Download Contrast enhanced and functional magnetic resonance imaging for

Imaging/Diagnostic Testing
Contrast enhanced and functional magnetic
resonance imaging for the detection of viable
myocardium after infarction
Paul Dendale, MD,a Philippe R. Franken, MD,b Pierre Block, MD,a Yiannis Pratikakis,c and Albert De Roos, MDd
Brussels, Belgium, and Leiden, The Netherlands
Purpose Viable myocardium after acute myocardial infarction may be characterized by magnetic resonance imaging
(MRI) either by demonstration of recovery of wall motion under dobutamine stress or by perfusion patterns after contrast
medium administration. This study examines the relation between the two techniques.
Materials and Methods Gradient-echo MRI at rest and under low-dose dobutamine stress was performed in 28
patients within the first 2 weeks after acute myocardial infarction. In addition, spin-echo MRI was performed after gadolinium-DOTA administration. Wall motion at rest and under stress was scored to assess the contractile reserve of the infarct
regions. Infarct enhancement patterns were classified as subendocardial, transmural, or as a doughnut pattern.
Result Subendocardial or absent infarct enhancement was related to functional recovery under stress in 31 of 37 infarct
segments. Transmural infarct enhancement was correlated with the absence of functional recovery in 10 of 17 infarct segments
(p < 0.002), indicating nonviability. The doughnut pattern was exclusively associated with the absence of viability (five of five).
Conclusion Contrast enhancement patterns are related to residual myocardial viability. (Am Heart J 1998:135:875-80.)
In the era of widespread use of thrombolysis and
direct coronary artery dilation, noninvasive techniques
for detecting jeopardized but viable myocardium are
becoming important for patient management.
Two different approaches can be used that rely on
the alterations that can be found after infarction. Metabolic disturbances with an increased reliance on glycolysis are known to occur frequently in ischemic
myocardium. Metabolic imaging techniques such as
positron emission tomography or thallium or fatty acid
scintigraphy rely on these changes to differentiate
viable tissue from infarction. Conversely, contractile
reserve determination with low-dose dobutamine
From the Departments of aCardiology, bNuclear Medicine, and cElectronics, Free
University of Brussels (VUB); and the dDepartment of Radiology, University Hospital
Leiden.
This work was supported by a grant from the Belgische Cardiologische Liga and the
Nationaal Fonds voor Wetenschappelijk Onderzoek.
The Gadolinium contrast (Dotarem R) was provided by Guerbet.
Submitted June 20, 1997; accepted Dec. 17, 1997.
Reprint requests: Paul Dendale, MD, Heart Centre, Virga Jesse Hospital, Stadsomvaart 11, 3500 Hasselt, Belgium.
Copyright © 1998 by Mosby, Inc.
0002-8703/98/$5.00 + 0 4/1/88881
echocardiography or magnetic resonance imaging
(MRI) is based on the knowledge that the depressed
contractile function of viable myocardium can be
recruited by inotropic stimulation.
MRI allows different approaches for viability detection,
depending on the imaging technique used. Spin echo
imaging is well suited to measure precisely the wall
thickness of the infarct region.1 A minimal wall thickness
of more than 6 mm suggests viable tissue in long-term
situations.2 The use of contrast agents during spin echo
or ultrafast perfusion imaging3-19 is useful in the more
acute stages of the infarction. Functional imaging with
wall thickening and the contractile response to dobutamine is the most recently described technique.20,21 In
this study, the relation between contrast enhancement
patterns and functional studies is analyzed.
Methods
Patient population
Twenty-eight patients presenting with an acute myocardial
infarction (creatine kinase peak >500 IU/L) and treated with
thrombolysis or direct percutaneous transluminal coronary
angioplasty (PTCA) were prospectively entered in the study.
American Heart Journal
May 1998
876 Dendale et al.
Figure 1
Twenty-two male and six female patients with a mean age of
64 ± 12 years were scanned within the first 2 weeks after
infarction. Eighteen patients presented with an inferior or lateral infarction and 10 patients with an anterior or septal
infarction based on ECG. The presumed infarct region was
determined on the basis of the ECG evolution and the results
of the coronary angiogram.
Twenty-five patients underwent coronary angiography in
the first 2 weeks after the infarction. One patient showed
normal coronary arteries, three had single-vessel, seven had
two-vessel, and 14 had three-vessel disease. Seven of the 25
patients examined had occluded infarct-related arteries. The
culprit artery was the left anterior descending in nine
patients, left circumflex artery in 12, and right coronary
artery in four.
Exclusion criteria were hemodynamic instability, irregular
heart rate, contraindications for MRI (pacemakers or ferromagnetic vascular clips), or contrast medium injection. The
protocol was approved by the Ethical Committee of the Free
University of Brussels, and the patients gave informed consent before entering the study.
MRI examination
All patients were examined between 6 and 14 days after
the infarction with a gradient echo and a spin echo MRI
examination after discontinuation of β-blockers for at least
48 hours. All MRI images were acquired with a 1.0 Tesla
Siemens Magnetom in double oblique orientation, resulting
in short-axis representation.
A spin echo series was obtained 15 minutes after injection of 0.1 mmol/kg gadolinium-DOTA (Dotarem, Guerbet,
Aulnay-sous-Bois, France). The whole of the left ventricle
was scanned with slices of 8 mm thickness and 2 mm gap.
All MRIs were made with a field of view of 300 using the
body coil and an acquisition matrix of 128 × 256 lines. Each
series was made with two acquisitions, a repetition time
(TR) equivalent to the RR interval and an echo time (TE) of
25 msec.
A gradient echo series was made with two midventricular
short-axis slices of 10 mm thickness and a gap of 20 mm.
The TR was 50 msec; the TE was 6 msec, with a flip angle of
30 degrees. The number of phases was dictated by the RR
interval (12 to 20 phases).
After this series, a continuous infusion of dobutamine in a
dose of 5 mg/kg/min was started and the gradient echo MRI
was repeated with exactly the same parameters 5 minutes
after the beginning of the perfusion.
Image analysis
Subendocardial anteroseptal infarct enhancement: spin echo
image (top). Transmural lateral infarct enhancement pattern
(middle). Transmural lateral infarct enhancement with central
hypointense zone: “doughnut pattern” (bottom).
All MRIs were stored on optical disk and transferred to a
SUN spark station (Sun Microsystems) for image analysis.
From the spin echo series, the two slices corresponding to
the level of the gradient echo images were selected for
analysis. The analysis of the wall thickening and the contrast
American Heart Journal
Volume 135, Number 5, Part 1
Dendale et al. 877
Table I. Relation between infarct enhancement on spin echo imaging and diagnosis of viability on dobutamine gradient echo
MRI (p < 0.0003)
Dobutamne negative
Transmural
Subendocardial
Normal
Total
Dobutamine positive
10
4
2
16
enhancement patterns was done by one observer on separate days.
The spin echo images were examined for a signal
enhancement in the region of the infarction induced by the
contrast agent by use of the method described by Bouchard
et al.22 The window width was set to zero, and the window
level was adjusted to completely null the signal of the
myocardium opposite the area with increased signal intensity. The borders of the remaining zone of high intensity
were then drawn and the window width and level were
returned to baseline. The relation between the contour
drawn and the endocardium and epicardium were analyzed. The signal enhancement pattern was classified as
transmural, subendocardial, or absent. Subendocardial
infarct enhancement (Fig. 1, top) was defined as an
enhancement that in no place reached the epicardium.
Transmural enhancement was defined as a signal increase
extending to the epicardium (Fig. 1, middle). All regions
with an increased signal intensity were examined for the
presence of central zones of reduced signal intensity (Fig. 1,
bottom). A signal enhancement surrounding a region without signal was defined as a “doughnut pattern.”
The gradient echo images were analyzed separately from
the spin echo images. The myocardium in the region of
interest was classified as showing normal contractility,
hypokinesis (diminished but not absent wall thickening in
comparison to the other segments in the same slice), or akinesis (absent wall thickening).
The dobutamine-stimulated images were classified as
unchanged wall thickening or improved wall thickening
(hypokinesis to normal; akinesis to hypokinesis or
normokinesis).
3
11
2
16
Normal
Total
4
7
11
22
17
22
15
54
segments. Of the 19 segments with transmural signal
enhancement, five were found to have a central
region of low-signal intensity, whereas none of the
segments with a subendocardial pattern showed this
feature.
In the gradient echo study, 22 segments were akinetic, 12 were hypokinetic, and the remaining 22 were
normokinetic.
Fifty-four of the dobutamine-stimulated dynamic
images were of sufficient quality to be analyzed: they
showed improvement in wall motion in 16 of the 32
slices with abnormal wall motion at rest and
unchanged wall motion in 16.
Relation between infarct enhancement pattern and
coronary anatomy
In transmural infarct enhancement, 22% of the vessels were occluded compared with 30% in the subgroup with subendocardial enhancement and 25% in
the subgroup with absence of infarct enhancement, a
difference that was not significant.
Relation between infarct enhancement pattern and
basal wall thickening
The chi-square test with Yates correction was used to
analyze the significance of the difference between subgroups.
In segments with transmural infarct enhancement,
abnormal wall thickening was present in 15 of 19
(79%) segments, most of them being akinetic (12 of
15). Four of the five segments with a central
hypointense region were akinetic. Among the 22 segments with subendocardial enhancement, 15 (68%)
showed abnormal wall thickening, with a slight predominance of hypokinesis (9 of 15). Segments without
infarct enhancement were most often normokinetic
(11 of 15).
Results
Relation between infarct enhancement pattern and
contractile response to low-dose dobutamine
On the spin echo images, an increased signal intensity was observed in 41 (73%) of the 56 infarct segments: a transmural enhancement was found in 19
segments, and a nontransmural in the remaining 22
Table I shows a significant relation between contrast
MRI and the response to low-dose dobutamine stimulation (p = 0.0003). Transmural infarct enhancement
corresponded with dobutamine-negative myocardium
Statistical analysis
878 Dendale et al.
in 10 of 17 (59%) cases, whereas subendocardial
infarct enhancement was related to normal or dobutamine-positive myocardium in 18 of 22 (82%) cases.
The five segments with a doughnut pattern were all
dobutamine negative (nonviable). Only two of 15 segments without infarct enhancement were shown as
nonviable by dobutamine MRI.
Combination of the different imaging techniques
The combination of transmural infarct pattern with
akinesis was related to an absence of contractile
reserve in 9 of 10 (90%) segments, whereas nontransmural or absent infarct enhancement in combination
with hypokinesis predicted a positive response to
dobutamine in 7 of 9 (78%) (p < 0.001). In all cases, a
doughnut pattern was predictive of nonviability, independently of the wall motion.
Discussion
In this study, two major infarct enhancement patterns were detected after contrast spin echo MRI:
subendocardial and transmural. In the transmural
group, a homogeneous pattern and a pattern with a
central hypointense zone (doughnut pattern) were
found. A significant association was found between
contrast pattern, wall thickening, and contractile
reserve. Transmural infarct enhancement was found
predominantly in akinetic segments, whereas subendocardial or absent infarct enhancement was found
more in hypokinetic segments. Almost half of the
homogeneous transmural patterns showed improved
wall thickening during dobutamine stimulation, suggesting viability, whereas the doughnut pattern was
specific for necrotic myocardium. Subendocardial or
absent contrast patterns showed contractile reserve in
more than 80%. The combination of the information of
contrast and wall motion analysis resulted in the distinction of subgroups with very high (>75%) and very
low (<15%) probability of viability.
Infarct enhancement in myocardial infarction
In our study, 25 of the 28 patients (89%) showed an
infarct enhancement on spin echo imaging. The published literature confirms the accuracy of contrast
enhanced MRI in the detection and quantification of a
recent infarction. In a study by Van Dijkman et al.,16,17
a clear increase in signal intensity was seen after
gadolinium contrast in the acute phase of myocardial
infarction. This change remained present for at least 6
weeks after the infarction.
American Heart Journal
May 1998
The absence of enhancement in three of our
patients can be explained by the size23 and localization of the infarction (apical images are more difficult
to analyze), or by the presence of ischemic or
stunned, but not necrotic, tissue. Also, the variability
in signal intensity in normal myocardium reduces the
contrast between infarction and normal tissue, and
this might decrease the sensitivity for the detection of
infarction. The slice selection was done without
knowledge of the localization of the infarction; therefore we cannot exclude the possibility that a study
using more slices to image the heart would have
found zones with signal enhancement.
The data in the literature are not conclusive in
determining the exact nature of infarct enhancement
in spin echo imaging: the presence of edema in the
infarct region does not explain the time course of the
signal increase. Although at 6 weeks edema in the
infarct region has certainly disappeared, some studies16,17 still show a clear increase in signal intensity in
the infarct region. The influence of uptake and
washout kinetics of the contrast product might play
an important role. As studies with contrast enhanced
CT scanning24 and MRI25,26 have already shown, the
signal intensity increases first in the normal
myocardium, and the infarct is visualized only several
minutes later, probably due in part to the slower
blood flow in the infarct region.
Infarct enhancement patterns and viability
In our study, two main infarct enhancement patterns
were identified: transmural and subendocardial. Most
segments with a transmural infarct enhancement pattern were found to be akinetic (12 of 19), with 41% (7
of 17) showing viability (defined as the presence of
contractile reserve during dobutamine). Of the five
segments with a doughnut pattern, four were akinetic,
and none showed signs of contractile reserve. In contrast, subendocardial enhancement was related to
hypokinesis or normal wall thickening in most segments, with a very high probability of viability (82%).
However, in our study, a few important discrepancies
were found between the results of the contrast study
and the wall thickening studies: in four segments with
transmural infarct enhancement, the wall thickening
was completely normal, and four other segments without enhancement were akinetic, showing a positive
response to dobutamine in two cases.
For the former four segments, the timing of the contrast study could be an explanation: recovery of func-
American Heart Journal
Volume 135, Number 5, Part 1
tion (of stunned myocardium) might precede the disappearance of infarct enhancement. The lack of specificity
of gadolinium contrast for necrosis is clearly demonstrated by these extreme cases. The absence of infarct
enhancement in akinetic segments might be related to
the presence of hibernating myocardium. Wall motion
abnormalities are known to exist even in the absence of
necrosis, and only a part will show recovery of function
during stimulation with dobutamine. The four segments
with akinesis and absence of infarct enhancement in
our study were all dependent on severely stenosed
(≥90%) or occluded vessels. Obviously, technical reasons such as imperfect matching of the segments might
also explain part of the discrepancies observed, even
though the side-to-side matching of the segments guaranteed the best possible concordance.
Several other studies analyzed the relation between
infarct enhancement and viability. In one study,5 a
subendocardial signal increase was 100% predictive for
wall motion improvement in the very short term (2
weeks). Transmural or nonhomogeneous infarct segments showed no improvement in wall motion.
In an animal study using gadolinium,9 a good relation was found between signal increase and nonreversible jeopardized myocardium at 6 and 48 hours of
reperfusion. In our study, several clear cases were
found of patients with transmural infarct enhancement
who showed viable myocardium on stimulation.
As in our study, Judd et al.26 showed that hypoenhanced regions with a hyperenhanced rim were very
sensitive for necrotic myocardium. The higher sensitivity in the latter study might be explained by the
dynamic nature of their imaging sequence. The spin
echo sequence in our study was done 15 minutes after
injection of contrast, whereas Judd et al.26 followed
the contrast evolution in the first 15 minutes using
ultrafast imaging. In patients with low blood flow to
the infarcted region, more time might be needed to
allow accumulation of the contrast agent, resulting in
different patterns depending on the time after infusion.
This is supported by the time intensity curves of the
study by Judd et al.,26 showing overlapping of the
hyperenhanced and hypoenhanced curves at 14 minutes after the contrast injection.
Clinical implications
Contrast patterns might eventually preclude the
necessity for low-dose dobutamine stimulation. As
shown in our study, a negative contrast MRI or a
subendocardial infarct enhancement would have a
Dendale et al. 879
high positive predictive value for viability. In this subgroup of patients, viability studies by low-dose dobutamine would not be needed. In the subgroup of
transmural infarct enhancement, the differentiation
should be based on the presence or absence of
hypointense zones and on wall motion analysis at rest
and during stimulation. This way a rapid and accurate
prediction of viability could be obtained. A larger trial,
however, is needed to confirm these data before the
technique is used in clinical practice.
Limitations of the study
No follow-up of wall motion was available, so the
diagnosis of viability was made only by low-dose
dobutamine MRI. However, a large body of literature
exists showing the accuracy of low-dose dobutamine
echocardiography to predict residual myocardial viability early after myocardial infarction, and recent studies have shown a comparable accuracy of MRI.20,21
Only two slices were analyzed in each patient,
which could result in an underdiagnosis of myocardial
infarction in our patients. However, most of the
patients (46 of 56 slices, or 26 of 28 patients) showed
an abnormality on at least one of the two slices. Also,
because we directly compared the same slice on the
different imaging techniques, the conclusions regarding the relation between contrast enhancement patterns and viability remain valid. Inasmuch as one
investigator analyzed all images, no data about reproducibility of the analysis are available from this study.
However, reproducibility was shown to be very good
in an earlier publication.20
Conclusion
Subendocardial and transmural signal intensity
increase during contrast-enhanced spin echo MRI is
often seen in the infarct region. These patterns can be
of use in the determination of viability after infarction.
The subendocardial infarct enhancement is accurate in
predicting viability, whereas the transmural pattern
can be seen in viable and in nonviable infarct regions.
The presence of hypointense zones in the infarct
region (doughnut pattern) predicts necrosis.
References
1. Akins EW, Hill JA, Sievers KW, Conti CR. Assessment of left ventricular
wall thickness in healed myocardial infarction by magnetic resonance
imaging. Am J Cardiol 1987;59:24-8.
2. Dubnow MH, Burchell HB, Titus JL. Postinfarction ventricular aneurysm:
a clinicomorphologic and electrocardiographic study of 80 cases. Am
Heart J 1965;70:753-60.
American Heart Journal
May 1998
880 Dendale et al.
3. De Roos A, Matheijssen NAA, Doornbos J, Van Dijkman P, Van
Voorthuizen AE, Van Der Wall EE. Myocardial infarct size after reperfusion therapy: assessment by gadolinium-DTPA-enhanced MR imaging. Radiology 1990;176:517-21.
4. Eichtstaedt HW, Felix R, Dougherty FC, Langer M, Rutsch W, Scmutzler H. Magnetic resonance imaging in different stages of myocardial
infarction using the contrast agent gadolinium-DTPA. Clin Cardiol
1986;9:527-35.
5. Fukuzawa S, Watanabe H, Shimada K, Katagiri N, Ozawa S. Distribution patterns of Gd-DTPA-enhanced magnetic resonance imaging
after intravenous tissue plasminogen activator therapy for acute
myocardial infarction. Jpn Circ J 1994;58:199-205.
6. Matheijssen NAA, De Roos A, Van Der Wall EE, Doornbos J, Van
Dijkman PR, Bruschke AV, et al. Acute myocardial infarction: comparison of T2-weighted and T1-weighted gadolinium-DTPA-enhanced MR
imaging. Magn Reson Med 1991;17:460-9.
7. Nishimura T, Kobayaski H, Ohara Y, Yamada N, Haze K, Takamiya
M, et al. Serial assessment of myocardial infarction by using gated
MR imaging and Gd-DTPA. Am J Roentgenol 1989;153:715-20.
8. Nishimura T, Yamada Y, Hayashi M, Kozuka T, Nakatani T, Noda H,
et al. Determination of infarct size of acute myocardial infarction in
dogs by magnetic resonance imaging and gadolinium-DTPA: comparison with indium-111 antimyosin imaging. Am J Physiol Imag 1989;
4:83-8.
9. Ovize M, Revel D, De Lorgeril M, Pichard JB, Dandis G, Delaye J, et
al. Quantitation of reperfused myocardial infarction by Gd-DOTAenhanced magnetic resonance imaging: a experimental study. Invest
Radiol 1991;26:1065-70.
10. Peshock RM, Malloy CR, Buja LM, Nunnally RL, Parkey RW, Willerson
JT. Magnetic resonance imaging of acute myocardial infarction:
gadolinium diethylenetriaminepentaacetic acids as a marker of reperfusion. Circulation 1986;74:1434-40.
11. Ryan T, Tarver RD, Duerk JL, Sawada SG, Hollenkamp NC, Johnson J,
et al. Distinguishing viable from infarcted myocardium after experimental ischemia and reperfusion by using nuclear magnetic resonance
imaging. J Am Coll Cardiol 1990;15:1355-64.
12. Schaefer S, Malloy CR, Katz J, Parkey RW, Buja M, Willerson JT, et al.
Gadolinium-DTPA-enhanced nuclear magnetic resonance imaging of
reperfused myocardium: identification of the myocardial bed at risk. J
Am Coll Cardiol 1988;12:1064-72.
13. Sechtem U, Voth E, Baer F, Schneider C, Theissen PK, Schicha H.
Assessment of residual viability in patients with myocardial infarction
using magnetic resonance techniques. Int J Card Imaging 1993;
9(Suppl 1):31-40.
14. Tscholakoff D, Higgins CB, Sechtem U, McNamara MT. Occlusive
and reperfused myocardial infarcts: effect of Gd-DTPA on ECG-gated
MR imaging. Radiology 1986;160:515-9.
15. Van Der Wall EE, Van Dijkman PRM, De Roos A, Doornbos J, Van Der
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
Laarse A, Manger Cats V, et al. Diagnostic significance of gadoliniumDTPA enhanced magnetic resonance imaging in thrombolytic treatment
for acute myocardial infarction: its potential in assessing reperfusion. Br
Heart J 1990;63:12-7.
Van Dijkman PR, Hold KM, Van Der Laarse A, Holman ER, Ozdemir
HI, Van Nat TH, et al. Sequential analysis of infarcted and normal
myocardium in piglets using in vivo gadolinium-enhanced MR images.
Magn Reson Imaging 1993;11:207-18.
Van Dijkman PRM, Van Der Wall EE, De Roos A, Doornbos J, Van Der
Laarse A, Manger Cats V, et al. Acute, subacute and chronic myocardial infarction: quantitative analysis of gadolinium-enhanced MR
images. Radiology 1991;180:147-51.
Van Rossum AC, Visser FC, Van Eenige MJ, Sprengers M, Valk J, Verheugt FW, et al. Value of gadolinium-diethylene-triamine pentaacetic
acid dynamics in magnetic resonance imaging of acute myocardial
infarction with occluded and reperfused coronary arteries after thrombolysis. Am J Cardiol 1990;65:845-51.
Wolfe CL, Moseley ME, Wikstrom MG, Sievers RE, Wendland MR,
Dupon JW, et al. Assessment of myocardial salvage after ischemia
and reperfusion using magnetic resonance imaging and spectroscopy.
Circulation 1989;80:969-82.
Dendale P, Franken PR, Waldman GJ, De Moor D, Tombeur D, Block
P, et al. Low dose dobutamine magnetic resonance imaging as an
alternative for echocardiography in the detection of viable
myocardium after acute infarction. Am Heart J 1995;130:34-140.
Baer FM, Voth E, Schneider CA, Theissen P, Schicha H, Sechtem U.
Comparison of low-dose dobutamine-gradient-echo magnetic resonance imaging and positron emission tomography with 18F-fluorodeoxyglucose in patients with chronic coronary artery disease. Circulation 1995;91:1006-15.
Bouchard A, Reeves RC, Cranny G, Bishop SP, Pohost GM. Assessment of myocardial infarct size by means of T2 weighted 1H nuclear
magnetic resonance imaging. Am Heart J 1989;117:281-9.
Holman ER, Van Jongergen HPW, Van Dijkman RM, Van Der Laarse A,
De Roos A, Van Der Wall E. Comparison of magnetic resonance imaging studies with enzymatic indexes of myocardial necrosis for quantification of myocardial infarct size. Am J Cardiol 1993;71:1036-40.
Doherty PW, Lipton MJ, Berninger WH, Skioldebrand CG, Carlsson
E, Redington RW. Detection and quantification of myocardial infarction
in vivo using transmission computed tomography. Circulation
1981;63:597-606.
Wesbey G, Higgins CB, Lanzer P, Botvinick E, Lipton MJ. Imaging and
characterization of acute myocardial infarction in vivo by gated
nuclear magnetic resonance. Circulation 1984;69:125-30.
Judd RM, Lugo-Olviieri CH, Arai M, Kondo T, Croisille P, Lima JA, et al.
Physiological basis of myocardial contrast enhancement in fast magnetic resonance images of 2-day-old reperfused canine infarcts. Circulation 1995;92:1902-10.