Virtual Colonoscopy

Virtual colonoscopy: Is it ready for prime time?
Virtual colonoscopy (VC) is a new method of screening the colon for precancerous polyps. Using a CAT or CT scanner and new computer methods of rendering or reconstructing the images, the colon can be evaluated without a colonoscope and without sedation.

What is it?
Virtual colonoscopy takes the information produced by a CT scanner and processes this information to produce an image of the colon's inner surface. The examination is possible because of new, very fast CT scanners and the refinement of computer hardware and software that have been used to produce modern digital movies, such as "Star Wars" or "The Matrix." When the colon is properly cleansed and distended with room air or carbon dioxide, and when the CT information is processed, we can then look at the inner lining or surface to detect polyps.

How does it work?
Modern CT scanners produce their images as the patient lies on a "bed" that is rapidly pulled through a gantry, or short tunnel. Within the gantry there is on one side a rotating x-ray tube and on the other a set of detectors that receive the transmitted x-rays after they have passed through the body. A continuous volume of information is obtained from this exam. This information can be sliced and diced in infinite ways. The information from these exams can be processed in virtual reality computers, much like digital movies are made. Because the inside of the colon has been distended, we can look inside the colon and have a view almost completely the same as the colonoscopists.

Why do we need another test to detect precancerous colon polyps?
It is estimated that a million people in the United States alone have not been properly screened for colorectal polyps. The best method for screening is a regular colonoscopy. There are simply not enough endoscopists to screen such a large number of patients. Other methods such, as sigmoidoscopy, evaluate only a part of the colon. This would be equivalent to performing a mammogram on only one breast. Tests to detect small amounts of blood in the stool can miss polyps and even cancers. The barium enema is not nearly as sensitive as a regular colonoscopy. If a noninvasive test could be developed that could select those patients with a polyp, then fewer and more focused colonoscopies could be performed. The combination of newer, faster CT scanners with computer software that would produce images equivalent to an endoscope seems to be a logical step in the development of such a test.

How is it done?
As with standard colonoscopy, the patient must undergo a preparation that includes a liquid diet and a cathartic that cleanses the colon. This starts the day before the procedure. In addition, the patient has to take small barium tablets during the day, which tags any stool that may remain in the colon. This aids our ability to differentiate stool from a small polyp.

• On the morning of the procedure, the patient arrives in the radiology department where he or she then changes into an examination gown.
• A small tube is placed in the rectum and gas is slowly pumped into the colon manually or using an automated insufflator, which maintains a constant pressure. With manual insufflation, room air is used. With the automated insufflator, CO2 (carbon dioxide) is used to inflate the colon. CO2 is thought by many to be more comfortable.
• Once the appropriate amount of distension is achieved, the patient is scanned with the Computer Tomography (CT) machine. The scan time takes approximately 15 seconds, during which the patient holds his breath. The patient is scanned both on his stomach as well as on his back. We do this to allow fluid and stool to fall away from the dependent portion of the colon.
• After the exam, the patient goes to the bathroom to expel the gas and then gets dressed.
• The patient waits for approximately 45 minutes to 1 hour in order to determine whether a polyp has been identified. We need this time to carefully examine the colon using our computer renderings. If a polyp is identified, the patient then goes to the colonoscopy suite where a formal colonoscopy is performed to remove the polyp. If no polyp is found, the patient goes home.
• No sedation is given for the virtual colonoscopy. However, the patient must still have a chaperone come with him in case a formal colonoscope is performed (sedation will be given for this).
• The patient may resume normal activity after the procedure.
• A repeat VC is recommended every 3 years.

Our philosophy has been to develop a virtual colonoscopy program that positions this technology as one option in a wider program of colon cancer screening. Because virtual colonoscopy cannot characterize a polyp when identified, it is necessary to then perform a standard colonoscopy. Because a bowel prep is necessary for both, it seems ideal that any patient with an identifiable polyp on a VC should then have a standard colonoscopy the same day. Thus, we are planning a comprehensive program of same-day virtual colonoscopy followed by same-day standard colonoscopy if a polyp is identified.

Myths about virtual colonoscopy
Virtual colonoscopy requires a bowel prep just like a regular colonoscopy. Many centers currently are investigating the accuracy of the test without a prep or with a modified prep, with stool tagging. However, that data is so far not positive. Virtual colonoscopy is uncomfortable as the colon is distended. Published data shows that more patients prefer regular colonoscopy to VC, as they are sedated. However, most patients state that the discomfort from VC is well tolerated.

While the accuracy of VC is equivalent to regular colonoscopy in polyps greater than 7 to 8 mm, VC is not as good as regular colonoscopy for smaller polyps. Fortunately, these smaller polyps are either noncancerous or benign. VC cannot differentiate a non-precancerous polyp from a precancerous polyp. Neither can regular colonoscopy. Only a biopsy can make this differentiation. Biopsy can only be performed via a regular colonoscopy. If an abnormality is detected on a VC, the patient will need a regular colonoscopy to determine the nature of that polyp.

Pros: A VC is relatively non-invasive and does not require sedation. It provides the community with one more way to screen for a preventable disease, colon cancer. The unique and "high-tech" nature of the study may be more acceptable to some people. If it is normal, a regular colonoscopy is not necessary. If normal, the patient can resume normal activity that day. A VC may detect abnormalities outside the colon, such as a kidney cancer, an abdominal aortic aneurysm, or enlarged lymph nodes. Studies would suggest that a significant abnormality will be detected in 5% of patients scanned.

Cons: VC is not as sensitive as standard colonoscopy for polyps less than 7 mm. Some of these polyps may be precancerous. Thus, we do not know when a patient should have a repeat VC. Currently, it is suggested that a repeat VC be performed every three years. A negative regular colonoscopy requires only a repeat in 10 years.

In addition, here is a very steep learning curve for radiologists to master interpretation of VC. Most consider this to take at least 75 cases with regular colonoscopy confirmation. We can easily miss flat adenomas. Fortunately, these are uncommon to rare.

Lastly, we may detect a potential abnormality outside the colon that requires further evaluation, often with more imaging. If it turns out that this is not significant, time and resources will be consumed unnecessarily.

How good is virtual colonoscopy?
The following is a summary of several reports:

HR= high risk patient population; S= screening population

It should be noted that much of this experience used CT scanners that were slower and produced images that were much thicker than can be produced now. For instance, 3 years ago, we had a scanner that took 25 to 30 seconds to obtain a dataset that produced 2 mm slices. Now we have scanners that take 15 seconds to obtain a dataset with 1 mm slices. Further, there have been significant advances in computer hardware and software, allowing much faster processing of the CT scan data. Lastly, there have been significant changes in how patients are prepared before the CT.

The most recent paper published in 2003 in the New England Journal of Medicine is more promising. The following is a summary of that data:

VC = virtual colonoscopy; OC = ocular or regular colonoscopy

This work was the largest series to date and involved multiple radiologists and colonoscopists. More than 1,200 people were studied, most of whom were in the low-risk screening category. The radiologists used the virtual portion of the exam as the primary means of interpretation, something other workers had not done. Further, as part of the preparation, the patients ingested barium tablets and oral contrast agents. The software that was used automatically subtracted out the barium and contrast. This had never been done before as well.

At The Cleveland Clinic, in our relatively small study of 83 patients, we have found the following:

• Any patient over the age of 50 who has never had a colon screening exam and is not in the exclusion group; it is controversial as to whether anyone with a first-degree relative with colon carcinoma should have a VC over a regular colonoscopy
• Any patient who has had a regular colonoscopy, and in whom polyps have been identified and removed in the past

Who should not have a virtual colonoscopy?

• Any patient with familial polyposis
• Any patient with blood in their stool
• Any patient with ulcerative colitis or Crohn's disease

Does private insurance or Medicare pay for a virtual colonoscopy?
Currently, neither Medicare nor insurance companies pay for this technology. There is hope that with recent evidence, these parties will soon reimburse for the study.

Can a virtual colonoscopy detect abnormalities outside the colon?
The "whole body screening" trend with CT has touted the virtues of detecting undiscovered abnormalities. While the data to date is inconclusive, it is likely that a virtual colonoscopy will discover an abnormality outside the colon in a small percentage of patients. These abnormalities may include a renal cancer, an aortic aneurysm, enlarged lymph nodes, or even a pelvic mass in a woman. Published rates for significant abnormalities are between 5 and 15%. You should also know that the CT may detect a potential abnormality that will require further investigation before a certain diagnosis is reached.

What future advances will improve virtual colonoscopy?
The field of computer aided or assisted diagnosis (CAD) is rapidly expanding in radiology. Already, the standard of care for mammography is to use CAD to assist the radiologist in detecting breast cancer. Research has proven that CAD is efficacious in mammography. CAD is now being tested in virtual colonoscopy and the preliminary results are encouraging.