Dobutamine stress echocardiography (DSE) is a very good method to analyze wall motion abnormalities during stress, but in about 10-15% of patients it is not possible to obtain adequate image quality despite very new methods such as second harmonic imaging. Additionally, there is a large observer dependency for the results; much training and practice is required to obtain reproducible results with echocardiography (ECG). In basal lateral and basal inferior segments there is worse image quality and these segments can not be analyzed very frequently.

Dobutamine stress MR (DSMR) can be used to look at wall abnormalities at rest and at stress. With the new imaging techniques, 5 different views are taken: 3 short axis, a 4-chamber view, a 2-chamber view. Complete cardiac motion of the entire heart can be analyzed with these 5 slices taken within 5 breath-holds of 12-15 seconds. Nagel showed one DSMR image from the cineloop showing the complete cardiac motion. No contrast agent was used, as there is a very strong contrast between the blood and myocardium can be used to visualize motion.

Turbo gradient echo techniques or echo planar imaging (EPI), a very new technique that allows obtaining a complete movie of the heart within 8-12 seconds, are now being used in their laboratory. Real-time imaging has now been developed. Breath-hold technique allows for using the same stress protocols with dobutamine as with ECG. Each step is performed within 3 minutes and then the dobutamine dose increased to a maximum dose of 40 mcg, plus atropine as needed.

Image analysis

Image analysis is mainly done by visual assessment, although quantitative assessment is possible. The American Society of Echocardiography criteria of looking at 16 segments is used, and ischemia is defined as lack of improvement or worsening of endocardial motion or wall thickening during stress. One example illustrated that high doses of dobutamine are needed to induce myocardial ischemia. The wall motion improved from rest to 20 mcg dobutamine but a severe wall motion abnormality was seen with 40 mcg dobutamine stress.

To perform an analysis it is best to view all the stress levels simultaneously. This can be done in a quad-screen format as in ECG showing the images only at rest, or using a newly developed tool by Hundley to show all different stress levels simultaneously. The studies of dobutamine stress for the diagnosis of ischemia performed through 1999 used a maximum dobutamine dose of 20 mcg, as the gradient techniques allowing an image to be obtained within one breath-hold was not available. The ability now to use high dobutamine stress has increased the sensitivity and specificity of DSMR imaging to about 85%, improving diagnostic accuracy.

Influence of image quality

A comparison of moderate echocardiographic image quality obtained with DSE to images obtained with DSMR shows a statistically significant difference in terms of sensitivity (62% vs 90%, respectively) and specificity (55% vs 82%) [Fig. 1]. Comparing good and very good image quality obtained with DSE to DSMR there is little difference in sensitivity and specificity, all at about 90%. However, even with second harmonic imaging, good or very good images were only obtained in about 50% of patients with DSE. Therefore, in about 50% of patients much better diagnosis can be obtained with DSMR than with DSE.

Limitations of DSMR

Patients with claustrophobia or metallic implants (pacemaker, implantable cardioverters) can not be examined. This problem may be solved with new pacemaker techniques. The problem of reduced image quality in patients with frequent premature ventricular complex or atrial fibrillation has been addressed using the breath-hold technique and the images are re-constructed within a few seconds after acquisition, which might be a safety problem. However, real-time techniques may be useful, requiring about 55-60 ms to acquire one image. ECG triggering is not required. The real-time images compare well with those obtained with turbo field echo (TFE) or EPI, with wall thickening, wall motion, and endocardial motion assessable. Real-time DSMR has been shown to capture 97% of the wall abnormalities seen with the standard technique.

Steady state technique

These images do not require inflow contrast, whereas standard techniques are dependent on fresh blood inflow, providing a 3-fold increase in contrast between the blood and myocardium in contrast to standard technique. This is advantageous in patients with low ejection fraction or for long axis views.

At the onset of ischemia a perfusion deficit is seen, and later in the course wall motion abnormalities, ECG changes and anginal pain are seen. Wall motion abnormalities are in principle less sensitive than looking at perfusion defect itself. Problems with the techniques routinely used, positron emission tomography (PET) and single-photon emission computed tomography (SPECT), include low spatial resolution, attenuation artifacts, and radiation. MR is highly promising technique for assessing myocardial perfusion, based on preliminary study results. One example showed that not only could perfusion be assessed, but that the defect was subendocardial, not transmural, which would not have been seen with nuclear techniques.

Five to seven short-axis views per heart beat are acquired--an important improvement in ability. A contrast injection is given which goes first into the right ventricle and then the left ventricle and then into myocardium.

An alteration of the upslope can be obtained if the contrast is given before or after vasodilation with adenosine or dipyridamole (Fig. 2). This is done for 6 segments per slice, for a total of 30 segments per patient. This technique has been validated in a highly selected group of patients with high-grade proximal single vessel stenoses. A good differentiation between segments supplied by high-grade stenotic coronary artery and control segments was used. From this study they developed a cut-off value of 1.5 MP/L which was then used for a prospective study. They have also shown that the upslope parameter is highly reproducible.

The recent studies of detection of coronary disease with MR have sensitivity (90-92%) and specificity (84-87%) rates comparable to those with nuclear imaging. A further development is the intravascular contrast agents that may further refine this technique. Current contrast agents show both perfusion and diffusion as the contrast agent moves from the vascular to the interstitial space during the first pass.

MR stress studies are possible today that can be used routinely for the detection of wall motion abnormalities that are much better than echocardiography, particularly if the echo images are suboptimal. MR perfusion studies can be done routinely with results that are highly promising and better than those with nuclear medicine. Wall motion and perfusion studies during stress can be combined into a single stress test, allowing CMR to be used routinely.

At his institution, a 35 ms temporary resolution is used to ensure that end systolic images are not missed. A time of 15 ms might be too short to observe relaxation disturbances such as early diastolic filling; better resolution is required, which is possible by either increasing measurement time or reducing spatial resolution. In response to a question about intrinsic motion of infarcted areas, Nagel stated that can not be done with normal wall motion analysis because passive movement can not be discerned as well from actively moving areas. However, there are two criteria that help. One, wall thickening itself because there is good contrast between the blood and endocardium with MR and also between the epicardium and fat allowing good visualization of wall thickening. This helps to discern passively moving areas that are just pulled towards the center versus actively contracting areas that thicken by themselves. Two, myocardial tagging that shows intrinsic shortening can be used.

Commenting on whether wall imaging during dobutamine or perfusion imaging during vasodilation is easiest to use today considering both acquisition and post-processing, Nagel stated there was no definite answer. In his institution it is currently easier to perform wall motion analysis due to their large experience with dobutamine. If one begins to do stress studies, one may select not to do high dose dobutamine studies in the MR scanner as some experience is needed. The stress study itself is easier with adenosine or dipyridamole. However, because post-processing is much more difficult for the perfusion imaging very good software is needed for the analysis, and this is not yet routinely available.