Download Proton and sodium MRI assessment of myocardial viability

Proton and sodium MRI assessment
of myocardial viability
Eissa Nadi Eissa Agour
Proton and sodium MRI assessment
of myocardial viability
Over het afbeelden van de levensvatbaarheid van hartspier
middels proton en natrium ion magnetische resonantie
(met een samenvatting in het Nederlands)
Proefschrift
ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de
rector magnificus, prof.dr. G.J. van der Zwaan, ingevolge het besluit van het college
voor promoties in het openbaar te verdedigen op woensdag 31 oktober 2012 des
ochtends te 10.30 uur
door
Eissa Nadi Eissa Agour
geboren op 16 January 1979, Riyadh, Saudi Arabia
Promotoren: Prof. dr. P.A.F.M. Doevendans
Prof. dr. C.J.A. van Echteld
Co-promotoren: Dr.ir. G.J. Strijkers
Dr. F. Arslan
The research described in this thesis was supported by a grant of the Dutch Heart
Foundation (DHF-2006B066) and by the Netherlands Heart Institute (ICIN).
Financial support by the Dutch Heart Foundation for the publication of this thesis is
gratefully acknowledged
Financial support by the Heart & Lung Foundation Utrecht for the publication of this
thesis is gratefully acknowledged
ISBN: 978-90-8891-506-2
Printed & lay-out by: Proefschriftmaken.nl || Printyourthesis.com
Published by: Uitgeverij BOXPress, Oisterwijk
This thesis is dedicated to my parents
Content
Chapter 1
General introduction
9
Part I - Proton MRI
Chapter 2
Quantitative T2* assessment of acute and chronic myocardial
ischemia/reperfusion injury in mice
Chapter 3
Quantitative changes in T2* reflect remodeling of both
remote and ischemic myocardium in a murine heart failure model
Chapter 4
25
47
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
63
Part II - Sodium MRI
Chapter 5 Na Chemical Shift Imaging and Gd Enhancement of
23
myocardial edema
Chapter 6
81
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
105
Part III
Chapter 7
Summarizing discussion
131
Brief summary in Dutch - Korte samenvatting in het Nederlands
141
List of publications
143
Acknowledgements - Dankwoord
147
Curriculum vitae
151
7
Chapter 1
General Introduction
This chapter provides a brief introduction of myocardial injury, a brief summary of
cardiac imaging methods, followed by an overview of 1H MRI and 23Na MRI methods to
assess myocardial viability. Finally, study aim and thesis outline are presented.
9
Chapter 1
1.1 Myocardial injury
Ischemic heart disease is a leading cause of morbidity and mortality globally. Ischemic
heart disease is characterized by partial or total lack of oxygen-rich blood supply to
the myocardium. Although early reperfusion can salvage myocardium at risk, the
beneficial effect after reperfusion is not always apparent.
The severity of myocardial injury is influenced by many factors, not only the occlusion
location within the coronary arteries, but also the ischemic duration. If the blood
flow is restored (reperfusion) after a brief, mildly ischemic period [1], the myocardial
cellular injury may still be reversible. In contrast, a prolonged, totally ischemic period
ultimately causes irreversible cellular injury and leads to cell death. Such severe injury
likely results in impairment of cardiac function. After ischemia and reperfusion, the
myocardium undergoes a complex cascade of dynamic changes, both the necrotic
and viable myocardium [2]. Therefore, in ischemic heart disease, distinguishing
between viable and non-viable myocardium is of high clinical importance, as viable
myocardium may benefit from revascularization whereas scar tissue will not. Equally
important is the differentiation from non-ischemic heart disease. Sensitive and
accurate characterization of the myocardial status is therefore essential for accurate
diagnosis and suitable therapy planning. Ultimately, this may significantly improve
survival and outcome [3;4]
Rodent models of myocardial infarction
To mitigate the risk of myocardial infarction and morbidity in patients, the search
for alternative therapeutic interventions and in suitable experimental myocardial
infarction (MI) models is ongoing. Research using animal models helps identifying
novel molecular mechanisms and therapeutic targets as well as suitable intervention
methods, providing valuable information for clinical translation. Many researchers
have investigated the progress of myocardial infarction in well-defined, in vivo
small animal models of infarction [5]. Already In 1895, Oscar Langendorff pioneered
the isolated perfused rodent heart. Since then Langendorff perfused hearts have
10
General Introduction
yielded valuable data, increasing our understanding of the fundamental physiology
of the heart [6]., despite the difference in cardiac anatomy in these animal models,
as well as the accelerated infarct development process and the higher heart beat, in
comparison to humans
1.2
Imaging methods for assessing myocardial viability
Many studies have shown that performing ECG tests [7] and probing specific blood
biomarkers [8] are clinically valuable for the detection of myocyte death. Such tests
therefore are usually performed as the first diagnostic assessment for suspected
MI. However, more precise and detailed information about the extent of MI can be
obtained using medical imaging.
Many clinical and pre-clinical studies have demonstrated the potential of non-invasive
imaging to characterize myocardium throughout the development of MI [9;10],
including ultrasound [11], computed tomography CT [12], single photon emission
computed tomography (SPECT) [13], positron emission tomography (PET) [14], and
magnetic resonance imaging (MRI) [15]. Each of which has unique strengths and
limitations. Furthermore, combining different techniques might lead to a better
understanding of the disease and help in more precise and accurate estimation of
the extent of infarction [9;15;16]. Echocardiography has emerged as a powerful
tool for imaging of cardiac function and structure. Echocardiography techniques
such as tissue Doppler and strain imaging provide interesting new readouts [17]. Still
the interpretation of images remains a challenge and operator experience is critical
[11], particularly for imaging small animals. Parallel to the recent advances in clinical
and research fields, computed tomography (CT) has evolved into a reliable cardiac
imaging modality [10]. Advanced multi slice CT is a powerful modality in cardiac
imaging and is preferred above MRI for visualizing coronary arteries and quantifying
calcium content. SPECT and PET are widely used nuclear medical imaging methods
for assessing myocardial perfusion and metabolism. On the downside, these imaging
methods require injection of radiotracers and suffer from poor spatial resolution
11
Chapter 1
[18;19]. MRI has a lot to offer in clinical and experimental cardiovascular research. It
combines high spatial resolution with acceptable imaging time. Cardiovascular MRI
is capable of providing a variety of anatomical, functional and physiological readouts
of the heart in health and disease, which aid in clinical decision-making and provide
fundamental insights in heart function. A number of these readouts are illustrated in
figure 1 and detailed in the next section.
1.3 Cardiac Magnetic Resonance imaging (CMRI)
Distinguishing viable from non-viable myocardium is essential for prognosis of patient
with ischemic heart disease. The available clinical diagnostic modalities to detect
myocardial viability and test sarcolemmal function include PET and MRI. However, PET
scans suffer from poor spatial resolution and involve appreciable radiation exposure
for the patient, in particular in serial examinations. On the other hand, MRI shows
better spatial resolution and does not use ionizing radiation. Moreover, a variety
of cardiovascular MRI methods is available, both clinically and pre-clinically. Most
cardiovascular MR images are acquired using hydrogen nuclei (protons). However, MR
images can also be obtained using other nuclei, such as 13C, 31P and 23Na. In this thesis,
we will focus on proton and sodium nuclei for MRI.
Proton (1H) MRI
Many researchers have investigated the potential of 1H based cardiovascular MRI
methods, mainly because of its relatively high signal to noise ratio. The flexibility
to adjust MR pulse sequence makes it possible to obtain high spatial resolution MR
images with different contrast for different soft tissues.
Cine MRI is the most commonly used MRI tool which provided valuable information
related to cardiac morphology and function. Cine MRI is commonly used to yield a
movie of a single or multiple slices of the heart. Valuable information about heart
size and function can be obtained; such as LV volume, ejection fraction (EF), and wall
thickness and thickening.
12
General Introduction
To monitor myocardial perfusion, a sequence of MR images are acquired directly after
intravenously injected bolus of Gd-based contrast agent through the myocardium, in
a so-called first pass myocardial perfusion MRI.
Figure 1. Various MRI readouts of mouse hearts with
(A) Cine MRI at baseline and 28 days after Ischemia Reperfusion (I/R) injury. (B)T2*
weighted MRI 2 days after injection with micro RNA-55 transfected CMPCs with
SPIOS labeling. The red arrows indicate SPIOS labeled CMPCs. (C) Quantitative T2*
mapping of and Late Gadolinium enhancement (LGE) MRI of the same mouse at day
1 post I/R injury. Red arrows indicate infarct areas.
MR longitudinal relaxation time (T1) and transverse relaxation time (T2) have different
values for different components and biological structures, which can be identified
using T1- weighted and T2‑weighted MRI. In addition, MR sequence parameters
can be adapted to achieve specific resolution and contrast to track pathological
changes and identify injured myocardium. For instance, a T1-weighted MRI sequence
in combination with a gadolinium contrast agent injection is suitable to determine
infarct size after a delay between contrast injection and MRI scan, in so-called Late
Gadolinium Enhancement (LGE) MRI [20]. LGE MRI shows enhancement of the injured
myocardium. However, the enhancement pattern in LGE MRI can be complicated,
by the formation of interstitial edema, which may occur in the peri-infarct zone or in
other cardiac diseases [21-23]. The sensitivity of the T2 relaxation time to myocardial
water content makes T2‑weighted MRI suitable to image myocardial edema as an
13
Chapter 1
indicator of the area at risk (AAR) following myocardial infarction (MI) in mice [24] and
in human [25]. T2* MRI has been used as an imaging tool to diagnose the presence of
hemorrhage and myocardial iron[26].
Apparently, T1-weighted and T2‑weighted MRI show great promise to delineate injured
myocardium both in patients and in animal studies[27-32]. However, both these
methods depend on relative regional differences in myocardial signal intensity and
are perhaps less suitable to identify globally edematous tissue associated with other
disease states, such as hypertension or myocarditis.
Alternatively, quantitative relaxation time measurement by MR is an emerging tool in
the assessment of myocardial infarction. In quantitative MRI, relaxation times on a
standard scale (in ms) are related to specific pathological conditions. Quantitative MR
methods, such asT1 mapping,T2 mapping andT2* mapping, enable a quantitative measure
of different pathological status of the myocardium, thus eliminating the problem
of choosing threshold values. For example, cardiac T1 mapping has proven useful in
differentiating healthy from infarcted myocardial tissue and enables quantification of
local contrast agent concentration [33]. In addition, cardiac T1 mapping has been used
to quantify diffuse myocardial fibrosis in cardiac diseases [34]. T2 mapping was used to
distinguish reversible from irreversible myocardial edema [35] and has been recently
used to measure the extent of myocardial fibrosis in diabetic mice [34]. Recently
O’Regan et al. [36] explored the use of T2* mapping to quantify post‑reperfusion
hemorrhage of the heart following percutaneous primary coronary intervention after
MI. T2* mapping has also been used to correlate myocardial iron and systolic function
in thalassemia patients, who are at risk for cardiomyopathy. Recently, we [37] used T2*
mapping in mouse models to assess myocardial infarction (figure 1G) which will be the
main subject of chapters 3 and 4 of this thesis.
14
General Introduction
Sodium (23Na) MRI
Sodium and cell viability
Healthy cardiomyocytes maintain ionic gradients across the sarcolemma. Maintaining
+
this ionic homeostasis is essential to maintain cellular viability and function. The Na +
K -ATPase is considered as the main force to maintain these gradients across the
+
+
sarcolemma, This Na -K -ATPase is electrogenic and transports 2 K+ ions into the cell
against 3 Na+ out of the cell. Another important exchanger involved in Na+ transport
across the cell membrane is the Na+/Ca2+ exchanger, which transport 3 Na+ into the cell
in exchange for 1 Ca2+. These and other transport mechanism such as Na+/H+ exchange
and Na+/HCO3- are essential to maintain a stable gradients and play a role in maintain
cellular volume regulation, but are also involved in dysregulation of ion gradients in
myocardial disease states.
Viable cell maintain a Na+ concentration gradient across the sarcolemmal with
extracellular/intracellular ratio of 140 mM/10 mM. During the very first few minutes
+
of myocardial ischemia, Nai already starts to rise due to inhibition of the sarcolemmal
+
+
+
+
+
Na -K -ATPase and continued or increased influx through the Na -channel, via the
+
Na -H exchanger and possibly through other routes [38]. This very early rise in Nai
+
makes Nai a very sensitive indicator of ischemia. Reperfusion of viable myocardium
+
+
+
causes an immediate reactivation of Na -K -ATPase function and a decrease in Nai .
+
However, if reperfusion is delayed, Nai recovers but remains elevated above the
+
baseline level. Therefore, Nai is a very sensitive indicator of myocardial viability.
Imaging of myocardial sodium (23Na MRI)
Proton based MRI is commonly used in the clinic. However, proton based MRI
methods, such as T1-weighted and T2-weighted MRI do not provide information on
the pathophysiological state of the myocardium at the biochemical level [3]. Imaging
of myocardial sodium (23Na) MRI is an alternative method for evaluating myocardial
viability via assessment of Na+ homeostasis [39] and cell volume regulation.
15
Chapter 1
The relatively low 23Na concentrations in biological tissues (table 1) and the fast biexponential signal decay make sodium MRI a challenge. However, recent advances
in MRI and the availability of higher MR field strengths enable in vivo 23Na MRI in
reasonable scan times. Still, 23Na MRI has poor spatial resolution, which is the reason
that it is always combined with 1H MRI (figure 2).
1
H
Na
23
Nucleus Natural abundance (%)
99.99
100.00
Relative MRI sensitivity (%)
100
9.3
In vivo concentration (mM)
H2O: 110 000
Intracellular: 10
Extracellular: 140
Overall in vivo signal
1
4 x1-5
Table 1 Myocardial tissue concentrations and MR sensitivity of 1H and 23Na MRI [40]
Clinical 23Na MRI examinations are extremely limited and are mostly concerned with
imaging of total sodium (Nat+) [41]. An elevation of total sodium Nat+ signal has
been observed after myocardial infarction in a number of studies, both in humans
and animals [41-43]. This elevated signal might reflect different patho-physiological
conditions., For example, in acute, low-flow ischemia, the Nat+ signal is expected to
increase because of unbalanced ion homeostasis and edema formation [41]. Similarly,
the Nat signal is significantly elevated during chronic scar formation, because of the
expansion of the extracellular space in scar [40]. Using Nat+ MRI alone would not
discriminate between these two conditions.
16
General Introduction
Figure 2 A typical example of perfused rat heart with
Na Chemical Shift Imaging CSI
23
representing extracellular Na (Nae) and H Late gadolinium enhanced (LGE) MRI,
1
arrows show injured myocardium in the LGE MRI.
More information on the sodium gradient and cellular membrane integrity can be
+
+
obtained by imaging Nai and Nae cellular separately. This requires a paramagnetic
chemical shift agent, such as [TmDOTP]5- and a 23Na chemical shift imaging (CSI)
sequence. 23Na CSI will be the main topic in the second part of this thesis. As mentioned
23
earlier, due to rapid change of Nai+ during ischemia and reperfusion, Na CSI appears
to be an ideal diagnostic modality for early detection of myocardial viability.
1.4 Aim of the study and thesis outline
The aim of the studies presented in this thesis was to explore new MRI methods for
assessment of myocardial viability. This thesis constitutes of two main parts. Part 1
demonstrates the assessment of myocardial viability using proton MRI methods.
Chapter 2 explores the utility of quantitative T2* mapping as a noninvasive technique
to characterize the myocardium in the acute and chronic phases of I/R injury in an
in vivo mouse model. Chapter 3 explores the utility of quantitative T2* mapping to
characterize the myocardium in the acute and chronic phases of permanent coronary
artery ligation in a similar in vivo mouse model. Chapter 4 gives an example of how
17
Chapter 1
cine MRI can demonstrate prevention of adverse remodeling after infarction in TLR2/-
mice.
Part II is dedicated to Sodium MRI in comparison to Proton MRI methods. Chapter 5,
describes the development of a “non-infarct model” of extracellular edema in isolated
rat hearts based on different perfusion pressures. Subsequently, the influence of
myocardial edema is investigated with 23Na CSI in combination with a shift reagent
and 1H MRI in combination with a contrast agent. In Chapter 6, 23Na CSI is used to
study cellular integrity and ion homeostasis during and after myocardial ischemia in
rat heart. Three conditions were studied: acute ischemia and reperfusion, MI 2 hours
after reperfusion and chronic MI. Finally, this thesis concludes with a summarizing
discussion in chapter 7.
Reference List
1. Miura M, Matsuoka H, Saito T, Kanazawa T (1991) The Pathophysiology of Myocardial
Stunning - Reversibility, Accumulation and Continuity of the Ischemic Myocardial Damage
After Reperfusion. Japanese Circulation Journal-English Edition 55:868-877
2. Rogers WJ, Kramer CM, Geskin G, Hu YL, Theobald TM, Vido DA, Petruolo S, Reichek N
(1999) Early contrast-enhanced MRI predicts late functional recovery after reperfused
myocardial infarction. Circulation 99:744-750
3. Jansen MA, Van Emous JG, Nederhoff MG, van Echteld CJA (2004) Assessment of myocardial
viability by intracellular 23Na magnetic resonance imaging. Circulation 110:3457-3464
4. Iskandrian AS, Heo J, Stanberry C (1994) When is myocardial viability an important clinical
issue? J Nucl Med 35:4S-7S
5. Gao XM, White DA, Dart AM, Du XJ (2012) Post-infarct cardiac rupture: recent insights on
pathogenesis and therapeutic interventions. Pharmacol Ther 134:156-179
6. Bell RM, Mocanu MM, Yellon DM (2011) Retrograde heart perfusion: the Langendorff
technique of isolated heart perfusion. J Mol Cell Cardiol 50:940-950
7. Bang LE, Ripa RS, Grande P, Kastrup J, Clemmensen PM, Wagner GS (2008) Comparison of
infarct size changes with delayed contrast-enhanced magnetic resonance imaging and
electrocardiogram QRS scoring during the 6 months after acutely reperfused myocardial
infarction. J Electrocardiol 41:609-613
8. Jaffe AS, Babuin L, Apple FS (2006) Biomarkers in acute cardiac disease: the present and
the future. J Am Coll Cardiol 48:1-11
18
General Introduction
9. Schuster A, Morton G, Chiribiri A, Perera D, Vanoverschelde JL, Nagel E (2012) Imaging in
the management of ischemic cardiomyopathy: special focus on magnetic resonance. J
Am Coll Cardiol 59:359-370
10. Collins R, Cranny G, Burch J, guiar-Ibanez R, Craig D, Wright K, Berry E, Gough M, Kleijnen
J, Westwood M (2007) A systematic review of duplex ultrasound, magnetic resonance
angiography and computed tomography angiography for the diagnosis and assessment
of symptomatic, lower limb peripheral arterial disease. Health Technol Assess 11:iii-xiii, 1
11. Citro R, Salustri A, Trambaiolo P, Gregorio G (2002) [Tissue Doppler in the assessment of
myocardial function in stress echocardiography]. Ital Heart J Suppl 3:161-169
12. Paech DC, Weston AR (2011) A systematic review of the clinical effectiveness of 64-slice
or higher computed tomography angiography as an alternative to invasive coronary
angiography in the investigation of suspected coronary artery disease. BMC Cardiovasc
Disord 11:32
13. Misko J, Dziuk M, Skrobowska E, Szalus N, Pietrzykowski J, Warczynska A (2006) Coregistration of cardiac MRI and rest gated SPECT in the assessment of myocardial perfusion,
function and viability. J Cardiovasc Magn Reson 8:389-397
14. Schindler TH, Schelbert HR, Quercioli A, Dilsizian V (2010) Cardiac PET imaging for the
detection and monitoring of coronary artery disease and microvascular health. JACC
Cardiovasc Imaging 3:623-640
15. Schulz-Menger J, Gross M, Messroghli D, Uhlich F, Dietz R, Friedrich MG (2003)
Cardiovascular magnetic resonance of acute myocardial infarction at a very early stage.
Journal of the American College of Cardiology 42:513-518
16. Bax JJ (2005) Assessment of myocardial viability in ischemic cardiomyopathy. Heart Lung
Circ 14 Suppl 2:S8-13
17. Prakken NH, Velthuis BK, Cramer MJ, Mosterd A (2009) Advances in cardiac imaging: the
role of magnetic resonance imaging and computed tomography in identifying athletes at
risk. Br J Sports Med 43:677-684
18. Mari C, Strauss WH (2002) Detection and characterization of hibernating myocardium.
Nucl Med Commun 23:311-322
19. Petretta M, Costanzo P, Acampa W, Imbriaco M, Ferro A, Filardi PP, Cuocolo A (2008)
Noninvasive assessment of coronary anatomy and myocardial perfusion: going toward an
integrated imaging approach. J Cardiovasc Med (Hagerstown ) 9:977-986
20. Cummings KW, Bhalla S, Javidan-Nejad C, Bierhals AJ, Gutierrez FR, Woodard PK (2009)
A pattern-based approach to assessment of delayed enhancement in nonischemic
cardiomyopathy at MR imaging. Radiographics 29:89-103
21. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
19
Chapter 1
22. Zagrosek A, Wassmuth R, bdel-Aty H, Rudolph A, Dietz R, Schulz-Menger J (2008) Relation
between myocardial edema and myocardial mass during the acute and convalescent
phase of myocarditis - a CMR study. Journal of Cardiovascular Magnetic Resonance 10:
23. Dall’Armellina E, Karia N, Lindsay AC, Karamitsos TD, Ferreira V, Robson MD, Kellman P,
Francis JM, Forfar C, Prendergast BD, Banning AP, Channon KM, Kharbanda RK, Neubauer
S, Choudhury RP (2011) Dynamic changes of edema and late gadolinium enhancement
after acute myocardial infarction and their relationship to functional recovery and salvage
index. Circ Cardiovasc Imaging 4:228-236
24. Beyers RJ, Smith RS, Xu Y, Piras BA, Salerno M, Berr SS, Meyer CH, Kramer CM, French BA,
Epstein FH (2012) T(2) -weighted MRI of post-infarct myocardial edema in mice. Magn
Reson Med 67:201-209
25. Stork A, Muellerleile K, Bansmann PM, Graessner J, Kaul M, Kemper J, Adam G, Lund
GK (2007) Value of T2-weighted, first-pass and delayed enhancement, and cine CMR to
differentiate between acute and chronic myocardial infarction. Eur Radiol 17:610-617
26. van den Bos EJ, Baks T, Moelker AD, Kerver W, van Geuns RJ, van der Giessen WJ, Duncker
DJ, Wielopolski PA (2006) Magnetic resonance imaging of haemorrhage within reperfused
myocardial infarcts: possible interference with iron oxide-labelled cell tracking? Eur Heart
J 27:1620-1626
27. Aletras AH, Tilak GS, Natanzon A, Hsu LY, Gonzalez FM, Hoyt RF, Arai AE (2006) Retrospective
determination of the area at risk for reperfused acute myocardial infarction with T2weighted cardiac magnetic resonance imaging - Histopathological and displacement
encoding with stimulated echoes (DENSE) functional validations. Circulation 113:18651870
28. Edwards NC, Routledge H, Steeds RP (2009) T2-weighted magnetic resonance imaging to
assess myocardial oedema in ischaemic heart disease. Heart 95:1357-1361
29. Eitel I, Friedrich MG (2011) T2-weighted cardiovascular magnetic resonance in acute
cardiac disease. J Cardiovasc Magn Reson 13:13
30. Raman SV, Simonetti OP, Winner MW, Dickerson JA, He X, Mazzaferri EL, Ambrosio G
(2010) Cardiac Magnetic Resonance With Edema Imaging Identifies Myocardium at Risk
and Predicts Worse Outcome in Patients With Non-ST-Segment Elevation Acute Coronary
Syndrome. Journal of the American College of Cardiology 55:2480-2488
31. Ubachs JFA, Engblom H, Erlinge D, Jovinge S, Hedstrom E, Carlsson M, Arheden H (2010)
Cardiovascular magnetic resonance of the myocardium at risk in acute reperfused
myocardial infarction: comparison of T2-weighted imaging versus the circumferential
endocardial extent of late gadolinium enhancement with transmural projection. Journal
of Cardiovascular Magnetic Resonance 12:
32. Wright J, Adriaenssens T, Dymarkowski S, Desmet W, Bogaert J (2009) Quantification of
Myocardial Area at Risk With T2-Weighted CMR Comparison With Contrast-Enhanced CMR
and Coronary Angiography. Jacc-Cardiovascular Imaging 2:825-831
20
General Introduction
33. Coolen BF, Geelen T, Paulis LE, Nicolay K, Strijkers GJ (2011) Regional contrast agent
quantification in a mouse model of myocardial infarction using 3D cardiac T1 mapping. J
Cardiovasc Magn Reson 13:56
34. Bun SS, Kober F, Jacquier A, Espinosa L, Kalifa J, Bonzi MF, Kopp F, Lalevee N, Zaffran S,
Deharo JC, Cozzone PJ, Bernard M (2012) Value of in vivo T2 measurement for myocardial
fibrosis assessment in diabetic mice at 11.75 T. Invest Radiol 47:319-323
35. Gouya H, Vignaux O, Le RP, Chanson P, Bertherat J, Bertagna X, Legmann P (2008) Rapidly
reversible myocardial edema in patients with acromegaly: assessment with ultrafast T2
mapping in a single-breath-hold MRI sequence. AJR Am J Roentgenol 190:1576-1582
36. O’Regan DP, Ahmed R, Karunanithy N, Neuwirth C, Tan Y, Durighel G, Hajnal JV, Nadra I,
Corbett SJ, Cook SA (2009) Reperfusion hemorrhage following acute myocardial infarction:
assessment with T2* mapping and effect on measuring the area at risk. Radiology 250:916922
37. Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, Van Echteld CJ,
Pasterkamp G, Strijkers GJ (2012) Quantitative T (2) (*) assessment of acute and chronic
myocardial ischemia/reperfusion injury in mice. MAGMA
38. Van Emous JG, Schreur JH, Ruigrok TJ, Van Echteld CJ (1998) Both Na+-K+ ATPase and Na
+-H+ exchanger are immediately active upon post-ischemic reperfusion in isolated rat
hearts. J Mol Cell Cardiol 30:337-348
39. Ouwerkerk R, Bleich KB, Gillen JS, Pomper MG, Bottomley PA (2003) Tissue sodium
concentration in human brain tumors as measured with 23Na MR imaging. Radiology
227:529-537
40. Ridgway JP (2010) Cardiovascular magnetic resonance physics for clinicians: part I. J
Cardiovasc Magn Reson 12:71
41. Kim RJ, Lima JA, Chen EL, Reeder SB, Klocke FJ, Zerhouni EA, Judd RM (1997) Fast 23Na
magnetic resonance imaging of acute reperfused myocardial infarction. Potential to
assess myocardial viability. Circulation 95:1877-1885
42. Constantinides CD, Kraitchman DL, O’Brien KO, Boada FE, Gillen J, Bottomley PA (2001)
Noninvasive quantification of total sodium concentrations in acute reperfused myocardial
infarction using 23Na MRI. Magn Reson Med 46:1144-1151
43. Kim RJ, Judd RM, Chen EL, Fieno DS, Parrish TB, Lima JA (1999) Relationship of elevated
23Na magnetic resonance image intensity to infarct size after acute reperfused myocardial
infarction. Circulation 100:185-192
21
Part I
Proton MRI
23
Chapter 2
Quantitative T2* assessment of acute and
chronic myocardial ischemia/reperfusion
injury in mice
Eissa N. E. Aguor, Fatih Arslan, Cees W.A. van de Kolk, Marcel G.J. Nederhoff,
Pieter A. Doevendans, Cees J.A. van Echteld, Gerard Pasterkamp, Gustav J. Strijkers
MAGMA. 2012, DOI 10.1007/s10334-012-0304-0
25
Chapter 2
Abstract
Object Imaging of myocardial infarct composition is essential to assess efficacy of
emerging therapeutics. T2* mapping has the potential to image myocardial hemorrhage
and fibrosis by virtue of its short T2*. We aimed to quantify T2* in acute and chronic
myocardial ischemia/reperfusion (I/R) injury in mice.
Materials and methods I/R‑injury was induced in C57BL/6 mice (n=9). Sham‑operated
mice (n=8) served as controls. MRI was performed at baseline, and 1, 7 and 28 days after
surgery. MRI at 9.4 T consisted of Cine, T2*‑mapping and late-gadolinium-enhancement
(LGE). Mice (n=6) were histologically assessed for hemorrhage and collagen in the
fibrotic scar.
Results Baseline T2*‑values were 17.1±2.0 ms. At day 1, LGE displayed a homogeneous
infarct enhancement. T2* in infarct (12.0±1.1 ms) and remote myocardium (13.9±0.8 ms)
was lower than at baseline. On days 7 and 28, LGE was heterogeneous. T2* in the infarct
decreased to 7.9±0.7 ms and 6.4±0.7 ms, whereas T2*‑values in the remote myocardium
were 14.2±1.1 ms and 15.6±1.0 ms. Histology revealed deposition of iron and collagen
in parallel with decreased T2*.
Conclusion T2*‑values are dynamic during infarct development and decrease
significantly during scar maturation. In the acute phase, T2* values in infarcted
myocardium differ significantly from those in the chronic phase. T2*‑mapping was able
to confirm the presence of a chronic infarction in cases where LGE was inconclusive.
Hence, T2* may be used to discriminate between acute and chronic infarctions.
Keywords cardiovascular magnetic resonance, myocardial infarction, T2*, LGE, mouse,
fibrosis, hemorrhage
26
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
2.1 Introduction
Despite recent advances in therapeutics and diagnostics, morbidity and mortality
after myocardial infarction (MI) remains a great socio-economic burden in Western
societies. Early identification of patients with sub-optimal cardiac recovery is of
great importance to improve clinical management. For this reason, both functional
and structural characterization of the infarcted myocardium is important for risk
stratification of patients after MI.
Technological advances in Cardiovascular Magnetic Resonance (CMR) have
demonstrated different ways to track and characterize MI at different time points
[1-6]. CMR has been applied to analyze inflammation, edema, necrosis, fibrosis,
left-ventricular (LV) remodeling and metabolism in myocardial infarction [7-15].
Pathological changes in the injured myocardium, such as the presence of fibrosis,
hemorrhage and edema, influence the signal intensity on T1‑ and T2‑weighted imaging
[7,16-18]. CMR is thus considered as one of the most important imaging tools in the
assessment of the heart after MI.
Typically, T1-weighted imaging is performed some time after injecting a
gadolinium‑based contrast agent to delineate injured myocardium in a Late
Gadolinium Enhancement (LGE) scan [19]. The enhancement mechanism of infarcted
tissue is related to the extent of the extracellular volume, which is associated with the
delayed washout relative to the healthy myocardium. LGE is widely used and without
competition considered the gold standard to assess infarct size. Nevertheless, there
are certain limitations of LGE related to pathological processes after the onset of
infarction. In the acute and sub-acute phase after successful reperfusion of the ischemic
myocardium, the transmural infarct extent is not stable and an overestimation of
infarct size has been reported [20]. In the chronic phase, excessive and heterogeneous
collagen deposition during scar maturation may hamper signal enhancement by LGE.
In view of the above limitations, several alternative contrast mechanisms are explored
to study the pathophysiological changes in the injured myocardium. The sensitivity
27
Chapter 2
of the T2 relaxation time to myocardial water content has led to the application of
T2‑weighted CMR protocols to image myocardial edema as an indicator of inflammation
and necrosis in acute MI [13,15,21-24]. T2* contrast has been explored as a tool to
diagnose the presence of hemorrhage and myocardial iron [1-5,25-28]. Recently, it was
demonstrated that T2* has the potential to image fibrosis in the infarct zone using an
ultra-short echo time (UTE) technique [29].
In this paper, we hypothesize that the structural and pathological changes in acute
and chronic MI, such as the presence and development of myocardial hemorrhage
and fibrotic tissue progressively influence T2* values in the infarcted myocardium.
To investigate this, we quantified T2* using a T2*‑mapping protocol longitudinally in
acute and chronic ischemia/reperfusion (I/R) infarcts in mice. We explored the utility of
T2* mapping as a complementary tool to LGE to characterize the infarcted myocardium.
2.2 Materials and methods
Experimental design
C57Bl6/J mice (n=23, body weight=20–25 g) were used in this study. CMR measurements
were performed at baseline (BL), 1 day (acute infarct), 7 days and 28 days (chronic
infarct) after induction of myocardial infarction (MI) by 30 min transient ligation of
the left coronary artery (n=9). A sham‑operated group (n=8) served as control. The
CMR protocol for the infarct group consisted of a multi‑gradient‑echo quantitative
T2*‑mapping sequence, traditional LGE to assess the location and size of the infarct,
and gradient‑echo Cine imaging to determine cardiac function. The sham-operated
(control) group underwent gradient‑echo Cine imaging only to evaluate cardiac
function. In the first group of 9 mice with MI, 1 mouse died at day 28 during the CMR
scan. Another 3 mice of this group presented a small infarct on LGE images at day 1,
but no chronic myocardial infarct at day 7 and 28 (as exemplified in Fig. 3). A separate
set of mice with myocardial infarction (n=6) were sacrificed either at day 1 (n=2),
7 (n=2), or 28 (n=2) for the histopathological characterization of myocardial tissues.
All animal experiments were performed in accordance with the national guidelines
28
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
on animal care and with prior approval by the Animal Experimentation Committee of
Utrecht University.
Mouse model of myocardial I/R injury
Mice underwent myocardial I/R surgery as previously described [30]. Briefly, mice
were anesthetized with a mixture of Fentanyl (Jansen-Cilag) 0.05 mg/kg, Dormicum
(Roche) 5 mg/kg, and medetomidine 0.5 mg/kg through an intraperitoneal injection.
A core body temperature of about 37 °C was maintained during surgery by continuous
monitoring with a rectal thermometer and an automatic heating blanket. Mice were
intubated and ventilated (Harvard Apparatus Inc, Holliston, Mass) with 100% oxygen.
The left coronary artery (LCA) was ligated for 30 min with an 8-0 Ethilon (Ethicon)
suture with a section of polyethylene-10 tubing placed over the left coronary
artery. Ischemia was confirmed by bleaching of the myocardium and ventricular
tachyarrhythmia. Reperfusion was initiated by release of the ligature and removal of
the polyethylene-10 tubing. In sham-operated animals, the suture was placed beneath
the left coronary artery without ligation. Reperfusion of the endangered myocardium
was characterized by typical hyperemia in the first few minutes. The chest wall was
closed, and the animals subcutaneously received Antisedan (Pfizer) 2.5 mg/kg, Anexate
(Roche) 0.5 mg/kg, and Temgesic (Schering-Plough) 0.1 mg/kg.
CMR protocol
A vertical 9.4 T and 89-mm-diameter bore scanner was used, equipped with 1500 mT/m
gradients (Bruker BioSpin GmbH, Ettlingen, Germany). Mice were anesthetized with
1.5‑3 vol% isoflurane in a 2:1 mixture of air (0.3 l/min) and oxygen (0.15 l/min). The
mice were put on a custom‑built cradle and positioned inside a 3-cm‑diameter
quadrature volume coil. During the CMR examination anesthesia was maintained at
1.5‑2.5 vol% isoflurane to keep the respiratory rate at 50-60 bpm. We did not use ECG
leads for cardiac triggering but derived both cardiac and respiratory motion signals
from a pressure pad placed under the chest of the mouse.
29
Chapter 2
The CMR protocol consisted of the following steps. After global shimming and RF pulse
calibration, 2- and 4-chamber view scout scans were made to plan a single mid-cavity
short-axis slice. Cine imaging was performed with a retrospectively‑triggered (selfgated) gradient‑echo sequence [31], with the following parameters: TR = 6.8 ms,
TE = 1.9 ms, number of movie frames = 15, slice thickness = 1 mm, matrix = 256x256,
field‑of‑view = 3x3 cm2. Because of the slight difference of hearts size between
mice and the dilation of each heart after infarction, seven to 9 short‑axis slices with
inter‑slice distance of 1 mm and no slice gap were measured to cover the heart from
apex to base. Based on the Cine short‑axis images, a mid axial slice was selected
that included an area of remote viable tissue as well as a substantial infarct area.
T2* mapping was performed in this slice using a cardiac‑triggered multi‑gradient-echo
sequence with the following parameters: TR = R‑R interval, TE = 1.22, 3.45, 5.68, 7.91,
10.14, and 12.37 ms, slice thickness = 1 mm, matrix = 128x128, field‑of‑view = 3x3 cm2.
The trigger delay was chosen in the mid-diastolic rest period of the heart cycle
observed in the Cine series. The acquisition window of less than 13 ms resulted in
images, which were sufficiently co-registered for the different echo times to allow
for a pixel‑wise T2* determination (Fig. 1) [32]. LGE scans were performed in the same
slice with a cardiac‑triggered inversion‑recovery segmented gradient‑echo sequence,
with the following parameters: TR = 5.8 ms, TE = 2.2 ms, number of segments = 16,
slice thickness = 1 mm, matrix = 256x256, field‑of‑view = 3x3 cm2. The inversion time
(TI) was optimized to null the signal from remote myocardium [33,34]. Gadobutrol
(Gadovist, Bayer Schering Pharma AG, Berlin, Germany) dose of 1 mmol/ml was used.
To reduce volume related variability, contrast agent was diluted 1:4 with a standard
saline solution. This dilution resulted in administered weight-adapted volume of 2 µl/g
body weight, resulting in an equivalent dose of 0.5 mmol/kg body weight [33]. The
contrast agent was injected intravenously through the tail vein, using a 24-gauge
catheter.
CMR data analysis
T2* maps were generated in Mathematica 7 (Wolfram Research, Champaign, IL, USA)
by pixel‑wise fittings of the multi-gradient-echo signal intensities to the equation
S = S0 e –TE/T2*. The LV myocardium was manually segmented by drawing epicardial and
30
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
endocardial contours, excluding the papillary muscles. LGE‑based single-slice infarct
size is reported as percentage of the LV myocardial cross‑sectional area. A region of
interest in the infarct in the myocardial free wall (infarct ROI) was drawn based on
gadolinium enhancement and presence of akinetic myocardium. A second ROI was
drawn in the non‑enhancing remote tissue in the septum (remote ROI). For the control
mice, ROIs were drawn in the septum and the free wall. These ROIs were drawn within
the zone free from susceptibility artifacts induced by lungs. T2* values are reported
as group average and standard deviation of the mean T2* in the ROIs. Cine images
were analyzed with Qmass digital imaging software (Medis, Leiden, the Netherlands).
Semi‑automatic segmentation was used to determine end-diastolic volume (EDV),
end‑systolic volume (ESV), wall thickness (WT) and systolic wall thickening (SWT).
Ejection fraction (EF) was calculated as 100%(EDV‑ESV)/EDV.
Histopathology
For histological evaluation a 1.5‑mm‑thick slice containing infarcted myocardium was
fixed in 4% formalin and paraffin‑embedded. Hereafter, 5‑µm‑thick sections were
mounted on glass slides and stained with Picrosirius red staining for collagen, Prussian
blue staining for iron, and Hematoxylin-Eosin (HE) for cellular structures. The stained
sections were photographed with a digital image microscopy using white or polarized
light.
Statistical analysis
All statistical analysis was evaluated using (SPSS 17 for Windows; SPSS). Results were
expressed as mean ± standard deviation for each group of animals. An independent
two-tailed sample t-test and the Mann-Whitney exact two-tailed U test were
performed to compare mean values in subjects for each time point post MI versus
baseline. Results were considered statistically significant for a p‑value of less than 0.05.
The correlation analysis of T2* values against wall thickness (WT) was performed using
linear regression.
31
Chapter 2
2.3Results
An image series of a mid-axial slice with varying TE for a representative mouse at
baseline is shown in Fig. 1a. The LV myocardium was well co‑registered for the different
images. Signal loss due to susceptibility differences was occasionally observed
for the images with the longest TE. Pixel‑wise fitting of the signal intensities in the
LV myocardium as function of TE resulted in a T2* map as shown in Fig. 1b. T2* values
were homogeneous throughout the left-ventricular myocardial wall and consistent
between mice, with a mean T2* value of 17.1±2.0 ms in the free wall.
Figure 2 displays a collection of images, consisting of LGE, T2* maps, and Cine images at
diastole, for a representative mouse at day 1, 7 and 28 after induction of the myocardial
infarction. On day 1, LGE resulted in a homogeneous enhancement of the infarct in
Figure 1. Representative series of cardiac images with varying TE and corresponding T2*
map. (a) Gradient-echo CMR with varying TE for a mid‑ventricle short-axis slice of a
representative mouse at baseline. (b) Top: Signal intensity S in the myocardium as
function of TE. The solid line is an exponential fit to the data. Bottom: T2* map. T2* in
the LV myocardium at end‑diastole is color-coded between 0 and 20 ms according
to the pseudo color scale on the right. The background image is a regular gradient
echo image with TE = 1.22 ms.
32
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
the mid-anterior and mid-anterolateral wall, whereas on days 7 and 28 enhanced area
was smaller, heterogeneous, and difficult to detect. Corresponding T2* maps revealed
a different picture. At day 1, a significant change in T2* could be detected in the LGEbased infarct area. Further, at days 7 and 28, T2* dramatically decreased, particularly
in the mid‑anterolateral region indicating the existence of a significant infarct. The
presence of an infarct was further corroborated by the Cine images, which showed
increased wall thickness (WT) in the mid-anterior and mid-anterolateral wall at day 1,
followed by progressive thinning below baseline levels at days 7 and 28, respectively.
Figure 2. Mid-ventricle short-axis slice of a mouse heart at 1, 7, and 28 days after I/R
myocardial infarction. (left) Cine images at end‑diastole. At day 1 wall thickness
in the free wall is increased, whereas at days 7 and 28 the free wall decreased in
thickness. (middle) Corresponding T2* maps color-coded from 0 to 20 ms. T2*
is decreased in the infarct in the free wall at day 7 and 28 (red arrows). (right)
Corresponding LGE images. Yellow arrows point to the location of the infarct. Infarct
location was confirmed by the presence of an akinetic area and locally reduced wall
thickening in Cine imaging.
Figure 3 shows LGE and T2* maps of one of three mice with a small infarct in the midanterolateral wall, as indicated by a small area of LGE at day 1. Gadolinium enhancement
33
Chapter 2
could not be detected anymore at days 7 and 28. In the first group of n=9 mice with MI,
n=3 mice displayed this atypical behavior. Presumably, LGE at day 1 revealed a small
area at risk that could not be detected in the chronic phase. Importantly, T2* values in
the mid-anterolateral wall also remained unaffected throughout infarct healing and
did not deviate from baseline values.
Figure 3. T2* maps and LGE of a mouse with
a small LGE positive area at day 1
but no detectable infarct at later
days. LGE showed enhancement
in the mid‑anterior wall at day
1 (yellow arrow).
Enhancement
however, was absent at 7 and 28
days. In parallel, T2* values in the
infarct area were maintained at
baseline levels. T2* maps displayed
spurious lowering in areas with
no infarct related to susceptibility
artifacts (red arrows).
Figure 4. Quantitative T2* values in the
infarct and remote myocardium
at baseline (BL) and up to 28
days after surgery. n=5 mice, * =
significantly different from baseline
(p < 0.05), and ** = significantly
different from baseline (p < 0.005).
Quantitative T2* values (n=5 mice, Fig. 4) were dynamic throughout infarct development.
In the infarct, quantitative T2* values decreased significantly (p < 0.05) from baseline
(17.1±2.0 ms) to day 1 (12.0±1.1 ms) followed by a further decrease (p < 0.005) both
34
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
at day 7 (7.9±1.0 ms) and day 28 (6.4±0.7 ms). In addition, the mid-septal (remote)
area showed a significant (p < 0.05) decrease of T2* values from baseline (18.2±2.0 ms)
to day 1 (13.9±0.8 ms), then slightly recovered at day 7 (14.2±1.1 ms) and day 28
(15.6±1.0 ms).
Global cardiac function of both infarct and sham-operated groups was assessed by
Cine imaging (Figs. 5a-c). The infarct group displayed a progressive and significant
increase in end-diastolic volume (EDV) and end-systolic volume (ESV), indicative for a
considerable dilation of the heart due to the infarct. For the sham-operated group no
significant changes in EDV and ESV upon time after surgery were observed. The infarct
group presented a decline in ejection fraction (EF) from 54±3 % at baseline to 39±2 %,
28 days after surgery (Fig. 5c). For the sham-operated group, EF was slightly elevated at
day 1, however normalized at day 7 and 28 after surgery. Group-averaged (n=5) infarct
sizes from LGE analysis of a mid-ventricular slice as function of days after surgery are
Figure 5. Left ventricle functional parameters and infarct size for infarct (n=5) and shamoperated (n=8) groups at baseline (BL) and up to 28 days after surgery. (a) End
diastolic volume. (b) End systolic volume. (c) Ejection fraction. (d) Single-slice infarct
size. * = significantly different from baseline (p < 0.05) , and ** = significantly different
from baseline (p < 0.005).
35
Chapter 2
presented in Fig. 5d. At day 1, LGE revealed an average infarct size of 41±6 %, dropping
to 26±3 % and 22±4 % (p < 0.05, compared to day 1) of the LV area at days 7 and 28,
respectively, which could be related to infarct resorption over time [35].
Figure 6. Regional wall thickness (WT) and systolic wall thickening (SWT) for infarct
(n=5) and sham-operated (n=8) groups. (a) WT for infarct group in remote and
infarct wall. (b) SWT for infarct group in remote and infarct wall. (c) WT for shamoperated group in septal and free wall. (d) SWT for sham‑operated group in septal
and free wall. * = significantly different from baseline (p < 0.05) , and ** = significantly
different from baseline (p < 0.005).
In the infarct group, WT in the free wall (infarct area) showed a significant (p < 0.005)
transient increase day 1 after myocardial infarction, followed by progressive thinning
(Fig. 6a). Decreased WT in the infarct paralleled a decrease of the quantitative T2*
values (Fig. 4). The mid-septal (remote) area in the infarct group showed no significant
changes in WT. In line with the increased WT of the free wall at day 1, the systolic wall
thickening (SWT), as a measure of local systolic function, also displayed a transient
depression 1 day after surgery, only partially recovering at days 7 and 28 (Fig. 6b). Also,
36
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
SWT of the remote area was slightly depressed at all time points compared to baseline.
The sham-operated group revealed no significant changes in both WT and SWT as
function of days after surgery (Figs. 6c and d). WT in the infarct correlated positively
and significantly with T2* (Fig. 7). Thus, the chronic MI is characterized by a lower T2*
value compared to the acute MI.
Figure 7. Correlation between infarct wall thickness (WT) and T2*.
Figure 8. Representative histological images of infarct myocardial tissue at day 1, 7 and
28 post MI. (a) Picrosirius red. (b) Prussian-blue. (c) Hematoxylin-Eosin. Scale bar
equals 200 μm.
37
Chapter 2
Since T2* changes are related to the structural and compositional changes in the
infarct myocardial tissue, we performed histology at each time point after I/R injury.
Figure 8 shows images of 3 representative mice at day 1, 7 and 28 after MI. Picrosirius
staining (Fig. 8a) revealed no collagen deposits in the infarct area at day 1. At day 7
substantial collagen deposition was observed, followed by a significant increase at day
28. Sections stained with Prussian Blue showed scattered spots of iron as early as 1
day after MI (Fig. 8b). Iron deposition became more prominent at day 7 and further
increased at day 28. Finally, on sections stained with Hematoxylin-Eosin (Fig. 8c) the
infarct area on day 1 was characterized by infiltration of inflammatory cells.
2.4Discussion
In this study we aimed to systematically study quantitative changes in T2* of the
mouse myocardium up to 28 days post I/R injury and to evaluate its added value in
characterizing chronic infarcts in comparison to commonly applied LGE and Cine
imaging. Combining different CMR techniques, if interpreted carefully, are a potential
advantage in the evaluation of the pathophysiological changes in the injured
myocardium. Baseline T2* values were comparable to T2 values found at 9.4 T [36] and
11.75 T [37], which suggests that decay due to static magnetic field inhomogeneities
is largely absent in the healthy myocardium. We found that quantitative T2* values
decreased progressively during infarct development. Throughout the observation
period, the local decrease of T2* in the infarct was accompanied by globally and locally
reduced ventricular function (Figs. 4-7). The reduction in T2* with infarct age could
be explained by the presence of iron (Fig. 8a) [3,38]. Additionally, the formation of
significant amounts of collagen (Fig. 8b) may contribute to the progressive decline of
T2* with infarct age [29].
LGE imaging is currently unchallenged the gold standard technique to assess infarct
size. Nevertheless, in certain cases when LGE imaging is inconclusive, additional
characterization of the infarct by T2* imaging could provide complementary
information. In the acute phase after I/R injury LGE‑based infarct size may be inaccurate
38
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
[39], when there is loss of cell membrane integrity [16], an inflammatory response
[40,41], necrosis, hemorrhage, micro‑vascular obstruction (MVO) and edema. Indeed,
for the acute infarcts at day 1 we found a large LGE-based infarct size accompanied
by increased wall-thickness, suggesting the presence of edema. We did not observe a
transient increase in T2* at day 1 by the presence of this edema, as previously observed
with T2‑weighted imaging [24]. The effect of edema on T2* is apparently lower than
the decrease of T2* induced by iron or collagen (Fig. 8). This observation is critical in
the clinical setting where early assessment of viability is significantly hampered by
the presence of infarct-induced edema. T2*-imaging may be a tool to circumvent this
inevitable problem after acute MI.
In the chronic infarct, the formation of scar tissue could be responsible for a reduced
distribution volume for the extracellular contrast agent. In turn, reduced distribution
volume leads to an apparent reduction in infarct size and inhomogeneous
enhancement (Figs. 2 and 5). Additionally, the extremely low T2* in the infarct (Fig. 2),
which locally decreased down to a few milliseconds, could effectively null the signal of
the inversion recovery gradient‑echo sequence with an echo time of 2.2 ms, obscuring
the late gadolinium enhancement [42].
At day 1, a significant decrease of T2* (Fig. 4) and a transient depression of systolic
performance (SWT; Fig. 6) were found not only in the infarct, but also to a lesser extent
in the remote myocardium. This could be a secondary response in parallel to the
structural changes in the infarct region. These findings are in agreement with those
of Bogaert et al., who observed a dysfunction in the remote area at 5±2 days after
reperfusion in patients with transmural anterior MI [43]. Others have also shown that
remodeling is initiated in the remote myocardium as early as 1 day after I/R injury [4447].
Comparing Figs. 2 and 3 makes an interesting case. Figure 2 presented an example of a
mouse with a considerable infarct at day 1, based on a large LGE positive area. However,
at days 7 and 28, the infarct was difficult to detect on LGE, from which one could jump
to the conclusion that a considerable part of the area at risk at day 1 recovered at
39
Chapter 2
later time points. However, the T2* maps at day 7 and 28 revealed low T2* values in the
infarct area, indicating that the infarct was not recovered. The presence of non-viable
myocardium (not detected by LGE) was confirmed by akinesia of that area and locally
suppressed systolic thickening (SWT) in Cine imaging. Figure 3, conversely, showed one
of three mice that presented a small infarct by LGE at day 1, but displayed no positive
enhancement at day 7 and 28. For this mouse T2* in the infarct region was preserved
during follow-up, indicating that for this mouse the initial area at risk recovered at later
time points, as confirmed by healthy LV function. Thus, T2* is able to provide additional
information, complementary to LGE, about the myocardial viability after infarction.
Taken together, these findings suggest that T2* can be used as a surrogate marker for
myocardial recovery during follow-up after myocardial infarction.
A limitation to the use of T2* as a diagnostic marker for myocardial injury is contamination
of intrinsic T2* values by the presence of macroscopic susceptibility differences, mainly
present at the air-lung interfaces [26,27]. Occasionally, we observed lowered T2* values
at the lung-heart interface (see for example Fig. 3) in areas with no infarct and no wallmotion abnormalities. Estimations of infarct sizes on the basis of T2* reduction were
therefore inaccurate. Also the orientation of the heart with respect to the magnetic
field could alter absolute T2* values. Improvements in local shimming techniques
might alleviate some of these susceptibility contaminations. Still, use of T2* mapping
as a stand-alone technique without the complementary information from LGE is
undesired.
2.5Conclusions
Imaging of the myocardium with a T2* mapping technique in a mouse model of
myocardial I/R injury provides useful and complementary information to LGE and Cine
imaging. The changes of the T2* were associated with the infarct age, reversibility of the
injury, and gave an indication of the structural variation in the myocardium throughout
infarct maturation. T2* decrease in the infarcts was associated with iron and collagen
depositions. In cases where LGE was inconclusive, T2* mapping was able to confirm
40
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
the presence of a chronic myocardial scar in mice. Taken together, T2* mapping may
provide important additional information on myocardial viability in the acute and
chronic phase and scar maturation after acute MI.
Acknowledgements
This research is supported by the Netherlands Heart Foundation (NHS) grant 2006B066
(CvE) and 2010T001 (FA), as well as by the Netherlands Heart Institute (ICIN). The
work of GJS is supported by the Dutch Technology Foundation STW, Applied Science
Division of NWO and the Technology Program of the Ministry of Economic Affairs
(grant number 07952). We would like to thank Maringa Emons and Chaylendra Strijder
for biotechnical assistance and histology.
Reference List
1.
Saraste A, Nekolla S, Schwaiger M (2008) Contrast-enhanced magnetic resonance imaging
in the assessment of myocardial infarction and viability. J Nucl Cardiol 15:105-117
2.
Schulz-Menger J, Gross M, Messroghli D, Uhlich F, Dietz R, Friedrich MG (2003)
Cardiovascular magnetic resonance of acute myocardial infarction at a very early stage. J
Am Coll Cardiol 42:513-518
3.
van den Bos EJ, Baks T, Moelker AD, Kerver W, van Geuns RJ, van der Giessen WJ, Duncker
DJ, Wielopolski PA (2006) Magnetic resonance imaging of haemorrhage within reperfused
myocardial infarcts: possible interference with iron oxide-labelled cell tracking? Eur Heart
J 27:1620-1626
4.
Walsh TF, Hundley WG (2007) Assessment of ventricular function with cardiovascular
magnetic resonance. Magn Reson Imaging Clin N Am 15:487-50410
5.
Sosnovik DE, Nahrendorf M, Weissleder R (2007) Molecular magnetic resonance imaging
in cardiovascular medicine. Circulation 115:2076-2086
6.
Bilchick KC, Dimaano V, Wu KC, Helm RH, Weiss RG, Lima JA, Berger RD, Tomaselli GF,
Bluemke DA, Halperin HR, Abraham T, Kass DA, Lardo AC (2008) Cardiac magnetic
resonance assessment of dyssynchrony and myocardial scar predicts function class
improvement following cardiac resynchronization therapy. JACC Cardiovasc Imaging
1:561-568
7.
Eitel I, Friedrich MG (2011) T2-weighted cardiovascular magnetic resonance in acute
cardiac disease. J Cardiovasc Magn Reson 13:13
41
Chapter 2
8.
Raman SV, Simonetti OP, Winner MW, 3rd, Dickerson JA, He X, Mazzaferri EL, Jr., Ambrosio
G (2010) Cardiac magnetic resonance with edema imaging identifies myocardium at risk
and predicts worse outcome in patients with non-ST-segment elevation acute coronary
syndrome. J Am Coll Cardiol 55:2480-2488
9.
Edwards NC, Routledge H, Steeds RP (2009) T2-weighted magnetic resonance imaging to
assess myocardial oedema in ischaemic heart disease. Heart 95:1357-1361
10.
Bandettini WP, Arai AE (2008) Advances in clinical applications of cardiovascular magnetic
resonance imaging. Heart 94:1485-1495
11. Saborowski O, Saeed M (2007) An overview on the advances in cardiovascular
interventional MR imaging. Magn Reson Mater Phy 20:117-127
12. Gouya H, Vignaux O, Le Roux P, Chanson P, Bertherat J, Bertagna X, Legmann P (2008)
Rapidly reversible myocardial edema in patients with acromegaly: assessment with
ultrafast T2 mapping in a single-breath-hold MRI sequence. AJR Am J Roentgenol
190:1576-1582
13.
Mikami Y, Sakuma H, Nagata M, Ishida M, Kurita T, Komuro I, Ito M (2009) Relation between
signal intensity on T2-weighted MR images and presence of microvascular obstruction in
patients with acute myocardial infarction. AJR Am J Roentgenol 193:W321-326
14.
Giri S, Chung YC, Merchant A, Mihai G, Rajagopalan S, Raman SV, Simonetti OP (2009) T2
quantification for improved detection of myocardial edema. J Cardiovasc Magn Reson
11:56
15. Stork A, Muellerleile K, Bansmann PM, Graessner J, Kaul M, Kemper J, Adam G, Lund
GK (2007) Value of T2-weighted, first-pass and delayed enhancement, and cine CMR to
differentiate between acute and chronic myocardial infarction. Eur Radiol 17:610-617
16.
Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
17.
Berry C, Kellman P, Mancini C, Chen MY, Bandettini WP, Lowrey T, Hsu LY, Aletras AH, Arai AE
(2010) Magnetic resonance imaging delineates the ischemic area at risk and myocardial
salvage in patients with acute myocardial infarction. Circ Cardiovasc Imaging 3:527-535
18.
Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA (2011) Assessment of myocardial fibrosis
with cardiovascular magnetic resonance. J Am Coll Cardiol 57:891-903
19. Ibrahim T, Hackl T, Nekolla SG, Breuer M, Feldmair M, Schomig A, Schwaiger M (2010)
Acute myocardial infarction: serial cardiac MR imaging shows a decrease in delayed
enhancement of the myocardium during the 1st week after reperfusion. Radiology
254:88-97
20. Merten C, Steen H, Kulke C, Giannitsis E, Katus HA (2008) Contrast-enhanced magnetic
resonance imaging reveals early decrease of transmural extent of reperfused acute
myocardial infarction. Clin Res Cardiol 97:913-916
42
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
21. Wright J, Adriaenssens T, Dymarkowski S, Desmet W, Bogaert J (2009) Quantification of
myocardial area at risk with T2-weighted CMR: comparison with contrast-enhanced CMR
and coronary angiography. JACC Cardiovasc Imaging 2:825-831
22.
Ibanez B, Prat-Gonzalez S, Speidl WS, Vilahur G, Pinero A, Cimmino G, Garcia MJ, Fuster V,
Sanz J, Badimon JJ (2007) Early metoprolol administration before coronary reperfusion
results in increased myocardial salvage: analysis of ischemic myocardium at risk using
cardiac magnetic resonance. Circulation 115:2909-2916
23. Friedrich MG, Abdel-Aty H, Taylor A, Schulz-Menger J, Messroghli D, Dietz R (2008)
The salvaged area at risk in reperfused acute myocardial infarction as visualized by
cardiovascular magnetic resonance. J Am Coll Cardiol 51:1581-1587
24.
Beyers RJ, Smith RS, Xu Y, Piras BA, Salerno M, Berr SS, Meyer CH, Kramer CM, French BA,
Epstein FH (2012) T2-weighted MRI of post-infarct myocardial edema in mice. Magn Reson
Med 67:201-209
25.
Beek AM, Nijveldt R, van Rossum AC (2010) Intramyocardial hemorrhage and microvascular
obstruction after primary percutaneous coronary intervention. Int J Cardiovasc Imaging
26:49-55
26.
Ghugre NR, Ramanan V, Pop M, Yang Y, Barry J, Qiang B, Connelly KA, Dick AJ, Wright GA
(2011) Quantitative tracking of edema, hemorrhage, and microvascular obstruction in
subacute myocardial infarction in a porcine model by MRI. Magn Reson Med 66:11291141
27.
Leonardi B, Margossian R, Colan SD, Powell AJ (2008) Relationship of magnetic resonance
imaging estimation of myocardial iron to left ventricular systolic and diastolic function in
thalassemia. JACC Cardiovasc Imaging 1:572-578
28. O’Regan DP, Ahmed R, Karunanithy N, Neuwirth C, Tan Y, Durighel G, Hajnal JV, Nadra I,
Corbett SJ, Cook SA (2009) Reperfusion hemorrhage following acute myocardial infarction:
assessment with T2* mapping and effect on measuring the area at risk. Radiology 250:916922
29.
de Jong S, Zwanenburg JJ, Visser F, der Nagel R, van Rijen HV, Vos MA, de Bakker JM, Luijten
PR (2011) Direct detection of myocardial fibrosis by MRI. J Mol Cell Cardiol 51:974-979
30. Arslan F, Smeets MB, O’Neill LA, Keogh B, McGuirk P, Timmers L, Tersteeg C, Hoefer IE,
Doevendans PA, Pasterkamp G, de Kleijn DP (2010) Myocardial ischemia/reperfusion injury
is mediated by leukocytic toll-like receptor-2 and reduced by systemic administration of a
novel anti-toll-like receptor-2 antibody. Circulation 121:80-90
31.
Heijman E, de Graaf W, Niessen P, Nauerth A, van Eys G, de Graaf L, Nicolay K, Strijkers GJ
(2007) Comparison between prospective and retrospective triggering for mouse cardiac
MRI. NMR Biomed 20:439-447
32.
Coolen BF, Moonen RP, Paulis LE, Geelen T, Nicolay K, Strijkers GJ (2010) Mouse myocardial
first-pass perfusion MR imaging. Magn Reson Med 64:1658-1663
43
Chapter 2
33.
Bohl S, Lygate CA, Barnes H, Medway D, Stork LA, Schulz-Menger J, Neubauer S, Schneider
JE (2009) Advanced methods for quantification of infarct size in mice using threedimensional high-field late gadolinium enhancement MRI. Am J Physiol Heart Circ Physiol
296:H1200-1208
34. Protti A, Sirker A, Shah AM, Botnar R (2010) Late gadolinium enhancement of acute
myocardial infarction in mice at 7T: cine-FLASH versus inversion recovery. J Magn Reson
Imaging 32:878-886
35. Lund GK, Stork A, Muellerleile K, Barmeyer AA, Bansmann MP, Knefel M, Schlichting U,
Muller M, Verde PE, Adam G, Meinertz T, Saeed M (2007) Prediction of left ventricular
remodeling and analysis of infarct resorption in patients with reperfused myocardial
infarcts by using contrast-enhanced MR imaging. Radiology 245:95-102
36. Coolen BF, Simonis FF, Moonen RP, Geelen T, Paulis LE, Nicolay K, Strijkers GJ (2011) T2
mapping of the mouse heart using segmented MLEV supercycle preparation. In: ISMRM
19th Scientific Meeting & Exhibition. Montreal, Canada
37. Schneider JE, Cassidy PJ, Lygate C, Tyler DJ, Wiesmann F, Grieve SM, Hulbert K, Clarke
K, Neubauer S (2003) Fast, high-resolution in vivo cine magnetic resonance imaging
in normal and failing mouse hearts on a vertical 11.7 T system. J Magn Reson Imaging
18:691-701
38.
Pepe A, Positano V, Santarelli MF, Sorrentino F, Cracolici E, De Marchi D, Maggio A, Midiri M,
Landini L, Lombardi M (2006) Multislice multiecho T2* cardiovascular magnetic resonance
for detection of the heterogeneous distribution of myocardial iron overload. J Magn
Reson Imaging 23:662-668
39. Schwitter J, Saeed M, Wendland MF, Derugin N, Canet E, Brasch RC, Higgins CB (1997)
Influence of severity of myocardial injury on distribution of macromolecules: extravascular
versus intravascular gadolinium-based magnetic resonance contrast agents. J Am Coll
Cardiol 30:1086-1094
40. Murphy E, Steenbergen C (2008) Mechanisms underlying acute protection from cardiac
ischemia-reperfusion injury. Physiol Rev 88:581-609
41. Arslan F, de Kleijn DP, Timmers L, Doevendans PA, Pasterkamp G (2008) Bridging innate
immunity and myocardial ischemia/reperfusion injury: the search for therapeutic targets.
Curr Pharm Des 14:1205-1216
42.
Bremerich J, Colet JM, Giovenzana GB, Aime S, Scheffler K, Laurent S, Bongartz G, Muller
RN (2001) Slow clearance gadolinium-based extracellular and intravascular contrast
media for three-dimensional MR angiography. J Magn Reson Imaging 13:588-593
43. Bogaert J, Bosmans H, Maes A, Suetens P, Marchal G, Rademakers FE (2000) Remote
myocardial dysfunction after acute anterior myocardial infarction: impact of left
ventricular shape on regional function: a magnetic resonance myocardial tagging study. J
Am Coll Cardiol 35:1525-1534
44
Quantitative T2* assessment of acute and chronic myocardial ischemia/reperfusion injury in mice
44. Lessick J, Sideman S, Azhari H, Marcus M, Grenadier E, Beyar R (1991) Regional threedimensional geometry and function of left ventricles with fibrous aneurysms. A cinecomputed tomography study. Circulation 84:1072-1086
45.
Kramer CM, Lima JA, Reichek N, Ferrari VA, Llaneras MR, Palmon LC, Yeh IT, Tallant B, Axel
L (1993) Regional differences in function within noninfarcted myocardium during left
ventricular remodeling. Circulation 88:1279-1288
46. Kramer CM, Rogers WJ, Theobald TM, Power TP, Petruolo S, Reichek N (1996) Remote
noninfarcted region dysfunction soon after first anterior myocardial infarction. A magnetic
resonance tagging study. Circulation 94:660-666
47. Rademakers F, Van de Werf F, Mortelmans L, Marchal G, Bogaert J (2003) Evolution of
regional performance after an acute anterior myocardial infarction in humans using
magnetic resonance tagging. J Physiol 546:777-787
45
Chapter 3
Quantitative changes in T2* reflect
remodelling of both remote and ischemic
myocardium in a murine heart failure model
Eissa N. E. Aguor†, Cees W.A. van de Kolk†, Gustav J. Strijkers, Marcel G.J. Nederhoff,
Pieter A. Doevendans, Cees J.A. van Echteld, Fatih Arslan
†
These authors contributed equally to this study
In preparation
47
Chapter 3
Abstract
Heart failure is an increasing burden to western societies due to aging and increased
survival of patients suffering from myocardial infarction. Early identification of
adverse structural changes in the myocardium may optimize clinical management.
Late Gadolinium enhancement (LGE)-MRI is widely used to assess infarct size and
myocardial fibrosis. We have recently shown that quantitative T2* mapping can provide
additional information on infarct status and changes in the infarcted myocardium in
relatively small murine infarctions after ischemia/reperfusion injury, In this study, we
further explored quantitative changes in T2* in both infarcted and remote myocardium
in a murine heart failure model induced by severe myocardial infarction.
Myocardial infarction (MI) was surgically induced in C57Bl/6J mice (n=12) by permanent
ligation of the left coronary artery. A sham‑operated group (n=8) served as control.
MRI at 9.4 T was performed at baseline, 1, 7 and 28 days after surgery. The protocol
consisted of Cine-MRI, multi‑gradient‑echo T2*‑mapping and LGE. Histological heart
sections were stained to identify hemorrhage and fibrosis.
Baseline myocardial T2* was 15.0±1.1 ms in the remote myocardium (septal wall) and
14.6±1.0 ms in the free wall. At day 1, LGE displayed a homogeneous enhancement of
the infarction. In addition, T2* values in the infarcts substantially decreased (5.7±0.4 ms),
whereas a slight decrease was observed in the remote myocardium (13.0±1.5 ms). On
days 7 and 28, LGE area of enhancement was smaller and heterogeneous compared to
LGE at day 1. After 7 and 28 days, T2* values in the infarct remained low (5.5±0.5 ms and
5.0±0.5 ms, respectively). Interestingly, T2* values in the remote myocardium continued
to decrease during follow-up (9.3±1.3 ms and 8.4±0.3 ms at day 7 and 28, respectively).
Histological analysis revealed progressive deposition of collagen in the infarct and to a
lesser degree in the remote myocardium, in parallel with decreased T2*.
Quantitative T2* values changed dynamically in this murine heart failure model. T2*
in infarcted areas exhibited a significant decrease starting from day 1, and further
decreased with scar maturation. In contrast to our previous observations in relatively
48
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
small murine infarctions, T2* in remote tissue also decreased significantly from baseline,
most likely as ,a result of adverse ventricular remodeling of non-infarcted areas
after MI. Serial LGE scans revealed merely changes in the infarct area, whereas the
remote myocardium did not exhibit any dynamics in LGE assessments. In conclusion,
quantitative T2* assessment may provide a valuable readout of both the infarct and
remote areas in heart failure after MI.
49
Chapter 3
3.1Background
Over the past decade, optimized reperfusion therapy has resulted in improved survival
of patients after myocardial infarction (MI). However, morbidity and mortality of
surviving patients is significantly increased due to aging and heart failure development
[1]. Heart failure is one of the most common complications after infarction and an
increasing socio-economic burden to Western societies. Early characterization of
structural changes within the infarcted and non-ischemic remote myocardium may
be of great importance for clinical management. In this respect, magnetic resonance
imaging (MRI) is a powerful tool to visualize, characterize and follow the injured
myocardium after an initial ischemic insult. Traditionally, MRI has been applied to
assess functional and pathological changes of the injured myocardium. Typically, T1weighted MRI is performed some time after injecting a Gadolinium‑based contrast
agent to delineate injured myocardium in a late Gadolinium enhancement (LGE) scan
[2]. The signal enhancement in LGE-MRI is influenced by many factors, such as the used
MRI protocol, MR sequence parameters, scan timing, the presence of edema [your
paper and others], and wash-in and wash-out kinetics of the contrast agent [2-4]. Thus,
LGE-MRI accuracy especially in acute MI has been subject to debate.
After the onset of myocardial ischemia, the post-injured myocardium undergoes
dynamic changes. Necrosis, increased extracellular volume, decreased perfusion,
microvascular obstruction, hemoglobin degradation products and the presence
of scar tissue influence the T1 and T2(*) tissue relaxation times [5]. T2*-weighted
imaging has been used before as a tool to diagnose the presence of hemorrhage
[6;7] and myocardial iron [8]. Also, T2*‑weighted imaging was used to track targeted
superparamagnetic iron oxide contrast agents in the myocardium [6] and to localize
iron oxide-labeled stem cells [9].
In general, quantitative relaxation time mapping MRI might be preferred over
relaxation time weighted imaging, because it is less subject to variations in protocol
and physiological parameters. Several quantitative T1-mapping [10], T2-mapping [11]
and T2*-mapping [5;12] protocols were found useful in the assessment of infarcts.
50
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
Specifically, T2* values may provide important information on the condition of the
post-injured myocardium related to the presence of edema [12;13], hemorrhage [12],
and the formation of collagenous scar tissue [14].
Previously, we have shown the benefit of cardiac T2* mapping in mice with relatively
small infarcts[15]. Subjecting mice to 30 minutes of ischemia followed by reperfusion,
we observed that T2* values in the infarct decreased during scar maturation, suggesting
that T2* values can differentiate acute from chronic MI. However, we did not observe a
change in T2* values in the remote myocardium. This may be due to the fact that the
relatively small infarct size and only 30 minutes ischemia followed by reperfusion does
not result in significant structural alterations within the remote myocardium. In the
present study, we used a heart failure model to address whether T2* mapping has any
added value in assessing non-ischemic remote myocardium after MI.
3.2Methods
Experimental design
C57Bl/6J mice (n=12, body weight=20-25 g)
were used in this study. Cardiac
MRI measurements were performed at baseline, 1, 7 and 28 days after myocardial
infarction (MI) induced by permanent ligation of the left coronary artery (LCA). The
MRI protocol consisted of a multi‑gradient‑echo quantitative T2*‑mapping sequence,
traditional LGE‑MRI to assess the location and size of the infarction, and gradient‑echo
Cine imaging to determine cardiac function. All animal experiments were performed
in accordance with the national guidelines on animal care and with prior approval by
the Animal Experimentation Committee of Utrecht University.
Myocardial infarction
Mice underwent permanent LCA ligation as previously described [16]. Briefly, mice
were anesthetized with a mixture of Fentanyl (Jansen-Cilag) 0.05 mg/kg, Dormicum
(Roche) 5 mg/kg, and medetomidine 0.5 mg/kg through an intraperitoneal injection.
51
Chapter 3
A core body temperature of about 37 °C was maintained during surgery by continuous
monitoring with a rectal thermometer and an automatic heating blanket. Mice were
intubated and ventilated (Harvard Apparatus Inc, Holliston, Mass) with 100% oxygen.
The left coronary artery was ligated permanently with an 8-0 Ethilon (Ethicon) suture.
Ischemia was confirmed by bleaching of the myocardium and transient tachycardia.
The chest wall was closed, and the animals subcutaneously received Antisedan (Pfizer)
2.5 mg/kg, Anexate (Roche) 0.5 mg/kg, and Temgesic (Schering-Plough) 0.1 mg/kg.
Magnetic resonance imaging
A vertical 9.4 T and 89-mm-diameter bore scanner was used, equipped with 1500 mT/m
gradients (Bruker BioSpin GmbH, Ettlingen, Germany). Mice were anesthetized with
1.5‑3 vol% isoflurane in a 2:1 mixture of air (0.3 l/min) and oxygen (0.15 l/min). The
mice were put on a custom‑built cradle and positioned inside a 3-cm‑diameter
quadrature volume coil. During the MRI examination anesthesia was maintained
between 1.5‑2.5 vol% isoflurane to keep the respiratory rate at 50-60 bpm. For cardiac
triggering and respiratory gating, a pressure balloon was placed under the chest, at
the level of the heart.
The MRI protocol consisted of the following steps. Two- and 4-chamber view scout
scans were made to plan a single mid-ventricular short-axis slice. Cine imaging
was performed with a retrospectively‑triggered gradient‑echo sequence, with
the following parameters: TR = 6.8 ms, TE = 1.9 ms, number of movie frames = 15,
slice thickness = 1 mm, matrix = 256x256, field‑of‑view = 3x3 cm2. Seven to 9
short‑axis slices with inter‑slice distance of 1 mm were measured to cover the heart
from apex to base. Based on the Cine axial images, a position of the axial midventricle lower papillary muscle slice was imported to include an area of remote
viable tissue as well as a substantial infarct area. T2* mapping was performed using
a cardiac‑triggered multi‑gradient-echo sequence with the following parameters:
TR = R‑R interval, TE = 1.22, 3.45, 5.68, 7.91, 10.14, and 12.37 ms, slice thickness = 1 mm,
matrix = 128x128, field‑of‑view = 3x3 cm2. The trigger delay was chosen according to
the mid-diastolic rest period of the heart cycle observed in the Cine series. LGE‑MRI
scans were performed in the same slice with a cardiac‑triggered inversion‑recovery
52
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
segmented gradient‑echo sequence, with the following parameters: TR = 5.8 ms,
TE = 2.2 ms, number of segments = 16, slice thickness = 1 mm, matrix = 256x256,
field‑of‑view = 3x3 cm2. The inversion time (TI) was 160 ms to null the signal from
remote myocardium. A Gadobutrol solution (Gadovist, Bayer Schering Pharma AG,
Berlin, Germany) of 1 mmol/ml was used. To reduce volume related variability, the
contrast agent was diluted 1:4 with a standard saline solution. This dilution resulted
in an administered weight-adapted volume of 2 µl/g body weight (equivalent dose of
0.5 mmol/kg body weight). The contrast agent was injected intravenously through the
tail vein, using a 24-gauge catheter.
Histopathology
On termination (at 1, 7 and 28 days after MI) the hearts were excised for histological
evaluation. A 1.5-mm-thick mid‑axial slice was fixed in 4% formalin and paraffinembedded. Hereafter, 5‑µm‑thick sections were mounted on glass slides and stained
with Picrosirius red staining for collagen, Hematoxylin-Eosin (HE) for cellular structures,
and Prussian blue staining for iron. The stained sections were photographed with a
digital image microscopy using white or polarized light.
MR Image analysis
T2* maps were generated in Mathematica 7 (Wolfram Research, Champaign, IL, USA)
by pixel‑wise fittings of the multi-gradient-echo MR signal intensities to the equation
S = S0 e -TE/T2*. The LV myocardium was manually segmented by drawing epicardial
and endocardial contours, excluding the papillary muscles. A region of interest in the
infarct in the myocardial free wall (infarct ROI) was drawn based on the enhanced
region in the LGE images. LGE‑based infarct size is reported as percentage of the LV
myocardial cross-sectional area (%LV). A second ROI was drawn in the non‑enhancing
remote tissue in the septum (remote ROI). For control mice, an ROI was drawn in the
septum only. These ROIs were drawn within the zone free from susceptibility artifacts
induced by lungs. T2* values are reported as group average and standard deviation of
the mean T2* in the ROIs. Cine MR Images were analyzed with Qmass digital imaging
software (Medis, Leiden, the Netherlands). Semi‑automatic segmentation was used to
53
Chapter 3
determine end-diastolic volume (EDV) and end‑systolic volume (ESV). Ejection fraction
(EF) was calculated as 100% (EDV‑ESV)/EDV.
Statistical analysis
All statistical analysis was done using (SPSS 17 for Windows; SPSS). Results are expressed
as mean ± standard deviation for each group of animals. Both independent twotailed sample t-test and the Mann-Whitney exact two-tailed U test were performed to
compare mean values in subjects for each time point post MI versus baseline. Results
were considered significant at P < 0.05. The correlation analysis was performed using
Pearson correlation with two‑tailed P values at all time points post MI.
3.3Results
In Figure 1 mid-axial T2* maps at baseline are shown for three representative mice.
T2* values were homogeneous throughout the left-ventricular myocardial wall and
consistent between mice, with a mean T2* value of 15.0±1.1 ms in the septum and
14.6±1.0 ms in the free wall.
Figure 1. Representative T2* maps of a mid-ventricle short-axis slice of three healthy
mice at baseline. T2* in the left-ventricular myocardium at end‑diastole is colorcoded between 0 and 20 ms according to the pseudo color scale on the left. The
background image is a regular gradient echo image with TE = 1.22 ms.
54
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
Figure 2. Mid-ventricle short-axis slice of a mouse heart at 1, 7, and 28 days, following
permanent ligation of the left coronary artery.
(bottom) LGE images. Red arrows point to the location of the infarct core (1) and
peri-infarct regions (2). Corresponding T2* maps color-coded from 0 to 20 ms. Note
the considerable T2* decrease in the free wall at day 7 and 28 (red arrows). (right)
Corresponding Prussian blue staining at day 28.
Figure 2 displays LGE and T2* maps at mid-diastole, for two mice at day 1, 7 and 28,
following the induction of MI. On day 1, LGE showed a homogeneous enhancement
of the infarct tissue in the mid-anterior and mid-anterolateral wall, whereas on days
7 and 28 the enhanced area was smaller, heterogeneous, and difficult to detect.
Corresponding quantitative T2* values were dynamic during infarct development and
maps showed a good match to the LGE area. T2* values in the mid-anterolateral wall
slightly decreased on day 1, and then decreased further on day 7, and 28 post MI.
Figure 3 shows the quantitative T2* values in the remote vs. infarct at baseline and
throughout infarct development. At day 1, T2* values in the infarcts dramatically
55
Chapter 3
Figure 3. Quantitative T2* values and infarct size
Quantitative T2* values in the infarct and remote myocardium as function of days
after surgery (A) and the infarct size from LGE-MRI (B). * = significantly different from
baseline (P < 0.05), and ** = significantly different from baseline (P < 0.005).
decreased to 5.7±0.4 ms (p<0.05, compared to baseline). A slight decrease was
observed in remote tissue (13.0±1.5 ms). At day 7 and 28, T2* values in the infarct
remained low (5.5±0.5 ms and 5.0±0.5 ms, respectively). T2* values in remote tissue
did not normalize but also decreased further at day 7 and 28 towards 9.3±1.3 ms
and 8.4±0.3 ms, respectively. In parallel to the decreased T2* values, LGE decreased
during follow-up. At day 1, LGE in infarct area was homogeneous and infarct size was
48.4±2.8% of the LV area. After 7 and 28 days post-MI, LGE became heterogeneous and
infarct size decreased (28.0±2.6 and 21.0±3.0% of the LV, respectively).
Global cardiac function was assessed using left ventricular end-diastolic volume (EDV)
and end-systolic volume (ESV) (Figure 4). During post-MI follow-up, EDV and ESV
showed a progressive and significant increase, indicative for a significant dilatation
of the heart, accompanied by a decline in ejection fraction (EF) from 54.7±2.3 % at
baseline to 21.3±10.8 %, 28 days after surgery.
A positive correlation was found between LGE-based infarct size and quantitative T2*
values in the infarcts (Pearson correlation 0.90, P = 0.045).
56
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
Figure 4. Left ventricle functional parameters as function of time
(A) EDV and ESV. (B) Ejection fraction. * = significantly different from baseline
(P < 0.05) , and ** = significantly different from baseline (P < 0.005).
Figure 5. Representative histological images for a mid‑axial slice of a mouse heart at day
1, 7 and 28 post MI. (A) T2* mapping. (B) Hematoxylin-eosin. (C) Picrosirius red under
polarized light. (D) Prussian-blue. (magnification is x 200)
57
Chapter 3
We performed immunohistochemistry in heart sections at each time point. Figure 5
shows representative images of 3 mice at baseline, 1 day and 28 days after infarction.
Picrosirius staining revealed no collagen deposition in the free wall at baseline and in
the infarct at day 1. At day 28, significant collagen deposition was observed. Sections
stained with Prussian blue showed scattered spots of iron as early as day 1. Iron
deposition became more prominent at day 28. T2* was lower in the peri-infarct zone in
comparison to the infarct core. Correspondingly, in the peri-infarct zone more iron was
found than in the core (Figure 5). Finally, on sections stained with hematoxylin-eosin
(Figures 5D) the infarct area was characterized by infiltration of inflammatory cells.
3.4Discussion
We have shown that T2* values in both ischemic and non-ischemic myocardium are
dynamic in time in a murine heart failure model. T2* dynamics reflect structural and
compositional changes within the tissue. The reduction in T2* with infarct is likely related
to the presence of iron and the formation of collagen. Furthermore, the reduction of T2*
in the infarct was paralleled by reduced LV function and infarct size.
T2* traditionally is used to assess iron deposition in the heart [17]. Despite the large
decrease of T2*, however, we did not observe significant amounts of iron and collagen
by histology in the infarct at day 1. Nevertheless, iron deposits by hemorrhage may
be present in different forms of hemoglobin degradation products [12;18-20], which
might have different sensitivity to the histological iron staining. The same is true for
the sensitivity of T2* to iron. The reduction of T2* at day 7 and 28 is most likely related
to the presence of iron and collagen in the infarct. Lower T2* values were found in the
peri-infarct zone in comparison to the core, in agreement with the findings by van den
Bos et al. [5].
Interestingly, T2* values in the infarct at day 1 were lower (5.7±0.4 ms) than previously
found in I/R infarcts (12.0±1.1 ms) [15], possibly related to the severity of injury at day
1 as a consequence of permanent occlusion. This suggests that the quantitative T2*
58
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
values could be used not only to image iron deposition but also to assess the condition
of the infarction. Higher T2* values in acutely I/R injured myocardium might also be
related to the presence of edema.
We also found a significant change of T2* in the remote myocardium, in agreement
with finding of Bogaert et al. [21], who observed a dysfunction in the remote area at
5±2 days after reperfusion in patients with transmural anterior MI. Others have also
shown that remodeling is initiated in the remote myocardium [22;23] as early as 1 day
after injury.
LGE showed different degree of enhancement for different mice and at different scan
points, reflecting the continuous pathological and cellular changes after infarction.
In acute myocardial injury with loss of cell membrane integrity [5;24], inflammatory
response, [5;25], necrosis, hemorrhage, micro-vascular obstruction (MVO) and
particularly edema, LGE-based infarct size may be overestimated by the presence
of edema [5;26]. In the chronic infarction, the formation of scar tissue and the larger
extent of vascular obstructions could be responsible for a reduced distribution volume
for the extracellular contrast agent, leading to an inhomogeneous enhancement and
reduction of LGE-based infarct size. Additionally, the extremely low T2* in the infarct
(Figure 3), which locally decreases down to a few milliseconds, could effectively null the
signal of the inversion recovery gradient-echo sequence with an echo time of 2.2 ms,
obscuring LGE [27]. These findings are in agreement with the observed reduction in
LGE infarct size in patients with one week old MI [2].
A drawback of cardiac T2* as a diagnostic marker for myocardial injury is contamination
of intrinsic T2* values by the presence of macroscopic susceptibility variation caused
by air-lung interfaces [12;13]. Occasionally, we observed lowered T2* values at the
lung-heart interface outside areas that could be associated with known locations of
myocardial infarction. Also the orientation of the heart with respect to the magnetic
field could alter absolute T2* values. Improvements in local shimming techniques
might alleviate some of these susceptibility variations. Therefore, use of T2* mapping
as a stand-alone technique without the complementary information from LGE might
still be undesired.
59
Chapter 3
3.5Conclusions
Imaging of the myocardium with a T2* mapping technique in a mouse model of
heart failure provides useful and complementary information to LGE and Cine MRI.
The T2* value decreases dramatically in infarcts, and correlates to the infarct size. For
the first time, we have demonstrated that T2* also decreases in non-ischemic remote
myocardium of the failing heart. The decrease of T2* was paralleled by iron and collagen
depositions. In addition to LGE, T2* mapping delineated the extent of myocardial injury
in both acute and chronic conditions compared with the cine MRI and histological
findings.
Reference List
1. Juenger J, Schellberg D, Kraemer S, Haunstetter A, Zugck C, Herzog W, Haass M (2002)
Health related quality of life in patients with congestive heart failure: comparison with
other chronic diseases and relation to functional variables. Heart 87:235-241
2. Ibrahim T, Hackl T, Nekolla SG, Breuer M, Feldmair M, Schomig A, Schwaiger M (2010)
Acute myocardial infarction: serial cardiac MR imaging shows a decrease in delayed
enhancement of the myocardium during the 1st week after reperfusion. Radiology
254:88-97
3. Rogers WJ, Kramer CM, Geskin G, Hu YL, Theobald TM, Vido DA, Petruolo S, Reichek N
(1999) Early contrast-enhanced MRI predicts late functional recovery after reperfused
myocardial infarction. Circulation 99:744-750
4. Saraste A, Nekolla S, Schwaiger M (2008) Contrast-enhanced magnetic resonance imaging
in the assessment of myocardial infarction and viability. Journal of Nuclear Cardiology
15:105-117
5. van den Bos EJ, Baks T, Moelker AD, Kerver W, van Geuns RJ, van der Giessen WJ, Duncker
DJ, Wielopolski PA (2006) Magnetic resonance imaging of haemorrhage within reperfused
myocardial infarcts: possible interference with iron oxide-labelled cell tracking? Eur Heart
J 27:1620-1626
6. Sosnovik DE, Nahrendorf M, Weissleder R (2007) Molecular magnetic resonance imaging
in cardiovascular medicine. Circulation 115:2076-2086
7. Ghugre NR, Ramanan V, Pop M, Yang Y, Barry J, Qiang B, Connelly KA, Dick AJ, Wright GA
(2011) Quantitative tracking of edema, hemorrhage, and microvascular obstruction in
60
Quantitative changes in T2* reflect remodelling of both remote
and ischemic myocardium in a murine heart failure model
subacute myocardial infarction in a porcine model by MRI. Magn Reson Med 66:11291141
8. Leonardi B, Margossian R, Colan SD, Powell AJ (2008) Relationship of magnetic resonance
imaging estimation of myocardial iron to left ventricular systolic and diastolic function in
thalassemia. JACC Cardiovasc Imaging 1:572-578
9. Amsalem Y, Mardor Y, Feinberg MS, Landa N, Miller L, Daniels D, Ocherashvilli A, Holbova
R, Yosef O, Barbash IM, Leor J (2007) Iron-oxide labeling and outcome of transplanted
mesenchymal stem cells in the infarcted myocardium. Circulation 116:I38-I45
10. Turkbey EB, Gai N, Lima JA, van der Geest RJ, Wagner KR, Tomaselli GF, Bluemke DA,
Nazarian S (2012) Assessment of Cardiac Involvement in Myotonic Muscular Dystrophy by
T1 Mapping on Magnetic Resonance Imaging. Heart Rhythm
11. Thavendiranathan P, Walls M, Giri S, Verhaert D, Rajagopalan S, Moore S, Simonetti OP,
Raman SV (2012) Improved detection of myocardial involvement in acute inflammatory
cardiomyopathies using T2 mapping. Circ Cardiovasc Imaging 5:102-110
12. O’Regan DP, Ahmed R, Karunanithy N, Neuwirth C, Tan Y, Durighel G, Hajnal JV, Nadra I,
Corbett SJ, Cook SA (2009) Reperfusion hemorrhage following acute myocardial infarction:
assessment with T2* mapping and effect on measuring the area at risk. Radiology 250:916922
13. Ghugre NR, Ramanan V, Pop M, Yang Y, Barry J, Qiang B, Connelly KA, Dick AJ, Wright GA
(2011) Quantitative tracking of edema, hemorrhage, and microvascular obstruction in
subacute myocardial infarction in a porcine model by MRI. Magn Reson Med 66:11291141
14. Pepe A, Positano V, Santarelli MF, Sorrentino F, Cracolici E, De MD, Maggio A, Midiri M,
Landini L, Lombardi M (2006) Multislice multiecho T2* cardiovascular magnetic resonance
for detection of the heterogeneous distribution of myocardial iron overload. J Magn
Reson Imaging 23:662-668
15. Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, van Echteld CJ,
Pasterkamp G, Strijkers GJ (2012) Quantitative T (2) (*) assessment of acute and chronic
myocardial ischemia/reperfusion injury in mice. MAGMA
16. Arslan F, Smeets MB, O’Neill LAJ, Keogh B, McGuirk P, Timmers L, Tersteeg C, Hoefer IE,
Doevendans PA, Pasterkamp G, de Kleijn DPV (2010) Myocardial Ischemia/Reperfusion
Injury Is Mediated by Leukocytic Toll-Like Receptor-2 and Reduced by Systemic
Administration of a Novel Anti-Toll-Like Receptor-2 Antibody. Circulation 121:80-90
17. Wood JC, Ghugre N (2008) Magnetic resonance imaging assessment of excess iron in
thalassemia, sickle cell disease and other iron overload diseases. Hemoglobin 32:85-96
18. Husser O, Monmeneu JV, Sanchis J, Nunez J, Lopez-Lereu MP, Bonanad C, Chaustre F, Gomez
C, Bosch MJ, Hinarejos R, Chorro FJ, Riegger GA, Llacer A, Bodi V (2012) Cardiovascular
magnetic resonance-derived intramyocardial hemorrhage after STEMI: Influence on long-
61
Chapter 3
term prognosis, adverse left ventricular remodeling and relationship with microvascular
obstruction. Int J Cardiol
19. Lotan CS, Bouchard A, Cranney GB, Bishop SP, Pohost GM (1992) Assessment of
postreperfusion myocardial hemorrhage using proton NMR imaging at 1.5 T. Circulation
86:1018-1025
20. Beek AM, Nijveldt R, van Rossum AC (2010) Intramyocardial hemorrhage and microvascular
obstruction after primary percutaneous coronary intervention. Int J Cardiovasc Imaging
26:49-55
21. Bogaert J, Bosmans H, Maes A, Suetens P, Marchal G, Rademakers FE (2000) Remote
myocardial dysfunction after acute anterior myocardial infarction: impact of left
ventricular shape on regional function: a magnetic resonance myocardial tagging study. J
Am Coll Cardiol 35:1525-1534
22. Lessick J, Sideman S, Azhari H, Marcus M, Grenadier E, Beyar R (1991) Regional threedimensional geometry and function of left ventricles with fibrous aneurysms. A cinecomputed tomography study. Circulation 84:1072-1086
23. Kramer CM, Lima JA, Reichek N, Ferrari VA, Llaneras MR, Palmon LC, Yeh IT, Tallant B, Axel
L (1993) Regional differences in function within noninfarcted myocardium during left
ventricular remodeling. Circulation 88:1279-1288
24. bdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
25. Murphy E, Steenbergen C (2008) Mechanisms underlying acute protection from cardiac
ischemia-reperfusion injury. Physiol Rev 88:581-609
26. Schwitter J, Saeed M, Wendland MF, Derugin N, Canet E, Brasch RC, Higgins CB (1997)
Influence of severity of myocardial injury on distribution of macromolecules: extravascular
versus intravascular gadolinium-based magnetic resonance contrast agents. J Am Coll
Cardiol 30:1086-1094
27. Bremerich J, Colet JM, Giovenzana GB, Aime S, Scheffler K, Laurent S, Bongartz G, Muller
RN (2001) Slow clearance gadolinium-based extracellular and intravascular contrast
media for three-dimensional MR angiography. J Magn Reson Imaging 13:588-593
62
Chapter 4
Lack of circulating Toll-like receptor 2
promotes survival and prevents adverse
remodeling after myocardial infarction
Eissa N.E. Aguor, Mirjam B. Smeets, Pieter A. Doevendans, Gustav J. Strijkers, Gerard
Pasterkamp & Dominique P. de Kleijn, Fatih Arslan
In preparation
63
Chapter 4
Abstract
Background Excessive inflammation is a known mediator of adverse cardiac
remodeling after myocardial infarction. Toll-like receptors (TLRs) have been shown to
play a crucial role in inflammatory responses after tissue injury. In the present study,
we explored the functional and biochemical consequences of TLR2 absence in left
ventricular remodeling after myocardial infarction.
Methods and Results C57Bl6 wild-type (WT), TLR2-/- and chimeric mice underwent
permanent ligation of the left coronary artery. Cardiac function and geometry were
assessed using 9.4T mouse-MRI at baseline, 7 and 28 days after infarction. Despite
similar infarct size between WT and TLR2-/- mice (39±1.4% vs. 42±1.5%, respectively;
p=.256), the absence of TLR2 promoted survival and prevented left ventricular
dilatation during 28 days follow-up. Furthermore, systolic performance of both
infarcted and remote myocardium was better in TLR2-/- mice compared to WT animals.
Chimeric mice experiments revealed that expansive remodeling after infarction was
mediated by bone marrow-derived TLR2 expression. Characterization of peripheral
blood cells showed that the absence of TLR2 reduced white blood cell count 7 days
after infarction. Peripheral blood monocytes of TLR2-/- mice exhibited similar CD11b
expression, but reduced CD49d levels compared to WT animals 7 days after infarction.
The functional and cellular changes in TLR2-/- mice were preceded by significant
reduction in tissue levels of TNFα, IL-1α, GM-CSF and IL-10 at 3 days after infarction.
Conclusions Lack of TLR2 improves survival and cardiac function after myocardial
infarction and is determined by leukocytic TLR2 deficiency via reduced inflammation.
Hence, TLR2 is a potential therapeutic target to prevent adverse remodeling after
myocardial infarction.
Keywords: myocardial infarction, heart failure, remodeling, immune system, Toll-like
receptor
64
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
4.1 Background
Optimized reperfusion strategies have resulted in improved survival of patients
after myocardial infarction (MI). Unfortunately, quality of life of surviving patients is
significantly decreased due to increased morbidity after infarction [1]. Heart failure is
one of the most common complications after infarction and inflict an increasing socioeconomic burden on Western societies. Currently, the post-infarct healing process
is of increasing importance for therapeutic modulation to enhance cardiac repair
and function after MI. Inflammation, proliferation and maturation are chronological
processes during infarct healing that can be targeted independently from each other
for therapeutic modulation [2]. Inflammation is an interesting therapeutic target, since
it is initiated early and has profound effects for the later stages. Cytokine and chemokine
production and subsequent leukocyte recruitment are initiated after the ischemic
injury. Leukocyte migration is a critical first step in the removal of debris and matrix
breakdown and subsequent scar formation [2]. The extent and type of inflammation
has great impact on processes like angiogenesis and fibrosis in the (sub)chronic phase
after infarction [3-5]. For this reason, modulation of the early inflammatory process may
have great therapeutic value to regulate adaptive repair processes hereafter. However,
inflammation is not a random binary process rather a well-orchestrated dynamic
response to tissue injury. For this reason, therapeutic modulation is challenging and
difficult. More importantly, inflammation has shown to have ‘double-edged sword’
characteristics thereby complicating therapeutic interventions [6;7].
Recent studies have clearly demonstrated that Toll-like receptor (TLR) deficiency or
inhibition enhance the ‘good’ and blunt the ‘bad’ of the inflammatory reaction in
cardiac ischemia. Traditionally known as pathogen recognition receptors, TLRs have
been shown to recognize endogenous activators released after cardiovascular injury
[8]. TLR2 and 4 have been extensively studied in cardiac ischemia [9]. TLR4 deficiency
or inhibition protects against ischemia/reperfusion (I/R) injury and adverse left
ventricular remodeling [10-12]. We have recently shown that circulating TLR2 mediates
myocardial I/R injury and that its inhibition is cardioprotective. TLR2 inhibition reduces
monocytes activation after reperfusion, thereby inhibiting cytokine and chemokine
production and subsequent reduction of apoptosis [13]. Furthermore, previous studies
65
Chapter 4
have shown that TLR2 deficiency prevents maladaptive remodeling [14]. However, the
relative contribution of parenchymal and hematopoiesis-derived TLR2 expression to
cardiac repair responses remain unknown.
In the present study, we show for the first time that circulating TLR2 expression mediates
adverse remodeling after myocardial infarction. Furthermore, we investigated the
functional and biochemical differences between TLR2 deficient and wild type animals
during adverse remodeling after infarction.
4.2Methods
Animals and Experimental Design
TLR2-/- animals were backcrossed for 6 generations into a C57Bl6 background. Male
C57Bl6/J and C57Bl6-TLR2-/- mice (10-12 wks, 25-30 g) received standard diet and
water ad libitum. Myocardial infarction was induced by permanent left coronary artery
ligation, just below the left atrial appendage as described previously. Digital images of
the infarcts were encrypted before being analyzed by the researcher. Heart function
and geometry assessments were done by a technician blinded to treatment. All animal
experiments were performed in accordance with the national guidelines on animal
care and with prior approval by the Animal Experimentation Committee of Utrecht
University.
Myocardial Infarction In Vivo
Mice were anesthetized with a mixture of Fentanyl (Jansen-Cilag) 0.05 mg/kg, Dormicum
(Roche) 5 mg/kg and medetomidine 0.5 mg/kg through an intraperitoneal injection.
Core body temperature was maintained around 37°C during surgery by continuous
monitoring with a rectal thermometer connected to an automatic heating blanket.
Mice were intubated and ventilated (Harvard Apparatus Inc.) with 100% oxygen. The
left coronary artery (LCA) was permanently ligated using an 8-0 vicryl suture. Ischemia
was confirmed by bleaching of the myocardium and ventricular tachyarrhythmia. In
sham operated animals the suture was placed beneath the LCA without ligating. The
66
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
chest wall was closed and the animals received subcutaneously Antisedan (Pfizer) 2.5
mg/kg, Anexate (Roche) 0.5 mg/kg and Temgesic (Schering-Plough) 0.1 mg/kg.
Infarct Size
for another 15 minutes. Viable tissue stains red and infarcted tissue appears white.
Heart sections were digitally photographed (Canon EOS 400D) under a microscope
(Carl Zeiss®). IS, AAR and total LV area were measured using ImageJ software (version
1.34). Infarct size was corrected for the weight of the corresponding heart slice.
Generation of Chimeric Mice
We generated chimeric mice to study the relative contribution of TLR2 expression in
blood and parenchymal cells to cardiac remodeling. Donor bone marrow (BM) cells
were collected from wild-type (WT) C57Bl6 and TLR2 knock-out (TLR2-/-) mice by
flushing humerus, femurs and tibiae with RPMI-1640 medium. Recipient mice received
5x106 BM cells after receiving a single dose of 7 Gy. Mice recovered for 6 weeks to
ensure stable engraftment of the donor bone marrow cells. Hereafter, chimerization
was confirmed by phenotyping TLR2 expression on peripheral blood samples
with Cytomics FC500 (Beckman Coulter) analysis. Successful chimerization (>95%
circulating donor cells) was achieved in all mice (data not shown). Irradiated WT mice
with TLR2-/- bone marrow are referred as WT/TLR2 KO BM, and TLR2-/- mice with WT
bone marrow as TLR2 KO/WT BM.
Magnetic Resonance Imaging
Twenty-three mice (n=5 in WT, n=6 in TLR2-/- and n=6 in sham operated mice) underwent
serial assessment of cardiac dimensions and function by high resolution magnetic
resonance imaging (MRI, 9.4T, Bruker, Rheinstetten, Germany) under isoflurane
anesthesia before, 7 and 28 days after MI. Long axis and short axis images with 1.0 mm
interval between the slices were obtained and used to compute end-diastolic volume
(EDV, largest volume) and end-systolic volume (ESV, smallest volume). The ejection
fraction (EF) was calculated as 100*(EDV-ESV)/EDV. Wall thickness (WT) and systolic
wall thickening (SWT) were assessed from both the septum (remote myocardium) and
67
Chapter 4
free wall (infarct area). All MRI data are analyzed using Qmass digital imaging software
(Medis, Leiden, The Netherlands).
Protein Isolation
Total protein was isolated from snap frozen infarcted heart sections (infarct and remote
area separated) using 40 mM Tris pH 7.4.
Flow Cytometry
Tumor necrosis factor (TNF)-α, IL-1α, IL-10, and granulocyte macrophage-colony
stimulating factor (GM-CSF) levels in isolated tissue protein samples were measured
by flow cytometry (Cytomics FC500, Beckman Coulter) using the Th1/Th2 customized
multiplex kit (Bender MedSystems, Vienna, Austria). The protein samples were diluted
1:1 in assay buffer, and the protocol is further followed according to the manufacturer’s
instructions.
CD11b and CD49d expression was assessed on circulating monocytes of EDTA
anticoagulated blood by flow cytometry. Whole blood was stained for CD11b (FITC,
eBioscience, San Diego, Calif ), and CD49d (Alexa Fluor 488, Serotec, Oxford, UK) and
F4/80 for monocytes (Alexa Fluor 647, Serotec, Oxford, UK).
Statistical Analysis
Data are represented as Mean±SEM. One-way ANOVA with post-hoc LSD test was used
for comparison >2 groups. Non-parametric t-test for 2 group comparisons. KaplanMeier survival analysis with log-rank test was used to evaluate mortality differences
between groups. All statistical analyses were performed using SPSS 15.1.1. and p<0.05
was considered significant.
The authors had full access to and take full responsibility for the integrity of the data.
All authors have read and agree to the manuscript as written.
68
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
4.3Results
Circulating TLR2 Expression Mediates Adverse Remodeling After
Infarction
There were no differences between WT and TLR2-/- animals in cardiac function and
geometry at baseline. Infarct size, determined as a percentage of the left ventricle (IS/
LV), showed no differences between TLR2-/- and WT mice 2 days after infarction (Figure
1A). Despite similar infarct size, TLR2-/- animals showed improved survival compared
to WT mice after infarction (Figure 1B). Most animals died between day 4 and 8 postMI due to cardiac ruptures. Serial cardiac MRI assessments revealed a significant
preservation of both function and LV dimensions during 28 days follow-up (Figure 1C
and D; Table 1). TLR2-/- mice were relatively protected against expansive remodeling
of the left ventricle when compared to WT animals. We performed bone marrow
transplantation experiments from WT to irradiated knock-out mice and vice versa,
in order to dissect the relative contribution to adverse remodeling of parenchymal
and circulating TLR2 expression. Chimeric mice experiments revealed that adverse
remodeling is relatively prevented in WT mice with TLR2-/- bone marrow (WT/TLR2 KO
BM). In contrast, TLR2-/- mice with WT bone marrow (TLR2 KO/WT BM) exhibit similar LV
dilatation compared to irradiated WTs with WT bone marrow (WT/WT BM; Figure 1E).
69
Chapter 4
Figure 1. Circulating TLR2 mediates adverse remodeling after myocardial infarction.
(A) Infarct size (IS) as a percentage of the left ventricle (LV) 2 days after infarction.
(B) Kaplan-Meier survival curve. (C) and (D), End-diastolic and end-systolic volume,
respectively (EDV, ESV); p-values see Table 1. (E) EDV and ESV in chimeric mice 28
days after infarction; *p=0.005 †p=0.007 compared to TLR2 KO/WT BM. Each bar
represents Mean±SEM, n=5/group for infarct size assessment, n=12 in WT and n=13
in TLR2-/- group for survival analysis. TLR2 KO/WT BM=irradiated TLR2-/- mice with WT
bone marrow, WT/TLR2 KO BM= irradiated WT mice with TLR2-/- bone marrow, WT/
WT BM=irradiated WT with WT bone marrow.
70
25.4±4.1
59.1±4.1
0.90±0.01
0.90±0.02
47.1±1.9
52.0±1.9
ESV, μl
EF, %
WT septum (remote, mm)
WT free wall (infarct, mm)
SWT septum (remote, %)
SWT free wall (infarct, %)
-34.0±5.7*
20.6±5.0*
0.60±0.08*
0.67±0.06*
7.2±2.8*
156.7±10.9*
168.6±7.2*
-56.6±3.6**
16.1±4.7**
0.45±0.04**
0.75±0.02**
4.3±2.3**
207.4±15.7**
216.3±11.8**
54.2±1.8
47.1±2.1
0.89±0.01
0.85±0.02
57.9±4.3
23.7±4.3
54.1±5.1
-18.6±4.2*
36.9±3.1*†
0.55±0.03*
0.72±0.01*
16.0±1.1*
93.1±6.0*†
110.7±6.6*†
7 days MI
NS
0.009
NS
NS
NS
<0.001
<0.001
†p
-26.2±4.0**†
35.8±3.2**†
0.45±0.07**
0.71±0.03**
11.9±1.0**
112.4±6.2**†
127.5±6.3**†
28 days MI
0.027
0.012
NS
NS
NS
<0.001
<0.001
†p
systolic volume, EF=ejection fraction, WT=wall thickness, SWT=systolic wall thickening
One-way ANOVA post hoc Dunnett t-test (2-sided), *p<0.05 compared to baseline, †p compared to WT mice. EDV=end-diastolic volume, ESV=end-
60.9±3.7
Baseline
28 days MI
Baseline
7 days MI
TLR2-/-
C57BL6/J WT
EDV, μl
Table 1.
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
71
Chapter 4
Figure 2. Lack of TLR2 reduces tissue cytokine levels and leukocytosis after infarction.
Tissue levels of (A) TNFα; *p=0.032 B, (B) IL-1α; *p=0.032 C, (C) IL-10; *p=0.063 and (D)
GM-CSF; *p=0.046 compared to WT animals. (E) White blood cell counts (WBC) after
infarction; *p=0.05 compared to WT animals. Each bar represents Mean±SEM, n=6/
group/time point.
72
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
TLR2 Deficiency Reduces Inflammation After Infarction
Myocardial infarction is characterized by cytokine and chemokine
induction and leukocytosis. In our experiments, TLR2 deficiency resulted
in a significant reduction of TNFα, IL-1α, IL-10 and GM-CSF 3 days after
infarction (Figure 2A – D). White blood cell counts revealed that TLR2/animals exhibited reduced leukocytosis compared to WT mice after
infarction (Figure 2E).
TLR2 Mediates Integrin-α4 Expression On Circulating Monocytes
After Infarction
Monocyte migration to the heart is an integrin-dependent process[15]. The expression
of CD11b did not differ between the groups after infarction (Figure 3A). However,
compared to WT animals, circulating monocytes of TLR2-/- mice expressed decreased
levels of CD49d after infarction (Figure 3B).
Figure 3. Lack of TLR2 reduces integrin-α4, but not CD11b expression on monocytes
after myocardial infarction. (A) CD11b and (B) CD49d on circulating monocytes;
*p=0.023 compared to WT animals. Each bar represents Mean±SEM, n=6/group/
time point. MFI=mean fluorescent intensity
73
Chapter 4
4.4Discussion
Toll-like receptors as therapeutic targets may hold great therapeutic value in the near
future. We have shown in both murine and porcine (unpublished data) models that
TLR2 inhibition is a good candidate for infarct size reduction after ischemia/reperfusion
[13]. The question remains whether TLR2 inhibition has adverse effects on remodeling
of the scar and remote myocardium in the long-term. There are tragic examples in
which inflammatory interventions with beneficial effects in the acute phase turned
out to be devastating in the long-term. Corticosteroid administration reduced infarct
size in animal experiments, but appeared to result in aneurysm formation and cardiac
ruptures in patients after infarction [6;7]. The main issue of down toning inflammation
is impaired healing resulting in reduced collagen deposition in the scar and ventricular
dilatation. Severe suppression of the immune response is associated with a so-called
“mummified myocardium”, characterized by delayed debris removal and decrease
infarct collagen content [16]. In addition, reduced infarct collagen is associated with
more pronounced LV dilatation and dysfunction [17]. A report by Mersmann J et al.
suggests that TLR2 deficiency results in adverse remodeling after infarction, despite
smaller infarct size compared to WT animals [18]. Unfortunately, in their study the
background of the TLR2-/- and WT mice were different: C57Bl6 vs. C3HeN, respectively.
For this reason, it is difficult to draw firm conclusion from these experiments.
In the present study, we investigated the healing process after infarction in TLR2/-
and WT mice, both on a C57Bl6/J background. Despite similar infarct size, cardiac
MRI assessments revealed that LV dilatation and functional deterioration is relatively
prevented in TLR2-/- mice. The improved cardiac function and geometry is likely the
reason for improved survival in TLR2-/- animals. Moreover, in our experiments lack
of TLR2 was relatively protective against bulging of the infarcted myocardium. It is
known that bulging of the infarct (i.e. aneurysm formation) is associated with reduced
collagen density in the scar [17].
We have previously shown that circulating TLR2 expression mediates ischemia/
reperfusion injury of the heart [13]. In addition, it is well-established that leukocytes
migrate to the heart to initiate the process of myocardial healing [2;5]. Hence,
we hypothesized that TLR2 expression on circulating blood cells is also the main
74
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
determinant for adverse remodeling. WT mice with TLR2 knock-out bone marrow were
indeed relatively protected against adverse remodeling compared to TLR2-/- mice with
WT bone marrow. Findings by Shishido et al. [14] and the present study are in strong
contrast to the study by Mersmann et al. in which TLR2 deficiency was associated with
impaired cardiac function and geometry. We believe that the different strains used
in their experiments (C3HeN background was compared to C57Bl6 background) are
the main reason for the conclusions drawn by Mersmann et al. rather than the TLR2
knock-out. It is known that mouse strains strongly differ in myocardial healing after
infarction [19]. Since improvement in cardiac function is a relative measure, it is of
utmost importance that the intervention (e.g. knock-out) is the only differing variable
and that confounding factors like background strain are absent.
Furthermore, we demonstrated that TLR2-/- mice exhibit reduced
Infarct size (IS) as a percentage of the left ventricle (LV) was determined using Evans’
blue dye injection and TTC staining, 2 days after infarction (n=5/group). By assessing
infarct size in the acute phase (at 2 days), one can determine whether differences
are present between WT and TLR2-/- mice in myocardial perfusion. Hence, 4% Evans
blue dye was injected via the thoracic aorta in a retrograde fashion. By doing so, one
can demarcate the area-at-risk (AAR), the extent of myocardial tissue that underwent
ischemia (i.e. endangered myocardium). Hearts were rapidly explanted, rinsed in 0.9%
saline and put in -20ºC freezer for 1 hour. Hereafter, hearts were mechanically sliced into
four 1-mm cross sections. Heart sections were incubated in 1% triphenyltetrazoliumchloride (Sigma-Aldrich) at 37ºC for 15 minutes before placing them in formaldehyde
cytokine levels and leukocytosis compared to WT animal after infarction. These
findings are in line with the pro-inflammatory actions of TLR2 upon activation. CD49d
is an important integrin for leukocyte migration and myofibroblast transdifferentiation
[15;20]. Whether leukocyte infiltration and myofibroblast transdifferentiation is altered
in TLR2-/- mice remains to be addressed. However, the reduced CD49d expression on
circulating TLR2-/- monocytes after infarction suggests that leukocyte migration and
myofibroblast transdifferentiation may be altered in TLR2-/- mice.
In conclusion, we have shown that circulating TLR2 mediates maladaptive remodeling
after myocardial infarction. Lack of TLR2 improves survival and prevents post-infarct
heart function deterioration via decreased inflammation and reduced migration
75
Chapter 4
capacity of monocytes. The exact mechanisms by which TLR2 mediates leukocyte
migration and myofibroblast transdifferentiation in the healing heart remains to be
addressed. Our study demonstrates that chronic absence of TLR2 is beneficial after
myocardial infarction and is a good candidate as a therapeutic target in patients after
myocardial infarction.
Acknowledgements
We would like to thank the following persons for their assistance: Ben J. van Middelaar,
Cees van der Kolk and Arjan Schoneveld (all from the UMC Utrecht, The Netherlands).
Funding Sources
This work is supported by research grants from Netherlands Organisation for Scientific
Research (NWO) Mozaïek grant (contract 017.004.004 to F.A.), Netherlands Heart
Foundation (contract 2010.T001 to F.A.) and Bekales Foundation (P.A.D.)
76
Lack of circulating Toll-like receptor 2 promotes survival and
prevents adverse remodeling after myocardial infarction
Reference List
1. Juenger J, Schellberg D, Kraemer S, Haunstetter A, Zugck C, Herzog W, Haass M (2002)
Health related quality of life in patients with congestive heart failure: comparison with
other chronic diseases and relation to functional variables. Heart 87:235-241
2. Frangogiannis NG (2008) The immune system and cardiac repair. Pharmacol Res 58:88-111
3. Arroyo AG, Iruela-Arispe ML (2010) Extracellular matrix, inflammation, and the angiogenic
response. Cardiovasc Res 86:226-235
4. Frantz S, Vincent KA, Feron O, Kelly RA (2005) Innate immunity and angiogenesis. Circ Res
96:15-26
5. Frangogiannis NG (2004) Chemokines in the ischemic myocardium: from inflammation to
fibrosis. Inflamm Res 53:585-595
6. Libby P, Maroko PR, Bloor CM, Sobel BE, Braunwald E (1973) Reduction of experimental
myocardial infarct size by corticosteroid administration. J Clin Invest 52:599-607
7. Roberts R, DeMello V, Sobel BE (1976) Deleterious effects of methylprednisolone in
patients with myocardial infarction. Circulation 53:I204-I206
8. Ionita MG, Arslan F, de Kleijn DP, Pasterkamp G (2010) Endogenous inflammatory
molecules engage Toll-like receptors in cardiovascular disease. J Innate Immun 2:307-315
9. Chao W (2009) Toll-like receptor signaling: a critical modulator of cell survival and ischemic
injury in the heart. Am J Physiol Heart Circ Physiol 296:H1-12
10. Arslan F, Keogh B, McGuirk P, Parker AE (2010) TLR2 and TLR4 in ischemia reperfusion
injury. Mediators Inflamm 2010:704202
11. Riad A, Jager S, Sobirey M, Escher F, Yaulema-Riss A, Westermann D, Karatas A, Heimesaat
MM, Bereswill S, Dragun D, Pauschinger M, Schultheiss HP, Tschope C (2008) Toll-like
receptor-4 modulates survival by induction of left ventricular remodeling after myocardial
infarction in mice. J Immunol 180:6954-6961
12. Timmers L, Sluijter JP, van Keulen JK, Hoefer IE, Nederhoff MG, Goumans MJ, Doevendans
PA, van Echteld CJ, Joles JA, Quax PH, Piek JJ, Pasterkamp G, de Kleijn DP (2008) Toll-like
receptor 4 mediates maladaptive left ventricular remodeling and impairs cardiac function
after myocardial infarction. Circ Res 102:257-264
13. Arslan F, Smeets MB, O’Neill LA, Keogh B, McGuirk P, Timmers L, Tersteeg C, Hoefer IE,
Doevendans PA, Pasterkamp G, de Kleijn DP (2010) Myocardial ischemia/reperfusion injury
is mediated by leukocytic toll-like receptor-2 and reduced by systemic administration of a
novel anti-toll-like receptor-2 antibody. Circulation 121:80-90
14. Shishido T, Nozaki N, Yamaguchi S, Shibata Y, Nitobe J, Miyamoto T, Takahashi H, Arimoto T,
Maeda K, Yamakawa M, Takeuchi O, Akira S, Takeishi Y, Kubota I (2003) Toll-like receptor-2
modulates ventricular remodeling after myocardial infarction. Circulation 108:2905-2910
77
Chapter 4
15. Sopel M, Ma I, Gelinas L, Oxner A, Myers T, Legare JF (2010) Integrins and monocyte
migration to the ischemic myocardium. J Invest Surg 23:79-86
16. Kloner RA, Fishbein MC, Lew H, Maroko PR, Braunwald E (1978) Mummification of the
infarcted myocardium by high dose corticosteroids. Circulation 57:56-63
17. Jugdutt BI, Joljart MJ, Khan MI (1996) Rate of collagen deposition during healing and
ventricular remodeling after myocardial infarction in rat and dog models. Circulation
94:94-101
18. Mersmann J, Habeck K, Latsch K, Zimmermann R, Jacoby C, Fischer JW, Hartmann C,
Schrader J, Kirschning CJ, Zacharowski K (2010) Left ventricular dilation in toll-like receptor
2 deficient mice after myocardial ischemia/reperfusion through defective scar formation.
Basic Res Cardiol
19. van den Borne SW, van dS, V, Strzelecka AE, Vervoort-Peters HT, Lijnen PM, Cleutjens JP,
Smits JF, Daemen MJ, Janssen BJ, Blankesteijn WM (2009) Mouse strain determines the
outcome of wound healing after myocardial infarction. Cardiovasc Res 84:273-282
20. Wang Q, Wang Y, Hyde DM, Gotwals PJ, Lobb RR, Ryan ST, Giri SN (2000) Effect of antibody
against integrin alpha4 on bleomycin-induced pulmonary fibrosis in mice. Biochem
Pharmacol 60:1949-1958
78
Part II
Sodium MRI
79
Chapter 5
Na Chemical Shift Imaging and
Gd Enhancement of myocardial edema
23
Eissa N.E. Aguor, Cees W.A. van de Kolk, Fatih Arslan, Marcel G.J. Nederhoff,
Pieter A.F.M. Doevendans, Gerard Pasterkamp, Gustav J. Strijkers, Cees J.A. van Echteld
International Journal of Cardiovascular Imaging, in press
DOI: 10.1007/s10554-012-0093-6
81
Chapter 5
Abstract
Myocardial edema can arise in several disease states. MRI contrast agent can accumulate
in edematous tissue, which complicates differential diagnosis with contrast-enhanced
(CE)-MRI and might lead to overestimation of infarct size. Sodium Chemical Shift
Imaging (23Na-CSI) may provide an alternative for edema imaging. We have developed
a non-infarct, isolated rat heart model with two levels of edema, which was studied
with 23Na-CSI and CE-MRI. In edematous, but viable tissue the extracellular sodium
(Na+e) signal is hypothesized to increase, but not the intracellular sodium (Na+i) signal.
Isolated hearts were perfused at 60 mmHg (n=6) and 140 mmHg (n=5). Dimethyl
methylphosphonate (DMMP) and phenylphosphonate (PPA) were used to follow
edema formation by
P-MR Spectroscopy. In separate groups, Thulium(III)1,4,7,10
31
tetraazacyclododecane-N,N /,N //,N ///-tetra(methylenephosphonate (TmDOTP5- )and
Gadovist were used for 23Na-CSI (n=8) and CE-MRI (n=6), respectively.
PPA normalized signal intensity (SI) was higher at 140 mmHg vs. 60 mmHg, with a ratio
of 1.27±0.12 (p<0.05). The (DMMP-PPA)/dry weight ratio, as a marker of intracellular
volume, remained unchanged. The mid-heart cross sectional area (CSA) of the left
ventricle (LV) was significantly increased at 140 mmHg. In addition, at 140 mmHg, the
LV Na+e SI increased with a 140 mmHg/60 mmHg ratio of 1.24±0.18 (p<0.05). Na+i SI
remained essentially unchanged. With CE-MRI, a subendocardially enhanced CSA was
identified, increasing from 0.20±0.02 cm2 at 60 mmHg to 0.31±0.02 cm2 at 140 mmHg
(p<0.05).
Edema shows up in both CE-MRI and Na+e. High perfusion pressure causes more
edema subendocardially than subepicardially. 23Na-CSI is an attractive alternative for
imaging of edema and is a promising tool to discriminate between edema, acute and
chronic MI.
82
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
5.1Introduction
Myocardial interstitial edema can arise in several disease states or following clinical
interventions, including myocardial ischemia and infarction [1-11], arterial [12]
and pulmonary hypertension [13], myocarditis [14], cardio-pulmonary bypass and
cardioplegic arrest [15] and cardiac transplantation [16]. The causes of edema may
vary: in patients with hypertension, differences in hydrostatic pressure may dominate,
whereas in patients with coronary artery disease, differences in osmotic pressure and
microvascular permeability are probably more important [17]. 1H-based MRI methods
such as contrast-enhanced (CE)-MRI and T2-weighted MRI are valuable tools to assess
injured myocardium. Indeed, CE-MRI has become the gold standard to delineate
chronic myocardial infarction [18;19]. However, the accuracy of CE-MRI to delineate
acute myocardial infarction has been questioned because of the accumulation of
contrast agent in the expanded extracellular compartment of both infarct and periinfarct regions [1-4;14]. Apparently, in this phase CE-MRI much resembles T2‑weighted
MRI, which has shown great promise to delineate the area at risk (AAR) and to quantify
myocardial salvage in patients with acute coronary syndromes after events related
to the formation of edema [5-11]. However, both these methods depend on relative
regional differences in myocardial signal intensity and are perhaps less suitable to
identify globally edematous tissue associated with any of the other disease states
mentioned above, such as hypertension or myocarditis.
An alternative approach to assess myocardial edema is imaging of myocardial sodium
with 23Na-MRI, since we expect the extracellular sodium (Na+e) signal to increase with
increasing interstitial edema because of the expansion of the extracellular space. In
a number of studies an elevated total sodium (Na+t) signal has been observed after
myocardial infarction, both in humans and animals [19-22] Particularly, imaging intra(Na+i) and extra- (Na+e) cellular sodium separately can provide more precise information
to assess the sodium gradient and cell membrane integrity [23]. This can be achieved
with chemical shift imaging (23Na-CSI) in combination with a paramagnetic chemical
shift agent to separate Na+i from Na+e signals [23;24].
83
Chapter 5
We hypothesize that, in viable tissue with cell membranes still intact but with a high
level of extracellular edema, Na+i signal intensity is normal because viable cells should
be able to maintain a normal Na+ gradient and a normal intracellular volume , whereas
Na+e and CE-MRI signal intensities will be increased because of the larger extracellular
space. To characterize edema independent from ischemia, this study consisted of
two parts. First, using phosphorus-31 magnetic resonance spectroscopy (31P MRS), we
developed an isolated heart model with two levels of extracellular edema based on
crystalloid perfusion at different pressures. Secondly, we aimed to characterize edema
in non-ischemic myocardial tissue with CE-MRI and 23Na-CSI.
5.2Methods
Animal and heart preparation
Male Wistar rats (n=25, body weight=300 to 350 g) were anesthetized with isoflurane
inhalation and given heparin (500 IU/Kg i.v.). Subsequently, the hearts were rapidly
excised and placed in modified, ice-cold Krebs-Henseleit (KH) buffer. After cannulation
of the aorta, perfusion was initiated according to Langendorff at a constant pressure
(60 mmHg) with KH at 37°C. To further prepare the heart, a drain was inserted through
the apex via the mitral valve to remove Thebesian flow. A latex balloon was inserted
into the left ventricle through the mitral valve and filled with water to reach a left
ventricular end-diastolic pressure (LVEDP) of 7.5 mmHg. Two pressure transducers
(MP-15, Micron Instruments) were used to monitor perfusion pressure and to measure
the isovolumic contractile pressure of the left ventricle via the inserted balloon. Left
ventricular developed pressure (LVPD) was determined as the difference between left
ventricular systolic pressure (LVSP) and LVEDP. To maintain a heart rate of 300 beats/
min, the outflow tract of the right ventricle was connected with two copper-wire
electrodes to a (4-7 V, 0.5 ms) stimulator (model S88, Grass Instruments, Quincey, MA,
USA). Heart rate was determined from the pressure curve. Coronary flow was measured
with a flow probe (Skalar instruments, Delft, The Netherlands). The data were recorded
with a PowerLab Data Acquisition System (ADI Instruments, Australia). Hearts were
placed in a 20-mm-diameter NMR tube and positioned inside the MR scanner. The
84
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
experimental protocol was in accordance with the guidelines of the Committee for
Animal Experiments of the University Medical Center Utrecht, The Netherlands.
Perfusion solutions
The modified KH buffer contained (in mmol/L): 119.0 NaCl, 4.7 KCl, 1.0 MgCl2, 24.0
NaHCO3, 1.3 CaCl2, 11.0 glucose, and 5.0 Na-pyruvate. Before perfusion, the KH buffer
was filtered through 8 µm (Millipore, Bedford, MA) filters. The solution was equilibrated
with 95% O2 and 5% CO2, resulting in a pH of 7.4 at 37°C, prior to the addition of the
CaCl2. For 31P-MRS, we included in the buffer 10 mM dimethyl methylphosphonate
(DMMP), which distributes evenly in intra- and extracellular spaces, as a marker of
total water space and 10 mM phenylphosphonate (PPA), which is non-permeant, as an
extracellular space marker [25]. In the 31P spectra, the DMMP and PPA signals can be easily
discriminated from endogenous phosphate signals. The difference between DMMP
and PPA signals was calculated to determine the intracellular space. After preparations,
hearts were allowed to stabilize at 60mmHg. For
tetraazacyclododecane-N,N ,N ,N
/
//
Na-CSI, Thulium(III)1,4,7,10
23
-tetra(methylenephosphonate) (TmDOTP5-, 3.5
///
mM) [26] was included in the buffer as a shift reagent (SR), to separate intra and extracellular Na+ signals. TmDOTP5- is also non-permeant and therefore only interacts with
Na+e, causing a shift in resonance frequency and leaving the Na+i resonance at its
original frequency.
To correct for Ca2+ binding by the shift reagent, the total Ca2+ added to the buffer was
increased to 3.42 mM, which resulted in a free [Ca2+] of 0.85 mmol/L as measured by
a HI 4004-51 Ca2+ Ion Selective Electrode (ISE) connected to a pH/ISE meter (HI 3221,
Hanna instruments). For CE-MRI scans, 1.3 mM Gadobutrol (Gadovist, Bayer Schering
Pharma AG, Berlin, Germany) was included in the buffer.
Modification of Langendorff set up
Usually, we aspirate the effluent from a level above the heart to keep the hearts
submerged for better magnetic susceptibility matching. However, the effluent
generates unwanted extra-cardiac 23Na and 31P signals. Two approaches were used to
85
Chapter 5
reduce these unwanted signals. For the 31P-MRS study, the effluent was removed from
the bottom of the NMR tube, sacrificing the susceptibility matching. For the 23Na-CSI
study, an additional perfusion line ending at the bottom of the tube was used to flush
the tube at a rate of 30 ml/min with a buffer in which Na+ was replaced by Li+. In this
case, the hearts were kept submerged.
MRS and MRI protocols
A vertical 9.4 T, 89 mm diameter bore scanner was used, equipped with a 1500 mT/m
gradient system, which was interfaced with an AVANCE 400 DRX spectrometer (Bruker,
Germany). All heart preparation including coil tuning and matching and optimization
of magnetic field homogeneity was performed at a perfusion pressure of 60 mmHg
and took 30-34 min as indicated in Figure 1.
P-MRS
31
Phosphorus spectroscopy was performed with a 20 mm multi-nuclear probehead
(Bruker, Germany).
P spectra were acquired for 10 min by accumulation of 32
31
consecutive free induction decays (FIDs) following 90o pulses, using 2048 data points
and 12.9 KHz spectral width (SW). Pulse repetition time was 18.75 s to achieve a fully
T1 relaxed signal.
CE-MRI and 23Na-CSI
The imaging protocol consisted of the following steps. First, 2- and 4-chamber view
scout scans were obtained with a 1H 20-mm-diameter birdcage coil (Bruker, Germany)
to plan a single mid-ventricular short-axis slice. Next, cine imaging was performed
with a self-gated gradient‑echo sequence (TE = 1.9 ms, TR = 5.2 ms, number of movie
frames = 15, slice thickness = 2.5 mm, matrix = 256x256, field of view (FOV) = 2x2 cm2).
Gated CE-MRI scans were performed with T1-weighted short-axis multi-slice FLASH
sequence, with the following parameters: TR = 63 ms, TE = 2.8 ms, FA = 75o, NA=8,
6 slices of 2.5 mm, matrix = 256x256, FOV = 2x2 cm2. To monitor the arrival of Gd
containing buffer, eight serial CE-MRI scans were performed. The last scan in this series
was used for edema analysis.
86
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
Figure 1: Schematic representation of experimental protocols
(A) 31P-MRS, (B) 23Na-chemical shift imaging (23Na-CSI), (C) Contrast-enhanced MRI
(CE-MRI).
For 23Na-CSI, the 1H birdcage coil was replaced with a 23Na 20-mm-diameter birdcage
coil (Bruker, Germany). To verify the heart position, scout images were obtained and
subsequently a 5 mm mid-axial ventricular slice was acquired as a reference image
with a 23Na self-gated FLASH sequence (TE = 2.4 ms, TR= 25 ms and 128 repetitions,
50o FA and SW = 7714 Hz, matrix = 32x32 and FOV = 2x2 cm2). To monitor the shift of
Na+e after switching to the TmDOTP containing buffer, 23Na MR spectra were acquired
23
with a 75o pulse, SW = 5.5 kHz, TR = 105 ms, and 512 data points. Subsequently,
acquisition weighted 2D-23Na CSI was performed with the following parameters:
87
Chapter 5
10,000 acquisitions, number of averages (NA) in center of k-space = 55, TR = 30 ms, 90o
0.5 ms sinc pulse, BW = 12420 Hz, FOV = 2x2 cm2, matrix = 16x16 and reconstruction
matrix = 32x32. FIDs were collected with 128 complex data points and 100.8 µs dwell
time resulting in an acquisition time of 25.8 ms/FID. Gradient switching resulted in an
acquisition delay of 0.91 ms.
Experimental protocol
Two studies were performed. First, we used
P-MRS to monitor the formation of
31
edema in 2 groups of isolated hearts. Secondly, edema was characterized with CE-MRI
and 23Na-CSI. Schematic diagrams of the experimental protocols for the 2 studies are
shown in Fig. 1.
P-MRS group
31
After preparing the heart at 60 mmHg perfusion pressure, hearts were continuously
perfused with PPA/DMMP-containing buffer for 70 min (Fig. 1A) either at 60 mmHg,
n=6 or at 140 mmHg, n=5, to induce different levels of edema. As depicted in figure1A,
P MR spectra were acquired as an average over 10 min to track the change of PPA
31
and DMPP signals and to measure the extent of extra-cellular edema during these 70
minutes of perfusion.
Na-CSI group
23
Hearts (n=8) were perfused at 60 mmHg and subsequently at 140 mmHg for 60 min
(Fig. 1B). After stabilization of the isolated hearts at 60 mmHg, 1H scout images were
acquired followed by 2- and 4-chamber cine MRI. Thereafter, a cine scan was acquired
for a mid-axial slice. Subsequently, the hearts were perfused with buffer containing
SR and 23Na spectra were acquired to monitor the arrival of buffer containing SR. 23Na
scout images were obtained to verify the anatomical position of the heart. Then a
reference image of a mid-axial slice was acquired, followed by CSI of the same slice.
The same scan protocol was repeated after 60 min of perfusion at 140 mmHg.
88
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
CE-MRI group
Hearts (n=6) were perfused at 60 mmHg (Fig. 1C) and subsequently at 140 mmHg for
60 min. After stabilization of the isolated hearts at 60 mmHg, 1H scout images were
acquired, followed by 2-chamber and 4-chamber cine MRI. Next, a cine scan was
acquired for a mid axial slice. Subsequently, the hearts were perfused with buffer
containing Gadovist. Eight serial CE-MR scans were acquired every 2 min at 60 mmHg
to monitor the arrival of Gadovist containing buffer. The same MR scan protocol was
repeated after 60 min of perfusion at 140 mmHg.
Data analysis
For 31P-MRS, all spectra were analyzed with jMRUI 4.0 software. After a polynomial
baseline correction, the spectra were fitted with AMARES, a time-domain based fitting
algorithm. The 23Na-CSI data were zero-filled in both the spatial and spectral domain
resulting in a 256x64x64 reconstruction size. To optimize peak separation a LorentzianGaussian multiplication was applied. Epi- and endocardial contours of the left ventricle
were manually drawn on end-diastolic 1H images to calculate the left ventricular cross
sectional area (CSA). Subsequently, these contours were copied to the 23Na-CSI images,
to determine the 23Na+e and 23Na+i average signal intensities of the entire left ventricle.
For CE-MRI, Epi- and endo-cardial LV contours were drawn on the 1H axial cine images
and then copied to CE-MR images. Subsequently, the CSA of the hyperenhanced areas
were delineated manually. The CSA and the signal intensities of both the entire LV
myocardium and the hyperenhanced areas were determined with Bruker Paravision
4 software.
Tissue water content
At the end of the MRI measurements, the hearts were cut open and blotted dry
with a soft tissue. The hearts were then weighed to determine their wet weights.
Subsequently, the dry weights of the hearts were measured after desiccation in an
oven at 50 oC for 48 hours. The wet/dry (W/D) weights of the isolated rat hearts were
then determined.
89
Chapter 5
Statistical analysis
Results were expressed as mean±standard error of the mean for each group of
animals. A paired t-test was performed to compare mean values for each time point
post perfusion. Results were considered significantly different for a two-tailed p value
of less than 0.05. The correlation test of PPA values against W/D ratios was performed
using a Pearson correlation with two‑tailed p values.
5.3Results
P-MRS
31
The addition of PPA and DMMP to the perfusate and the non-submerged state of the
heart did not affect the stability of the heart, in terms of steady heart rate and coronary
flow. The “dry” setup reduced the shimming quality as expected. The coronary flow was
24.0±0.5 ml/min for the 140 mmHg group and 9.9±0.4 ml/min for the 60 mmHg group.
Figure 2 shows examples of
P MR spectra of an isolated rat heart (Fig. 2A) still
31
submerged in the NMR tube and (Fig. 2B) after the effluent was removed from the
bottom of the tube, both during perfusion with perfusate containing PPA and DMMP.
Figure 3A summarizes the PPA signal intensities normalized by heart dry weight
(PPA/dry weight), whereas Fig. 3B depicts the difference between DMMP and PPA
signal intensities normalized by heart dry weight [(DMMP-PPA)/dry weight], both as
function of time for the 60 and 140 mmHg groups. Ten minutes after switching to
the phosphonates-containing buffer, PPA/dry weight was 30.6±3.2 and 34.7±2.7 for
60 and 140 mmHg, respectively. From 50 min onward, PPA/dry weight differences
between the 60 mmHg and 140 mmHg group were significant (p<0.05). After 70
min PPA/dry weight was 51.5±4.5 at 140 mmHg vs. 40.6±1.4 at 60 mmHg, with a
140 mmHg/60 mmHg ratio of 1.27±0.12. The calculated (DMMP-PPA)/dry weight,
as a marker for the intracellular space, remained at approximately baseline values
(29.9±2.1) during the entire perfusion period for both groups.
90
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
Figure 2: Typical examples of 31P-MRS spectra of an isolated rat heart, (A) while the heart is
submerged in buffer and (B) after removing the effluent.
The signal of the three phosphate groups of ATP, phosphocreatine (PCr),
phenylphosphonate (PPA) as extracellular space marker, Methylphosphonate (MPA),
a solution in a small glass capillary as a reference, and dimethyl methylphosphonate
(DMMP) as a marker of total water space are pointed out. MPA reference volume in
(B) was different from that in (A).
The higher level of edema, deduced from a larger PPA signal at 140 mmHg, was
corroborated by an increase in heart wet/dry weight ratio (W/D) of the 140 mmHg
group in comparison to the 60 mmHg group (Fig. 3C, 6.5±0.1 vs. 6.0±0.1, respectively,
p=0.05). Moreover, a positive, significant correlation was found between W/D and PPA/
dry weight (Fig. 3D, Pearson correlation coefficient 0.71, p=0.02).
Na-CSI
23
Baseline cine MRI and 23Na-MRI were performed at 60 mmHg perfusion pressure, after
which the perfusion pressure was increased to 140 mmHg (Fig. 1B). The hearts remained
stable during perfusion with SR (TmDOTP). However, immediately after switching to
SR-containing buffer, the hearts left ventricular developed pressure (LVDP) dropped
from 60.0±4.0 mmHg to 34.0±4.3 (p<0.05) at 60 mmHg and from 99.4±6.5 to 48.0±4.3
mmHg (p<0.05) at 140 mmHg as a result of the reduction of the free Ca2+ concentration
due to the affinity of the shift reagent to Ca2+ [26]. During perfusion with SR-containing
buffer, two 23Na peaks were observed in the acquired spectra (Fig. 4). A large peak
91
Chapter 5
Figure 3: 31P-MRS data on extra- and intracellular space markers and heart wet /dry weights
from isolated rat hearts perfused at 60 mmHg and 140 mmHg
(A) PPA/heart dry weight as a function of time. (B) (DMMP-PPA)/heart dry weight as
a function of time. (C) Normalized heart wet/dry and PPA/heart dry weight ratios 70
min after switching to the phosphonates containing buffer. (D) Heart wet/dry weight
ratio vs. PPA/heart dry weight ratio.
downshifted by 2.2±0.2 ppm corresponding to Na+e and mainly representing signal
from the vasculature and interstitium, and the Na+i intracellular sodium peak at 0 ppm.
Similar Na+i and Na+e peaks were seen in the localized spectra of the CSI data.
Figure 5 shows 1H longitudinal and axial views of an isolated rat heart and corresponding
Na+e and Na+i images at 60 mmHg and 140 mmHg. The balloon in the left ventricle
was filled with Na+-free solution and therefore appeared dark in 23Na-CSI images. The
right ventricular cavity had very high signal intensity, because it was filled with Na+containing effluent. The Na+i images were of considerable lower intensity than the Na+e
images because of the lower Na+i concentrations in the myocardium. As can be readily
seen in the cine 1H MR images (and in the 23Na images), the CSA of the mid-axial LV
92
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
myocardium showed a considerable increase at 140 mmHg, going from 0.66 ± 0.03
cm2 at 60 mmHg to 0.91±0.03 cm2 at 140 mmHg (p<0.05) with a 140 mmHg/60 mmHg
ratio of 1.40±0.08, in good agreement with the PPA data.
Figure 4: Na-NMR spectra acquired with control buffer (A) and with shift reagent (SR)
23
TmDOTP containing buffer.
In the absence of TmDOTP (control buffer, A), the on-resonance was seen at 0 ppm.
After 10 min of continuous perfusion with TmDOTP, Na+e peak was separated from
Na+i peak. Reference: a small glass capillary, which contained a known amount of
TmDOTP.
In addition to the increase of the LV CSA, also an increase of LV Na+e relative signal
intensity was observed, going from 17.4±1.7 at 60 mmHg to 21.6±2.2 at 140
mmHg (p<0.05), with a 140 mmHg/60 mmHg ratio of 1.24±0.18 (p<0.05), in good
agreement with CSA and PPA data. We also found an increase in LV Na+i relative signal
intensity, going from 1.8±0.2 at 60 mmHg to 2.0±0.2 at 140 mmHg (p<0.05), with a
140 mmHg/60 mmHg ratio of 1.11±0.17. However, the shift reagent resulted in a clear
but not complete separation between Na+e and Na+i CSI signals. The Na+e signal is at
least an order of magnitude larger than the Na+i signal, and spectral deconvolution
93
Chapter 5
Figure 5: Representative MR images of isolated rat heart perfused at 60 mmHg and 140
mmHg
Single frame 1H-cine MRI and extracellular and intracellular
Na-CSI. B=balloon,
23
CSA=cross sectional area of the LV, RV=right ventricular lumen.
shows that the increase in the Na+i signal can be explained for > 80% by the overlap
with the increased Na+e signal. Therefore, the corrected LV Na+i signal shows no change
when going from 60 mmHg to 140 mmHg.
CE-MRI
After switching to the buffer with Gadovist, heart function remained stable. Baseline
cine MRI and CE-MRI were performed at 60 mmHg, after which the perfusion pressure
was increased to 140 mmHg. Figure 6 shows a series of T1‑weighted MR images every
two minutes after switching to Gd-containing buffer at 60 mmHg and 140 mmHg. The
Gd started to appear in both groups after 4 min. In addition, CE-MR images showed an
increased subendocardial enhanced area compared to the rest of the myocardium at
60 mmHg and even more so at 140 mmHg. From 10 min onwards the signal intensity
of this hyper-enhanced area, delineated with a red dashed line, remained constant.
Figure 7 shows typical CE-MR images at the beginning of the 60 mmHg perfusion
and later at 140 mmHg and corresponding signal profiles across the images. As can
be readily seen in the CE-MR image, the CSA of the hearts during perfusion at 140
mmHg was increased. From the cine images, the CSA of the mid-axial slice of the LV
94
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
Figure 6: Representative T1-weighted MR images at 60 mmHg and 140 mmHg of a midaxial slice of an isolated rat heart.
Scans were acquired every 2 min after switching to Gadovist containing buffer.
B=balloon.
myocardium was significantly (p=0.03) increased from 0.74±0.04 cm2 at 60 mmHg to
0.90±0.03 cm2 at 140 mmHg, with a 140 mmHg/60 mmHg ratio of 1.22±0.08.
The enhanced subendocardial area (Fig. 7, region outlined by the red dashed line)
at 140 mmHg was significantly larger than at 60 mmHg (0.20±0.02 cm2 vs. 0.31±0.02
cm2, p=0.01), with a ratio of 1.55±0.07. Relative signal intensities in these enhanced
areas were 15.1±0.7 at 60 mmHg and 16.3±1.0 at 140 mmHg (p>0.05), with a ratio
of 1.08±0.05. The relative signal intensities between the subendocardial area and
the subepicardial area were significantly different at both 60 mmHg (subendocardial
15.1±0.7 vs. subepicardial 13.4±0.6, p<0.03, ratio 1.13±0.01) and 140 mmHg
(subendocardial 16.3±1.0 vs. 14.3±0.9, p<0.03, ratio 1.14±0.01).
95
Chapter 5
Figure 7: Representative CE-MR images for the same mid-axial slice of an isolated rat
heart perfused (A) at 60 mmHg and (B) at 140 mmHg. (C) The signal profiles
corresponding to the red lines across the images.
The enhanced subendocardial area (marked with red dashed line) at 140 mmHg was
larger than at 60 mmHg (p=0.01), which can also be appreciated from the signal
profile (C). The signal intensity of the subendocardial area was higher than the
epicardial area both at 60 mmHg and 140 mmHg (p<0.03).
96
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
5.4Discussion
Using
P-MRS we have developed a non-infarct model of extracellular edema in
31
isolated rat hearts. With this model, we have demonstrated that CE-MRI can identify
edematous myocardial tissue, most likely as a result of an increased distribution
volume of Gd. Surprisingly, we found more edema in the subendocadial area than in
the subepicardial area. We have demonstrated in this same edema model that the Na+e
signal increases with increasing edema, confirming that Na+e is a useful endogenous
marker of extracellular volume. Interestingly, a similar transmural distribution of signal
intensities as observed in the contrast-enhanced images was also observed in the Na+e
images. Na+i essentially remained unchanged in edema.
Using 31P-MRS we have shown that perfusion of hearts for an hour with crystalline buffers
at high perfusion pressures leads to formation of interstitial edema, as evidenced by
an increased distribution volume for PPA (Fig. 3). In addition, we observed a significant
correlation between the increased PPA signal and the heart wet/dry weight ratio at
the end of experiment. The increased PPA signal and the heart wet/dry ratio in our
experiment are in line with findings by Li et al. [27]. The intracellular space remained
unchanged, as evidenced by the stability of the difference between DMMP and PPA
signals. This is in contrast to the results of Li et al., who observed intracellular edema
as well, but in addition to simultaneous arterial/venous perfusion, they have used a
hyperkalemic buffer, which may have contributed to the intracellular edema.
The presence of edema could also be confirmed by the cine 1H-MRI. The CSA in both
Na-CSI (Fig. 5) and CE-MRI (Fig. 7) groups showed a significant increase at 140 mmHg
23
compared to 60 mmHg perfusion pressure, indicating an expanded volume of the
extracellular compartment caused by myocardial interstitial edema. The expanded
extracellular compartment is in line with the increased PPA signal and the heart wet/
dry weight ratio of the 31P-MRS study.
In parallel to the increased CSA, the LV Na+e signal intensity increased at 140 mmHg,
corroborating an increased level of interstitial edema. The absence of intracellular
97
Chapter 5
edema in the used model, as found with the 31P-MRS intracellular space measurements,
was corroborated by the essentially unchanged LV Na+i signal, after correction for
spectral overlap. Previously, we have used 23Na-CSI to study myocardial ischemia and
found increases in Na+i signal of 200-300%, where overlap with the Na+e signal had
negligible effects on Na+i signal intensity [23;28].
With CE-MRI at 60mmHg and 140 mmHg (Fig. 6), the myocardium appeared dark in
the T1 weighted image prior to Gd perfusion, whereas it appeared brighter after Gd
perfusion, confirming the arrival of Gd in the myocardium. In addition, higher signal
intensities of the mid-axial slice of the LV myocardium and the hyperenhancedsubendocardial area (Fig. 7) were observed at 140 mmHg compared to perfusion at 60
mmHg. One might assume a similar increase in extracellular Gd content as observed for
Na+e. However, Gd affects both intra- and extracellular water, and therefore the effects
of the increased extracellular Gd content are diluted compared to Na+e. Furthermore,
the relationship between [Gd] and signal intensity is complex and non-linear.
We consistently observed transmural heterogeneous enhancement of the LV
myocardium (Figs. 6-7) after perfusion with Gd. The subendocardial area showed
higher signal intensity, suggesting more edema in the subendocardial area than
in the subepicardial area. This is most likely related to the intramyocardial pressure
gradient, with this pressure being higher in the subendocardial area. In the model
used, hydrostatic pressure difference is the most important, if not the only, source
of edema, whereas in vivo differences in microvascular permeability and osmotic
pressure may also contribute to the formation of edema [17]. The subendocardial area
with higher signal intensity was significantly larger in hearts perfused at 140 mmHg
than at 60 mmHg, confirming more edema at higher perfusion pressure. This finding
might provide an explanation for observed late enhancement in patients with arterial
hypertension in the absence of myocardial infarction [29]. Interestingly, a similar
transmural distribution of signal intensities as observed in the CE-MR images can also
be seen in the Na+e images (Fig. 5). However, since these hearts belong to different
groups and, in addition, differences in spatial resolution, slice thickness and cardiac
98
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
gating exist between CE-MRI and 23Na-CSI groups, direct correlation of these patterns
is not possible.
Whether 23Na-CSI is indeed better capable than CE-MRI to distinguish between normal,
edematous and infarcted tissue, in acute and chronic myocardial infarction, requires
further experiments in an infarct model.
Methodological considerations
––
In addition to intracellular and interstitial contributions, the signals from the
total water space marker DMMP and the extracellular space marker PPA include
also contributions from vasculature, ventricular cavities, and perfusion and
suction tubing adjacent to the rat heart. Nevertheless, our results clearly suggest
a significant increase of the myocardial extracellular space within an hour of
perfusion at higher pressure.
––
In biological tissue, 23Na (in Na+ ions) is present in a much lower concentration
than 1H in H2O. This explains, together with the lower MR sensitivity of 23Na, the
low spatial and temporal resolution of our 23Na-images compared to the 1H-MR
images. To increase the SNR, we have chosen a thicker slice and bigger voxel size,
and we have used k-space weighted acquisition to avoid Gibbs ringing. Moreover,
the fact that 23Na has short relaxation times enabled us to use short repetition
times. To partially overcome the low spatial resolution in the 23Na-images, we have
used 1H-MR images as anatomical reference.
––
In this model, myocardial interstitial edema might already be present at 60
mmHg, resulting in a larger extracellular space than observed in vivo. However,
the observed increase of Na+e at 140 mmHg clearly suggests a further expansion
of the extracellular compartment, which was corroborated by 31P- and 1H-MR data.
––
Na+e mostly shows mono-exponential T2 relaxation (T2 = 38.7 ms, Van Emous et
al. [24]) whereas Na+i shows bi-exponential T2 relaxation with usually 60% of the
signal fast relaxing (T2f = 2.3 ms [24]) and 40% slow relaxing (T2s =18.9 ms [24]).
The CSI method requires an acquisition delay (0.91 ms), which is in the range of
the T2f component, causing a non-negligible loss of Na+i signal. However, although
Van Emous et al. have observed changes in T2 relaxation during ischemia and
99
Chapter 5
reperfusion, no such changes have been observed during normal perfusion such
as used in the current experiments. Therefore, no correction was applied. Similar
arguments apply with respect to T1 relaxation.
––
It is important to mention that our CE-MRI protocol is different from that used
clinically to assess infarct size. We have used continuous perfusion of contrast
agent, whereas in the clinic a bolus of contrast agent is administered. Due to the
absence of an infarct and due to time constraints caused by the high rat heart rate,
nulling of ‘remote’ myocardium is not possible in our model.
Clinical implications
With a non-infarct model of edema, we have demonstrated that tissue edema shows
up in CE-MRI. This suggests that CE-MRI might lead to wrongful diagnosis of infarction
in other cardiac disease states exhibiting edema or to overestimation of infarct size
[1;2]. 23Na-CSI might be an attractive tool to resolve these issues. We have shown that
viable – but edematous myocardium – is characterized by increased Na+e and normal
Na+i signals. Previously, we have shown that during global low flow and complete
regional ischemia, Na+i signals increase several-fold, with partial recovery upon
reperfusion [23]. Furthermore, we have previously observed in a model of chronic
infarction after coronary artery ligation, a substantial increase of the Na+e signal and a
complete absence of Na+i signal in the infarct area [30;31]. These characteristics make
Na-CSI a promising tool to discriminate between edema, acute and chronic MI. So
23
far, only total Na+-MRI has been used clinically. However, to make 23Na-CSI a clinically
feasible, high field (7 T) MRI scanners are recommended and a shift reagent safe for
human use is required, which is not yet available, but could be developed, since MR
contrast agents and MR shift reagents are chemically rather similar. Alternatively, the
use of multiple quantum filtered MR methods [32;33] or ultrashort-echo time (UTE)
sequences [34] could be investigated.
100
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
Acknowledgements
We thank Steven S. A. Chamuleau MD, PhD (UMC, Utrecht, the Netherlands) for useful
discussions and Thomas Oerther PhD (Bruker BioSpin GmbH, Germany) for technical
assistance with MR imaging.
Funding Sources
ICIN and the Netherlands Heart Foundation (NHS) grant 2006B066 (CvE) and (EA), and
NHS grant 2010T001 (FA). The Dutch Technology Foundation STW, Applied Science
Division of NWO and the Technology Program of the Ministry of Economic Affairs grant
07952 (GS).
Disclosures
None.
Reference List
1. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
2. Berry C, Kellman P, Mancini C, Chen MY, Bandettini WP, Lowrey T, Hsu LY, Aletras AH, Arai AE
(2010) Magnetic Resonance Imaging Delineates the Ischemic Area at Risk and Myocardial
Salvage in Patients With Acute Myocardial Infarction. Circulation-Cardiovascular Imaging
3:527-535
3. Giri S, Chung YC, Merchant A, Mihai G, Rajagopalan S, Raman SV, Simonetti OP (2009) T2
quantification for improved detection of myocardial edema. Journal of Cardiovascular
Magnetic Resonance 11:56
4. Schaefer S, Malloy CR, Katz J, Parkey RW, Buja LM, Willerson JT, Peshock RM (1988)
Gadolinium-Dtpa-Enhanced Nuclear Magnetic-Resonance Imaging of Reperfused
Myocardium - Identification of the Myocardial Bed at Risk. Journal of the American College
of Cardiology 12:1064-1072
5. Aletras AH, Tilak GS, Natanzon A, Hsu LY, Gonzalez FM, Hoyt RF, Arai AE (2006) Retrospective
determination of the area at risk for reperfused acute myocardial infarction with T2-
101
Chapter 5
weighted cardiac magnetic resonance imaging - Histopathological and displacement
encoding with stimulated echoes (DENSE) functional validations. Circulation 113:18651870
6. Edwards NC, Routledge H, Steeds RP (2009) T2-weighted magnetic resonance imaging to
assess myocardial oedema in ischaemic heart disease. Heart 95:1357-1361
7. Eitel I, Friedrich MG (2011) T2-weighted cardiovascular magnetic resonance in acute
cardiac disease. J Cardiovasc Magn Reson 13:13
8. Raman SV, Simonetti OP, Winner MW, Dickerson JA, He X, Mazzaferri EL, Ambrosio G
(2010) Cardiac Magnetic Resonance With Edema Imaging Identifies Myocardium at Risk
and Predicts Worse Outcome in Patients With Non-ST-Segment Elevation Acute Coronary
Syndrome. Journal of the American College of Cardiology 55:2480-2488
9. Ubachs JFA, Engblom H, Erlinge D, Jovinge S, Hedstrom E, Carlsson M, Arheden H (2010)
Cardiovascular magnetic resonance of the myocardium at risk in acute reperfused
myocardial infarction: comparison of T2-weighted imaging versus the circumferential
endocardial extent of late gadolinium enhancement with transmural projection. Journal
of Cardiovascular Magnetic Resonance 12:
10. Wince WB, Kim RJ (2010) Molecular imaging: T2-weighted CMR of the area at risk--a risky
business? Nat Rev Cardiol 7:547-549
11. Wright J, Adriaenssens T, Dymarkowski S, Desmet W, Bogaert J (2009) Quantification of
Myocardial Area at Risk With T2-Weighted CMR Comparison With Contrast-Enhanced CMR
and Coronary Angiography. Jacc-Cardiovascular Imaging 2:825-831
12. Mehlhorn U, Davis KL, Laine GA, Geissler HJ, Allen SJ (1996) Myocardial fluid balance in
acute hypertension. Microcirculation 3:371-378
13. Davis KL, Mehlhorn U, Laine GA, Allen SJ (1995) Myocardial edema, left ventricular
function, and pulmonary hypertension. J Appl Physiol 78:132-137
14. Korkusuz H, Esters P, Naguib N, Eldin NEN, Lindemayr S, Huebner F, Koujan A, Bug
R, Ackermann H, Vogl TJ (2009) Acute myocarditis in a rat model: late gadolinium
enhancement with histopathological correlation. European Radiology 19:2672-2678
15. Mehlhorn U, Davis KL, Burke EJ, Adams D, Laine GA, Allen SJ (1995) Impact of
cardiopulmonary bypass and cardioplegic arrest on myocardial lymphatic function. Am J
Physiol 268:H178-H183
16. Kong XQ, Wang LX, Kong DG (2007) Cardiac lymphatic interruption is a major cause for
allograft failure after cardiac transplantation. Lymphat Res Biol 5:45-47
17. Levick JR, Michel CC (2010) Microvascular fluid exchange and the revised Starling principle.
Cardiovascular Research 87:198-210
18. Mahrholdt H, Wagner A, Holly TA, Elliott MD, Bonow RO, Kim RJ, Judd RM (2002)
Reproducibility of chronic infarct size measurement by contrast-enhanced magnetic
resonance imaging. Circulation 106:2322-2327
102
Na Chemical Shift Imaging and Gd Enhancement of myocardial edema
23
19. Kim RJ, Judd RM, Chen EL, Fieno DS, Parrish TB, Lima JA (1999) Relationship of elevated
23Na magnetic resonance image intensity to infarct size after acute reperfused myocardial
infarction. Circulation 100:185-192
20. Constantinides CD, Kraitchman DL, O’Brien KO, Boada FE, Gillen J, Bottomley PA (2001)
Noninvasive quantification of total sodium concentrations in acute reperfused myocardial
infarction using 23Na MRI. Magn Reson Med 46:1144-1151
21. Kim RJ, Lima JA, Chen EL, Reeder SB, Klocke FJ, Zerhouni EA, Judd RM (1997) Fast 23Na
magnetic resonance imaging of acute reperfused myocardial infarction. Potential to
assess myocardial viability. Circulation 95:1877-1885
22. Ouwerkerk R, Bottomley PA, Solaiyappan M, Spooner AE, Tomaselli GF, Wu KC, Weiss RG
(2008) Tissue sodium concentration in myocardial infarction in humans: A quantitative
Na-23 MR imaging study. Radiology 248:88-96
23. Jansen MA, Van Emous JG, Nederhoff MG, van Echteld CJA (2004) Assessment of myocardial
viability by intracellular 23Na magnetic resonance imaging. Circulation 110:3457-3464
24. Van Emous JG, Van Echteld GJA (1998) Changes of intracellular sodium T-2 relaxation times
during ischemia and reperfusion in isolated rat hearts. Magnetic Resonance in Medicine
40:679-683
25. Clarke K, Anderson RE, Nedelec JF, Foster DO, Ally A (1994) Intracellular and extracellular
spaces and the direct quantification of molar intracellular concentrations of phosphorus
metabolites in the isolated rat heart using 31P NMR spectroscopy and phosphonate
markers. Magn Reson Med 32:181-188
26. Buster DC, Castro MMCA, Geraldes CFGC, Malloy CR, Sherry AD, Siemers TC (1990)
Tm(Dotp)5--A 23Na+ Shift Agent for Perfused Rat Hearts. Magnetic Resonance in Medicine
15:25-32
27. Li G, Xiang B, Dai GP, Shaw A, Liu HY, Yang BF, Jackson M, Deslauriers R, Tian GH (2005) Tissue
edema does not change gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA)enhanced T-1 relaxation times of viable myocardium. Journal of Magnetic Resonance
Imaging 21:744-751
28. Van Emous JG, Nederhoff MGJ, Ruigrok TJC, van Echteld CJA (1997) The role of the
Na+ channel in the accumulation of intracellular Na+ during myocardial ischemia:
Consequences for post-ischemic recovery. Journal of Molecular and Cellular Cardiology
29:85-96
29. Yang EH, Lerman A (2005) Angina pectoris with a normal coronary angiogram. Herz 30:1725
30. Jansen, M. A., Nederhoff, M. G., and van Echteld, C. J. A. Intracellular Sodium Imaging
after Coronary Artery Ligation in Rats (2003) ISMRM 11th Scientific Meeting & Exhibition.
International Society for Magnetic Resonance in Medicine , 188. Ref Type: Conference
Proceeding
103
Chapter 5
31. Jansen, M. A., Nederhoff, M. G., and van Echteld, C. J. A. (2004) Intracellular Sodium MRI in
Chronic Myocardial Infarction. Journal of Cardiovascular Magnetic Resonance 6(1), 318.
Ref Type: Abstract
32. Navon G, Werrmann JG, Maron R, Cohen SM (1994) 31P NMR and triple quantum filtered
23Na NMR studies of the effects of inhibition of Na+/H+ exchange on intracellular sodium
and pH in working and ischemic hearts. Magn Reson Med 32:556-564
33. Dizon JM, Tauskela JS, Wise D, Burkhoff D, Cannon PJ, Katz J (1996) Evaluation of triplequantum-filtered 23Na NMR in monitoring of Intracellular Na content in the perfused rat
heart: comparison of intra- and extracellular transverse relaxation and spectral amplitudes.
Magn Reson Med 35:336-345
34. de Jong S, Zwanenburg JJ, Visser F, van der Nagel R, van Rijen HV, Vos MA, de Bakker JM,
Luijten PR (2011) Direct detection of myocardial fibrosis by MRI. Journal of Molecular and
Cellular Cardiology 51:974-979
104
Chapter 6
Differentiation of myocardial injury
in various infarct models with
23
Na Chemical Shift Imaging and
Late Gd Enhancement MRI
Eissa Aguor†, Maurits A. Jansen†, F. Arslan, Marcel G.J. Nederhoff, Cees W.A. van de Kolk,
Pieter A.F.M. Doevendans, Gerard Pasterkamp, Gustav J. Strijkers, Cees J.A. van Echteld
†
These authors contributed equally to this study
Submited
105
Chapter 6
Abstract
Using
Na-Chemical Shift Imaging (CSI) complemented by 1H-Late Gadolinium
23
Enhanced (LGE)-MRI we aimed to differentiate injured myocardium under three
conditions: acute ischemia and reperfusion (I/R) (group1), 2 hours after I/R injury
(group2), and chronic infarction (MI) (group3).
Isolated rat hearts (n=6) underwent 30, 40 or 50 min of ischemia, followed by 40 min of
reperfusion (group1). Seven rats underwent 50 min of in vivo ischemia, followed by 2
hours of in vivo reperfusion (group2). Chronic MI was induced in six rats by 3 weeks of
LCA-ligation (group3). Following interventions, group 2 and 3 hearts were also isolated
and perfused.
During ischemia, Na+i signal intensity (SI) increased to 24.4±1.4, 28.2±1.5, and 33.7±1.4
(a.u.), at 30, 40 and 50 min, respectively, but decreased during 40 min of reperfusion
(group1). In group2, Na+e appeared higher in infarct than remote (0.51±0.07 vs.
0.46±0.05, p=0.17), whereas Na+i was equal (0.0138±0.003 vs. 0.0143±0.003, p=0.75).
In chronic MI, Na+e SI was elevated in the infarct (0.51±0.05 vs. 0.32±0.03, p=0.002),
whereas Na+i was decreased (0.006±0.001 vs. 0.012±0.002, p=0.003). Infarct size (IS) from
TTC-staining corresponded to automatically determined IS from LGE and Na+e images,
at approximately 5SD thresholds (acute-MI), but in chronic-MI at 2SD thresholds.
Na+ tracking is valuable for indentifying the pathophysiological status of the heart:
highly elevated Na+i is a unique characteristic of acute ischemia, whereas the absence
of Na+i signal in combination with elevated Na+e signal is a marker of scar tissue. The
rapid normalization of highly elevated Na+i in infarcted area upon reperfusion provides
unique insights in the dynamics of infarct development. Combination of 23Na-CSI and
H-LGE-MRI shows added value for differentiation of various pathophysiological states
1
after MI.
Keywords Magnetic resonance imaging, sodium MRI, myocardium, myocardial
infarction, reperfusion, chemical shift imaging
106
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
6.1Background
MRI plays an important role in the prognostic assessment of patients with heart
disease. It is a powerful noninvasive modality to image cardiac anatomy, regional and
global function, myocardial perfusion, and viability. The pathophysiological tissue
status alters longitudinal relaxation time (T1) and transverse relaxation time (T2). Thus,
an adequate MR protocol may enable identification of the pathophysiological changes
after a myocardial infarction (MI). Commonly, these are mainly 1H-MRI techniques which
have been extensively investigated and applied to assess edema [1-3], inflammation
[4], apoptosis [5], necrosis, fibrosis [6], inflammatory-cell infiltration to myocardial
infarction sites [7], and hemorrhage, as well as iron overload in non ischemic heart
disease [7;8].
For instance, T1-weighted MRI after the injection of a gadolinium‑based contrast
agent is a common method to delineate injured myocardium [9] in so-called Late
Gadolinium Enhancement (LGE) MRI. LGE of the injured myocardium occurs because
of altered wash-in/wash-out kinetics of the contrast agent in comparison to healthy
myocardium. In this respect, the presence of edema is a confounding factor, which may
lead to LGE in the absence of necrosis and thus an overestimation of true infarct size
[10]. Moreover, selecting a standard signal threshold offset for accurate differentiation
between injured but salvageable and non-viable myocardium, both in acute and
chronic MI, is often problematic [11]. Accurate distinction between stunned and scar
tissue is clinically important to decide whether revascularization may still be beneficial
after an ischemic insult. Based on a LGE-MRI scan alone, it is difficult to discriminate
between salvageable and scar tissue. Recently, T2-weighted MRI was introduced as a
complementary technique to image the salvageable myocardium [12-14]. Although
promising, T2-weighted MRI requires further optimization [4].
Viable myocytes sustain a constant gradient between the extra-cellular sodium
(Na+e) and the intra-cellular sodium (Na+i) concentrations with an approximate ratio
of 150/10 (mM/mM). Alteration of this gradient indicates dysfunction of the Na+/K+ATPase pumps and thus relates to changes in cellular homeostasis occurring during
107
Chapter 6
and after MI [15]. Therefore, measurement of the sodium gradient by 23Na-MRI is of
particular interest and provides a direct window on cellular viability [16;17].
Data on 23Na-MRI are scarce due to the relatively lower MR sensitivity and lower in
vivo concentration in comparison to 1H-MRI [18]. Still, 23Na-MRI was investigated in
a number of studies. In human heart, total 23Na (Na+t) MR signal was significantly
elevated in injured myocardium [19]. A more complete way to assess the Na+/K+ATPase pump function for myocardial viability is to monitor Na+i and Na+e separately
facilitated by the use of a shift reagent and a chemical shift imaging (CSI) sequence.
The Na+i concentration rises shortly after myocardial ischemia because of inhibition of
sarcolemmal Na+/K+-ATPase. Na+ may enter the myocytes via many pathways, perhaps
most importantly via the Na+ channel, and the Na+-H+ exchanger (NHE) [20]. Therefore,
monitoring extracellular and intracellular Na+ separately could be more sensitive to
distinguish different pathophysiological changes states after MI than Na+t-MRI and
H-MRI alone.
1
We have previously shown that during reperfusion after ischemia induced by application
of low coronary flow in an isolated rat heart model, the elevated Na+i signal intensity
correlated to the severity of ischemia [16]. Furthermore, we have shown that a higher
level of interstitial edema in isolated heart model also increases extracellular volume
distribution in LGE-MRI and 23Na-CSI [21]. However, the complex pathophysiological
changes in acute and chronic myocardial injury remain to be fully addressed. The aim
of this study was therefore to apply 23Na-CSI complemented by 1H-LGE-MRI in isolated
rat hearts to characterize the injured myocardium in three different conditions: in
acute MI during ischemia and reperfusion, 2 hours after ischemia/reperfusion (IR)
injury, and in chronic MI.
108
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
6.2 Materials and Methods
Experimental protocol
All experimental protocols were approved by and performed according to the
guidelines of the animal experiments committee of the University Medical Center
Utrecht, the Netherlands. Male Wistar rats (body weight=300 to 350 g) were used for
this study. Three groups of animals were used. Group 1 was studied during acute MI,
group 2 two hours after reperfusion, and group 3 in the chronic phase. Details on the
several groups are given below.
Group 1: Acute MI
Six rats were used for this group. Hearts were excised under anesthesia and perfused in
a Langendorff set-up with Krebs-Henseleit (KH) buffer, using a dual perfusion cannula
which allowed independent perfusion of the left and right coronary bed [22]. 1H-MRI
survey scans were performed in axial and longitudinal orientations, followed by shortaxis cine 1H-MRI. Next, the perfusate was switched to buffer containing the shift reagent
(SR) Tm-DOTP5-. Subsequently, the left side of the hearts was subjected to regional
ischemia for 30, 40, or 50 min (flow to the other side remained unaltered), followed by
40 min of reperfusion, and in parallel, sequential 23Na-CSI short-axis scans were acquired.
After that, the left side of the heart was perfused with buffer containing a Gd contrast
agent and a T1-weighted MRI scan was acquired to assess the spatial extent of injured
myocardium. A schematic diagram of the experimental protocol is shown in Figure 1.
After 2 hours, the right side of the heart was perfused with methylene blue to determine
the area at risk for histology. After that, the whole heart was perfused with 1% triphenyl
tetrazolium chloride (TTC) to stain the viable tissue.
Group 2: Two hours reperfusion
Induction of myocardial injury
Rats (n=6) were anesthetized with isoflurane in 2:1 mixture of air (0.3 l/min) and
oxygen (0.15 l/min). A core body temperature of 37 °C was maintained during surgery
109
Chapter 6
Groups
1:
Ctrl
SR
Gd
Buffer
Preparation
+
1H MRI
25
Ctrl
23Na CSI
Ischemia
(30-50)
SR
T1 W
MRI
23Na CSI
Reperfusion
40
Ctrl
10
Time (min)
Gd
Buffer
2:
Preparation
+
1H MRI
25
23Na
CSI Washout
T1 W
MRI
20
Ctrl
SR
16
Time (min)
Gd
Buffer
3:
Preparation
+
1H MRI
25
23Na
20
CSI
T1 W
MRI
5
Time (min)
Figure 1: Schematic representation of experimental protocols
Group 1: during acute MI, group 2: two hours after reperfusion, group 3: chronic MI
group. Isolated hearts were perfused with control (ctrl) buffer, with shift reagent (SR)
TmDOTP containing buffer, and with contrast agent (Gd) containing buffer.
by continuous monitoring with a rectal thermometer and an automatic heating
blanket. The rats subcutaneously received Antisedan (Pfizer) 2.5 mg/kg, Anexate
(Roche) 0.5 mg/kg, and Temgesic (Schering-Plough) 0.1 mg/kg. Rats were intubated
and ventilated (Harvard Apparatus Inc, Holliston, Mass) with 100% oxygen. The left
coronary artery was ligated for 50 min with an 8-0 Ethilon (Ethicon) suture with a
section of polyethylene-10 tubing placed over the left coronary artery. Ischemia
was confirmed by bleaching of the myocardium and ventricular tachyarrhythmia.
Reperfusion was initiated by release of the ligature and removal of the polyethylene-10
tubing. Reperfusion of the endangered myocardium was characterized by typical
hyperemia in the first few minutes. After 50 minutes of continuous reperfusion in vivo,
the rats were injected with heparin (500 IU/Kg i.v.) via the femoral vein.
Perfusion
After a total of 2 hours of reperfusion, hearts were excised and perfused in a Langendorff
set-up with KH buffer, using a single perfusion cannula. Subsequently, 1H-MRI survey
110
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
scans were performed in axial and longitudinal orientations, followed by short-axis cine
H-MRI. Thereafter, hearts were perfused with buffer containing SR and 23Na spectra
1
were acquired to monitor the arrival of SR. 23Na-CSI was performed in a mid-axial slice.
Subsequently, the perfusate was switched to control KH buffer to washout the SR for
16 min. Next, the hearts were perfused with buffer containing a Gd contrast agent and
eight sequential 1H-LGE-MRI scans were acquired every two minutes to monitor the
arrival of contrast agent. Then the perfusion was switched to KH buffer to washout the
Gd, monitored by sequential T1-weighted 1H-MRI. At the end of the scan, hearts slices
were stained with 1% TTC for viable tissue. A schematic diagram of the experimental
protocol is shown in Figure 1.
Group 3: Chronic MI
Chronic MI of rat hearts (n=6) was induced by ligation of the left coronary artery as
described above. After permanent occlusion of the left coronary artery, the thorax was
closed and animals were allowed to recover. Three weeks after ligation, hearts were
excised and perfused in a Langendorff setup. The remainder of the protocol was the
same as for the rats in group 2. A schematic diagram of the experimental protocol is
shown in Figure 1.
Perfusates and Langendorff setup
As perfusate, a modified KH buffer was used containing (in mmol/L): 119.0 NaCl, 4.7 KCl,
1.0 MgCl2, 24.0 NaHCO3, 1.3 CaCl2, 11.0 glucose, and 5.0 Na-pyruvate. Before perfusion,
the KH buffer was filtered through 8 µm (Millipore, Bedford, MA) filters. The solution
was equilibrated with 95% O2 and 5% CO2, resulting in a pH of 7.4 at 37°C, prior to the
addition of the CaCl2.
For
Na-CSI, Thulium(III)1,4,7,10 tetraazacyclododecane-N,N
23
,N
/
,N
//
-tetra
///
(methylenephosphonate) (Tm-DOTP , 3.5 mM) [23] was included in the buffer as SR
5-
to separate Na+i and Na+e. To correct for Ca2+ binding by the SR, the total Ca2+ added
to the buffer was increased to 3.42 mM, which resulted in a free [Ca2+] of 0.85 mmol/L
as measured by a HI 4004-51 Ca2+ Ion Selective Electrode (ISE) connected to a pH/
111
Chapter 6
ISE meter (HI 3221, Hanna instruments). For T1-weighted 1H-LGE-MRI, 1.3 mM of Gd
contrast agent Gadobutrol (Gadovist) was added to the buffer.
For group 2, an additional perfusion line ending at the bottom of the NMR-tube was
used to flush the tube at a rate of 30 mL/min with a buffer in which Na+ was replaced
by Li+, keeping the extracardiac Na+ signal low and the hearts submerged for good
magnetic susceptibility matching.
Hearts were rapidly excised and placed in ice-cold KH buffer. The hearts were mounted
via the aorta with a single perfusion cannula (groups 2 and 3) or via a home-built dual–
perfusion cannula (group 1), which allowed independent perfusion of the left and right
coronary beds of the heart. After cannulation, perfusion was initiated according to
Langendorff with KH buffer at 37oC. To further prepare the heart, a drain was inserted
through the apex via the mitral valve to remove Thebesian flow. A latex balloon was
inserted into the left ventricle through the mitral valve and filled with water to reach a
left ventricular end-diastolic pressure (LVEDP) of 7.5 mmHg. Two pressure transducers
(MP-15, Micron Instruments) were used, one to monitor perfusion pressure and a
second one to measure the isovolumetric contractile pressure of the left ventricle
via the inserted balloon. Left ventricular developed pressure (LVPD) was determined
as the difference between left ventricular systolic pressure (LVSP) and LVEDP. To
maintain a heart rate of 300 beats/min, the outflow tract of the right ventricle was
connected with two copper-wire electrodes to a (4-7 V, 0.5 ms) stimulator (model S88,
Grass Instruments, Quincey, MA, USA). Heart rate was determined from the pressure
curve. Coronary flow was measured with a flow probe (Skalar instruments, Delft, The
Netherlands). The data were monitored and recorded using data Acquisition System
(PowerLab, ADI Instruments, Australia). MRI scans was initiated after placing the hearts
in a 20 mm NMR tube diameter that was positioned inside the MR scanner.
MR acquisition
A vertical 9.4 T, 89-mm-diameter bore magnet was used, equipped with a 1500 mT/m
gradient system and interfaced to an AVANCE 400 DRX spectrometer (Bruker, Germany).
112
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
1
H-MRI
Tuning and global shimming was done using a 20-mm-diameter 1H birdcage coil
(Bruker, Germany). Subsequently, scout (2- and 4-chamber view) MR scans were
acquired to plan a single mid-ventricular short-axis slice. Cine imaging was performed
with a retrospectively‑triggered gradient‑echo (IntraGateFlash), with the following
parameters: TE=1.9 ms, TR=5.2 ms, FA=10o, number of movie frames=15, slice
thickness=2.5 mm, matrix=256´256, field of view (FOV)=2´2 cm2. 1H-LGE-MRI was
performed with (group 2) an ECG-gated, short axis multi-slice FLASH sequence, using
TR=63 ms, TE=2.8 ms, FA=75o, NA=8, 6 slices of 2.5 mm, matrix=256´256, FOV=2´2 cm2
or (groups 1 and 3) with a multi-slice MSME sequence, using TR=1003 ms, TE=6.1 ms,
NA=1, 2 slices of 2.5 mm, matrix=256´256, FOV=2´2 cm2.
Na-MRS and CSI
23
For
Na MR scans, the 1H birdcage coil was replaced with a
23
Na birdcage coil
23
(Bruker, Germany). A 5 mm mid-axial slice was acquired as a reference image using
a retrospectively gated 23Na FLASH sequence, with TE=2.4 ms, TR=25 ms and 128
repetitions, FA=50o, matrix=32x32, and FOV=2´2 cm2. To monitor the shift of Na+e
after switching to the buffer with SR, 23Na MR spectra were acquired using a pulseand-acquire sequence, with TR=105 ms, FA=75o, SW=5.5 kHz, and 512 data points.
Acquisition-weighted
Na-CSI was performed with 10,000 acquisitions, NA=55
23
in center of k-space, TR= 30 or 27 ms, FA=90o, pulse=0.5 ms Sinc, BW=12420 Hz,
matrix=16x16, reconstruction matrix=32x32, and FOV=2´2 cm2. FIDs were collected
with 128 complex data points and 100.8 µs dwell time resulting in an acquisition
time of 25.8 ms/FID. The 23Na-CSI data were zero-filled in both the spatial and spectral
domain resulting in a (256x32x32) reconstruction size. To optimize peak separation, a
Lorentzian-Gaussian multiplication was applied.
Analysis
To determine average Na+ signal intensities, regions-of-interest (ROIs) were manually
drawn to delineate the myocardial free wall (injured myocardium) and septal wall
(remote) in the 1H-MR images and subsequently copied into the 23Na images. The
113
Chapter 6
average signal intensities in the ROIs for myocardial Na+i and Na+e were calculated with
Paravision 4 software (Bruker, Germany).
To semi-automatically determine infarct size in groups 2 and 3, the 1H and
Na
23
signal intensities in the remote myocardium were used. The enhanced areas in the
Na-CSI and 1H-LGE-MR images were quantified with a semiautomatic algorithm
23
(Segment, Medviso AB, Sweden) based on signal-intensity thresholds (2, 4, and 6
standard deviations (SDs) above remote myocardium) as proposed by Kim et al. [24].
To avoid contamination by extra-cardiac and intra-ventricular signals, epicardial and
endocardial contours of the left ventricle were manually drawn on end-diastolic 1HMR images. Subsequently, these contours were copied to the 23Na-CSI and 1H-LGE-MR
images.
TTC stained tissue sections were photographed under a microscope with a digital image
camera. Infarct size (IS) and total LV area were manually delineated and quantified
using ImageJ software (version 1.42). The IS was then expressed as percentage of the
total LV area.
Statistical analysis
Results were expressed as mean ± standard error of the mean. A paired t-test was
performed to compare mean values. Results were considered significantly different for
a two-tailed p value < 0.05. The correlation test of the infarct size for TTC vs. LGE was
performed using a Pearson correlation with two‑tailed p values.
114
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
6.3Results
Group 1: Acute MI
Figure 2A shows a collection of short-axis 23Na-CSI and 1H-LGE-MR images, and the
corresponding TTC stained section for an isolated rat heart subjected to 40 min of
ischemia of the left side, followed by 40 min of reperfusion. As expected, regions with
high sodium concentration such as the balloon, the right ventricular cavity and the
heart-surrounding fluid appeared bright in Na+e-CSI. During ischemia, Na+i MR image
intensity increased significantly at the left side (infarct) whereas that of the right side
(remote) remained unaltered. The hyperenhanced region of Nai at the end of ischemia
Figure 2: Representative 23Na CSI and 1H MR images of a mid-axial slice of an isolated
rat heart of the acute MI (group 1) and intra- and extracellular Na+ signal in the
infarct, before ischemia, during ischemia and reperfusion
(A) Example of extracellular and intracellular 23Na CSI, LGE MRI and corresponding
TTC staining of isolated rat heart, and the corresponding 23Na CSI signal intensities
(B) before ischemia, during ischemia and reperfusion
115
Chapter 6
appears to correlate better to the TTC unstained region, than the hyperenhanced
region of Nai at the end of reperfusion. As expected, Na+e image intensity in the
infarct showed a slight decrease during ischemia whereas it increased again during
reperfusion. TTC infarct areas also corresponded to 1H-LGE-MRI infarct areas.
Quantitative analysis of the 23Na-CSI data revealed the dynamic changes of the Na+i
and Na+e signals over time (Figure 2). Prior to ischemia, relative Na+i and Na+e signal
intensity values, derived from magnitude images, were in remote tissue 6.5±0.5 and
286±23, respectively, and 6.4±0.6 and 265±28 in the region that would become infarcted
at later time points. Values in the remote region did not change during the course of
the experiment. During ischemia, Na+i increased linearly (R=0.97) with relative signal
intensities of 24.4±1.4, 28.2±1.5, and 33.7±1.4, at 30, 40 and 50 min, respectively. 40
min after reperfusion, Na+i values had decreased towards baseline values of 7.6±0.1 and
5.3±2.3, respectively, for the 30 and 40 min hearts, whereas for the heart that suffered
50 min of ischemia Na+i remained elevated at 13.10, suggesting more severe myocardial
damage or poor reflow.
Relative Na+e signal intensities in the infarct (Figure 2B) decreased to 201±58, 183±39,
and 173, at 30, 40, and 50 min of ischemia, respectively. After 40 min of reperfusion, the
Na+e signal intensities had decreased but remained slightly higher than baseline values,
suggesting the presence of extracellular edema or cell damage.
Group 2: Two hours reperfusion
The isolated rat hearts that underwent 50 min of in vivo ischemia followed by 2 hours of
reperfusion had a heart rate of 305.96±0.05 beats/min during the whole experiment. The
left ventricular developed pressure (LVDP) decreased from 29.1±1.4 mmHg for control
buffer to 22.1±2.7 mmHg for buffer with SR. the LVDP partly recovered to 25.7±2.4 mmHg
after switching back to control buffer. For the three conditions, the coronary flow rates
were 7.87±0.45, 7.01±0.86, 6.73±1.03 mL/min, respectively.
Figure 3 depicts a time-series of T1‑weighted MR images every two minutes during
control perfusion (A), after switching to Gd-containing buffer (B) and again after
116
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
washout with control buffer (J). The enhancement in T1‑weighted MRI confirmed Gd
arrival in the myocardium. The enhancement was more obvious in the mid-anterior
and mid-anterolateral wall. After switching to the control buffer, the Gd enhanced
area started to decrease until it finally disappeared, indicating that the dynamics of Gd
wash in and wash out influence LGE MRI. The best contrast of infarct vs. remote was
observed at the 3rd MR scan (L) (approx. 6 minutes) after Gd washout.
Figure 3: Representative T1-weighted MR images during infusion and washing-out of Gd
containing buffer of a mid-axial slice of an isolated rat heart (group 2).
Scans were acquired every 2 min after switching to Gd containing buffer. A good
contrast between infarct and remote is observed after washout.
Figure 4A presents a representative example of corresponding mid short-axis 23Na-CSI,
H-LGE-MRI, and TTC slices, after in vivo ischemic injury and 2 hours in vivo reperfusion.
1
The balloon in the left ventricle contained a Na+-free solution, thus appeared dark in
Na-CSI. Regions of high sodium concentration such as in the right ventricular cavity,
23
which filled with effluent, had high signal intensity. The Na+i images displayed lower
SNR than those of Na+e because of the lower Na+i content. The Na+e signal was hyperenhanced in the mid-anterior and mid-anterolateral wall in the same region of 1H-LGE
and the unstained region in TTC, indicating the presence of myocardial injury in this area.
117
Chapter 6
Figure 4: Representative example of a mid-axial slice of an isolated rat heart with acute
inferior myocardial infarction (group 2).
(A) Extracellular and intracellular Na CS images, LGE image and corresponding TTC
staining. (B) Normalized signal intensity of extracellular and intracellular Na+ and LGE
in infarct vs. remote. (C) Infarct size delineation based on signal intensity thresholds
2, 4, and 6 SDs above remote normal myocardium. The infarct region is marked with
a red dashed line in the enhanced area of the LGE MRI.
We have specifically evaluated the changes of the Na+e and Na+i signal intensity in
the infarct and remote tissue (Figure 4B). The signal intensity was normalized to the
buffer Na+ signal. Na+e was slightly higher in infarct than remote tissue (0.51±0.07 vs.
0.46±0.05, p=0.17), whereas Na+i values were equal (0.0138±0.003 vs. 0.0143±0.003,
p=0.75). 1H-LGE-MRI corresponded to Na+e, with higher intensity in the infarct ((6.3±0.5)
x104 vs. (4.3±0.9)x104, p<0.05).
The infarct size determined from the TTC (percentage of the LV) was 17.8±3.2%. To
further characterize the MRI enhancement of the injured myocardium in more detail,
we determined the Infarct size based on 2, 4, and 6 SD cut-off thresholds above remote
118
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
signal intensity (Figure 4C). The best agreement between TTC determined infarct size,
on the one hand, and infarct size from LGE and Na+e, on the other hand, was observed
for thresholds of 4 and 6 SD. The high correspondence between LGE and Na+e confirms
that these two readouts are both markers of the extracellular volume fraction.
Figure 5: Representative example of a mid-axial slice of an isolated rat heart with chronic
myocardial infarction (group 3).
(A) Extracellular and intracellular Na CS images, LGE image and corresponding TTC
staining. (B) Normalized signal intensity of extracellular and intracellular Na+ in the
infarct vs. remote. (C) Infarct size delineation based on signal intensity thresholds 2,
4, and 6 SDs above remote normal myocardium. The scar region is marked with red
dashed line in the enhanced area of the LGE MRI and in the hypo-enhanced area of
the Nai CSI
Group 3: Chronic MI
Axial 23Na-CSI, 1H-LGE-MRI, and corresponding TTC slices of isolated rat heart 3 weeks
after LAD ligation are shown in Figure 5A. 23Na+e and LGE revealed similar enhanced
119
Chapter 6
infarcts corresponding well to the TTC infarct areas. Quantitative analysis (Figure 5B)
revealed that the normalized Na+e signal intensity was significantly elevated in the
infarct versus remote tissue (0.51±0.05 vs. 0.32±0.03, p=0.002), indicating a significant
increase in the extracellular volume fraction in the infarct scar. Na+i was significantly
decreased in the infarct as compared to remote tissue (0.006±0.001 vs. 0.012±0.002,
p=0.003).
Similar to group 2, the infarct size estimation based on signal intensity was determined
using (2, 4, and 6 SD) threshold above the remote signal intensity for LGE and Na+e
(Figure 5C). TTC based infarct size (as percentage of the LV) was 25.4±3.7%. Best
agreement with TTC staining was obtained for 2 SD. For LGE the estimated infarct size
rapidly decreased for higher thresholds. The estimation of infarct size based on Na+i
failed because of the very low Na+i signal intensity in the chronic infarct.
6.4Discussion
We have investigated the dynamic behaviour of intra- and extra-cellular sodium
using cardiac
Na CSI, combined with 1H LGE MRI under different myocardial
23
pathophysiological conditions, including acute ischemia, reperfusion, and chronic
infarcts using in vivo and in vitro rat models of myocardial ischemia and infarction.
We have demonstrated that
Na CSI is a powerful tool to identify sarcolemmal
23
dysfunction related to myocardial pathophysiological status in the time-course after
MI. Acute, totally ischemic myocardium is characterized by highly elevated Na+i levels,
caused by Na+/K+ ATPase dysfunction and continued Na+ influx, and concomitant
decrease of Na+e (group 1). In contrast, chronic MI (group 3) was characterized by
an almost complete absence of Na+i, indicating the loss of viable myocardium and
replacement by scar tissue, corroborated by the observed increased Na+e. However,
the situation in acutely reperfused myocardium (groups 1 and 2) is less clear. The
(partial) recovery of Na+i may signify recovery of Na+/K+ ATPase function, but could
be equally well explained by heterogeneous reperfusion with areas of no reflow and
120
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
persistent high Na+i in combination with areas showing advanced cell deterioration
and sarcolemmal dysfunction/rupture with loss of cellular content. The slightly
elevated Na+e could signfy edema, but could also be explained by loss of sarcolemmal
integrity. These findings are summarized below:
Status of injured myocardium
Na+i
Na+e
Acute Ischemia (without reperfusion)
+++
-
Acute reperfusion
-/0/+
0/+
---
+++
Chronic infarct (scar tissue)
Group 1: Acute MI
In the acute MI animals (group 1) the Na+i infarct signal increase during ischemia was
directly proportional to the ischemia duration. The linear Na+i increase during ischemia
is in line with previously published data from our group [20;25;26]. During reperfusion
after 30 or 40 minutes of ischemia, Na+i levels normalized. However, reperfusion after
50 minutes of ischemia resulted in a persistently elevated Nai+, which could be
interpreted as more severe injury caused by the prolonged ischemia. In parallel, Na+e
levels decreased during ischemia and were slightly higher than remote after reperfusion.
We have demonstrated before [25] that recovery of Na+i levels after a substantial ischemic
rise, is not necessarily correlated with recovery of function and that post-ischemic Na+-H+
exchange blockade is not cardioprotective, since most of the rise of Na+i has already taken
place during ischemia, resulting anyway in detrimental reverse Na+-Ca2+ exchange and
subsequent Ca2+ overload. In view of the good correspondence of TTC infarct area and
the area of end-ischemic enhanced Na+i and the poor correspondence with the smaller
area of end-reperfusion enhanced Na+i, a more likely interpretation of the observed
changes in Na+i is that the recovery of Na+i signifies detrimental Ca2+ overload and the
persistently elevated Na+i in the core of the infarct represents an area of poor reflow.
121
Chapter 6
Group 2: Two hours reperfusion
We have examined isolated rat hearts with 1H-LGE-MRI and 23Na-CSI following in vivo
coronary ligation and in vivo reperfusion. We did not observe a significant difference
of the Nai signal intensity between the injured myocardium and the remote region.
This might be construed as full recovery, but in view of the discussion on group 1, it is
more likely that the dynamics of 2 hours of reperfusion are such that the entire infarct
area has been reperfused, leading to recovery of Na+i at the expense of Ca2+ overload,
setting the stage for further infarct development. Alternatively, the findings may also be
explained by a combination of myocardial cells in which Ca2+ overload has resulted in
sarcolemmal disintegration, with myocardial cells exhibiting persistently elevated Na+i
due to poor reflow, resulting in no net change in Na+i. However, we have to keep in
mind that, in contrast to group 1, in case of areas of complete no reflow in group 2,
the shift reagent would not be able to reach these areas, which would result in Na+e
resonating at the Na+i position and a concomitant strong increase in signal intensity,
which we obviously have not observed.
H-LGE-MRI was performed during continuous infusion of Gd-based contrast agent
1
as well as during (delayed) wash-out (Figure 3). Higher contrast between infarct and
remote tissue was observed in the wash-out phase, supporting the notion that LGE
is very sensitive to timing of the experiment [10;27;28]. Infarct size was estimated
with 23Na-CSI and 1H-LGE-MRI using different signal threshold levels and compared
to TTC staining(Figure 4B). Different values of infarct size based on different cutoff
threshold values have been also observed by others [29]. Although 1H-LGE-MRI has
obvious advantages in terms of SNR and spatial resolution, infarct size from 23Na-CSI,
compared well to both LGE and TTC. Na+e and 1H signals were elevated in the infarct,
but, as mentioned above, Na+i in the infarct area was not different from remote tissue,
and could not be used to estimate infarct size. Interestingly, the best correspondence
between TTC infarct size and both LGE and Nae was found for areas defined by approx. 5
SD’s above remote, indicating an area of lower enhanced LGE and Nae signal intensities,
not corresponding to the true infarct. Edema is pathognomonic for IR injury [30] and
we have demonstrated before in a non-ischemic model of edema that both LGE and
122
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
Nae signal intensities are elevated in edema. Therefore, the current results could be
well explained by edema in a peri-infarct zone.
Group 3: Chronic MI
In chronic infarcts, 3 weeks after MI, Na+i was very low and Na+e levels were significantly
increased. These findings confirm the absence of viable cells and increased interstitial
space in fibrotic scar (Figure 5A-B). Interestingly, 2-3 SD’s thresholds seem to have a
better match to the TTC infarct area in this group, which could be explained by lower
or absent edema levels.
Clinical perspective
LGE MRI has become the current gold standard to determine chronic infarct size, but
Na-CSI has not yet been used clinically. To make 23Na-CSI a clinical reality, high field
23
(7 T) MRI scanners are recommended. Futhermore, a shift reagent suitable for use in
human is required, which is not yet available, but, given the similarity between contrast
agents and shift reagents, could be developed.
Alternatively, multiple quantum filtered MR methods [31;32] or ultrashort-echo time
(UTE) sequences [33] could be used, but further investigations are still needed for
clinical implementation of these methods.
Na-CSI of normal Na+i suffers from poor SNR and low spatial resolution, but the
23
absence of Na+i signal in combination with strongly elevated Na+e signal unambiguously
indicates a chronic infarct, unlike LGE, which would also show enhancement with
edema. Highly elevated Na+i levels indicate ischemia, which requires immediate
reperfusion, preferably accompanied by interventions to prevent reverse Na+-Ca2+
exchange .
123
Chapter 6
6.5Conclusion
We have demonstrated the feasibility of cardiac imaging of intra- and extra-cellular
sodium to assess myocardial viability in different rat models of myocardial infarction.
Na+ tracking is a valuable tool to indentify the pathological status of the heart: highly
elevated Na+i is a unique characteristic of acute ischemia, whereas the absence of Nai
signal in combination with elevated Na+e signal is a marker of scar tissue. The rapid
normalization of highly elevated Na+i in infarcted area upon reperfusion provides unique
insights in the dynamics of infarct development. The combination of
Na-CSI and
23
H-LGE-MRI shows added value for accurate evaluation, understanding and distinction
1
of different pathophysiological states after myocardial injury and helps to interpret
results mutually.
Competing interests
None.
Acknowledgements
This research is supported by the Netherlands Heart Foundation (NHS) grant 2006B066
(CvE) and 2010T001 (FA), as well as by the Netherlands Heart Institute (ICIN). The
work of GJS is supported by the Dutch Technology Foundation STW, Applied Science
Division of NWO and the Technology Program of the Ministry of Economic Affairs (grant
number 07952). We would like to thank Maringa Emons for biotechnical assistance.
Reference List
1. Walls MC, Verhaert D, Min JK, Raman SV (2011) Myocardial edema imaging in acute
coronary syndromes. J Magn Reson Imaging 34:1243-1250
2. Beyers RJ, Smith RS, Xu Y, Piras BA, Salerno M, Berr SS, Meyer CH, Kramer CM, French BA,
Epstein FH (2012) T(2) -weighted MRI of post-infarct myocardial edema in mice. Magn
Reson Med 67:201-209
124
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
3. Edwards NC, Routledge H, Steeds RP (2009) T2-weighted magnetic resonance imaging to
assess myocardial oedema in ischaemic heart disease. Heart 95:1357-1361
4. Eitel I, Friedrich MG (2011) T2-weighted cardiovascular magnetic resonance in acute
cardiac disease. J Cardiovasc Magn Reson 13:13
5. Chen HH, Josephson L, Sosnovik DE (2011) Imaging of apoptosis in the heart
with nanoparticle technology. Wiley Interdisciplinary Reviews-Nanomedicine and
Nanobiotechnology 3:86-99
6. Mewton N, Liu CY, Croisille P, Bluemke D, Lima JA (2011) Assessment of myocardial fibrosis
with cardiovascular magnetic resonance. J Am Coll Cardiol 57:891-903
7. Yang YD, Yang YH, Yanasak N, Schumacher A, Hu TCC (2010) Temporal and Noninvasive
Monitoring of Inflammatory-Cell Infiltration to Myocardial Infarction Sites Using
Micrometer-Sized Iron Oxide Particles. Magnetic Resonance in Medicine 63:33-40
8. Rajappan K, Bellenger NG, Anderson L, Pennell DJ (2000) The role of cardiovascular
magnetic resonance in heart failure. European Journal of Heart Failure 2:241-252
9. Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, Van Echteld CJ,
Pasterkamp G, Strijkers GJ (2012) Quantitative T (2) (*) assessment of acute and chronic
myocardial ischemia/reperfusion injury in mice. MAGMA
10. Dall’Armellina E, Karia N, Lindsay AC, Karamitsos TD, Ferreira V, Robson MD, Kellman P,
Francis JM, Forfar C, Prendergast BD, Banning AP, Channon KM, Kharbanda RK, Neubauer
S, Choudhury RP (2011) Dynamic Changes of Edema and Late Gadolinium Enhancement
After Acute Myocardial Infarction and Their Relationship to Functional Recovery and
Salvage Index. Circulation-Cardiovascular Imaging 4:228-U67
11. Neizel M, Katoh M, Schade E, Rassaf T, Krombach GA, Kelm M, Kuhl HP (2009) Rapid
and accurate determination of relative infarct size in humans using contrast-enhanced
magnetic resonance imaging. Clinical Research in Cardiology 98:319-324
12. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
13. Friedrich MG, bdel-Aty H, Taylor A, Schulz-Menger J, Messroghli D, Dietz R (2008)
The salvaged area at risk in reperfused acute myocardial infarction as visualized by
cardiovascular magnetic resonance. Journal of the American College of Cardiology
51:1581-1587
14. Stork A, Muellerleile K, Bansmann PM, Graessner J, Kaul M, Kemper J, Adam G, Lund
GK (2007) Value of T2-weighted, first-pass and delayed enhancement, and cine CMR to
differentiate between acute and chronic myocardial infarction. European Radiology
17:610-617
15. Rosano GMC, Fini M, Caminiti G, Barbaro G (2008) Cardiac Metabolism in Myocardial
Ischemia. Current Pharmaceutical Design 14:2551-2562
125
Chapter 6
16. Jansen MA, Van Emous JG, Nederhoff MG, van Echteld CJA (2004) Assessment of myocardial
viability by intracellular 23Na magnetic resonance imaging. Circulation 110:3457-3464
17. Weidensteiner C, Horn M, Fekete E, Neubauer S, von Kienlin M (2002) Imaging of
intracellular sodium with shift reagent aided Na-23 CSI in isolated rat hearts. Magnetic
Resonance in Medicine 48:89-96
18. Pabst T, Sandstede J, Beer M, Kenn W, Neubauer S, Hahn D (2003) Evaluation of sodium T1
relaxation times in human heart. J Magn Reson Imaging 17:726-729
19. Ouwerkerk R, Bleich KB, Gillen JS, Pomper MG, Bottomley PA (2003) Tissue sodium
concentration in human brain tumors as measured with 23Na MR imaging. Radiology
227:529-537
20. Ten Hove M, Van Emous JG, van Echteld CJA (2003) Na+ overload during ischemia and
reperfusion in rat hearts: comparison of the Na+/H+ exchange blockers EIPA, cariporide
and eniporide. Mol Cell Biochem 250:47-54
21. Aguor, E. N., van de Kolk, C. W., Nederhoff, M. G., Doevendans, P. A., Pasterkamp, G., Strijkers,
G. J., Arslan, F., and van Echteld, C. J. A. (2012) 23Na Chemical Shift Imaging of myocardial
edema. ISMRM 19th Annual Meeting. International Society for Magnetic Resonance in
Medicine (ISMRM). 5-11 Ref Type: Conference Proceeding
22. Avkiran M, Curtis MJ (1991) Independent Dual Perfusion of Left and Right CoronaryArteries in Isolated Rat Hearts. American Journal of Physiology 261:H2082-H2090
23. Buster DC, Castro MMCA, Geraldes CFGC, Malloy CR, Sherry AD, Siemers TC (1990)
Tm(Dotp)5--A 23Na+ Shift Agent for Perfused Rat Hearts. Magnetic Resonance in Medicine
15:25-32
24. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J, Finn JP, Klocke FJ, Judd
RM (1999) Relationship of MRI delayed contrast enhancement to irreversible injury, infarct
age, and contractile function. Circulation 100:1992-2002
25. ten Hove M, van Echteld CJA (2004) Limited effects of post-ischemic NHE blockade on
[Na+]i and pHi in rat hearts explain its lack of cardioprotection. Cardiovasc Res 61:522-529
26. ten Hove M, Nederhoff MG, van Echteld CJA (2005) Relative contributions of Na+/H+
exchange and Na+/. Am J Physiol Heart Circ Physiol 288:H287-H292
27. Saborowski O, Saeed M (2007) An overview on the advances in cardiovascular interventional
MR imaging. Magnetic Resonance Materials in Physics Biology and Medicine 20:117-127
28. Schulz-Menger J, Gross M, Messroghli D, Uhlich F, Dietz R, Friedrich MG (2003)
Cardiovascular magnetic resonance of acute myocardial infarction at a very early stage.
Journal of the American College of Cardiology 42:513-518
29. Varga-Szemes A, Ruzsics B, Kirschner R, Singh SP, Kiss P, Brott BC, Simor T, Elgavish A,
Elgavish GA (2012) Determination of Infarct Size in Ex Vivo Swine Hearts by Multidetector
Computed Tomography Using Gadolinium as Contrast Medium. Investigative Radiology
47:277-283
126
Differentiation of myocardial injury in various infarct models
with 23Na Chemical Shift Imaging and Late Gd Enhancement MRI
30. Arai AE (2011) Magnetic Resonance Imaging for Area at Risk, Myocardial Infarction, and
Myocardial Salvage. Journal of Cardiovascular Pharmacology and Therapeutics 16:313320
31. Dizon JM, Tauskela JS, Wise D, Burkhoff D, Cannon PJ, Katz J (1996) Evaluation of triplequantum-filtered 23Na NMR in monitoring of Intracellular Na content in the perfused rat
heart: comparison of intra- and extracellular transverse relaxation and spectral amplitudes.
Magn Reson Med 35:336-345
32. Navon G, Werrmann JG, Maron R, Cohen SM (1994) 31P NMR and triple quantum filtered
23Na NMR studies of the effects of inhibition of Na+/H+ exchange on intracellular sodium
and pH in working and ischemic hearts. Magn Reson Med 32:556-564
33. de Jong S, Zwanenburg JJ, Visser F, van der Nagel R, van Rijen HV, Vos MA, de Bakker JM,
Luijten PR (2011) Direct detection of myocardial fibrosis by MRI. Journal of Molecular and
Cellular Cardiology 51:974-979
127
Part III
129
Chapter 7
Summarizing discussion
131
Chapter 7
Ischemic heart disease is a leading cause of morbidity and mortality globally. Many
studies have demonstrated the potential of proton based cardiac MR in fundamental
and clinical research. Still, very limited data are available on sodium cardiac MRI,
although the technique potentially provides very promising imaging readouts of
myocardial tissue status. In this thesis, we have assessed myocardial viability using
novel cardiac MRI techniques based on proton (1H) MRI in the first part of this thesis
and on sodium (23Na) combined with 1H MRI in the second part. Different models were
used to induce changes in myocardial tissue status, e.g. ischemia,in mouse heart in
vivo and in rat heart under perfused conditions. In this final chapter, the results will
be summarized and further possible improvements in experimental research and
possibilities for translating some of the imaging techniques into the clinic will be
discussed.
Part I
Late gadolinium enhancement MRI cannot discriminate between necrotic, fibrotic and
edematous tissue[1;2]. Therefore, acute myocardial infarction assessment with LGE
MRI is complicated. The aim of the first part of this thesis was to explore the potential
value of quantitative T2* mapping in combination with commonly used cine and LGE
MRI methods in mouse myocardium with different models of myocardial infarction.
T2* MRI traditionally is used to assess iron deposition in the heart [3]. In chapter 2,
quantitative T2* mapping was compared to cine MRI and LGE MRI in a mouse model
of ischemia reperfusion (I/R) injury, starting from 1 day to 28 days post MI. The reason
for selecting this I/R infarct model was not only to induce a relatively small infarct, but
also to induce myocardial edema, known to accompany I/R injury [4;5]. This model
was meant to mimic a similar myocardial status as found in patients after coronary
revascularization, in which edema can complicate the diagnosis. We found that
quantitative T2* values in the infarcted tissue decreased progressively during infarct
development [6]. Indeed, for acute infarcts at day 1 we found a large LGE-based infarct
size accompanied by increased wall-thickness, suggesting the presence of edema.
Surprisingly, we did not observe a transient increase in T2* at day 1 accompanying
this edema, as previously observed with T2‑weighted imaging [7]. Apparently, the
132
Summarizing discussion
effect of edema on T2* is less than the decrease of T2* induced by iron or collagen.
This observation is critical in the clinical setting where early assessment of viability
is significantly hampered by the presence of infarct-induced edema. T2*-imaging
may be a tool to circumvent this inevitable problem after acute MI. The changes of
T2* were associated with infarct age, reversibility of the injury, and gave an indication
of the structural variation in the myocardium throughout infarct maturation. The
local decrease of T2* in the infarct was accompanied by globally and locally reduced
ventricular function. In addition, the reduction in T2* with infarct age was associated
also with the presence of iron and the formation of significant amounts of collagen.
Following the assessment of small, myocardial I/R injury in mice with T2* mapping
(chapter 2), we aimed in chapter 3, to further explore quantitative changes in T2* after
more severe MI in a mouse heart failure model, induced by permanent ligation of
the left coronary artery. In comparison to the results of the mild I/R injury (chapter
2), T2* in infarcted areas exhibited a more dramatic decrease starting from day 1, but
did not decrease further on day 7 and 28. In contrast to the I/R injury with relatively
small infarcts, T2* in remote tissue also decreased significantly and progressively from
baseline accompanied by adverse ventricular remodeling of non-infarcted areas
after MI, in agreement with findings of Bogaert et al. [8], who observed dysfunction
in remote myocardium in patients with transmural anterior MI. Serial LGE scans
only revealed changes in the infarct area and not in the remote myocardium. Future
research will include the quantification of fibrosis in the remote area, to see if T2* MRI
can identify interstitial fibrosis in non-ischemic remote tissue. We concluded that,
quantitative T2* assessment provides a valuable readout, not only for infarct, but for
remote areas as well.
Also the more classical Cine MRI for assessing cardiac function can provide valuable
information on left ventricular remodeling after myocardial infarction. In Chapter 4,
this approach was used to assess cardiac function after myocardial infarction in TLR2/-
mice. Many studies have suggested that Toll-like receptors (TLRs) play a crucial role
in inflammatory responses after tissue injury [9]. Cardiac MRI analysis showed that the
lack of TLR2 in a mouse model of MI indeed improved survival and cardiac function.
133
Chapter 7
Hence, TLR2 is a potential therapeutic target to prevent adverse remodeling after
myocardial infarction.
Part II
The second part of this thesis was dedicated to sodium MRI in comparison to proton
MRI methods. Accumulation of edema is not only a result of myocardial I/R injury, but
may arise from many causes and with different extent, such as arterial hypertension
[10] and myocarditis [11], which complicates differential diagnosis with LGE MRI. To
address this problem, a non-infarct model of different levels of edema in isolated
rat hearts based on different perfusion pressures was developed in chapter 5.
Subsequently, the influence of myocardial edema was investigated with 23Na CSI in
combination with a shift reagent and 1H MRI in combination with a contrast agent. With
this model, we have demonstrated that contrast enhanced (CE) MRI with a Gd contrast
agent can identify edematous myocardial tissue, most likely as a result of an increased
distribution volume of Gd. Surprisingly, we found more edema in the subendocardial
area than in the subepicardial area. We have also demonstrated in this same edema
model that the Na+e signal increases with increasing edema, confirming that Na+e is a
useful endogenous marker of extracellular volume. Interestingly, a similar transmural
distribution of signal intensities as observed in the contrast-enhanced images was
observed in the Na+e images. Na+i essentially remained unchanged in edema. We
concluded that edema shows up both in CE MRI and Na+e, confirming that CE MRI may
overestimate infarct size. In addition, we could conclude that high perfusion pressure
causes more edema subendocardially than subepicardially.
After assessing myocardial edema in edematous but viable myocardium, in chapter
6, we used 23Na CSI to study cellular integrity and ion homeostasis during and after
myocardial ischemia in rat hearts to differentiate injured myocardium under three
conditions: in acute ischemia and reperfusion (I/R), 2 hours after I/R injury, and in
chronic infarction (MI). In this study, we demonstrated that 23Na CSI is a powerful tool to
identify sarcolemmal dysfunction related to the myocardial pathophysiological status in
the time-course after MI. Acutely, the ischemic myocardium was characterized by highly
elevated Na+i levels, caused by Na+/K+ ATPase dysfunction and continued Na+ influx, and
134
Summarizing discussion
concomitant decrease of Na+e. In contrast, chronic MI was characterized by an almost
complete absence of Na+i, indicating the loss of viable myocardium and replacement
by scar tissue, corroborated by the observed increased Na+e. However, the situation in
2 hours reperfused myocardium was less clear. The (partial) recovery of Na+i may signify
recovery of Na+/K+ ATPase function, but could be equally well explained by heterogeneous
reperfusion with areas of no reflow and persistent high Na+i in combination with areas
showing advanced cell deterioration and sarcolemmal dysfunction/rupture with loss
of cellular content. The slightly elevated Na+e could signify edema, but could also be
explained by loss of sarcolemmal integrity. It is known that infarct size estimation with
LGE is influenced by threshold selection [12]. Similarly, the determination of infarct size
was very much dependent for both CE MRI and Nae MRI on threshold selection, which
is not surprising since both techniques monitor the same compartment. Infarct size
determination based on Nai images is also similarly dependent on threshold selection.
Furthermore, we have demonstrated that wash-out of contrast agent from edematous
tissue occurs much faster than from infarcted tissue, providing further explanation for
the threshold dependence of infarct size determination.
Further development and clinical translation
Part I
MRI is steadily becoming the preferred imaging tool for characterizing the heart in
health and disease. Several new cardiac imaging methods, focusing on quantitative
readouts of heart function and physiology as well as tissue MR relaxation time
parameters, are under development
As shown in this thesis T2* may have added value in the assessment of infarct
and remote tissue. A serious limitation to the use of T2* at 9.4 Tesla as a diagnostic
marker for myocardial injury is contamination of intrinsic T2* values by the presence
of macroscopic susceptibility differences, mainly induced by the lung-air interfaces.
Using a lower magnetic field scanner should reduce such an artifact. MRI scanners
with lower magnetic field are commonly available in the clinic, where T2* mapping is
135
Chapter 7
commonly used for thalassemia patients. Moreover, using shorter echo times and using
different excitation pulses might solve or at least reduce this problem. In addition,
application of local shimming techniques might alleviate some of these susceptibility
contaminations.
We showed here that T2* mapping is not sensitive to edema in acute MI, unlike LGE
MRI, which makes T2* mapping a valuable tool to assess myocardium,in acute MI.
Furthermore, in contrast to LGE MRI, T2* mapping does not require any contrast agent
injection, which may be risky for some patients [13]. Recently, it was demonstrated
that T2* has the potential to image fibrosis in the infarct zone using an ultra-short echo
time (UTE) technique [14]. In vivo application of such a technique should be possible
in the clinic using radial and spiral k-space acquisitions. A particular application of
T2* mapping is the tracking of superparamagnetic iron oxide labeled stem cells after
delivery in cardiovascular disease [15]. Methods utilizing magnetization transfer have
also been explored for assessing edema in infarcted areas [16] This method may give
valuable information on specific pathological conditions of the myocardium, however,
optimizing for small animals with a fast beating heart is not trivial.
Part II
The studies conducted in this part of this thesis were performed to get more insights
in the potential of 23Na CSI in evaluating myocardial injury. Although the techniques
have provided valuable information on myocardial tissue status, their transition to
clinic is a challenge. For instance, 23Na CSI of normal Na+i suffers from poor SNR and low
spatial resolution. To make 23Na CSI a clinical reality, high field (7 T) MRI scanners are
recommended [14]. Furthermore, a shift reagent suitable for use in human is required,
which is not yet available, but, given the similarity between contrast agents and shift
reagents, could be developed. Alternatively, multiple quantum filtered MR methods
[17;18] or ultrashort-echo time (UTE) sequences [14] could be used, but further
investigations are still needed for clinical implementation of these methods.
136
Summarizing discussion
Conclusion
Quantitative relaxation time measurement by MR is an emerging tool in the
assessment of myocardial infarction, offering a quantitative standard MRI scale related
to specific pathological conditions. Our approach using T2* mapping promises to be a
powerful MRI method to monitor remodeling post-infarction and provides useful and
complementary information to LGE and Cine imaging.
In addition, we have demonstrated the feasibility of cardiac imaging of intra- and
extra-cellular sodium to assess myocardial viability. 23Na-CSI is an attractive alternative
for imaging of edema. In addition, Na+ tracking is a valuable tool to indentify the
pathological status of the heart: highly elevated Na+i is a unique characteristic of acute
ischemia, whereas the absence of Nai signal in combination with elevated Na+e signal is
a marker of scar tissue. The rapid normalization of highly elevated Na+i in infarcted area
upon reperfusion provides unique insights in the dynamics of infarct development. The
combination of 23Na-CSI and 1H-LGE-MRI shows added value for accurate evaluation,
understanding and distinction of different pathophysiological states after myocardial
injury and helps to interpret results mutually.
Further development of both sodium and proton MRI will allow better understanding
of myocardial viability.
Reference List
1. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
2. Abdel-Aty H, Zagrosek A, Schulz-Menger J, Taylor AJ, Messroghli D, Kumar A, Gross M, Dietz
R, Friedrich MG (2004) Delayed enhancement and T2-weighted cardiovascular magnetic
resonance imaging differentiate acute from chronic myocardial infarction. Circulation
109:2411-2416
137
Chapter 7
3. Wood JC, Ghugre N (2008) Magnetic resonance imaging assessment of excess iron in
thalassemia, sickle cell disease and other iron overload diseases. Hemoglobin 32:85-96
4. Lockie T, Nagel E, Redwood S, Plein S (2009) Use of cardiovascular magnetic resonance
imaging in acute coronary syndromes. Circulation 119:1671-1681
5. Oyama J, Blais C, Jr., Liu X, Pu M, Kobzik L, Kelly RA, Bourcier T (2004) Reduced myocardial
ischemia-reperfusion injury in toll-like receptor 4-deficient mice. Circulation 109:784-789
6. Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, Van Echteld CJ,
Pasterkamp G, Strijkers GJ (2012) Quantitative T (2) (*) assessment of acute and chronic
myocardial ischemia/reperfusion injury in mice. MAGMA
7. Beyers RJ, Smith RS, Xu Y, Piras BA, Salerno M, Berr SS, Meyer CH, Kramer CM, French BA,
Epstein FH (2012) T(2) -weighted MRI of post-infarct myocardial edema in mice. Magn
Reson Med 67:201-209
8. Bogaert J, Bosmans H, Maes A, Suetens P, Marchal G, Rademakers FE (2000) Remote
myocardial dysfunction after acute anterior myocardial infarction: impact of left
ventricular shape on regional function: a magnetic resonance myocardial tagging study. J
Am Coll Cardiol 35:1525-1534
9. Arslan F, Smeets MB, O’Neill LA, Keogh B, McGuirk P, Timmers L, Tersteeg C, Hoefer IE,
Doevendans PA, Pasterkamp G, de Kleijn DP (2010) Myocardial ischemia/reperfusion injury
is mediated by leukocytic toll-like receptor-2 and reduced by systemic administration of a
novel anti-toll-like receptor-2 antibody. Circulation 121:80-90
10. Mehlhorn U, Davis KL, Burke EJ, Adams D, Laine GA, Allen SJ (1995) Impact of
cardiopulmonary bypass and cardioplegic arrest on myocardial lymphatic function. Am J
Physiol 268:H178-H183
11. Korkusuz H, Esters P, Naguib N, Eldin NEN, Lindemayr S, Huebner F, Koujan A, Bug
R, Ackermann H, Vogl TJ (2009) Acute myocarditis in a rat model: late gadolinium
enhancement with histopathological correlation. European Radiology 19:2672-2678
12. Jahrsdoerfer RA, Gardner G, Yeakley JW, Karam CR (1986) Lethal cholesteatoma.
Otolaryngol Head Neck Surg 95:507-510
13. Stacul F, van der Molen AJ, Reimer P, Webb JA, Thomsen HS, Morcos SK, Almen T, Aspelin P,
Bellin MF, Clement O, Heinz-Peer G (2011) Contrast induced nephropathy: updated ESUR
Contrast Media Safety Committee guidelines. Eur Radiol 21:2527-2541
14. de JS, Zwanenburg JJ, Visser F, der NR, van Rijen HV, Vos MA, de Bakker JM, Luijten PR
(2011) Direct detection of myocardial fibrosis by MRI. J Mol Cell Cardiol 51:974-979
15. Kedziorek DA, Kraitchman DL (2010) Superparamagnetic iron oxide labeling of stem cells
for MRI tracking and delivery in cardiovascular disease. Methods Mol Biol 660:171-183
16. Henkelman RM, Stanisz GJ, Graham SJ (2001) Magnetization transfer in MRI: a review. NMR
Biomed 14:57-64
138
Summarizing discussion
17. Navon G, Werrmann JG, Maron R, Cohen SM (1994) 31P NMR and triple quantum filtered
23Na NMR studies of the effects of inhibition of Na+/H+ exchange on intracellular sodium
and pH in working and ischemic hearts. Magn Reson Med 32:556-564
18. Dizon JM, Tauskela JS, Wise D, Burkhoff D, Cannon PJ, Katz J (1996) Evaluation of triplequantum-filtered 23Na NMR in monitoring of Intracellular Na content in the perfused rat
heart: comparison of intra- and extracellular transverse relaxation and spectral amplitudes.
Magn Reson Med 35:336-345
139
Korte samenvatting in het Nederlands
Wanneer een deel van het hart niet goed samentrekt is het cruciaal te weten of een
bypass operatie of een dotterbehandeling dit zou kunnen verhelpen. Is dit deel van
het hart nog levensvatbaar of is het al littekenweefsel? Veelal wordt hiervoor MRI
met contrastmiddel (DCE-MRI) gebruikt. Een nadeel hiervan is dat de infarctgrootte
vaak overschat wordt, waarschijnlijk vanwege vocht rondom het infarct. M.b.v.
proefdieren zijn alternatieve MRI methoden onderzocht. Gebleken is dat een van de
beeldparameters van het op water gebaseerde MRI beeld, de T2*, bij een klein infarct in
de loop der tijd daalt als gevolg van afzetting van collageen en ijzer, terwijl de T2* in het
gezonde hartspierweefsel normaliseert. Bij een groot infarct daalt de T2* in het infarct
sneller, terwijl nu in het ‘gezonde’ weefsel T2* ook daalt door remodelleren (hartfalen).
Hartfunctie MRI onderzoek liet zien dat de ‘Toll-like receptor2’ in het bloed een
belangrijke rol speelt bij dit remodelleren. M.b.v. een ‘shift reagens’ is het ook mogelijk
het natrium in (Na+i) en buiten (Na+e) de hartspiercellen af te beelden. Door (bloed)
drukverhoging werd vochtophoping in de hartspier veroorzaakt, waarmee bevestigd
werd dat deze vochtophoping inderdaad door DCE-MRI wordt afgebeeld en ook door
Na+e-MRI, maar niet door Na+i-MRI. Aan de binnenkant van de hartspier was er meer
vochtophoping dan aan de buitenkant. Tenslotte bleek dat Na+i sterk verhoogd is bij
acute infarcten, terwijl in het chronische infarct Na+e verhoogd is en Na+i nagenoeg
verdwenen. Voor toepassing van deze nieuwe MRI-methoden bij patiënten is nader
onderzoek vereist.
141
List of publications
Manuscripts
Na chemical shift imaging and Gd enhancement of myocardial edema Aguor EN,
23
van de Kolk CW, Arslan F, Nederhoff MG, Doevendans PA, Pasterkamp G, Strijkers GJ,
van Echteld CJ
Int J Cardiovasc Imaging. 2012 Jul 12. [Epub ahead of print]
Quantitative T2 * assessment of acute and chronic myocardial ischemia/reperfusion
injury in mice Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA, van
Echteld CJ, Pasterkamp G, Strijkers GJ
MAGMA. 2012 Feb 11. [Epub ahead of print]
MicroRNA-155 inhibits cell migration in human cardiomyocyte progenitor cells
(CMPCs) via targeting MMP-16 Liu J, van MA, Aguor EN, Siddiqi S, Vrijsen K, Jaksani S,
Metz C, Zhao J, Strijkers GJ, Doevendans PA, Sluijter JP
J Cell Mol Med. 2012 Feb 20
Faraday Effect Position Sensor for Interventional Magnetic Resonance Imaging Bock
M, Umathum R, Sikora J, Brenner S, Aguor EN, Semmler WA
Phys Med Biol. 2006 Feb 51[4], 999-1009.
Differentiation of myocardial injury in various infarct models with 23Na Chemical Shift
Imaging and Late Gd Enhancement MRI Aguor EN, van de Kolk CW, Arslan F, Nederhoff
MG, Doevendans PA, Pasterkamp G, Strijkers GJ, van Echteld CJ, submitted
MSC-derived exosomes enhance myocardial viability after ischemia/reperfusion injury
Arslan F, Ruenn CL, Choo A, Aguor EN, Smeets M b, Akeroyd L, van Rijen HV, Timmers
L, Doevendans PA, Pasterkamp G, Lim KS , de Kleijn DP, submitted
143
Molecular Imaging in Stem Cell therapy Siddiqi S, Aguor EN, Sluijter JPG, Chamulaeu
SA, Doevendans PA, submitted
Quantitative T2* assessment of acute and chronic myocardial permanent ligation injury
in mice Aguor EN, van de Kolk CWA, Nederhoff MGJ, Doevendans PA, van Echteld CJA,
Pasterkamp G, Strijkers GJ, Arslan F, Submitted
Lack of circulating Toll-like receptor 2 promotes survival and prevents adverse
remodeling after myocardial infarction Aguor EN, Arslan f, Smeets MB, Doevendans
PA, Strijkers GJ, Pasterkamp G, de Kleijn DP, In preparation
Abstracts and Conference Proceedings 1st author only
Na Chemical Shift Imaging and Late Gadolinium enhanced (LGE) MRI of acute ischemia
23
reperfusion myocardial injury Aguor EN, Arslan F, Strijkers GJ, Nederhoff MGJ, van de
Kolk CW, Pasterkamp G, Doevendans PA, van Echteld CJA 20th Intl. Soc. Mag. Reson.
Med. (ISMRM), Annual meeting and exhibition Intl. 2012 May 8, Melbourne, Australia.
“Presentation”
Na Chemical Shift Imaging and Late Gadolinium enhanced (LGE) MRI of acute
23
ischemia reperfusion myocardial injury Aguor EN, Arslan F, Strijkers GJ, Nederhoff
Marcel GJ, van de Kolk CW, Pasterkamp G, Doevendans PA, Van Echteld CJA 4th annual
meeting of the ISMRM Benelux, 2012 Jan 16, the Leuven Institute for Ireland in Europe,
Leuven, Belgium. “Presentation”
Na Chemical Shift Imaging of Myocardial Edema Aguor EN
23
ImagO Promovendimiddag, 2011 Dec 9, presentation, University medical centre,
Utrecht, the Netherlands. “Presentation”
Assessment of myocardial viability with Magnetic Resonance Imaging techniques
Aguor EN
2nd ICIN- Netherlands Heart Institute Masterclass, 2011 Sep 6, Utrecht, the Netherlands
“Presentation”
144
Longitudinal assessment of T2* changes in mouse myocardium following ischemiareperfusion injury Aguor EN
3rd annual meeting of the ISMRM Benelux, 2011 Jan 19, Roosendaal, the Netherlands.
“Presentation”
Na Chemical Shift Imaging of Myocardial Edema Aguor EN, van de Kolk CW,
23
Nederhoff MGJ, Doevendans PA, Pasterkamp G, Strijkers GJ, Arslan F, van Echteld CJA
3rd annual meeting of the ISMRM Benelux, 2011 Jan 19, Roosendaal, the Netherlands.
“Presentation”
Na Chemical Shift Imaging of Myocardial Edema Aguor EN
23
Masterclass, 2010 Dec 15, Biomedical NMR (TU/e), Eindhoven Technical University,
Eindhoven, the Netherlands. “Presentation”
Lack of Toll-like receptor 2 promotes survival and preserves cardiac function and
geometry after acute MI Aguor EN, Arslan F, van de Kolk CW, Doevendans PA, de Kleijn,
G. Pasterkamp. de Kleijn DP
Eur Heart J 2010;31 (Suppl 1): 9(abstract 186), Stockholm, Sweden. “Presentation”
Assessment of myocardial viability with Magnetic Resonance Imaging techniques
Aguor EN
Bruker MRI /MRS User’s Meeting 2008 Oct 13, Germany. “Presentation”
A Reflection Mode Fiber-optic Faraday Sensor for Device Localization in Interventional
MRI Aguor EN, Brenner S, Umathum R, Semmler W, Bock M.
Proc. 14th Scientific meeting of ISMRM 2006, Seattle, USA
A Reflection Mode Fiber-optic Faraday Sensor for Device Localization in Interventional
MRI Aguor EN, Brenner S, Umathum R, Semmler W, Bock M .
Proc. ISMRM. 14 (2006), Washington, USA. “Presentation”
145
Na Chemical Shift Imaging of Myocardial Edema Aguor EN, van de Kolk CW, Arslan F,
23
Nederhoff MG, Doevendans PA, Pasterkamp G, Strijkers GJ, van Echteld CJ
Proc. ISMRM, 19(2011), (Abstract 1353), Montréal, Canada. “Poster”
Longitudinal assessment of T2* changes in mouse myocardium following ischemiareperfusion injury Aguor EN, Arslan F, van de Kolk CW, Nederhoff MG, Doevendans PA,
van Echteld CJ, Pasterkamp G, Strijkers GJ
Proc. ISMRM,19(2011), (Abstract 1363), Montréal, Canada. “Poster”
Assessment of myocardial viability with Magnetic Resonance Imaging techniques
Aguor EN Heart foundation the Netherlands (NHS), 2008 Sep, Amsterdam, the
Netherlands, “Poster”
146
Acknowledgment
After all this years of ups, downs, doubts, deadlines, mice and MR scans it’s time to look
back and think of everyone in this long, exciting, challenging journey. Still I am glad I
have come to the Netherlands and decided to start a PhD study in Utrecht.
First I want to thank Prof. dr. C.J.A. van Echteld. Dear Cees, thank you for accepting me
as a member in your group and thanks for giving me the freedom to investigate on
this “Sodium MRI”. Just a few months later, you went to Switzerland. Being busy on a
new carrier and in a different place, you were still able to support me in the project.
Committed as you were, you always found a moment for consultation via Skype or
phone, you even came to Utrecht on a regular base. I am grateful for your scientific
criticisms and honest opinion about the results. I have always enjoyed discussions with
you about science and numerous other topics. Without your help in all aspects of this
project, it would never have resulted in the nice thesis as it lies here in front of you.
The next to thank is Prof. dr. P.A.F.M. Doevendans. Dear Pieter, after Prof. van Echteld
left, you took over the supervision of our MRI group and became my promoter. It has
been good to know that you were always there to support me after the big changes
in our lab and all the obstacles I met during this journey. Thanks for your trust and for
giving me the opportunity to continue this research. It was always nice meeting you to
discuss the progress of my project. You always kept asking me about the articles I was
writing. It was very important to have you on my side all the time. You won’t have to
worry any more about this project
Prof. dr. G. Pasterkamp, dear Gerard, you became the supervisor of our lab and I am
very thankful for your fast and efficient action whenever things went wrong. Thanks
for your assistance on my research. Your group “the Experimental Cardiology” provided
me a stimulating scientific and social environment. I learned a lot from the variety of
fellow researchers within your group. It was always nice to discuss my research and
what was missing from a clinical point of view. Dr.ir. G.J. Strijkers, dear Gustav, you were
the savior, you showed up in a critical moment. I remember the first time I met you
giving a lecture in Wageningen University about MRI. I was truly amazed about your
147
lecture and your Cardiac-MRI-dedicated group. Your contribution certainly speeded
up things. Thank you for the critical remarks, suggestions and corrections. Not to forget
you presenting my abstracts at ISMRM. Also thanks for preparing the nice fresh coffee
in the early morning at our lab. Dr. Fatih Arslan, already involved in so many projects,
you still found the time to step into a joint study with me. Thanks for the excellent
surgery on my mice and rats, for revising the articles and for all the good advise during
the years.
Dear Marcel Nederhoff, thanks for your support and tricks with the Langendorff, but
also for your tricks for inserting the catheter and these tiny tubes. I enjoyed very much
our conversation, it was always interesting and special to know your opinion about
topics such as God, martial arts, dancing, music, culture and many other. I liked the way
you think about things. Thanks for your advice, discussion and humors both in science
but also in personal level. It was great to have you in the lab for all these years. Thanks
for sharing my thoughts and feeling when many hearts did not work, you were still
positive and saying “of course that is normal in science”!! Who knows maybe one day
you will have your own company brewing your own beer. I am glad I do not have to see
strange photos in my drawer anymore. Drs. Ing. Kees van de Kolk, dear Kees, you were
the MRI-operator as well as the computer system manager, therefore always there.
I still remember calling you at night and having you remotely-connected to the MR
scanner. Thanks for your support also for the buffer recycling and analysis, especially
at the end when I was really running out of time. Thanks for your commitment. You
were the man I could talk to about MRI. Thanks for the good time. I really wish you lots
of success in the coming years. Sandra and Bernard, we were the last PhD students in
the MRI-lab, it was nice to have you in the lab and I wish you good luck too. Krista den
Ouden, thank you very much for all the imaging work. Nice to find someone already in
the lab made coffee when I arrived in the morning. It was great to have you in the lab.
Sanne Willems, it was great to have you as my neighbor in the Lange Nieuwstraat. We
always had so many topics to talk about and you were the first person I could think
to choose to be my Paranymph. Thanks so much for your friendship and for all these
educational Tennis matches we had. I wish you lots of success in your PhD too.
148
Thanks for other members in the lab. Arjan, Sander, Loes, Noortje, Chaylendra, Joost S.,
Evelyn, Cees, Joyce, Marlijn, Ben, Joost F. Bas, Joyce, Vince, Paul, Vincent, Pleunie, Dave,
Verena, Esther, Roberto, Sailay, Ellen, Geert, Marten, Krijn, Alain, Dries, Zhiyong, Hamid,
Claudia, Marish, Diana, Frebus, Hendrik, Imo, Karlijn, Mirjam, Jiongwei, Frederieke,
and all the interns, thanks for making my project here at the Experimental Cardiology
a very pleasant times. Dear Maringa, special thanks for your help in completing the
imaging and analysis of the last groups. I wish you all success. Special thanks of course,
to Marjolijn and Ineke, Margo, Tamara, Jantine and Sylvia for your pro-activity, you’re
always being there when necessary.
It was also nice to be working next to the in vivo lab, Lissette, Geralda, Annette, Gerhard,
Erwin, Ivo, Maurits, Kaio, Rick, Mark, Umesh, Pavel. For you who did not graduate yet,
go for it. And thanks for the “borrels” that I sadly enough were not able to attend all of
them.
From the GDL, thanks for, Helma, Anja, Willeke, Romy, Sabine, Gerard, Kiki and all the
others.
Many thanks to ICIN, starting from Jan Weijers, Eelco Soeteman, Rachida Tallahi and
Marijan de jonge. Thank for all your help for the management of my project. You were
always there when it was necessary and thanks for all the nice social activity.
Special thanks to my very dear friends, Raffa, Tina, Gianpiero. It all began in the kitchen
next to my room in biltstraat’s house. Raffa: great to see progressing in your scientific
carrier ‘even more to the north’ in the UK as a married but still stylish lady. Great to see
you with Hazem. Good luck Hazem with your new life in Italy, it was great to met you
in Siegen! Tina: I would say you mad it very clear that you love Germany very much
being there for the second time. So you guys stop complaining. Good luck also to
Gianpie: it was always nice talking to you and the nice food you always cooked. I can
say you’re doing fine now since Raffa is next to you to you now. You were the most
popular guy in Biltstraat house. Gene was also amazing to get to know you, a happy,
polite, generous person. Well finally I am the one who stayed here and your guys were
149
gone to different places now, but remarkably we are still friends. Thanks also to Edu
and his food specialties and the good time. I hope now you are a bit faster in cooking
in Switzerland. Thanks for the friendship and good time for the people I met in Lange
Nieuwstraat house, Tobias, Emma, Setfiee, Natasha, Pradeep, Wouter, Metta, Astrid,
Gudrun, Susana, Ariana, Jeroen and Eneda. Michaela and Augustine, thanks for the
good time and the nice trips to the black forest, it is nice to have you here too, I wish
you all much success.
Thanks for my Salsa gang: Axel, Alex Juko, Natasja, Trevor, Sandra, Marleen, Arwen,
Ellen, Valendard, Emrah, Glenda, Marjolein, Amy, Karin, Michille and Loes. Gerhard,
thanks for your friendship and I am very glad I met you, we have the similar opinion
most of the time, thanks for your friendship.
Marijke Coster and Els thanks for your friendship and for the support you offer. I always
think of you as a family here. Neeltje thanks for everything and thanks for your support
and sharing many things together.
Finally, thanks for the support of my family from far away. It was always difficult to
meet, but you were always in my soul and heart. I dedicate this thesis to my mother
and my late father ‘may his soul rest in peace’.
150
Curriculum Vitae
Eissa N. E. Agour was born on January 16, 1979 in Riyadh, Saudi Arabia. After graduating
from higher secondary school in, Riyadh, Saudi Arabia in 1996, he obtained his Higher
Diploma’s degree in Electronic and Electrical Engineering with specialization in
computer science at the Islamic Institute of Technology, Dhaka, Bangladesh in 2000
(currently Islamic University of Technology). Hereafter he worked as a service engineer/
imaging application specialist for Hitachi medical imaging systems in Saudi Arabia.
From 2004-2006 he worked as a research assistant at the interventional MRI division
of the German Cancer Research Center (DKFZ), Heidelberg, Germany. He obtained his
Master’s degree in Imaging Physics at the University of Siegen, Germany in 2006. In
2007 he started as a PhD candidate at the Department of Cardiology of the University
Medical Center Utrecht (UMCU), the Netherlands under the supervision of prof. dr.
C.J.A. van Echteld and prof. dr. P.A.F.M. Doevendans. The results of his PhD research are
described in this thesis. Recently he joined Medis medical imaging systems bv, Leiden,
the Netherlands as an Imaging product specialist.
151