Journal of Clinical Imaging 29 (2005) 200 – 206
CT colonography features of sigmoid diverticular disease Marc J. Golluba,*, Sujata Jhaveria, Eric Schwartza, Howard Feldermana, Cathleen Coopera, Arnold J. Markowitzb, Robert C. Kurtzb, Howard Thalerc a
Department of Radiology, Memorial Sloan-Kettering Cancer Center, Room C276F, 1275 York Avenue, New York, NY 10021, United States b Department of Medicine (Gastroenterology and Nutrition Service), Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, United States c Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021, United States Received 1 May 2004; received in revised form 10 May 2004
Abstract Objective: The objective of this study is to assess the sigmoid distensibility during CT colonography (CTC) in patients with diverticular disease. Methods: Consecutive patients without a history of pelvic radiation or neoplasms underwent 150 CTC. Three radiologists in consensus evaluated axial images for colonic distention, luminal diameters (mm), diverticula, and muscular thickening. Results: The minimum colon diameter in patients with muscular thickening was significantly smaller, irrespective of the presence of diverticula ( P = .009). Conclusion: Muscular thickening with diverticular disease was associated with significantly less sigmoid colon distension. D 2005 Elsevier Inc. All rights reserved. Keywords: CT colonography; Diverticula; Distensibility; Quality
1. Introduction Diverticular disease of the colon is common in the United States and is estimated to affect 33–50% of adults by the age of 50 [1]. It most frequently occurs in the sigmoid colon and, when present in moderate to advanced severity, can interfere with the detection of colonic neoplasm. Due to the anatomic distortion from either the associated muscular prominence and spasm or the numerous diverticular outpouchings, difficulty in accurately examining a diseased sigmoid colon may be encountered at the time of colonoscopy [2], barium enema [3], and even during surgery [4]. This entity thus presents a formidable obstacle to the detection of colon cancer, in addition to the associated complications of diverticulitis, obstruction, abscess formation, and perforation.
* Corresponding author. Tel.: +1 212 639 2183; fax: +1 212 794 4010. E-mail address:
[email protected] (M.J. Gollub). 0899-7071/05/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.clinimag.2004.07.004
CT colonography (CTC), introduced in 1994, continues to undergo intensive investigation and has moved progressively into the clinical setting. Its appropriateness as a screening test for colorectal cancer has not yet been established. However, in multiple trials of mixed populations—predominantly consisting of symptomatic or surveillance-type patients—the sensitivity of CTC was in the range of 90–95% per polyp and 95–100% per patient for polyps equal to or greater than 10 mm [5]. In our institution, CTC is now used instead of barium enema, when feasible, as the standard of care for patients who have undergone incomplete colonoscopy. Potential indications for CTC in our patients include the evaluation of the colon in patients who have contraindications to colonoscopy, such as poor pulmonary capacity, and those who cannot stop or need to remain on anticoagulation. Finally, some patients request to have CTC instead of colonoscopy out of personal preference. Experience with CTC in patients who have had prior surgery or a combination of surgery and radiation has been previously published. CTC in these patients can be
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
201
completed without any compromise in quality [6]. In our desire to promote efficient patient triage, we set out to formally assess patients who undergo CTC after incomplete colonoscopy, in whom we have discovered diverticular disease involving the sigmoid colon. In these patients, we hypothesized that the quality of CTC would be compromised by the inability to achieve good distension secondary to the muscular pathology associated with diverticular disease.
2. Materials and methods 2.1. Patient and imaging data Between July 1999 and February 2002, 150 CTC examinations were performed in 147 consecutive adult patients at our cancer center, after the administration of a standard colon cleansing preparation (phosphosoda, 24-h Fleet 1; Fleet Pharmaceuticals, Lynchburg, VA). The indication for CTC was colorectal cancer surveillance (n = 43), screening (n = 28), and evaluation of symptoms (e.g., rectal bleeding; n = 28). Four CT colonographies were performed on a singledetector helical CT scanner (HiSpeed Advantage; General Electric Medical Systems, Milwaukee, WI; table speed, 10 mm/rot, 1.0-s scan rotation speed; pitch, 2; slice thickness, 5.0 mm; overlap interval, 2.5 mm) before the availability of our multidetector scanner (LightSpeed QX/i; General Electric Medical Systems), on which the remaining 146 CT colonographies were performed [table speed, 11.25 mm/rot, 0.8-s scan rotation speed; pitch 0.75; slice thickness, 3.75 mm (effective, 3.9 mm); overlap interval, 1.95 mm]. All patients were scanned in the prone and then the supine positions, and all patients received 0.9 mg iv of glucagon hydrochloride (Eli Lilly, Indianapolis, IN) immediately
Fig. 2. Supine 2D axial CT image from CTC exam in a patient with 11 or more (not all shown) diverticula (arrowheads) and muscular thickening (myochosis, arrows).
before air insufflation. A fellow or attending radiologist inserted a 16-gauge flexible rubber catheter 15 cm into the rectum, followed by the insufflation of air using a blue manual compression bulb to patient tolerance (c30 – 60 pumps, total volume 1 – 2 l). Patient incontinence occasionally required the use of a balloon tip catheter with 10 cc balloon insufflation. Initial and repeat scout images were obtained to assess the quality of air distention. Insufficient colonic distention led to additional attempts at insufflation to patient tolerance. Our Institutional Review Board deemed this retrospective study exempt from the requirement of patient informed consent. 2.2. Interpretation All examinations had been previously officially interpreted. Unaware of patient medical and surgical history,
Fig. 1. Supine 2D axial CT image from CTC exam in a patient with three sigmoid diverticula (arrow).
Fig. 3. Prone 2D axial image from CTC exam in a patient with muscular wall thickening (arrow) in association with diverticular disease (not shown) of the colon.
202
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
Table 1a Mean maximum and minimum sigmoid colon diameters (mm; means are weighted) Diverticula (+/ ) Absent Maximum diam prone Maximum diam supine Maximum diam (P + S mean) Minimum diam prone Minimum diam supine Minimum diam (P + S mean)
Mch (+/ ) Present
Absent
Present
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
36.45 37.07 36.76 10.22 6.24 8.23
12.81 14.08 12.49 9.54 9.93 8.04
31.57 33.02 32.30 9.57 3.75 6.66
9.27 9.90 8.36 8.93 7.40 7.15
35.92 35.65 35.78 11.15 6.38 8.76
11.39 12.63 11.33 9.57 9.83 8.20
29.44 34.16 31.80 6.32 1.44 3.88
10.97 12.29 9.66 7.18 3.66 4.06
Diam = diameter; P = prone; S = supine; Mch = muscular hypertrophy.
three board-certified radiologists, with experience interpreting CTC examinations (ranging from 5 to 80 colonoscopically proven cases), retrospectively performed consensus analyses of the two-dimensional (2D) images on a picture archiving and communications system (General Electric PACS; General Electric Medical Systems). Prone and supine 2D data sets were evaluated for the degree of colonic air distention and degree of colonic fluid or stool retention. Two of three radiologists simultaneously evaluated the cases, while the third recorded the data. The preparation of the colon was judged to be bpoorQ (defined as greater than 10 particles of stool or two or greater segments where fluid amounted to greater than one half the luminal diameter or greater than five segments with fluid in less than one half the luminal diameter), bgoodQ (5 to 10 particles of stool or one segment with greater than one half the luminal diameter filled with fluid or four or five segments with less than one half the luminal diameter filled with fluid), or bexcellentQ (fewer than 5 particles of stool, no segment with greater than one half the lumen filled with fluid, or three or less segments with less than one half the lumen filled with fluid). Distention was subjectively rated 1– 4 as follows: 1 = no distention; 2 = poor distention (minimal wall separation and inadequate fold or mucosal flattening); 3 = good
distention (most or all mucosal redundancy effaced); and 4 = optimal distention. Ratings were applied to eight colonic segments, including the cecum, ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid, and rectum. The rating of the least distended area of each segment was recorded as the overall rating of that segment. The evaluation of the sigmoid colon also included measurements of the diameter in the least and in the most distended loop from the inner-to-inner wall, in millimeters, on both prone and supine images. The minimum value was 0 mm at points of collapse. The presence of diverticula in the sigmoid was categorized into none, those containing 1–10 diverticula, or more than 11 diverticula (Figs. 1 and 2). The presence of muscular hypertrophy, also known as myochosis (Mch), was defined as focal areas of abnormal thickening of interhaustral folds, often jutting into the lumen and giving a bsaw-toothedQ appearance to the lumen (see Figs. 2 and 3). These areas with a primary wall abnormality were distinguished from loops of bowel adjacent or nearby that had a similar amount of distension but without disproportionate wall thickening, indicating pseudothickening. Colonoscopy reports were reviewed for the (1) purpose of the exam (screening, surveillance or symptoms); (2) cause
Table 1b Mean of maximum and minimum diameters (mm) of sigmoid colon (raw data) Diverticula (+/ ) Absent
Present
Mch (+/ )
Mch (+/ )
Absent Maximum diam prone Maximum diam supine Maximum diam (P + S mean) Minimum diam prone Minimum diam supine Minimum diam (P + S mean)
Present
Absent
Present
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
36.82 36.34 36.58 10.38 6.66 8.52
12.27 13.85 12.33 9.55 10.25 8.28
32.80 44.40 38.60 8.60 2.00 5.30
18.78 15.88 15.38 10.38 4.47 4.60
34.04 34.21 34.13 12.75 5.79 9.27
9.24 9.73 8.91 9.61 9.08 8.18
28.60 31.60 30.10 5.75 1.30 3.53
8.60 0.16 7.26 6.38 3.56 3.97
Diam = diameter; P = prone; S = supine; Mch = muscular hypertrophy.
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
203
Table 2a Mean qualitative ratings of the sigmoid and whole colon (1 = no distension, 2 = minimal distension, 3 = good distension, 4 = optimal; means are weighted) Diverticula (+/ ) Absent Prone sigmoid Supine sigmoid P + S mean sigmoid Prone whole colon Supine whole colon P + S mean whole colon
Mch (+/ ) Present
Absent
Present
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
2.49 1.82 2.15 3.29 3.10 3.19
1.22 1.17 0.97 0.53 0.50 0.43
2.16 1.53 1.85 3.27 3.06 3.16
0.97 0.93 0.83 0.48 0.51 0.42
2.49 1.86 2.18 3.31 3.08 3.19
1.13 1.18 0.99 0.50 0.52 0.44
1.92 1.20 1.56 3.18 3.08 3.13
1.00 0.41 0.46 0.51 0.48 0.38
P = prone; S = supine; Mch = muscular hypertrophy.
of the incomplete exam (diverticular disease, stricture/mass, surgery/radiation, tortuosity/redundancy/looping, patient unable to tolerate/pain, poor prep or unknown); and (3) presence of sigmoid diverticular disease of mild, moderate, or severe degree (colonoscopist’s subjective description). Both parametric and nonparametric statistical analyses were employed as appropriate and detailed below. Statistical significance was considered present at a P value of .05 or less. All patients in whom a known or subsequently proven malignancy was the cause of colonic narrowing and in whom there was a documented history of pelvic radiation were excluded from analysis for this study. Ninety-seven patients who underwent 99 exams were thus eligible. It should be noted that the sensitivity and specificity for the detection of polyps could not be determined in these patients due to the infrequent availability of colonoscopic or surgical follow-up.
3. Results Ninety-seven adult patients, with mean age 62 years (range 30–94 years), underwent 99 CTC examinations. There were 32 males and 65 females. Incomplete colonoscopy within one year preceded CTC in 78 exams. The reason that the colonoscopy was unable to be completed was cited by the colonoscopist as tortuosity (n = 27),
inability to tolerate procedure (n = 16), unknown (n = 8), bowel fixation secondary to a history of surgery and/or radiation leading to fixation (n = 8), diverticular disease (n = 8), poor prep (n = 5), stricture or mass (n = 3), and one each of the following: reaction to the sedative, respiratory problems, and ventral hernia. Patients had a history of neoplasm in this selected oncologic patient group as follows: none (n = 30), colon (n = 28), breast (n = 10), ovarian (n = 5), bladder (n = 3), two patients each with unknown primary neoplasm, prostate cancer, nonHodgkin’s lymphoma, and cholangiocarcinoma, and one patient each with melanoma, thyroid cancer, stomach cancer, tongue cancer, skin cancer, rectal cancer, parathyroid cancer, pancreatic cancer, osteosarcoma, lymphoma, pituitary adenoma, endometrial cancer, and appendix cancer. The colonic cleansing was judged to be excellent in 48 exams, good in 29 exams, and poor in 22 exams. Sigmoid diverticular disease was detected at colonoscopy in 27 patients (27%) and was described as bmildQ (n = 11), bmoderateQ (n = 4), and bsevereQ (n = 12). The minimum sigmoid diameter (average of prone and supine positions) for patients without myochosis, irrespective of the presence of diverticula (note: all averages are weighted to reflect the differing size of groups being compared), was 8.76 mm. This was statistically significantly greater than the minimum sigmoid diameter (average of prone and supine positions) for patients with Mch, 3.88 mm, irrespective of the presence of diverticula ( P = .009). This
Table 2b Mean of qualitative ratings of sigmoid and whole colon (raw data) Diverticula (+/ ) Absent
Present
Mch (+/ )
Mch (+/ )
Absent Prone sigmoid Supine sigmoid P + S mean sigmoid Prone whole colon Supine whole colon P + S mean whole colon
Present
Absent
Present
Mean
S.D.
Mean
S.D.
Mean
S.D.
Mean
S.D.
2.48 1.884 2.18 3.31 3.09 3.20
1.20 1.21 1.00 0.51 0.51 0.44
2.60 1.20 1.90 3.08 3.18 3.13
1.52 0.45 0.65 0.72 0.51 0.42
2.52 1.83 2.17 3.32 3.06 3.19
0.99 1.15 0.98 0.50 0.55 0.46
1.75 1.20 1.48 3.21 3.05 3.13
0.79 0.41 0.38 0.47 0.48 0.38
P = prone; S = supine; Mch = muscular hypertrophy.
204
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
difference was significant for the prone and supine positions as well ( P = .027 and .025, respectively; Wilcoxon sign rank test; Table 1a). The minimum sigmoid diameter (average of prone and supine data sets) for patients with diverticula, 6.6 mm, did not differ significantly from the averaged minimum sigmoid diameter of the prone and supine data sets of patients without diverticula, 8.23 mm (Table 1b). The qualitative rating (average of supine and prone data) of the distention of the sigmoid colon in patients who did have myochosis (1.56 F 0.46) was significantly lower than for those without myochosis (2.18 F 0.99, P b.01 Wilcoxon test). No significant difference was present in the qualitative rating (average of prone and supine data) for the whole colon between these groups. (3.13 F 0.38 Mch vs. 3.19 F 0.44, P = NS; Table 2a). No significant difference was present in maximum sigmoid diameter either, irrespective of the presence or absence of diverticula or myochosis (Table 2b). The quality of the colonoscopy cleansing preparation did not differ between patients with or without diverticula or with or without myochosis. Neither the presence or absence of diverticula nor the number, when categorized as 1–10 or 11 or more, correlated with the minimum sigmoid diameter. However, a trend was present, which positively related the presence or absence of diverticula and their increased number (i.e., none vs. 1–10 vs. 11 or more) with a greater likelihood of myochosis ( P b.01, chi square for contingency tables).
4. Discussion Diverticular disease of the colon is a very common affliction, especially in the United States. It is estimated that more than one third of individuals over the age of 50 years have some evidence of diverticular disease of varying severity. Using anatomical and histological correlation, the true nature of the disease has been well characterized and found to represent primarily a muscular disorder with resultant formation of diverticular outpouchings. The muscular disorder is characterized by the deposition of elastin in longitudinal muscle fibers (teniae coli), resulting in the foreshortening of the colon. This causes a bunching up of the circular muscle fibers. No true hyperplasia or hypertrophy is found on histologic examination. This has been termed bmyochosisQ and is commonly referred to by the misnomer bcircular muscle hypertrophyQ. At points of weakness in the wall adjacent to the perforating vessels, pseudodiverticula form because of the herniation of the mucosa. The muscular thickening also causes the bunching up of the mucosa within the lumen in a redundant pattern, giving a bconcertina-likeQ appearance. The disease is most common in the sigmoid colon because it is the narrowest segment and the area of greatest intersegmental peristaltic pressure. In a small percentage of cases, diverticular disease can occur without
diverticula, confirming the theory that the disease is primarily one of muscle with incidental formation of diverticula. In lesser degrees of inflammation and severity, the muscle abnormality can presumably resolve, leaving only sigmoid diverticula without noticeable muscular pathology, whereas in other cases, severe muscle abnormality is seen in the absence of diverticula [7]. Although substantial morbidity is associated with diverticular disease, including obstruction and inflammation, perhaps more importantly, it presents a formidable obstacle to the detection of concurrent neoplasm [8,9]. Although the distribution of cancer in the colon has changed, with a greater percentage of tumors being discovered more proximally, the segment harboring the highest number of cancers is still the sigmoid colon, the same area of greatest frequency of occurrence of diverticula [10]. Unfortunately, these entities can present with similar symptoms, such as bleeding, perforation, and sepsis, or obstruction. This has led to a large body of literature examining the radiologist’s ability to distinguish, e.g., perforated diverticulitis from perforated carcinoma, realizing the importance of overlooking a cancer diagnosis [11]. In more advanced degrees of diverticular disease, where there is significant myochosis, radiographic methods such as the contrast enema and CT are limited because appearances may mimic malignancy or, at least, limit the detection of concurrent malignancy [3,11]. Similarly, more advanced degrees of diverticular disease may impede the colonoscope from both examining the entire sigmoid colon and reaching the more proximal colon [2]. Even simpler degrees of diverticular disease have been shown to interfere with the diagnosis of polypoid cancer on barium examinations in over half of cases [12]. In fact, intraoperatively, the surgeon may not recognize a cancer within a significantly inflamed indurated sigmoid colon with surrounding abscesses or mesenteric reaction [4]. Although we were unable to compare the effects of diverticular disease on the detection of polyps due to too few colonoscopic follow up exams, our data indicate significant limitation in the visualization of the sigmoid colon in these patients because of the inadequate luminal distension achieved as compared with normals. Inadequate luminal distension is one of the major limitations in the detection of polyps. Glucagon was used in all patients, a policy we have abandoned because of evidence in the literature that it makes no significant difference. Further investigation to compare sigmoid distensibility in this subset of patients with and without the use of glucagon may be a worthwhile avenue of research. In patients with muscular hypertrophy associated with diverticular disease, the minimum sigmoid colon diameter of 3.88 mm was significantly smaller than for patients without muscular hypertrophy, whether averaging prone and supine positions or considering them separately. The difference was even appreciable qualitatively, where the average
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
score was 1.56 for those with muscular hypertrophy compared with 2.18 for those without. Interestingly, the presence of simple diverticula did not alter the minimum diameter if muscular hypertrophy was absent. But indirectly, the increasing numbers of diverticula were correlated with the increasing likelihood of muscular hypertrophy, and this, in turn, was a risk for poor luminal distention. To assess whether the presence of myochosis with diverticular disease interferes with the normal evacuation of stool and fluid, perhaps requiring more rigorous cleansing for colon screening exams in those with diverticulosis, we compared the cleanliness of the colon between those with and without myochosis and found no difference. Diverticular disease can occasionally be the cause for an incomplete colonoscopy. In this scenario, the muscular rigidity and angulation of the sigmoid may prevent the colonoscope from negotiating turns and advancing into the more proximal colon. Alternatively, numerous widemouthed diverticula are occasionally encountered, impeding advancement into the true lumen and contributing to the overall duration and risk of the procedure [13]. The more recently developed CTC is touted to have many advantages in addition to a high sensitivity for the detection of polyps measuring 10 millimeters or greater. Among these advantages are the detection of incidental extracolonic abnormalities because of the cross-sectional depiction of anatomy, the ability to examine those portions of the colon not reached during an incomplete colonoscopy, and the visualization of the actual wall of the colon and associated pericolonic fat and lymph nodes assisting in the staging of neoplasm. We feel that the visualization of the actual wall of the colon may actually be disadvantageous in the setting of myochosis. Specifically, the appearance may be confusing and cause difficulty in distinguishing between myochosis, tumor, or underdistension in the sigmoid colon. Primary adenocarcinoma of the colon typically begins with a premalignant mucosal lesion (adenomatous polyp) and extends along the mucosa, with various combinations of elevated growth and ulceration at the same time, as it may grow submucosally. Analysis of the mucosa is key because this is where a primary epithelial tumor begins and manifests on all imaging examinations used to date. For this reason, the double contrast barium enema, which directly depicts the mucosa and avoids visualizing the actual bowel wall (e.g., confusing wall appearances), should provide an advantage in this clinical scenario. Furthermore, the ability to interactively control the degree of distension during the barium examination, either with air or additional barium, should overcome the tendency toward undersdistension noted on CTC exams. Results from numerous published studies reveal that nearly one half of false positive [14] and numerous false negative results can be attributed to segments of the sigmoid colon of patients with diverticular disease [15,16]. These limitations support our findings.
205
A limitation of our study is the lack of determination of a difference in actual detection rates for polyps in the sigmoid colon between those with and without diverticular disease. Nonetheless, undersdistension is a key limitation in CTC for the detection of polyps and neoplasm. As such, clinicians should be aware that in this setting, CTC may not provide further assistance but rather delay the collection of conclusive data on those sigmoid colons affected by diverticulosis. Further study is warranted to determine if the limitations in the quality of visualization of these segments affected by myochosis result in limitations in diagnostic accuracy. Further investigation to confirm whether barium enema is superior in this setting would be ideal and will require a prospective study wherein a patient receives all three examinations within a short time period, a trying, but not impossible request to make of patients in a research endeavor. We have demonstrated that the presence of diverticular disease with associated muscular abnormality (myochosis) compromises the quality of distension of the sigmoid colon during CTC. Poor distension is a known pitfall in the evaluation for polyps and cancer. In addition, the appearance of this muscle thickening leads to a confusing and often worrisome appearance, which leads to further work-up. For this reason, we recommend alternative methods, such as barium enema, when patients require further evaluation of the sigmoid colon after incomplete colonoscopy in the setting of sigmoid diverticulosis. If the sigmoid colon was adequately assessed, but the muscular rigidity precluded advancement of the endoscope to the more proximal colon, CTC has been well established as superior to barium enema in the examination of the remainder of the colon [17].
References [1] Balthazar EJ. Diverticular disease. In: Gore RM, Levine MS, Laufer I, editors. Textbook of gastrointestinal radiology. Philadelphia7 WB Saunders, 1994. p. 1072. [2] Glerum J, Agenant D, Tytgat GN. Value of coloscopy in the detection of sigmoid malignancy in patients with diverticular disease. Endoscopy 1977;9:228 – 30. [3] Baker SR, Alterman DD. False-negative barium enema in patients with sigmoid cancer and coexistent diverticula. Gastrointest Radiol 1985;10:171 – 3. [4] Corman ML. Colon and rectal surgery, 4th ed. Philadelphia7 Lippincot-Raven, 1998. pp. 693 – 5. [5] Yee J, Akerkar GA, Hung RK, Steinauer-Gebauer AM, Wall SD, McQuaid KR. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219: 685 – 92. [6] Gollub MJ, Flaherty F. Barium enema following incomplete colonoscopy. Clin Imaging 1999;23:367 – 74. [7] Williams I. Diverticular disease of the colon without diverticula. Radiology 1967;89:401 – 12. [8] Whiteway J, Morson BC. Pathology of the ageing—diverticular disease. Clin Gastroenterol 1985;14:829 – 46. [9] Morson BC. The muscle abnormality in diverticular disease of the sigmoid colon. Br J Radiol 1963;36:385 – 92.
206
M.J. Gollub et al. / Journal of Clinical Imaging 29 (2005) 200–206
[10] Thoeni RF, Laufer I. Polyps and cancer. In: Gore RM, Levine MS, Laufer I, editors. Textbook of gastrointestinal radiology. Philadelphia7 WB Saunders, 1994. pp. 1160 – 5. [11] Chintapalli KN, Chopra S, Ghiatas AA, Esola CC, Fields SF, Dodd GD III. Diverticulitis versus colon cancer: differentiation with helical CT findings. Radiology 1999;210:429 – 35. [12] Schnyder P, Moss AA, Thoeni RF, Margulis AR. A double-blind study of radiologic accuracy in diverticulitis, diverticulosis, and carcinoma of the sigmoid colon. J Clin Gastroenterol 1979;1: 55 – 66. [13] Webb WA. Colonoscoping the ddifficultT colon. Am Surg 1991;57: 178 – 82. [14] Fenlon HM, Nunes DP, Schroy PC, Barish MA, Clarke PD, Ferrucci JT. A comparison of virtual and conventional colonoscopy
for the detection of colorectal polyps. N Engl J Med 1999;341: 1496 – 503. [15] Macari M, Bini EJ, Xue X, Milano A, Katz SS, Resnick D, Chandarana H, Krinsky G, Klingenbeck K, Marshall CH, Megibow AJ. Colorectal neoplasms: prospective comparison of thin-section low-dose multidetector row CT colonography and conventional colonoscopy for detection. Radiology 2002;224:383 – 92. [16] Neri E, Giusti P, Battolla L, Vagli P, Boraschi P, Lencioni R, Caramella D, Bartolozzi C. Colorectal cancer: role of CT colonography in preoperative evaluation after incomplete colonoscopy. Radiology 2002; 223:615 – 9. [17] Macari M, Berman P, Dicker M, Milano A, Megibow AJ. Useful of CT colonography in patients with incomplete colonoscopy. AJR 1999;173:561 – 4.
