Author + information
- Received August 12, 2008
- Revision received January 12, 2009
- Accepted January 15, 2009
- Published online May 1, 2009.
- Yasutaka Ichikawa, MD⁎,⁎ (, )
- Kakuya Kitagawa, MD‡,
- Shuji Chino, MD⁎,
- Masaki Ishida, MD‡,
- Koji Matsuoka, MD†,
- Takashi Tanigawa, MD†,
- Tomoaki Nakamura, MD†,
- Tadanori Hirano, MD⁎,
- Kan Takeda, MD‡ and
- Hajime Sakuma, MD‡
- ↵⁎Reprint requests and correspondence:
Dr. Yasutaka Ichikawa, Department of Radiology, Matsusaka Central Hospital, 102 Kobou, Kawai, Matsusaka, Mie 515-8566, Japan
Objectives Our aim was to investigate the frequency of left ventricular (LV) and right ventricular adipose tissue on multislice computed tomography (CT) in patients with a history of myocardial infarction (MI) and to determine correlations with infarct age.
Background Fat deposition in the ventricular wall has frequently been observed in post-infarct myocardial tissue. However, the in vivo relevance of adipose tissue in MI on CT and correlations with infarct age have not been determined.
Methods Fifty-three patients with a history of MI (mean age 66 ± 10 years; 38 men, 15 women) and 63 subjects with no history of MI or coronary revascularization (mean age 65 ± 12 years; 37 men, 26 women) were retrospectively studied for intramyocardial fat on 64-slice cardiac CT. Presence or absence, distribution, and correlations with infarct age of LV adipose tissue were evaluated.
Results Compared with noninfarct control subjects, the MI group showed a significantly higher prevalence of fat deposition within LV myocardium on CT (MI group, 62% [33 of 53] vs. control group, 3% [2 of 63]; p < 0.0001). In 32 of 33 patients (97%) with MI and LV fat deposition on CT, adipose tissue was observed in the region perfused by the infarct-related artery and was located in the subendocardium in 30 patients (94%), the middle layer in 1 patient (3%), and the subepicardium in 1 patient (3%). Mean infarct age was significantly higher in patients with LV adipose tissue (8.2 ± 4.4 years) than in those without adipose tissue (2.2 ± 2.6 years, p < 0.001). Thirty of 35 patients (89%) with infarct age ≥3 years showed adipose tissue in MI. Conversely, none of 9 patients with infarct age <10 months showed fatty replacement.
Conclusions Myocardial adipose tissue is common in patients with infarct age ≥3 years. CT evaluation of myocardial adipose tissue may be important for accurate interpretation of CT perfusion and infarct imaging of the heart.
The presence of adipose tissue in the ventricular wall has been well recognized as a feature of arrhythmogenic right ventricular (RV) dysplasia (1–3). However, the pathology literature indicates that fat deposition in the ventricular wall is seen more frequently in post-infarction myocardial tissue (4). A previous study that evaluated histological sections of explanted hearts from 91 patients undergoing heart transplantation due to ischemic heart disease found adipose tissue in 84% of healed myocardial infarctions (MIs) (5).
Recent progress in multislice computed tomography (CT) technology has greatly improved the ability of CT to visualize the heart and coronary arteries. With higher spatial and temporal resolution as compared with that seen with conventional single-slice CT, coronary artery stenoses can be reliably ruled out by multislice CT (6–9). Moreover, multislice CT can be used to evaluate left ventricular (LV) function and myocardial tissue component (10–17). The purpose of this study was to investigate the frequency of LV adipose tissue as detected by multislice CT in patients with a history of MI and to determine relationships with infarct age.
Subjects for this retrospective study comprised 53 patients with a history of MI (mean age 66 ± 10 years; 38 men, 15 women) and 63 control subjects with no history of MI or coronary revascularization (mean age 65 ± 12 years; 37 men, 26 women). In all patients, cardiac CT studies were performed due to chest pain or equivocal stress test, and to exclude significant coronary artery disease from December 2005 to April 2007. Subject characteristics are summarized in Table 1. A diagnosis of MI was made based on the presence of chest pain lasting ≥20 min associated with changes on electrocardiogram (ECG) (ST-segment elevation or depression, pathologic Q waves, new onset of left bundle branch block) and abnormal levels of cardiac enzymes. Mean infarct age was 5.8 ± 4.8 years (range 1.5 months to 16.6 years). All subjects provided informed consent to the procedures, and the study protocols were approved by the internal review board before initiation of the study.
Cardiac CT was performed using a 0.625 mm × 64-detector scanner (Light Speed VCT, General Electric Medical Systems, Milwaukee, Wisconsin). Oral beta-blocker medication was used if heart rate was >65 beats/min (metoprolol 20 to 60 mg, 1 h before examination). Isosorbide dinitrate (5 mg) was administered sublingually before cardiac CT acquisition in all patients. Patients were scanned in a supine position, and ECG was monitored during scanning. Cardiac CT was performed twice: first without contrast medium, then with contrast medium for coronary angiography. Pre-contrast CT was originally performed to determine coronary calcium score. The scanning protocol for pre-contrast CT was a retrospective ECG-gated helical scan with low tube current (150 mA) (18). Other scan parameters were gantry rotation time, 350 ms; pitch, 0.18; tube voltage, 120 kV; and scanning field of view, 250 mm. Contrast transit time was determined using a test bolus injection. For post-contrast cardiac CT, 65 to 85 ml of iohexol (350 mg/ml) was injected into the antecubital vein at an injection speed of 4 ml/s, followed by a 20-ml saline bolus (injection rate 4 ml/s). Post-contrast cardiac CT acquisition started 20 to 35 s after starting contrast agent injection. Post-contrast cardiac CT was obtained with retrospective gating and scan parameters as follows: gantry rotation time, 350 ms; pitch, 0.24; tube voltage, 120 kV; tube current, 680 to 750 mA; and scanning field of view, 180 mm. Pre-contrast CT images were reconstructed with a slice thickness of 2.5 mm at cardiac phase of 75% for evaluation. Post-contrast CT images were reconstructed with a slice thickness of 0.625 mm at cardiac phase of 75% for evaluation. The effective radiation dose was estimated using the method proposed by the European Working Group for Guidelines on Quality Criteria in CT as derived from the product of the dose-length product and a conversion coefficient for the chest as the investigated anatomic region (k = 0.017 mSv·mGy−1·cm−1) averaged between male and female models (19). Dose-length product was obtained from the protocol summarizing each CT. In our study, the estimated effective radiation dose was 6.3 ± 0.7 mSv for pre-contrast CT and 23.7 ± 5.0 mSv for post-contrast CT.
CT image analysis
All CT images were transferred to an image analysis workstation (Advanced Workstation 4.2, General Electric Medical Systems), then axial images and multiplanar reformation images were analyzed by consensus of 2 observers blinded to clinical information. The observers first identified the presence or absence of adipose tissue in LV and RV myocardium on pre-contrast CT images by visually comparing myocardial CT density with that of subcutaneous fat. A region of interest was then placed by manually tracing the border of the LV adipose tissue to measure the CT attenuation value and volume of this region. The CT attenuation value of remote LV myocardium was also measured. In addition, distribution of LV adipose tissue within the LV wall on pre-contrast CT was determined by referring to post-contrast CT images, and was classified as subendocardium, middle layer, or subepicardium.
Statistical analyses were performed using SPSS 11.5J statistical analysis software (SPSS Inc., Chicago, Illinois). Continuous variables are presented as the mean ± SD. Differences between patient groups were tested using Student t test or the Mann-Whitney U test, as appropriate. The chi-square test was performed for noncontinuous variables where appropriate. All tests were 2-tailed, and values of p < 0.05 were considered statistically significant.
Adipose tissue within LV myocardium was detected in 33 of 53 patients (62%) with a history of MI on multislice CT images (Figs. 1 to 3).⇓⇓⇓ Mean CT attenuation value of LV adipose tissue (−33 ± 24 Hounsfield units [HU]; range −82 to −1 HU) was significantly lower than the CT attenuation value of remote myocardium on pre-contrast images without overlap (34 ± 7 HU; range 23 to 51 HU; p < 0.001), with a mean difference of 67 ± 23 HU. CT attenuation value of LV adipose tissue on post-contrast CT was −30 ± 28 HU, showing no significant difference from that on pre-contrast CT (p = 0.57). Mean volume of LV adipose tissue as determined by pre-contrast multislice CT was 2.4 ± 3.6 cm3 (range 0.15 to 17.43 cm3). In 32 of the 33 patients (97%) with MI and LV adipose tissue on CT, adipose tissue was located in the region perfused by the infarct-related artery as documented in the medical record. In these 32 patients, LV adipose tissue was seen in the subendocardium in 30 patients (94%), middle layer in 1 patient (3%), and subepicardium in 1 patient (3%) (Table 2). In the remaining 1 patient who had infarct in the inferior wall, adipose tissue was observed in the subepicardium of the anteroseptal wall. The infarct group showed a significantly higher prevalence of fat deposition within LV myocardium on CT (62%, 33 of 53) compared with the noninfarct control group (3%, 2 of 63; p < 0.0001). In both control subjects with LV fat deposition, adipose tissue was observed in the apical segment.
No difference in the presence of LV adipose tissue was seen according to gender or age of patients, infarct territory, coronary risk factors, or presence or absence of abnormal Q waves on ECG (Table 3). Infarct age of MI with LV adipose tissue (8.2 ± 4.4 years; range 10 months to 16.6 years) was significantly higher than that of MI without adipose tissue (2.2 ± 2.6 years; range 1.5 months to 9.8 years; p < 0.001) (Fig. 4).Figure 5 shows the relationship between presence of LV adipose tissue in MI and infarct age. None of the 9 subjects who had MI within 10 months before multislice CT study showed fat deposition in the corresponding infarct area. LV adipose tissue was observed in 2 of 9 patients (22%) with infarct age of 10 months to 3 years, in 13 of 16 patients (81%) with infarct age of 3 to 6 years, and in 17 of 19 patients (89%) with infarct age ≥6 years. In 2 of all 53 patients (4%), calcification was detected in the subendocardium of the corresponding infarct area. In these patients with myocardial calcification, adipose tissue was located adjacent to myocardial calcification on CT. Infarct age for these 2 patients was 9.9 and 13.3 years. One patient with anteroseptal MI and a history of ventricular tachyarrhythmias after MI showed adipose tissue in the subendocardium of the anteroseptal wall.
No significant difference in the prevalence of RV fat deposition was seen between the MI group (34%, 18 of 53) and control subjects (29%, 18 of 63; p = 0.53). In the MI group, mean age was 72.7 ± 6.6 years for patients with RV fat deposition, compared with a mean age of 62.3 ± 10.4 years for patients without (p < 0.001). Mean age was 71.2 ± 5.5 years for control subjects with RV fat deposition, compared with a mean age of 61.8 ± 12.3 years for control subjects without (p < 0.01). LV fatty replacement in the infarct area was observed in 12 of 18 patients (67%) with RV adipose tissue and in 20 of 35 patients (57%) without, showing no significant association between presence of RV and LV adipose tissue after MI (p = 0.50).
The current study found that adipose tissue was present in the LV wall for 33 of 53 MI patients (62%) and identified a significant relationship to infarct age in patients with old MI, particularly among patients with infarct age ≥3 years.
An autopsy study by Baroldi et al. (4) frequently observed mature adipose tissue in LV myocardial segments from patients with a history of MI. A recent ex vivo histological study by Su et al. (5) confirmed a high prevalence (84%) of adipose tissue in healed MI. The presence of fat has been identified not only ex vivo at the microscopic level, but also in vivo at the macroscopic level by CT. A previous case report described CT demonstration of fatty changes in LV myocardium for 4 patients with a history of MI (20). Jacobi et al. (21) reported myocardial fat corresponding to the site of prior MI in 8 of 9 patients (89%) on pre-contrast CT. A study by Zafar et al. (22) reported that the frequency of LV myocardial fat in 47 patients with CT evidence of old MI was 51%. Fat in LV myocardium is considered a common finding in patients with old MI on cardiac CT.
LV fat after MI and correlations with infarct age have been described in a magnetic resonance imaging study reported by Goldfarb et al. (23). That study used T1-weighted imaging to detect fat deposition in MI, revealing that the presence of LV fat in MI is associated with infarct age (23). LV adipose tissue was observed in 18 of 23 patients (78%) with MI with infarct age ≥6 months, while no LV adipose tissue was found in any of the 7 patients with recent MI (infarct age <6 months) (23). These results are in line with our study performed using a 64-slice CT scanner. In the current study, LV adipose tissue was observed in 32 of 44 MI patients (73%) with infarct age ≥10 months, while no LV adipose tissue was observed in any of the 9 patients with infarct age <10 months. Although Su et al. (5) reported that the frequency of adipose tissue in histopathological specimens was increased in elderly patients and male patients, no difference in the presence of LV adipose tissue according to age or sex was apparent in our study. This discrepancy may be explained by the much lower spatial resolution of CT compared with histopathological examination.
As speculated by Goldfarb et al. (23), infarcted myocardium seems to be replaced by fibrous scar tissue first, and then populated with adipocytes. However, the exact mechanism underlying the formation of mature adipose tissue in MI scar remains unclear. Baroldi et al. (4) hypothesized that reduced myocardial contractility is involved in the pathogenesis of adipose tissue in healed MI. Su et al. (5) suggested that modern therapy for coronary artery disease creates conditions that promote transdifferentiation of cells to adipocytes in MI scar, since adipose tissue in healed MI was not described until 1997. However, the present study found no significant difference in the prevalence of LV adipose tissue between reperfused and nonreperfused infarcts.
The clinical significance of intramyocardial LV fat after infarction has yet to be established. A previous report suggested that the replacement of myocardium by fat may change the structural integrity of the heart, affecting the remodeling process, regional ventricular mechanics, and global ventricular function (23). Fat may interfere with the conduction system of the ventricle, causing arrhythmias and possible sudden death (24). From the perspective of cardiac imaging, recognition of LV adipose tissue on pre-contrast chest CT may be the first opportunity to diagnose silent MI. In addition, LV adipose tissue may be used as a hallmark of old MI. Furthermore, the presence of fat in areas of scarred myocardium may affect interpretation of CT myocardial perfusion imaging and late contrast-enhanced infarct imaging of the heart. For example, CT attenuation value of LV adipose tissue in MI on post-contrast CT was −30 ± 28 HU in this study. Since this value is less than that of water (i.e., <0 HU), old MI may be visualized as a substantially lower attenuation area in comparison with ischemic myocardium on CT myocardial perfusion imaging.
A previous study (25) demonstrated an association between age and fatty replacement in the RV free wall. Electron beam CT was used to detect RV fat infiltration in asymptomatic patients, revealing RV fat deposition as a common incidental finding in up to 17% of patients and increasing prevalence with advancing age. Kim et al. (26) reported that RV fat infiltration occurred in about 17% of asymptomatic subjects on CT, with more severe findings in elderly subjects. In the current study, fat infiltration in the RV free wall was detected in approximately 30% of subjects and was frequently observed in elderly patients.
The first major limitation of this study was the lack of pathological validation for LV adipose tissue detected on CT. Ideally, CT findings would be correlated to histological specimens, but such examinations are extremely difficult to perform in vivo in human hearts as opposed to experimental animal models (13). Second, since CT cannot reliably depict myocardial fibrosis after infarction, we could not compare the size of myocardial fat detected on CT with that of fibrous tissue after MI. Third, inter- and intraobserver variabilities were not evaluated in this study, as all images were interpreted concurrently by 2 observers. Fourth, the number of patients with MI in this study was relatively small. Further prospective studies including a larger cohort of patients are required to confirm the present findings. Lastly, although a wide range of infarct ages was included in this study, other aspects of patient selection such as the relatively good health of participating subjects may have influenced the results.
In summary, adipose tissue is frequently observed in the subendocardial LV wall of the region perfused by the infarct-related artery on multislice CT. LV adipose tissue showed a significant association with infarct age in patients with a history of MI, particularly with infarct age ≥3 years. Since recognition of such adipose tissue may affect the interpretation of CT perfusion imaging and late contrast-enhanced infarct imaging of the heart, evaluation of myocardium on pre-contrast CT may be important. Further studies to investigate associations between fat deposition and structural, mechanical, and functional properties of the myocardium are warranted.
- Abbreviations and Acronyms
- computed tomography
- left ventricle/ventricular
- myocardial infarction
- right ventricle/ventricular
- Received August 12, 2008.
- Revision received January 12, 2009.
- Accepted January 15, 2009.
- American College of Cardiology Foundation
- Basso C.,
- Ronco F.,
- Marcus F.,
- et al.
- Tandri H.,
- Castillo E.,
- Ferrari V.A.,
- et al.
- Hamon M.,
- Biondi-Zoccai G.G.,
- Malagutti P.,
- et al.
- Hausleiter J.,
- Meyer T.,
- Hadamitzky M.,
- et al.
- Budoff M.J.,
- Dowe D.,
- Jollis J.G.,
- et al.
- Lardo A.C.,
- Cordeiro M.A.,
- Silva C.,
- et al.
- Habis M.,
- Capderou A.,
- Ghostine S.,
- et al.
- Bongartz G.,
- Golding S.J.,
- Jurik A.G.
- Lie J.T.