MR imaging in cardiovascular medicine: Part One

Use is increasing to assess myocardial function, valvular heart disease, myocardial structure and presence or absence of infarction or ischemia.

Cardiology Today held a wide-ranging discussion before a live audience in Atlanta during the recent ACC meeting. Moderated by Carl J. Pepine, MD, and Roger Blumenthal, MD, the round table highlighted the use of MI, whether it is ready for more widespread use, and issues related to cost of incorporating its use into a medical practice. Part one is printed below; part two will be in our July issue.

Moderators
Carl J. Pepine, MD Eminent Scholar, Professor and Chief, Division of Cardiovascular Medicine, University of Florida, Gainesville, and Chief Medical Editor of Cardiology Today.	Roger S. Blumenthal, MD Director, Ciccarone Preventive Cardiology Center, Johns Hopkins University, and member of Today in Cardiology’s Editorial Board, Preventive Cardiology Section.
James A. Hill, MD Professor of Medicine and Pharmacology, Co-director of CTMR for CV Medicine, University of Florida, Gainesville.	Greg Hundley, MD Associate Professor of Radiology and Medicine (Cardiology), Wake Forest University Health Sciences, Winston Salem.
Nathaniel Reichek, MD Director, Cardiac Imaging Research Program, St. Francis Hospital-The Heart Center, Roslyn, New York, andProfessor of Medicine and Biomedical Engineering, Stony Brook University.	Samuel Wann, MD Clinical Professor of Medicine, University of Wisconsin and the Medical College of Wisconsin, and Director of Cardiology, Wisconsin Heart Hospital, Milwaukee
Norbert Wilke, MD Associate Professor in Radiology and Medicine, Chief of the CV MR/CT, University of Florida, Jacksonville.

CARL J. PEPINE, MD: Use of cardiac imaging procedures is increasing in the United States. Can you comment on the extent of that growth?

SAMUEL WANN, MD: There has been a huge growth in the number of studies done with echocardiography and nuclear medicine over the past three years. In 2005, there were more than 7 million nuclear myocardial perfusion studies, almost 25 million echoes, 2.5 million diagnostic caths, but only 5,000 cardiac MR studies. It is still early to have meaningful data on use of CT, but some data show that 35,000 coronary CTs were performed in 2005.

In my practice, we do lots of nuclear, even more echoes, a lot of cardiac caths, not many MRs, and we’re starting to do more and more CTs. Despite very exciting technologic advances in MR, MR has not had great penetration into clinical cardiology. This is in part related to economics and in part related to relatively few numbers of patients in whom MR provides unique and essential diagnostic information. MR has lots of competition.

NORBERT WILKE, MD: Those numbers may not reflect the real diagnostic capabilities of CT and/or MR as compared to nuclear imaging. Cardiac nuclear imaging in the United States is driven by its revenues. Cardiac nuclear medicine is the great revenue generator in cardiology compared to CT and MR where currently no solid reimbursement structure exists for these cardiac tests.

In other parts of the world, such as Asia and Europe, there are different reimbursement models. Therefore, these MR/CT numbers are quite different and clearly favor cardiac MR and CT much more. Nevertheless the numbers quoted give an idea how big the potential U.S. market could be, especially for MR rest-stress perfusion imaging.

PEPINE: Under what circumstances do you think MR offers us the best choice for cardiovascular imaging?

GREG HUNDLEY, MD: We are seeing an increased use in patients with heart failure. With a single test we can obtain information about myocardial function, valvular heart disease, myocardial structure or composition, and the presence or absence of infarction or ischemia.

PEPINE: I see our participants nodding. There is agreement that in the setting of heart failure or suspected ventricular dysfunction, cardiac MR offers unique information. We are now seeing many more patients with heart failure who also have devices. Is that a problem with MR?

JAMES HILL, MD: Yes, it is a problem. All of the current device manufacturers are looking at MR down the road and trying to decide how best to make their devices MR compatible. For patients with poor LV function, it’s almost certain that patients are going to have devices or will have them soon. Following them with MR is going to be a problem.

ROGER BLUMENTHAL, MD: Dr. Reichek, if you have a patient with a pacemaker or ICD, are there any of those patients that you’d feel comfortable sending for an MR and what precautions would you take?

NATHANIEL REICHEK, MD: We haven’t really had an occasion to do it. I’ve certainly followed very carefully the work that has been published on that issue. I would not be uncomfortable using MR on somebody who’s not pacer dependent and in whom there is an absolute need for information from an MR. The protocol would include essentially turning off the sensing and turning off the pacer if the patient’s not pacer dependent. It shouldn’t be done terribly often and one needs to be prepared to determine the thresholds afterward and to reprogram the device in an appropriate way if necessary.

As MR for noncardiac patients becomes so integral to the practice of medicine in patients with devices, I’m hopeful that there will be more effort made to solve the problem because it’s certainly not insoluble.

BLUMENTHAL: Are any pacemakers MR compatible at this stage?

Reichek: I think the published data suggest that recent pacemaker models are relatively impervious to adverse effects from being stuck in an MR scanner. If the patient is not pacer dependent and you absolutely need the data and you’re prepared to monitor the patient and do all the right things after the imaging study is complete, you can get away with it.

However, that’s a complicated set of events and not something that’s going to happen at your local outpatient imaging center.

WANN: The echo guys are very keen on using tissue Doppler and color kinesis to look at wall motion to select patients who might benefit from biventricular pacers. Unlike MR, echo does not require device makers to make magnetic-compatible devices so that follow-up studies after pacing can be performed.

Echo is here-and-now for evaluating valvular and ventricular function and for selecting people who might be appropriate candidates for biventricular pacing. It’s inexpensive, widely available, and can even be done in the intensive care unit or operating room. Why do we need MR?

Reichek: More could be said about the biventricular pacing story because it’s murky right now in terms of selection criteria. The most common reasons that we get referrals for clinical cardiac MR nowadays are really patients being seen by our arrhythmia group for one thing or another and it’s a constellation of different problems.

The young person with left bundle pattern PVC is a very common source of referral because of the concern about arrhythmogenic right ventricular dysplasia, which far outstrips the prevalence of the disorder.

Thanks to some recent clinical trials, ejection fraction has become the end-all for the population with advanced ischemic heart disease. Our EP group is getting tired of having problems due to the variability of the ejection fraction estimates from other sources.

We get a fair number of referrals for looking at viability in patients who may or may not be candidates for revascularization, and we are beginning to get a lot of referrals of problem cases in stress perfusion where MR has some capabilities due to improved spatial resolution and very short duration of study that lends itself to answering problems.

Commonly, we get referrals that say ‘stenosis of this diagonal, but is it really the flow-limiting thing that’s causing this patient’s chest pain or not?’ With MR you can do a pretty good job of looking for things like that with relatively limited areas of visible hypoperfusion.

The cardiomyopathy story also is one where MR has a lot to offer generally; we see a lot of reclassification of patients among people with supposed cardiomyopathy who turn out to be different from what their clinicians suspected.

BLUMENTHAL: If you suspect congenital heart disease, would you use MR?

Reichek: Actually, we use it extensively. Probably the single most common referral is just follow-up on somebody who had coarct, had it fixed, but is still hypertensive.

Nathaniel Reichek, MD

It’s an extraordinary tool for congenital heart disease, particularly in the adult and particularly in the postoperative patient. The pediatric cardiologists can do wonders with ultrasound because the acoustic window is so much better than in adults, but those patients keep coming back well into adult life by and large, and the ones who have prosthetic material placed also can be a real challenge to evaluate. MR has been very, very effective in those areas.

HILL: There are still areas with echo — no matter how small a child is — that are better imaged with MR. If you have a pediatric cardiologist that has specific interest in this as we do, you can get some amazing information. You can get a lot of things that you can’t get with angiography or echo.

So, I think that’s a very specialized but extremely useful thing and there’s no radiation for the child. You just have to sedate them. You don’t have to worry about heart rate for CT, you don’t have to worry about the risks of cath and so forth. It works extremely well.

PEPINE: Do you think MR technology could be applied to a broad range of patients?

Reichek: Right now, MR is far behind CT in performance characteristics for coronaries. There has been some real progress and there is the promise of more, but it is uncertain whether or not it will get to a point where it really is as good.

So far, the limits on the sensitivity of MR make it unlikely that you will get the slice thickness and spatial resolution that can be achieved with CT.

PEPINE: Tell us what you’re doing in terms of a MR stress protocol.

HUNDLEY: We’ve spent some time developing MR for stress applications at our institution. We try to utilize the advantages of MR — whether it be a wall motion assessment or a perfusion assessment or an infarct detection assessment — and make that available to those that need the information, whether that be a surgeon, an internist, a family practitioner or a cardiologist.

MR is well suited for performing stress testing. You don’t have, in general, limitations for body habitus. Once you get up to a certain size where you can’t fit in the hole, it doesn’t work well. But generally this technology is imaging protons and the more there are in that hole, the better the images.

We don’t have some of the acoustic window limitations with echocardiography or the attenuation issues with radionuclide scintigraphy. Moreover, the spatial and temporal resolution of the images is very high with MR and that can be very helpful for assessing different parameters during stress testing. For example, with temporal resolutions of 15 milliseconds or so, not only can one appreciate changes during systole and they can be visualized qualitatively or quantified, but one can also appreciate changes during diastole. With diastole generally being more sensitive to the presence of ischemia, that’s very important information.

Greg Hundley, MD

A big advantage for MR is the spatial resolution. Previously with echocardiography or radionuclide scintigraphy, there was no heightened spatial resolution to appreciate, for example, differences with subendocardial defects.

MR is well suited for stress testing. Something that we’re now working on and I think may become more pervasive is not just looking at the left heart, but looking at the left heart in relationship to blood flow in the aorta and the relationship of the left heart simultaneously with the right heart.

We have adenosine perfusion and dobutamine pharmacologic perfusion. This is an area of emphasis for us, but we’ve also begun doing treadmill testing.

Generally we’ll take images very similar to a stress echocardiogram: A patient gets off and gets on the treadmill and then goes back into the scanner. We collect those images within a minute of cessation of the exercise. In Europe, others are doing the exercise with a bike affixed to the leads.

PEPINE: Dr. Reichek, what about your stress protocol?

Reichek: We’re mainly doing adenosine stress perfusion and have found it to be an extremely practical, straightforward approach.

WANN: As much as I would like to, we do not use MR in our routine clinical practice, as the clinical volume of MR is too low to justify the expense in our dedicated cardiac private practice environment. Current multislice CT technology is superior to MR in visualizing the coronary arteries. Coronary arteriography provides the volume of patients that makes CT angiography viable in our small hospital.

However, we do recognize that one of the limitations of CT is lack of functional information about perfusion, ischemia and viability. There is a lot of interest among my nuclear colleagues in adopting CT, combining CT and SPECT or PET for hybrid imaging in the same patient to give us information about both coronary anatomy and tissue perfusion.

MR is clearly more elegant, and no radiation is involved, but there are a lot of gamma cameras in cardiology offices and hospitals across the country and very few cardiac MR machines. Perfusion imaging with CT/SPECT or CT/PET will be a competitor of MR. I think there are limitations with MR for looking at coronary anatomy, and there are limitations with CT for looking at perfusion. I suspect both of those will be overcome in the future. I don’t know which will come first, but certainly you can look at perfusion and do stress testing with CT just like you can with MR.

Iodinated contrast material, which is imaged on CT, is distributed in ischemic and infarcted tissue in a similar manner to gadolinium, which is detected by MR. Fast CT scanners that deliver a low radiation dose are being developed for serial studies of perfusion and viability, as well as coronary angiography.

PEPINE: What changes in technology would you like to see in the near-term to facilitate emergence of MR for more widespread use?

HILL: The major limitation with MR is the relative complexity of obtaining the image and interacting with the machinery. It’s not nearly as plug-and-play as CT is, although that’s not perfectly plug-and-play either.

If it’s going to have widespread use in the cardiology community and among cardiologists, that needs to be simplified.

WILKE: Use of cardiovascular MR over the last couple of years has drastically improved compared with when we started 15 years ago. There have been major improvements, particularly toward cardiovascular MR imaging. Making it more intuitive for cardiovascular MR imaging is an effort that industry is currently undertaking. Nevertheless, it takes substantially more training to get into cardiovascular MR imaging than it does to get into cardiac CT. .

Reichek: A lot of things have not been made user-friendly about MR, both in data acquisition and especially in the data analysis. Some of these things were done years ago for echocardiography and nuclear. Cardiac MR should be a lot more user-friendly for both acquiring images at the console and analyzing the data.

BLUMENTHAL: Dr. Hundley, can you comment about the use of MR in large epidemiology-type studies, such as MESA and ARIC and others, where the focus has been more on echo as an imaging modality. What advantages could MR bring to some of the ongoing research studies?

HUNDLEY: One of the benefits of the larger population-based studies is that we’re applying quantitative methods that heretofore we haven’t had the capability to utilize; for example, tissue tagging. Tissue tagging allows us to track shape and deformity changes and very small quantities of both the right and the left ventricle during a resting exam.

We’re beginning to glean data that there are early changes in shape deformities on a regional level, on a global level, very small changes in the weight of the heart, the mass of the heart, and changes in its volume that we haven’t had the reproducibility or the accuracy with another technique to appreciate. I think we will begin to learn about new predictors for cardiovascular events. That data will be forthcoming in the next two or three years.

Another leap forward with this technology will be the ability to look at some of our traditional risk factors and link them to an outcome in the heart that will define mechanisms of disease that we haven’t been able to appreciate before because we haven’t had a technology suitable for picking up those changes.

MR is noninvasive. It doesn’t use ionizing radiation, and there is no nephrotoxic contrast. That allows us to not only sample someone once but to perform longitudinal studies over time. I think we will see applications not only in clinical trials but also in managing our patients.

Reichek: Some of the things that MR has been used for in MESA are really extraordinary and will rewrite our understanding of pathophysiology. A paper from Michael Jerosch-Herold’s lab on stress perfusion in a subset of a MESA population has shown direct associations with risk factors — not with established coronary disease but with risk factors — such that smokers or hypertensives have impaired perfusion reserve. But we don’t know the mechanism.

We were accustomed to thinking that may well be true but is mediated by coronary atherosclerosis and stenosis. However, evidence emerging from MESA stress perfusion studies in that regard and also in some data that has not yet been published fully from Hopkins on mechanical function show reductions in contraction of myocardium in patients with risk factors but not clinically overt stenotic coronary disease.

These are extraordinary developments that may well rewrite our understanding of the natural history of cardiac dysfunction in the time course of atherosclerosis and change strategies that emerge for secondary prevention.

James A. Hill, MD

HILL: MR is such a multipotential technique that it becomes confusing to the average person who looks at this. People are going to say, I don’t really want to think about all the different things that MR can do for me because I’m really not sure what they are necessarily. Let’s just do an echo. Let’s just do a nuclear study or what have you. I think that’s the real challenge at this point.

BLUMENTHAL: One of the most highly publicized studies out of the [March] ACC meeting is ASTEROID. Dr. Steve Nissen has been one of the greatest proponents of IVUS to look at ways that we can affect cardiovascular disease progression. What are your thoughts about MR’s potential to show the usefulness of some of these new technologies and treatments?

Reichek: MR is probably the best method if you don’t have to do it in the coronaries. Work at Johns Hopkins is looking at plaque progression with MR showing that it can be done extremely well in the carotids and the thoracic aorta. Surveying widely the coronary tree for plaque with MR is really just not practical because you get limited exams in a few selected places. It’s an area for research for that purpose.

If the strategy evolves of using, say, carotid plaque burden as your biologic marker for indications for and effects of aggressive therapy, then MR is a superb way of doing it. MR can give you not only plaque volume, but it’s also extremely good at doing plaque composition in those vessels.

BLUMENTHAL: Dr. Joao Lima and others have shown stabilization and perhaps some modest regression in aortic plaques. Do we have any data in animals that plaque regression in the aorta corresponds to some plaque stabilization or regression in the coronaries?

Reichek: That data don’t exist right now. Actually, what probably needs to be done is figuring out some way to couple the coronary plaque capabilities of CT and the plaque composition and other artery capabilities of MR and to use them in concert to get some answers to these kinds of questions.

BLUMENTHAL: Dr. Chesebro [in the audience], you’ve been a leader in the imaging field and in atherosclerosis progression for a long time. What do you think?

JAMES H. CHESEBRO, MD: It seems that the current strength of MR could be in seeing about a positive stress test in a very accurate way. There’s attenuation with nuclear; we see a number of some false positives with the echoes and the echo, in part, is dependent upon who is reading the wall motion. If you see a plaque that’s high grade with CT, you’ve still got to do the coronary angio to get the stent in. The surgeon isn’t going to want to just see what’s on the CT because he can’t necessarily see the distal vessels well enough or there’s so much calcification he can’t necessarily assess some of the lesions.

TY J. GLUCKMAN, MD, (in the audience): What do you see as the future of real-time MR, for instance, in a cath lab or EP suite? We hear about it and we read about it on the experimental level, but how far away are we from incorporating it into everyday use?

HILL: Clearly cost is going to be an issue in terms of the cost benefit of adding another piece of expensive equipment to an already expensive lab. If there’s no change in radiation, if there’s no change in outcome from a success rate standpoint, it’s not clear to me that that’s going to be beneficial.

It may be a little bit better in the EP lab because it may save you some radiation in that respect, but at this point it’s a little early to tell.

MAHFOUZ EL SHAHAWAY, MD, (in the audience): What is the difference in cost for diagnosing diastolic heart failure with MR vs. echo and the cost for the practice ?

HUNDLEY: For echo, the 2-D is right around $320 with the technical and professional fee, but the MR is about $480.

If you add the 3-D echo charge, the echo cost goes up to $420 or so. For patients with diastolic heart failure, we’ve found MR to be useful for precise determination of left ventricular mass. It can quantify relaxation and contraction. We’re also spending some time looking at the relationship between the vasculature and the heart in diastolic heart failure patients, and we can collect information on aortic wall stiffness, extensibility wall composition, as well as peripheral endothelial function.

You can’t reliably acquire vascular information with echo or the pulse wave velocities in all patients due to body habitus. The tonometry requires a whole other sitting in a different exam. The measures can be acquired in the same exam with MR. For looking at patients with a diastolic heart failure syndrome, we’ve found MR to be an advantage.

With optimal acoustic windows and sonographers, you can do some of the quantitative techniques that we discussed, and I think you can obtain good information on diastolic and systolic function and mass with an echocardiogram.

In our experience, with an MR you can obtain reliable information in 95% of men and women, regardless of body habitus. At our institution, we’re seeing some use in this area.

Audience participants whose comments are included: James H. Chesebro, MD, is professor of medicine at the Mayo Medical School, Mayo Clinic, St. Luke’s Hospital; Ty J. Gluckman, MD, is a fourth-year fellow at Johns Hopkins and a member of Today in Cardiology’s Fellows Advisory Board; and Mahfouz El Shahawy, MD, is at the Cardiovascular Center of Sarasota.