| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of California at San Francisco, San
Francisco, CA.
2 Department of Radiology, Louis Pradel Hospital, Lyon, France.
3 Present address: Department of Radiology, VA Medical Center, 4150 Clement St.,
San Francisco, CA 94121.
4 Department of Cardiology, Louis Pradel Hospital, Lyon, France.
Received November 9, 2007;
accepted after revision February 11, 2008.
L. Boussel is supported by the Société Française de
Radiologie, 20 Avenue RAPP, 75007 Paris, France.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Nineteen consecutive patients presenting with primary acute myocardial infarction underwent delayed enhanced MDCT immediately after coronary angioplasty and underwent delayed enhanced MRI within 8 days of angioplasty. Only patients with a thrombolysis in myocardial infarction (TIMI) score of 0 or 1 of the culprit coronary artery before endovascular angioplasty and TIMI score of 2 or 3 after angioplasty were selected. Comparison of delayed enhanced MDCT and delayed enhanced MRI was performed by three observers and focused on identifying the involved segments and determining the transmural extent of enhancement and infarct size.
RESULTS. The mean signal intensity was significantly higher in the
involved territory than in healthy myocardium: 197 ± 81 H versus 71
± 20 H, respectively (p < 0.0001). We found significant
agreement between delayed enhanced MDCT and delayed enhanced MRI for the
number of involved segments, transmural extent of enhancement, and infarct
size (r2 = 0.74, 0.76, and 0.67, respectively; p
< 0.0001) with good interobserver reproducibility (
= 0.8).
CONCLUSION. The results of our study show that delayed enhanced MDCT allows accurate visualization of early myocardial contrast uptake compared with delayed enhanced MRI and does not require an additional contrast injection after PCI.
Keywords: cardiac imaging • delayed enhanced MDCT • delayed enhanced MRI • heart disease • myocardial infarction • percutaneous coronary intervention
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Delayed enhanced MDCT has been proposed as an alternative imaging technique for the noninvasive evaluation of myocardial infarct size [4–6]. Traditionally, images are obtained 5 minutes after the injection of IV iodinated contrast material. Recently, the feasibility of obtaining images immediately after percutaneous coronary intervention (PCI) without the need for an additional contrast injection was described by Habis et al. [7]. In that study, the authors also found an inverse correlation between the size of the delayed myocardial contrast uptake in infarcts and myocardial viability as assessed by segmental contraction on sonography 2–4 weeks after PCI. Delayed enhanced MDCT is a promising alternative for early infarct imaging that is fast, is widely available, and may provide information similar to that provided by MRI. There are, however, no other data to our knowledge about the quantification of myocardial delayed contrast uptake after cardiac catheterization comparing CT with other techniques, particularly MRI.
The aim of this study was to compare delayed enhanced MDCT with delayed enhanced MRI in the evaluation of myocardial contrast uptake in acute myocardial infarction after PCI without the need for an additional contrast injection.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Creatine kinase level at admission and maximum troponin value were recorded for each patient.
All angioplasty procedures were performed at our institution using iodinated contrast material (ioxaglate meglumine [Hexabrix, Guerbet]); a mean volume of 10 mL of contrast material was injected at a rate of 2 mL/s.
Informed consent was obtained from all participants after the nature of the procedure had been fully explained. The study was performed in compliance with the requirements of the institutional review board.
CT Protocol
CT scans were obtained on a Brilliance 40 scanner (Philips Healthcare). IV
contrast material was administered during angioplasty, but additional contrast
material was not given specifically for CT. Patients were transported from the
cardiac catheterization laboratory to the CT scanner as quickly as possible
after PCI. The distance between the two rooms is approximately 50 m. A
low-dose retrospective ECG-triggered CT examination of the entire heart was
performed using the following parameters: number of detectors, 40; individual
detector width, 0.625 mm; gantry rotation time, 420 milliseconds; pitch, 0.2;
half scan reconstruction mode; and craniocaudal imaging direction. The tube
current was fixed to 80 kV and 600 mAs per slice. ECG tube modulation was not
used. Reconstruction parameters for axial slices were a 2-mm effective section
thickness, 1-mm increment, standard intermediate reconstruction filter (kernel
CB), and adapted field of view. Retrospective ECG-gated reconstruction in the
middiastolic phase (75% of the R-R interval) was performed.
Cardiac MDCT evaluation was performed on a 3D workstation (Brilliance Workstation, Philips Healthcare). Only the middiastolic phase (75% of the cardiac cycle) was analyzed using a 15-mm-thick multiplanar reconstruction in the short axis and horizontal long axis.
The total volume of contrast material used during the revascularization procedure and the delay between the last contrast injection and the CT acquisition were recorded. In addition, radiation exposure was recorded for each CT examination.
MR Examination
MR studies were performed on a 1.5-T scanner (Intera, Philips Healthcare;
or Avento, Siemens Medical Solutions) using a dedicated cardiac coil (5 and 8
channels, respectively). First, ECG-gated steady-state free precession cine
images were acquired in the two-chamber, four-chamber, and short-axis views.
Second, delayed contrast-enhanced MRI was performed after the injection of
0.15 mmol/kg of a gadolinium-based contrast agent (gadoterate dimeglumine
[Dotarem, Guerbet]). A 3D inversion recovery segmented gradient-echo sequence
was used in the two-chamber and short-axis views. Imaging started 10 minutes
after contrast administration, and the inversion time (range, 260–340
milliseconds) was optimized to obtain adequate nulling of normal myocardial
signal. The imaging sequence para meters included an in-plane voxel size of
from 1.25 x 1.25 mm2 to 1.5 x 1.5 mm2; slice
thickness, 5 mm; flip angle, 25° (Philips unit) or 10° (Siemens unit);
and TR range/TE range, 1.4–3/4.3–9.
Delayed enhanced MRI evaluation was performed on a workstation (Advantage Windows, GE Healthcare). The delay between the CT and MR examinations was recorded.
Image Evaluation and Statistical Analysis
All angiographic images were reviewed by an experienced interventional
cardiologist for TIMI score evaluation before and after reperfusion.
Three blinded experienced observers evaluated the CT scans in random order. Observers were free to adjust the window width and level values. The regional extent of delayed enhancement was assessed using a 17-segment model [10]. Each segment was described as involved or healthy, and the percentage of transmural extent of enhancement was graded using a 4-point scale (1–4): 0–25%, 26–50%, 51–75%, or 76–100% [11]. Furthermore, to assess the total infarct size, infarcted myo cardium was delineated on each slice by two ob servers and the total volume of infarcted myo cardium was calculated.
Interobserver correlation was estimated between each pair of observers using a Cohen's kappa for the number of involved segments classification and a weighted Cohen's kappa for the transmural extent grading to account for the importance of the discrepancies between the observers [12]. Interobserver reproducibility of the measurement of infarct volume was calculated as the SD of the differences in measurements between each pair of observers relative to the global mean value of the measurements [13].
CT image quality of the myocardium was recorded using a 5-point scale: 0, not assessable; 1, motion or bandlike artifacts limiting interpretation; 2, motion or bandlike artifacts that did not limit interpretation; 3, no artifacts, slight blurring at the edges of enhanced areas; and 4, no artifacts, sharply defined enhanced areas. To evaluate microvascular obstruction, observers rated sparing of the immediate subendocardial myocardium by contrast enhancement in regions of infarct in the left ventricle as present (1) or absent (0). Similarly, involvement of the right ventricle was reported.
Delayed enhanced MRI was analyzed using the same criteria by two blinded experienced observers.
Finally, a consensus analysis was performed by the three CT observers and two MRI observers. A linear regression was then calculated between the CT and MRI results regarding involved and healthy segments and grading of transmural extent of enhancement. Sensitivity, specificity, and predictive values of delayed enhanced MDCT were also calculated using MRI as the gold standard.
Pearson's correlation coefficients comparing the CT analysis with the creatine kinase value at admission and maximum troponin value were also calculated. The normality of the variables was previously checked using a skewness and kurtosis test.
The mean attenuation in Hounsfield units of enhanced myocardium on CT—measured within a manually contoured region of interest that included all the visible hyperenhanced area on a slice located in the center of the myocardial infarction—was compared with values in the left ventricular cavity and healthy myocardium using a paired Student's t test. This value was also correlated with the total volume of contrast material used during the revascularization procedure and the delay between the last contrast injection and the CT acquisition. The mean signal-to-noise ratios for MRI in enhanced myocardium, healthy myocardium, and the left ventricular cavity were also provided.
All statistical analysis was performed using statistics software (Intercooled Stata 9, StataCorp).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The mean effective dose for the CT examination was 3.89 ± 0.66 mSv (range, 2.6–5.1 mSv). All CT scans were assessable by all observers. The mean CT image quality score was 3.2 ± 0.5 on the 5-point scale (0–4).
The mean attenuation was significantly higher in the involved territory than in healthy myocardium or the left ventricular cavity: 197 ± 81 H (range, 105–394 H) for the involved territory versus 71 ± 20 H (range, 43–110 H) for healthy myocardium (p < 0.0001) and 102 ± 22 H (range, 70–145 H) for the left ventricular cavity (p < 0.0001) (Fig. 1). The mean signal-to-noise ratio on MRI is provided on this same graph. Mean values are, respectively, 64.3 ± 44.68 (range, 13.9–195.6), 12.2 ± 11 (range, 2.5–40.9), and 54.8 ± 66.6 (range, 6.8–304.3) for involved myocardium, healthy myocardium, and left ventricular cavity.
|
Enhanced segments were in a territory supplied by the occluded coronary artery as seen on angiography. Consensus analysis of all delayed enhanced MDCT images showed that the mean number of involved segments was 5.6 ± 2.4 segments (range, 2–11 segments). The grades of transmural extent of enhancement for each segment were summed to give a total grade of transmural extent. The mean total grade of transmural extent per patient was 21.1 ± 9.6 (range, 5–39). The mean infarct volume was 28.3 ± 13.1 cm3 (range, 8–54.6 cm3).
|
|
|
|
|
|
|
The mean delay between delayed enhanced MDCT and delayed enhanced MRI was 3.5 ± 2 days (range, 1–8 days). MR interobserver reproducibility was very good with a kappa of 0.88 for classification of the number of involved segments and a weighted kappa of 0.88 for grading transmural extent. Interobserver reproducibility was 87.5% for the measurement of infarcted volume.
There was good agreement about the number of involved segments (p < 0.0001, r2 = 0.74), grade of transmural extent (p < 0.0001, r2 = 0.76), and volume of myocardial infarct (p < 0.0001, r2 = 0.67) for both techniques (Figs. 2A, 2B, and 2C). The sensitivity, specificity, positive predictive value, and negative predictive value of CT were, respectively, 90.1%, 96.7%, 93.5%, and 94.9% for the classification of involved versus healthy segments, and 87.6%, 97.7%, 95%, and 93.9% for the classification of transmural extent.
Involvement of the right ventricle was observed in five of 19 patients (26%) with perfect agreement between delayed enhanced MDCT and delayed enhanced MRI (Figs. 3A, 3B, 3C, and 3D).
Sparing of the subendocardial myocardium in regions of infarct was found in 16 patients on MRI and in only nine patients on delayed enhanced MDCT (same patients) with, in every case, visual underestimation on delayed enhanced MDCT of its extent (Figs. 4A, 4B, 4C, and 4D).
|
|
|
|
Finally, a significant correlation was found between the number of involved segments and the initial creatine kinase level and maximal troponin level (p < 0.002). Similarly, correlations between grade of transmural extent and infarct size and between creatine kinase and maximal troponin levels were in the same range (p < 0.002) (Figs. 5A, 5B, 5C, 5D, 5E, and 5F).
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Delayed contrast uptake in infarcted myocardium after IV iodine contrast injection was initially reported by Paul et al. [4]. Lardo et al. [5] reported both excellent localization of contrast uptake to the region of infarct and excellent determination of the extent of infarction in an animal model when compared with histology results. Furthermore, CT has been reported to show good correlation with MRI in both animal and human models [6, 14, 15] with respect to uptake of IV contrast material (gadolinium vs iodine) in infarcted myocardium [16]. Given that MRI has already been validated in the assessment of the extent of myocardial infarct [2], this suggests that CT may also be useful in this evaluation.
The results of our study confirm that in the emergency setting delayed enhanced MDCT can be used to assess the number of involved segments and to quantify transmural extent and infarct size with the same accuracy as delayed enhanced MRI and without the need for an additional contrast injection.
Also, the strong correlation between the extent of myocardial contrast uptake on delayed enhanced MDCT and peak creatine kinase and troponin levels indirectly supports this conclusion because these laboratory values are related to infarct size [17–19].
Despite good general agreement between the MRI and CT findings, some
discrepancies exist. The sensitivity of CT (
90%) is somewhat lower than
the specificity, indicating that some segments showing delayed enhancement on
MRI were missed on CT. This finding is likely because CT is less able to
distinguish involved myocardium from healthy myocardium (mean
involved-to-healthy myocardium signal ratio is smaller on CT than MRI)
particularly at the periphery of infarcted regions. Also, despite its
excellent specificity, CT overestimates infarct size in some cases. Blooming
artifact, as seen in Figures
3A,
3B,
3C, and
3D, can lead to overestimation
of infarct size, particularly with regard to the extent of transmural
enhancement.
Analysis of sparing of the subendocardial myocardium in the infarcted territory did not show the same correlation as hyperenhancement. Delayed enhanced MDCT appears to underestimate these regions compared with MRI. There are several possible explanations for this occurrence. First, the intensity of contrast uptake on delayed enhanced MDCT in the infarcted zone may partially mask the relatively hypoenhanced subendocardial region because of blooming artifact. Also, delayed enhanced MDCT and delayed enhanced MRI were not performed on the same day. Microvascular obstruction, which produces regions of hypoenhancement within an infarct, may not have been present initially when delayed enhanced MDCT was performed because no-reflow is a dynamic process [20, 21]. Finally, direct injection of a coronary artery during angiography could possibly produce deeper diffusion of contrast material into the infarcted area compared with the IV injection used with MRI. Furthermore, the delay between injection and imaging is longer for MDCT than for MRI. This longer delay may lead to increased diffusion of contrast material on MDCT because the recirculation time is longer than that of MRI (average = 22 minutes for CT vs 10 minutes for MRI), particularly for low-reflow zones [22].
The intensity of myocardial enhancement depends on multiple factors. In our study, because the examinations were performed after PCI without an additional contrast injection, these variables include the volume of contrast material given and the time between the last injection and image acquisition. But the lack of correlation between these factors and the intensity of myocardial enhancement underline the fact that other elements, such as a delay in patient management or previous development of collateral flow, may play an important role. However, injection of high doses of contrast material directly into the infarcted region during angiography may counterbalance potential inadequacies of the delay between injection and scanning. Indeed, good image quality was obtained in this study up to 45 minutes after contrast injection as compared with the typical delay of 5–10 minutes with standard CT infarct imaging [16].
Given that the time to perform most examinations was only approximately 10 minutes, MDCT using our technique did not appear to significantly interfere with acute patient care. Radiation, however, is still an important issue. Although MRI involves no ionizing radiation, except for a few highly specialized units [23] MRI is not a practical technique to use in an acute setting. With that limitation in mind, CT appears to be a useful tool to fill the role of immediate postinfarct imaging because nuclear medicine techniques and stress echocardiography, also used clinically for the assessment of viability and risk stratification in acute myocardial infarction [24], are also not practical in the emergency setting. Using a protocol specifically designed for postinfarct imaging, we were able to lower radiation exposure to less than 5.1 mSv. An additional reduction of the dose should be achievable using prospective gating instead of retrospective gating as in our study.
One of the limitations of this study is the inclusion of only patients with total or subtotal coronary artery obstruction (TIMI 0 or 1 at admission). A larger study including patients with higher TIMI scores is needed to assess the accuracy of CT in patients with small infarcts. Another limitation of our study is the variability of IV contrast material administered and the time between contrast injection and acquisition of images. However, because this study was performed in the emergency setting, these parameters were not adjustable.
Another limitation is that we did not assess wall motion and left ventricular function on CT. This element could be of interest because Sato et al. [25] recently reported that left ventricular function at acute phase and at 6-month follow-up was significantly lower in patients with extensive transmural enhancement of the myocardium on CT. A dedicated study should be conducted to assess this point because there are numerous different issues including the relatively low temporal resolution of CT or the low spatial resolution due to the low kilovoltage we used.
In conclusion, when performed immediately after PCI and successful angioplasty of an occluded coronary artery, MDCT appears to be as accurate as MRI in the assessment of myocardial infarct size in patients with a low TIMI score on admission. Myocardial enhancement remains detectable even 45 minutes after directed contrast injection during angioplasty, obviating further contrast injection. Despite the fact that MDCT may underestimate the no-reflow phenomenon, MDCT appears to be a useful technique with which to assess myocardial infarct size in the emergency setting.
Acknowledgments
We thank Philippe Coulon from Philips Healthcare, Suresnes, France. We also
thank Delphine Gamondes and Mohamed Aissat from the Department of
Radiology-Louis Pradel Hospital, Lyon, France.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
