- Email: [email protected]
Multidetector computed tomography shows intramyocardial fat deposition
Journal of Cardiovascular Computed Tomography (2008) 2, 152–163
Original Research Article
Multidetector computed tomography shows intramyocardial fat deposition Aidan R. Raney, MD, Farhood Saremi, MD*, Satish Kenchaiah, MD, Swaminatha V. Gurudevan, MD, Jagat Narula, MD, Navneet Narula, MD, Stephanie Channual, BS Department of Radiologic Sciences, University of California, Irvine, UCI Medical Center, 101 The City Drive, Route 140, Orange, CA 92868-3298, USA KEYWORDS: ARVD, arrhythmogenic right ventricular dysplasia; Myocardial fat; Myocardial infarction
BACKGROUND: Intramyocardial fat deposition occurs as an age-related process and in multiple pathologic processes. OBJECTIVE: We evaluated the presence of left ventricular (LV) and right ventricular (RV) intramyocardial fat with 64-slice multidetector computed tomography (MDCT). METHODS: One hundred persons with no history of coronary artery disease (47 women, 53 men; mean age [⫾ SD], 53 ⫾ 12.2 years) and 25 patients with CT findings of myocardial infarction (17 men, 8 women; mean age, 71.3 ⫾ 9.6 years) were studied for intramyocardial fat in defined segments of the ventricles (17 LV and 10 RV segments) at 3 levels. Fat deposition was defined as density range of ⫺30 to ⫺190 Hounsfield units on images both before and after contrast. RESULTS: In healthy persons, LV intramyocardial fat was primarily located in the basal segments (5% anteroseptal, 5% inferior), and RV intramyocardial fat was primarily located in the anterolateral (24% of base, 23% of mid) and inferolateral (27% base, 27% mid) segments. Older age was associated with an increased odds of RV (sex-adjusted odds ratio [OR] per decade increment, 1.61; 95% confidence interval [CI], 1.11–2.33; P ⫽ 0.012) but not LV (OR, 0.97; 95% CI, 0.67–1.40; P ⫽ 0.85) intramyocardial fat. Compared with women, men had a lower risk of LV (95% CI, 0.1– 0.64; P ⫽ 0.004) but not RV (95% CI, 0.35–1.87; P ⫽ 0.62) intramyocardial fat. Patients with old myocardial infarction (⬎3 years) had increased percentage of fat in infarcted left ventricles at all 3 levels (P ⱕ 0.004). CONCLUSIONS: Intramyocardial fat can be detected by MDCT and is common in healthy and infarcted myocardium. © 2008 Society of Cardiovascular Computed Tomography. All rights reserved.
Introduction Intramyocardial fat deposition occurs as an age-related process1,2 and in multiple pathologic processes, including Conflict of interest: The authors report no conflicts of interest. Presented at the 56th annual scientific session of the American College of Cardiology, New Orleans, LA, March 24 –27, 2007. * Corresponding author. E-mail address: [email protected] Submitted August 29, 2007. Accepted for publication January 24, 2008.
arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C),3 myocardial infarction (MI),4,5 and dilated cardiomyopathy.5 Previous imaging studies of cardiac fat tissue have focused on pericardial and epicardial fat,6 and there was limited evaluation of adipose tissue within the myocardium itself because of limitations in spatial and temporal resolution. This is especially true in the right ventricle where the free wall thickness approaches a few millimeters. The introduction of multidetector computed tomography (MDCT) has made it possible to assess cardiac disease with
1934-5925/$ -see front matter © 2008 Society of Cardiovascular Computed Tomography. All rights reserved. doi:10.1016/j.jcct.2008.01.004
Raney et al
MDCT shows intramyocardial fat deposition
153
Figure 1 Differentiating artifact from fat deposition in a 49-year-old man with no history of heart disease. A review of consecutive images in different projections, and particularly in different phases, allows for verification of the presence of intramyocardial fat. (A) It is difficult to differentiate whether the lesion represents fat or artifact. Artifact-free images in systole at 30% of R-R interval (B) and late diastole at 80% of R-R interval (C) show the true presence of endocardial fat deposition in the anteroseptal wall of the left ventricle.
a high level of detail, including myocardial infarct and small atherosclerotic plaques.7-9 CT was also shown to be accurate in the assessment and quantification of fat tissue because of attenuation values distinct from those of other anatomical structures.10-15 The purpose of this study was to combine the improved spatial and temporal resolution of MDCT with the unique attenuation values of fat tissue to assess the spatial distribution and structural pattern of fat deposition within the myocardium. We evaluated intramyocardial fat deposition in the left and right ventricles of healthy persons with no clinical history or CT evidence of cardiac disease. In the left ventricle, we also included a group of patients with history of MI to show differences in the distribution of intramyocardial fat deposition with the healthy population.
Methods The study was conducted with the approval of the institutional review board at the University of California Irvine Medical Center and was compliant with Health Insurance Portability and Accountability Act regulations. Informed consent was not required for this retrospective analysis.
Patient selection To identify healthy persons, we reviewed 130 consecutive electrocardiogram (ECG)– gated coronary MDCT studies (64-slice scanner; Toshiba Aquilion, Tustin, CA) performed at our institution from December 2005 to January 2006. Patients completed a questionnaire, and patients with coronary artery disease (defined as a clinical history of significant coronary artery disease [CAD], coronary artery bypass graft surgery, or intracoronary stent placement) were excluded. We also excluded patients with previously unrec-
ognized CAD that was apparent on the evaluation of images. Studies were then evaluated in multiple views (short axis, long axis, and 3-dimensional) and were given a subjective score of good, average, and poor, based on the presence of metal artifact (pacemakers, sternal wires, etc), noise, calcification, streak artifact related to contrast-enhanced cardiac chambers, beam hardening that caused false attenuation, and motion artifact. After excluding poor quality studies, 100 studies with no history of CAD or infarct were reviewed for fat deposition of the right ventricular (RV) and the left ventricular (LV) myocardium. For the infarct group, we evaluated patients with a clinical history of MI by questionnaire who received cardiac CT angiography (CTA) between December 2005 and November 2006 (50 patients). We then selected the first 25 cases with characteristics of infarction on CTA (ie, hypokinesis, akinesis, and thinning of myocardium) who had fatty areas (⬍ ⫺30 Hounsfield units [HU]) within the infarcted segment.
Scan protocol and image reconstruction An ECG was completed, and vital signs were obtained for all patients on arrival at the imaging suite. For pa-
Table 1 Key patient characteristics for the healthy and infarct groups
Age (y), mean ⫾ SD Sex, % male (n) Weight (kg), mean ⫾ SD BMI (kg/m2), mean ⫾ SD BMI, body mass index.
Healthy group (n ⫽ 100)
Infarct group (n ⫽ 25)
53.0 ⫾ 12.2 53 (53) 167.9 ⫾ 34.0 26.1 ⫾ 3.6
71.3 ⫾ 9.6 68 (17) 172.5 ⫾ 42.2 27.9 ⫾ 6.9
154
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 2 (A) Segmental distribution of intramyocardial fat in the left ventricle of infarct (top) and healthy participants (middle), showing the percentage of each segment positive for fat deposition, based on the previously described 17-segment model.16 (B) Segmental distribution of intramyocardial fat in the right ventricle of healthy persons, showing demonstrating the percentage of each segment positive for fat deposition. (Note: description of segments is available in Table 2.)
tients with a heart rate greater than 65 beats/min, oral or intravenous metoprolol was used to achieve a heart rate of ⬍65 beats/min. All patients received sublingual nitroglycerin (0.4 – 0.8 mg) 1 minute before image acquisition, unless contraindicated. All scans were performed on a scanner with 64 parallel detector rows and an individual detector width of 0.5 mm. With the use of a prospectively ECG-gated (mid-diastole) technique, precontrast 3-mm nonspiral images were obtained from the level of tracheal bifurcation to the diaphragm to cover the entire heart. Contrast enhancement was achieved using 70 – 80 mL of iohexol (Omnipaque 350 mg/mL; Amersham Health, Cork, Ireland) injected at a rate of 4 –5 mL/sec, followed by an injection of 50 mL of saline at 5 mL/s through an 18-gauge catheter in an antecubital vein to allow washout of contrast from the right heart and superior vena cava. Scan parameters included 120 kVP tube voltage, 400 mA tube current, 64 ⫻ 0.5 mm collimation, 7.2 mm table feed per rotation, and a gantry rotation time of 400 milliseconds. Scan start was automatically initiated 4 seconds
after reaching a threshold of 180 HU in the descending aorta at the tracheal bifurcation level. A retrospective ECG-gated volumetric data set was acquired during a single breathhold. Depending on the heart rate throughout the examination, axial slices were reconstructed synchronized to the ECG by a nonsegmented (ⱕ65 beats/ min) or segmented (⬎65 beats/min) image reconstruction algorithm. When necessary, R-wave indicators were manually repositioned to improve the quality of synchronization. Diastolic axial images were reconstructed based on a relative-delay strategy at 70%, 75%, and 80% of R-R intervals, respectively. In case of persistent artifacts in the area of interest, a second reconstruction approach was performed, and systolic images were reconstructed (25%–35% R-R interval). To evaluate regional myocardial function, 2-mm slices were reconstructed at 10% R-R intervals. Axial slices with a thickness of 0.5 mm (increment, 0.3 mm) and a cardiac CTA algorithm were used for evaluation of the myocardium. The data set least affected by cardiac motion was transferred to an offline
Raney et al
MDCT shows intramyocardial fat deposition
Table 2 Segmental prevalence of intramyocardial fat deposition in the left and right ventricles Percentage of involved segments
Left ventricle segment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 — — Right ventricle segment 1 2 3 4 5 6 7 8 9 10 — — —
Corresponding area
Healthy population
Infarct group
Base anterior Base anteroseptal Base inferoseptal Base inferior Base inferolateral Base anterolateral Mid anterior Mid anteroseptal Mid inferoseptal Mid inferior Mid inferolateral Mid anterolateral Apical anterior Apical septal Apical inferior Apical lateral Apex Anterior papillary Posterior papillary
2 5 3 5 2 2 1 2 5 0 2 0 1 3 1 1 1 14 12
8 8 8 16 24 28 20 12 16 24 28 12 24 20 28 32 16 — —
Base outflow tract Base inferolateral Base inferoseptal Base anteroseptal Mid anterolateral Mid inferolateral Mid inferoseptal Mid anteroseptal Apical free Apical septal Anterior insertion Inferior insertion Moderator band
24 27 2 6 23 27 3 1 12 3 1 23 16
— — — — — — — — — — — — —
3D Vitrea workstation (Vital Images, Inc, Minnetonka, MN) for further analysis.
CT data analysis Multiplanar reformations of the axial images (short and long axis views) after contrast were rendered and evaluated for the presence of intramyocardial fat deposition by consensus of 2 authors (F.S. and S.G.) with 16 and 2 years of experience in CT interpretation, respectively. All measurements were performed by a resident (A.R.) with 1 year of experience in cardiac CT imaging and confirmed by author F.S. The right and left ventricles
155 were separated into segments and analyzed for the presence of fat tissue in the myocardium. Anatomic localization of fat was performed with high resolution postcontrast data and then confirmed by precontrast images. To differentiate low attenuation areas as intramyocardial fat deposition compared with artifacts (motion, beam hardening, or blooming), studies with uncertain findings were reviewed in at least 2 different reconstructions of the R-R interval of the cardiac cycle (Fig. 1). A segment was considered positive for fat deposition based on the presence of a linear lesion with threshold attenuation values between ⫺30 and ⫺190 HU on series both before and after contrast, a range used in previous studies of fat tissue.10-15 The maximum width of the visualized fatty density was measured at each segment. LV myocardial segments were analyzed based on a 17segment model of the left ventricle16 at 3 levels of the ventricle (base, midventricular, and apical). The presence or absence of fat in the papillary muscles was also evaluated. The right ventricle was evaluated in 10 segments at the same 3 levels. Base segments (4) included the anterolateral (pulmonary outflow tract), inferolateral, inferior septal, and anterior septal walls; the midventricular segments (4) included the anterolateral and inferolateral wall, inferior septal wall, and anterior septal wall; and the apical segments (2) included the free and septal walls. We assessed fat deposition at the anterior and posterior RV insertion (insertion of free wall of the right ventricle with the interventricular septum) and categorized these separately from the above segments. We also evaluated fat deposition in the moderator band and its septal band insertion. The relative location (endocardial side compared with epicardial side) of intramyocardial fat deposition in both groups was assessed. The maximum visible thickness of the RV wall was measured at 5 segments, corresponding to 3 levels of the ventricle, including segments 1 and 2 of the base, segments 5 and 6 of the mid-ventricle, and segment 9 of the apical right ventricle. The anatomic location of fat deposition along major coronary arteries in healthy and infarct groups was recorded. As per the 17-segment model,16 coronary distribution included segments 1, 2, 7, 8, 13, 14, and 17 for the left anterior descending artery; segments 5, 6, 11, 12, and 16 for the left circumflex artery; and segments 3, 4, 9, 10, and 15 for right coronary artery.
Statistical analysis We computed mean ⫾ standard deviation (SD) and ranges for continuous variables and proportions (expressed as percentages) for categorical variables. We evaluated fat as both dichotomous (presence or absence) and continuous (defined as number of segments involved in the 17-segment model of left ventricle and 10-segment model of right ventricle) variables. We used linear and logistic regression analyses, as appropriate, to examine the cross-sectional association of age and sex with intramyocardial fat in the right ventricle only and the left
156
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 3 Common locations of fat deposition in healthy persons at the base of the left ventricle, including the anteroseptal (A), inferoseptal (B), and inferior (C) segments with HU values at each location. Evaluation of images before and after contrast reduces the possibility that low HU lesions are due to artifact. Note that the values for precontrast images are generally lower than are postcontrast images and are most likely because of volume averaging of small areas of fat deposition with normal myocardium in thick noncontrast (3 mm) images compared with postcontrast images (0.5 mm). Intramyocardial fatty deposition typically involves the endocardial aspect of the myocardium, although we did observe intramyocardial fat in the inferior and inferolateral segments of the base of the left ventricle in 3 patients (C). Table 3 Results of logistic regression analyses examining the predictors of presence or absence of fat in myocardium of healthy participants. Univariate analyses Variables Left ventricle only (excluding papillary muscles) Age (per decade increase) Sex (male/female) Right ventricle only Age (per decade increase) Sex (male/female) OR, odds ratio; 95% CI, 95% confidence intervals.
Multivariable analyses
OR (95% CI)
P
OR (95% CI)
P
1.10 (0.78–1.56) 0.25 (0.10–0.64)
0.59 0.003
0.97 (0.67–1.40) 0.25 (0.10–0.64)
0.85 0.004
1.64 (1.14–2.36) 0.63 (0.28–1.39)
0.007 0.25
1.61 (1.11–2.33) 0.81 (0.35–1.87)
0.012 0.62
Raney et al
MDCT shows intramyocardial fat deposition
157
Figure 4 Fat deposition in infarcted myocardium. Images before and after contrast of the heart in the left circumflex (LCx; A) artery, left anterior descending (LAD; B) artery, and the right coronary artery (RCA; C) distribution of 3 different patients are shown. Fat deposition is typically subendocardial and large enough to cover a significant portion of infarcted myocardium. Associated findings included thinning of myocardium (A and B), abnormal myocardial function, and corresponding coronary artery disease (not shown but present in all persons). The findings help differentiate fat deposition of infarct scar from fat deposition as a normal variant.
ventricle only excluding papillary muscles. To compare the presence or absence and extent (number of involved myocardial segments) of intramyocardial fat among healthy participants and patients with MI, we constructed unadjusted and age- and sex-adjusted logistic and linear regression models, respectively. We expressed results of logistic regression analyses as odds ratios, 95% confidence intervals, and P values, and linear regression as P values. We considered a 2-sided P value of ⬍0.05 as statistically significant. We performed all analyses using SAS software version 9.1 (SAS Institute, Cary, NC).17
Results Patients The patients evaluated in the healthy group included 47 women and 53 men with an age of 53.0 ⫾ 12.2 years (Table 1). Twenty-one patients with CAD (including 6 by history and 15 after evaluation of images) and 9 of the studies with poor quality as defined above were excluded from the study. In the infarct group, all 25 cases of infarct were at least 3 years old by clinical history (7.6 ⫾ 3.7 years). The patients
158
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Table 4 Results of logistic regression analyses comparing presence or absence of fat in various regions of the left myocardium among healthy persons and patients with myocardial infarction Unadjusted
Left ventricular level Base Mid Apical Coronary distribution LAD LCx RCA
Age and sex-adjusted
Healthy, %
Infarct, %
OR (95%CI)
P
OR (95%CI)
P
20 10 6
64 60 60
6.0 (2.3–15.2) 15.2 (5.3–43.5) 36.0 (10.0–129.6)
0.0002 ⬍0.0001 ⬍0.0001
11.7 (3.2–43.0) 14.4 (3.6–57.7) 18.8 (4.0–87.9)
0.0002 0.0002 0.0002
15 14 7
52 68 48
5.3 (2.1–13.6) 15.5 (5.5–43.8) 12.3 (4.1–36.8)
0.0005 ⬍0.0001 ⬍0.0001
5.9 (1.7–19.9) 19.8 (4.9–79.7) 8.8 (2.2–34.8)
0.004 ⬍0.0001 0.002
OR, odds ratio; 95% CI, 95% confidence intervals; LAD, left anterior descending coronary artery; LCX, left circumflex coronary artery; RCA, right coronary artery. According to the 17-segment model,16 coronary distribution included segments 1, 2, 7, 8, 13, 14, and 17 for the LAD; segments 5, 6, 11, 12, and 16 for the LCx; and segments 3, 4, 9, 10, and 15 for the RCA.
included 17 men and 8 women with an age of 71.3 ⫾ 9.6 years. Ninety-two percent of these patients had a history of coronary intervention, defined as ⱖ1 of the following: angioplasty/stent (81%) or coronary artery bypass graft (52%). The mean (⫾SD) heart rate for all patients before data acquisition was 58 ⫾ 5 beats/min (range: 47– 68 beats/min), and the mean scan time was 9.1⫾1.4 seconds (range: 8 –13 seconds).
LV fat deposition A total of 1700 LV segments in 100 healthy persons and 425 LV segments in 25 patients with clinical history of MI were analyzed for the presence of intramyocardial fat deposition. Both groups contained intramyocardial fat; however, the structure, segmental location (Fig. 2; Table 2), and lowest Hounsfield units in the lesions were different in the healthy participants compared with patients with a history of MI. In the healthy group, LV intramyocardial fat was primarily located at the basal level, with ⱖ1 involved segments in 19% of patients. The anteroseptal (5% of patients) and inferior (5% of patients) segments were the most commonly involved individual segments in the basal region. Fat deposition was uncommon in the mid and apical levels of the left ventricle, with ⱖ1 involved segments seen in 10% and 6% of the healthy population, respectively. The average maximum width was 2.2 ⫾ 0.2 mm at the base, 1.8 ⫾ 0.2 mm at the midventricular level, and 2.1 ⫾ 1.0 mm at the apical level. All intramyocardial fat deposition was located on the endocardial aspect with a preserved layer of myocardium on the epicardial side of the myocardium (Fig. 3), with the exception of 3 lesions that had preserved myocardium on both sides (Fig. 3C) located in the inferior (2) and inferolateral (1) segments at the basal level. We commonly observed round foci of fat densities in LV papillary muscles and at their attachments to the chordae tendineae, which were evenly distributed anteriorly (14%) and posteriorly (12%) on postcontrast images. Fatty foci in
papillary muscles were seen in only 3 patients on images both before and after contrast. Compared with women, men had a lower risk of LV (age-adjusted odds ratio [OR], 0.25; 95% confidence interval [CI], 0.1– 0.64; P ⫽ 0.004) but not RV (age-adjusted OR, 0.81; 95% CI, 0.35–1.87; P ⫽ 0.62) intramyocardial fat (Table 3). The higher prevalence of fat deposition in ⱖ1 LV myocardial segments of female participants was seen at all 3 levels (unadjusted P ⫽ 0.015, and age-adjusted P ⫽ 0.022). The infarct group showed a higher prevalence of fat deposition of the myocardium at all levels of the left ventricle (Figs. 2 and 4; Table 4), with ⱖ1 affected segment in all but one study (96%). Compared with the healthy group, greater LV intramyocardial fat deposition in the infarct group was evident at each level of the ventricle, including 64% of basal, 60% of mid, and 60% of apical levels (Table 4). The most common individual segments with fat deposition included the inferior and inferolateral segments of the base;, the anterior, inferior, and inferolateral segments of the mid-ventricle; and all segments of the apical level of the ventricle. The location of fat deposition in the ventricular wall was primarily on the endocardial side in the infarct and healthy groups, although areas of transmural extension were only seen in the infarct group (Fig. 4). The lesion density was lower in the infarct group, in which the average of the lowest density in each lesion was ⫺127.3 ⫾ 55.4 HU (range, ⫺51 to ⫺291 HU) compared with ⫺77.2 ⫾ 32.3 HU (range, ⫺30 to ⫺180 HU) in the healthy group (unadjusted P ⬍ 0.0001, age-adjusted P ⫽ 0.0013) on postcontrast images. The average density was lower in precontrast images than in postcontrast images ion both the infarct and noninfarct groups (Figs. 3 and 4).
RV fat deposition A total of 1000 RV myocardial segments were assessed in the healthy population. One or more segments positive for fat deposition were observed in 36% of basal,
Raney et al
MDCT shows intramyocardial fat deposition
159
Figure 5 Short axis views showing typical anatomic distribution of myocardial fat in the right ventricle of 4 healthy persons. Note the deposition of fat on the endocardial aspect of the myocardium with preserved myocardium on the epicardial side of the pulmonary outflow tract (1), lateral free wall at the base (2), inferior free wall (3), septal and moderator bands (4), and insertion point of the right ventricle with the septum (5). LV indicates left ventricle.
34% of mid, and 13% of apical levels of the right ventricle. The most commonly involved individual segments of the right ventricle were the anterolateral (pulmonary outflow tract; 24%) and inferolateral (27%) segments of the base and the anterolateral (23%) and inferolateral (27%) segments of the mid-ventricle (Fig. 2). Involvement of the RV septal wall was less common, with the basal anteroseptal segment seen as the most commonly affected septal segment (6%). The presence of fat in the moderator band was seen in 16% of patients, and fat deposition at the inferior septal insertion of the RV free wall was seen in 23 patients (23%), whereas it was only detected anteriorly in 1 patient (1%). The structure of RV intramyocardial fat deposition frequently showed the spread of fat deposition beginning on the endocardial side
of the RV wall and extending outward toward the epicardial side of the free wall, with a thin layer of preserved myocardium on the subepicardial aspect (Fig. 5). The average maximum width of the measurable fatty area was 2.7 ⫾ 1.6 mm at the base, 2.4 ⫾ 1.4 mm at the midventricular level, and 2.0 ⫾ 1.2 mm at the apical level. The maximum visible thickness of the RV wall was 4.0 ⫾ 0.86 mm in segment 1, 3.4 ⫾ 0.71 mm in segment 2, 4.3 ⫾ 1.01 mm in segment 5, 3.4 ⫾ 0.78 mm in segment 6, and 3.6 ⫾ 0.95 mm in segment 9 of the right ventricle. Older age (per decade increase) was associated with an increased odds of RV (sex-adjusted OR per decade increment, 1.61; 95% CI, 1.11–2.33; P ⫽ 0.012) but not LV (sex-adjusted OR, 0.97; 95% CI, 0.67–1.40; P ⫽ 0.85) intramyocardial fat (Table 3).
160
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 6 Fat deposition of infarcted (A) compared with healthy myocardium (B). Microphotograph of a myocardial specimen (original magnification, ⫻2; hematoxylin-eosin stain) from a 67-year-old man with severe triple vessel coronary artery disease showing fat deposition in an area of healed infarct (long arrows). This patient had multiple areas of patchy replacement fibrosis (arrowheads) in the distribution of the left anterior descending artery (not included in present study). Some of these foci show entrapped fat. (B) Photomicrograph of myocardial specimen (original magnification, ⫻4) from a remote area shows the presence of fat in the interstitial area (arrows) between cardiomyocytes without accompanying interstitial or replacement fibrosis.
Discussion In this study we evaluated fat deposition within the RV and LV myocardium of healthy persons (without CAD) by MDCT. To provide clinical utility for this data, we then compared our findings with the primary differential for fat deposition in each ventricle. This included a separate group of our patients with MI for the left ventricle and reported pathologic studies of ARVD/C for comparison of RV intramyocardial fat.
LV fat deposition On evaluation of LV intramyocardial fat, we found that women had a significantly higher number of LV segments with fat deposition than did the male patients. The presence of LV intramyocardial fat was briefly described in pathology reports,6 and, although RV intramyocardial fat was correlated with female sex,1 we did not encounter any reports comparing age or sex with fat deposition in the LV myocardium by disease or in vivo imaging methods. Overall, the presence of LV intramyocardial fat in the healthy group was relatively rare in comparison to the right ventricle. When present, fat-density lesions were more common in basal segments than in apical portions of the left ventricle, which may be in part related to the volume averaging of extracardiac fat in the atrioventricular groove or the mitral ring. Myocardial infarct is the primary pathologic process that should be considered in the differential of low-density lesions within the LV myocardium. Pathologic data from recent literature show that fat is deposited in healed infarcted myocardium, including 68%5 to 84%4 of myocardial
infarct scars (Fig. 6). For this reason, we also evaluated a group of persons with known infarcts to show differences in segmental distribution of fat deposition. In general, fat deposition was much more common in the infarct group at all levels and segments and was correlated with coronary artery distribution. We also found that the lesion density, as measured in HU, was significantly lower in regions of infarct than was fat deposition in noninfarct patients. Considerable overlap was observed in the 2 groups, however; thus, decreased density alone should not be used as a differentiating factor. The presence of artifact should also be considered in the evaluation of low-density lesions in the left ventricle, including artifacts because of motion, volume averaging, beam hardening, and blooming effects (Fig. 7). Therefore, it is important to confirm the presence of each lesion on noncontrast images as well as review images from different phases of the image acquisition to exclude the presence of artifact. As fat deposition was relatively rare in the healthy group, an LV lesion with density in the range of fat tissue that correlates with coronary distribution and has other imaging characteristics (ie, myocardial thinning, regional dysmotility, or both) likely represents fat deposition secondary to chronic scar of myocardial infarct.
RV fat deposition Our analysis shows that fat deposition of the right ventricle is common in the general population and increases with age. These findings correlate with previously published pathology reports,2,18 in which the youngest persons had minimal or no fat-density lesions, whereas fat deposition was common in older persons.1 The close correlation be-
Raney et al
MDCT shows intramyocardial fat deposition
161
Figure 7 Artifactual low attenuation foci with densities in the range of fat or lower are common in cardiac studies (arrows). These artifacts are usually secondary to a combination of motion and beam hardening. Common locations for such artifacts include blood-myocardium interfaces, mitral leaflet attachments to chordae tendineae (A), and adjacent to high-density structures such as calcified mitral ring (B), descending aorta (C), sternal wires, or pacemaker catheter. Artifacts are particularly common at the base of the left ventricle where the range of motion is the greatest (D).
tween pathologic studies and our investigation provides evidence that MDCT is an accurate method for the evaluation of intramyocardial fat. We observed that RV fat deposition was primarily seen in the basal and midventricular levels and involved the anterolateral and inferolateral aspects of the ventricle. These findings are also noted on pathology studies, with the greatest degree of fatty deposition in both male and female subjects seen in the lateral followed by the anterior RV wall.1 Fat deposition was also commonly observed at the insertion of the inferior right ventricle with the interventricular septum (23%), findings corroborated by Shirani et al.6
The presence of RV fat deposition is of primary importance in the need to distinguish normal fat deposition from ARVD/C, a process with clinical sequelae that include arrhythmias and sudden death.19-21 In our analysis, the segmental distribution of intramyocardial fat by MDCT in the healthy population closely resembles the distribution of fat deposition in ARVD/C as seen by pathology. Burke et al3 showed intramyocardial fat deposition in 62% of the lateral wall (apex and base), 40% of the apical anterior wall, 36% of the anterior and posterior basal right ventricle, and 32% of the posterior apical segments of patients with ARVD/C. The close similarity to our findings suggests that the re-
162
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
gional distribution RV intramyocardial fat is not useful in the differentiation of ARVD/C from fat deposition as a physiologic process. The presence of fat in the RV myocardium should not be diagnosed as ARVD/C without characteristic functional, structural, and electrophysiologic abnormalities that collectively establish the diagnosis of ARVD/ C,19-21 including MDCT findings of functional abnormalities, thinning of the myocardium, aneurysmal changes of the RV wall, and dilatation of the right ventricle.20 The mean age of patients with ARVD is also lower than the population we studied. One important distinction between normal RV fat deposition and ARVD/C may be the subendocardial compared with the subepicardial location of fat within the myocardium. In our study, we observed that RV fat deposition in healthy persons typically began on the endocardial side of the RV free wall and extended outward toward the epicardium, a finding previously reported in healthy persons.6 MDCT images frequently showed the visualization of a thin rim of intact myocardium that separated subepicardial fat from intramyocardial fat, which originated on the endocardial side of the ventricle and involved the inner side of the myocardium (Fig. 5). This finding is in contrast with pathologic reports of ARVD/C, in which fat deposition typically starts from a subepicardial position, extends into the myocardium, and infiltrates only the outer third of the myocardium.22 It is important to note that the latter form of fatty infiltration into the thin wall of RV myocardium could be difficult to diagnose, given the limited spatial resolution of current scanners (0.4 mm).
Limitations In this study, we defined intramyocardial fat deposition as a lesion with density in the range of adipose tissue, but we do not have histologic evidence that these regions contain adipose tissue. Evidence that these regions contain adipose tissue can be seen because our analysis of fat deposition in the right ventricle is similar to previously published pathologic reports. We also believe these lesions are unlikely to represent artifact because we showed the presence of fatty lesions on postcontrast images at different phase intervals and verified on precontrast images. Another limitation was that we found it difficult to ascertain a quantitative measurement of fat deposition in the myocardium. Although we were able to clearly delineate certain segments with lesions of low Hounsfield units because of fat deposition, the transition of fat deposition into regular myocardium frequently occurred without defined margins. In areas of extensive involvement, the overall Hounsfield units of the involved myocardium were generally decreased. On the basis of this observation, areas that did not reach our cutoff for fat (⬍ ⫺30 HU) likely contained fat deposition but could not be measured quantitatively. With current scanners, it is possible to differentiate intramyocardial from normal epicardial fat only if they are separated by intact myocardium. This pattern was common
in our study because in most of our cases fat deposition started from the endocardial aspect of the myocardium, leaving the myocardium intact on the pericardial side. Finally, the mean age of the healthy population was notably younger than the infarct group, and may have implications on the presence of fat in the infarct group as a direct comparison to healthy persons. In addition, the mean age of patients in our healthy population does not represent the mean age of patients with ARVD/C and may affect correlations with these patients.
Conclusion Intramyocardial fat deposition is a common finding in healthy persons corresponding with age in the right ventricle and sex in the left ventricle, and it can be assessed by MDCT using scanning parameters identical to routine CTA.
References 1. Tansey DK, Aly Z, Sheppard MN: Fat in the right ventricle of the normal heart. Histopathology. 2005;46:98 –104. 2. Lorin de la Grandmaison G, Le Bihan C, Durigon M: Assessment of right ventricular lipomatosis by histomorphometry in control adult autopsy cases. Int J Legal Med. 2001;115:105– 8. 3. Burke AP, Farb A, Tashko G, Virmani R: Arrhythmogenic right ventricular cardiomyopathy and fatty replacement of the right ventricular myocardium: are they different diseases? Circulation. 1998;97: 1571– 80. 4. Su L, Siegel JE, Fishbein MC: Adipose tissue in myocardial infarction. Cardiovasc Pathol. 2004;13:98 –102. 5. Baroldi G, Silver MD, De Maria R, Parodi O, Pellegrini A: Lipomatous metaplasia in left ventricular scar. Can J Cardiol. 1997;13:65–71. 6. Shirani J, Berezowski K, Roberts WC: Quantitative measurement of normal and excessive (cor adiposum) subepicardial adipose tissue, its clinical significance, and its effect on electrocardiographic QRS voltage. Am J Cardiol. 1995;76:414 – 8. 7. Leschka S, Alkadhi H, Plass A, Desbiolles L, Grunenfelder J, Marincek B, Wildermuth S: Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J. 2005; 26:1482–7. 8. Mollet NR, Cademartiri F, van Mieghem CA, Runza G, McFadden EP, Baks T, Serruys PW, Krestin GP, de Feyter PJ: High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation. 2005;112: 2318 –23. 9. Raff GL, Gallagher MJ, O’Neill WW, Goldstein JA: Diagnostic accuracy of noninvasive coronary angiography using 64-slice spiral computed tomography. J Am Coll Cardiol. 2005;46:552–7. 10. Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, Silbert JE: Assessment of abdominal fat content by computed tomography. Am J Clin Nutr. 1982;36:172–7. 11. Sjostrom L, Kvist H, Cederblad A, Tylen U: Determination of total adipose tissue and body fat in women by computed tomography, 40K, and tritium. Am J Physiol. 1986;250:736 – 45. 12. Yoshizumi T, Nakamura T, Yamane M, Islam AH, Menju M, Yamasaki K, Arai T, Kotani K: Abdominal fat: standardized technique for measurement at CT. Radiology. 1999;211:283– 6. 13. Baumgartner RN, Heymsfield SB, Roche AF, Bernardino M: Abdominal composition quantified by computed tomography. Am J Clin Nutr. 1988;48:936 – 45.
Raney et al
MDCT shows intramyocardial fat deposition
14. Rossner S, Bo WJ, Hiltbrandt E, Hinson W, Karstaedt N, Santago P, Sobol WT, Crouse JR: Adipose tissue determinations in cadavers--a comparison between cross-sectional planimetry and computed tomography. Int J Obes. 1990;14:893–902. 15. Zhao B, Colville J, Kalaigian J, Curran S, Jiang L, Kijewski P, Schwartz LH: Automated quantification of body fat distribution on volumetric computed tomography. J Comput Assist Tomogr. 2006;30: 777– 83. 16. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS: American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging: Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Int J Cardiovasc Imaging. 2002;18:539 – 42. 17. SAS Institute Inc: SAS/STAT User’s Guide VC. Cary, NC: SAS Institute Inc; 2004. 18. Fontaine G, Fontaliran F, Zenati O, Guzman CE, Rigoulet J, Berthier JL, Frank R: Fat in the heart. A feature unique to the human species?
163
19.
20.
21.
22.
Observational reflections on an unsolved problem. Acta Cardiol. 1999;54:189 –94. McKenna WJ, Thiene G, Nava A, Fontaliran F, Blomstrom-Lundqvist C, Fontaine G, Camerini F: Diagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myocardial and Pericardial Disease of the European Society of Cardiology and of the Scientific Council on Cardiomyopathies of the International Society and Federation of Cardiology. Br Heart J. 1994; 71:215– 8. Kimura F, Sakai F, Sakomura Y, Fujimura M, Ueno E, Matsuda N, Kasanuki H, Mitsuhashi N: Helical CT features of arrhythmogenic right ventricular cardiomyopathy. Radiographics. 2002;22:1111– 24. Marcus FI, Fontaine GH, Guiraudon G, Frank R, Laurenceau JL, Malergue C, Grosgogeat Y: Right ventricular dysplasia: a report of 24 adult cases. Circulation. 1982;65:384 –98. Basso C, Thiene G, Corrado D, Angelini A, Nava A, Valente M: Arrhythmogenic right ventricular cardiomyopathy. Dysplasia, dystrophy, or myocarditis? Circulation. 1996;94:983–91.
Original Research Article
Multidetector computed tomography shows intramyocardial fat deposition Aidan R. Raney, MD, Farhood Saremi, MD*, Satish Kenchaiah, MD, Swaminatha V. Gurudevan, MD, Jagat Narula, MD, Navneet Narula, MD, Stephanie Channual, BS Department of Radiologic Sciences, University of California, Irvine, UCI Medical Center, 101 The City Drive, Route 140, Orange, CA 92868-3298, USA KEYWORDS: ARVD, arrhythmogenic right ventricular dysplasia; Myocardial fat; Myocardial infarction
BACKGROUND: Intramyocardial fat deposition occurs as an age-related process and in multiple pathologic processes. OBJECTIVE: We evaluated the presence of left ventricular (LV) and right ventricular (RV) intramyocardial fat with 64-slice multidetector computed tomography (MDCT). METHODS: One hundred persons with no history of coronary artery disease (47 women, 53 men; mean age [⫾ SD], 53 ⫾ 12.2 years) and 25 patients with CT findings of myocardial infarction (17 men, 8 women; mean age, 71.3 ⫾ 9.6 years) were studied for intramyocardial fat in defined segments of the ventricles (17 LV and 10 RV segments) at 3 levels. Fat deposition was defined as density range of ⫺30 to ⫺190 Hounsfield units on images both before and after contrast. RESULTS: In healthy persons, LV intramyocardial fat was primarily located in the basal segments (5% anteroseptal, 5% inferior), and RV intramyocardial fat was primarily located in the anterolateral (24% of base, 23% of mid) and inferolateral (27% base, 27% mid) segments. Older age was associated with an increased odds of RV (sex-adjusted odds ratio [OR] per decade increment, 1.61; 95% confidence interval [CI], 1.11–2.33; P ⫽ 0.012) but not LV (OR, 0.97; 95% CI, 0.67–1.40; P ⫽ 0.85) intramyocardial fat. Compared with women, men had a lower risk of LV (95% CI, 0.1– 0.64; P ⫽ 0.004) but not RV (95% CI, 0.35–1.87; P ⫽ 0.62) intramyocardial fat. Patients with old myocardial infarction (⬎3 years) had increased percentage of fat in infarcted left ventricles at all 3 levels (P ⱕ 0.004). CONCLUSIONS: Intramyocardial fat can be detected by MDCT and is common in healthy and infarcted myocardium. © 2008 Society of Cardiovascular Computed Tomography. All rights reserved.
Introduction Intramyocardial fat deposition occurs as an age-related process1,2 and in multiple pathologic processes, including Conflict of interest: The authors report no conflicts of interest. Presented at the 56th annual scientific session of the American College of Cardiology, New Orleans, LA, March 24 –27, 2007. * Corresponding author. E-mail address: [email protected] Submitted August 29, 2007. Accepted for publication January 24, 2008.
arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C),3 myocardial infarction (MI),4,5 and dilated cardiomyopathy.5 Previous imaging studies of cardiac fat tissue have focused on pericardial and epicardial fat,6 and there was limited evaluation of adipose tissue within the myocardium itself because of limitations in spatial and temporal resolution. This is especially true in the right ventricle where the free wall thickness approaches a few millimeters. The introduction of multidetector computed tomography (MDCT) has made it possible to assess cardiac disease with
1934-5925/$ -see front matter © 2008 Society of Cardiovascular Computed Tomography. All rights reserved. doi:10.1016/j.jcct.2008.01.004
Raney et al
MDCT shows intramyocardial fat deposition
153
Figure 1 Differentiating artifact from fat deposition in a 49-year-old man with no history of heart disease. A review of consecutive images in different projections, and particularly in different phases, allows for verification of the presence of intramyocardial fat. (A) It is difficult to differentiate whether the lesion represents fat or artifact. Artifact-free images in systole at 30% of R-R interval (B) and late diastole at 80% of R-R interval (C) show the true presence of endocardial fat deposition in the anteroseptal wall of the left ventricle.
a high level of detail, including myocardial infarct and small atherosclerotic plaques.7-9 CT was also shown to be accurate in the assessment and quantification of fat tissue because of attenuation values distinct from those of other anatomical structures.10-15 The purpose of this study was to combine the improved spatial and temporal resolution of MDCT with the unique attenuation values of fat tissue to assess the spatial distribution and structural pattern of fat deposition within the myocardium. We evaluated intramyocardial fat deposition in the left and right ventricles of healthy persons with no clinical history or CT evidence of cardiac disease. In the left ventricle, we also included a group of patients with history of MI to show differences in the distribution of intramyocardial fat deposition with the healthy population.
Methods The study was conducted with the approval of the institutional review board at the University of California Irvine Medical Center and was compliant with Health Insurance Portability and Accountability Act regulations. Informed consent was not required for this retrospective analysis.
Patient selection To identify healthy persons, we reviewed 130 consecutive electrocardiogram (ECG)– gated coronary MDCT studies (64-slice scanner; Toshiba Aquilion, Tustin, CA) performed at our institution from December 2005 to January 2006. Patients completed a questionnaire, and patients with coronary artery disease (defined as a clinical history of significant coronary artery disease [CAD], coronary artery bypass graft surgery, or intracoronary stent placement) were excluded. We also excluded patients with previously unrec-
ognized CAD that was apparent on the evaluation of images. Studies were then evaluated in multiple views (short axis, long axis, and 3-dimensional) and were given a subjective score of good, average, and poor, based on the presence of metal artifact (pacemakers, sternal wires, etc), noise, calcification, streak artifact related to contrast-enhanced cardiac chambers, beam hardening that caused false attenuation, and motion artifact. After excluding poor quality studies, 100 studies with no history of CAD or infarct were reviewed for fat deposition of the right ventricular (RV) and the left ventricular (LV) myocardium. For the infarct group, we evaluated patients with a clinical history of MI by questionnaire who received cardiac CT angiography (CTA) between December 2005 and November 2006 (50 patients). We then selected the first 25 cases with characteristics of infarction on CTA (ie, hypokinesis, akinesis, and thinning of myocardium) who had fatty areas (⬍ ⫺30 Hounsfield units [HU]) within the infarcted segment.
Scan protocol and image reconstruction An ECG was completed, and vital signs were obtained for all patients on arrival at the imaging suite. For pa-
Table 1 Key patient characteristics for the healthy and infarct groups
Age (y), mean ⫾ SD Sex, % male (n) Weight (kg), mean ⫾ SD BMI (kg/m2), mean ⫾ SD BMI, body mass index.
Healthy group (n ⫽ 100)
Infarct group (n ⫽ 25)
53.0 ⫾ 12.2 53 (53) 167.9 ⫾ 34.0 26.1 ⫾ 3.6
71.3 ⫾ 9.6 68 (17) 172.5 ⫾ 42.2 27.9 ⫾ 6.9
154
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 2 (A) Segmental distribution of intramyocardial fat in the left ventricle of infarct (top) and healthy participants (middle), showing the percentage of each segment positive for fat deposition, based on the previously described 17-segment model.16 (B) Segmental distribution of intramyocardial fat in the right ventricle of healthy persons, showing demonstrating the percentage of each segment positive for fat deposition. (Note: description of segments is available in Table 2.)
tients with a heart rate greater than 65 beats/min, oral or intravenous metoprolol was used to achieve a heart rate of ⬍65 beats/min. All patients received sublingual nitroglycerin (0.4 – 0.8 mg) 1 minute before image acquisition, unless contraindicated. All scans were performed on a scanner with 64 parallel detector rows and an individual detector width of 0.5 mm. With the use of a prospectively ECG-gated (mid-diastole) technique, precontrast 3-mm nonspiral images were obtained from the level of tracheal bifurcation to the diaphragm to cover the entire heart. Contrast enhancement was achieved using 70 – 80 mL of iohexol (Omnipaque 350 mg/mL; Amersham Health, Cork, Ireland) injected at a rate of 4 –5 mL/sec, followed by an injection of 50 mL of saline at 5 mL/s through an 18-gauge catheter in an antecubital vein to allow washout of contrast from the right heart and superior vena cava. Scan parameters included 120 kVP tube voltage, 400 mA tube current, 64 ⫻ 0.5 mm collimation, 7.2 mm table feed per rotation, and a gantry rotation time of 400 milliseconds. Scan start was automatically initiated 4 seconds
after reaching a threshold of 180 HU in the descending aorta at the tracheal bifurcation level. A retrospective ECG-gated volumetric data set was acquired during a single breathhold. Depending on the heart rate throughout the examination, axial slices were reconstructed synchronized to the ECG by a nonsegmented (ⱕ65 beats/ min) or segmented (⬎65 beats/min) image reconstruction algorithm. When necessary, R-wave indicators were manually repositioned to improve the quality of synchronization. Diastolic axial images were reconstructed based on a relative-delay strategy at 70%, 75%, and 80% of R-R intervals, respectively. In case of persistent artifacts in the area of interest, a second reconstruction approach was performed, and systolic images were reconstructed (25%–35% R-R interval). To evaluate regional myocardial function, 2-mm slices were reconstructed at 10% R-R intervals. Axial slices with a thickness of 0.5 mm (increment, 0.3 mm) and a cardiac CTA algorithm were used for evaluation of the myocardium. The data set least affected by cardiac motion was transferred to an offline
Raney et al
MDCT shows intramyocardial fat deposition
Table 2 Segmental prevalence of intramyocardial fat deposition in the left and right ventricles Percentage of involved segments
Left ventricle segment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 — — Right ventricle segment 1 2 3 4 5 6 7 8 9 10 — — —
Corresponding area
Healthy population
Infarct group
Base anterior Base anteroseptal Base inferoseptal Base inferior Base inferolateral Base anterolateral Mid anterior Mid anteroseptal Mid inferoseptal Mid inferior Mid inferolateral Mid anterolateral Apical anterior Apical septal Apical inferior Apical lateral Apex Anterior papillary Posterior papillary
2 5 3 5 2 2 1 2 5 0 2 0 1 3 1 1 1 14 12
8 8 8 16 24 28 20 12 16 24 28 12 24 20 28 32 16 — —
Base outflow tract Base inferolateral Base inferoseptal Base anteroseptal Mid anterolateral Mid inferolateral Mid inferoseptal Mid anteroseptal Apical free Apical septal Anterior insertion Inferior insertion Moderator band
24 27 2 6 23 27 3 1 12 3 1 23 16
— — — — — — — — — — — — —
3D Vitrea workstation (Vital Images, Inc, Minnetonka, MN) for further analysis.
CT data analysis Multiplanar reformations of the axial images (short and long axis views) after contrast were rendered and evaluated for the presence of intramyocardial fat deposition by consensus of 2 authors (F.S. and S.G.) with 16 and 2 years of experience in CT interpretation, respectively. All measurements were performed by a resident (A.R.) with 1 year of experience in cardiac CT imaging and confirmed by author F.S. The right and left ventricles
155 were separated into segments and analyzed for the presence of fat tissue in the myocardium. Anatomic localization of fat was performed with high resolution postcontrast data and then confirmed by precontrast images. To differentiate low attenuation areas as intramyocardial fat deposition compared with artifacts (motion, beam hardening, or blooming), studies with uncertain findings were reviewed in at least 2 different reconstructions of the R-R interval of the cardiac cycle (Fig. 1). A segment was considered positive for fat deposition based on the presence of a linear lesion with threshold attenuation values between ⫺30 and ⫺190 HU on series both before and after contrast, a range used in previous studies of fat tissue.10-15 The maximum width of the visualized fatty density was measured at each segment. LV myocardial segments were analyzed based on a 17segment model of the left ventricle16 at 3 levels of the ventricle (base, midventricular, and apical). The presence or absence of fat in the papillary muscles was also evaluated. The right ventricle was evaluated in 10 segments at the same 3 levels. Base segments (4) included the anterolateral (pulmonary outflow tract), inferolateral, inferior septal, and anterior septal walls; the midventricular segments (4) included the anterolateral and inferolateral wall, inferior septal wall, and anterior septal wall; and the apical segments (2) included the free and septal walls. We assessed fat deposition at the anterior and posterior RV insertion (insertion of free wall of the right ventricle with the interventricular septum) and categorized these separately from the above segments. We also evaluated fat deposition in the moderator band and its septal band insertion. The relative location (endocardial side compared with epicardial side) of intramyocardial fat deposition in both groups was assessed. The maximum visible thickness of the RV wall was measured at 5 segments, corresponding to 3 levels of the ventricle, including segments 1 and 2 of the base, segments 5 and 6 of the mid-ventricle, and segment 9 of the apical right ventricle. The anatomic location of fat deposition along major coronary arteries in healthy and infarct groups was recorded. As per the 17-segment model,16 coronary distribution included segments 1, 2, 7, 8, 13, 14, and 17 for the left anterior descending artery; segments 5, 6, 11, 12, and 16 for the left circumflex artery; and segments 3, 4, 9, 10, and 15 for right coronary artery.
Statistical analysis We computed mean ⫾ standard deviation (SD) and ranges for continuous variables and proportions (expressed as percentages) for categorical variables. We evaluated fat as both dichotomous (presence or absence) and continuous (defined as number of segments involved in the 17-segment model of left ventricle and 10-segment model of right ventricle) variables. We used linear and logistic regression analyses, as appropriate, to examine the cross-sectional association of age and sex with intramyocardial fat in the right ventricle only and the left
156
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 3 Common locations of fat deposition in healthy persons at the base of the left ventricle, including the anteroseptal (A), inferoseptal (B), and inferior (C) segments with HU values at each location. Evaluation of images before and after contrast reduces the possibility that low HU lesions are due to artifact. Note that the values for precontrast images are generally lower than are postcontrast images and are most likely because of volume averaging of small areas of fat deposition with normal myocardium in thick noncontrast (3 mm) images compared with postcontrast images (0.5 mm). Intramyocardial fatty deposition typically involves the endocardial aspect of the myocardium, although we did observe intramyocardial fat in the inferior and inferolateral segments of the base of the left ventricle in 3 patients (C). Table 3 Results of logistic regression analyses examining the predictors of presence or absence of fat in myocardium of healthy participants. Univariate analyses Variables Left ventricle only (excluding papillary muscles) Age (per decade increase) Sex (male/female) Right ventricle only Age (per decade increase) Sex (male/female) OR, odds ratio; 95% CI, 95% confidence intervals.
Multivariable analyses
OR (95% CI)
P
OR (95% CI)
P
1.10 (0.78–1.56) 0.25 (0.10–0.64)
0.59 0.003
0.97 (0.67–1.40) 0.25 (0.10–0.64)
0.85 0.004
1.64 (1.14–2.36) 0.63 (0.28–1.39)
0.007 0.25
1.61 (1.11–2.33) 0.81 (0.35–1.87)
0.012 0.62
Raney et al
MDCT shows intramyocardial fat deposition
157
Figure 4 Fat deposition in infarcted myocardium. Images before and after contrast of the heart in the left circumflex (LCx; A) artery, left anterior descending (LAD; B) artery, and the right coronary artery (RCA; C) distribution of 3 different patients are shown. Fat deposition is typically subendocardial and large enough to cover a significant portion of infarcted myocardium. Associated findings included thinning of myocardium (A and B), abnormal myocardial function, and corresponding coronary artery disease (not shown but present in all persons). The findings help differentiate fat deposition of infarct scar from fat deposition as a normal variant.
ventricle only excluding papillary muscles. To compare the presence or absence and extent (number of involved myocardial segments) of intramyocardial fat among healthy participants and patients with MI, we constructed unadjusted and age- and sex-adjusted logistic and linear regression models, respectively. We expressed results of logistic regression analyses as odds ratios, 95% confidence intervals, and P values, and linear regression as P values. We considered a 2-sided P value of ⬍0.05 as statistically significant. We performed all analyses using SAS software version 9.1 (SAS Institute, Cary, NC).17
Results Patients The patients evaluated in the healthy group included 47 women and 53 men with an age of 53.0 ⫾ 12.2 years (Table 1). Twenty-one patients with CAD (including 6 by history and 15 after evaluation of images) and 9 of the studies with poor quality as defined above were excluded from the study. In the infarct group, all 25 cases of infarct were at least 3 years old by clinical history (7.6 ⫾ 3.7 years). The patients
158
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Table 4 Results of logistic regression analyses comparing presence or absence of fat in various regions of the left myocardium among healthy persons and patients with myocardial infarction Unadjusted
Left ventricular level Base Mid Apical Coronary distribution LAD LCx RCA
Age and sex-adjusted
Healthy, %
Infarct, %
OR (95%CI)
P
OR (95%CI)
P
20 10 6
64 60 60
6.0 (2.3–15.2) 15.2 (5.3–43.5) 36.0 (10.0–129.6)
0.0002 ⬍0.0001 ⬍0.0001
11.7 (3.2–43.0) 14.4 (3.6–57.7) 18.8 (4.0–87.9)
0.0002 0.0002 0.0002
15 14 7
52 68 48
5.3 (2.1–13.6) 15.5 (5.5–43.8) 12.3 (4.1–36.8)
0.0005 ⬍0.0001 ⬍0.0001
5.9 (1.7–19.9) 19.8 (4.9–79.7) 8.8 (2.2–34.8)
0.004 ⬍0.0001 0.002
OR, odds ratio; 95% CI, 95% confidence intervals; LAD, left anterior descending coronary artery; LCX, left circumflex coronary artery; RCA, right coronary artery. According to the 17-segment model,16 coronary distribution included segments 1, 2, 7, 8, 13, 14, and 17 for the LAD; segments 5, 6, 11, 12, and 16 for the LCx; and segments 3, 4, 9, 10, and 15 for the RCA.
included 17 men and 8 women with an age of 71.3 ⫾ 9.6 years. Ninety-two percent of these patients had a history of coronary intervention, defined as ⱖ1 of the following: angioplasty/stent (81%) or coronary artery bypass graft (52%). The mean (⫾SD) heart rate for all patients before data acquisition was 58 ⫾ 5 beats/min (range: 47– 68 beats/min), and the mean scan time was 9.1⫾1.4 seconds (range: 8 –13 seconds).
LV fat deposition A total of 1700 LV segments in 100 healthy persons and 425 LV segments in 25 patients with clinical history of MI were analyzed for the presence of intramyocardial fat deposition. Both groups contained intramyocardial fat; however, the structure, segmental location (Fig. 2; Table 2), and lowest Hounsfield units in the lesions were different in the healthy participants compared with patients with a history of MI. In the healthy group, LV intramyocardial fat was primarily located at the basal level, with ⱖ1 involved segments in 19% of patients. The anteroseptal (5% of patients) and inferior (5% of patients) segments were the most commonly involved individual segments in the basal region. Fat deposition was uncommon in the mid and apical levels of the left ventricle, with ⱖ1 involved segments seen in 10% and 6% of the healthy population, respectively. The average maximum width was 2.2 ⫾ 0.2 mm at the base, 1.8 ⫾ 0.2 mm at the midventricular level, and 2.1 ⫾ 1.0 mm at the apical level. All intramyocardial fat deposition was located on the endocardial aspect with a preserved layer of myocardium on the epicardial side of the myocardium (Fig. 3), with the exception of 3 lesions that had preserved myocardium on both sides (Fig. 3C) located in the inferior (2) and inferolateral (1) segments at the basal level. We commonly observed round foci of fat densities in LV papillary muscles and at their attachments to the chordae tendineae, which were evenly distributed anteriorly (14%) and posteriorly (12%) on postcontrast images. Fatty foci in
papillary muscles were seen in only 3 patients on images both before and after contrast. Compared with women, men had a lower risk of LV (age-adjusted odds ratio [OR], 0.25; 95% confidence interval [CI], 0.1– 0.64; P ⫽ 0.004) but not RV (age-adjusted OR, 0.81; 95% CI, 0.35–1.87; P ⫽ 0.62) intramyocardial fat (Table 3). The higher prevalence of fat deposition in ⱖ1 LV myocardial segments of female participants was seen at all 3 levels (unadjusted P ⫽ 0.015, and age-adjusted P ⫽ 0.022). The infarct group showed a higher prevalence of fat deposition of the myocardium at all levels of the left ventricle (Figs. 2 and 4; Table 4), with ⱖ1 affected segment in all but one study (96%). Compared with the healthy group, greater LV intramyocardial fat deposition in the infarct group was evident at each level of the ventricle, including 64% of basal, 60% of mid, and 60% of apical levels (Table 4). The most common individual segments with fat deposition included the inferior and inferolateral segments of the base;, the anterior, inferior, and inferolateral segments of the mid-ventricle; and all segments of the apical level of the ventricle. The location of fat deposition in the ventricular wall was primarily on the endocardial side in the infarct and healthy groups, although areas of transmural extension were only seen in the infarct group (Fig. 4). The lesion density was lower in the infarct group, in which the average of the lowest density in each lesion was ⫺127.3 ⫾ 55.4 HU (range, ⫺51 to ⫺291 HU) compared with ⫺77.2 ⫾ 32.3 HU (range, ⫺30 to ⫺180 HU) in the healthy group (unadjusted P ⬍ 0.0001, age-adjusted P ⫽ 0.0013) on postcontrast images. The average density was lower in precontrast images than in postcontrast images ion both the infarct and noninfarct groups (Figs. 3 and 4).
RV fat deposition A total of 1000 RV myocardial segments were assessed in the healthy population. One or more segments positive for fat deposition were observed in 36% of basal,
Raney et al
MDCT shows intramyocardial fat deposition
159
Figure 5 Short axis views showing typical anatomic distribution of myocardial fat in the right ventricle of 4 healthy persons. Note the deposition of fat on the endocardial aspect of the myocardium with preserved myocardium on the epicardial side of the pulmonary outflow tract (1), lateral free wall at the base (2), inferior free wall (3), septal and moderator bands (4), and insertion point of the right ventricle with the septum (5). LV indicates left ventricle.
34% of mid, and 13% of apical levels of the right ventricle. The most commonly involved individual segments of the right ventricle were the anterolateral (pulmonary outflow tract; 24%) and inferolateral (27%) segments of the base and the anterolateral (23%) and inferolateral (27%) segments of the mid-ventricle (Fig. 2). Involvement of the RV septal wall was less common, with the basal anteroseptal segment seen as the most commonly affected septal segment (6%). The presence of fat in the moderator band was seen in 16% of patients, and fat deposition at the inferior septal insertion of the RV free wall was seen in 23 patients (23%), whereas it was only detected anteriorly in 1 patient (1%). The structure of RV intramyocardial fat deposition frequently showed the spread of fat deposition beginning on the endocardial side
of the RV wall and extending outward toward the epicardial side of the free wall, with a thin layer of preserved myocardium on the subepicardial aspect (Fig. 5). The average maximum width of the measurable fatty area was 2.7 ⫾ 1.6 mm at the base, 2.4 ⫾ 1.4 mm at the midventricular level, and 2.0 ⫾ 1.2 mm at the apical level. The maximum visible thickness of the RV wall was 4.0 ⫾ 0.86 mm in segment 1, 3.4 ⫾ 0.71 mm in segment 2, 4.3 ⫾ 1.01 mm in segment 5, 3.4 ⫾ 0.78 mm in segment 6, and 3.6 ⫾ 0.95 mm in segment 9 of the right ventricle. Older age (per decade increase) was associated with an increased odds of RV (sex-adjusted OR per decade increment, 1.61; 95% CI, 1.11–2.33; P ⫽ 0.012) but not LV (sex-adjusted OR, 0.97; 95% CI, 0.67–1.40; P ⫽ 0.85) intramyocardial fat (Table 3).
160
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
Figure 6 Fat deposition of infarcted (A) compared with healthy myocardium (B). Microphotograph of a myocardial specimen (original magnification, ⫻2; hematoxylin-eosin stain) from a 67-year-old man with severe triple vessel coronary artery disease showing fat deposition in an area of healed infarct (long arrows). This patient had multiple areas of patchy replacement fibrosis (arrowheads) in the distribution of the left anterior descending artery (not included in present study). Some of these foci show entrapped fat. (B) Photomicrograph of myocardial specimen (original magnification, ⫻4) from a remote area shows the presence of fat in the interstitial area (arrows) between cardiomyocytes without accompanying interstitial or replacement fibrosis.
Discussion In this study we evaluated fat deposition within the RV and LV myocardium of healthy persons (without CAD) by MDCT. To provide clinical utility for this data, we then compared our findings with the primary differential for fat deposition in each ventricle. This included a separate group of our patients with MI for the left ventricle and reported pathologic studies of ARVD/C for comparison of RV intramyocardial fat.
LV fat deposition On evaluation of LV intramyocardial fat, we found that women had a significantly higher number of LV segments with fat deposition than did the male patients. The presence of LV intramyocardial fat was briefly described in pathology reports,6 and, although RV intramyocardial fat was correlated with female sex,1 we did not encounter any reports comparing age or sex with fat deposition in the LV myocardium by disease or in vivo imaging methods. Overall, the presence of LV intramyocardial fat in the healthy group was relatively rare in comparison to the right ventricle. When present, fat-density lesions were more common in basal segments than in apical portions of the left ventricle, which may be in part related to the volume averaging of extracardiac fat in the atrioventricular groove or the mitral ring. Myocardial infarct is the primary pathologic process that should be considered in the differential of low-density lesions within the LV myocardium. Pathologic data from recent literature show that fat is deposited in healed infarcted myocardium, including 68%5 to 84%4 of myocardial
infarct scars (Fig. 6). For this reason, we also evaluated a group of persons with known infarcts to show differences in segmental distribution of fat deposition. In general, fat deposition was much more common in the infarct group at all levels and segments and was correlated with coronary artery distribution. We also found that the lesion density, as measured in HU, was significantly lower in regions of infarct than was fat deposition in noninfarct patients. Considerable overlap was observed in the 2 groups, however; thus, decreased density alone should not be used as a differentiating factor. The presence of artifact should also be considered in the evaluation of low-density lesions in the left ventricle, including artifacts because of motion, volume averaging, beam hardening, and blooming effects (Fig. 7). Therefore, it is important to confirm the presence of each lesion on noncontrast images as well as review images from different phases of the image acquisition to exclude the presence of artifact. As fat deposition was relatively rare in the healthy group, an LV lesion with density in the range of fat tissue that correlates with coronary distribution and has other imaging characteristics (ie, myocardial thinning, regional dysmotility, or both) likely represents fat deposition secondary to chronic scar of myocardial infarct.
RV fat deposition Our analysis shows that fat deposition of the right ventricle is common in the general population and increases with age. These findings correlate with previously published pathology reports,2,18 in which the youngest persons had minimal or no fat-density lesions, whereas fat deposition was common in older persons.1 The close correlation be-
Raney et al
MDCT shows intramyocardial fat deposition
161
Figure 7 Artifactual low attenuation foci with densities in the range of fat or lower are common in cardiac studies (arrows). These artifacts are usually secondary to a combination of motion and beam hardening. Common locations for such artifacts include blood-myocardium interfaces, mitral leaflet attachments to chordae tendineae (A), and adjacent to high-density structures such as calcified mitral ring (B), descending aorta (C), sternal wires, or pacemaker catheter. Artifacts are particularly common at the base of the left ventricle where the range of motion is the greatest (D).
tween pathologic studies and our investigation provides evidence that MDCT is an accurate method for the evaluation of intramyocardial fat. We observed that RV fat deposition was primarily seen in the basal and midventricular levels and involved the anterolateral and inferolateral aspects of the ventricle. These findings are also noted on pathology studies, with the greatest degree of fatty deposition in both male and female subjects seen in the lateral followed by the anterior RV wall.1 Fat deposition was also commonly observed at the insertion of the inferior right ventricle with the interventricular septum (23%), findings corroborated by Shirani et al.6
The presence of RV fat deposition is of primary importance in the need to distinguish normal fat deposition from ARVD/C, a process with clinical sequelae that include arrhythmias and sudden death.19-21 In our analysis, the segmental distribution of intramyocardial fat by MDCT in the healthy population closely resembles the distribution of fat deposition in ARVD/C as seen by pathology. Burke et al3 showed intramyocardial fat deposition in 62% of the lateral wall (apex and base), 40% of the apical anterior wall, 36% of the anterior and posterior basal right ventricle, and 32% of the posterior apical segments of patients with ARVD/C. The close similarity to our findings suggests that the re-
162
Journal of Cardiovascular Computed Tomography, Vol 2, No 3, May/June 2008
gional distribution RV intramyocardial fat is not useful in the differentiation of ARVD/C from fat deposition as a physiologic process. The presence of fat in the RV myocardium should not be diagnosed as ARVD/C without characteristic functional, structural, and electrophysiologic abnormalities that collectively establish the diagnosis of ARVD/ C,19-21 including MDCT findings of functional abnormalities, thinning of the myocardium, aneurysmal changes of the RV wall, and dilatation of the right ventricle.20 The mean age of patients with ARVD is also lower than the population we studied. One important distinction between normal RV fat deposition and ARVD/C may be the subendocardial compared with the subepicardial location of fat within the myocardium. In our study, we observed that RV fat deposition in healthy persons typically began on the endocardial side of the RV free wall and extended outward toward the epicardium, a finding previously reported in healthy persons.6 MDCT images frequently showed the visualization of a thin rim of intact myocardium that separated subepicardial fat from intramyocardial fat, which originated on the endocardial side of the ventricle and involved the inner side of the myocardium (Fig. 5). This finding is in contrast with pathologic reports of ARVD/C, in which fat deposition typically starts from a subepicardial position, extends into the myocardium, and infiltrates only the outer third of the myocardium.22 It is important to note that the latter form of fatty infiltration into the thin wall of RV myocardium could be difficult to diagnose, given the limited spatial resolution of current scanners (0.4 mm).
Limitations In this study, we defined intramyocardial fat deposition as a lesion with density in the range of adipose tissue, but we do not have histologic evidence that these regions contain adipose tissue. Evidence that these regions contain adipose tissue can be seen because our analysis of fat deposition in the right ventricle is similar to previously published pathologic reports. We also believe these lesions are unlikely to represent artifact because we showed the presence of fatty lesions on postcontrast images at different phase intervals and verified on precontrast images. Another limitation was that we found it difficult to ascertain a quantitative measurement of fat deposition in the myocardium. Although we were able to clearly delineate certain segments with lesions of low Hounsfield units because of fat deposition, the transition of fat deposition into regular myocardium frequently occurred without defined margins. In areas of extensive involvement, the overall Hounsfield units of the involved myocardium were generally decreased. On the basis of this observation, areas that did not reach our cutoff for fat (⬍ ⫺30 HU) likely contained fat deposition but could not be measured quantitatively. With current scanners, it is possible to differentiate intramyocardial from normal epicardial fat only if they are separated by intact myocardium. This pattern was common
in our study because in most of our cases fat deposition started from the endocardial aspect of the myocardium, leaving the myocardium intact on the pericardial side. Finally, the mean age of the healthy population was notably younger than the infarct group, and may have implications on the presence of fat in the infarct group as a direct comparison to healthy persons. In addition, the mean age of patients in our healthy population does not represent the mean age of patients with ARVD/C and may affect correlations with these patients.
Conclusion Intramyocardial fat deposition is a common finding in healthy persons corresponding with age in the right ventricle and sex in the left ventricle, and it can be assessed by MDCT using scanning parameters identical to routine CTA.
References 1. Tansey DK, Aly Z, Sheppard MN: Fat in the right ventricle of the normal heart. Histopathology. 2005;46:98 –104. 2. Lorin de la Grandmaison G, Le Bihan C, Durigon M: Assessment of right ventricular lipomatosis by histomorphometry in control adult autopsy cases. Int J Legal Med. 2001;115:105– 8. 3. Burke AP, Farb A, Tashko G, Virmani R: Arrhythmogenic right ventricular cardiomyopathy and fatty replacement of the right ventricular myocardium: are they different diseases? Circulation. 1998;97: 1571– 80. 4. Su L, Siegel JE, Fishbein MC: Adipose tissue in myocardial infarction. Cardiovasc Pathol. 2004;13:98 –102. 5. Baroldi G, Silver MD, De Maria R, Parodi O, Pellegrini A: Lipomatous metaplasia in left ventricular scar. Can J Cardiol. 1997;13:65–71. 6. Shirani J, Berezowski K, Roberts WC: Quantitative measurement of normal and excessive (cor adiposum) subepicardial adipose tissue, its clinical significance, and its effect on electrocardiographic QRS voltage. Am J Cardiol. 1995;76:414 – 8. 7. Leschka S, Alkadhi H, Plass A, Desbiolles L, Grunenfelder J, Marincek B, Wildermuth S: Accuracy of MSCT coronary angiography with 64-slice technology: first experience. Eur Heart J. 2005; 26:1482–7. 8. Mollet NR, Cademartiri F, van Mieghem CA, Runza G, McFadden EP, Baks T, Serruys PW, Krestin GP, de Feyter PJ: High-resolution spiral computed tomography coronary angiography in patients referred for diagnostic conventional coronary angiography. Circulation. 2005;112: 2318 –23. 9. Raff GL, Gallagher MJ, O’Neill WW, Goldstein JA: Diagnostic accuracy of noninvasive coronary angiography using 64-slice spiral computed tomography. J Am Coll Cardiol. 2005;46:552–7. 10. Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, Silbert JE: Assessment of abdominal fat content by computed tomography. Am J Clin Nutr. 1982;36:172–7. 11. Sjostrom L, Kvist H, Cederblad A, Tylen U: Determination of total adipose tissue and body fat in women by computed tomography, 40K, and tritium. Am J Physiol. 1986;250:736 – 45. 12. Yoshizumi T, Nakamura T, Yamane M, Islam AH, Menju M, Yamasaki K, Arai T, Kotani K: Abdominal fat: standardized technique for measurement at CT. Radiology. 1999;211:283– 6. 13. Baumgartner RN, Heymsfield SB, Roche AF, Bernardino M: Abdominal composition quantified by computed tomography. Am J Clin Nutr. 1988;48:936 – 45.
Raney et al
MDCT shows intramyocardial fat deposition
14. Rossner S, Bo WJ, Hiltbrandt E, Hinson W, Karstaedt N, Santago P, Sobol WT, Crouse JR: Adipose tissue determinations in cadavers--a comparison between cross-sectional planimetry and computed tomography. Int J Obes. 1990;14:893–902. 15. Zhao B, Colville J, Kalaigian J, Curran S, Jiang L, Kijewski P, Schwartz LH: Automated quantification of body fat distribution on volumetric computed tomography. J Comput Assist Tomogr. 2006;30: 777– 83. 16. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS: American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging: Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Int J Cardiovasc Imaging. 2002;18:539 – 42. 17. SAS Institute Inc: SAS/STAT User’s Guide VC. Cary, NC: SAS Institute Inc; 2004. 18. Fontaine G, Fontaliran F, Zenati O, Guzman CE, Rigoulet J, Berthier JL, Frank R: Fat in the heart. A feature unique to the human species?
163
19.
20.
21.
22.
Observational reflections on an unsolved problem. Acta Cardiol. 1999;54:189 –94. McKenna WJ, Thiene G, Nava A, Fontaliran F, Blomstrom-Lundqvist C, Fontaine G, Camerini F: Diagnosis of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Task Force of the Working Group Myocardial and Pericardial Disease of the European Society of Cardiology and of the Scientific Council on Cardiomyopathies of the International Society and Federation of Cardiology. Br Heart J. 1994; 71:215– 8. Kimura F, Sakai F, Sakomura Y, Fujimura M, Ueno E, Matsuda N, Kasanuki H, Mitsuhashi N: Helical CT features of arrhythmogenic right ventricular cardiomyopathy. Radiographics. 2002;22:1111– 24. Marcus FI, Fontaine GH, Guiraudon G, Frank R, Laurenceau JL, Malergue C, Grosgogeat Y: Right ventricular dysplasia: a report of 24 adult cases. Circulation. 1982;65:384 –98. Basso C, Thiene G, Corrado D, Angelini A, Nava A, Valente M: Arrhythmogenic right ventricular cardiomyopathy. Dysplasia, dystrophy, or myocarditis? Circulation. 1996;94:983–91.