DOI: 10.1148/rg.235035701
(Radiographics. 2003;23:1073-1091.)
© RSNA, 2003
Filling Defects at CT Colonography: Pseudo- and Diminutive Lesions (The Good), Polyps (The Bad), Flat Lesions, Masses, and Carcinomas (The Ugly)1
Michael Macari, MD,
Edmund J. Bini, MD,
Stacy L. Jacobs, MD,
Nick Lange, BS and
Yvonne W. Lui, MD
1 From the Departments of Radiology (M.M., S.L.J., N.L., Y.W.L.) and Medicine (E.J.B.), New York University Medical Center, 560 First Ave, Suite HW 207, New York, NY 10016; and the Department of Medicine, Veterans Administration Medical Center, New York, NY (E.J.B.). Received January 7, 2003; revision requested March 6 and received April 1; accepted April 4. Address correspondence to M.M. (e-mail: michael.macari@med.nyu.edu).
 |
Abstract
|
|---|
Numerous filling defects may be detected in the colon during interpretation of data sets obtained with computed tomographic (CT) colonography. A series of 230 patients were evaluated with thin-section multidetector row CT colonography immediately before conventional colonoscopy. In all cases, the interpreting radiologist and gastroenterologist reviewed the imaging findings as well as the results of histologic analysis of biopsy specimens to determine the causes of filling defects. In many cases, the cause of a filling defect can be confidently determined at CT colonography by using combinations of two- and three-dimensional images. However, lesions will occasionally be indeterminate because of overlapping features and will require further evaluation with endoscopy. With knowledge of the morphologic and attenuation characteristics of the various filling defects in the colon, one should be able to differentiate those filling defects detected at CT colonography that require no further evaluation from those that require endoscopic interrogation.
© RSNA, 2003
Index Terms: Colon, CT, 75.1211 • Colon neoplasms, CT, 75.1211, 75.30 • Colon neoplasms, diagnosis, 75.30 • Computed tomography (CT), three-dimensional, 75.12117
|
LEARNING OBJECTIVES FOR TEST 2
|
|---|
After reading this article and taking the test, the reader will be able to:
- Describe the technique of thin-section CT colonography.
- Discuss how two-and three-dimensional CT colonography techniques are complementary.
- Identify the morphologic and attenuation characteristics that enable differentiation of filling defects in the colon.
|
Introduction
|
|---|
Computed tomographic (CT) colonography is rapidly becoming a viable option in the imaging armamentarium in the evaluation of the colon (1). For CT colonography to ultimately succeed as a colon imaging tool, studies need to show not only a high sensitivity in the detection of colorectal polyps but also a high specificity. There is debate among gastroenterologists and radiologists about what constitutes a clinically significant colorectal polyp (2–4). Some argue that 10 mm is a good target, since the prevalence of carcinoma is extremely low in polyps smaller than this size (3). Others are concerned that many polyps measuring between 6 and 9 mm are tubular or villous adenomas and believe these should be removed if detected (4). In practice, some gastroenterologists biopsy every colonic mucosal elevation detected during colonoscopy. Ultimately, a 6-mm colorectal polyp may be the lesion that requires detection with high sensitivity and specificity for virtual colonoscopy to be accepted as an imaging tool by the clinical community.
In addition to a high sensitivity and specificity for colorectal polyps, techniques of patient preparation, data acquisition, and image interpretation need to be disseminated into the general radiology community for CT colonography to make a widespread impact. In an attempt to optimize data interpretation, recent studies have focused on improving patient preparation and data acquisition techniques (5–14).
As regards image interpretation, a primary two-dimensional (2D) image review with multiplanar reformation (MPR) and endoluminal (three-dimensional [3D]) imaging for problem solving is the preferred method for data interpretation, allowing evaluation to be performed in a time-efficient manner (15–17). However, with thin-section multidetector row CT and enhanced computer processing workstations, it is certainly possible to perform image interpretation with a primary 3D display technique. One recent study showed improved detection of diminutive lesions in an ideally prepared and distended pig colon with use of 3D and MPR views compared with use of axial images alone (18). However, in a human subject undergoing routine bowel preparation and distention, the ability of endoluminal imaging to demonstrate these diminutive filling defects may lead to decreased specificity as a result of the detection of adherent bits of residual fecal material.
As for interpretation of CT colonographic data, there has been little published about the characteristics of detected colorectal filling defects that can lead to increased sensitivity and the important factor of improved specificity. By improving specificity, unnecessary endoscopy can be avoided. Several pictorial essays based primarily on single-section CT colonography have been published. These essays have demonstrated some of the characteristic morphologic features of filling defects at CT colonography (19–21). With use of thin-section multidetector row CT colonography, near-isotropic data sets are available for image review. This may lead to improved specificity by improving the z-axis resolution, thereby facilitating differentiation of the three most common filling defects detected at CT colonography: stool, bulbous or irregular folds, and colorectal polyps and neoplasms.
This review demonstrates the 2D and 3D morphologic characteristics of good, bad, and ugly filling defects in the colon. Techniques that facilitate differentiation of these filling defects are reviewed. Moreover, the limitations of CT colonography in distinguishing one filling defect from another are emphasized.
The case material presented in this article is from a series of 230 patients who underwent thin-section multidetector row CT colonography immediately prior to conventional colonoscopy. In all cases, the interpreting radiologist and gastroenterologist reviewed the imaging findings as well as histologic results of biopsied material to determine the cause of a filling defect.
|
CT Colonography Technique
|
|---|
Optimal CT colonography technique relies on careful cleansing and distention of the colon, as well as acquisition with thin-section multidetector row CT. On the day prior to the examination, one of three different bowel preparations is administered. These consist of oral hydration with two 45-mL doses of phosphosoda (Fleet Prep 1; Fleet Pharmaceuticals, Lynchburg, Va), 4 L of polyethylene glycol electrolyte solution (Golytely; Braintree Laboratories, Braintree, Mass), or magnesium citrate (LoSo Preparation; E-Z-Em, Westbury, NY). On site and immediately prior to CT colonography, the patient is asked to evacuate any residual fluid from the rectum.
All CT colonography was performed on a multidetector row CT system (Plus 4 Volume Zoom; Siemens Medical Systems, Forchheim, Germany). No bowel relaxant is used. A flexible rubber catheter is inserted into the rectum, and the colon is insufflated with room air to patient tolerance by an experienced technologist or nurse practitioner. The catheter is left in the rectum, and a supine scout CT image is obtained to verify adequate bowel distention. If adequate bowel distention is present, the CT examination is performed. If adequate bowel distention is not achieved, additional air is insufflated into the rectum. After air insufflation, CT colonography is performed with the patient supine in a cephalocaudal direction encompassing the entire colon and rectum. The patient is then placed in the prone position. Several additional puffs of air are administered, and the rectal catheter is removed. After acquisition of a second scout localizing image, the process is repeated over the same z-axis range.
CT parameters used are 4 x 1-mm section detector collimation, 120 kV, 0.5-second gantry rotation, and 50 mAs (effective). Pitch (table feed per gantry rotation/nominal section thickness) is varied between 6 and 7 so that the entire abdomen and pelvis can be covered within a 30-second breath hold. CT images are reconstructed as 1.25-mm-thick sections with a 1-mm reconstruction interval. Data are reviewed and images are interpreted on a Vitrea 2 workstation (Vital Images, Plymouth, Minn) or a Leonardo workstation (Siemens Medical Systems).
|
Pseudo- and Diminutive Lesions (The Good)
|
|---|
In reality, no filling defect is a "good" filling defect because every time an abnormality is detected, it requires careful analysis to determine if it is something that can be ignored or something that may require endoscopic confirmation and removal. Therefore, the importance of bowel preparation cannot be stressed enough to patients. Currently, the ideal situation for interpretation of CT colonographic data is a clean, dry, distended colon that allows rapid data evaluation (Fig 1). However, despite the most optimal bowel and patient preparation residual fecal material may persist, and irregular bulbous folds will always require further investigation.
View larger version (82K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Optimal preparation of the colon for virtual colonoscopy. (a) Axial CT image shows multiple colonic segments that are well distended, dry, and without residual fecal material. (b) Endoluminal CT image of the descending colon shows a well-distended colon. Note the interhaustral fold (arrow).
|
|
View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Optimal preparation of the colon for virtual colonoscopy. (a) Axial CT image shows multiple colonic segments that are well distended, dry, and without residual fecal material. (b) Endoluminal CT image of the descending colon shows a well-distended colon. Note the interhaustral fold (arrow).
|
|
When interpreting CT colonography data sets, the filling defects that can be ignored if they are confidently differentiated from colorectal polyps are as follows: residual fecal material, bulbous folds, lipomas, prominent ileocecal valves, extrinsic compression defects, and foreign bodies. In addition, diminutive irregularities that cannot be accurately classified and measure less than 5 mm can likely be dismissed given the fact that they are overwhelmingly going to be lesions that have no clinical significance (adherent solid residual fecal material, hyperplastic polyps, or small adenomas that do not harbor colorectal cancer).
Residual Fecal Material
Residual fecal material accounts for the vast majority of false-positive findings at CT colonography (22). However, by using supine and prone thin-section multidetector row CT colonography and a combination of 2D and 3D imaging, most residual fecal material can be distinguished from polyps (5). The three characteristics that are used to determine if a filling defect is residual fecal material include mobility, morphology, and internal attenuation characteristics.
In general, mobility implies residual fecal material (Fig 2). However, it is important to realize that several segments of the colon are quite mobile due to the presence of long attached mesenteries. The long mesentery may allow the colon to rotate when the patient is moved from the supine to the prone position. These segments include the cecum, transverse colon, and sigmoid colon. Occasionally, a filling defect in one of these segments will appear to change position relative to the colonic surface when the patient is moved from the supine to the prone position (Fig 3). However, on careful inspection it is the colon that has changed position, not the filling defect. In these cases, evaluation of the quality of bowel preparation, morphology, and attenuation characteristics of the lesion can be helpful in differentiating fecal material from polyp. A second reason why a polyp may appear mobile is if it is pedunculated and has a long stalk. In these cases, the head of the polyp will change position with gravity when the patient is moved from the supine to the prone position. Recognizing the morphology of the polyp (head and stalk) is helpful in differentiating these lesions from mobile fecal material.
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Mobility as an indicator of residual fecal material. (a) Supine axial CT image shows a 15-mm homogeneously attenuating filling defect (arrow) on the dorsal aspect of the rectum. (b) Prone axial CT image shows that the filling defect (arrow) is now on the ventral aspect of the rectum, a finding indicative of mobility. In general, mobility indicates residual fecal material.
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Mobility as an indicator of residual fecal material. (a) Supine axial CT image shows a 15-mm homogeneously attenuating filling defect (arrow) on the dorsal aspect of the rectum. (b) Prone axial CT image shows that the filling defect (arrow) is now on the ventral aspect of the rectum, a finding indicative of mobility. In general, mobility indicates residual fecal material.
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Pitfall of apparent mobility of a polyp. (a) Supine axial CT image shows a homogeneously attenuating pedunculated filling defect (arrow) in the sigmoid colon. Note that the filling defect is on the ventral, nondependent wall. (b) Prone axial CT image shows that the filling defect (arrow) is now on the dorsal wall. Again, the filling defect is on the nondependent wall. (c) Endoluminal CT image shows the filling defect (arrow), which measures 7 mm. (d) Image from conventional colonoscopy shows that the filling defect is a tubular adenoma (arrow). Apparent mobility of a filling defect, especially in the cecum, transverse colon, and sigmoid colon should be interpreted cautiously, since these areas of the colon may rotate when the patient is turned from supine to prone. If the colon is well prepared and the filling defect is homogeneously attenuating and polypoid, it should be considered a polyp.
|
|
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Pitfall of apparent mobility of a polyp. (a) Supine axial CT image shows a homogeneously attenuating pedunculated filling defect (arrow) in the sigmoid colon. Note that the filling defect is on the ventral, nondependent wall. (b) Prone axial CT image shows that the filling defect (arrow) is now on the dorsal wall. Again, the filling defect is on the nondependent wall. (c) Endoluminal CT image shows the filling defect (arrow), which measures 7 mm. (d) Image from conventional colonoscopy shows that the filling defect is a tubular adenoma (arrow). Apparent mobility of a filling defect, especially in the cecum, transverse colon, and sigmoid colon should be interpreted cautiously, since these areas of the colon may rotate when the patient is turned from supine to prone. If the colon is well prepared and the filling defect is homogeneously attenuating and polypoid, it should be considered a polyp.
|
|
View larger version (175K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. Pitfall of apparent mobility of a polyp. (a) Supine axial CT image shows a homogeneously attenuating pedunculated filling defect (arrow) in the sigmoid colon. Note that the filling defect is on the ventral, nondependent wall. (b) Prone axial CT image shows that the filling defect (arrow) is now on the dorsal wall. Again, the filling defect is on the nondependent wall. (c) Endoluminal CT image shows the filling defect (arrow), which measures 7 mm. (d) Image from conventional colonoscopy shows that the filling defect is a tubular adenoma (arrow). Apparent mobility of a filling defect, especially in the cecum, transverse colon, and sigmoid colon should be interpreted cautiously, since these areas of the colon may rotate when the patient is turned from supine to prone. If the colon is well prepared and the filling defect is homogeneously attenuating and polypoid, it should be considered a polyp.
|
|
View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3d. Pitfall of apparent mobility of a polyp. (a) Supine axial CT image shows a homogeneously attenuating pedunculated filling defect (arrow) in the sigmoid colon. Note that the filling defect is on the ventral, nondependent wall. (b) Prone axial CT image shows that the filling defect (arrow) is now on the dorsal wall. Again, the filling defect is on the nondependent wall. (c) Endoluminal CT image shows the filling defect (arrow), which measures 7 mm. (d) Image from conventional colonoscopy shows that the filling defect is a tubular adenoma (arrow). Apparent mobility of a filling defect, especially in the cecum, transverse colon, and sigmoid colon should be interpreted cautiously, since these areas of the colon may rotate when the patient is turned from supine to prone. If the colon is well prepared and the filling defect is homogeneously attenuating and polypoid, it should be considered a polyp.
|
|
The morphology of a filling defect is the second characteristic that should be evaluated when attempting to differentiate residual fecal material and polyps. Colorectal polyps are round, oval, or lobulated. They do not contain geometric borders or sharp angles (Fig 4). Although residual fecal material may be round, oval, or lobulated, it often contains angled borders or geometric morphology. When this morphology appears on 2D or 3D images, residual fecal material is present and the lesion can be ignored. When thin-section multidetector row CT is used, morphology can often be better appreciated by using endoluminal imaging (Fig 5).
View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Eight-millimeter round filling defect in the rectum. (a) Axial CT image shows a well-circumscribed, homogeneously attenuating, round filling defect (arrow) on an interhaustral fold. Arrowhead = rectal catheter. (b) Endoluminal CT image shows the round polypoid morphology of the lesion (arrow). Arrowhead = rectal catheter. (c) Image from conventional colonoscopy shows the identical morphology (arrow). Histologic analysis revealed a tubular adenoma.
|
|
View larger version (176K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Eight-millimeter round filling defect in the rectum. (a) Axial CT image shows a well-circumscribed, homogeneously attenuating, round filling defect (arrow) on an interhaustral fold. Arrowhead = rectal catheter. (b) Endoluminal CT image shows the round polypoid morphology of the lesion (arrow). Arrowhead = rectal catheter. (c) Image from conventional colonoscopy shows the identical morphology (arrow). Histologic analysis revealed a tubular adenoma.
|
|
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Eight-millimeter round filling defect in the rectum. (a) Axial CT image shows a well-circumscribed, homogeneously attenuating, round filling defect (arrow) on an interhaustral fold. Arrowhead = rectal catheter. (b) Endoluminal CT image shows the round polypoid morphology of the lesion (arrow). Arrowhead = rectal catheter. (c) Image from conventional colonoscopy shows the identical morphology (arrow). Histologic analysis revealed a tubular adenoma.
|
|
View larger version (176K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5. Geometric morphology as an indicator of fecal material. Endoluminal CT image of the sigmoid colon shows a filling defect with angled edges (arrow). This appearance is not consistent with a polyp. Small polyps are round, oval, or lobulated but do not contain geometric edges. Visualization of 3D morphology is facilitated by thin-section multisection CT.
|
|
Finally, the internal attenuation characteristics of a filling defect can be helpful in distinguishing residual fecal material and polyps. Small polyps are homogeneous in attenuation since they do not ulcerate. Although fecal material may be homogeneous in attenuation, it frequently contains areas of low or high attenuation. Areas of low attenuation represent trapped gas (Fig 6), and areas of high attenuation represent high-density food particles or tagging agents. Occasionally, larger polyps may ulcerate. In these cases, the ulceration is on the surface of the lesion and internally the lesion will remain homogeneously attenuating.
View larger version (166K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Internal heterogeneity as an indicator of fecal material. (a) Endoluminal CT image shows a round well-circumscribed filling defect (arrow) in the cecum. (b) Prone axial CT image shows that the filling defect (arrow) has internal low attenuation, which indicates gas. In small lesions, internal heterogeneity is an indicator of residual fecal material.
|
|
View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Internal heterogeneity as an indicator of fecal material. (a) Endoluminal CT image shows a round well-circumscribed filling defect (arrow) in the cecum. (b) Prone axial CT image shows that the filling defect (arrow) has internal low attenuation, which indicates gas. In small lesions, internal heterogeneity is an indicator of residual fecal material.
|
|
A major barrier to widespread colon cancer screening is bowel preparation. There is currently investigation into improving the preparation that is required for CT colonography. Indeed, it has been suggested by gastroenterologists that if CT colonography could be performed without bowel preparation, it would be the screening test of choice (23). Methods of fecal and fluid tagging as well as segmentation techniques to subtract out "tagged fecal material" are being investigated (9–11). By homogeneously labeling fecal material with previously ingested high-attenuation barium, one study showed 100% sensitivity and specificity for colorectal polyps 10 mm and larger (11). However, in that study a laxative, consisting of magnesium citrate, was administered in addition to the barium fecal tagging.
When CT data are interpreted after fecal tagging, residual fecal material is easily differentiated from folds and polyps since it has homogeneous high attenuation (Fig 7). Fecal tagging may also allow decreased interpretation times, since tagged residual fecal material can be rapidly differentiated from polyps. More investigation is required for fecal tagging and segmentation techniques to be used without prior bowel preparation.
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7. Fecal tagging with barium as an aid in differentiation of fecal material from polyps. Axial CT image shows multiple lesions with homogeneous high attenuation (arrow) in the sigmoid colon, an appearance compatible with fecal material.
|
|
Bulbous and Irregular Interhaustral Folds
On axial images, interhaustral folds may appear as rounded or pedunculated filling defects. This is frequently the case in segments of the colon that are not well distended. However, by scrolling through contiguous axial images at a workstation, the linear nature of the filling defect can usually be confirmed. If difficulty persists in differentiating a bulbous interhaustral fold from a polyp on 2D images, coronal, sagittal, and endoluminal images may be used for problem solving by displaying the true linear morphology of the fold (Fig 8).
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Bulbous fold mimicking a pedunculated polyp. (a) Supine axial CT image shows a pedunculated lesion (arrow) in the sigmoid colon. (b) Coronal CT image shows the linear morphology of the lesion (arrow). (c) Endoluminal CT image shows that the lesion is an interhaustral fold (arrow). Note the small diverticula (arrowheads).
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Bulbous fold mimicking a pedunculated polyp. (a) Supine axial CT image shows a pedunculated lesion (arrow) in the sigmoid colon. (b) Coronal CT image shows the linear morphology of the lesion (arrow). (c) Endoluminal CT image shows that the lesion is an interhaustral fold (arrow). Note the small diverticula (arrowheads).
|
|
View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. Bulbous fold mimicking a pedunculated polyp. (a) Supine axial CT image shows a pedunculated lesion (arrow) in the sigmoid colon. (b) Coronal CT image shows the linear morphology of the lesion (arrow). (c) Endoluminal CT image shows that the lesion is an interhaustral fold (arrow). Note the small diverticula (arrowheads).
|
|
Occasionally, even with use of MPR and endoluminal images, it may be difficult to differentiate a bulbous interhaustral fold from a polyp. These cases are usually located in the hepatic or splenic flexural regions and are likely due to converging folds. In some cases, it may be impossible to differentiate these processes; if the lesion is larger than 5 mm, endoscopy may be warranted to exclude adenoma.
Lipomas and the Ileocecal Valve
With any imaging study (double-contrast barium enema examination, colonoscopy, or CT colonography), it is very difficult, if not impossible, to determine the histologic nature of a neoplasm. An exception is the lipoma (Fig 9). Colonic lipomas are the most common submucosal tumor of the colon, occurring in approximately 1.2% of patients at colonoscopy (24). During interpretation of CT colonographic data sets, it is important to carefully interrogate all detected lesions with standard colon window settings (width, 1,500 HU; level, -200 HU) and with abdominal window settings (width, 400 HU; level, 40 HU). Lipomas are not uncommon and should not prompt endoscopic examination. Occasionally, large colonic lipomas may serve as a lead point for colonic intussusceptions. In these cases, endoscopic or surgical resection may be required (25).
View larger version (183K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Usefulness of multiple window settings in evaluation of filling defects. (a) Endoluminal CT image shows an 11-mm filling defect of indeterminate origin (arrows) in the cecum. (b) Image from conventional colonoscopy shows the lesion (arrows). (c) Supine axial CT image obtained with a window width of 1,500 HU and window level of -200 HU shows that the lesion (arrow) is indeterminate. (d) Supine axial CT image obtained with a window width of 400 HU and window level of 10 HU shows that the lesion (arrow) is a lipoma. The presence of adipose tissue in a filling defect is confirmation of a lipoma.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Usefulness of multiple window settings in evaluation of filling defects. (a) Endoluminal CT image shows an 11-mm filling defect of indeterminate origin (arrows) in the cecum. (b) Image from conventional colonoscopy shows the lesion (arrows). (c) Supine axial CT image obtained with a window width of 1,500 HU and window level of -200 HU shows that the lesion (arrow) is indeterminate. (d) Supine axial CT image obtained with a window width of 400 HU and window level of 10 HU shows that the lesion (arrow) is a lipoma. The presence of adipose tissue in a filling defect is confirmation of a lipoma.
|
|
View larger version (118K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Usefulness of multiple window settings in evaluation of filling defects. (a) Endoluminal CT image shows an 11-mm filling defect of indeterminate origin (arrows) in the cecum. (b) Image from conventional colonoscopy shows the lesion (arrows). (c) Supine axial CT image obtained with a window width of 1,500 HU and window level of -200 HU shows that the lesion (arrow) is indeterminate. (d) Supine axial CT image obtained with a window width of 400 HU and window level of 10 HU shows that the lesion (arrow) is a lipoma. The presence of adipose tissue in a filling defect is confirmation of a lipoma.
|
|
View larger version (171K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9d. Usefulness of multiple window settings in evaluation of filling defects. (a) Endoluminal CT image shows an 11-mm filling defect of indeterminate origin (arrows) in the cecum. (b) Image from conventional colonoscopy shows the lesion (arrows). (c) Supine axial CT image obtained with a window width of 1,500 HU and window level of -200 HU shows that the lesion (arrow) is indeterminate. (d) Supine axial CT image obtained with a window width of 400 HU and window level of 10 HU shows that the lesion (arrow) is a lipoma. The presence of adipose tissue in a filling defect is confirmation of a lipoma.
|
|
The ileocecal valve may have a variable appearance at imaging (19,26). It manifests as a filling defect along the medial aspect of the cecum in most patients (Fig 10). Following the terminal ileum into the colon is often helpful in confirming that the filling defect is the ileocecal valve. When the cecum is on a persistent mesentery, the valve may be on the lateral aspect of the cecum. Often, the ileocecal valve contains adipose tissue, which again is helpful in distinguishing the valve from other filling defects. At virtual colonoscopy, the valve may have one of three appearances: labial, in which the mouth of the valve is set between an upper and lower fold; papillary, in which the valve appears as a domelike protrusion; and mixed, with features of both types (Fig 11) (19). A recent study that evaluated the appearance of the ileocecal valve during double-contrast barium enema examination showed that the mean height of the valve was 1.7 cm and the mean width was 2.8 cm (26). It is important to remember that the valve is covered by colonic mucosa, and polyps and cancers may develop on the valve. Therefore, if the appearance of the ileocecal valve is abnormal or unusual, endoscopy should be performed to exclude a neoplasm arising on the valve.
View larger version (204K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Ileocecal valve appearing as a prominent filling defect. (a) Endoluminal CT image shows a large filling defect (arrow) in the cecum. (b) Coronal CT image shows a prominent ileocecal valve (arrow). The ileocecal valve may have a number of appearances at endoluminal imaging.
|
|
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Ileocecal valve appearing as a prominent filling defect. (a) Endoluminal CT image shows a large filling defect (arrow) in the cecum. (b) Coronal CT image shows a prominent ileocecal valve (arrow). The ileocecal valve may have a number of appearances at endoluminal imaging.
|
|
View larger version (198K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11. Ileocecal valve appearing as a thickened fold. Endoluminal CT image shows a thickened fold with a central opening (large arrow), which represents the ileocecal valve. Note the small polyp in the ascending colon (small arrow).
|
|
Extrinsic Compression Defects, Diverticula, and Foreign Bodies
If CT data interpretation is performed with a primary endoluminal viewing technique, extrinsic compression defects may also manifest as substantial filling defects in the colon (19). These compression defects may be from adjacent bowel loops (Fig 12), the surface of the liver or spleen, the aorta in thin patients, or accessory spleens. When only endoluminal imaging is used, these filling defects cannot be distinguished from true polyps and neoplasms. Whether one relies on a primary 2D or 3D viewing technique, both must be available to facilitate adequate characterization of a filling defect. Therefore, extrinsic compressions, although potentially confusing at 3D imaging, are easily characterized with a combination of axial and MPR imaging.
View larger version (185K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Extrinsic compression appearing as a filling defect in the transverse colon. (a) Endoluminal CT image shows a large smooth filling defect (arrow). (b) Prone axial CT image shows that the filling defect (arrow) is partially extrinsic to the colon. (c) Coronal CT image shows the filling defect as a tubular structure (arrow), which represents a jejunal loop compressing the transverse colon.
|
|
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Extrinsic compression appearing as a filling defect in the transverse colon. (a) Endoluminal CT image shows a large smooth filling defect (arrow). (b) Prone axial CT image shows that the filling defect (arrow) is partially extrinsic to the colon. (c) Coronal CT image shows the filling defect as a tubular structure (arrow), which represents a jejunal loop compressing the transverse colon.
|
|
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12c. Extrinsic compression appearing as a filling defect in the transverse colon. (a) Endoluminal CT image shows a large smooth filling defect (arrow). (b) Prone axial CT image shows that the filling defect (arrow) is partially extrinsic to the colon. (c) Coronal CT image shows the filling defect as a tubular structure (arrow), which represents a jejunal loop compressing the transverse colon.
|
|
Another potential pitfall of 3D navigation is mistaking a diverticulum as a polyp (19). However, on careful inspection diverticula usually have a complete dark ring around them, whereas a raised filling defect will have an incomplete ring shadow (Fig 13). Occasionally, a diverticulum may not have this classic appearance at 3D imaging; in these cases, further evaluation with 2D imaging will be necessary.
View larger version (186K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13a. Difference between a polyp and a diverticulum at virtual and conventional colonoscopy. (a) Image from virtual colonoscopy shows a polyp (black arrow) and two small diverticula (white arrows) in the cecum. In general, there is an incomplete border around a polyp, whereas a diverticulum has a complete ring around the orifice. If there is uncertainty at endoluminal imaging, axial images are helpful in differentiating these entities. (b) Image from conventional colonoscopy shows the polyp (black arrow), which measures 7 mm, and the diverticula (white arrows).
|
|
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13b. Difference between a polyp and a diverticulum at virtual and conventional colonoscopy. (a) Image from virtual colonoscopy shows a polyp (black arrow) and two small diverticula (white arrows) in the cecum. In general, there is an incomplete border around a polyp, whereas a diverticulum has a complete ring around the orifice. If there is uncertainty at endoluminal imaging, axial images are helpful in differentiating these entities. (b) Image from conventional colonoscopy shows the polyp (black arrow), which measures 7 mm, and the diverticula (white arrows).
|
|
The most common foreign body is retained fecal material. Other foreign bodies that are always seen are the rectal catheters. These should pose no problem for data interpretation, especially when thin catheters without balloons are used for insufflation. Note that optimal colonic distention is not dependent on placement of large-caliber rectal catheters with balloons. These catheters are quite uncomfortable for many patients, and thin rubber catheters can be reliably used to distend the colon (Fig 14). Moreover, large balloons may obscure low rectal lesions at any imaging study.
View larger version (176K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14a. Rectal catheters. (a) Endoluminal CT image shows a large-caliber plastic catheter (diameter, 16 mm) (arrow). (b) Endoluminal CT image shows a small-caliber rubber catheter (diameter, 5 mm) (arrow). Both catheters allow good colonic distention, with less discomfort for the patient when the thin catheter is used.
|
|
View larger version (174K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 14b. Rectal catheters. (a) Endoluminal CT image shows a large-caliber plastic catheter (diameter, 16 mm) (arrow). (b) Endoluminal CT image shows a small-caliber rubber catheter (diameter, 5 mm) (arrow). Both catheters allow good colonic distention, with less discomfort for the patient when the thin catheter is used.
|
|
Diminutive Filling Defects
Published studies that evaluated CT colonography with colonoscopic correlation have consistently demonstrated that polyp detection rates are directly related to polyp size (Fig 15) (1,27,28). Moreover, when a diminutive filling defect is seen, it could represent adherent fecal material, a normal colonic mucosal protrusion, a hyperplastic polyp, or a small adenoma. The first three lesions have no malignant potential. The significance of missing a diminutive adenoma is controversial (5,29–31). Small adenomas will be missed with virtual colonoscopy (5). However, if follow-up imaging or screening is performed on an interval basis, missing these small lesions is likely not important, since the dwell time of these lesions is long (up to 10 years) in the development and progression of the adenoma-carcinoma sequence (3,32).
View larger version (127K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 15. Diminutive filling defect at colonoscopy. Image from conventional colonoscopy shows a 3-mm raised filling defect (arrows), which could not be visualized even in retrospect with 2D or 3D CT colonography. At histologic analysis, the filling defect was a hyperplastic polyp. Most such small lesions are not seen at CT colonography; however, their clinical significance is minimal.
|
|
Occasionally, small filling defects will be seen at CT colonography. As a result, an important question is what to do when these diminutive filling defects are identified at CT colonography. Since the differential diagnosis includes homogeneously attenuating adherent fecal material, small hyperplastic polyps, and small adenomas, some centers performing CT colonography no longer report these lesions because of the high probability that they are not significant (Fig 16) (15). In fact, it has been suggested by some gastroenterologists that attention should be shifted away from finding and removing all diminutive polyps and toward strategies that will allow more reliable detection of the less common, but more dangerous, larger advanced adenomas (4).
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16a. Advantage of performing both supine and prone imaging. (a, b) Prone (a) and supine (b) axial CT images show a small filling defect (arrow) in the descending colon. The filling defect is not mobile. (c) Endoluminal CT image shows that the lesion is polypoid (arrow). At colonoscopy, a 5-mm hyperplastic polyp was found.
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16b. Advantage of performing both supine and prone imaging. (a, b) Prone (a) and supine (b) axial CT images show a small filling defect (arrow) in the descending colon. The filling defect is not mobile. (c) Endoluminal CT image shows that the lesion is polypoid (arrow). At colonoscopy, a 5-mm hyperplastic polyp was found.
|
|
View larger version (186K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 16c. Advantage of performing both supine and prone imaging. (a, b) Prone (a) and supine (b) axial CT images show a small filling defect (arrow) in the descending colon. The filling defect is not mobile. (c) Endoluminal CT image shows that the lesion is polypoid (arrow). At colonoscopy, a 5-mm hyperplastic polyp was found.
|
|
|
Polyps (The Bad)
|
|---|
There is controversy regarding what should be considered a clinically significant polyp. As stated earlier, some argue that we should disregard diminutive polyps and concentrate on larger lesions (4). Many radiologists point to the low prevalence of carcinoma in polyps less than or equal to 10 mm as a rationale for stating that 10 mm should be the target lesion we are trying to detect. However, some gastroenterologists disagree with this recommendation, since some polyps measuring 6–9 mm may contain severe dysplasia or very rarely carcinoma. If these polyps are left in place for a period of 5 years, some may progress to larger lesions or even invasive cancer. Therefore, others argue that 6 mm should be the target lesion for CT colonography and when detected, should prompt endoscopic removal. For CT colonography to become a competitive imaging tool, it needs to show a moderate to good sensitivity for polyps measuring 6–9 mm and excellent sensitivity for polyps greater than or equal to 10 mm.
The ability of CT colonography to demonstrate lesions measuring 6–9 mm is decreased compared with the ability to demonstrate polyps greater than or equal to 10 mm (1). However, with combination 2D and 3D interpretation techniques, the ability to detect lesions in the 6–9-mm range may be increased compared with the detection ability of primary axial review only (18,27). A study of 100 patients who underwent back-to-back virtual and conventional colonoscopy showed a per lesion sensitivity of 100% for colorectal carcinoma, 91% for polyps that were 10 mm or larger, and 82% for polyps that were 6–9 mm (27). Unfortunately, use of complete combinations of 2D and 3D imaging will result in long interpretation times and likely more false-positive findings. Computer-assisted detection algorithms hold promise in improving detection rates of polyps in this size range with only minimal amounts of increased interpretation times (28).
Although there are few polyps measuring 6–9 mm that contain carcinoma, up to 50% of lesions in this size range will be adenomas (29,30). Therefore, polyps in this size range should be considered potentially "bad" lesions, since some have malignant potential and studies have demonstrated that removing lesions such as these at sigmoidoscopy may decrease the risk of colorectal cancer (31). It is well established that colorectal carcinogenesis occurs in most cases as a polyp progresses from a tubular adenoma to villous adenoma and finally to carcinoma by means of a series of mutations in the lesion (32). It is not possible to differentiate the histologic nature of colorectal polyps at conventional or virtual colonoscopy. However, size is clearly important, and the larger a lesion is, the more likely it is to harbor dysplasia or carcinoma (30,32). As regards the morphologic assessment of polyps, pedunculated lesions are less likely to harbor carcinoma than sessile lesions; however, the histologic nature cannot be predicted on the basis of morphology alone (Figs 3, 4, 13, 17).
View larger version (179K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17a. Large pedunculated polyp. (a) Axial CT image shows a pedunculated polyp (arrow) in the sigmoid colon. With the stalk, it measures 37 mm. However, such a polyp should be measured according to the size of the head, which is 13 mm. (b, c) Images from conventional (b) and virtual (c) colonoscopy show the head of the polyp (arrow). Histologic analysis revealed a hyperplastic polyp.
|
|
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17b. Large pedunculated polyp. (a) Axial CT image shows a pedunculated polyp (arrow) in the sigmoid colon. With the stalk, it measures 37 mm. However, such a polyp should be measured according to the size of the head, which is 13 mm. (b, c) Images from conventional (b) and virtual (c) colonoscopy show the head of the polyp (arrow). Histologic analysis revealed a hyperplastic polyp.
|
|
View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 17c. Large pedunculated polyp. (a) Axial CT image shows a pedunculated polyp (arrow) in the sigmoid colon. With the stalk, it measures 37 mm. However, such a polyp should be measured according to the size of the head, which is 13 mm. (b, c) Images from conventional (b) and virtual (c) colonoscopy show the head of the polyp (arrow). Histologic analysis revealed a hyperplastic polyp.
|
|
Therefore, if a lesion greater than or equal to 6 mm is detected at CT colonography and is not fecal residue, a fold, a lipoma, an extrinsic defect, or a foreign body, colonoscopic removal should be performed unless there is a clinical contraindication. Obviously, larger lesions (
10 mm) should be considered bad, and endoscopic removal should be performed when such lesions are detected at CT colonography. Note that, when pedunculated polyps are measured, it is the head of the polyp that should be measured, not the head and stalk.
|
Flat Lesions, Masses, and Carcinomas (The Ugly)
|
|---|
Although most "ugly" lesions are easy to identify at CT colonography, several are not easy to visualize and many are difficult to categorize. Ugly lesions include flat adenomas; tumors hidden behind folds, which can be difficult to detect at colonoscopy or endoluminal CT colonography; carcinomas; unusual-appearing tumors; and large irregular hemorrhoids.
Flat Adenomas and Tumors Behind or Adjacent to Folds
Flat colorectal lesions are difficult to identify at CT colonography even when thin-section techniques are used (5). A study of 46 patients who underwent screening CT colonography and colonoscopy found that only one of four flat adenomas greater than or equal to 2 cm that were seen at colonoscopy were identified at CT colonography (33). In our experience, truly flat lesions are very difficult to visualize and even in retrospect are usually not detected (Fig 18). However, these lesions are relatively rare in western populations. The truly flat adenoma is not raised and may in fact have central areas of depression. These lesions should be differentiated from slightly raised lesions (Fig 19). Although these lesions are much broader than they are raised, they are not flat tumors and should always be detected at CT colonography.
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18a. Flat adenoma in the rectum. (a) Image from conventional colonoscopy shows a slightly discolored area (arrows), which is not raised from the background mucosa. Analysis of the biopsy specimen revealed a tubular adenoma. (b) Axial CT image of the same area shows no abnormality (arrow). Even in retrospect, the abnormality could not be seen with 2D or 3D imaging. Truly flat adenomas are nearly impossible to visualize at CT colonography.
|
|
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 18b. Flat adenoma in the rectum. (a) Image from conventional colonoscopy shows a slightly discolored area (arrows), which is not raised from the background mucosa. Analysis of the biopsy specimen revealed a tubular adenoma. (b) Axial CT image of the same area shows no abnormality (arrow). Even in retrospect, the abnormality could not be seen with 2D or 3D imaging. Truly flat adenomas are nearly impossible to visualize at CT colonography.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 19a. Carpetlike filling defect. (a) Prone axial CT image shows fluid (arrow) obscuring the ventral wall of the rectum. (b) Supine axial CT image shows redistribution of the fluid and an irregular carpetlike filling defect (arrow) along the ventral rectal wall. (c) Endoluminal CT image shows the irregular carpetlike morphology of the rectal surface (arrow). Arrowhead = rectal catheter. (d) Image from conventional colonoscopy shows the irregular morphology of the rectal wall (arrow). Histologic analysis revealed a villous adenocarcinoma.
|
|
Top
Abstract
LEARNING OBJECTIVES FOR TEST...
