Abstract
Background
High-dose dobutamine stress magnetic resonance imaging (DS-MRI) is safe and feasible for the diagnosis of coronary artery disease (CAD) in humans. However, the assessment of cine scans relies on the visual interpretation of regional wall motion, which is subjective. Recently, Strain-Encoded MRI (SENC) has been proposed for the direct color-coded visualization of myocardial strain. The purpose of our study was to compare the diagnostic value of SENC to that provided by conventional wall motion analysis for the detection of inducible ischemia during DS-MRI.
Methods and Results
Stress induced ischemia was assessed by wall motion analysis and by SENC in 101 patients with suspected or known CAD and in 17 healthy volunteers who underwent DS-MRI in a clinical 1.5T scanner. Quantitative coronary angiography deemed as the standard reference for the presence or absence of significant CAD (≥50% diameter stenosis). On a coronary vessel level, SENC detected inducible ischemia in 86/101 versus 71/101 diseased coronary vessels (p<0.01 versus cine), and showed normal strain response in 189/202 versus 194/202 vessels with <50% stenosis (p=NS versus cine). On a patient level, SENC detected inducible ischemia in 63/64 versus 55/64 patients with CAD (p<0.05 versus cine), and showed normal strain response in 32/37 versus 34/37 patients without CAD (p=NS versus cine).Quantification analysis demonstrated a significant correlation between strain rate reserve (SRreserve) and coronary artery stenosis severity (r²=0.56, p<0.001), and a cut-off value of SRreserve=1.64 deemed as a highly accurate marker for the detection of stenosis≥50% (AUC=0.96, SE=0.01, 95% CI = 0.94–0.98, p<0.001).
Conclusions
The direct color-coded visualization of strain on MR-images is a useful adjunct for DS-MRI, which provides incremental value for the detection of CAD compared to conventional wall motion readings on cine images.
Keywords: myocardial strain response, Strain-Encoded MRI, dobutamine stress MRI, inducible ischemia, strain rate reserve, quantification analysis
Introduction
High-dose dobutamine/atropine stress cardiac magnetic resonance imaging (DS-MRI) is a safe and sensitive modality for the diagnosis of inducible myocardial ischemia in humans1,2. Regional wall motion abnormalities (WMA) during dobutamine stress, precede the development of ST-segment depression and anginal symptoms3 and are (i) accurate for the detection of anatomically significant CAD1,2 and (ii) predictive of clinical outcomes4. However, the assessment of cine images relies on the visual interpretation of regional wall motion, which is subjective and depends on the experience and training of the readers.
In the field of dobutamine stress echocardiography (DSE) objective methods have been introduced, which can quantify regional myocardial strain and strain rate. These techniques demonstrated incremental value for the objective detection of inducible ischemia during stress testing compared to conventional wall motion assessment5. However, due to its dependency on the acoustic window, echocardiography shows lower accuracy for the detection of inducible ischemia compared to DS-MRI1. Nonetheless, objective approaches for the quantification of myocardial strain during DS-MRI have been very limited so far.
Recently, Strain-Encoded-MRI (SENC) has been proposed for the objective color-coded evaluation of regional myocardial strain. The ability of SENC to quantify myocardial strain has been validated in experimental and in clinical settings6,7. The purpose of our study was to investigate regional myocardial strain and strain rate response during DS-MRI in healthy volunteers and in a cohort of subjects with suspected or known CAD. The results were compared with conventional wall motion readings, and coronary angiography deemed as the standard reference for the presence or absence of anatomically significant CAD.
Methods
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Patient population
Consecutive patients with suspected or known CAD (n=101) underwent DS-MRI prior to clinically-indicated coronary angiography due to chest pain of possible coronary origin (n=83) or exertional dyspnoea (n=18). Patients were excluded because of non-sinus rhythm, unstable angina, severe arterial hypertension (>200/120 mmHg), moderate or severe valvular disease, reduced ejection fraction<35%, and general contraindications to MR-examination. Cardiac medications were not discontinued. Seventeen age-matched healthy volunteers also underwent high-dose DS-MRI, in order to acquire normal values for myocardial strain and strain rate response. Exclusion criteria for control subjects comprised any history, symptoms, electrocardiographic signs or biochemical findings indicative of cardiovascular disease (normal brain natriuretric peptide and troponin T levels), evidence of systemic hypertension (baseline blood pressure>140/85), diabetes, abnormal glucose tolerance, hyperlipidemia (LDL>130mg/dl) or the presence of WMA at baseline or during DS-MRI. All procedures complied with the Declaration of Helsinki, were approved by our local ethic committee and all patients gave written informed consent.
Cardiovascular MR-Examination
Subjects were examined in a clinical 1.5-T whole-body MR-scanner Achieva system (Philips Medical Systems, Best, The Netherlands) using a 5-element cardiac phased-array receiver coil. Stress cardiovascular MR-images were acquired at rest and during a standardized high-dose dobutamine/atropine protocol1 involving short breath holds, and using a vector electrocardiogram for R-wave triggering. A 4-, 2-, and 3-chamber and 3 short-axis views (apical, midventricular, and basal) were used. Electrocardiographic rhythm and symptoms were monitored continuously, and blood pressure was measured every 3 minutes. Dobutamine was infused intravenously during 3-minute stages, at incremental doses of 10, 20, 30, and 40 µg per kilogram of body weight per minute until at least 85% of the age-predicted heart rate was reached (220-age). If necessary, the last dobutamine stage was prolonged accordingly (typically from 3.0 to ∼3.5–4.0 minutes), in order to allow for the acquisition of all cine and SENC sequences. If at the peak dose of dobutamine infusion the target heart rate was not achieved, atropine was administrated in 0.25-mg increments up to a maximal dose of 2.0 mg. Stress testing was discontinued when the target heart rate was achieved, or when one of the following occurred: new or worsening WMA in at least one segment, severe chest pain or dyspnea, decrease in systolic blood pressure of more than 40mmHg, severe arterial hypertension (blood pressure>220/120 mm Hg), or severe arrhythmias. In the absence of ischemia, failure to attain 85% of age-predicted maximal heart rate was considered as a non-diagnostic result.
Cine Imaging
A steady-state-free precession sequence was used to obtain the cine images of the 4-, 2-, and 3-chamber view and 3 short-axis planes (apical, mid-ventricular and basal) with an 8mm slice thickness. Typical parameters were: field-of-view (FOV)=350×350mm2, matrix size=160×160, flip angle=60°, repetition time/echo time (TR/TE)=2.8/1.4ms and acquired voxel size=2.2×2.2×8mm3. The temporal resolution was 21–28ms and the total scan duration was 7–12s. Cine images were acquired at baseline and acquisitions were repeated during each stage including the peak level of dobutamine/atropin administration.
SENC acquisition and implementation
The SENC pulse sequence is a modified spatial modulation of magnetization (SPAMM) tagging pulse sequence and an extension of the stimulated echo acquisition mode (STEAM) sequence8. In contrast to conventional tagging, where the tagging modulation gradient is applied in the phase- or frequency-encoding direction, in SENC the gradient is applied in the slice selection direction, orthogonal to the imaging plane. Briefly, radiofrequency (RF) pulses with ramped flip angles are applied to compensate for the tag fading caused by longitudinal relaxation and to maintain constant myocardial signal intensity throughout the cardiac cycle. By combining SENC magnitude images at 2 different tuning frequencies (low- and high-tuning), a SENC strain map is computed, which is then overlaid pixel-by-pixel on the anatomical image. By this way color-coded SENC functional images (maximum contraction illustrated as red, and lack of contraction illustrated as white) are generated. SENC images of the 4-, 2-, and 3-chamber view and 3 short-axis planes were acquired at the same plane levels of that used for cine scans with 10mm slice thickness. Typical parameters were: FOV=350×350mm2, matrix size=80×80, flip angle=30°, TR/TE=25/0.9ms and acquired voxel size=4.4×4.4×10mm3. The temporal resolution was set at 25ms during baseline and at 15ms during peak dobutamine infusion. The number of cardiac phases was adapted accordingly (typically 26–38 during baseline and 22–27 during stress), to cover ∼85–90% of the cardiac cycle. The total scan duration was 10–14s during baseline and 5–8s during peak dobutamine stress. Strain-encoded images were acquired at baseline and during the peak level of dobutamine/atropin administration.
Visual interpretation of cine and SENC images
For interpretation of wall motion, corresponding rest and peak stress cine images were displayed using View Forum software (Philips Medical Systems, Best, The Netherlands). Segmental wall motion was graded semi-quantitatively using a 3-point scale (0=normal wall motion, 1=hypokinesia, 2=a- or dyskinesia)9–11, and inducible ischemia was considered present in cases of new or worsening WMA of≥1 grade during stress. In addition, resting WMA that improved during intermediate stress and subsequently deteriorated during peak stress (‘biphasic response’), were considered as diagnostic of inducible ischemia. Corresponding baseline and peak stress SENC images were displayed using Diagnosoft SENC (Version 1.06, Diagnosoft Inc., Palo Alto, CA), a software package, which allows the color-encoded interpretation of myocardial strain on SENC images. An analysis using HARP algorithms was not performed with this software. Similar to wall motion analysis, a 16-segment model was used for the analysis of regional myocardial strain, which was graded semiquantitatively using a 3-point color scale (0=normal strain corresponding to red myocardium, 1=reduced strain corresponding to faded orange/yellowish myocardium, 2=severely reduced or absent strain corresponding to faded yellow/white myocardial tissue on strain-encoded images). For visual interpretation differences in the color-scale were carefully evaluated at baseline and during peak stress in corresponding myocardial segments, and inducible ischemia was considered present in case of strain reduction of≥1 grade. Cine and SENC images from short and long axis views were evaluated visually by 2 independent observers with at least 5 years experience in cardiovascular imaging (G.K. and D.L.), in a clinical high volume cardiac MR-laboratory (∼25–30 DS-MRI studies per week). Both observers evaluated cine and SENC images separately, and were blinded to clinical data and to the results of invasive angiography.
Quantitative analysis of circumferential strain on SENC images
Because with SENC, the tagging modulation gradient is applied in the slice selection direction, for quantitative estimation of circumferential strain, round regions of interest (4mm in diameter) were placed in basal-, mid- and apical-segments of the left ventricle from the 4-, 2-, and 3-chamber view at baseline and during peak dobutamine stress. For each segment, the temporal course of regional myocardial strain was registered throughout the cardiac cycle, and quadratic interpolation was used to estimate the time derivative of the systolic strain and to calculate the peak systolic strain (Ecc expressed in %) and the peak systolic strain rate (SR expressed in 1/s)12. The ratios: Sreserve=Speak-dobutamine/Sbaseline and SRreserve=SRpeak-dobutamine/SRbaseline were generated in order to evaluate strain and strain rate response, respectively during DS-MRI. Normal values were assessed in healthy volunteers, and we anticipated that reduced values for strain and strain rate reserve would be indicative of stress induced ischemia in patients with coronary artery disease. Registration of regional myocardial strain in a patient without CAD (Fig. 1a–b) showed that myocardial strain remained unchanged during peak stress (Sreserve∼1, curved black arrow in Fig. 1c), while strain rate increased by ∼2-fold (SRreserve∼2, curved red arrow in Fig. 1c). On the other hand, in an ischemic segment, which demonstrated stress induced hypokinesia, in a patient with coronary artery disease (Fig. 1d–e), myocardial strain decreased during peak dobutamine stress (Sreserve<1, black arrow in Fig. 1f), while strain rate remained almost unchanged (SRreserve∼1, dotted red circle in Fig. 1f).
Figure 1. Illustration of the strain and strain rate response during inotropic stimulation.
Quadratic interpolation was used to generate strain curves in a patient without significant CAD by coronary angiography (a–c) and in a subject with single-vessel CAD of the LAD (d–f). In the first subject myocardial strain rate increased by ∼2-fold (curved red arrow in c), and peak systolic strain remained constant (black arrow in c), while in the ‘ischemic territory’, myocardial strain decreased during peak stress (black arrow in f), and strain rate remained unchanged (dotted red circle in f).
Quantitative coronary angiography and comparison to cine and SENC images
Coronary angiography deemed as the standard reference for the detection of CAD and angiograms were obtained in all patients within 3 weeks from the DS-MRI study. The procedure was done according to the angiographic guidelines and at least two orthogonal views of every major coronary vessel and its side branches were acquired. Quantification of coronary artery stenosis was performed off-line using commercially available software (Centricity QCA, GE Medical Systems, Milwaukee, WI). Myocardial segments evaluated on MR images were assigned to coronary vessels according to AHA guidelines13 (see Appendix 1 for details).
Statistical analysis
Statistical analysis was performed using commercially available software (SPSS, version 11.5 for Windows; SPSS, Chicago, Ill) and data are presented as mean±standard deviation. Agreement between the 2 observers interpreting cine and SENC images was assessed using kappa statistics14. Intra- and interobserver variability for quantification analysis of circumferential strain were calculated by repeated analysis of 40 representative regions of interest for strain and strain rate and was expressed in percentages relative to the strain and strain rate values. The readings were separated by 8 weeks to minimize recall bias. Differences in sensitivity and specificity were tested using Mc-Nemar χ²-square statistics15. Differences in strain and strain rate by stenosis severity (0%; 1–20%; 21–40%; 41–60%; 61–80% and 81–100% i.e 6 groups) were assessed with clustered regression and using Bonferroni's correction for multiple comparisons. The relation between Sreserve and SRreserve with the stenosis severity by angiography was assessed by second order polynomial regression analysis. According to quantitative coronary angiography each segment was dichotomously classified as non-ischemic or ischemic (i.e. supplied by a vessel with <50% or ≥50% stenosis) and receiver operating characteristics (ROC) were used to determine the diagnostic value of Sreserve and SRreserve for the detection of segmental ischemia. To compare ROC curves, pair-wise comparisons of areas under the curve was assessed, using Lehmann curves to account for the clustered nature of our data16. Differences were considered statistically significant at p<0.05.
Results
Demographic and haemodynamic parameters
Diagnostic DS-MRI examinations (positive for ischemia or negative but with achievement of the target heart rate) were achieved in 101 of 114 consecutive patients. Thirteen patients were excluded from analysis for the following reasons: in 4 cases, the target heart rate was not achieved in the absence of ischemia and despite maximal pharmacologic stimulation; in 4 further patients stress testing was discontinued due to asymptomatic increase in systolic blood pressure >220mmHg (n=1) or on patient’s request (nausea in 2 cases and panic attack in one case), while repeated extrasystoles during stress resulted in non-diagnostic image quality in 5 patients (4%). In all 17 healthy subjects the target heart rate was achieved and diagnostic images were acquired. Overall, no severe adverse events were recorded. The demographics of our patients are summarized in Table 1.
Table 1.
Demographic and haemodynamic characteristics.
| Parameters | Healthy volunteers | Patients |
|---|---|---|
| Demographics | ||
| No of subjects | n=17 | n=101 |
| Age (yrs) | 61±6 | 63±11 |
| Male sex | 12/17 (71%) | 75/101 (74%) |
| Coronary risk factors | ||
| Arterial hypertension | 0/17 (0%) | 89/101 (88%) |
| Hypercholesterolemia | 0/17 (0%) | 66/101 (65%) |
| Diabetes mellitus | 0/17 (0%) | 22/101 (22%) |
| Family history | 0/17 (0%) | 30/101 (30%) |
| Smoker | 0/17 (0%) | 25/101 (25%) |
| Prior percutaneous coronary intervention | 0/17 (0%) | 35/101 (35%) |
| Prior myocardial infarction | 0/17 (0%) | 25/101 (25%) |
| Prior coronary artery bypass grafting | 0/17 (0%) | 5/101 (5%) |
| Reasons for referral | ||
| Chest pain | (n/a) | 83 (82%) |
| Exertional dyspnoea | (n/a) | 18 (18%) |
| Baseline haemodynamics | ||
| Mean blood pressure (mmHg) | 92±6 | 89±12 |
| Heart rate (1/min) | 67±10 | 66±18 |
| Double product (mmHg/min) | 8339±1233 | 8298±2604 |
| Peak stress haemodynamics | ||
| Mean blood pressure (mmHg) | 92±12 | 109±22 |
| Heart rate (1/min) | 149±6 | 138±11 |
| Double product (mmHg/min) | 19514±2275 | 20029±4327 |
| Baseline MR-parameters | ||
| LV end-diastolic diameter (mm) | 49±4 | 52±5 |
| LV end-systolic diameter (mm) | 32±4 | 35±6 |
| LV end-diastolic volume (ml) | 144±29 | 147±39 |
| LV end- systolic volume (ml) | 47±14 | 58±28 |
| Baseline ejection fraction (%) | 67±6 | 61±10 |
| No of segments with resting wall motion abnormalities by cine | 0/272 (0%) | 72/1616 (4%) |
Data presented as number of patients or as mean±standard deviation; n/a indicates not applicable.
Results of coronary angiography
Coronary angiography showed ≥50% stenosis in 101 of 303 coronary vessels (33%), including 44 vessels with LAD, 30 with LCX and 27 with RCA lesions. On a patient level 64 (63%) patients had CAD by angiography, including 32 with single-vessel, 27 with 2-vessel and 5 with 3-vessel disease.
Detection of CAD by SENC versus cine imaging
Analysis by vessels
On a coronary vessel level, SENC detected inducible ischemia in 86/101 versus 71/101 diseased coronary vessels (sensitivity of 85% by SENC versus 70% by cine, p<0.01), and showed normal strain response in 189/202 versus 194/202 vessels with <50% stenosis (specificity of 94% by SENC versus 96% by cine, p=NS), (Table 2).
Table 2.
Detection of coronary artery stenosis during dobutamine stress MRI.
| Sensitivity | Specificity | Accuracy | |
|---|---|---|---|
| Assessment by coronary vessels (n=303) | |||
| Cine images (all vessels) | 70% (71/101) | 96% (194/202) | 87% (265/303) |
| Single-vessel CAD (32 diseased vessels) | 88% (28/32) | 94% (60/64) | 92% (88/96) |
| Multi-vessel CAD (69 diseased vessels) | 62% (43/69) | 100% (27/27) | 73% (70/96) |
| SENC (all vessels) | 85%† (86/101) | 94% (189/202) | 91%* (275/303) |
| Single-vessel CAD (32 diseased vessels) | 97% (31/32) | 97% (62/64) | 97% (93/96) |
| Multi-vessel CAD (69 diseased vessels) | 80% (55/69) | 81% (22/27) | 80% (77/96) |
| Assessment by patients (n=101) | |||
| Cine imaging (all patients) | 86% (55/64) | 92% (34/37) | 88% (89/101) |
| Patients with resting WMA by cine (n=25) | 83% (20/24) | 100% (1/1) | 84% (21/25) |
| Patients without resting WMA by cine (n=76) | 86% (35/40) | 92% (33/36) | 89% (68/76) |
| SENC (all patients) | 98%* (63/64) | 86% (32/37) | 94% (95/101) |
| Patients with resting WMA by cine (n=25) | 100% (24/24) | 100% (1/1) | 100% (25/25) |
| Patients without resting WMA by cine (n=76) | 98% (39/40) | 86% (31/36) | 92% (70/76) |
Data are presented as number of patients or percentages.
p<0.05 and
p<0.01 for Strain-encoded MRI versus wall motion analysis on cine images (by Mc-Nemar χ2-square test).
Analysis by patients
On a patient level, SENC detected inducible ischemia in 63/64 versus 55/64 patients with CAD (sensitivity of 98% by SENC versus 86% by cine, p<0.05), and showed normal strain response in 32/37 versus 34/37 patients without CAD (specificity of 86% by SENC versus 92% by cine, p=NS). Thus, SENC detected abnormal strain response in 8 additional patients, who were missed by cine, while no patients were correctly diagnosed with CAD by cine and missed by SENC. Furthermore, cine and SENC showed similar diagnostic characteristics for ischemia detection in patients with/without resting WMA (Table 2).
Fig. 2 illustrates a patient, where wall motion analysis correctly detected stress induced ischemia in the LAD, but in contrast to SENC, overlooked ischemia in the LCX perfusion territory.
Figure 2. Strain-Encoded MRI detects ischemic myocardium in the lateral wall, which is overlooked by wall motion assessment.
Wall motion analysis detected stress induced ischemia in the left ventricular apex (f, red arrows) and in the mid-ventricular septum (h, red arrow). SENC confirmed the presence of inducible ischemia in these segments (g, red arrows) and additionally detected reduced strain response in the mid-ventricular lateral wall (i, yellow arrows). Coronary angiography yielded 2-vessel CAD, with high grade lesions (hatched circles) in both the left anterior descending (e), and in the left circumflex coronary artery (j).
Quantitative assessment of strain response during DS-MRI
Segments in healthy volunteers and in non-ischemic segments (supplied by vessels with <50% diameter stenosis) had significantly higher strain/strain rate response during inotropic stimulation compared to ischemic segments (Fig. 3a and c). Further differentiation showed that ischemic segments without new/worsening WMA during stress had significantly higher strain and strain rate response compared to those with WMA (Fig. 3b and d). Weak, albeit significant, correlations were observed between the stenosis severity on coronary angiograms with SRreserve and with Sreserve (p<0.001 for both), (Fig. 4a and d). Interestingly, segments with new/worsening WMA on cine images were located at the bottom right corner (purple dots within the hatched red circles in Fig. 4a and d), showing higher degree of stenosis (87±9% versus 68±11 diameters stenosis, p<0.001) and lower SRreserve (0.92±0.33/s versus 1.41±0.33/s, p<0.001). Thus, a considerable amount of segments was classified as ‘normal’ by wall motion readings had impaired SRreserve by quantitative analysis (dotted red square in Fig. 4d). SRreserve was reduced already with moderate coronary stenosis (41–60%), while a significant reduction in Sreserve required higher grade stenotic lesions (>60%), (Fig. 4b and e). ROC analysis showed that SRreserve, using a cut-off value of SRreserve>1.64 is a highly accurate parameter for the detection of anatomically significant coronary artery stenosis≥50% (AUC=0.96, SE=0.01, 95% CI = 0.94–0.98, p<0.001) (Fig. 4c and f), (Table 2). Pair-wise comparison of ROC curves yielded significantly higher accuracy for detection of CAD by SRreserve compared to Sreserve (Difference between areas=0.15, SE=0.02, 95%CI=0.08 to 0.16, p<0.001). This difference remained statistically significant when using Lehmann curves to account for the clustered nature of our data (p<0.001).
Figure 3. Quantification of myocardial strain/strain rate.
Normal and non-ischemic segments (supplied by vessels with <50% diameter stenosis) had significantly higher strain and strain rate response during inotropic stimulation compared to ischemic segments (a and c, p<0.001 for both). Furthermore, ischemic segments without new/worsening WMA during stress had significantly higher strain/strain rate response compared to those with new/worsening WMA (b and d, p<0.001).
Figure 4. Correlation between myocardial strain and stenosis severity.
Scatter plots showing the relation of myocardial strain to stenosis severity demonstrated weak non-linear correlations for Sreserve (a), (r2=0.38), and for SRreserve (d), (r2=0.56). SRreserve was reduced already with moderate coronary stenosis (e), while a significant reduction in Sreserve required higher grade stenotic lesions (b). ROC analysis showed that SRreserve, with a cut-off value of SRreserve>1.64, is a highly accurate parameter for the detection of coronary artery stenosis≥50% (f).
Differentiating between segments with (n=72 in 25 patients with prior myocardial infarction) versus segments without resting WMA, SRreserve correlated in both groups with stenosis severity (Fig. 5a and c), and detected anatomically significant CAD with similarly high sensitivity and overall accuracy (Fig. 5b and d, and Table 2).
Figure 5. SRreserve for the detection of inducible ischemia in segments with resting WMA.
SRreserve correlated significantly with stenosis severity both in segments with and without baseline WMA, (a and c). Furthermore, in segments with resting WMA, the detection of anatomically significant CAD was similar to that in segments without resting WMA (b and d).
Observer agreement and variabilities
Agreement between observers interpreting wall motion on cine images and myocardial strain on SENC images was 87% (κ=0.78) versus 84% (κ=0.75), respectively on test positivity versus test negativity, and 86% (κ=0.77) versus 83% (κ=0.73), respectively for visual scoring of regional wall motion on cine and of color-shades on SENC images. SENC allowed for reproducible quantification of regional myocardial strain, showing relatively low intra- and interobserver variabilities (8.1% and 12.2% for strain and 9.7% and 13.1% for strain rate, respectively).
Discussion
The results of our study comprise (i) the ability of strain-encoded MRI to objectively assess and quantify regional myocardial strain during dobutamine stress testing and (ii) enhanced sensitivity of strain-encoded imaging for the detection of anatomically significant CAD compared to conventional wall motion reading. In particular impaired strain rate response is a highly sensitive marker of inducible ischemia, which seems to occur early in the ischemic cascade preceding the development of inducible WMA.
Previous studies utilized objective echocardiographic methods like tissue Doppler and strain rate imaging to evaluate strain response during dobutamine stress5, demonstrating higher accuracy than 2D-echocardiography for the detection of ischemic myocardium. These techniques are generally associated with higher temporal resolution of >100frames/sec, compared to that provided by regular tagged MR-sequences (<30 frames/s). The acquisition of tagged images on the other hand requires long series of breath holds, which may not be feasible in some patients. However, echocardiographic strain imaging is associated with problems that are inherent in the use of Doppler, and are related to signal noise and angle dependency, particularly in the apical segments17. Both imaging modalities however, have potential for technical improvements. Thus, speckle tracking methodologies have been recently introduced, which avoid the angle dependency of the conventional Doppler techniques18. Newer MR-techniques on the other hand, allow the quantitative assessment of regional myocardial strain within a single heart beat with fast SENC sequences, obviating the need of prolonged breathholds7. Furthermore, correction of through plane motion was recently demonstrated with SENC19. This may allow for even more accurate estimation of regional myocardial strain in future studies.
Although DS-MRI is a feasible and widely available technique for the detection of stress induced ischemia, objective approaches for the quantification of strain response during DS-MRI have been very limited so far. Paetsch and collegues20 previously reported on the value of a diastolic parameter to detect ischemia during low-dose inotropic stimulation. Kuijpers et al21 on the other hand, demonstrated that the use of myocardial tagging during high-dose DS-MRI adds incremental value to conventional wall motion analysis for the detection of inducible ischemia. However, the analysis of tagged MR-images was purely based on visual assessment of the in-plane movement of the tags, while a quantification analysis of myocardial strain was not performed. Despite the use of newly introduced harmonic phase (HARP) concepts to analyze regional strain, the time-spent required for the analysis of tagged images is still higher than that needed for SENC7.
In our study, the direct color-coded visualization of myocardial strain allowed for the objective detection of ‘subtle’ differences in regional myocardial strain during high-dose DS-MRI, yielding significantly higher sensitivity for the detection of CAD, both by coronary vessely and by patients, compared to conventional wall motion readings. Furthermore, quantification of myocardial strain during the inotropic challenge showed that particularly strain rate is a highly sensitive marker for the detection of ischemic myocardium, which may precede the development of WMA in the ischemic cascade. Thus, strain rate reserve already decreased with moderate coronary lesions and in the absence of new WMA. This is in accordance with data from echocardiographic studies, demonstrating a significant relation between strain rate and functional significance of coronary artery stenosis22 and with experimental data, which showed that strain rate but not strain values are related to regional contractility independent of the heart rate23. In addition, inducible diastolic dysfunction during intermediate stages of inotropic stimulation (time to peak untwist and diastolic rotation velocity)20 may be another early marker of inducible ischemia in patients with CAD, which may merits further investigation in future studies.
Quantification of myocardial strain was assessed in the circumferential direction from long-axis SENC images, because circumferential shortening is considered to be more sensitive compared to longitudinal and radial strain for the detection of abnormal myocardial function in human24–26 and in experimental studies27. In agreement with these observations, in our study, the circumferential strain rate allowed for the more precise detection of ischemic myocardium during inotropic stress compared with the assessment of wall thickening by the human eye, which is more equivalent to radial strain. Especially in segments with resting WMA, where the visual recognition of worsening wall motion represents a challenge even for experienced readers, quantification of circumferential strain rate identified subtle changes in contractile reserve and proved almost as accurate for the detection of ischemia, as in segments without resting WMA.
In accord with general recommendations for stress testing, dobutamine infusion was discontinued when newly developed or of worsening WMA occurred. Therefore, in patients with multi-vessel disease, segments assigned to the ‘less severe’ coronary stenosis may have been classified as ‘normal’, although supplied by >50%-stenotic coronary lesions. In agreement with this assumption, sensitivity of both cine and strain-encoded imaging was relatively low when assessed on a coronary territory vessel in patients with multi-vessel disease. However, when all 3 territories were considered on a patient level, which is most meaningful for clinical decision making, sensitivity was higher and in agreement with the current literature.
Limitations
The number of healthy volunteers was relatively low, while coronary angiography was not performed to exclude CAD. However, due to the high diagnostic accuracy of DS-MRI on a patient level, in combination with the low risk profile of these subjects, it is unlikely that subclinical atherosclerotic disease influenced the calculations on strain/strain rate reserve in these subjects. Furthermore, the majority of our patients was male and showed high-grade coronary lesions by angiography. This may represent a potential limitation for extrapolation of our findings in a population with lower pre-test probability for CAD. As part of the present protocol, in contrast to cine images, SENC was not performed during every stage of inotropic stimulation. Therefore, segments exhibiting a biphasic response were only considered on cine, but not on SENC images. Furthermore, both spatial and temporal resolution of SENC was relatively low, while the estimation of radial strains was not possible with this sequence. Therefore, our method comparison may suffer from a methodological mismatch, in that visual assessment of wall motion may be most sensitive to changes in radial thickening, while SENC was used to quantify only circumferential shortening. In addition the apical cap (segment 17) was excluded from analysis, which may have lead to an underestimation of both SENC and cine for the detection of inducible ischemia. The significance of the current findings however, now warrants the initiation of multi-center trials, where technical improved SENC sequences (i.e. with higher temporal and spatial resolution and with through-plane motion tracking19) should be applied in a large population and possibly during intermediate stages of inotropic stimulation. Non-discontinuation of anti-anginal medications prevented the achievement of the target heart rate in 4 of 114 (<5%) consecutive patients despite maximal pharmacologic stimulation. However, the target-rate rule was applied in all 101 subjects, which constituted our patient population. The normal LV is characterized both morphologically and functionally by a high degree in regional nonuniformity, attributed to variations in transmural fiber orientation and local differences in ventricular morphology28. The investigation of regional heterogeneity in strain/strain response on SENC images, however, was beyond the scope of the present study and merits further investigation in future trials.
Conclusions
Due to its tomographic nature, MRI allows for sequential acquisition of identical slices (each segment at baseline serves as its own control during peak stress), providing the reproducible and accurate assessment of strain rate response during dobutamine stress testing. This is to our knowledge the first study to demonstrate that the direct color-coded visualization of strain on MR-images is a useful adjunct for DS-MRI, which provides incremental value for the detection of CAD compared to conventional wall motion analysis on cine images. Particularly an impaired strain rate reserve appears to be a highly sensitive and accurate marker for the detection of ischemic myocardium during inotropic stimulation.
CLINICAL PERSPECTIVE.
The assessment of inducible regional wall motion abnormalities during high-dose dobutamine stress magnetic resonance imaging (DS-MRI) is an established clinical method for the evaluation of patients with coronary artery disease (CAD). However, the assessment of cine images relies on subjective interpretation of regional wall motion, and objective approaches for the detection of inducible ischemia with DS-MRI are lacking. We used Strain-Encoded-MRI (SENC) for the objective color-coded evaluation of regional myocardial strain during DS-MRI in a cohort of subjects with suspected or known CAD (n=101). SENC yielded significantly higher sensitivity than visual wall motion analysis (85% vs. 70%, respectively, p<0.01) for detection of obstructive CAD as assessed by coronary angiography. In addition, SENC identified significant CAD in 8 subjects that were missed by visual wall motion analysis (p<0.05). Quantification analysis demonstrated that impaired strain rate response during DS-MRI is a highly sensitive marker of inducible ischemia, which seems to occur early in the ischemic cascade than wall motion abnormalities. The increased sensitivity of SENC may help with the evaluation of patients with known or suspected CAD.
Acknowledgements
Dr. Nael F. Osman is a founder and shareholder in Diagnosoft Inc., the software used for the analysis of the acquired SENC images.
Funding Sources: None.
Footnotes
Conflicts of Interest Disclosures: None.
REFERENCES
- 1.Nagel E, Lehmkuhl HB, Bocksch W, Klein C, Vogel U, Frantz E, Ellmer A, Dreysse S, Fleck E. Noninvasive diagnosis of ischemia-induced wall motion abnormalities with the use of high-dose dobutamine stress MRI: comparison with dobutamine stress echocardiography. Circulation. 1999;99:763–770. doi: 10.1161/01.cir.99.6.763. [DOI] [PubMed] [Google Scholar]
- 2.Paetsch I, Jahnke C, Wahl A, Gebker R, Neuss M, Fleck E, Nagel E. Comparison of dobutamine stress magnetic resonance, adenosine stress magnetic resonance, and adenosine stress magnetic resonance perfusion. Circulation. 2004;110:835–842. doi: 10.1161/01.CIR.0000138927.00357.FB. [DOI] [PubMed] [Google Scholar]
- 3.Upton MT, Rerych SK, Newman GE, Port S, Cobb FR, Jones RH. Detecting abnormalities in left ventricular function during exercise before angina and ST-segment depression. Circulation. 1980;62:341–349. doi: 10.1161/01.cir.62.2.341. [DOI] [PubMed] [Google Scholar]
- 4.Jahnke C, Nagel E, Gebker R, Kokocinski T, Kelle S, Manka R, Fleck E, Paetsch I. Prognostic value of cardiac magnetic resonance stress tests: adenosine stress perfusion and dobutamine stress wall motion imaging. Circulation. 2007;115:1769–1776. doi: 10.1161/CIRCULATIONAHA.106.652016. [DOI] [PubMed] [Google Scholar]
- 5.Voigt JU, Exner B, Schmiedehausen K, Huchzermeyer C, Reulbach U, Nixdorff U, Platsch G, Kuwert T, Daniel WG, Flachskampf FA. Strain-rate imaging during dobutamine stress echocardiography provides objective evidence of inducible ischemia. Circulation. 2003;107:2120–2126. doi: 10.1161/01.CIR.0000065249.69988.AA. [DOI] [PubMed] [Google Scholar]
- 6.Pan L, Stuber M, Kraitchman DL, Fritzges DL, Gilson WD, Osman NF. Real-time imaging of regional myocardial function using fast-SENC. Magn Reson Med. 2006;55:386–395. doi: 10.1002/mrm.20770. [DOI] [PubMed] [Google Scholar]
- 7.Korosoglou G, Youssef AA, Bilchick KC, Ibrahim el S, Lardo AC, Lai S, Osman NF. Real-time fast strain-encoded magnetic resonance imaging to evaluate regional myocardial function at 3.0 Tesla: Comparison to conventional tagging. J Magn Reson Imaging. 2008;27:1012–1018. doi: 10.1002/jmri.21315. [DOI] [PubMed] [Google Scholar]
- 8.Frahm J, Hanicke W, Bruhn H, Gyngell ML, Merboldt KD. High-speed STEAM MRI of the human heart. Magn Reson Med. 1991;22:133–142. doi: 10.1002/mrm.1910220114. [DOI] [PubMed] [Google Scholar]
- 9.Korosoglou G, Dubart AE, DaSilva KG, Jr, Labadze N, Hardt S, Hansen A, Bekeredjian R, Zugck C, Zehelein J, Katus HA, Kuecherer H. Real-time myocardial perfusion imaging for pharmacologic stress testing: added value to single photon emission computed tomography. Am Heart J. 2006;151:131–138. doi: 10.1016/j.ahj.2005.02.046. [DOI] [PubMed] [Google Scholar]
- 10.Korosoglou G, Hansen A, Hoffend J, Gavrilovic G, Wolf D, Zehelein J, Haberkorn U, Kuecherer H. Comparison of real-time myocardial contrast echocardiography for the assessment of myocardial viability with fluorodeoxyglucose-18 positron emission tomography and dobutamine stress echocardiography. Am J Cardiol. 2004;94:570–576. doi: 10.1016/j.amjcard.2004.05.018. [DOI] [PubMed] [Google Scholar]
- 11.Korosoglou G, da Silva KG, Jr, Labadze N, Dubart AE, Hansen A, Rosenberg M, Zehelein J, Kuecherer H. Real-time myocardial contrast echocardiography for pharmacologic stress testing: is quantitative estimation of myocardial blood flow reserve necessary? J Am Soc Echocardiogr. 2004;17:1–9. doi: 10.1016/j.echo.2003.08.004. [DOI] [PubMed] [Google Scholar]
- 12.Ennis DB, Epstein FH, Kellman P, Fananapazir L, McVeigh ER, Arai AE. Assessment of regional systolic and diastolic dysfunction in familial hypertrophic cardiomyopathy using MR tagging. Magn Reson Med. 2003;50:638–642. doi: 10.1002/mrm.10543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS. 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. Circulation. 2002;105:539–542. doi: 10.1161/hc0402.102975. [DOI] [PubMed] [Google Scholar]
- 14.Kramer MS, Feinstein AR. Clinical biostatistics. LIV. The biostatistics of concordance. Clin Pharmacol Ther. 1981;29:111–123. doi: 10.1038/clpt.1981.18. [DOI] [PubMed] [Google Scholar]
- 15.Mantel N, Fleiss JL. The equivalence of the generalized McNemar tests for marginal homogeneity in 2(3) and 3(2) tables. Biometrics. 1975;31:727–729. [PubMed] [Google Scholar]
- 16.Gönen M. Analyzing Receiver Operating Characteristic Curves with SAS 1st printing. 2007 [Google Scholar]
- 17.Sutherland GR, Di Salvo G, Claus P, D'Hooge J, Bijnens B. Strain and strain rate imaging: a new clinical approach to quantifying regional myocardial function. J Am Soc Echocardiogr. 2004;17:788–802. doi: 10.1016/j.echo.2004.03.027. [DOI] [PubMed] [Google Scholar]
- 18.Delgado V, Ypenburg C, van Bommel RJ, Tops LF, Mollema SA, Marsan NA, Bleeker GB, Schalij MJ, Bax JJ. Assessment of left ventricular dyssynchrony by speckle tracking strain imaging comparison between longitudinal, circumferential, and radial strain in cardiac resynchronization therapy. J Am Coll Cardiol. 2008;51:1944–1952. doi: 10.1016/j.jacc.2008.02.040. [DOI] [PubMed] [Google Scholar]
- 19.Ibrahim el SH, Stuber M, Fahmy AS, Abd-Elmoniem KZ, Sasano T, Abraham MR, Osman NF. Real-time MR imaging of myocardial regional function using strain-encoding (SENC) with tissue through-plane motion tracking. J Magn Reson Imaging. 2007;26:1461–1470. doi: 10.1002/jmri.21125. [DOI] [PubMed] [Google Scholar]
- 20.Paetsch I, Foll D, Kaluza A, Luechinger R, Stuber M, Bornstedt A, Wahl A, Fleck E, Nagel E. Magnetic resonance stress tagging in ischemic heart disease. Am J Physiol Heart Circ Physiol. 2005;288:H2708–H2714. doi: 10.1152/ajpheart.01017.2003. [DOI] [PubMed] [Google Scholar]
- 21.Kuijpers D, Ho KY, van Dijkman PR, Vliegenthart R, Oudkerk M. Dobutamine cardiovascular magnetic resonance for the detection of myocardial ischemia with the use of myocardial tagging. Circulation. 2003;107:1592–1597. doi: 10.1161/01.CIR.0000060544.41744.7C. [DOI] [PubMed] [Google Scholar]
- 22.Weidemann F, Jung P, Hoyer C, Broscheit J, Voelker W, Ertl G, Stork S, Angermann CE, Strotmann JM. Assessment of the contractile reserve in patients with intermediate coronary lesions: a strain rate imaging study validated by invasive myocardial fractional flow reserve. Eur Heart J. 2007;28:1425–1432. doi: 10.1093/eurheartj/ehm082. [DOI] [PubMed] [Google Scholar]
- 23.Weidemann F, Jamal F, Sutherland GR, Claus P, Kowalski M, Hatle L, De Scheerder I, Bijnens B, Rademakers FE. Myocardial function defined by strain rate and strain during alterations in inotropic states and heart rate. Am J Physiol Heart Circ Physiol. 2002;283:H792–H799. doi: 10.1152/ajpheart.00025.2002. [DOI] [PubMed] [Google Scholar]
- 24.Tanaka H, Oishi Y, Mizuguchi Y, Emi S, Ishimoto T, Nagase N, Tabata T, Yamada H, Oki T. Three-dimensional evaluation of dobutamine-induced changes in regional myocardial deformation in ischemic myocardium using ultrasonic strain measurements: the role of circumferential myocardial shortening. J Am Soc Echocardiogr. 2007;20:1294–1299. doi: 10.1016/j.echo.2007.03.010. [DOI] [PubMed] [Google Scholar]
- 25.Winter R, Jussila R, Nowak J, Brodin LA. Speckle tracking echocardiography is a sensitive tool for the detection of myocardial ischemia: a pilot study from the catheterization laboratory during percutaneous coronary intervention. J Am Soc Echocardiogr. 2007;20:974–981. doi: 10.1016/j.echo.2007.01.029. [DOI] [PubMed] [Google Scholar]
- 26.Chan J, Hanekom L, Wong C, Leano R, Cho GY, Marwick TH. Differentiation of subendocardial and transmural infarction using two-dimensional strain rate imaging to assess short-axis and long-axis myocardial function. J Am Coll Cardiol. 2006;48:2026–2033. doi: 10.1016/j.jacc.2006.07.050. [DOI] [PubMed] [Google Scholar]
- 27.Reant P, Labrousse L, Lafitte S, Bordachar P, Pillois X, Tariosse L, Bonoron-Adele S, Padois P, Deville C, Roudaut R, Dos Santos P. Experimental validation of circumferential, longitudinal, and radial 2-dimensional strain during dobutamine stress echocardiography in ischemic conditions. J Am Coll Cardiol. 2008;51:149–157. doi: 10.1016/j.jacc.2007.07.088. [DOI] [PubMed] [Google Scholar]
- 28.Bogaert J, Rademakers FE. Regional nonuniformity of normal adult human left ventricle. Am J Physiol Heart Circ Physiol. 2001;280:H610–H620. doi: 10.1152/ajpheart.2001.280.2.H610. [DOI] [PubMed] [Google Scholar]