| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Diagnostic Radiology, Severance Hospital, Seodaemun-gu, Shinchon-dong 134, Seoul 120-752, South Korea.
Received July 30, 2008;
accepted after revision September 11, 2008.
Address correspondence to B. W. Choi
(bchoi{at}yuhs.ac).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We included 161 patients (129 men, 32 women; mean age, 60.7 years) who had previously been diagnosed with MI and had undergone CT coronary angiography between February 2003 and April 2005. We analyzed the characteristics of myocardial fat, if present, and compared the clinical parameters of the patients with and those without myocardial fat.
RESULTS. Myocardial fat was found in the subendocardial region in 36 (22.4%) patients with MI. In all cases, the myocardial fat was located in the subendocardial region and was typically detected in the left anterior descending artery territory (75%, n = 27). The mean attenuation value of myocardial fat was –29.6 HU on unenhanced CT. Myocardial fat was more frequently associated with a greater infarct age, milder coronary artery stenosis, and fewer number of diseased vessels. Patients with myocardial fat had more severe regional wall motion abnormalities on follow-up echocardiography. Age, sex, the presence of ST elevation or Q wave, elevated levels of cardiac enzymes, ejection fraction, and end-diastolic left ventricular dimension on follow-up echocardiography, as well as the presence of arrhythmia, were not significantly different between the two groups.
CONCLUSION. Myocardial fat was detected in 22.4% of MI patients and was more frequently associated with a longer postinfarct period, milder coronary artery stenosis, fewer number of diseased vessels, and more severe regional wall motion abnormalities.
Keywords: cardiac imaging • coronary arteries • CT • heart disease • myocardial fat • myocardial infarction
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several cases of incidentally detected myocardial fat in the subendocardial left ventricular wall related to myocardial infarction (MI) have been reported [1, 2]. Myocardial fat at the site of an MI is a common but unappreciated entity, and it is frequently observed as a subendocardial low attenuation on both unenhanced and contrast-enhanced cardiac CT.
Histologic evidence of myocardial fat has been observed at left ventricular myocardial scars in explanted hearts in 68% of patients with ischemic heart disease and 84% of patients with a history of MI [3, 4]. With recent advances in imaging techniques, myocardial fat has also been depicted on cardiac MRI through the use of a fat-suppression pulse [5–8].
In this study, we set out to systemically evaluate myocardial fat in relation to MI in a large group of patients. This study aimed to provide information regarding the incidence of fat in infarcted myocardium, characteristics of subendocardial low attenuation, conditions in which the deposition of fat might be more readily induced, and implications of myocardial fat in a clinical setting. We analyzed and characterized myocardial fat of the left ventricle and evaluated potential correlations between findings of myocardial fat and clinical diagnoses.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MDCT
Unenhanced CT was performed according to the coronary calcium scan protocol
for coronary artery calcium scoring and to achieve accurate localization for
CT coronary angiography and simultaneous evaluation of extracardiac
abnormalities. CT coronary angiography was performed immediately after
unenhanced CT.
CT was performed using a 16-MDCT scanner (Sensation 16, Siemens Medical Solutions). A β-blocker (propranolol hydrochloride [Pranol, Daewoong]) was administered 1 hour before initiation of scanning for all subjects who had a baseline heart rate of > 65 beats per minute. Unenhanced CT was performed using the following scanning parameters: 16 x 1.5 mm collimation, 420-millisecond gantry rotation time, 18-mm-per-scan table feed, 120-kV tube voltage, and 30 mAs. Scanning was performed using an established prospective ECG-triggering protocol for low-dose scanning.
After unenhanced CT was complete, 100–120 mL of nonionic iodinated contrast agent (iopamidol, 370 mg I/mL [Iopamiro, Bracco]) was injected via an 18- or 20-gauge needle at a flow rate of 4 mL/s. An automatic bolus-triggering method was applied with a circular region of interest (ROI) that was positioned at the level of the ascending aorta with a trigger threshold for data acquisition preset at 100 HU. On triggering of the threshold, a volume data set was subsequently acquired (16 x 0.75 mm collimation, 420-milli second gantry rotation time, 3.4-mm-per-rotation table feed, 120-kV tube voltage, and 550-mA tube current).
Axial images were reconstructed with retrospective ECG gating. Initially, one data set was reconstructed by setting the reconstruction window to an arbitrary level of 65% before the onset of the following R wave. If motion artifacts were present, image reconstruction was repeated with the reconstruction window offset by 5% toward either the beginning or the end of the cardiac cycle. Images were reconstructed with a 1-mm slice thickness at increments of 0.5 mm.
Myocardial Fat Analysis
Axial images and reformatted short-axis and vertical long-axis views were
used to detect and evaluate myocardial fat. The presence of myocardial fat was
concluded when two radiologists with 7 and 4 years of experience in cardiac
CT, respectively, agreed on the detection of focal low attenuation within the
myocardium on unenhanced CT analysis. Although there was no pathologic
confirmation of myocardial fat in these patients, we assumed that the lesion
that apparently showed low attenuation on unenhanced CT contained fat
components. After initial identification, we analyzed the attenuation,
distribution, and transmural extent of the myocardial fat for each case.
During unenhanced CT analysis, two independent ROIs for the regional low-attenuation area and remote myocardium were drawn by two radiologists within an area of the mid slice that they agreed by consensus showed the largest area of low attenuation (Fig. 1A, 1B). The transitional zone between the low-attenuation and normal myocardium was excluded from the ROIs. The sizes of the ROIs ranged from 7 to 55 mm2 (mean ± SD, 15.3 ± 9.0 mm2). For each patient, three ROIs were repeatedly drawn for a single region of the myocardial fat and a mean CT attenuation (in Hounsfield units) for each region of myocardial fat was calculated.
|
|
Myocardial fat distribution was determined by analyzing the territories of corresponding major coronary arteries, including the left anterior descending (LAD) artery, right coronary artery (RCA), and left circumflex artery (LCX), all of which were based on coronary angiographic findings.
The transmural extent of myocardial fat was estimated by dividing the fat thickness by the wall thickness in the middle of the slice that had the largest area of low attenuation in the short-axis view for all segments except the apex, for which the vertical long-axis view was used. The transmural extent of myocardial fat was graded as 1, 2, 3, or 4 based on its occupation of 0–25%, 26–50%, 51–75%, or 76–100% of the myocardium, respectively.
We also compared the following clinical information of patients with and those without myocardial fat: first, the levels of maximal cardiac biomarkers CKMB and troponin T based on laboratory tests; second, the presence of ST-segment elevation or Q waves on ECG; and, third, the degree of culprit artery stenosis and number of diseased vessels present on conventional coronary angiography on diagnosis of MI. We also com pared the time elapsed since the onset of MI and follow-up echocardiography in terms of regional wall motion abnormalities, ejection fraction, and left ventricular end-diastolic diameter. Regional wall motion abnormalities were graded as 0 for normal, 1 for mild to moderate hypokinesia, 2 for severe hypokinesia, 3 for akinesia, and 4 for dyskinesia. In addition, the results of follow-up SPECT performed with 99mTc-methoxyisobutyl isonitrile were compared between the two groups.
Statistical Analysis
Comparisons between the two groups with respect to the continuous data were
analyzed using independent two-sample Student's t tests. Categoric
variables were compared using the chi-square test or Fisher's exact test.
Logistic regression was performed to identify independent predictors of
myocardial fat accumulation after MI. A p value of less than 0.05 was
considered statistically significant.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
We also performed territory-based analysis: 27 patients had myocardial fat (30.3% of the total number of LAD territories with MI) in the LAD territory and nine (10.5% of the total RCA territories with MI) had myocardial fat in the RCA territory, whereas there was no myocardial fat in the LCX territory. The differences in MI territories between the patients with and those without myocardial fat were statistically significant (p = 0.0108).
In all cases, myocardial fat was located in the subendocardial layer. The transmural extension of myocardial fat within the MI territories was 0–25% in five lesions (13.9%), 25–50% in 25 lesions (69.4%), and 50–75% in six lesions (16.7%). There were no cases of myocardial fat in which its transmural extent was more than 75% of the myocardial thickness.
Comparison of Patients With and Those Without Myocardial Fat
Our analysis included 36 patients who had myocardial fat in the infarction
territory and 125 patients who did not; the demographic data and examination
results are summarized in Table
1. The time interval between the diagnosis of MI and CT was
significantly greater in the group with myocardial fat than in the group
without myocardial fat (5.6 ± 2.1 vs 2.4 ± 3.6 years,
respectively).
|
The incidence of myocardial fat increased according to the postinfarct elapsed time: 9.1% (7/77) within the first year after infarction, 20.7% (6/29) between the second and third years, 37.0% (10/27) between the fourth and seventh years, and 46.4% (13/28) thereafter. Culprit artery stenosis was milder in patients with myocardial fat than in patients without myocardial fat. In addition, patients with myocardial fat had fewer diseased vessels than patients without myocardial fat. In terms of previous treatments, patients with myocardial fat had more frequently undergone CABG than stent insertion at the vessels deemed necessary for supplying fat-containing territories. However, the proportion of stent insertion was greater in patients with myocardial fat than in patients without myocardial fat.
The presence of myocardial fat did not correlate significantly with demographic factors including sex and age, risk factors for cardiovascular disease (hypertension, diabetes mellitus, and hyperlipidemia), infarct age, the presence of ST elevation or Q wave, or elevated CKMB or troponin T levels after the initial diagnosis.
An echocardiography examination was performed within 30 days of obtaining the CT scans for 17 patients with myocardial fat and 64 patients without myocardial fat. The proportion of patients with akinesia was significantly larger in the patient group with myocardial fat than the group without myocardial fat (p = 0.0044), whereas there were no significant differences between the two groups with respect to the ejection fraction or left ventricular end-diastolic diameter. One case each of ventricular tachycardia and atrial fibrillation was noted among the patients with myocardial fat, whereas two cases each of atrial fibrillations and ventricular fibrillations were detected in patients without myocardial fat. On the basis of these findings, we concluded that arrhythmogenic potency related to myocardial fat was not clear in this study.
Cardiac SPECT was performed for 34 patients, eight of whom had myocardial fat and 26 who did not have myocardial fat. Four of the eight patients (50%) with myocardial fat had a fixed perfusion defect on SPECT, whereas only eight of the 26 patients (30.8%) without myocardial fat had a similar finding; however, this difference was not statistically significant (p = 0.42).
Logistic regression analysis (n = 141) revealed that the degree of culprit artery stenosis was the only significant independent predictor of myocardial fat (p = 0.0405).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We detected myocardial fat in 22.4% of patients with a previous MI. In an autopsy study, Baroldi et al. [3] observed "lipomatous metaplasia" in 68% of cases of ischemic heart disease. In another study, myocardial fat was identified in 84% of healed MIs in 91 explanted hearts for heart transplantation; however, only 22% of the healed MI lesions were composed of more than 25% fat [4], which is similar to the prevalence of myocardial fat observed in our study population as detected on CT scans. We assume that a small amount of microscopic myocardial fat that is undetectable by CT might also be in the infarcted myocardium.
The mean attenuation value of myocardial fat was –29.6 HU (range, –90.5 to 8.8 HU) on unenhanced CT and –11.2 HU (range, –75.6 to 33.7 HU) on contrast-enhanced CT. All lesions with myocardial fat except three had negative mean attenuation values; however, these three lesions exhibited a relatively high mean attenuation (range, 7.7–8.8 HU) and also showed minimum attenuation values of less than –19 HU on unenhanced CT. We used the largest ROI as long as it revealed low attenuation; thus, the three lesions discussed might have comprised a mixture of scar tissue and fat rather than pure fat. Although perfusion defects due to residual myocardial ischemia may also be seen as low-attenuating areas on contrast-enhanced CT, they are not similarly observed on unenhanced CT. In addition, the prevalence of reversible perfusion defects on SPECT was not significantly different between the two groups (p = 0.42). Therefore, we assume that residual ischemia at the sites of the previous MIs likely did not affect the detection of fat on unenhanced CT and did not significantly influence our analysis of the differences between the two groups.
Although several cases of myocardial fat related to MI have been reported, the pathogenesis of myocardial fat is uncertain. The results of experimental animal studies have shown that myocardial fat in MI may result from the inability of ischemic myocytes to metabolize free fatty acids, which play a role in normal myocardium nutrition [10]. Indeed, impaired metabolism results in the accumulation of circulating fatty acids in the border zones of infarcted myocardium. In our study, myocardial fat was associated with milder coronary artery stenosis and fewer diseased vessels. In addition, the results of logistic analysis showed that a milder degree of culprit artery stenosis was the only independent predictor of myocardial fat. In previous studies, investigators have reported that surviving cells in the border zone of infarcted myocardium might be degenerated with fat that is accumulated by relative ischemia [11, 12]. Thus, blood flow through vessels with mild-degree coronary artery stenosis, as well as collateral vessels, might play a role in the accumulation of fat at the infarcted myocardium. Perfusion to the lesion, which is responsible for delivery of circulating fatty acids to the border zones of infarcted myocardium, may predispose individuals to this condition.
Regional wall motion abnormalities were observed more frequently on follow-up echocardiography in patients with myocardial fat than in those without myocardial fat. ST elevation was noted in 78.3% (18/23) of patients with myocardial fat and in 56.2% (59/105) of patients without myocardial fat, and although this difference did not reach statistical significance (p = 0.0502), this may have been because our study lacked sufficient statistical power.
Recently, Jacobi et al. [13] reported that fat in the left ventricle detected on unenhanced CT is frequently related to prior MI, which had been already reported in previous studies [4–8]. However, the prevalence of myocardial fat in the patients with MI could not be assessed because only the patients (n = 26) with ventricular myocardial fat were included in their study. In another study, Zafar et al. [14] reported that left ventricular myocardial fat was found in 51% of patients (n = 47) with typical imaging findings of chronic MI such as myocardial thinning or calcification, which might have overestimated the frequency of myocardial fat in patients with previous MI. On the other hand, we used clinical criteria of MI, which enables us to evaluate the prevalence of myocardial fat in MI. In these two studies a nongated chest CT protocol was used in most of the patients, some of whom underwent unenhanced CT only or contrast-enhanced CT only, whereas we used unenhanced CT and contrast-enhanced CT with ECG gating in all patients. In addition, only our study analyzed the correlation between the clinical parameters of MI and myocardial fat on CT in a detailed and practical manner.
Our study design had a few limitations. First, there were no histologically confirmed cases of myocardial fat. Furthermore, because this review was retrospective, the clinical information was collected from documented charts that were often incomplete beyond the information relevant to the criteria used to diagnose MI. The patient group was heterogeneous in terms of the treatment methods that were provided for MI. In addition, many patients had received a CABG, which made it difficult to predict the relationship between previous treatments and fat deposition. With the increased use of and additional technical advances in cardiac MRI, the clinical implications of the presence of myocardial fat may be better assessed in the future by evaluating the degree of microvascular obstruction in the acute phase, transmural extent of infarction, and residual ischemia.
In summary, myocardial fat was detected in the infarcted myocardium of 22.4% of patients with MI. Myocardial fat was more frequently associated with a longer postinfarct elapsed time, milder coronary artery stenosis, fewer number of diseased vessels, and more severe regional wall motion abnormalities. Further investigation to determine the clinical significance of myocardial fat findings in MI as a predictor for prognosis and ventricular wall remodeling is warranted.
Acknowledgments
We thank Eun Hee Choi, Department of Biostatistics, Severance Hospital, for
statistical consultation and analysis of the data.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
