Coronary CTA takes giant leap with 64-slice scanners

October 5, 2005
David A. Dowe, MD

Coronary CT angiography (CCTA) was initially performed on four-slice multidetector CT scanners. These examinations were confined to evaluation of the proximal coronary arteries because the studies were limited in spatial and temporal resolution and were plagued by long acquisition times that required intolerably long breath-holds. Sixteen-slice scanners produced revolutionary improvements in CCTA technique: For the first time, complete coronary circulation exams with breath-holds in the 15 to 20-second range could be routinely obtained.

Coronary CT angiography (CCTA) was initially performed on four-slice multidetector CT scanners. These examinations were confined to evaluation of the proximal coronary arteries because the studies were limited in spatial and temporal resolution and were plagued by long acquisition times that required intolerably long breath-holds. Sixteen-slice scanners produced revolutionary improvements in CCTA technique: For the first time, complete coronary circulation exams with breath-holds in the 15 to 20-second range could be routinely obtained.

Breath-holds of this length are easily tolerable in patients who do not have chronic obstructive pulmonary disease (COPD). Submillimeter slice thickness significantly improved spatial resolution, while gantry speeds of less than a half-second improved temporal resolution. The most important improvement, however, is the increased signal-to-noise ratios achieved with high-milliamperage generators. These generators provide enhanced small vessel visualization, better spatial resolution, and a significant reduction in blooming of calcified plaques.

Sixteen-slice CCTA achieved very favorable results in detecting a > 50% stenosis compared with catheter angiography, far exceeding any sensitivity achieved with nuclear stress tests.1-4 More important, 16-slice CCTA demonstrated negative predictive values > 97%, effectively ruling out coronary artery disease. Sixty-four slice scanners are building on the improvements seen with 16-slice CT scanners. They promise to make CCTA the first diagnostic exam in the workup of patients with chest pain and to replace catheter angiography as the diagnostic tool for evaluation of the coronary arteries. This development is similar to the replacement of catheter angiography by noninvasive imaging in every vascular bed in the body with the exception of the heart.

TECHNICAL IMPROVEMENTS

I use a 64-slice LightSpeed volume CT (VCT) scanner (GE Healthcare) to perform all CCTA examinations. A number of technical modifications have dramatically improved image quality. The most immediate of these is the reduction in acquisition time that results from the ability to obtain 64 x 0.625-mm MSCT scans. This detector provides true 4-cm coverage. Acquisition times and, hence, breath-hold times for CCTA are now five to six seconds. Even patients with severe COPD can tolerate this breath-hold, but all patients benefit from the reduction in phase misregistration because we use data from no more than six heartbeats. The subsequent reduction in reconstruction time has greatly improved workflow. Gantry speeds continue to decline; the gantry speed of the VCT is 0.35 second. This reduces temporal resolution to 175 msec with single-sector analysis, 87.5 msec with two-sector analysis, and 44 msec with four-sector analysis. Higher heart rates are no longer an obstacle to scanning and reconstruction using single-sector analysis (Figure 1).

Due to the dramatic reduction in acquisition time for CCTA, exams can be accomplished at a maximum mA of 800 on serial patients without confronting tube cooling issues. Image quality has thus improved for CCTA's most difficult patients, the morbidly obese.

Blooming of calcified plaques has been the major limitation in evaluating the coronary arteries, accounting for the persistent positive predictive value in the 80% range compared with the negative predictive value of > 97%. Technical advances in spatial and temporal resolution with CCTA have reduced the blooming artifact (Figure 2).

EXAM TECHNIQUE

Patient selection remains the same for 64-slice CCTA as for 16-slice examinations. The top four indications are chest pain with CAD risk factors, chest pain without CAD risk factors, patients being considered for a diagnostic coronary catheterization, and asymptomatic high-risk patients.

Patient preparation is also the same. I still use oral beta blockers for every CCTA exam. With the 64-slice scanner, my goal heart rate is < 70 bpm as opposed to < 60 bpm with a 16-slice MSCT scanner. Higher heart rates can easily be accommodated without the use of beta blockers, but I prefer to use single-sector analysis, which renders superior image quality. Beta blockers have advantages in addition to decreasing the heart rate: They limit variability, diminish anxiety, and assist in decreasing radiation dose when combined with routine use of ECG-gated dose modulation.

I favor oral rather than intravenous beta blockers. Oral beta-blockers are well tolerated, the patient can take them at home prior to the exam, and the patient does not need the extra time in the CT scanner and extra attention from a nurse that use of intravenous beta blockers requires.

Given the true 4-cm coverage of the detector, 0.625-mm slice thickness is possible in all patients, even those with bypass grafts (Figure 3). With 16-slice CCTA, the volume of tissue coverage in a bypass graft case required a breath-hold of > 25 seconds. For this reason, 16-slice CCTA scanned bypass grafts at 1.25 mm, which reduced the acquisition time by 50% but sacrificed spatial resolution in the z-axis.

True 4-cm coverage also has affected CCTA of the morbidly obese patient. Using 16-slice CCTA, I would sacrifice spatial resolution by acquiring data at 1.25 mm in order to shorten the acquisition time, which in turn allowed me to use higher mA. The higher mA greatly improved my signal-to-noise ratio at the expense of increased radiation exposure. On the 64-slice scanner, 0.625-mm slice thickness in obese patients is routinely possible at 800 mA, due to the five to six-second acquisition time made possible by this detector's 4-cm coverage.

The shorter acquisition time enables me to use a lower contrast volume. I currently use a triphasic injection technique (40 cc contrast at 5 cc/sec, 20 cc contrast plus 30 cc saline admixture at 5 cc/sec, 20 cc saline at 5 cc/sec), for a total administration of 60 cc of contrast. When combined with the manual timing bolus run, the patient receives a maximum of 80 cc of contrast. I do not taper the injection rate. Once circulation time has been calculated, my goal is to get the highest peak injection in the coronary arteries by maintaining a high injection rate.

TRIPLE RULE-OUT EXAMS

A triple rule-out examination is a single acquisition that results in simultaneous arterial phase opacification of the pulmonary arteries, coronary arteries, and aorta. The radiologist can rule out pulmonary embolism, CAD, and aortic dissection or aneurysm with a single 12-second scan. Triple rule-out exams require the use of 110 cc of contrast. Postprocessing includes routine reconstruction of the coronary arteries (Figure 4), wide field-of-view coronal reconstructions of the lungs (Figure 5) and aorta, and "candy cane" oblique sagittal reconstructions of the thoracic aorta.

Radiation dose for this exam is increased secondary to the necessary coverage. In my opinion, this exam will revolutionize the imaging of patients with acute chest pain. This represents an enormous volume of work originating in the emergency room, and radiology groups will struggle to manage this work. The only factors that may control use of this examination are the risks imposed by the increased radiation dose and increased volume of contrast.

VULNERABLE PLAQUE

Noncalcified, or soft, plaques are more vulnerable to rupture than calcified, or hard, plaques. Coronary artery calcium scoring is limited to the detection and quantification of calcified plaque burden. The presence of coronary atherosclerosis can be verified based on the presence of calcified plaques. Atherosclerotic plaque burden, measured with the calcium score, in turn implies the likelihood of a flow-limiting stenosis. The calcium score is also an independent predictor of mortality and is therefore considered additive to the Framingham score.5 It is limited, however, in younger patients whose disease has not existed long enough to calcify, and it does not detect the soft, more vulnerable plaques that may be present in any age group regardless of calcium score.

Patients of any age with any calcium score value may have stenoses from soft plaques or any degree of overall soft plaque burden (Figure 6). CCTA has the advantage of detecting hard and soft plaques and assessing the degree of stenosis for the entire coronary arterial circulation with minimal increased risk above and beyond the calcium score. Compared with intravascular ultrasound, 16-slice CCTA has shown an acceptable sensitivity for detecting soft plaques,6 which greatly increased when the study was limited to detection of plaque in the proximal vessels.7 The importance of this finding is that the risk of sudden death and large, disabling myocardial infarcts decreases dramatically when a plaque ruptures distal to the first septal perforator branch of the left anterior descending artery.

Plaque characterization by Hounsfield unit (HU) measurement has been criticized as inaccurate, based on signal-to-noise ratio limitations. Although this criticism is partly valid, the plaque composition itself contributes to this variation. Plaques are heterogeneous in their composition, reflecting various rates of healing. Visualization of atheromatous tissue within a heterogeneous plaque is possible thanks to the improved temporal and spatial resolution of modern scanners (Figure 7). Applying a color scale to HU measurements may increase the accuracy of soft plaque detection and characterization and improve the ability to assess the degree of stenosis by the application of color to contrast HU ranges (Figure 8). This technique requires further validation by intravascular ultrasound. In the future, it will be possible to quantify plaque burden and subdivide that value into the percentages of atheromatous, fibrous, and calcified tissue present within the total plaque burden.

STENT AND BYPASS GRAFT EVALUATION

Detection of in-stent and stent margin restenosis first became possible with 16-slice scanners compared with cardiac catheterization8 and should improve still further with 64-slice scanners. However, stents that are old and overgrown by endothelium, distal stents, and small stents must be treated with caution. In these cases, it may be difficult to directly visualize the stent lumen. An early observation with 64-slice scanners is that they may greatly enhance our ability to evaluate stents compared with 16-slice CCTA (Figure 9).

CCTA imaging of bypass grafts has always been accurate and fast. The saphenous vein grafts are large and extracardiac and therefore are less prone to motion artifact and have no branches. Arterial conduits, although smaller in caliber, share these imaging advantages. The major advancement with 64-slice CCTA is that we now image bypass grafts with the superior spatial resolution of 0.625-mm slice thickness made possible by the 4-cm-long detector. Visualizing bypass graft stenoses is equally easy, but the enhanced spatial resolution results in more accurate evaluation of stenoses (Figure 10). With more complete assessment of stents and bypass grafts, we are able to triage patients with new symptoms or new stress test findings into two categories: patients with widely patent stents and grafts showing progression of CAD within their native vessels, and those with stenoses in their interventions.

RADIATION DOSE

Radiation dose remains a serious issue with CCTA. Increased dose is inevitable because of the very low pitch values (0.20 to 0.26:1) and retrospective gating. ECG dose modulation decreases dose by reducing the mA of the x-ray beam during portions of the cardiac cycle when we are unlikely to reconstruct the coronary arteries. I use peak mA during the 40% to 80% window of the cardiac cycle only. This limits the radiation dose outside of this time window but still affords me the flexibility I need for multiple phase reconstruction, should the patient's heart rate change significantly during the exam. Many other dose-reducing technical advances are available on the 64-slice scanners, including cardiac filter modes and bowtie filters, but discussion of these is beyond the scope of this article.

Effective radiation dose is calculated by multiplying the dose length product by a k factor that represents different tissues' sensitivities to radiation-induced damage. The result is expressed in units of millisievert. The effective radiation dose of other cardiac and noncardiac imaging studies yields a few surprises when compared with the much maligned dose from a CCTA exam (see table). The dose from a typical cardiac catheterization varies from 3 mSv for a routine diagnostic study with only a few projections to 30 mSv for a congenital heart disease exam. A stress SPECT thallium exam is typically around 25 mSv, due to the long half-life of thallium, and a stress SPECT sestamibi exam dose is about 12 mSv.9

Using a 64-slice CCTA scanner with an acquisition time of five seconds produces a radiation dose between 6 and 13 mSv, depending on the mA used during the study and the patient's heart rate. This places 64-slice CCTA well within the historically acceptable range of radiation exposure for other invasive and noninvasive cardiac imaging exams. This fact, combined with the additional fact that CCTA is the only exam that noninvasively images the coronary artery wall, the end organ that causes CAD, speaks very well for its widespread use in the future. But such use must be tempered by the effects of radiation to the individual and to society as a whole. This issue becomes more significant in children and women, given the increased sensitivity of breast tissue to radiation damage. We are not a pediatric center, and I use CCTA in pediatric patients only when a coronary anomaly is a concern (Figures 11 and 12). Furthermore, the effective dose of a triple rule-out exam is approximately 20 mSv, and the uncontrolled, ubiquitous use of this CT application in patients who do not need it should be prohibited.

Conclusion

The revolution in noninvasive imaging of the heart has taken a giant leap with the production of 64-slice, 4-cm-wide, thin-slice detectors. The resultant superior spatial resolution, combined with the improved temporal resolution made possible by faster gantry rotation speeds, makes CCTA feasible for almost every patient. The markedly improved sensitivity, specificity, and positive and negative predictive values seen with 16-slice CCTA compared with cardiac catheterization will improve still further. Sixty-four-slice CCTA will soon qualify as a potential replacement for nuclear stress tests as the first-line imaging test for potential CAD patients, at similar or reduced radiation dosages. CCTA should precede every diagnostic catheterization, based on the following: 91% of diagnostic cardiac catheterizations are elective, and of these 40% to 50% are normal, and CCTA has a negative predictive value > 99% in the most recent papers. Using CCTA in this manner would eliminate unnecessary negative diagnostic caths with their high cost and associated morbidity and mortality.

The triple rule-out examination holds great promise for revolutionizing the imaging of patients with acute chest pain. Hospitals are certain to embrace this technique as a noninvasive exam that improves patient care, decreases ER costs, and reduces the medicolegal risk of missing a potentially lethal diagnosis. The use of this exam must be moderated by its requirement for a higher iodinated contrast volume and its effective radiation dose.

References

1. Nieman K, Cademartiri F, Lemos P, et al. Reliable noninvasive coronary angiography with fast submillimeter multislice spiral computed tomography. Circulation 2002;106: 2051-2054.

2. Ropers D, Baum U, Pohle K, et al. Detection of coronary artery stenosis with thin slice multidetector row spiral CT with multiplanar reconstruction. Circulation 2003;107:664-666.

3. Kuettner A, et al. Diagnostic accuracy of noninvasive coronary imaging using 16 detector slice spiral computed tomography with 188ms temporal resolution. JACC 2005; 45:123-127.

4. Mollet N, Cademartiri F, Kreslin G, et al. Improved diagnostic accuracy with 16-row multislice CT coronary angiography. JACC 2005;45:128-132.

5. Shaw LJ, Raggi P, Schisterman E, et al. Prognostic value of cardiac risk factors and coronary artery calcium screening for all cause mortality. Radiology 2003:3:826-833.

6. Leber AW, Knez A, Becker A, et al. Accuracy of multidetector spiral CT in identifying and differentiating the composition of coronary atherosclerotic plaques: a comparative study with coronary ultrasound. JACC 2004;43(7):1241-1247.

7. Achenbach S, Moselewski F, Ropers D, et al. Detection of calcified and noncalcified coronary atherosclerotic plaque by contrast enhanced, submillimeter spiral computed tomography: a segment-based comparison with intravascular ultrasound. Circulation 2004;109(1):14-17.

8. Kyodo General Hospital. Coronary CT angiography compared with coronary catheterization for detecting in-stent restenosis. Circulation 2004;110(17):suppl III-563, abstract 2622.

9. Perisnakis K, Theocharopoulos N, Karkavitasis N, Damilakin T. Patient effective radiation dose and associated risk from transmission scans using Gd-153 line sources in cardiac SPECT studies. Health Phys 2002:83(1):66-74.

Dr. Dowe is chief operating officer and medical director of Atlantic Medical Imaging in Galloway, NJ. He has received grants/research support from GE Healthcare and Medrad and is a member of the speakers' bureau for both companies.

EFFECTIVE DOSE OF SELECTED RADIOLOGIC EXAMINATIONS

Head CT : 1-2 mSv

Chest CT : 5-7 mSv

Abdomen and pelvis CT : 8-11 mSv

Diagnostic coronary angiogram : 3-30 mSv

SPECT thallium : 25 mSv

SPECT sestamibi : 12 mSv

Coronary CTA : 6-13 mSv