cause iscemia. JACC Cardiovasc Imaging
2015;8:1-10.
McCabe J, Croce K. Optical coherence tomography. Circulation 2012;126:2140-2143.
Thygesen K, Alpert JS, White HD. Universal definition of myocardial infarction.
Eur Heart J 2007;28:2525-2538.
Thygesen K, Alpert JS, Jaffe A, et al. Third Universal Definition of Myocardial
Infarction. Circulation 2012;126:2020-2035.
DeWood MA, Spores J, Notske R, et al. Prevalence of total coronary occlusion during
the early hours of transmural myocardial infarction. N Engl J Med 1980;303:897-902.
DeWood MA, Stifter WF, Simpson CS, et al. Coronary arteriographic findings soon
after non-Q-wave myocardial infarction. N Engl J Med 1986;315:417-423.
Stensaeth KH, Fossum E, Hoffmann P, et al. Clinical characteristics and role of early
cardiac magnetic resonance imaging in patients with suspected ST-elevation
myocardial infarction and normal coronary arteries. Int J Cardiovasc Imaging
2011;27:355-365.
Gehrie ER, Reynolds HR, Chen AY, et al. Characterization and outcomes of women
and men with non-ST-segment elevation myocardial infarction and nonobstructive
coronary artery disease: Results from the Can Rapid Risk Stratification of Unstable
Angina Patients Suppress Adverse Outcomes with Early Implementation of the
ACC/AHA Guidelines (CRUSADE) quality improvement initiative. Am Heart J
2009;158:688-694.
Larsen AI, Nilsen DW, Yu J, Mehran R, et al. Long-term prognosis of patients
presenting with ST-segment elevation myocardial infarction with no significant
coronary artery disease (from the HORIZONS-AMI trial). Am J Cardiol 2013;111:643648.
Agewall S, Daniel M, Eurenius L, et al. Risk factors for myocardial infarction with
normal coronary arteries and myocarditis compared with myocardial infarction with
coronary artery stenosis. Angiology 2012;63:500-503.
Chokshi NP, Iqbal SN, Berger RL, et al. Sex and race are associated with the absence
of epicardial coronary artery obstructive disease at angiography in patients with acute
coronary syndromes. Clin Cardiol 2010;33:495-501.
Redfors B, Angerås O, Råmunddal T, et al. Trends in gender differences in cardiac care
and outcome after acute myocardial infarction in western Sweden: A report from the
Swedish Web System for Enhancement of Evidence-Based Care in Heart Disease
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
Evaluated According to Recommended Therapies (SWEDEHEART).
J Am Heart Assoc 2015;4:e001995.
Sethi A, Bajaj A, Malhotra G, et al. Diagnostic accuracy of senstitive or high-sensitive
troponin on presentation for myocardial infarction: A meta-analysis and systematic
review. Vasc Health Risk Manag 2014:10;435-450.
Ong P, Athanasiadis A, Hill S, et al. Coronary artery spasm as a frequent cause of acute
coronary syndrome: The CASPAR (Coronary Artery Spasm in Patients With Acute
Coronary Syndrome) Study. J Am Coll Cardiol 2008;52:523-527.
Reynolds HR, Srichai MB, Iqbal SN, et al. Mechanisms of myocardial infarction in
women without angiographically obstructive coronary artery disease. Circulation
2011;124:1414-1425.
Kardasz I, De Caterina R. Myocardial infarction with normal coronary arteries: a
conundrum with multiple aetiologies and variable prognosis: An update. J Int Med
2007;261:330-348.
Caussin C, Ohanessian A, Lancelin B, et al. Coronary plaque burden detected by
multislice computed tomography after acute myocardial infarction with near-normal
coronary arteries by angiography. Am J Cardiol 2003;92:849-852.
Aldrovandi A, Cademartiri F, Menozzi A, et al. Evaluation of coronary atherosclerosis
by multislice computed tomography in patients with acute myocardial infarction and
without significant coronary artery stenosis: A comparative study with quantitative
coronary angiography. Circ Cardiovasc Imaging 2008;1:205-211.
Aldrovandi A, Cademartiri F, Arduini D, et al. Computed tomography coronary
angiography in patients with acute myocardial infarction without significant coronary
stenosis. Circulation 2012;126:3000-3007.
Sato H, Tateishi H, Uchida T. Takotsubo-type cardiomyopathy due to multivessel
spasm. In: Kodama K, Haze K, Hon M, eds. Clinical Aspect of Myocardial Injury:
From Ischemia to Heart Failure (in Japanese). Tokyo, Japan: Kagakuhyouronsya Co
1990:56-64.
Singh NK, Rumman S, Mikell FL et al. Stress cardiomyopathy: Clinical and
ventriculographic characteristicts in 107 North American subjects. Int J Cardiol
2010; 141(3):297-303.
Haghi D, Athanasiadis A, Papavassiliu T, et al. Right ventricular involvement in
Takotsubo cardiomyopathy. Eur Heart J 2006;27(20):2433-2439.
Omerovic E. How to think about stress-induced cardiomyopathy? – Think “out of the
box”! Scand Cardiovasc J 2011;45(2):67-71.
Y-Hassan S. Takotsubo syndrome triggered by acute coronary syndrome in a cohort of
20 patients; An often missed diagnosis. Int J Cardiol Res 2015;02(2):28-33.
Y-Hassan S. Myocarditis and takotsubo syndrome: Are they mutually exclusive?
Int J Cardiol 2014;177:149-151.
Yoshikawa T. Takotsubo cardiomyopathy, a new concept of cardiomyopathy: Clinical
features and pathophysiology. Int J Cardiol 2015;182:297-303.
Singh K, Carson K, Hibbert B, et al. Natural history of cardiac arrest in patients with
Takotsubo cardiomyopathy. Am J Cardiol 2015;115:1466-1472.
51
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
52
Brinjikji W, El-Sayed AM, Salka S. In-hospital mortality among patients with
takotsubo cardiomyopathy: A study of the National Inpatients Sample 2008 to 2009.
Am Heart J 2012;164:215-221.
Redfors B, Vedad R, Angerås O, et al. Mortality in takotsubo syndrome is similar to
mortality in myocardial infarction- A report from the SWEDEHEART registry.
Int J Cardiol 2015;185:282-289.
Samuels MA. The brain-heart connection. Circulation 2007;116:77-84.
Redfors M, Shao Y, Ali A, et al. Current hypotheses regarding the pathophysiology
behind the takotsubo syndrome. Int J Cardiol 2014;177:771-779.
Kawecki D, Morawiec B, Monney P, et al. Diagnostic contribution of cardiac magnetic
resonance in patients with acute coronary syndrome and culprit-free angiograms.
Med Sci Monit 2015;21:171-180.
Dastidar AG, Rodrigues J, Ahmed N et al. The role of cardiac MRI in patients with
troponin-positive chest pain and unobstructed coronary arteries.
Curr Cardiovasc Imaging Rep 2015;8:28.
Irvin RG. The angiographic prevalence of myocardial bridging in man. Chest
1982;81:198-202.
Leschka S, Koepfli P, Husmann L, et al. Myocardial bridging: Depiction rate and
morphology at CT coronary angiography– comparison with conventional coronary
angiography. Radiology 2008;246:754-762.
Nakanishi R, Rajani R, Ishikawa Y, et al. Myocardial bridging on coronary CTA: An
innocent bystander or a culprit in myocardial infarction? J Cardiovasc Comput Tomogr
2012;6:3-13.
Kim PJ, Hur G, Kim SY, et al. Frequency of myocardial bridges and dynamic
compression of epicardial coronary arteries: A comparison between computed
tomography and invasive coronary angiography. Circulation 2009;119:1408-1416.
Rubinshtein R, Gaspar T, Lewis BS, et al. Long-term prognosis and outcome in
patients with a chest pain syndrome and myocardial bridging: A 64-slice coronary
computed tomography angiography study. Eur Heart J Cardiovasc Imaging
2013;14:579-585.
Wang MH, Sun AJ, Qian JY, et al. Myocardial bridging detection by non-invasive
multislice spiral computed tomography: Comparison with intravascular ultrasound.
Chin Med J 2008;121:17-21.
Kramer JR, Kitazume H, Proudfit WL, et al. Clinical significance of isolated coronary
bridges: Benign and frequent condition involving the left anterior descending artery.
Am Heart J 1982;103:283-288.
Juilliere Y, Berder V, Suty-Selton C, et al. Isolated myocardial bridges with
angiographic milking of the left anterior descending coronary artery: A long-term
follow-up study. Am Heart J 1995;129:663-665.
Corban MT, Hung OY, Eshtehardi P, et al. Myocardial bridging: Contemporary
understanding of pathophysiology with implications for diagnostic and therapeutic
strategies. J Am Coll Cardiol 2014;63:2346-2355.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
Teragawa H, Fukuda Y, Matsuda K, et al. Myocardial bridging increases the risk of
coronary spasm. Clin Cardiol 2003;26:377-383.
D'Agostino RB, Sr., Vasan RS, Pencina MJ, et al. General cardiovascular risk profile
for use in primary care: the Framingham Heart Study. Circulation 2008;117:743-753.
Executive summary of the Third Report of the National Cholesterol Education Program
(NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults (Adult Treatment Panel III). J Am Med Assoc 2001;285:24862497.
Catapano A, Reiner Z, De Backer G, et al. ESC/EAS Guidelines for the management of
dyslipidaemias The Task Force for the management of dyslipidaemias of the European
Society of Cardiology (ESC) and the European Atherosclerosis Society (EAS).
Atherosclerosis 2011;217:3-46.
Conroy RM, Pyorala K, Fitzgerald AP, et al. Estimation of ten-year risk of fatal
cardiovascular disease in Europe: The score project. Eur Heart J 2003;24:987-1003.
Ridker PM, Buring JE, Rifai N, et al. Development and validation of improved
algorithms for the assessment of global cardiovascular risk in women: The Reynolds
risk score. J Am Med Assoc 2007;297:611-619.
Wang T. Assessing the role of circulating, genetic, and imaging biomarkers in
cardiovascular risk prediction. Circulation 2011;123:551-565.
Khot U, Khot M, Bajzer C, et al. Prevalence of conventional risk factors in patients
with coronary heart disease. J Am Med Assoc 2003;290:898-904.
Lorenz MW, Markus HS, Bots ML, et al. Prediction of clinical cardiovascular events
with carotid intima-media thickness: A systematic review and meta-analysis.
Circulation 2007;115:459-467.
Inaba Y, Chen JA, Bergmann SR. Carotid plaque, compared with carotid intima-media
thickness, more accurately predicts coronary artery disease events: a meta-analysis.
Atherosclerosis 2012;220:128-133.
Naqvi T, Lee M. Carotid intima-media thickness and plaque in cardiovascular risk
assessment. JACC Cardiovasc Imaging 2014;7:1025-1038.
Lorenz MW, Polak JF, Kavousi M, et al. Carotid intima-media thickness progression to
predict cardiovascular events in the general population (the prog-imt collaborative
project): A meta-analysis of individual participant data. Lancet 2012;379:2053-2062.
Goff D, Lloyd-Jones D, Bennett G, et al. 2013 ACC/AHA guideline on the assessment
of cardiovascular risk: A report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014;63:2935-2948.
Park K, Park W. Endothelial dysfunction: Clinical implications in cardiovascular
disease and therapeutic approaches. J Korean Med Sci 2015;30:1213-1225.
Mordi I, Tzemos N. Is reversal of endothelial dysfunction still an attractive target in
modern cardiology? World J Cardiol 2014;6(8):824-835.
Frolow M, Drozdz A, Kowaleska A, et al. Comprehensive assessment of vascular
health in patients; towards endothelium-guided therapy. Pharmacol Rep 2015;67:786792.
53
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
54
Halcox JP, Schenke WH, Zalos G, et al. Prognostic value of coronary vascular
endothelial dysfunction. Circulation 2002;106:653-658.
Yeboah J, Folsom AR, Burke GL, et al. Predictive value of brachial flow-mediated
dilation for incident cardiovascular events in a population-based study: The multiethnic study of atherosclerosis. Circulation 2009;120:502-509.
Anderson TJ, Uehata A, Gerhard MD, et al. Close relation of endothelial function in the
human coronary and peripheral circulations. J Am Coll Cardiol 1995;26:1235-1241.
Bonetti PO, Pumper GM, Higano ST, et al. Noninvasive identification of patients with
early coronary atherosclerosis by assessment of digital reactive hyperemia.
J Am Coll Cardiol 2004;44:2137-2141.
Celermajer DS, Sorensen KE, Gooch VM, et al. Non-invasive detection of endothelial
dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340:11111115.
Kandhai-Ragunath JJ, Jorstad HT, de Man FH, et al. Approaches for non-invasive
assessment of endothelial function: Focus on peripheral arterial tonometry.
Neth Heart J 2013;21:214-218.
Konttinen J, Lindholm H, Sinisalo J, et al. Association between lowered endothelial
function measured by peripheral arterial tonometry and cardio-metabolic risk factors–
A cross-sectional study of Finnish municipal workers at risk of diabetes and
cardiovascular disease. BMC Cardiovasc Disord 2013;13:83.
Gilstrap L, Wang T. Biomarkers and cardiovascular risk assessment for primary
prevention: An update. Clin Chem 2012;58:72-82.
Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease meta-analysis of
prospective studies. Circulation 2000;102:1082-1085.
Luo J, Wang L, Hu H, et al. Cystatin C and cardiovascular or all-cause mortality risk in
the general population: A meta-analysis. Clin Chim Acta 2015;450:39-45.
Djaberi R, Schuijf JD, de Koning EJ, et al. Usefulness of carotid intima-media
thickness in patients with diabetes mellitus as a predictor of coronary artery disease.
Am J Cardiol 2009;104:1041-1046.
Schroeder B, Francis G, Leipsic J, et al. Early atherosclerosis detection in
asymptomatic patients: A comparison of carotid ultrasound, coronary artery calcium
score, and coronary computed tomography angiography. Can J Cardiol 2013;29:16871694.
Bamberg F, Truong QA, Koenig W, et al. Differential associations between blood
biomarkers of inflammation, oxidation, and lipid metabolism with varying forms of
coronary atherosclerotic plaque as quantified by coronary CT angiography.
Int J Cardiovasc Imaging 2012;28:183-192.
Caselli C, de Graaf M, Lorenzoni V, et al. HDL cholesterol, leptin and interleukin-6
predict high risk coronary anatomy assessed by CT angiography in patients with stable
chest pain. Atherosclerosis 2015;241:55-61.
Imai A, Komatsu S, Ohara T, et al. Serum cystatin C is associated with early stage
coronary atherosclerotic plaque morphology on multidetector computed tomography.
Atherosclerosis 2011;218:350-355.
104. Matsuda M, Tamura R, Kishida N, et al. Predictive Value of Adiponectin in Patients
with Multivessel Coronary Atherosclerosis Detected on Computed Tomography
Angiography. J Atheroscler Thromb 2013;20:767-776.
105. Thygesen K, Alpert JS, White HD. Universal definition of myocardial infarction.
Eur Heart J 2007;28:2525-2538.
106. Austen WG, Edwards JE, Frye RL, et al. A reporting system on patients evaluated for
coronary artery disease. Report of the ad hoc committee for grading of coronary artery
disease, Council on Cardiovascular Surgery, American Heart Association. Circulation
1975;51:5-40.
107. Budoff MJ, Cohen MC, Garcia MJ, et al. ACCF/AHA clinical competence statement
on cardiac imaging with computed tomography and magnetic resonance. Circulation
2005;112:598-617.
108. Kim PJ, Hur G, Kim SY, et al. Frequency of myocardial bridges and dynamic
compression of epicardial coronary arteries: A comparison between computed
tomography and invasive coronary angiography. Circulation 2009;119:1408-1416.
109. Hamburg NM, Benjamin EJ. Assessment of endothelial function using digital pulse
amplitude tonometry. Trends Cardiovasc Med 2009;19:6-11.
110. Hamburg NM, Palmisano J, Larson MG, et al. Relation of brachial and digital measures
of vascular function in the community: the Framingham heart study. Hypertension
2011;57:390-396.
111. Flammer AJ, Anderson T, Celermajer DS, et al. The assessment of endothelial
function: from research into clinical practice. Circulation 2012;126:753-767.
112. Pundziute G, Schuijf JD, Jukema JW, et al. Evaluation of plaque characteristics in
acute coronary syndromes: Non-invasive assessment with multi-slice computed
tomography and invasive evaluation with intravascular ultrasound radiofrequency data
analysis. Eur Heart J 2008;29:2373-2381.
113. Pflederer T, Marwan M, Schepis T, et al. Characterization of culprit lesions in acute
coronary syndromes using coronary dual-source CT angiography. Atherosclerosis
2010;211:437-444.
114. Virmani R, Burke AP, Farb A, et al. Pathology of the vulnerable plaque.
J Am Coll Cardiol 2006;47: C13-18.
115. Papadopoulou SL, Neefjes LA, Schaap M, et al. Detection and quantification of
coronary atherosclerotic plaque by 64-slice multidetector CT: A systematic head-tohead comparison with intravascular ultrasound. Atherosclerosis 2011;219:163-170.
116. Øvrehus KA, Marwan M, Bøtker HE, et al. Reproducibility of coronary plaque
detection and characterization using low radiation dose coronary computed
tomographic angiography in patients with intermediate likelihood of coronary artery
disease (ReSCAN study). Int J Cardiovasc Imaging 2012;28:889-899.
117. Maurovich-Horvat P, Ferencik M, Bamberg F, et al. Methods of plaque quantification
and characterization by cardiac computed tomography. J Cardiovasc Comput Tomogr
2009;3 Suppl 2:S91-98.
118. Park H, Lee B, Shin S, et al. Clinical feasibility of 3D automated coronary
atherosclerotic plaque quantification algorithm on coronary computed tomography
55
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
56
angiography: comparison with intravascular ultrasound. Eur Radiol 2015;25:30733083.
Fleg J, Stone G, Fayad Z, et al. Detection of high-risk atherosclerotic plaque: report of
the NHLBI working group on current status and future directions.
JACC Cardiovasc Imaging 2012;5(9)941-955.
Motoyama S, Sarai M, Harigaya H, et al. Computed tomographic angiography
characteristics of atherosclerotic plaques subsequently resulting in acute coronary
syndrome. J Am Coll Cardiol 2009;54:49-57.
Ripsweden J. (2011) Computed tomography of the coronary arteries. Ph.D thesis,
Karolinska Institutet. Retrieved from: https//openarchive.ki.se
Arbab-Zadeh A, Fuster V. The myth of the “vulnerable plaque”. J Am Coll Cardiol
2015;65:846-855.
Hazenberg AJ, Jessurun GA, Tio RA. Mechanisms involved in symptomatic
myocardial bridging: Value of sequential testing for endothelial function, flow reserve
measurements and dobutamine stress angiography. Neth Heart J 2008;16:10-15.
Gallagher S, Jones DA, Anand V, et al. Diagnosis and management of patients with
acute cardiac symptoms, troponin elevation and culprit-free angiograms. Heart
2012;98:974-981.
Lurz P, Eitel I, Adam J, et al. Diagnostic performance of CMR imaging compared with
EMB in patients with suspected myocarditis. JACC Cardiovasc Imaging 2012;5:513524.
Bergström G, Berglund G, Blomberg A, et al. The Swedish CardioPulmonary
BioImage Study; Objectives and design. J Intern Med 2015; doi: 10.1111/joim.12384.
[Epub ahead of print]
Stiermaier T, Desch S, Blazek S, et al. Frequency and significance of myocardial
bridging and recurrent segment of the left anterior descending coronary artery in
patients with takotsubo cardiomyopathy. Am J Cardiol 2014;114:1204-1209.
Cohen GI, Aboufakher R, Bess R, et al. Relationship between carotid disease on
ultrasound and coronary disease on CT angiography. JACC Cardiovasc Imaging
2013;6:1160-1167.
Rohani M, Jogestrand T, Ekberg M, et al. Interrelation between the extent of
atherosclerosis in the thoracic aorta, carotid intima-media thickness and the extent of
coronary artery disease. Atherosclerosis 2005;179:311-316.
Matsuzawa Y, Sugiyama S, Sugamura K, et al. Digital assessment of endothelial
function and ischemic heart disease in women. J Am Coll Cardiol 2010;55:1688-1696.
Rohani M, Jogestrand T, Kallner G, et al. Morphological changes rather than flowmediated dilatation in the brachial artery are better indicators of the extent and severity
of coronary artery disease. J Hypertens 2005;23:1397-1402.
Bugiardini R, Manfrini O, Pizzi C, et al. Endothelial function predicts future
development of coronary artery disease: a study of women with chest pain and normal
coronary angiograms. Circulation 2004;109:2518-2523.
133. Bauer M, Mohlenkamp S, Lehmann N, et al. The effect of age and risk factors on
coronary and carotid artery atherosclerotic burden in males-Results of the Heinz
Nixdorf Recall Study. Atherosclerosis 2009;205:595-602.
134. Luo J, Wang L, Hu H, et al. Cystatin C and cardiovascular or all-cause mortality risk in
the general population: A meta-analysis. Clin Chim Acta 2015;450:39-45.
135. Lee M, Saver JL, Huang WH, et al. Impact of elevated cystatin C level on
cardiovascular disease risk in predominantly high cardiovascular risk populations: A
meta-analysis. Circ Cardiovasc Qual Outcomes 2010;3:675-683.
136. Muntner P, Mann D, Winston J, et al. Serum Cystatin C and Increased Coronary Heart
Disease Prevalence in US Adults Without Chronic Kidney Disease. Am J Cardiol
2008;102:54-57.
137. Angelidis C, Deftereos S, Giannopoulous G, et al. Cystatin c: An emerging biomarker
in cardiovascular disease. Curr Top Med Chem 2013;13:164-179.
138. Niccoli G, Conte M, Della Bona R, et al. Cystatin C is associated with an increased
coronary atherosclerotic burden and a stable plaque phenotype in patients with
ischemic heart disease and normal glomerular filtration rate. Atherosclerosis
2008;198:373-380.
139. Kiyosue A, Hirata Y, Ando J, et al. Plasma Cystatin C Concentration Reflects the
Severity of Coronary Artery Disease in Patients Without Chronic Kidney Disease.
Circ J 2010;74:2441-2447.
140. Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease meta-analysis of
prospective studies. Circulation 2000;102:1082-1085.
141. Morrow D, De Lemos J. Benchmarks for the assessment of novel cardiovascular
biomarkers. Circulation 2007;115:949-952.
142. Beltrame J. Assessing patients with myocardial infarction and nonobstructed coronary
arteries (MINOCA). J Int Med 2013;273:182-185.
143. Bekkers S, Smulders M. High biomarkers, but normal angiogram: what next?
Heart Metab 2015;67:15-20.
144. Braunwald E. Progress in the noninvasive detection of high-risk coronary plaques.
J Am Coll Cardiol 2015;6:347-349.
57