What are “cardiac perfusion MRI” and “delayed enhancement imaging”?
Perfusion imaging with MRI is very similar to perfusion imaging with nuclear tracers (single photon emission computed tomography [SPECT] imaging or positron emission tomography [PET] imaging). Each test has an imaging (contrast) agent which can be tracked as it goes through the myocardium, so MRI is the same principle as a study using SPECT imaging or PET imaging. In cardiac perfusion MRI, you look at blood flow at rest and with a stressor, like adenosine. We use the contrast agent gadolinium, a standard MR contrast agent which washes in and out of the myocardium fairly quickly, and first-pass imaging, which means we are looking at the first pass of the contrast agent as it enters into the chambers and then into the myocardium itself. In the myocardium, we are looking at the perfusion. As with nuclear cardiology perfusion imaging, we evaluate blood flow changes from rest to stress.
The difference between MRI and SPECT imaging is that the spatial resolution of MRI is much better and there are no attenuation artifacts. MRI cannot be done in patients with implanted devices, like pacemakers and ICDs. With that exception, very good images are available in virtually all patients, without some of the attenuation issues one has with overweight patients or women who have breast artifacts, where you may see attenuation artifacts with SPECT imaging. It is possible to get around the attenuation artifacts with PET imaging, but PET is less available in most centers. Most large hospitals do have MRI. Of course, with MRI, there is no radiation exposure, but the real benefit is that the spatial resolution is so good that you can see the distribution of blood flow across the transmural extent of the left ventricular wall. With SPECT and PET, mild perfusion defects are visible, but it is not possible to determine if the entire wall of the heart is mildly underperfused, or if it is mostly a subendocardial perfusion defect. With MRI, you can see the subendocardial nature of the perfusion defects, and get a feeling of the transmural extent of the ischemia, as well as the circumferential extent around the ventricular wall. Therefore, MRI is quite sensitive at picking up very small perfusion defects that might be missed with other techniques. It’s a true advantage.
A disadvantage is that MRI is more difficult to do, because the test is done in a magnet and in a room where you cannot have any metallic objects. Stressing the patients in the magnet can be tricky, so it does take some logistical setup and a great deal of practice to get cardiac perfusion MRI to work well. Whether it will ever take off as a technique where large volumes of patients can be done is not certain. For example, in our hospital, we have three SPECT cameras — some larger centers have 5 or 6 cameras — going all day long. Labs can easily do 20-30 patients a day. It would be unlikely that we could ever do that volume of patients with MRI, only because it is more difficult logistically, given the constraints of doing the test in a large magnet, and also the fact that MRI is used for many other things.
Delayed enhancement imaging arose out of a phenomenon that has been demonstrated repeatedly. In irreversibly damaged myocardium, the contrast agent gadolinium washes out of normal heart tissue, but is retained within irreversibly damaged areas. This test is useful, for example, in a patient with an acute infarction that is imaged within a day or so of the infarct, or even in someone who had a chronic infarction many years ago. Delayed enhancement imaging is a way to visualize the presence, location and extent of myocardial infarcts. Even very small infarcts can be identified due to the spatial resolution of MRI. Virtually every MRI vendor now has this element in their software package, because the ability to have contrast-enhanced studies that can demonstrate infarct location is unique right now for MRI. This technique is also useful if a patient has left ventricular dysfunction with coronary artery disease, and you want to know if the area not working normally is irreversibly damaged myocardium or viable myocardium that is hibernating, i.e., it is underperfused and not contracting, but not dead. It is an excellent technique for identifying which dysfunctional areas are irreversibly damaged and which areas are viable, but underperfused. Ultimately, delayed enhancement imaging can identify which patients we can send for revascularization. There’s no sense in revascularizing dead areas, but if you return blood flow to the hibernating areas, function can improve. This is definitely an area where MRI can become a test one might consider clinically. It is unlikely that MRI perfusion imaging will ever eclipse nuclear cardiology, which has been around for a long time, and will continue to be around for a long time to diagnose coronary disease and identify the extent and severity of ischemia. However, MRI does have the ability to show the areas that are hyper-enhanced. Once you get the first-pass perfusion images, then you come back later and see where the tracer has not washed out of the heart, i.e., the hyper-enhancement image. MRI also provides information on the function of the heart. We can get gated cine images of the heart function, again, without the attenuation artifacts and without the poor ultrasound windows which you can get with echocardiograms. Good images of cardiac function are routinely possible. MRI fits in beautifully in our heart failure center and our heart failure specialists routinely order MRI studies. First, they can see the function — the volumes and the anatomy — of the heart. If patients have remodeled hearts with dilated ventricles, this is visible. MRI is also good for patients who have restrictive or hypertrophic cardiomyopathy, because you can get the function and the morphology. Next, perfusion imaging is done with MRI to determine if there is reversible ischemia. Then, with the contrast agents, you get the evidence — or not — of reversible ischemia. Delayed enhancement images can show you how much of the left ventricular dysfunction appears to be irreversibly damaged myocardium.
One other advantage of the perfusion studies, however, is that perfusion imaging is potentially quantifiable. There is a lot of work going on in our laboratory and others to not only visualize the perfusion defects but also quantify the blood flow. It could be important in the future when we are dealing with some of the very exciting research finally translating into clinical studies, things like stem cell research, where we may be able to grow new blood vessels using endothelial progenitor cells, for example. If you want to measure the effect, it is important to identify subtle changes in blood flow. It may not be a major, remarkable effect; it’s not, initially, in the early studies. What this type of research requires is a technique like MRI, which can get down to the subendocardial level — it goes to the transmural level — and not only visualize small areas of ischemia that might be improved, but quantify them.
Cardiac MR enables visualization of new entities like the ‘peri-infarct zone’;or ‘small unrecognized infarcts’; What do you think about these?
I am not sure yet that we know what to do with the peri-infarct zone. The way I would translate what you are saying is that the gadolinium hyper-enhancement images give you some areas that are clearly viable, and some that are clearly not. At certain points, there is some intermingling of viable and non-viable cells, and that is why a particular area would appear in the myocardium to be, if not totally bright, not irreversibly damaged, then not totally dark either. For me, the peri-infarct zone is just a border zone between the dead areas and not-so-dead areas. I do not think it is something which is highly relevant right now. What is much more relevant is the ability to obtain data regarding the infarct size as well as the area of viability versus non-viability. Cardiac MRI detects very small infarcts. When every other test is normal — normal EKG, normal stress tests, normal nuclear, normal echo tests — you can still see small infarcts missed by other techniques. It is quite good for detecting reversibly damaged myocardium with potential for improvement. That’s where MRI is most clinically relevant right now.
What do you think the role of MRI might be in the future?
It is a great test already. It will continue to be a great test, particularly for complex anatomy. In patients with congenital heart disease, MRI is spectacular for identifying abnormalities, and that will always be the case. Again, importantly, it is a technique which has no radiation involvement. I do think it has great potential in an era surrounded by radiation exposure through testing.
Sooner or later, there will be more work obtaining reproducible, routine, noninvasive coronary angiograms with MR, which has been an elusive target. In some patients, one can get spectacular images of the coronary arteries. Obviously, it’s difficult to do because the coronary arteries are small, tortuous and moving. There is more work being done in this area, using higher field strength magnets than we use in normal imaging: instead of using .5 tesla magnets we might be using 3 tesla magnets or greater to get better imaging of the coronary arteries. Obviously CT has taken over that area of imaging, but there is radiation exposure with CT, which is one concern. The other concern is clearly how it should be used and who the right patients are, which is an issue for MRI as well. Finally, the functional information from MRI can be quite good. We can already image plaques in the carotids and aorta, which are stationary blood vessels. There is a great deal of work going on right now to determine plaque composition, such as a fatty pool with a thin fibrous cap, for example, as opposed to what might be a more stable plaque. MRI can do this already in these larger blood vessels, and I think with more research, we will find that it can be done in the coronary blood vessels too.
What about reimbursement?
There’s already a slowdown in CT imaging because of reimbursement images. You may have seen, unless there is a change in legislation, that Medicare may no longer be approving clinical CT scans unless patients are part of a clinical registry or trial. This will, sooner or later, affect MR imaging as well, until we can clearly identify the right patients to study, and whether we can demonstrate that MR will lead to better outcomes than what we can achieve with more readily available and cheaper technology.