|
|
disembodied
human heart. “Watch this,” he says. vessels. The program then virtually unpeels the left anterior descending artery from the heart, straightens it out, and allows us to peer down its lumen. A few calcifications glint white.
The computer can measure a vessel’s diameter at different points along its length and then use this data to calculate its overall percent stenosis. A cardiologist examining an angiogram may give a more accurate estimate, admits Mezrich, but the radiologists are getting very close—thanks to new technology and the sweeping desire for noninvasive diagnostics.
A former electrical engineer with 25 patents to his name, Mezrich entered medical school at the age of 38. When he came from Boston to head the department two and half years ago, he brought not only a passion for research, but vision, momentum, and enormous enthusiasm for the existing and potential miracles of his field.
Only about thirty years have elapsed since the clinical introduction of CT scans, magnetic resonance imaging (MRI), positron emission tomography (PET scans), Doppler sonography (ultrasound), and digital imaging—technologies that revolutionized a specialty built upon the interpretation of plain film X-rays. Since the early 1970s, radiologists have “moved out of the basement,” says Mezrich. “We’re no longer hiding in the dark like mushrooms.”
Today, specialists throughout the hospital depend heavily on the department of diagnostic radiology—a reliance spoofed in one of Mezrich’s favorite lecture slides, depicting the main entrance to the hospital as a CT scanner. Actually, in the case of the R Adams Cowley Shock Trauma Center, the CT scan does play an immediate role in the evaluation of patients who arrive unconscious or in shock. It takes only thirty seconds to conduct a head-to-toe CT scan, revealing bleeds, breaks, or obstructions that might otherwise prove to be lethal diagnostic dilemmas.
Even in non-acute situations, radiology now plays an unprecedented role in the initial work-up of many patients. “That laying on of hands isn’t really done the way it used to be,” explains Mezrich. “We jump right to scan—which is good and bad. It’s good because it makes sense. It’s bad in a kind of romantic way, because it takes away a little bit of the intimacy between the doctor and the patient. It also drives up the cost of medicine.”
Assessing the Cost/Benefit Ratio
Understanding the cost-benefit ratio of diagnostic studies is the focus
of considerable research currently under way in the department. For example,
when a patient presents with atypical chest pain, might a quick cardiac
CT help prevent unnecessary cardiac catheterization? Might CT make sense
as a screening study for patients with certain risk factors for heart
disease? These are the questions driving the research of thoracic radiologist
Dr. Charles White.
 Charles White, MD |
White began his career in internal medicine and has maintained an interest in areas of overlap between the two fields. In his ongoing study of patients presenting to the emergency room with complaints of chest pain, the great majority of cardiac CT scans have been normal, providing families with needed reassurance. A normal chest CT rules out diagnoses such as coronary artery disease, aortic dissection, pulmonary embolus, pericarditis, pneumothorax, pneumonia, hiatal hernia, or rib fracture.
Beyond the emergency setting, White is also evaluating the practicality of cardiac CT as a screening tool. “For typical patients in their 40s or 50s who experience intermittent chest pain,” he explains, “the goal is to get a rapid look at the coronary arteries—and to this point, that has required catheterization.” CT holds the promise of identifying soft plaques, hard plaques, and vessel narrowing without the risks associated with invasive studies. In the coming year, the department is installing new 40-slice and later 64-slice CT scanners, capable of sub-millimeter imaging. “That’s pretty good when you consider that the left main artery is 5 millimeters in diameter. It’s great for any vessel bigger than 1.5 or 2 millimeters,” says White.
“A few years ago,” he points out, “a suspicion of pulmonary embolus was diagnosed with ventilation-perfusion scanning, followed if necessary with pulmonary angiography. But now, few if any suspected PEs are evaluated that way. At this point it’s multi-slice CT, and that’s a paradigm shift in the past few years. We can get pulmonary angiography noninvasively, and now coronary assessment is following that path.”
When combined with PET scans, which provide visual representation of differential metabolism within the body, the diagnostic power of cardiac CT is even greater. As Dr. Mezrich explains, “The idea is—and we’re not there yet, but we’re working on it—all right, this part of the heart muscle is not working hard. I suspect the coronary artery is not happy. Let’s go take a look at it. And I can do this in one swoop.” Increased reliance on this combined PET/CT study will likely increase the use of stenting, predicts Mezrich, as more people with coronary artery disease will be identified early. It may also reduce the need for emergency stent placement.
MRI simultaneously has a role to play in the evaluation of chest pain. Its ability to visualize perfusion and heart motion surpasses that of CT, but its ability to evaluate coronary arteries does not. “So we’re in this wonderful position of having multiple excellent tests to choose from, depending on what we suspect the patient’s problem is,” says Mezrich. “And as we’re trying to work out the best protocol for chest pain, we’re also trying to develop better communication among clinicians.”
Deciding which studies to order at each stage of a work-up is a challenge that has been complicated by the fast pace of technological advance. The best diagnostic study may not be the least expensive study, for example, yet cost cannot be ignored. Insurance companies rely on population-based cost-benefit analysis to determine coverage policies for high-tech screening. As a result, coverage does not always keep perfect step with the data supporting a particular study’s relevance for an individual patient.
For
example, consider virtual colonoscopy—or colonography—a CT scan of the
abdomen that Medicare currently covers as a second-line diagnostic study
when traditional colonoscopy fails (either because a lesion or a muscle
spasm blocks the scope from complete imaging), but which is not yet covered
as a first-line screening test.
Virtual colonoscopy still requires the patient to undergo laxative “bowel prep” in advance of the study, and to have the colon inflated with air, but then spares the discomfort and risks associated with an actual scope. The images allow the radiologist to visualize the interior of the colon and to examine lesions from any orientation; in addition, the computer simultaneously tracks the location of the virtual scope with great precision. As the virtual scope “advances” through the colon, a corresponding flashing dot tracks its movement on an image of the patient’s abdomen, orienting the viewer. If CT colonography becomes the standard screening test for colon cancer, then patients would require traditional colonoscopy only when a suspicious lesion demands biopsy. (And the departments of radiology and gastroenterology intend to coordinate care to allow same-day colonoscopy for patients who need biopsy, thus avoiding the need for repeat bowel prep.)
Improving
the power of diagnostic screening studies while minimizing their risks
and costs is the global aim behind many of the department’s research initiatives.
With Drs. Helen Mrose and Rao Gullapalli, Dr. Mezrich is exploring the
potential role of MRI in the early detection of breast cancer. It is an
approach that may ultimately replace mammography as a screening test,
particularly for younger women, whose denser breast tissue can limit the
diagnostic efficacy of mammograms.
Magnetic resonance imaging uses spectroscopy to detect choline, a substance often found in malignant tissue but never in normal tissue. The team’s research seeks to determine whether women who are at high risk for breast cancer, or who have had an abnormal mammogram followed by lumpectomy, might be better followed with routine MR studies than with mammograms. Currently, MR has a higher sensitivity than mammogram for detecting abnormalities, but cannot distinguish benign fibroadenoma from cancer. Continuing research aims to improve the specificity of the test; so that a patient can be told with confidence whether an abnormal finding is, or is not, malignant.
 Rao Gullapalli, MD |
Crossing Departmental Borders
In addition to developing better diagnostic and screening tests, other
department studies seek to apply new imaging technologies to the delivery
of treatment. The research of Raj Shekhar, Ph.D. and general surgeon Dr.
Adrian Park combines imaging modalities and puts them to work in the operating
room. By digitally coupling a patient’s pre-operative CT or MR study with
a real-time ultrasound conducted during an operation, a surgeon is able
to use the ultrasound as a quasi-translator, asking the computer to show
one or another angle of the CT image. The ultrasound linkage enables accurate
visualization of that particular patient’s anatomy from different orientations,
throughout the procedure.
Further research wedding imaging and treatment is taking place in the department’s division of nuclear medicine, where Dr. Bruce Line is investigating the anti-tumor potential of polonium, a substance that emits high energy alpha particles strong enough to kill tissue on contact, but with low enough penetrance that one could safely handle a jar of it. By attaching polonium to biologic molecules that naturally adhere to receptors found on the endothelial lining of new blood vessels, Dr. Line and his colleagues are developing ways to deliver polonium to malignant tumors, inhibiting new vessel growth and thus diminishing a tumor’s ability to grow.
Though most cancer treatments involving delivery of radiation come under the independent department of radiation oncology (which prior to 1981 was a division of the department of diagnostic radiology), certain cancer treatments involve collaboration between the two departments. Already in routine clinical use are the gamma knife, which delivers carefully targeted high-dose ionizing radiation to treat a range of intracranial tumors, and SIRT (selective internal radiation therapy), which delivers microscopic radioactive spheres, called SIR-Spheres®, directly to the site of non-resectable liver tumors. Both treatments avoid the risks of conventional surgery, while offering the benefit of precise radiation targeting.
 Helen Mrose, MD |
The hospital’s wide range of leading technology and clinical research attracts not only patients to the University of Maryland, but faculty and students as well. In the past two years, the department of diagnostic radiology has hired fifteen new clinicians and scientists, bringing the total faculty count to 50, plus 14 fellows. Efforts are also under way to expand the number of residents to 29. And though radiology is still an elective rotation for medical students, virtually all choose to take the hugely popular course directed by thoracic radiologist Dr. Robert Pugatch—a course affectionately known as “the Bob Show.”
Dr. Pugatch, whose reading room is always open to colleagues, residents, and medical students, sees education as a core imperative for the department. “No one should be afraid to wander in and ask for help interpreting a study,” he says, lamenting the loss of a more collaborative tradition. Clinicians, he explains, used to come in person to review their patients’ films with the radiologists, but hyperspecialization and the convenience of the PACS in every team room have hampered these consultations. (PACS stands for the filmless “picture archive and communication system” that enables treatment teams to access digital images and even preliminary radiology reports from terminals located throughout the hospital.)
Medical professionals outside of radiology need expert education about the field’s constant evolution. As Dr. Mezrich points out, “I can’t imagine that the surgeons and clinicians out there are able to keep up to date with all these new imaging studies when, to be honest with you, we’re having trouble keeping up with them!” Not surprisingly, fascination with diagnostic radiology abounds among medical students, whose comfort with digital technology reflects their lifetime in the company of computers. Last year the department received 550 applications for the seven available residency positions. It will take them four years of training to master the basic elements of their specialty.
 Bruce Line, MD |
“One of the things we’re trying to develop is computer-assisted ordering,” says Dr. Mezrich. “It used to be that in the old days the clinician would call and say, ‘Hey, Mezrich, what study do you think I should get for this patient?’ But there’s no time for these phone calls anymore. We’re trying to put some of this burden back on the computers; so that a doctor can tell the computer what’s wrong with the patient and suggest a study, and then the computer can say, ‘well, the last three hundred times someone ordered that study for this scenario, it didn’t work as well as this study. . . .’”
Typically it is the least informed doctors—new interns—who order the most diagnostic studies, usually beginning with the least expensive study and then working their way up. As technology has advanced to offer more and better study options, the downside to this tradition looms ever larger. Nobody wants to subject patients to unnecessary radiation exposure or to unnecessary delay in treatment. Informatics research applies computers to the challenge of establishing “best ordering practices”—protocols based on detailed analysis of patient outcomes. It is but one of the department’s many priority research fronts.
The hospital’s entrance may not be a CT scanner, but check out the first floor of the sunny, new Weinberg building, home to the clinical department of diagnostic radiology. Oh, and the emergency room. And the second floor. And the trauma center. You’ll find that radiology has outgrown the basement—an image simple to interpret.