|
|||||||||||||||
|
Home
|
Cardiac MRI —
A Radiology Today Interview With Ricardo C. Cury, MD Ricardo C. Cury, MD, is the director of clinical cardiac MRI at Massachusetts General Hospital (MGH). He is also director of teleradiology and education at the MGH Cardiovascular CT Core Lab and assistant professor of radiology at Harvard Medical School. Radiology Today (RT): How long have you been involved with cardiac MRI? Ricardo C. Cury, MD: I’ve been involved with cardiac MRI and CT for about five years. I did a dedicated fellowship in cardiac MR and CT, so I do both in clinical and research settings, which provides a good assessment without bias to one modality or the other. If you are only doing cardiac MR or only doing cardiac CT, you tend to be biased toward your modality. RT: So you started with both modalities basically at the same time? RCC: My training was with cardiac MRI first, back in 2003 when cardiac CT was just taking off. RT: What is the “state of the union” of cardiac MRI at this point in time? RCC: Basically, right now, we are using 1.5T MR with high gradients and parallel imaging so we can reduce the scan time. Compared with other cardiac imaging modalities, the longer length of time MR takes to scan is one of its drawbacks. Regarding clinical applications, there are well-established ones that we use on a daily basis. The clinical applications are to assess myocardial viability, assess myocardial disease overall for ischemic and nonischemic cardiomyopathies, and assess pericardial constriction, congenital heart disease, cardiac masses, and thrombi. These vary from site to site, but cardiac MRI is generally used as the last resource in the workup of cardiac patients. In general, patients undergo a cardiac echo, then a nuclear study, and then an invasive angiogram. After those exams, if there are still questions to be answered, then cardiac MRI is used. The efforts at our site and some others hope to bring MRI to the forefront of the diagnosis process in cardiac patients instead of MR being the last test to answer difficult questions. RT: Is there much likelihood of that happening in the next few years, or is it still a ways off? RCC: MRI is extremely good and probably considered the gold standard for assessing left and right ventricular function and volumes. Myocardial viability and stress perfusion MRI are gaining popularity; however, nuclear medicine has an advantage in this use, mainly because of reimbursement. One thing that MRI needs to get better at is coronary imaging. Cardiac CT has grown so fast in the past few years because it provides robust imaging of the coronary arteries for stenosis and plaque. MRI needs to improve in that area to move to the forefront in cardiac imaging. RT: Is there much work toward that objective, or is it an area that needs to be addressed? RCC: There is work being done on that, but the evolution of cardiac CT has been so fast that every other year brings a new generation of CT scanners. MRI is slowly catching up, but more progress will have to be made. The spatial resolution of cardiac MRI is about 1.5 millimeters, and there are few ways to improve it to 0.7 or 0.8 millimeters. I think the key issues are improving the spatial resolution and acquiring the entire image of the heart in one breathhold, as opposed to using navigator techniques for 15 minutes that can create a lot of variability and blurring of the coronary arteries. These are things MRI will need to evolve. RT: That sounds like quite a challenge. RCC: There is work in development such as whole heart MR angiography where you get a spatial resolution of approximately 0.7 or 0.8 millimeters. This view will not be quite as good as CT, but it will be getting close. Right now, the main problem is that the acquisition takes about 15 minutes because it is divided into several RR [respiratory rate] intervals. You need to have a very cooperative patient to obtain reliable images. You can also use a 2D approach where you obtain slabs of each individual coronary artery. RT: How fast do these exams need to be? Do they have to get down to a single breathhold? RCC: That would be ideal; then you’d really be competing with cardiac CT. By using new coil design, parallel imaging, and maybe going to 3T systems, that potential improvement might be possible for cardiac MR. One growing area is stress perfusion MRI, which is an indirect way of detecting coronary stenosis. For example, with coronary CTA [CT angiography] or invasive angiogram, you are looking at the coronary stenosis itself. With stress perfusion MRI, we are looking for ischemia and infarct, but it really represents a way to detect hemodynamically clinically significant stenosis. This is what nuclear medicine is doing now. There are several trials, including multicenter trials, showing that cardiac MR is better for stress perfusion, so MR could gain more popularity for stress perfusion imaging. But at this point, I believe one of the main reasons MR has not gained more popularity is that reimbursement is very good with nuclear medicine, and it’s easy for the physician to read several cases as opposed to MRI, which is more labor intensive to acquire and interpret the data. RT: So cardiac MRI is competing in a couple areas against established modalities with good reimbursement. Are there any other reasons why it hasn’t taken off? RCC: The other thing is, we are trying to develop protocols we can do in a short period of time. For example, we performed one research project here at Massachusetts General in which we studied patients with acute chest pain who come to the emergency department. There are several groups, including ours, that are proposing cardiac CT as a very fast test to look at the coronaries and discharge the patient or send to the cath lab. We developed a protocol that includes the evaluation of myocardial edema with T2-weight images, function, perfusion, and delayed enhancement. The MRI protocol took approximately 30 minutes and gave very accurate and reliable information in patients with myocardial damage, confirming the presence or ruling out acute coronary syndrome. Compare that with cardiac CT, which takes 10 to 15 minutes for the patient to go on and come off the table. That speed provides one major advantage for patient throughput—and also potentially for the business perspective when cardiac CT gets reimbursed across the board. But cardiac CT may lack positive predictive value and have a tendency for overestimation of stenosis, mainly in segments with dense calcifications. RT: Many researchers and clinicians have talked about how cardiac CT is excellent as a rule-out diagnosis. Will cardiac MR then come into play when there is not a rule-out diagnosis? RCC: I think they will be complementary modalities, and this is a very important aspect of how they’ll be used. Cardiac CT will be a first test to find out if a patient has coronary artery disease and, if so, what is its severity. The negative predictive value is very good, but the positive predictive value is moderate. If you have a positive test, sometimes you don’t know the hemodynamic significance of those lesions—mainly with moderate lesions from 40% to 70%. Those patients would benefit from a cardiac MRI to look at perfusion, to see if there is ischemia, and also to look at delayed enhancement for myocardial infarct. A combination of cardiac CT and MRI would provide enough information to better manage the patient and to really decide the therapeutics and guide the right patients to the cath lab. RT: Looking at the big economic picture, not sending people to the cath lab who don’t need to be there should more than justify the increased imaging costs because you’re sending far fewer people to the more expensive cath lab. RCC: Exactly. Also, in the setting of acute chest pain in the emergency department, if you have a reliable test that you can rule out acute coronary syndrome early on, you are also saving costs because the patient will not be admitted into the hospital. When you admit patients, they’ll stay in observational units for 24 hours and have serial cardiac enzyme tests, EKGs [electrocardiograms], and also a stress test with or without imaging. With a quick rule-out diagnosis you save a lot of costs right there. RT: In general, many exams start out being done on CT, and then researchers develop MRI protocols to perform them, getting the benefit of not exposing the patient to radiation. How much of a factor is that in the cardiac environment? RCC: There is increasing concern in healthcare about radiation exposure. Cardiac CT exposes the patient to a considerable amount of radiation. A cardiac patient may have a cardiac CT scan, a nuclear stress test, and an invasive angiogram—and even follow-up studies to these exams. That increases his or her radiation exposure quite a bit. That’s where I believe cardiac MRI has a unique advantage by not exposing the patient to radiation. And that’s why, I think, the challenge of cardiac MR is getting into coronary imaging. Because other than imaging the coronaries, MRI is extremely good for imaging left ventricular function, perfusion, and viability. Right now, we are doing three to five cases per day of cardiac MR, but compare that with CT and nuclear medicine where we do about 10 and 40 cases. I believe, in the future, we may see a shift to fewer nuclear stress tests—or maybe all the modalities will grow as the population ages and you have a greater prevalence of disease among the baby boomers. RT: Is there anything else that may be an obstacle to further cardiac MR development? RCC: The other thing that stands in the way of its widespread use is physician training. Right now, there are not many radiologists or cardiologists who are trained to do cardiac MRI because it’s challenging. It probably takes six months—if not a year—to be able to interpret these tests in an accurate way. So physician training is also a limiting factor. We need to train more physicians in this technique. RT: Is there a steeper learning curve than with cardiac CTA? RCC: Yes, that’s for sure. With CTA, I would say you need about three months of training to do it well. But cardiac MRI would take a minimum of six months and ideally, you should do a dedicated yearlong fellowship to train in this modality. RT: You mention that there aren’t enough radiologists and cardiologists out there trained to read these exams. As training expands, is that going to create a turf battle between the specialties, or is there going to be enough work for everyone? RCC: Depending on the setting, I would definitely expect to see some turf battles. I think the best approach is a combination of radiologists and cardiologists working together. I think both specialties bring a lot to the table. For example, radiologists completely understand both CT and MRI technique and the physics. Basically, they just need to be trained in the anatomy and physiology of cardiac disease. Having trained many of the fellows here—and we train both radiologists and cardiologists—I can see during the fellowships that radiologists learn faster. The cardiologists need more time because they are not familiar with axial raw data imaging and reconstructions. Working at an imaging workstation is bread and butter to a radiologist, where cardiologists need more training. In general, they are getting good between six and 12 months into the fellowship. In the end, both have good training. The cardiologists already know the other cardiac imaging modalities such as echo, nuclear medicine, and cardiac catheterization. And they certainly understand all the cardiac diseases, so it’s a matter of training in the modalities. The other point that is very important when you get to the evaluation of patients with acute chest pain is that you need a comprehensive evaluation. You need to examine not only the heart but other structures in the CT or the MRI. You need to evaluate the aorta, pericardium, mediastinum, and lungs to detect all the pathologies such as dissection, pneumonia, pulmonary embolism, etc. In this area, the radiologist will have a major advantage. It will be difficult for cardiologists to have training in pathologies outside the heart. RT: Some people say that’s the reason this will, or at least should, shake out as a collaborative effort. RCC: But the collaborative effort will not be cardiologists reading the cardiac portion and radiologists reading the rest. The collaborative effort will be the radiologists bringing their expertise in reading all the structures of the chest, including the heart, and discussing it with the cardiologist. ***
|
|
3801 Schuylkill Rd • Spring City, PA 19475 Publishers of Radiology Today All rights reserved. |