Clinical trials begin, but conclusions remain years away, while public demand heats up

Screening CT moves into controversial areas

By: Karen M. Horton, M.D., And Elliot K. Fishman, M.D.

Advancements in scanner and computer technology have led to major improvements in CT imaging systems since their introduction into clinical practice more than 20 years ago. The unprecedented speed possible on multidetector-row imaging systems, combined with their higher resolution and sophisticated image processing tools such as real-time 3D volume rendering, has resulted in unparalleled CT capabilities.

The clinical role of CT continues to expand along with the changes in technology. It is now possible to scan the entire body from neck to pelvis by 1-mm slices in 30 seconds, and newer 16-slice scanners will allow the entire body to be imaged with submillimeter slices in 10 seconds. These technical improvements will soon lead to new clinical applications including CT angiography and colonography. Examinations that were traditionally performed with other modalities, such as pulmonary angiograms, intravenous pyelography, and angiography for pancreatic cancer staging, are now routinely performed with CT.

CT has become so successful in detecting disease in symptomatic patients that it is being investigated as a screening tool to detect disease in the asymptomatic population. CT screening centers are beginning to open across the U.S., and the initial public response has been favorable, although controversy surrounds the use of CT as a screening modality in asymptomatic people.

CARDIAC SCANS

Based on data from electron-beam CT, it is now possible to detect and quantify calcified atherosclerotic plaque in the coronary arteries (Figure 1). Coronary artery calcification is a marker for atherosclerosis, and the coronary artery calcification score established by Agatston has been shown to be an accurate measure of coronary plaque burden and a potent predictor of future cardiac events.^1-4 The calcium score can then be used to identify patients who are at risk for future cardiac events so that appropriate interventions can be made.^2,5 Similarly, a negative coronary calcium score can rule out significant coronary artery disease in symptomatic patients with atypical chest pain.⁶ The score obtained with EBCT correlates with the score obtained with gated MDCT studies, and therefore coronary artery calcium scoring can be done at CT screening centers with either EBCT or MDCT scanners.^7,8

The technique for CT coronary artery calcium examinations depends on the type of scanner available. On EBCT scanners, images are obtained using a single breath-hold and electrocardiogram triggering. Patients are scanned in the supine position from the level of the pulmonary arteries through the base of the heart, using a 350-mm field-of-view with a 512 x 512 reconstruction matrix, as 3-mm slices are obtained every 3 mm with an image acquisition time of 100 msec.

With an eight-detector-row MDCT, prospective ECG gating is used in a sequential mode to obtain 2.5-mm slices through the heart at 2.5-mm intervals. We use 140 kV and 50 mAs. This corresponds to an image acquisition time of 360 msec. With a new 16-detector-row MDCT scanner, 3-mm slice thickness is reconstructed from data acquired using a 1.5-mm collimation. Slice reconstruction with overlap (0.5 mm) has been shown to improve reproducibility of the calcium quantity measurement.9 Because of the higher absorption, coronary calcium can be more easily detected with 80 kV. Obtaining 3-mm slices at 300 mAs allows for images with a high signal-to-noise ratio that enable reliable detection of coronary calcium. The radiation exposure for this examination is in the range of 1 mSv. The faster MDCT scanners available also allow high-resolution CT angiography of the coronary arteries.¹⁰

LUNG CANCER SCREENING

Lung cancer is the number one killer among cancers. It kills more people than colon, breast, and prostate cancer combined. Despite improvements in detection and treatment, the five-year survival rate for lung cancer has made only minimal improvement over the last 25 years. In the last two years, public demand for lung cancer screening has increased. The theory that if lung cancers are found at an earlier stage, before the patient is symptomatic, survival should be increased has not yet been proven, however. Opponents to lung cancer screening argue that finding cancers early may not result in improvement in long-term survival, since even small cancers may have already spread by the time they are visible on CT.

Large multicenter studies are under way to evaluate the effectiveness of screening high-risk populations such as smokers. It will be some time before the results of these studies are available, since patients must be followed for many years to determine if CT screening will improve long-term survival. Some short-term data available for the use of CT to screen high-risk smokers for lung cancer are encouraging. In a study by Henschke et al^11,12 of 1000 asymptomatic smokers, all patients underwent low-dose CT of the chest: 27 lung cancers were detected on CT, only seven of which were visible on chest x-ray. Of the 27 cancers detected, 23 were at stage I, indicating that the five-year survival rate should be 70%.¹²

Another controversial aspect of lung cancer screening is the detection of many benign lesions (Figures 2 and 3). In the same study by Henschke et al, for instance, 23 % of patients had nodules that required follow-up.¹² Most of these small detected nodules will be benign but will require serial follow-up scans to ensure stability. This process adds significant cost to the healthcare system as well as anxiety for the patients.

Because these studies are ongoing, the optimal CT protocol has not yet been determined. It is clear, however, that scans should be performed with thin collimation of 3 mm or less, single breath-hold to limit respiratory motion, and decreased mAs to decreased radiation exposure. We use MDCT for lung cancer screening with the 2.5-mm collimator setting to create 3-mm slices. The kV is 140 and mAs is 80, in an effort to decrease radiation dose.

A promising technology under development may enhance lung cancer screening with CT.^13,14 Computer-aided diagnosis systems from R2 Technology use sophisticated software programs to aid in lung nodule detection. These systems can detect nodules and give accurate diameter and volume measurements, and they may come to play a role at CT screening centers. This technology is not yet available commercially, but our experience has found it to be very promising.

WHOLE-BODY SCANS

Whole-body CT screening is probably the most controversial of the proposed CT screening examinations. To date, no studies have been published evaluating the efficacy of screening healthy, asymptomatic individuals with CT. Given the utility of CT in detecting disease in symptomatic patients and recent CT advances, however, it is likely that CT can function as a screening study to detect unsuspected disease at an early stage. Heavy publicity about whole-body CT scanning over the last year has created a public demand.

At most screening centers, whole-body CT is performed using 5 to 8-mm slices and no IV contrast. As every radiologist knows, however, a noncontrast CT of the abdomen and pelvis is limited in the information it provides. Small lesions of the liver, kidneys, and pancreas can be missed without contrast (Figure 4). Studies without IV contrast often detect benign liver and kidney lesions such as cysts and hemangiomas but cannot characterize them adequately, necessitating further imaging studies with IV contrast. In addition, it is impossible without IV contrast to assess organ function.

Our approach to the screening CT study is that it must be optimized to detect any disease present. We recognize from doing CT on a daily basis that the use of iodinated contrast agent is mandatory for detection and characterization of a wide range of abnormalities from simple cysts of the kidney to hypervascular renal cell carcinomas. Although the use of IV contrast does introduce a small risk to the patient of allergy, reduced kidney function, and so on, agents have become quite safe, thanks to the introduction of nonionic contrast. We believe that the benefit of IV contrast may outweigh the risks and that patients should be given the option of receiving it for their examination, provided there are no contraindications. We consider it wrong to do only noncontrast CT on screening patients just because it is easier and faster and requires no physician in attendance. Therefore, it is our policy to offer each patient the option of IV contrast and obtain written consent. If a patient prefers not to receive IV contrast, a noncontrast exam will be performed, for the same cost. The charge to the patient will be same whether or not IV contrast is given.

We exclusively use nonionic contrast agents such as Omnipaque-350 or Visapaque-320 from Amersham Health. Isosmolar agents like Visapaque, which have fewer minor reactions, should be ideal in this study population. When performing a full-body CT scan, we first perform a noncontrast cardiac CT for detection of coronary artery calcium scoring. Next, we inject 120 cc of Visapaque-320 at a rate of 3 cc/second. Scanning begins 30 seconds after the start of the injection and includes the neck, chest, abdomen, and pelvis (Figure 5). The 4 x 2.5-mm collimator setting is used to create 3-mm slices, and 3D imaging is performed in all patients.

VIRTUAL COLONOSCOPY

CT colonography is an exciting potential screening study for colon cancer. It offers the advantages of a noninvasive study with a low-risk of complication compared with conventional colonoscopy as well as a lower cost. Its main advantage may be an increase in the number of patients who undergo screening. There is still considerable variably in how the studies are performed and interpreted, but continued improvements in scanner technology and computer software should eliminate these problems. Additional investigation and large-scale multicenter trials with comparison to conventional colonoscopy are necessary to determine definitively the sensitivity and specificity for this technique.

The technique uses data acquired from a spiral CT scan combined with sophisticated computer software to generate both 2D and 3D views of the colon. Besides being less expensive and less invasive than conventional colonoscopy, CT colonoscopy offers a lower morbidity and mortality. The sensitivity and specificity of CT colonoscopy have improved as scanner and computer technology has advanced. Most trials have been performed with single-detector spiral scanners using a collimation of 5 mm or less. These studies use conventional colonoscopy as a gold standard.

One of the largest studies of 300 patients by Yee et al used 3-mm collimation.¹⁵ In that study, CT colonoscopy had a 100% sensitivity for detection of cancers (8/8) and an overall per patient sensitivity and specificity of 90.1% (164/182) and 72% (85/118), respectively, for polyp detection. The per-patient sensitivity for polyps 1 cm or greater was 100% (49/49); for polyps between 5 to 9.9 mm, it was 93% (50/54); and for polyps less than 5 mm, it was 82% (65/79).15 This study also included a significant number of asymptomatic patients.

One large trial has evaluated single-detector CT and multidetector-row CT in 237 high-risk patients.16 The authors noted that the advantages of multidetector CT included a shorter breath-hold with less respiratory and motion artifact and improved demonstration of colonic distention.¹⁶ The thinner collimation available with MDCT allows nearly isotropic images, with high-resolution results in all imaging planes as well as improved endoluminal views.

Multiple techniques exist for performing CT colonoscopy. It is essential that the colon be completely clean to avoid missing lesions obscured by retained stool and also to avoid mistaking adherence stool for polyps. Several colon cleansing routines are available. We currently use the Fleet prep kit #1, which includes 45 cc of phospho-soda as well as four bisacodyl tablets and one suppository. This is considered to be a saline cathartic and is better tolerated by patients and results in less retained fluid. It should be avoided in patients with renal failure or congestive heart failure, as it causes electrolyte problems. Several authors have investigated the use of stool tagging to decrease the number of false positives due to retained stool, but we do not use these agents.¹⁷

When the patient arrives on the day of the study, the colon must be completely inflated with either air or carbon dioxide. An underdistended colon will mask polyps or lead to false-positive diagnoses. Colonic distention is achieved by retrograde insufflation of the colon using a small rectal tube. We inflate the colon using room air. At least 1.5 to 2 L of air is usually necessary to completely distend the colon. Carbon dioxide is an alternative that has the advantage of being absorbed very quickly and results in less patient discomfort. We have generally found electronic pumps to assist colonic inflation to be less than ideal.

The CT colonoscopy protocol requires scanning the patient in both supine and prone positions.18 This allows for improved distention and better visualization of all segments of the colon and is especially important with retained fluid, as polyps can be obscured in the dependent pool of fluid. After inflation of the colon, the patient is scanned initially in a supine position from the level of the diaphragm through the symphysis pubis. We routinely use MDCT for CT colonoscopy at the 4 x 2.5-mm collimator setting to create 3-mm slices that can be reconstructed at 2-mm intervals. The mAs can be decreased to minimize radiation dose, and the entire acquisition is performed in one breath-hold. After the supine images are performed, the patient is placed in a prone position and, if necessary, additional air is added. The scan is then repeated.

Many software packages are available to assist in the interpretation. Initially, it is important to do a comprehensive review of the axial scans. With good distention, it is easy to review the studies, especially on a workstation with cine capabilities. Many of the polyps can be detected on the axial views. Next, simple multiplanar reconstruction can be performed to increase reader confidence and to visualize any abnormality in multiple planes. The use of MDCT scanners allows increased resolution on the multiplanar reconstructions. In addition to the axial scans and multiplanar reconstructions, the endoluminal views can be reviewed.

Most software packages allow automated travel through the colon (Figure 6). There is some controversy over whether the entire supine and prone data sets need to be reviewed using the endoluminal software. Dachman et al, for instance, evaluated 44 patients and found that the interpretation times were shorter if routine endoluminal views were not performed.¹⁹ In another study by Macari et al, interpretation times saw a significant decrease from 40 to 16 minutes when only selected endoluminal views were reviewed.²⁰ This did not decrease the sensitivity for polyp detection.

CT colonography has continued to make major advances since its introduction, but additional study is necessary to prove its performance as a purely screening study in asymptomatic subjects.

STUDY CONTINUES

The concept of using CT for screening healthy asymptomatic people is still in its infancy. Opponents stress that large multicenter trials are needed to prove its efficacy. Trials are under way for lung cancer and coronary artery screening, but it will be many years before the results are known. Another criticism involves the potential effects of radiation exposure on a healthy population. Although CT cannot be performed without the use of radiation, protocols are being developed to minimize the dose to patients, and new MDCT scanners will allow the radiation exposure to be individualized, based on the patient's body habitus.

It is likely that we will not know the long-term benefits of CT screening on the population for decades. Given the success of existing screening centers, however, it is clear that public demand exists. People are interested in taking control of their health, and they recognize the benefits of detecting disease even before symptoms develop. Used in coordination with the patient and the referring physician, CT screening may prove to be a valuable tool in preventive health care.

Dr. Fishman is a professor of radiology, and Dr. Horton is an assistant professor of radiology, both at Johns Hopkins University in Baltimore.

---

References

1. Agatston AS, Janowitz AS, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827-832.
2. Arad Y, Spadaro LA, Goodman K, et al. Prediction of coronary events with electron beam computed tomography. J Am Coll Cardiol 2000;36:1253-1260.
3. Baumgart D, Schmermund A, George G, et al. Comparison of electron beam computed tomography with intracoronary ultrasound and coronary angiography for the detection of coronary atherosclerosis. J Am Coll Cardiol 1997;30:57-64.
4. Haberl R, Becker A, Leber A, et al. Correlation of coronary calcification and angiographically documented stenoses in patients with suspected coronary artery disease: results on 1,764 patients. J Am Coll Cardiol 2001;37:451-457.
5. Raggi P, Callister TQ, Cooil B, et al. Identification of patients at increased risk of first unheralded acute myocardial infarction by electron-beam computed tomography. Circulation 2000;101:850-855.
6. Laudon DA, Vukov LF, Breen JF, et al. Use of electron-beam computed tomography in the evaluation of chest pain patients in the emergencey department. Ann Emerg Med 1999;33:15-21.
7. Becker CR, Jakobs TF, Aydemir S, et al. Helical and single slice conventional CT versus electron beam CT for the quantification of coronary artery calcification. AJR 2000;174:543-547.
8. Becker CR, Kleffel T, Crispin A, et al. Coronary artery calcium measurement: agreement of multi-row detector and electron beam CT. AJR 2001;176:1295-1298.
9. Kopp A, Ohnesorge B, Becker C, et al. Reproducibility and accuracy of coronary calcium measurements with multi-detector row versus electron-beam CT. Radiology online 2002;10.114B,
10. Vogl TJ, Abolmaali ND, Diebold T, et al. Techniques for the detection of coronary atherosclerosis: multi-detector row CT coronary angiography. Radiology 2002;223:212-220.
11. Henschke CI. Early lung cancer action project: overall design and findings on baseline screening. Cancer 2000;89:2474-2482.
12. Henschke CI, McCauley DI, Yankelevitz DF, et al. Early lung cancer action project: summary of the findings on baseline screening. Oncologist 2001;6:147-152.
13. Wormanns D, Fiebich M, Saidi M, et al. Automatic detection of pulmonary nodules at spiral CT: clinical application of a computer-aided diagnosis system. Eur Radiol 2002;12:1052-1057.
14. Reeves AP, Kostis WJ. Computer-aided diagnosis of small pulmonary nodules. Semin Ultrasound CT MR 2000;21:116-128.
15. Yee J, Akerkar GA, Hung RK, et al. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219:685-692.
16. Hara AK, Johnson CD, MacCarty RL, et al. CT colonography: single- versus multi-detector row imaging. Radiology 2001;219:461-465.
17. Pochaczevsky R. Stool tagging with barium for CT colonography. Radiology 2001;221:850-851.
18. Chen SC, Lu DS, Hecht JR, et al. CT colonography: value of scanning in both the supine and prone positions. AJR 1999;172:595-599.
19. Dachman AH, Kuniyoshi JK, Boyle CM, et al. CT colonography with three-dimensional problem solving for detection of colonic polyps. AJR 1998;171:989-995.
20. Macari MA, Lavelle M, Berman P, Megibow AJ. Comparison of time-efficient CT colonography with two- and three-dimensional colonic evaluation for detecting colorectal polyps. AJR 2000;174:1543-1549.