Colorectal cancer is the second leading cause of cancer-related deaths in the United States. Screening tests for colorectal cancer include testing for fecal occult blood, sigmoidoscopy, colonoscopy, barium enema, and digital rectal exam. Colonoscopy beginning at age 50 years and performed every 5-10 years accompanied by twice annual testing for fecal occult blood is probably the most effective way of detecting early colon caner.

Unfortunately, patient compliance with conventional colonoscopy or sigmoidoscopy screening and fecal blood testing remains low. Reasons for resistance to colonoscopy include the required use of sedation and the small risk of perforation of the large intestine. However, if all persons availed themselves of colonoscopy screening at appropriate times there would not be enough trained endoscopists to accommodate them. Thus, better means of detecting colon cancer and increasing compliance are needed. A recent study suggests that increasing screening to 70% of the target average risk population would decrease colorectal cancer deaths by 50% by the year 2020.¹

In an attempt to improve patient compliance, researchers have evaluated CTC as an alternative approach to screening for colon cancer. Persons undergoing CTC receive the same bowel prep as persons undergoing conventional colonoscopy. Accuracy of CTC depends on the thoroughness of the bowel preparation, the sophistication of the software, and skill of the observer in interpreting the study. Many studies have suggested that CTC technique is as accurate as conventional colonoscopy in detecting polyps 10 mm in size or greater but not for smaller polyps. However, both studies can miss significant pathology.

At the present time, CTC is not covered by Medicare in 49 states and is not covered by most private insurers. However, CTC is reported to be used as a primary colorectal cancer screening modality in some military hospitals and in some institutions that are studying this technique. One of the major limitations of CTC is that detected polyps cannot be removed immediately as is the case with conventional colonoscopy. This disadvantage can be offset if more patients accept CTC and comply with appropriate referral for conventional colonoscopy for polyp removal. Importantly, large numbers of individuals can be spared a screening by conventional colonoscopy.

Comparative Accuracy of CTC Versus Convention Colonoscopy

We have reviewed and summarized the results of 12 studies comparing CTC to conventional colonoscopy in Cancer News on the Cancerconsultants.com Web site since 2002 and these results are summarized in this section.

In the August 2002 issue of the journal Radiology, researchers from New York University Medical Center compared CTC with conventional colonoscopy.² In this study, 105 patients underwent CTC immediately before a conventional colonoscopy. Conventional colonoscopy detected 132 polyps in 59 patients. CTC had no false positive results in the 46 patients who had no detectable polyps on conventional colonoscopy. CTC detected 93% of polyps larger than 10 mm but only 12% of those smaller than 5 mm and 70% of those between 6 and 9 mm. These authors concluded that CTC was very accurate for detecting clinically significant polyps.

Researchers in Switzerland were able to confirm that they could reliably detect colonic lesions that were at least 10 mm in size using CTC. Their results were published in the August 2002 issue of the medical journal Gut.³ They compared CTC performed with a multi-detector array CT scan (MDCT) with conventional colonoscopy in 50 patients. MDCT allows for a more rapid procedure with less interference by motion than conventional CTC. This was a blinded study where patients were examined first by MDCT and then by colonoscopy by individuals unaware of the first results. Nine of 11 lesions that were 10 mm or greater were detected by MDCT but only 5 of 15 between 6 and 9 mm and 1 of 42 that were less than 5 mm were detected. They found that 19 of 21 patients who had no lesions by conventional colonoscopy had no lesions detected by MDCT colonoscopy. They concluded that MDCT colonography provides good data quality and has good sensitivity and specificity for the detection of colonic lesions of 10 mm or more.

Researchers from Wake Forrest conducted a trial to determine the ability of CTC to identify patients with colorectal lesions who need conventional colonoscopy. The details of this study were published in 2003 in Gastroenterology.⁴ They performed virtual colonoscopy on 205 patients prior to conventional colonoscopy. The overall results showed that CTC had a sensitivity of 90% and a specificity of 94.4% for lesions of 10 mm or greater but was less accurate for smaller lesions. They also reported that the negative predictive value of CTC was 95% for a 6-mm cutoff size and 98.9% for a 10-mm cutoff. They suggest that CTC colonoscopy precluded the need for conventional colonoscopy in 86% of patients with a 1% false-negative rate. The high negative predictive value may help reduce the number of negative screening colonoscopies required.

A second study published in Gastroenterology was a retrospective trial involving 341 patients who had a CTC performed prior to conventional colonoscopy in eight different medical centers.⁵ The purpose of this study was to compare different software and variability of interpretation by 18 different radiologists. For lesions 10 mm or greater, the average sensitivity and specificity were 75% and 73%, respectively. There observed a trend between performance and the experience of the observer. There were no apparent differences between the different software products. The researchers concluded that CTC performance compared favorably with reported performance of fecal occult blood testing, flexible sigmoidoscopy, and barium enema.

Researchers conducting a multicenter trial have concluded that CTC compares favorably with conventional colonoscopy as a screening test for colon cancer. The results of this study appeared in the December 4, 2003, issue of the New England Journal of Medicine.6 An accompanying editorial suggests that CTC may be the best screening test for colon cancer.⁷ In this multicenter study, 1,233 asymptomatic individuals underwent CTC followed by conventional colonoscopy. These researchers found that CTC could not be performed in 12 cases; 6 due to an inadequate preparation. The sensitivity for detecting polyps at least 10 mm in diameter was 93.8% for CTC and 87.5% for conventional colonoscopy. For lesions at least 8 mm in diameter, the sensitivity of CTC was 93.8% compared to 91.5% for conventional colonoscopy. For polyps 6 mm or less, the sensitivity of CTC was 88.7% compared to 92.3% for conventional colonoscopy. These researchers also reported that two polyps were malignant; both were detected on CTC, and one of them was missed on conventional colonoscopy before the results on CTC were revealed.

Researchers from NYU Medical Center have published a report documenting the detection of polyps 10 mm or greater in average risk individuals by CTC. The results of this study were published in the March 2004 issue of Radiology.⁸ In this study, 68 asymptomatic men over the age of 50 underwent CTC followed by conventional colonoscopy. They detected 98 polyps in 39 patients. Conventional colonoscopy results were normal in 29 of the 68 patients; CTC correctly identified as normal 26 of these 29 patients. However, CTC detected one polyp greater than 10 mm which was not detected by conventional colonoscopy. CTC detected all 3 polyps that were greater than 10 mm. CTC in this study, as in others, detected only 11.5% of polyps 1-5 mm, and 52.9% of those 6-9 mm. The interpretation time for CTC was 9 minutes.

Researchers from nine U.S. Medical Centers have concluded that “computed tomographic colonography by these methods is not yet ready for widespread clinical application. Techniques and training need to be improved.” The results leading to these conclusions were published in the April 14, 2004, issue of the Journal of the American Medical Association.⁹ This study compared CTC to conventional colonoscopy in 615 persons 50 years of age or older. Centers involved in this study were required to have performed only 10 CTCs prior to this study. They detected 827 lesions of which 104 were 6 mm or greater in size. CTC detected 39% of the 6 mm or greater lesions and 55% of those were at least 10 mm in size. However, CTC missed 2 of 8 cancers. The researchers also stated that “the accuracy of CTC varied considerably between centers and did not improve as the study progressed. Participants expressed no clear preference for either technique.” However, only one of the centers had substantial previous experience and this center detected 82% of patients with at least one of the smaller polyps compared to a 24% detection rate for the 8 other centers. This study is of interest primarily because 2 of 8 cancers in this study were missed by CTC and that the patients did not have a preference for CTC versus conventional colonoscopy. The center effect was very important in this study indicating that single high volume centers detect more lesions.

Researchers from the Netherlands have reported that CTC is as effective as conventional colonoscopy in detecting polyps 10 mm in size or greater in individuals at high risk for colorectal cancer.  The details of this report appeared in the July 2004 issue of Gastroenterology.¹⁰ A total of 249 patients at high-risk of colon cancer were evaluated by CTC followed by conventional colonoscopy and a second look colonoscopy when indicated.  Patients in this study had a history of colorectal polyps or were at increased risk due to family history.  They underwent CTC followed by conventional colonoscopy.  A second conventional colonoscopy was performed for discrepant findings. Conventional colonoscopy detected 141 polyps.  The sensitivity for CTC for detecting at least one large polyp was 84%. For patients with a polyp greater or equal to 6 mm the sensitivity was 78%.  For all polyps the sensitivity was 62%.  The specificity for detecting large polyps was 92% with a lower specificity for smaller polyps. These researchers concluded that 82% of patients would not have to undergo conventional colonoscopy to detect large polyps.  CTC was not effective in 16% of 31 patients with large polyps. However, in 19% of patients with large polyps detected by CTC, the polyps were only detected by a second conventional colonoscopy. These authors also pointed out that CTC is more likely to miss large flat lesions which were more likely to be malignant. These authors concluded that “our data indicate that CTC and conventional colonoscopy have a similar ability to identify large colorectal polyps or cancer.  However, the majority of flat lesion were overlooked, a finding that requires further study.”

A multi-center study concluded that conventional colonoscopy is more sensitive than CTC or air contrast barium enema in detecting polyps and cancers. This report appeared in the January 22-28, 2004, issue of The Lancet.¹¹ This study compared the findings of conventional colonoscopy, 2D CTC and air contrast barium enema in 614 patients who had occult fecal blood, hematochezia, iron deficiency anemia, or a family history of colon cancer. There were 9 cancers detected in this study and it was reported that barium enema detected 8, CTC detected 7 and conventional colonoscopy detected all 9. For lesions 10 mm or larger, barium enema detected 48%; CTC detected 59%; and conventional colonoscopy detected 98%. Smaller lesions were less well detected by barium enema and CTC than by conventional colonoscopy. There were 4 cancers detected outside the colon by CTC. There were 4 perforations from conventional colonoscopy, for a complication rate of 0.1%. These authors concluded that conventional colonoscopy was more sensitive than the other 2 tests for the detection of polyps and cancers.

In an accompanying editorial entitled “Unbiased studies are needed before virtual colonoscopy can be dismissed,” Drs. Halligan and Atkin point out the discrepancies between various studies of CTC.¹² They point out that CTC is an evolving technology with controversy over the superiority of 2D versus 3D techniques. They also point out that skill and experience is necessary for radiologists to correctly interpret CTCs. They suggest training and accreditation for this technique, as well as an evaluation of computer-aided detection. They also suggest that CTC may be most useful for use in symptomatic patients who may have colonic lesions or lesions outside the colon. The fact that 4 cancers were detected outside the colon is of major interest.

Researchers from the National Institutes of Health have reported that CTC coupled with computer-aided polyp detection (CAD) detects most polyps that are eight millimeters in size or larger. These results were comparable to results achieved with conventional colonoscopy. The details of this study were published in the December 2005 issue of Gastroenterology.¹³ These researchers evaluated 1186 patients undergoing colorectal cancer screening at three medical centers. CAD involves the use of a computer program to evaluate the CT image of the large intestine. CAD detected 83% of polyps that were 10 millimeters or larger, 81% of polyps that were eight millimeters or larger, and 61% of polyps that were six millimeters or larger. The false-positive rate was two false polyps per patient for polyps that were 10 millimeters or larger, seven false polyps per patient for polyps that were eight millimeters or larger, and eight false polyps per patient for polyps that were six millimeters or larger. These researchers concluded that the accuracy of CAD and CTC for colorectal cancer screening may prove to be acceptable. Polyps that were missed tended to be under fluid, small or flat, or on a fold.

What is the Comparable Tolerability of CTC to Other Screening Techniques?

A multi-center study has concluded that patients prefer conventional colonoscopy to CTC or barium enema (BE) as screening tests for colon cancer.¹⁴ It was suggested that patients prefer conventional colonoscopy over CTC and BE because they received conscious sedation during CC but not during CTC or BE. The details of this study appeared in the September 2006 issue of the American Journal of Medicine. This study reported patient experience with each of the techniques for detecting colon cancer. The test patients were most willing to repeat was conventional colonoscopy. The most uncomfortable procedure was the air contrast BE. Patients reported feeling more “more worn out” from conventional colonoscopy but also reported less pain and discomfort. Patients reported less worry and anxiety after CTC than after the other two procedures.

These appear to be very reliable observations and of importance to physicians and patients in the selection of colon cancer screening techniques. However, it must be kept in mind that conventional colonoscopy is usually done with morphine or Demerol and Versed, so naturally most patients have no recollection or poor recollection of the procedure.  Conventional colonoscopy requires a morning or afternoon in an out-patient surgery center and requires family assistance to get home as driving is prohibited.   There were no data presented in this paper on total lost time with each procedure but it must be greater with conventional colonoscopy than with CTC. Conventional colonoscopy is also the most risky because of the small but definite risk of perforation and drug reactions to sedative regimens. There is also evidence that the toleration of BE and CTC could be increased if operators used CO2 with a pressure regulator rather than air without a regulator. The administration of anti-spasmotics may also decrease the unpleasantness of air contrast BE and CTC. Both CTC and conventional colonoscopy are highly dependent on the skill and experience of the person performing the procedure. Currently, the most important thing to accomplish is to increase the number of persons undergoing any kind for screening for colon cancer on a regular basis beginning at age 50 years. 

What Is the Official Position on CTC?

The American College of Gastroenterology (ACG) Action Plan on Colorectal Cancer Prevention has determined that CTC is not ready for approval as an “effective and cost-effective strategy for colorectal cancer prevention.”¹⁵

The ACG has reviewed the results of 49 publications concerning the accuracy of CTC in detecting polyps and colon cancers. They have concluded that there is great variability between studies in the ability of detect polyps. Variability is attributed to differences in software and training in interpreting the studies. These studies also suggest that flat polyps are not detected by CTC colonoscopy. They also pointed out seven areas where there is uncertainty among experts and these are:

• Threshold of polyp size for referral for optical colonoscopy
• Appropriate screening interval
• Radiation risk of malignancy from repeat screening
• Effectiveness of low-dose virtual colonoscopy protocols
• Interpretation and management of extra-colonic findings
• Impact on adherence with and without laxative bowel preparation
• Cost-effectiveness of virtual colonoscopy at the intervals selected for clinical practice

These authors concluded that the “ACG will continue to follow developments in CTC closely and update its members on these developments.”

What Is the Status of Insurance Coverage for CTC?

Over 150 insurance companies cover the cost of non-invasive colon cancer screening, but the companies are highly selective, depending on an individual’s coverage limits.  It is also  unclear as to whether more patients would be agreeable to CTC if insurance covered the cost.  What is clear is that thousands of normal healthy people pay for CTC screening out of their own pockets.  Costs for consumer-based CTC screening (walk-in screening facilities) range from $600 to $1600.  In many cases this cost is less than the co-pay for a full in-hospital (or out-patient surgical facility) colonoscopy.

Researchers from the University of Wisconsin have recently reported the results of CTC screening of 1110 persons over a one year period.¹⁶ They stated that more than 99% of these procedures were covered by managed care agreements.

What is the role of Computer-Assisted Diagnosis (CAD) in CTC?

CAD is becoming increasing more popular for several reasons.  Firstly, for all of the FDA-approved CAD technologies it is more accurate than human detection; secondly it is a time-saver for the radiologist since it highlights areas for visual review by the radiologist, or it can be used by a technician to preview a study for the radiologist to review.  CAD for CTC is already in widespread use and it will likely become used more frequently as more radiologists become familiar with its advantages.

What Is the Appropriate Machine and Software for CTC?

Table 1 shows the radiation doses for CTC using different CT scanning technology.

Table 1: Radiation exposure during virtual colonoscopy*

Putting these radiation doses in perspective, Table 2 compares background radiation at one extreme with very high dose diagnostic studies.  Radiation dose is a very real consideration for any healthy person who may be getting repeated CTC screening. Normally colon visualization is required no more often than every 5-10 years (if the first screen is normal) and every 3-5 years (if there are small polyps under 10 mm which do not need to be removed).

Table 2: Radiation exposures for common events or procedures

* The U.S. National Academy of Sciences - Biological Effects of Ionizing Radiation committee has determined that a lifetime exposure to >100 millisieverts (>10,000 mRem) may cause cancer in 1 out of 100 people.

** http://www.jsonline.com/alive/news/jan06/384843.asp

Most computer software for visualizing the colon after CTC is standardized, and is available from many CT and MRI software vendors.  The rendering of a 3-D volume and reconstruction of the colon to resemble a real colonoscopy is more accurate then reading the individual slices in CTC, so most radiologists are using this method, often referred to as “virtual colonoscopy.”.

What Should Be Looked for When Choosing a Center to Perform CTC?

For individual patients interested in CTC the following should be considered when choosing a diagnostic center for self-referral.  These should also be discussed with the referring physician if the CTC is ordered by a doctor:

The diagnostic center should use the lowest possible radiation dose for the study.  This means using a low-dose CT scanner such as an Electron Beam scanner, or making sure that the CT scanning facility uses the lowest possible dose for the study if using a conventional multi-slice CT scanner.  This is especially important in people under the age of 55 years old where cumulative radiation dose can be an issue.

The diagnostic center should have experience performing CTC. Any center which as done more than 1000 CTC studies for colon cancer screening will usually have well-trained radiologists and technicians that use methods that are most comfortable for the patient.

The cost of CTC varies widely as does insurance coverage, so this may be an important consideration, particularly if the patient will be required to self-pay or have a significant co-pay obligation.

Conclusions

CTC at the present time is an acceptable procedure for screening for colon cancer in persons not wanting to undergo conventional colonoscopy. This technique could markedly improve the number of average risk patients undergoing meaningful screening for colorectal cancer. CTC is an evolving technology which will undoubtedly improve in accuracy due to improvements in software and experience in interpreting results. There are already discussions about whether or not CTC will replace conventional colonoscopy in the near future.¹⁷ One improvement that is being currently explored is CTC without colonic cleansing, which would take away the most important complaint about CTC and conventional colonoscopy.¹⁸