Organized radiology has struggled in recent years with consumer-driven imaging exams that are not supported by valid research. The most controversial are whole-body CT and MR screening, lung cancer CT screening, and coronary calcium screening.
At first, concerns about CT screening were hard to hear over the din of the public's excitement about discovering asymptomatic disease at a curable stage. Slowly, word spread. Public education, coupled with a downturn in the economy, forced many entrepreneurial screening centers to close.
Now another era of CT technology carries promises as well as pitfalls. Cardiac CT angiography has proved to be an effective tool to rule out coronary artery disease in patients with acute chest pain. It also has become the de facto test to rule out pulmonary embolism and aortic dissection. With 64-slice scanners, practitioners want to extend their diagnostic reach. They'd like to have a single 20-second scan for CAD, PE, and aortic dissection: the so-called triple ruleout. Is this possible? More important, is it necessary? The answers to both questions vary depending on whom you ask.
"Triple ruleout is the biggest catastrophe in the history of diagnostic imaging," said Dr. Lawrence Boxt, chief of cardiac imaging at North Shore Long Island Jewish Health System. "There are no data supporting it, and no one knows how to do it. It's a great idea, but it's becoming a holy grail."
Combining coronary CTA with CTA of the chest and aorta is challenging. Patients must be scanned through the chest and the heart, and the chest scan must be gated because of heart motion. With 16-slice scanners, a gated scan through the chest takes about 35 to 40 seconds, too long for people with shortness of breath. With the 64-slice scanners, the breath-hold decreases to 10 to 15 seconds, and the holy grail seems within reach.
One problem with the triple ruleout is the radiation dose. Using a 16-slice scanner, Dr. David A. Dowe, medical director of Galloway Coronary CTA in Galloway, NJ, routinely performed what he calls "the poor man's triple ruleout" by following a coronary CTA with a spiral scan of the chest. But when he learned that the dose for the procedure-between 8 and 10 mSv-carries a one-in-2000 risk of inducing cancer over a lifetime, the procedure ceased to be a routine option.
Another issue is that only 20% of patients who present with acute chest pain have CAD, said Dr. Udo Hoffmann, a radiology instructor at Harvard Medical School. It's important not to expose the other 80% to too much radiation or too much contrast. The 64-slice scanner makes this feasible with dose modulation techniques and faster scanning that requires less contrast material.
Even if the radiation dose is high, many radiologists say that the benefits of a CT scan for someone with acute chest pain generally outweigh the risks of radiation exposure, particularly if the patient is over 50 years old. Data should be available by the end of the year regarding protocols and techniques one can employ that optimize scan quality while reducing the effective dose delivered to the body by 64-slice scanners.
WILD, WILD WEST
The lure of cardiac CTA should not be underestimated. This month, two professional societies geared toward promoting the clinical use of cardiac CTA will meet at the American College of Cardiology meeting to see if they can overcome their differences and merge into one society. A focus of each group is to reach consensus on education, training, and quality assurance.
"One reason for standardizing protocols is to avoid the 'Wild West' mentality that accompanied the early days of calcium scoring and lung cancer screening," said Dr. U. Joseph Schoepf, director of CT research and development at the Medical University of South Carolina in Charleston.
The genesis of the triple ruleout stems from the ability of CT to diagnose or rule out PE, according to Dr. Charles White, director of thoracic imaging at the University of Maryland Medical Center in Baltimore. A few years ago, patients with suspected PE had to go to the nuclear medicine department. Now they can be immediately evaluated in the emergency room (where the aorta is also appraised). Recent studies, however, have questioned the ubiquity of the exam, in particular, the low percentage of positive findings and the high radiation dose delivered to the breasts of younger women.
But the ease of performing the exam and its accuracy in assessing PE, along with the negative predictive value of coronary CTA, have whetted the appetites of physicians. Scott Schubert, global product manager for CT at GE Healthcare, said that radiologists and ER physicians have told the company that it would be a significant breakthrough if all CAD, PE, and aortic dissection could be assessed in a single exam.
The challenges were daunting. The researchers had to find a way to scan through the apex of the lung in the highest resolution possible, through the heart, and then through base of diaphragm, with a single contrast injection and a short breath-hold.
"We heard that 20 seconds or longer would be inappropriate. And we couldn't just increase the pitch to shorten the time because that would have degraded the resolution," Schubert said.
Dr. W. Dennis Foley, a professor of radiology at the Medical College of Wisconsin, has been using a prototype of a GE 64-slice scanner since June and a commercial model since November.
He has adapted the timing of the contrast to take advantage of the 64-slice scanner's rapid acquisition. A 17-second gated acquisition can deliver images of the coronary arteries, lungs, and the thoracoabdominal aorta. He uses 6 mL of contrast per second, for a total of 80 mL, and delays the contrast for four seconds after the aortic arrival before imaging in order to have a sustained plateau of enhancement.
"You want aortic enhancement to be 300 HU or greater," Foley said.
The gated study can be extended to the aortic bifurcation, which will increase the contrast proportionally. Foley prefers to do the complete workup.
It's too soon for peer-reviewed studies involving 64-slice scanners, but the 2005 RSNA meeting will likely include a number of abstracts on the subject. Studies that Foley and his colleagues hope to present will cover:
- a comparison of CTA with conventional catheter coronary angiography;
- the reliability of the new scanners for imaging the aortic valve and proximal coronary arteries; and
- the diagnostic information that is included in the gated scans of patients with suspected aortic dissection.
While the term "triple ruleout" implies that only three indications are examined, radiologists actually look for other heart and chest abnormalities, and are responsible for incidental findings. At Massachusetts General Hospital, Hoffmann and colleagues call the test a high-resolution ECG-gated thorax. They look at function, perfusion defects, and other cardiac diseases such as pericardial effusion, aortic or mitral valve infection, pneumonia, PE, and aortic diseases. They image from the aortic root and move to the bottom of the heart, adding 4 or 5 cm to the normal coronary CT exam. On the 64-slice scanner, this examination takes 15 to 17 seconds and requires an average of 80 mL of contrast.
"Normally, for a 15-second scan, we would only need 60 mL of contrast, but we give the extra 20 mL to keep the pulmonary arteries enhanced," Hoffmann said.
WHO'S STAFFING THE ER?
The major value of CT in people with acute coronary syndrome is minimizing liability, according to Hoffmann. The threshold for admitting patients with chest pain is very low, and doctors in the ER are reluctant to discharge anyone. Many patients have 24-hour follow-up in the hospital; one-third of them have no findings in the coronaries. A patient with a normal CTA can be sent home immediately.
Although cost-effectiveness analyses of coronary CTA have not been done, other studies have shown that patients tested with myocardial perfusion imaging have a 15% decrease in hospital admission, leading to a cost reduction of about $1000 per patient, Hoffmann said.
Hospital administrators recognize that chest pain is an expensive clinical indication, said Sean McSweeney, CT product manager for Toshiba America Medical Solutions. If all the markers are there, the patient is sent to the cath lab. But if the ECG is nondiagnostic, additional tests, such as stress perfusion, are expensive and often nondiagnostic as well. The need for a definitive ruleout has huge potential, he said.
The newest generation scanner at the Medical University of South Carolina will definitely be cost-effective, Schoepf said. Most patients with acute chest pain are indeterminate and require a huge amount of time, space, and resources in the ER. To have a single test that would differentiate those who need acute intervention from those who do not would be extremely beneficial.
Some radiologists are not convinced. A facility would need to do 10 to 12 scans daily on a 64-slice scanner to pay the $1.8 million price tag, according to Boxt. Both radiologists and cardiologists are needed to evaluate the triple ruleout study, and the number of ancillary personnel will increase as the exam becomes available 24/7. The economics will then change dramatically.
"It's like a firehouse. The city has to have it whether there is a fire or not," he said. "Physicians and nurses get paid whether they are doing a study or not. A scanner now becomes a 24/7 machine, and the overhead rises dramatically."
Studies to prove the scanner's cost-effectiveness will require subjecting patients to CTA and conventional angiograms, two loads of contrast, and extra radiation. Is it worth it for one-third of patients who are nondiagnostic in ECG? That question will be answered in time, Boxt said.
Boxt and other skeptics are assuming that the scanner is used only for cardiac studies. But there is no reason a 64-slice scanner in the ER should be limited to cardiac imaging, said Dr. Elliot K. Fishman, director of diagnostic radiology and body CT at Johns Hopkins University. Almost every hospital has a CT scanner in its ER, and it's not that great a leap to purchase a 64-slice machine for the next upgrade.
"These scanners are ideal for the ER. They handle difficult patients and large volumes of data. We use them now in the ER to evaluate aortic aneurysms, vascular trauma, and perfusion for stroke. Cardiac is just another good reason to buy a 64-slice machine," Fishman said.
Mr. Kaiser is news editor of Diagnostic Imaging.
