Lynette Duncanson

MRI feature tracking strain profiles distinguish patients with left ventricular systolic and diastolic dysfunction with and without clinical heart failure

Background Characterizing myocardial mechanical properties is valuable in understanding cardiomyopathy with and without clinical heart failure (HF). We sought to examine the incremental value of strain assessment in patients with left ventricular systolic dysfunction without HF (LVSD), HF with reduced ejection fraction (HFrEF), LV diastolic dysfunction without HF (LVDD) and HF with preserved EF (HFpEF) during routine cardiac MR (CMR) evaluation.

Association of reduced right ventricular global and regional wall motion with abnormal right heart hemodynamics

Background The relationship of right ventricular (RV) regional and global systolic function to right heart hemodynamics is not well understood. In this study we used MRI feature tracking to assess regional RV wall motion, cine MR to evaluate global RV function and examined the relationship of regional wall motion and global RV function to right heart hemodynamics. Methods Fifty patients undergoing clinically indicated right heart catheterization were prospectively recruited to a research CMR within 5 hours of catheterization. Majority of the heart failure cases were due to left heart failure. SSFP cine images were acquired to evaluate RV regional and global function. RV longitudinal and long axis radial strains were derived from the 4-chamber cine, and mid wall septal circumferential and radial strains from the short-axis plane in mid ventricle using CIM feature tracking software (Auckland, NZ). Right heart hemodynamics was assessed during catheterization. Results Mean age was 64±13 years, mean RV ejection fraction (EF) 51±13%. Reduced longitudinal and radial displacement in 4 chamber view was significantly associated with reduced RVEF, r=-0.621 (p<0.001) and r=0.346 (p=0.014), respectively. Similarly, in short axis plane reduced septal circumferential and radial strain were also significantly correlated with RVEF, r=-0.488 (p=0.001) and r=0.527 (p<0.001) respectively. In regression analysis reduced RVEF had strongest association with increased pulmonary wedge pressure (r=-0.622, p<0.001) in univariate analysis and in multivariate analysis (p<0.001) after adjusting for all the right heart hemodynamic parameters. Pulmonary wedge pressure contributed to 41% of RVEF variation in this cohort with predominant left heart failure. While also significantly correlated with pulmonary wedge pressure, regional wall motion including longitudinal strain, septal circumferential and radial strain had strongest association with mean pulmonary arterial pressure, r=0.577 (p<0.001), r=0.440 (p=0.002) and r=-0.451 (p=0.001), respectively. Conclusions It is feasible to use MR feature tracking to characterize RV regional wall motion. Reduced global and regional RV systolic functions are associated with right heart hemodynamic abnormalities. In patients with predominant left heart failure, pulmonary wedge pressure contributes importantly to RVEF variation. Funding None.

Comparisons of myocardial strain and strain rate in patients with heart failure with preserved and reduced ejection fraction using Feature Tracking of cine MR images

Methods All participants were prospectively recruited and underwent CMR in a 1.5 T scanner. LVEF 400 pg/ml. Patients with myocardial infarction were excluded. Myocardial circumferential strain (CST) and strain rate (CSR) was analyzed in mid LV of the short axis plane and the longitudinal strain (LST) and strain rate (LSR) in 4-chamber view of the SSFP cine images using feature tracking (CIM software, Auckland, New Zealand). LV end diastolic pressure (LVEDP) was estimated using normalized left atrial transition time from time-intensity curves of the first pass perfusion images.

Association of increased extracellular volume expansion with elevated left ventricular end diastolic pressure in patients with nonischemic cardiomyopathy

Background Myocardial extracellular volume expansion is a marker of myocardial fibrosis which is associated with myocardial stiffness and impaired relaxation property. In this study we investigated the association of extracellular volume indices using T1 mapping and left ventricular end diastolic pressure (LVEDP) using left atrial transit time in patients with nonischemic cardiomyopathy. Methods All subjects were prospectively recruited to undergo cardiac MRI in a 1.5 T scanner. Modified Look-Locker Inversion recovery sequence was used with motion correction included in the post processing algorithm. Pre-contrast and post contrast (20 minutes after gadolinium injection at 0.15 mmol/kg) T1 maps of the blood pool and myocardium were assessed in basal, mid and apical segments of the short axis planes using Siemens software. Calculated T1 indices were the average value of the basal and mid slice which included the pre- and post-contrast myocardial T1, contrast partition coefficient using the ratio of signal change in blood and myocardium before and after contrast administration, and extracellular volume fraction (ECV) which was partition coefficient multiplied with one minus hematocrit to account for the blood contrast volume of distribution. LVEDP was assessed using normalized mean left atrial transit time derived from the time-signal intensity curve of the first pass perfusion image during gadolinium injection at 0.01 mmol/kg. Results Compared to the normal controls (N = 8) patients with nonischemic cardiomyopathy (N = 31) had lower LV ejection fraction (EF) (47 ± 12% vs 54 ± 3%, p = 0.021) and higher LVEDP (14 ± 5 mmHg vs 10 ± 2 mmHg, p = 0.014). Using Pearson’s correlation increased partition coefficient, ECV and pre-contrast myocardial T1 were significantly correlated with elevated LVEDP: r = 0.385 (p = 0.021), r = 0.355 (p = 0.029) and r = 0.335 (p = 0.04), respectively. In dichotomized analysis elevated LVEDP (≥13 mmHg, N = 13) was associated with significantly higher partition coefficient (0.51 ± 0.1 vs 0.45 ± 0.06, p = 0.004), higher ECV (0.30 ± 0.06 vs 0.26 ± 0.03, p = 0.002) and higher pre-contrast myocardial T1 (1040 ± 123 ms vs 983 ± 44 ms, p = 0.045) when compared to those with normal LVEDP ( < 13 mmHg, N = 26). Among patients with elevated LVEDP (average 19 ± 3 mmHg) there were no significant differences in partition coefficient, ECV and pre-contrast myocardial T1 between those with preserved or reduced LVEF. In contrast, there was a lack of significant association between post-contrast myocardial T1 and LVEDP. Conclusions Increased contrast partition coefficient, ECV and precontrast myocardial T1 were significantly associated with elevated LVEDP in patients with nonischemic cardiomyopathy thereby supporting the link between extracellular volume expansion and myocardial diastolic dysfunction. Quantitative tissue characterization combined with hemodynamic assessment underscores the value of cardiac MRI in the evaluation of nonischemic cardiomyopathy.

Comparison of myocardial mechanical properties in patients with dilated cardiomyopathy with and without acute heart failure

Background Dilated cardiomyopathy (DCM) is characterized by impaired myocardial contractile function and/or diastolic relaxation in the setting of dilated left or both ventricles that ultimately leads to heart failure (HF). In this study, we aim to compare myocardial mechanical properties in DCM patients with and without acute HF. Methods We prospectively enrolled 20 patients with DCM and 8 normal controls. Of the 20 patients with DCM, 10 patients had acute HF, defined as B-type natriuretic peptide (BNP) greater than 400 pg/ml, and the remaining 10 patients were determined to have no acute HF with BNP predominantly < 100 pg/ml. All subjects underwent cardiac MRI using 1.5-T scanner. Left ventricular end-diastolic pressure (LVEDP) was assessed noninvasively using mean left atrial circulation transit time normalized by RR-interval during the first pass perfusion imaging. LV circumferential strain (CST) and strain rate (CSR) of mid LV in short axis plane and longitudinal strain (LST) and strain rate (LSR) in 4-chamber view

Assessment of the trabeculated to compacted layer ratio of the left ventricle using short and longitudinal axis views in young patients presenting with ventricular dysrhythmias

Background Cardiac MRI (CMR) is increasingly being used clinically to assess the regional thickness of the compacted and trabeculated layers of the left ventricle (LV) in patients with suspected non-compaction. However, there is no consensus on whether the short or longitudinal axis view is preferred. We sought to assess the compacted and trabeculated layers in a young ventricular dysrhythmia population with age matched controls and to compare the measurements derived from both the short axis and longitudinal views. Methods The study consisted of 91 patients aged 20 to 49 years with ventricular dysrhythmia (VT) and 23 age-matched normal subjects. Regional wall thickness was assessed in short axis CMR cine views using a basal, mid and apical slice as well as in longitudinal views using the 2, 3, and 4 chamber view. The images were analyzed at end-diastole where the compacted layer was measured following the standardized 16-segment model. The apical segments were chosen 2 cm from the most apical portion of the LV for both short and longitudinal views. Within each wall segment, the most prominent trabeculation was measured, and the ratio of the trabeculated to compacted layer was compared between the VT group and controls in both views. Results The mean age was 38 years and 48 (42%) were men. In the VT group, 31 had premature ventricular contractions and 60 had ventricular tachycardia. VT patients were associated with larger LV end-diastolic volumes (89.7 ml/ m2 vs. 82.6 ml/m2, p=0.05), larger LV mass (58 g/m2 vs. 49.9 g/m2 p=0.001), and lower LV ejection fraction (50% vs. 60% p=0.015) as compared to the control group, respectively. Using the short axis view, the prevalence of trabeculated to compacted layer ratios ≥ 2.0, ≥ 2.1, ≥ 2.2 and ≥ 2.3 were 18.4%, 15.2%, 9.2% and 8.7% in the VT group, respectively while the prevalence in the control group was 13% for a ratio ≥ 2.0 with no segments ≥ 2.1. Using the same ratio cut points for the longitudinal view, the prevalence was 39.1%, 32.6%, 27.2% and 16.3% in the VT group and 43%, 39.1%, 21.7%, and 13% in controls, respectively. Conclusions In this young cohort, the trabeculated to compacted ratio ≥ 2.1 is only present in patients with ventricular dysrhythmias and not seen in normal controls using short axis planes, suggesting that increased thickness of the trabeculated layer in young subjects may be associated with the risk of ventricular dysrhythmia. However, the ratio ≥ 2.3 is equally prevalent in both groups using longitudinal planes. Our findings suggest that the short axis view may be more discriminatory then the longitudinal view in identifying normal subjects. More importantly, there is a clear need to standardize the assessment of the trabeculated to compacted layer ratio in clinical practice.

Regional thickness of the compacted and trabeculated layers of the left ventricle in young patients presenting with ventricular dysrhythmias, a Cardiac MRI study

Summary We sought to characterize the compacted and trabeculated layers of the left ventricle in patients presenting with ventricular dysrhythmias. Our results showed the dysrhythmic group had a statistically significant increase in the trabecular thickness as well as increased prevalence of trabecular to compacted thickness ratio. Background Cardiac MRI (CMR) is increasingly being used clinically to assess patients with ventricular dysrhythmias. Structural heart disease such as non-compaction is associated with dysrhythmic risk. Recent literature has characterized the regional thickness of the compacted and trabeculated layers of the left ventricle (LV) in normal subjects but not in patients with ventricular dysrhythmias. We sought to characterize the regional thickness of the compacted and trabeculated layers in a young community-based ventricular dysrhythmia population. Methods The study consisted of 98 patients under the age of 50 who had undergone CMR for ventricular dysrhythmia evaluation. These patients were compared to 26 agematched normal subjects. Patients who had known cardiomyopathy or myocardial infarction and those referred for AICD evaluation were excluded. SSFP cine images were used to assess cardiac structure and function. Intravenous gadolinium was given for late gadolinium enhancement (LGE) assessment. Regional thickness of the compacted and trabeculated layers were assessed in short axis cine views in 16 of the standard 17 segments at end-diastole. The trabeculated layer thickness and the ratio of trabeculated to compacted layer were compared between the dysrhythmic and control group. Results The mean age was 37 years and 56 (57%) were women. In the dysrhythmic group, 45 had premature ventricular contractions (PVC) and 55 had ventricular tachycardia (VT). Mean LV ejection fraction was 54±7% in the dysrhythmia group. LGE was present in 10 cases (10%) but none showed an infarct pattern. On average, the trabeculated layer was significantly larger in the dysrhythmic group than in control group: 3.7±1.4 mm vs. 3.0±1.0 mm at base (p<0.001), 4.6±1.3 mm vs. 3.4±1.0 mm in mid segment (p<0.001) and 5.4±1.5 mm vs. 3.5±0.9 mm in apical segment (p<0.001). The prevalence of trabeculated to compacted ratio ≥2.0 was 19% in controls, 20% in PVC group and 20% in VT group (p=NS). In contrast, the prevalence of a ratio ≥2.3 was 0 in controls, 16% in PVC group and 11% in VT group (p=0.01). The most common areas where the ratio exceeded 2.3 were mid-anterior, mid-anterolateral, apical anterior and apical lateral walls. Conclusions In this young cohort with ventricular dysrhythmias, the trabeculated layer was significantly thicker than agematched controls. Moreover, a trabeculated to compacted ratio greater than 2.3 was only present in patients with ventricular dysrhythmias, but not in normal controls. Our findings suggest that increased

The impact of coronary calcium score and cardiac risk factors on coronary endothelial function

Methods Following navigator coronary MRA scout images, the LAD lumen was imaged in short axis before and during the cold pressor test (CPT) in normal controls and subjects with one or more atherosclerotic risk factors using either supine breath-hold double inversion T2 weighted spin echo imaging (38 subjects, 1.5 T Siemens Sonata) or prone breath-hold SSFP retrospectively gated cine imaging with a 4 element phased array carotid coil (39 subjects, 1.5 T Siemens Avanto). The % change in LAD lumen area with CPT was calculated. CT calcium scores were available or obtained by protocol in each subject and those undergoing prone imaging also had coronary CT angiography to exclude coronary stenoses. Linear associations between continuous variables were measured by Pearson and Spearman (non-parametric) correlation coefficients. ROC analysis was also performed. A p-value < 0.05 was deemed statistically significant.

Differences in cardiac chamber volumes during prone versus supine imaging

Methods 11 subjects (27% females), age 54.9+17.3, 10 without clinical heart disease and 1 with coronary artery disease and normal ventricular function were imaged in supine and prone position using retrospectively gated SSFP cine imaging in contiguous short axis planes covering both right(RV) and left (LV) ventricles. Left and right ventricular end diastolic(EDV) and end systolic volumes(ESV), stroke volume(SV), cardiac output(CO) and ejection fraction (EF) were derived for both positions by a single highly experienced image analyst using Medis QMass MR software version 6.2.3 and compared using paired t tests. Heart rate (HR) and blood pressure (BP) were also compared. Results (Table 1) Both LV and RV EDV and SV were reduced in prone versus supine position with a small increase in heart rate and no change in EF or BP. Thus CO was also significantly reduced. LV ESV did not change significantly but RV ESV was reduced in prone position.