A CVMR practice demands training

Radiologists who want to start and maintain a successful cardiovascular MR practice need to complete adequate training, work collaboratively with cardiologists, and educate referring physicians about the benefits of MR over other modalities.

In 1999, Dr. Scott Flamm started the cardiovascular MR lab at St. Luke's Episcopal Hospital, home of the Texas Heart Institute, in Houston. The lab has grown from three employees to nearly a dozen and from 500 studies a year to over 3000. Case studies split evenly between cardiac and vascular are performed on two commercial 1.5T MR scanners. The ratio of outpatient to inpatient exams is 60:40.

In cardiac applications, MR is used for primary issues of ischemic heart disease: looking at viability, perfusion, coronary arteries, and various cardiomyopathies. The vascular aspect addresses the standard thoracic and abdominal aorta studies, some peripheral runoffs, and a large number of renal artery studies. The practice did very few renal studies five years ago, but the staff developed their MR techniques and were able to get very good high-resolution studies, and renals now account for about 20% of their studies.

"We know we're very good at imaging the renals," Flamm said. "We performed an internal study with 134 patients who also had conventional angiography. The sensitivity and specificity of MR were 91% and 90%, respectively."

The lab is researching the use of viability techniques in patients with sarcoidosis. Researchers want to determine how many of these patients actually have sarcoid involvement in the heart, a heretofore unknown statistic. They also are setting up protocols to use myocardial stress perfusion as a technique to determine myocardial ischemia on a routine clinical basis.

Flamm, who has an appointment in both radiology and cardiology at St. Luke's and Texas Heart, espouses a collaborative approach to working with cardiologists. The best way to maintain their trust and confidence is by being competent, staying on the cutting edge with the clinical research, and giving them the information they want and need. Having an understanding of cardiac physiology and pathophysiology is also important.

"Whether you're talking to a surgeon, congestive heart failure cardiologist, or electrophysiologist, you have to learn what their clinical scenarios are," he said.

In assessing left ventricular dysfunction, for example, Flamm supplies cardiologists with, among other data, the end diastolic and end systolic volume. The correlatives in echocardiography are end diastolic and end systolic diameter. Those are important parameters based on the MADIT II trial for determining whether a patient can receive a biventricular pacemaker. But the end diastolic diameter is a poor surrogate for end diastolic volume, Flamm said.

"The number you really want to have is end diastolic volume. However, since all of the MADIT II trial data were based on echocardiography, which included end diastolic diameter, we're stuck," he said.

Flamm includes both numbers in his reports, and although some cardiologists use the end diastolic volume diagnostically, it will take years before there is any wholesale change among the majority of cardiologists.

Adequate training is essential for radiologists interested in starting a CVMR lab. It's not enough to spend several months learning cardiac MR. The field should be viewed like any other subspecialty of radiology and require a full-year fellowship, he said.

Once the lab is staffed with well-trained technologists, a physicist, nurses, and an image processor, it needs patients. Flamm focuses mainly on cardiologists and cardiovascular surgeons for referrals. Advertising is personalized to each physician's specialty. To a heart failure cardiologist, for example, he will promote MR's ability to quantitatively assess left ventricular function in patients who often are difficult to analyze accurately by an echocardiogram. Cardiothoracic surgeons who perform many bypasses may be swayed by MR's ability to assess myocardial viability. Cardiovascular surgeons who do thoracic or abdominal aortic surgery will hear about CVMR's vascular prowess, including high-resolution images, 3D renditions, and potential additional data such as left ventricular function and aortic valvular function that would not be included with a CT scan.

Besides the direct advertising approach, Flamm is always selling the virtues of CVMR in his daily one-on-one interactions with physicians, at specialty society conferences, and at cardiology noon conferences. At the noon conferences, questions invariably arise about MRI's usefulness with certain patients.

"If no one is there to promote it, the question just hangs," he said.

He also lectures to cardiology and radiology fellows, an important educational endeavor because these physicians are the ones on the ward thinking about what test to order.

Cardiologists are not the only specialists who need to be educated about CVMR, however. Nephrologists want to know about the high-resolution studies and availability of non-nephrotoxic contrast agent. General internists and family practitioners order scores of echocardiograms to assess ventricular function or valvular function. These physicians can be schooled about MR as a tool to evaluate ischemia.

Another selling point is MR's ability to pick up incidental findings. A cardiac scan always includes axial images through the chest. Although these are not the same as high-resolution CT scans, they nonetheless can show pleural effusions, significant masses within the pulmonary parenchyma, or adenopathy within the mediastinum-information that can help determine patient management.