As baby boomers enter their prime heart attack years, the CT industry is racing to keep up. Scanner technology is maturing so swiftly, it has been predicted that patients will soon be able to undergo a quick heart scan that tabulates an arterial calcium score, produces a detailed angiogram, and assesses ventricular function-all in one appointment and with no more than the prick of the intravenous needle used to deliver contrast media.
It's an opportunity that is energizing CT research and development among vendors and academic institutions around the world.
"We see cardiovascular imaging as one of the fastest clinical procedure and application growth areas," said H. David He, Ph.D., who manages GE Medical System's cardiology CT business. "It is going to be one of the key applications driving CT development for the next five years."
The emphasis on cardiovascular CT is also reshaping companies. He's position overseeing cardiovascular CT, for example, did not even exist until January. Similarly, Marconi Medical Systems has reorganized to create high-level positions to oversee cross-modality cardiology innovations in CT, PET, nuclear cardiology, and MR.
Michael S. Sims, who directs CT product strategies for Marconi, said heart disease and cardiac imaging are a strategic priority for his division.
"It is driving a lot of the advancements in multislice CT, if not most. Heart disease is the critical pathology-the heart is the organ with the critical coverage parameters," he said.
How big the cardiovascular CT market will become is unclear, but sales of Imatron's electron-beam CT scanner indicate surging interest. In July, the company reported revenues of almost $27 million for the first half of the fiscal year, nearly double the $13.5 million in revenue booked in the same time period in 1999. Sims forecasts an even more robust market for dedicated cardiac multislice CT scanners, betting on sales of 200 to 300 units annually, for totals of $300 to $450 million in revenue.
CORONARY CALCIUM
Coronary calcium scoring has become the bread-and-butter application of cardiac CT. Developed and validated on Imatron's electron-beam scanners, it became the focus of research with conventional CT scanners once spiral technology was introduced, and interest has intensified with the development of multislice capabilities.
The multislice threat has sparked a vigorous defense by Imatron and EBCT researchers. Dr. John Rumberger, medical director at Ohio Heart Ultrafast CT on the outskirts of Columbus, is a well-known researcher and user of EBCT. In May he blasted multislice CT users for casually adopting without independent validation the calcium scores and risk algorithms developed by him for EBCT.
"Until independent validation with pathologic, angiographic, and clinical standards can be established using spiral CT scanners for purposes of coronary artery imaging, the practice of suggesting they are 'just as good' as EBCT must be stopped. The claims are unsubstantiated, unfounded, and irresponsible," Rumberger said.
Validation studies involving multislice scanners are under way, but results may not be available for several years.
Imatron is also fighting the oft-cited prediction that multislice CT will become the preferred calcium scoring technology simply by being available in more hospitals and imaging clinics. Borrowing a page from the early days of MR, Imatron introduced a mobile EBCT scanner in September at the Heart Institute of the Cascades at St. Charles Medical Center in Bend, OR, and announced plans to add a second unit a month later. A full-scale marketing campaign promoting the mobile scanners is expected at the RSNA meeting this month.
Although entrepreneurs in both radiology and cardiology have recognized coronary calcium scoring as a business opportunity and have vigorously marketed it to the public, many questions remain about the role it will ultimately play in screening for coronary artery disease.
In July, for example, an expert task force of the American College of Cardiology and the American Heart Association issued a consensus statement that amounted to lukewarm support for coronary calcium screening as a way of determining the risk of heart attack or other cardiac problems. The test might be useful in determining whether someone who clinically appeared to be at intermediate risk was actually at high risk, the panel concluded. It did not, however, endorse widespread use of the test in people without symptoms or multiple clinical risk factors.
ANGIOGRAPHY
Even as the future of coronary calcium screening is being worked out, industry has refocused its most ardent attention on a more tantalizing target: coronary CT angiography.
"Calcium scoring is the easy application. That's actually not our focus," said John Sandstrom, CT division manager for Siemens. "The leading edge is in looking at angiography, both at stenosis on the one hand and attempts to evaluate soft plaque on the other."
Calcification is considered to be an indicator of stable plaque, whereas soft, fatty plaque is the site of active disease and the potential source of rupture that leads to blood clot formation and myocardial infarction. Even Imatron is focused squarely on promoting EBCT's strengths in noninvasive angiography, an indication for which the company received FDA marketing clearance last November.
"This procedure is performed noninvasively with a venous injection of contrast and is acquired using prospective triggering in a single breath-hold," said Jeff Sorenson, marketing director at Imatron. "This is a unique capability of EBT. No other vendor can acquire prospectively triggered images and cover the entire heart in a single breath-hold."
Indeed, multislice technology does not yet measure up to the challenge of both covering the entire heart in a single breath-hold and achieving the extremely high resolution needed for coronary angiography.
"Right now you can image the entire heart with a 2.5-mm slice thickness in about 20 to 40 seconds, depending on someone's heartbeat. But if you want to do coronary angiography, where you're looking at coronary arteries that are only a couple of milllimeters in size, there is a significant advantage to getting thinner slices, 1.25-mm or 0.5-mm thick. When you do that, your image acquisition exceeds the time the average person can hold their breath," said Dr. J. Jeffrey Carr, an assistant professor of radiology and public health sciences at Wake Forest University in Winston-Salem, NC.
Synchronizing image acquisition with the phase of the cardiac cycle, particularly with real-time modifications for an erratic heart rhythm, remains a challenge as well.
"There are changes in the R to R interval, and those vary. People often can have aberrant beats, and those with cardiac problems are more prone to have anomalies in their cardiac rhythm. So what you'd ideally like is some kind of interactive ECG that could identify those situations and, on-the-fly, change your imaging acquisition protocol to match," Carr said.
Industry insiders say help is on the way, although it will likely take another year or two before coronary angiography is ready for routine use. Nearly everyone agrees that the list of needed improvements includes faster scan speeds, more detector channels, better Z-axis resolution, better signal-to-noise ratio, new image acquisiton and reconstruction algorithms, and more powerful image processing tools.
At GE, some early research is demonstrating that it's possible to achieve an effective temporal resolution of 60 to 80 msec and achieve good coronary images, according to He.
"I think this represents the front edge of multidetector CT technology," He said. "Using ECG-gated acquisition and reconstruction, coupled with some clever image reconstruction techniques, you're able to really push much further."
Future development may require deciding which improvements are essential to providing clinically useful information.
"It really comes down to the price points of the product in the end. If I can get twice as many slices for x, or I can get a 30% rotation time reduction for y, which one is the better return? Once I can do a 125-msec scan, which you can do on the Volume Zoom, what's better?" Sandstrom said.
With today's technology, Siemens researchers are able to visualize the first two-thirds of the coronary arteries well and with reasonable reliability, but the distal third is less clear, according to Sandstrom.
"The first two-thirds of the coronary arteries are the area of interest, because if you've got an occlusion in the distal third portion, are you really going to treat that? You're exposing the patient to risk with less benefit," he said.
Marconi is prepared to up the angiographic ante at this year's RSNA meeting with enhancements to the Mx8000 line of CT scanners that will bring temporal resolution below 100 msec.
"All we're ready to talk about at this point is that the next-generation scanners from Marconi will have the ability to cover the entire heart in around 10 seconds with temporal resolution that approaches EBCT speeds," Sims said.
He acknowledged that achieving this will require image acquisition with more than four detectors, submillimeter Z-axis resolution, and gantry rotation times of less than 500 msec.
FUNCTIONAL ANALYSIS
Retrospective tagging and reconstruction of the volumetric data set also opens the door for functional analyses of the heart, Sims said. By reconstructing multiple phases of the cardiac cycle, it's possible to analyze wall motion, for example.
"You can do dual-phase end-systole and end-diastole reconstructions and chamber volume assessments, and you can do things that open up avenues for other advanced physiologic analysis, like ejection fraction or myocardial perfusion," he said.
At the RSNA meeting, Marconi will demonstrate software needed to determine ejection fraction, as well as the ability to do perfusion imaging in a limited area of the heart; for example, in an area known to have suffered a myocardial infarction.
"Ultimately, we'd like to be able to cover the whole heart over a long period of time, but to do that, you'd have to cover 15 cm all at once continuously over a period of 20 seconds. Because we're using very thin slices-1 mm or less-you would need at least 150 slices. That's down the road a bit, maybe three to five years," Sims said.
