DOI:10.2214/AJR.04.1889
AJR 2006; 186:1611-1617
© American Roentgen Ray Society
Flat Polyps of the Colon: Detection with 16-MDCT Colonography—Preliminary Results
Seong Ho Park1,
Hyun Kwon Ha1,
Ah Young Kim1,
Kyoung Won Kim1,
Moon-Gyu Lee1,
Pyo Nyun Kim1,
Yong Moon Shin1,
Jeong-Sik Byeon2,
Suk-Kyun Yang2,
Jin Ho Kim2 and
Young Il Min2
1 Department of Radiology and Research Institute of Radiology, University of
Ulsan College of Medicine, Asan Medical Center, 388-1 Poongnap-Dong,
Songpa-Gu, Seoul, South Korea, 138-040.
2 Department of Internal Medicine, University of Ulsan College of Medicine, Asan
Medical Center, Seoul, South Korea.
Received December 13, 2004;
accepted after revision May 29, 2005.
Address correspondence to H. K. Ha
(hkha{at}amc.seoul.kr).
Abstract
OBJECTIVE. This study evaluates the ability of contrast-enhanced,
16-MDCT colonography to display flat colonic lesions when a very narrow slice
thickness (1 mm) is used.
CONCLUSION. Less than 50% of flat lesions in our population could be
visualized by blinded and unblinded review. Before they could be visualized,
flat lesions were 2 mm or greater in height and 7 mm or greater in diameter.
Lesions with a height of 1 mm or less were not seen on CT colonography.
Contrast enhancement, location on a haustral fold, and abnormal 2D and 3D
morphology contributed to lesion conspicuity.
Keywords: colon • colonography • colonoscopy • CT
Introduction
CT colonography is rapidly gaining acceptance as a viable option for
colon cancer screening. A recent study using MDCT and electronic cleansing
showed that the detection rate of CT colonography for adenomatous polyps 6 mm
or larger was similar to that of colonoscopy in the average-risk screening
population [1]. Although CT
colonography appears to have been successful in extending its role for
detecting colonic lesions, there are still some factors that impair lesion
detection. Among these factors, flat morphology of a colonic lesion may be an
important issue. Despite some controversy, a significant number of colonic
polyps are presumed to be flat, and two large series
[2,
3] showed that approximately
38% of the adenomatous lesions detected on colonoscopy were flat. The
malignant potential of flat adenomatous lesions of the colon is not clearly
understood. Some reports suggest that flat adenomatous lesions have a higher
risk of containing high-grade dysplasia or adenocarcinoma than sessile or
pedunculated lesions
[2-5];
however, the difference could be attributable in part to the variability of
the histologic diagnostic criteria used by pathologists
[6,
7]. In contrast to the
attention given to the sizes and histopathologic types (i.e., adenomatous or
hyper-plastic) of colonic polyps, however, the difficulty of diagnosing flat
lesions and the significance of missed flat lesions have not been sufficiently
addressed in the era of CT colonography. To our knowledge, there have only
been three large studies
[8-10]
of the detection rate of CT colonography for flat lesions. Because two of them
[8,
9] used a 5-mm slice thickness,
the results may not be relevant to those of the higher performance scanners
currently in widespread use. Also, those studies did not clearly address the
lesion height, which we hypothesized would be an important factor for
determining the visibility of flat colonic lesions. Therefore, the purpose of
this study was the preliminary evaluation of 16-MDCT colonography for the
detection of flat colonic lesions.
Materials and Methods
This study was a retrospective observational study for which our
institutional review board did not require informed consent.
Patient Population
Two hundred and thirteen consecutive patients underwent colonoscopy at our
institution during the period between January 2004 and July 2004 with the
following indications: patients referred for newly diagnosed colon cancer
(n = 180) or suspected of having colon cancer (n = 20),
routine screening (n = 5), and others (n = 8). Of those
patients, 32 flat colonic polyps were found in 21 patients. Six of the 21
patients did not undergo CT colonography, and five underwent CT colonography
after polypectomy. The remaining 10 patients, between 49 and 74 years old
(mean ± SD, 59.2 ± 8.4 years), with a total of 18 flat polyps,
comprised the study cohort. In those 10 patients, the time difference between
CT colonography and colonoscopy (mean ± SD) was 4.4 ± 7.7
days.
CT Colonography
Study technique—Bowel preparation was performed with a
low-residue diet followed by an oral magnesium carbonate preparation (Magcorol
Soln, Taejoon Pharm) and 10 mg of bisacodyl (Dulcolax, Boehringer Ingelheim
Korea) the evening before CT colonography. CT colonography was performed after
intramuscular injection of 20 mg of scopolamine N-butylbromide
(Buscopan, Boehringer Ingelheim Korea) if no contraindication was present. The
colon was gently insufflated with room air according to the patient's
tolerance in a left lateral decubitus position using a flexible rubber tube
placed in the rectum. The patient was first scanned in the prone position and
then in the supine position. Before scanning, an anteroposterior CT scout
image was obtained to ensure adequate bowel distention. Further air
insufflation was performed when collapsed bowel segments were identified.
During the scanning, the rectal tube was removed. CT was performed from the
top of the diaphragm down to the anus during a breath-hold with a 16-MDCT
scanner (Somatom Sensation 16, Siemens Medical Solutions) with the following
parameter settings: beam collimation, 16 x 0.75 mm; slice thickness, 1
mm; reconstruction interval, 0.7 mm; beam pitch, 1; gantry rotation time, 0.5
sec; table speed, 24 mm/sec; field of view to fit; 120 kV; and 150 mAs.
Scanning in the supine position was performed after IV injection of 150 mL of
iopromide (Ultravist 370, Schering) at a rate of 2.7 mL/sec through a 20-gauge
angiographic catheter inserted in the antecubital vein. Contrast enhancement
was aimed at detecting possible liver metastasis because most of the patients
either had colon cancer or were suspected of having colon cancer; the scanning
was performed 75 sec after initiation of contrast injection.
Image analysis—One board-certified gastrointestinal
radiologist who had experienced more than 100 CT colonography cases with
colonoscopic correlation, interpreted the CT colonography images at a
dedicated workstation using a software package with volume-rendering
capabilities (V-works 5.0, CyberMed). The observer was blinded to the
colonoscopic findings and was requested to report flat lesions. A 2D search
was done to select colonic lesions using transverse images and also, when the
transverse plane was not perpendicular to the bowel segments to be looked at,
oblique images that were perpendicular to the long axes of the bowel segments
as primary 2D views. The rationale for including oblique images as primary
views was that we assumed the flat lesions would be even more difficult to
detect when they were parallel to the review plane versus when they were
perpendicular to the review plane because some flat lesions would only
slightly project into the bowel lumen. To be classified as a flat lesion, the
lesion height had to be less than half the lesion diameter
[2,
6]. When a suspected lesion was
identified, the morphology was examined on coronal and sagittal reformations
and 3D endoluminal views. An irregular, angular, or barlike shape was
considered to suggest a false lesion. The observer examined the internal
attenuation of the suspected lesion by modifying the window width and level to
depict small gas bubbles, a finding that is consistent with stool. The images
were reviewed from both the prone and the supine data sets. Review of the CT
colonography images was performed during one session using a standard colon
window setting (width, 1,500 H; level, -200 H) and subsequently with an
intermediate soft-tissue window setting (width, 400 H; level, 20 H). Contrast
enhancement of a suspected lesion could be noted on the soft-tissue window
setting of the supine images, which was used to distinguish colonic lesions
from feces. The location of a lesion was specified as being one of the eight
segments: the cecum (C), ascending colon (A), hepatic flexure (HF), transverse
colon (T), splenic flexure (SF), descending colon (D), sigmoid colon (S), or
rectum (R). Locations other than the cecum were further recorded as proximal,
mid, or distal within each segment. The height from the luminal surface and
width of the polyp were measured on 2D images using an electronic ruler after
appropriate magnification.
CT Colonographic and Colonoscopic Comparison and Data Analysis
Colonoscopy was considered the reference standard in this study. Four
board-certified gastroenterologists, each of whom had performed more than
1,000 examinations, performed the colonoscopy using a standard
videocolonoscope (CF series, Olympus Optical). Lesion location was described
the same way as on CT colonography. Lesion morphology and size were described.
Flat lesions were defined as mucosal elevations with a height less than half
the lesion diameter [2,
6]. Lesion size, including the
height of the flat lesion, was determined by comparing the lesion with open
endoscopic biopsy forceps pushed against the lesion or by direct measure of
the resected specimen with a ruler when the lesion was retrieved in toto. For
a given flat lesion to be considered a true-positive match between CT
colonography and colonoscopy, the lesion had to be in the same colonic segment
or in the adjacent portion of the adjacent segments, and the two recorded
widths and heights had to be the same within a 50% margin of error. Flat
lesions that were missed on the blinded review were reassessed by two
radiologists in retrospect with the knowledge of the colonoscopic findings,
and judgments were made regarding the causes of the false-negative results.
Lesion conspicuity was qualitatively compared between the colon window 2D
view, soft-tissue window 2D view, and the 3D endoluminal view. For lesions
that could not be found on unblinded review but which were located in
distended and clean colonic segments, lesion height, width, location (haustral
fold vs haustra), and histopathology were compared with those of visualized
lesions. Fisher's exact test was used, and a p value less than 0.05
was considered statistically significant. For all the flat lesions that were
found on CT colonography but were not present at colonoscopy, all the
available information, such as surgical and pathologic reports or follow-up
colonoscopy results, was reviewed to verify if those were really
false-positive findings. The lesions heights, except for those of
false-positive lesions, referred to in the article were from colonoscopic
measurement.
Results
The summary of the 18 flat polyps with the CT colonographic results is
presented in Table 1. Four flat
lesions (22.2%, 4/18; 95% confidence interval [CI], 
47.8%) were detected
on the blinded review (Figs.
1A,
1B,
1C,
1D,
2A,
2B,
2C and
2D). On the unblinded review,
of the 14 lesions that were missed at the blinded review, a 7 x 2-mm
(width x height) hyperplastic polyp was found (Fig.
3A,
3B,
3C and
3D); six lesions could not be
visualized because of excessive luminal fluid, poor bowel preparation, and/or
poor bowel distention; and the remaining seven lesions could not be found
despite optimal preparation and distention of the relevant bowel segments.
Except for a 15 x 1-mm adenocarcinoma in situ that was not visualized
because of excessive luminal fluid and poor bowel distention, the other two
cancerous lesions were detected on blinded review (Fig.
1A,
1B,
1C and
1D). All the flat lesions that
were detected were found on the colon window setting, and no additional flat
lesion was found when the intermediate soft-tissue window setting was used.
However, three lesions could be confidently distinguished from feces by noting
their contrast enhancement when the soft-tissue window was used (Fig.
2A,
2B,
2C and
2D). Two lesions located on the
haustral fold showed higher conspicuity on the 3D endoluminal view than on the
colon window setting 2D view (Fig.
1A,
1B,
1C and
1D and
Table 1). For the lesions that
could not be found on unblinded review but which were located in distended and
clean colonic segments, lesion height was a significant factor that determined
lesion visibility; five of five lesions 2 mm or higher versus zero of seven
lesions 1 mm or less in height could be visualized (p = 0.001)
(Table 1). Lesion width
(p =1) and histopathology (p = 0.576) were not associated
significantly with lesion visibility (Table
1). In terms of location, three of three lesions located on the
haustral fold and two of nine lesions located on the haustra were visualized
(p = 0.046); however, all the three lesions on the haustral fold were
2 mm in height, whereas the seven haustral lesions not visualized were 1 mm or
less in height (Table 1).
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —52-year-old woman with 8 x 2-mm (width x height) flat
adenocarcinoma confined within mucosa in rectum. Colonoscopy shows profile
view of plaquelike, slightly elevated lesion (arrowheads) in
rectum.
|
|
View larger version (107K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —52-year-old woman with 8 x 2-mm (width x height) flat
adenocarcinoma confined within mucosa in rectum. 2D transverse images of colon
window setting (B: width, 1,500 H; level, -200 H) and soft-tissue
window setting (C: width, 400 H; level, 20 H) of CT colonography show
slightly elevated lesion (arrow, B and C) on haustral
fold of rectum. This lesion was correctly detected on blinded review.
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —52-year-old woman with 8 x 2-mm (width x height) flat
adenocarcinoma confined within mucosa in rectum. 2D transverse images of colon
window setting (B: width, 1,500 H; level, -200 H) and soft-tissue
window setting (C: width, 400 H; level, 20 H) of CT colonography show
slightly elevated lesion (arrow, B and C) on haustral
fold of rectum. This lesion was correctly detected on blinded review.
|
|
View larger version (125K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —52-year-old woman with 8 x 2-mm (width x height) flat
adenocarcinoma confined within mucosa in rectum. 3D endoluminal image of CT
colonography shows plaquelike, slightly elevated lesion (arrowheads)
on haustral fold of rectum. Lesion is more conspicuous on 3D endoluminal view
than on 2D views.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —52-year-old man with 8 x 2-mm flat tubular adenoma in sigmoid
colon. and C, 2D transverse images of colon window setting (B:
width, 1,500 H; level, -200 H) and soft-tissue window setting (C:
width, 400 H; level, 20 H) of contrast-enhanced CT colonography show slightly
elevated lesion (arrow, B and C) in distal sigmoid
colon. Lesion shows moderate contrast enhancement at venous phase (C).
This lesion was correctly detected on blinded review.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —52-year-old man with 8 x 2-mm flat tubular adenoma in sigmoid
colon. 2D transverse images of colon window setting (B: width, 1,500 H;
level, -200 H) and soft-tissue window setting (C: width, 400 H; level,
20 H) of contrast-enhanced CT colonography show slightly elevated lesion
(arrow, B and C) in distal sigmoid colon. Lesion shows
moderate contrast enhancement at venous phase (C). This lesion was
correctly detected on blinded review.
|
|
View larger version (101K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —63-year-old man with 7 x 2-mm flat hyperplastic polyp in
transverse colon. 2D transverse images of colon window setting (B:
width, 1,500 H; level, -200 H) and soft-tissue window setting (C:
width, 400 H; level, 20 H) of CT colonography show slightly elevated lesion
(arrow, B and C) on haustral fold of transverse colon.
This lesion was missed on blinded review because of its low conspicuity but
was found on unblinded review.
|
|
View larger version (127K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —63-year-old man with 7 x 2-mm flat hyperplastic polyp in
transverse colon. 2D transverse images of colon window setting (B:
width, 1,500 H; level, -200 H) and soft-tissue window setting (C:
width, 400 H; level, 20 H) of CT colonography show slightly elevated lesion
(arrow, B and C) on haustral fold of transverse colon.
This lesion was missed on blinded review because of its low conspicuity but
was found on unblinded review.
|
|
View larger version (122K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3D —63-year-old man with 7 x 2-mm flat hyperplastic polyp in
transverse colon. 3D endoluminal image of CT colonography shows plaquelike,
slightly elevated lesion (arrowheads) on haustral fold of transverse
colon.
|
|
Seven false-positive flat lesions were found in six patients, that is, in
the transverse colon (n = 1), sigmoid colon (n = 5), and the
rectum (n = 1). Lesion heights were 2 mm (n =5) and 3 mm
(n = 2), and the widths were from 6 to 10 mm (median, 8 mm). Of
those, four false-positives in the sigmoid colon and one in the rectum were
confirmed to be false-positive by pathologic inspection of the relevant
colonic segments after surgery; however, in two false-positives further
confirmation was unavailable.
Discussion
Considering the design of our study and the small number of lesions
included in it, this study is not appropriate to present the reliable
sensitivity of 16-MDCT colonography for flat colonic lesions. However, despite
the nearly optimal z-axis spatial resolution and the use of IV
contrast, more than 50% of flat lesions are radiographically occult even to
the unblinded reviewer. Lesion height was associated significantly with the
visibility of a flat lesion on CT colonography in our study. Considering that
none of the seven flat lesions l mm or less in height could be detected even
in retrospect despite optimal bowel preparation and distention, it appears
that a flat lesion must be at least 2 mm in height to be detected on 16-MDCT
colonography. Although it may be true that some flat lesions with high-grade
dysplasia or adenocarcinoma are very short in height, it is not yet known how
many flat lesions 1 mm or less in height are significant lesions. The
association between the malignant potential of a flat adenomatous lesion and
the lesion's height has not yet been described, although the malignant
potential is known to increase as the lesion width increases
[2,
3].
The difficulty of diagnosing flat lesions on CT colonography is also
thought to be related to several other factors. Plaquelike morphology may be
easily mistaken for feces. As shown in the three flat lesions of our study
that were confidently distinguished from feces with the help of contrast
enhancement, IV contrast may improve interpreter confidence and lesion
detection [11]. IV contrast
may be used as a problem-solving tool when a potential flat lesion is found
and there is doubt; however, the use of the fecal tagging with an oral
contrast agent [12] may be a
more feasible way to distinguish flat lesions from feces. Because lesions were
detected in this study using 2D search and conspicuity was better with 3D
endoluminal images in two lesions on the haustral fold, sensitivity of CT
colonography may be improved by using a search method of 3D endoluminal
fly-throughs. Our study is limited in that 3D fly-throughs were not used. The
roles of fecal tagging and 3D fly-throughs for detecting flat polyps may be
worth further study. Increasing awareness of the flat lesions and familiarity
with their morphologic features would also be helpful in correctly diagnosing
these lesions.
Regarding the negative impact of missed flat lesions on screening CT
colonography, a recent report
[10] suggested that flat
lesions were not a significant drawback for screening in an asymptomatic
low-risk population. According to some investigators
[13], hyperplastic polyps
accounted for most flat lesions in a low-risk screening population. It is also
not yet known how many flat lesions in a screening population would be 1 mm or
less in height. The scale of the negative impact of missed flat lesions on
screening CT colonography may need further study.
In two false-positive lesions, confirmation was lacking. These lesions
could be false-negative findings of colonoscopy. With regard to the detection
of flat polyps, colonoscopy, as performed with standard technique, is
relatively insensitive. Special techniques such as high-magnification
chromoscopy and dye spraying are necessary to maximize the detection of flat
polyps on colonoscopy [6].
Our study has limitations. First, only disease-positive flat-lesion cases
were included, and the reviewer was requested to report flat lesions rather
than adopting a conventional interpretation of CT colonography. Therefore,
although the reviewer was blinded to any relevant information, there could be
a bias in favor of a suspicion that the patients harbored flat polyps. Second,
as most of our study patients had advanced colon cancer, complete bowel
preparation was difficult and many flat lesions were inevitably unable to be
detected because of poor bowel preparation and/or excessive luminal fluid.
In conclusion, less than 50% of flat lesions in our population could be
visualized by blinded and unblinded review. Visualized flat lesions were 2 mm
or greater in height and 7 mm or greater in diameter. Lesions with a height of
1 mm or less were not seen on CT colonography.
References
- Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual
colonoscopy to screen for colorectal neoplasia in asymptomatic adults.
N Engl J Med 2003;349
: 2191-2200[Abstract/Free Full Text]
- Rembacken BJ, Fujii T, Cairns A, et al. Flat and depressed colonic
neoplasms: a prospective study of 1000 colonoscopies in the UK.
Lancet 2000; 355:1211
-1214[CrossRef][Medline]
- Hurlstone DP, Cross SS, Adam I, et al. A prospective
clinicopathological and endoscopic evaluation of flat and depressed colorectal
lesions in the United Kingdom. Am J Gastroenterol2003; 98:2543
-2549[CrossRef][Medline]
- Wolber RA, Owen DA. Flat adenomas of the colon. Hum
Pathol 1991; 22:70
-74[CrossRef][Medline]
- Teixeira CR, Tanaka S, Haruma K, et al. Flat-elevated colorectal
neoplasms exhibit a high malignant potential. Oncology1996; 53:89
-93[CrossRef][Medline]
- Hurlstone DP, Brown S, Cross SS. The role of flat and depressed
colorectal lesions in colorectal carcinogenesis: new insights from
clinicopathological findings in high-magnification chromoscopic colonoscopy.
Histopathology 2003;43
: 413-426[CrossRef][Medline]
- Schlemper RJ, Itabashi M, Kato Y, et al. Differences in the
diagnostic criteria used by Japanese and Western pathologists to diagnose
colorectal carcinoma. Cancer 1998;82
: 60-69[CrossRef][Medline]
- Gluecker TM, Fletcher JG, Welch TJ, et al. Characterization of
lesions missed on interpretation of CT colonography using a 2D search method.
AJR 2004; 182:881
-889[Abstract/Free Full Text]
- Fidler JL, Johnson CD, MacCarty RL, Welch TJ, Hara AK, Harmsen WS.
Detection of flat lesions in the colon with CT colonography. Abdom
Imaging 2002; 27:292
-300[Medline]
- Pickhardt PJ, Nugent PA, Choi JR, Schindler WR. Flat colorectal
lesions in asymptomatic adults: implications for screening with CT virtual
colonoscopy. AJR 2004;183
: 1343-1347[Abstract/Free Full Text]
- Morrin MM, Farrell RJ, Kruskal JB, Reynolds K, McGee JB,
Raptopoulos V. Utility of intravenously administered contrast material at CT
colonography. Radiology 2000;217
: 765-771[Abstract/Free Full Text]
- Pickhardt PJ, Choi JR. Electronic cleansing and stool tagging in CT
colonography: advantages and pitfalls with primary three-dimensional
evaluation. AJR 2003;181
: 799-805[Free Full Text]
- Pickhardt PJ, Choi JR, Hwang I, Schindler WR. Nonadenomatous polyps
at CT colonography: prevalence, size distribution, and detection rates.
Radiology 2004;232
: 784-790[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
S. S. Lee, S. H. Park, E. K. Choi, S. Y. Kim, M.-J. Kim, K. H. Lee, and Y. H. Kim
Colorectal Polyps on Portal Phase Contrast-Enhanced CT Colonography: Lesion Attenuation and Distinction from Tagged Feces
Am. J. Roentgenol.,
July 1, 2007;
189(1):
35 - 40.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. H. Park, S. S. Lee, E. K. Choi, S. Y. Kim, S.-K. Yang, J. H. Kim, and H. K. Ha
Flat Colorectal Neoplasms: Definition, Importance, and Visualization on CT Colonography
Am. J. Roentgenol.,
April 1, 2007;
188(4):
953 - 959.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
