Clinical Utility of Assessments of Left Ventricular Systolic Function and Coronary Arterial Blood Flow During Pharmacological Stress with Magnetic Resonance Imaging

Advances in computer software, scanner hardware, and hemodynamic monitoring equipment have permitted the development of magnetic resonance imaging (MRI) protocols that can be used to assess left ventricular (LV) systolic wall thickening and coronary arterial blood flow during intravenous dobutamine or adenosine infusions. Recent studies documented the safety profile and clinical utility of these protocols for assessing patients suspected of having coronary arterial atherosclerosis. These relatively new MRI cardiac stress testing techniques appear particularly useful in patients not well suited to other forms of noninvasive stress testing. ASSESSMENT OF LEFT VENTRICULAR SYSTOLIC THICKENING DURING INTRAVENOUS DOBUTAMINE/ATROPINE Although exercise testing is used widely to identify inducible LV myocardial ischemia, many patients are limited in their ability to perform exercise due to peripheral arteriosclerotic vascular disease, disabling arthritis, history of stroke, orthopedic constraints, or chronic pulmonary disease (1). Dobutamine, a synthetic catecholamine that increases LV myocardial contractility in normal subjects (2,3), can be substituted for exercise to provide myocardial inotropic stimulation (1). It is relatively inexpensive; its onset of action is within 2 min; and it achieves a maximal effect on contractility in normal subjects at a relatively low dose (10 μg/kg/min) (4,5). When a heart rate response ≥85% of the maximum predicted heart rate response (MPHRR) for age is not achieved with intravenous dobutamine (a target heart rate that has been shown to improve the diagnostic accuracy of wall motion stress testing) (6), atropine (a parasympatholytic drug) may be administered in 0.3-mg increments every 30 seconds up to a total dose of 1.5 mg to augment heart rate response. Combined dobutamine/atropine pharmacological stress protocols have a sensitivity and specificity of 80% and 84%, respectively, for detecting inducible ischemia (a diagnostic accuracy that is similar to exercise testing) (7). Since 1992, several investigators have documented the clinical utility of using dobutamine MRI (DMRI) for detecting coronary arterial luminal narrowings that induce myocardial ischemia during stress (Table 1) (8–10). The first three studies shown in Table 1 were performed to demonstrate the feasibility and diagnostic accuracy of DMRI stress testing. Importantly, these first three studies had limitations. They were performed in relatively small numbers of patients, the severity of coronary artery arteriosclerosis was known prior to MRI testing, wall motion was assessed at baseline and at peak stress (and not visualized continuously throughout the course of testing), and LV wall motion was not assessed during rapid heart rates at high levels of pharmacological stress (particularly 85% MPHRR by age). These factors and the perception of poor patient tolerance for stress testing during MRI limited widespread clinical application of these early reported techniques.TABLE 1: Sensitivity and specificity of dobutamine magnetic resonance imaging for detection of coronary arterial luminal narrowingsRecent software and hardware advances have enabled investigators to overcome some of these limitations. Specifically, scans incorporating shorter repetition times (on the order of 7 to 15 ms), phase-encoding grouping, and phased-array surface coils have allowed for acquisition of images with high temporal resolution and signal-to-noise ratio that permit clear delineation of the LV endocardial surface during rapid heart rates (11) in all tomographic planes of the LV (12). In addition, the introduction of hemodynamic monitoring (heart rate, oxygen saturation, blood pressure) and intravenous administration equipment, and wall motion display software (Fig. 1) have allowed physicians to safely monitor patients during the course of stress testing (13). This latter development of specialized software for image display during stress testing is particularly important for detecting inducible ischemia because ST segments are altered by the magnetic field gradient within the bore of the scanner; thus, monitoring them for ischemia during stress testing is not feasible.FIG. 1.: Selectable display format for viewing synchronized cine loops of the left ventricle (LV) during testing. Top: Three apical and short-axis views of the LV at end-diastole from a single dose of dobutamine are shown. Bottom: Single short-axis view acquired at end-systole at multiple doses of drug infusion are displayed. In each view, the LV myocardium appears gray and the flowing blood appears white. The viewing format is selectable to provide rapid switching (<1 second) between display formats continuously during stress testing. [Reproduced with permission from Lippincott Williams & Wilkins (13).]Recently, the latter two studies listed in Table 1 highlighted the utility of these technical advancements for the performance of DMRI in a clinical setting. In these studies, an 18-segment model incorporating three longitudinal and three short-axis views (a model derived from the 16-segment model developed by the American Society of Echocardiography) (14) was used for image acquisition, and, like echocardiography, a 4-point scoring system, in which 1 is normal, 2 is hypokinetic, 3 is akinetic, and 4 is dyskinetic, was used to assess LV endocardial thickening. Ischemia was identified when a deterioration in score ≥1 occurred during stress testing, or a hypokinetic segment at rest failed to improve contractility or displayed a biphasic response (Fig. 2) (15).FIG. 2.: Peak dobutamine infusion cine MR images of the left ventricle (LV) are shown in the middle short-axis and the apical LV outflow tract views. Several images from end-diastole (left) to end-systole (right) are displayed in a patient with a normal contractile response (top panels) and evidence of inducible ischemia (bottom panels). Inducible ischemia is identified by the development of hypokinesis in the middle and apical segments of the anteroseptum at end systole (white arrows). [Reproduced with permission from (15) Marcel Dekker, Inc.]Nagel et al. (16) compared DMRI and dobutamine stress echocardiography (DSE) in 208 patients referred for contrast coronary angiography. DMRI provided better sensitivity (89% vs 74%) and specificity (86% vs 70%) for detecting ≥50% coronary arterial luminal narrowings with coronary angiography compared with DSE. Hundley et al. (13) addressed the utility of DMRI in 163 patients who failed DSE. In general, these referrals were relatively heavy (mean weight 89 kg, range 39–135 kg) and short (average height 170 cm, range 140–196 cm). At baseline, 63% of patients had abnormal LV wall motion, 34% had prior myocardial infarction, 32% had prior coronary surgical revascularization, and 25% had prior percutaneous coronary interventions. When compared to contrast coronary angiography, the sensitivity and specificity were both 83% for detecting coronary arterial luminal narrowings >50%. Patients who had normal wall motion at rest and no evidence of inducible ischemia had a very good prognosis, with a negative predictive value of 93%. The results of these latter two studies highlight the clinical utility of these technique for the 5% to 15% of patients referred for stress echocardiograms who cannot undergo the test due to relatively poor acoustic windows (examples include patients with large body habitus, severe obstructive airway disease, or prior cardiothoracic surgery) (7). Although these results are encouraging, there are several areas that deserve further study. These include the study of patients with left ventricular hypertrophy and those with myocardial segments that are persistently hypokinetic throughout the course of stress testing. In addition, the application of rapid quantitative analysis techniques (myocardial tagging and velocity-encoded phase mapping) for image interpretation during stress may warrant further study (17,18). MRI MEASUREMENTS OF CORONARY FLOW RESERVE FOR ASSESSMENT OF THE PHYSIOLOGICAL IMPORTANCE OF CORONARY ARTERIAL STENOSES A potential advantage of MRI for assessing patients with coronary atherosclerosis is the ability to directly visualize epicardial coronary arteries and to assess the functional importance of epicardial luminal narrowings through measurement of coronary flow reserve (CFR) (19–21). CFR, the stress/rest ratio of coronary arterial blood flow in mL/min, has been shown with invasive (coronary sinus thermodilution and Doppler guidewires) and noninvasive (Doppler transesophageal echocardiography) techniques to be superior to coronary arterial lumen measurements for assessing the physiologic importance of coronary stenoses (20–25). CFR is particularly useful when stenoses of intermediate severity (50% to 70% luminal narrowing) are present (25,26). Direct measurements of CFR are made most frequently during MRI through analysis of the phase of the collected data (25,27–29). This technique, phase-contrast MRI (PC-MRI), is based on the principle that flow through a magnetic field gradient induces a phase shift that is proportional to velocity. Flow is calculated by multiplying mean velocity and the area of the vessel lumen (25). Several investigators reported the utility of measuring coronary artery velocity or flow reserve during pharmacological administration of adenosine (25,30) and dipyridamole (30–32). Using PC-MRI, Sakuma et al. (31) measured flow velocity in the left anterior descending (LAD) artery of eight healthy adult volunteers. Baseline flow velocity was 14.8 ± 1.9 cm/s, flow velocity after dipyridamole infusion was 46.3 ± 10.2 cm/s, and average coronary velocity reserve was 3.14 ± 0.59. Davis et al. (32) used PC-MRI to measure coronary blood flow and flow reserve in the LAD of 10 healthy subjects. Mean coronary blood flow was 38 ± 11 mL/min and increased to 169 ± 42 mL/min after dipyridamole infusion. Mean coronary flow reserve was 5.0 ± 2.61. Investigators also used measurements of CFR to identify coronary arterial stenoses (Fig. 3). Hundley et al. (20) demonstrated that CFR measured in the proximal and middle LAD correlated well with those by Doppler flow wires. The sensitivity and specificity of CFR ≤ 1.7 by PC-MRI for detecting coronary stenosis >70% were 83% and 100%, respectively (20). A particular clinical application for this methodology is the identification of coronary arterial restenosis (Fig. 4) (21). Using the same techniques, Hundley et al. (21) reported that CFR ≤2.0 correlated with coronary arterial luminal narrowings ≥70% in patients who had chest pain 3 months after successful percutaneous coronary interventions. In addition to native vessels, PC-MRI CFR measurements can be obtained in bypass conduits (33). Reduced blood flow and flow reserve has been associated with significant stenosis within bypass grafts (33,34). Hoogendoorn et al. (35) evaluated patients with saphenous vein grafts and found that blood flow in the patent grafts was 74 ± 17 mL/min and decreased to 9 ± 8 mL/min in the stenotic or occluded grafts. Voiglander et al. (36) found CFR differences in patients with stenotic bypass grafts to be 2.9 ± 1.9 versus 1.2 ± 0.5, respectively. Studies performed to date utilizing PC-MRI CFR measurements to identify coronary or bypass graft narrowings are listed in Table 2.TABLE 2: Studies performed to date utilizing PC-MRI to measure coronary artery flow reserve in the setting of coronary atherosclerosisFIG. 3.: Tangential magnetic resonance images of the left anterior descending (LAD) coronary artery obtained from two patients: one with 30% (top left) and one with 64% (top right) stenosis by computer-assisted QCA. White arrows demarcate site of stenosis appreciated with MRI. Arrows in the middle row demarcate corresponding cross-sectional views of LAD imaged 5 to 10 mm distal to each stenosis. Bottom: Corresponding velocity maps (cross-sectional images of vessels) from patients at rest and after pharmacological stress with adenosine. Intensity of the gray scale of each pixel within the region of interest represents velocity (darker indicates higher velocity). Black circles highlight pixels used for flow calculations in each patient. Left: Darker pixels indicate higher velocities after adenosine infusion. This patient had coronary flow reserve of 2.7 and 30% stenosis by computer-assisted QCA. Continued gray pixels after adenosine in patient on the right indicate impaired coronary flow reserve (value = 1.2) and presence of 64% stenosis by QCA. [Reproduced with permission from Lippincott Williams & Wilkins (20).]FIG. 4.: Tangential (top left panel) and cross-sectional (top right panel) MR angiograms of the left anterior descending coronary (LAD) artery. Right: Selected image from the patient's contrast coronary angiogram. In the tangential MRI and contrast coronary angiogram, the location of an intracoronary stent is shown. The white arrow on the MR cross-sectional image demarcates the corresponding cross-sectional view of the LAD imaged 5 to 10 mm distal to the intracoronary stent. From these cross-sectional images, a region of interest is selected for calculating the cross-sectional area of the vessel. This region of interest is transferred to the velocity maps, in which the intensity of the gray scale of each pixel within the region of interest represents velocity (darker = higher velocity). The bottom row of images are corresponding velocity maps (cross-sectional images of the vessels) from the patient at rest and after pharmacological stress with adenosine. On the right velocity map, the darker pixels within the black circles indicate higher velocities after adenosine infusion. This patient had a coronary flow reserve of 3.1% and 35% stenosis by QCA. Cx = circumflex; LM = left main. [Reproduced with permission from Lippincott Williams & Wilkins (21).]Although PC-MRI CFR measurements are safe and can be performed in an outpatient setting without intravenous contrast or ionizing radiation, there are some limitations to their clinical use. First, postprocedure analysis is time consuming. Automated analysis programs are necessary to heighten clinical applicability. Second, metallic clips or stents can cast artifacts that prohibit visualization of vessel segments. Measuring flow reserve proximal or distal to these artifacts may be helpful (21,33). Finally, conditions such as dilated or hypertrophic cardiomyopathies, moderate-to-severe valvular heart disease, severe hypertension, diabetes, previous anterior myocardial infarction, or syndrome X may reduce CFR (20,21,37); thus, using PC-MRI CFR measurements in these circumstances may falsely indicate the presence of a severe stenosis. CONCLUSIONS In the past decade, there has been an increase in the use of cardiac MRI for evaluating patients with, or suspected of having, coronary arterial atherosclerosis. DMRI has been used effectively to diagnose inducible ischemia in patients who have poor acoustic windows during transthoracic echocardiography. Recent studies indicate that it provides accuracy and that it is similar in accuracy to other widely accepted noninvasive imaging modalities. Measuring coronary arterial blood flow and flow reserve with PC-MRI is an accurate and reproducible technique that can be used to assess the physiological importance of coronary arterial luminal narrowings in patients with normal coronary arterial microcirculatory vasoreactivity. CFR measurements are particularly useful when evaluating patients with stenoses of intermediate severity (luminal narrowings of 50% to 70%). They also have been used to evaluate the functional importance of luminal narrowings within coronary arterial bypass conduits or native coronary artery segments that have undergone percutaneous intervention. Acknowledgments: Research supported in part by grants from North Carolina Baptist Medical Center Development Technology Fund (B-03-97/98R/OOR), NIH Clinical Associate Physician Award (3-MO1-RR07122-08S1), and General Electric Medical Systems (BG-09918987-9960).