| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cardiac Imaging |
1 From the Departments of Radiology (J.F.P., A.S.C., C.A.), Nuclear Medicine (M.W.), and Cardiology (C.C., B.L., G.D.), Centre Chirurgical Marie Lannelongue, 133 Avenue de la Résistance, 92350 Le Plessis-Robinson, France. Received January 26, 2004; revision requested April 6; revision received August 6; accepted September 8, 2004. Address correspondence to J.F.P. (e-mail: pauljf{at}ccml.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Institutional review board approval and informed consent were obtained. A total of 34 patients (29 men and five women; mean age, 56 years ± 13) underwent dual-phase 16-detector row CT within 3 days ± 3 after successful reperfusion of acute MI. Iodinated contrast medium (1.5 mL per kilogram of body weight) was injected at a flow rate of 3.5 mL/sec. A first arterial phase acquisition was followed 5 minutes later by a late acquisition, without reinjection of contrast medium. A radiologist and a cardiologist used a 17-segment model in a blind analysis of images obtained during late acquisition. For each segment, presence of late defect or late enhancement was recorded. Findings were compared with SPECT studies analyzed by a nuclear medicine physician and a cardiologist 6 weeks after the acute event. CT defects were compared with SPECT defects on a segmental and per-patient basis. Mean number of segments with late defects on multi–detector row CT scans was compared with infarct size on SPECT images by using the t test.
RESULTS: All patients had late enhancement in the infarcted myocardium. In 27 of 34 patients, a late defect surrounded by a subepicardial late enhancement was detected. Segments with late defect on CT scans were predictive of residual perfusion defects at 6-week follow-up, with sensitivity of 78%, specificity of 91%, and accuracy of 90%. On a per-patient basis, sensitivity was 93%, specificity was 100%, and accuracy was 94%. Mean number of segments with late defects on multi–detector row CT scans (ie, 3.1 segments) was not significantly different from infarct size on SPECT images (eg, 2.5 segments) (P = .2).
CONCLUSION: Late defect on multi–detector row CT scans indicates residual perfusion SPECT defect and infarct size after successfully reperfused MI, with sensitivity of 93%, specificity of 100%, and accuracy of 94%.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Multi–detector row computed tomography (CT) is a promising technique for imaging coronary arteries and myocardium. Some reports have indicated that CT may show myocardial contrast abnormalities (ie, early defect at arterial phase or late enhancement 5–10 minutes later, without reinjection) associated with MI (5,6). Early after reperfused acute MI, we frequently observed a late defect within late enhancement in the infarcted myocardium. We hypothesized that this late defect seen on multi–detector row CT scans may correspond to irreversible myocardial injury. Thus, the purpose of our study was to prospectively assess the accuracy of multi–detector row CT in the prediction of infarct size after successful reperfusion of acute MI by using SPECT scans obtained 6 weeks later as the reference standard.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Multi–Detector Row CT Technique and Evaluation
Examinations were performed with a 16-detector row CT scanner (Sensation 16; Siemens, Erlangen, Germany). A highly concentrated iodinated contrast medium, iopamidol (Iopamiron 370; Schering, Berlin, Germany) (1.5 mL per kilogram of body weight), was injected at a flow rate of 3.5 mL/sec. The acquisition protocol included both an arterial phase acquisition for analysis of coronary arteries and stent patency and a delayed phase acquisition performed 5 minutes after this first-pass acquisition for analysis of left ventricle myocardial enhancement. Imaging parameters for the delayed phase acquisition were as follows: rotation time, 420 msec; collimation, 1.5 mm; tube voltage, 80 kV; tube current, 500 mAs.
The multi–detector row CT scans were analyzed visually, in consensus, by one radiologist (J.F.P.) and one cardiologist (G.D.), with 7 and 2 years of experience with cardiac CT, respectively. The observers were blinded to clinical data at the time of the study. The contrast enhancement variations (ie, late defect or late enhancement seen on scans obtained with 5-minute delayed acquisition) were assessed with multiplanar long-axis (ie, two and four cavities) and short-axis (ie, basal, median, or apical) views by using 8-mm-thick slab sections. Variations were assessed by comparison with remote noninfarcted myocardium. Findings were recorded on a diagram representing the heart as a 17-segment model (6), and the number of segments with late defect was evaluated for comparison with infarct size on SPECT images. In addition, any late defect involving more than 25% of the left ventricle thickness was recorded. Involvement of myocardial thickness was measured directly on the screen with calipers. Only data from the delayed acquisition were used for this study, and observers did not use data from the arterial phase for interpretation.
SPECT Technique and Evaluation
SPECT findings were acquired 1 hour after injection of 350 MBq of 99mTc sestamibi (Bristol-Meyers Squibb, Rueil-Malmaison, France) with a dual-head camera (DST camera; GE Medical Systems, Milwaukee, Wis) with rectangular detectors equipped with low-energy high-spatial-resolution collimators. Images were acquired by using a 64 x 64 matrix, 32 projections, and a 180° circumferential orbit. Gated SPECT images were acquired with the stop condition of 60 seconds of accepted beats per projection and 16 views per cardiac cycle. Projection images were filtered through a low-pass Butterworth filter with a frequency cutoff of 0.25 cycles per pixel and an order of 5.0. Classical orthogonal tomographic sections were reconstructed. One cardiologist (G.D., with 2 years of experience) and one nuclear medicine physician (M.W., with 16 years of experience) evaluated, in consensus, SPECT sections for segmental sestamibi uptake. These physicians were blinded to multi–detector row CT data and clinical information. A threshold of 60% of peak counts was used to identify residual perfusion defects. Like CT reports, findings were drawn on a similar diagram representing the heart as a 17-segment model (7).
Data and Statistical Analysis
The number of segments showing late enhancement or late defect on multi–detector row CT scans and the number of segments with SPECT defects was recorded, and the mean was calculated. The sensitivity, specificity, and overall accuracy of late defects on multi–detector row CT scans in the prediction of SPECT defects at 6-week follow-up were calculated on a segmental basis by first considering segments with any late defect and then considering segments with at least 25% left ventricular thickness involvement. Thereafter, sensitivity, specificity, and overall accuracy were calculated on a per-patient basis. SPECT images obtained at 6-week follow-up served as the reference standard.
The mean number of segments with late defect on multi–detector row CT scans was compared with mean infarct size, which was defined as the number of segments with perfusion defects on SPECT images. The data are presented as mean ± standard deviation. The t test was used to compare means. A P value of .05 or less was considered to indicate a statistically significant difference. For comparison of infarct size on multi–detector row CT scans and SPECT images, a power analysis was conducted to determine (with 80% confidence) the minimum sample size necessary to detect a one-segment difference in means between the methods. Calculations were performed with software (nQuery Advisor, version 2; Statistical Solutions, Saugus, Mass) and revealed that the required minimum sample size was 34 patients.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
SPECT at 6-week Follow-up
A total of 578 segments in 34 patients were analyzed. A total of 82 SPECT defects in 29 patients were found. Five of the 34 patients (15%) did not have any residual defects.
CT at 3-day Follow-up
All 34 patients had late enhancement of the left ventricle on CT scans in the area of MI. The mean number of segments with late enhancement was 8 ± 3 (minimum number of segments, four; maximum number of segments, 13). Twenty-seven of 34 patients (79%) had late defects within the late enhanced areas (Fig 1).
|
|
Analysis by patient.—Twenty-seven patients with late defect were identified on the basis of multi–detector row CT scans. All 27 patients had true-positive findings. Five of seven patients had true-negative findings (ie, no late defect on 3-day multi–detector row CT scans and no residual perfusion defects on 6-week SPECT images) (Fig 2). Two patients had false-negative findings; for these two patients, there was only one segmental SPECT defect. On a per-patient basis, the sensitivity, specificity, and accuracy of late defect in the prediction of SPECT defects were as follows: 93% (27 of 29 patients), 100% (five of five patients), and 94% (32 of 34 patients), respectively.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The main findings of our study are as follows: The presence of any late defect on multi–detector row CT scans obtained 3 days after MI in patients is a predictor (sensitivity, 93%) of residual defects on SPECT images obtained 6 weeks after MI. For prediction of fixed infarction involving at least two segments, sensitivity was 100%. For each segment, the detection of a late defect had a sensitivity of 79% for prediction of infarction. The time after injection of contrast medium may be important in the detection of late defects. We arbitrarily chose a standard delay of 5 minutes after injection on the basis of initial cases. For ethical reasons, we did not perform repeated acquisitions at different times. A different interval between injection and acquisition may result in different findings, and further studies are needed to determine the optimal delay.
The specificity for prediction of infarcted segments was slightly greater for segments with a more transmural extent (>25% of myocardial thickness) of late defect (94% vs 91%), which suggests more irreversible myocardial injury when the myocardial thickness involved was greater. We found 42 false-positive segments, some of which could be due to partial recovery of segments with late defect between the two examinations. In all five patients without any late defect on multi–detector row CT scans, no residual defect was found on SPECT images (specificity, 100%). This suggests that segments with late enhancement but without late defect may have experienced reversible injury, unlike segments with late defect, which were associated with fixed infarcts.
Overall accuracy (ie, 90%) of multi–detector row CT at 3-day follow-up was very good with the 17-segment model analysis. Some of the discrepancies (eg, false-positive or false-negative segments) may have been due to differences in the orientations of multiplanar sections, which were obtained manually and independently by the two operators for each technique. The number of segments involved was slightly higher but not statistically different between multi–detector row CT and SPECT, despite the delay between the two examinations. Our study shows that shortly after MI, the number of cardiac segments with late defect on multi–detector row CT scans is a good predictor of fixed infarct size. Early determination of infarct size is an advantage of multi–detector row CT over 6-week SPECT because infarct size may be estimated before discharge. Early prognosis information may lead to changes in patient care, and late defect assessment can be used as a benchmark to compare postinfarction therapeutic strategies.
Persistence of hypoattenuation over time is responsible for late defects on multi–detector row CT scans and is probably due to microvascular obstruction, which results in poor penetration of contrast medium into the deep layers of the myocardium, despite reopened epicardial arteries. This "no reflow" phenomenon is well established, and it may be observed with other imaging techniques, including contrast echocardiography (13) and magnetic resonance (MR) imaging, where it appears as a central dark zone surrounded with contrast hyperenhancement. By using MR imaging, the presence of these dark zones has been correlated with poor recovery (14); however, to our knowledge, segmental analysis has not been used to compare MR imaging with other imaging techniques.
Multi–detector row cardiac CT is a quick and simple imaging technique. The mean time for late acquisition was 10 seconds. The standard radiation dose per acquisition with cardiac CT is relatively high (ie, about 4 mSv with the electrocardiographic pulsing technique) (15), but we have lowered the tube voltage to 80 kV for delayed acquisition to allow better contrast enhancement (16). Iodine has a high atomic number, which is responsible for more attenuation at lower tube voltage settings. Thus, in the delayed phase, the radiation dose we used was 65% lower than that used for standard image acquisition in the arterial phase, and it was responsible for greater image noise. Noise was reduced by using 8-mm-thick multiplanar section reconstructions and making consensual assessment of myocardial changes possible in all cases.
This preliminary study has some important limitations: SPECT is an imperfect reference standard because it is known to result in a large number of false-positive studies, primarily because of attenuation artifacts. Thus, some multi–detector row CT scans may have been classified as false-negative because SPECT images were false-positive. Recently, MR imaging has been proved to be more sensitive than SPECT in the detection of small MIs (17) because of poor partial resolution of SPECT images; however, our study population included patients with large Q-wave MIs, thus excluding the risk of not detecting small MIs. Lack of spatial resolution, however, may account for segmental discrepancies between multi–detector row CT and SPECT. Comparing multi–detector row CT with MR imaging or positron emission tomography would be of great interest; however, such comparisons were beyond the scope of this preliminary study. Interobserver variability was not assessed because difference of experience between the two CT readers may have been responsible for most of the interobserver variability. Hounsfield unit measurements in regions of interest were not used because values in hypoenhanced regions showed large local variations in the same area, as well as in remote noninfarcted myocardium between individuals. A visual segmental assessment was more relevant for this preliminary study; however, additional studies may be needed to determine if Hounsfield units may help in the detection of nonviable myocardium. Some artifacts may be responsible for false-positive segments, but most of them could be seen as transverse dark lines crossing the left ventricle on short-axis images. We did not obtain unenhanced images to avoid excessive radiation dose, and we cannot exclude relevant information about myocardium on unenhanced images.
To our knowledge, this study provides the first assessment of the accuracy of late defect on multi–detector row CT scans as a predictor of infarct size shortly after acute MI. Multi–detector row CT is a simple method, and it provides useful myocardial information with good accuracy. It appears to be potentially valuable for routine clinical use.
|
FOOTNOTES |
|---|
Abbreviations: MI = myocardial infarction
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, J.F.P., M.W., G.D.; study concepts and design, J.F.P., M.W., G.D.; literature research, J.F.P., M.W., G.D.; clinical studies, J.F.P., M.W., G.D., C.C., A.S.C.; data acquisition, J.F.P., M.W., G.D., A.S.C., C.C.; data analysis/interpretation, J.D., J.F.P., M.W.; statistical analysis, J.F.P.; manuscript preparation, J.F.P., M.W.; manuscript definition of intellectual content, J.F.P., M.W., G.D.; manuscript editing, J.F.P., M.W.; manuscript revision/review, all authors; manuscript final version approval, J.F.P.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Boussel, M. Ribagnac, E. Bonnefoy, P. Staat, B. M. Elicker, D. Revel, and P. Douek Assessment of Acute Myocardial Infarction Using MDCT After Percutaneous Coronary Intervention: Comparison with MRI Am. J. Roentgenol., August 1, 2008; 191(2): 441 - 447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lessick, R. Dragu, D. Mutlak, S. Rispler, R. Beyar, D. Litmanovich, A. Engel, Y. Agmon, M. Kapeliovich, H. Hammerman, et al. Is Functional Improvement after Myocardial Infarction Predicted with Myocardial Enhancement Patterns at Multidetector CT? Radiology, September 1, 2007; 244(3): 736 - 744. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Habis, A. Capderou, S. Ghostine, B. Daoud, C. Caussin, J.-Y. Riou, P. Brenot, C. Y. Angel, B. Lancelin, and J.-F. Paul Acute Myocardial Infarction Early Viability Assessment by 64-Slice Computed Tomography Immediately After Coronary Angiography: Comparison With Low-Dose Dobutamine Echocardiography J. Am. Coll. Cardiol., March 20, 2007; 49(11): 1178 - 1185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. T. Abada, C. Larchez, B. Daoud, A. Sigal-Cinqualbre, and J.-F. Paul MDCT of the Coronary Arteries: Feasibility of Low-Dose CT with ECG-Pulsed Tube Current Modulation to Reduce Radiation Dose Am. J. Roentgenol., June 1, 2006; 186(6_Supplement_2): S387 - S390. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Selvanayagam Imaging quantification of infarct size. Radiology, May 1, 2006; 239(2): 613 - 613. [Full Text] [PDF]
|
|
|
|
|
|
J. F. Bruzzi, M. Remy-Jardin, D. Delhaye, A. Teisseire, C. Khalil, and J. Remy When, Why, and How to Examine the Heart During Thoracic CT: Part 1, Basic Principles Am. J. Roentgenol., February 1, 2006; 186(2): 324 - 332. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Bruzzi, M. Remy-Jardin, D. Delhaye, A. Teisseire, C. Khalil, and J. Remy When, Why, and How to Examine the Heart During Thoracic CT: Part 2, Clinical Applications Am. J. Roentgenol., February 1, 2006; 186(2): 333 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A de Roos, L J M Kroft, J J Bax, H J Lamb, and J Geleijns Cardiac applications of multislice computed tomography Br. J. Radiol., January 1, 2006; 79(937): 9 - 16. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
