Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2442060837
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Copel, L.
Right arrow Articles by Morrin, M. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Copel, L.
Right arrow Articles by Morrin, M. M.
(Radiology 2007;244:471-478.)
© RSNA, 2007

Gastrointestinal Imaging

CT Colonography in 546 Patients with Incomplete Colonoscopy1

Laurian Copel, MD, Jacob Sosna, MD, Jonathan B. Kruskal, MD, PhD, Vassilios Raptopoulos, MD, Richard J. Farrell, MD, and Martina M. Morrin, MD

1 From the Department of Radiology, Assaf-Harofeh Medical Center, Sackler School of Medicine, Zerifin, Israel 70300 (L.C.); Department of Radiology, Hadassah Hebrew University Medical Center, Jerusalem, Israel (J.S.); and Departments of Radiology (J.S., J.B.K., V.R., M.M.M.) and Gastroenterology (R.J.F.), Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass. Received May 13, 2006; revision requested July 12; revision received October 23; accepted November 21; final version accepted January 8, 2007. Address correspondence to L.C. (e-mail: lcopel{at}gmail.com).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Purpose: To retrospectively evaluate the positive predictive value (PPV) of computed tomographic (CT) colonography performed in patients who were referred for further examination after incomplete colonoscopy.

Materials and Methods: This HIPAA-compliant retrospective study was approved by the institutional review board; informed consent was waived. We identified 546 consecutive patients (mean age, 64.1 years; 401 [73.4%] women) who underwent CT colonography after incomplete colonoscopy between November 1999 and December 2002. A retrospective chart review was performed if CT colonography depicted endoscopically nonvisualized lesions 6 mm or greater in diameter. Repeat colonoscopy rate, endoluminal findings, and PPV of CT colonography were determined. Subsequent colonoscopic findings were used as the reference standard.

Results: In 72 (13.2%) patients, CT colonography depicted 88 endoscopically nonvisualized lesions 6 mm or greater. Of 11 patients reported to have 12 masses (≥20 mm), at subsequent colonoscopy, one patient had no mass. Eighteen patients had 23 large (10–19-mm) polyps that they were suspected of having, and 47 patients had 53 medium (6–9-mm) polyps that they were suspected of having. At a median follow-up of 31 months (range, 6–42 months), 45 (63%) of 72 patients underwent follow-up colonoscopy because of their CT colonographic findings. Rates of repeat colonoscopy for masses, large polyps, and medium polyps were 100%, 94%, and 45%, respectively. Per-patient and per-lesion PPVs of CT colonography for masses, large polyps, and medium polyps were 90.9% and 91.7%, 64.7% and 70%, and 33.3% and 30.4%, respectively.

Conclusion: CT colonography has the potential to become an accepted technique for evaluation of the nonvisualized part of the colon after incomplete colonoscopy, and it can increase the diagnostic yield of masses and clinically important polyps in this part of the colon.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Colonoscopy is currently the investigation of choice for colorectal cancer screening and investigation of occult anemia and lower gastrointestinal tract symptoms, including a change in bowel habits and bleeding through the rectum (1–3). However, up to 10% of colonoscopic examinations are technically difficult even for experienced colonoscopists (4,5). In addition to poor bowel preparation, an experienced colonoscopist may be unable to complete the colonoscopy and intubate the cecal pole for a variety of reasons (redundant colon; colonic spasm; marked diverticulosis; obstructing masses or strictures; and angulation or fixation of colonic loops, most commonly due to previous pelvic surgery). The reported rates of incomplete colonoscopy from various U.S. and European series over the past 15 years range from 4% to 25% (4–7).

When colonoscopy is incomplete, double-contrast barium enema examination has been performed to complete the inspection of the nonvisualized part of the colon (8). However, double-contrast barium enema examination can be associated with considerable patient discomfort and may not be tolerated on the same day as colonoscopy (9–11). Furthermore, data from the National Polyp Study Work Group (12), in which comparison was made between paired colonoscopy and double-contrast barium enema examination, indicated that the latter depicted only 48% of adenomas greater than 10 mm in diameter, 53% of adenomas 6–9 mm in diameter, and 32% of adenomas 5 mm or smaller in diameter.

Computed tomographic (CT) colonography is a rapidly evolving technique for the detection of colorectal polyps and cancers. Its sensitivity is similar to (13) or better than (14) that of double-contrast barium enema examination for detection of colonic cancer and polyps. Several centers have shown that CT colonography has accuracy similar to that of conventional colonoscopy both in high-risk groups (15–18) and in a low-prevalence screening population (19). Although concerns about performance consistency, technical variability, and clinical implementation need to be resolved before CT colonography can be advocated for widespread colorectal cancer screening, even gastroenterologists agree that CT colonography is a reasonable alternative to double-contrast barium enema examination in patients with incomplete colonoscopy (20,21). In 1999, Morrin and colleagues (22) reported the utility of CT colonography in patients with incomplete colonoscopy; since then, other centers have reported their experience in several small series of patients (23–25). All these studies highlighted the usefulness of CT colonography in the detection of clinically important endoluminal lesions in the part of the colon that was not inspected with colonoscopy. The purpose of our study, therefore, was to retrospectively evaluate the positive predictive value (PPV) of CT colonography in patients who were referred for further examination after incomplete colonoscopy.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Patients
This Health Insurance Portability and Accountability Act–compliant retrospective study was approved by the institutional review board; informed consent was waived. Patient confidentiality was maintained. To overcome the issue of a sedated patient's ability to consent to CT colonography after incomplete colonoscopy, the text of our standard colonoscopy consent form was modified to include referral for CT colonography in the event of an adequately prepared incomplete colonoscopy. A review of the computerized history information system database at our university hospital revealed that 15 600 colonoscopic examinations and 843 CT colonographic examinations were performed between November 1999 and December 2002. During this period, 546 (64.8%) of 843 CT colonographic examinations were performed because of incomplete colonoscopy (3.5% of all colonoscopic examinations). All 546 patients were referred to the Department of Radiology at Beth Israel Deaconess Medical Center, Boston, Mass, by 19 gastroenterologists (all endoscopists with at least 5 years of experience performing colonoscopy—ie, all had performed >1300 colonoscopic examinations at the time of the study) who work in the same institution. The mean age of the 546 patients was 64.1 years (range, 39–88 years), and 401 (73.4%) patients were female. The main reason for incomplete colonoscopy was redundant or tortuous loops (39.9%) (Table 1), and the most frequent location of incomplete colonoscopy was the sigmoid colon (23.8%) (Table 2). CT colonography was not performed in patients whose incomplete colonoscopy was due to poor bowel preparation.


View this table:
[in this window]
[in a new window]

  		Table 1. Reasons for Incomplete Colonoscopy in 546 Patients

 

View this table:
[in this window]
[in a new window]

  		Table 2. Locations of Termination of Incomplete Colonoscopy in 546 Patients

 
Of 546 patients, 492 (90.1%) were at high risk for developing colorectal cancer, whereas 54 patients (9.9%) were considered a low-risk screening population. High-risk patients were defined as symptomatic or asymptomatic patients who had a history of colorectal neoplasia; were under surveillance because of family history of colorectal cancer; or recently had positive findings at a previous screening test, such as a fecal occult blood test or flexible sigmoidoscopy. CT colonography was not performed in patients who were suspected of having or had a previous diagnosis of familial polyposis syndromes or inflammatory bowel disease. CT colonography was performed a mean of 4.2 hours after the incomplete colonoscopy, with 518 (94.9%) examinations performed on the same date as the colonoscopy (overall range, 1.5 hours to 28 days).

Patient Preparation
All patients underwent standard bowel preparation 24 hours before colonoscopy and subsequent CT colonography. Most patients (n = 456) received two 45-mL doses of sodium phosphate (Fleet Prep Kit 1; Fleet Pharmaceutical, Lynchburg, Va) in addition to 10 mg of bisacodyl. Patients with chronic renal failure or congestive heart failure (n = 90) received a 4-L oral cathartic polyethylene glycol preparation (GoLytely; Braintree Laboratories, Braintree, Mass).

Acquisition of Data from CT Colonography
All CT scans were obtained by using multidetector CT scanners (GE Medical Systems, Milwaukee, Wisc): GE Lightspeed QX/i four-section scanner (n = 512) or GE Lightspeed Ultra eight-section scanner (n = 34). The scans were obtained with the following detector configurations: four detector rows with 2.5-mm section thickness (n = 162), four detector rows with 1.25-mm section thickness (n = 350), or eight detector rows with 1.25-mm section thickness (n = 34), with a 7.5–13.5-mm table speed per gantry rotation by using a high-speed mode with gantry rotation time of 0.5–0.8 seconds. The tube current was set to 100–200 mA and 120 kVp.

Patients were placed in the right decubitus position, and room air was insufflated through a 12-F soft-tipped rectal catheter. Patient tolerance was monitored, and two scout images (anteroposterior and lateral) were initially acquired with the patient in the supine position to confirm adequate colonic distention. This was done to facilitate full evaluation of the circumferential colonic wall (air distention of all the segments of the colon was assessed on a qualitative basis). Data acquisition of the entire colon was performed in one breath hold. The patient was then placed in the prone position, and a second set of scout images (anteroposterior and lateral) was obtained to confirm adequate colonic distention; additional air was insufflated if required. A second single–breath-hold data acquisition of the entire colon was performed. One hundred fifty milliliters of ioversal 68% (Optiray 320; Mallinckrodt Medical, St Louis, Mo) that contained 320 mg of iodine per milliliter was intravenously administered to 228 (41.8%) patients. Criteria for injection of intravenous contrast material included cases in which the fluid component of air-fluid levels was greater than 50% of the lumen cross section (n = 81; 35.5%), cases in which staging of suspected colorectal tumors was performed (n = 56; 24.6%), or cases in which the contrast agent was administered during parts of the study period as part of our routine protocol for CT colonography (n = 91; 39.9%). The determination to inject contrast material was made by the radiologist who monitored the examination. A mechanical injector (Medrad, Pittsburgh, Pa) was used for intravenous injection via an 18- to 21-gauge antecubital line at a rate of 3 mL/sec. When the intravenous contrast agent was injected, the scan was obtained 60 seconds after the initiation of the injection, the tube current was increased to 240—320 mA, with no change in the other scanning parameters.

Scan Interpretation
After the performance of each study, the raw data were sent to a commercial workstation (Advantage Windows, version 3.1 or 4.1; GE Medical Systems) for clinical interpretation. The reading technique was based on a cine mode assessment of all the two-dimensional transverse images that were acquired in the supine and the prone positions. Multiplanar reformations and three-dimensional endoluminal navigation were reserved for problem solving. All scans from examinations were read first by one of 14 abdominal imaging fellows (L.C., J.S.) with different experience in interpretation of CT colonographic scans (all fellows initially read 10–25 CT colonographic scans as part of basic training). Then, all scans from examinations were read by one of three board-certified radiologists experienced in interpretation of CT colonographic findings (M.M.M., J.B.K., and V.R.), who had interpreted scans from 720, 350, and 250 examinations, respectively. These three radiologists were blinded to the results of the first reading by the abdominal imaging fellow. In all cases, a consensus based on agreement was reached among the readers.

Data Collection
Two radiologists (L.C. and J.S.) reviewed all CT colonographic and colonoscopic reports. Images were reviewed to identify all endoscopically nonvisualized endoluminal lesions 6 mm or larger in diameter in the part of the colon that was not inspected with colonoscopy. A segmental comparison between CT colonographic findings and the colonoscopic report was performed to confirm that a lesion identified at CT colonography was not in a location reached by colonoscopy. Endoluminal lesions not visualized at endoscopy were categorized according to size: A lesion was defined as a mass if its diameter at CT colonography was 20 mm or larger, as a large polyp if its diameter was 10 mm or larger but smaller than 20 mm, and as a medium polyp if its diameter was 6–9 mm. Diminutive lesions (≤5 mm) were not included in this study. Lesions were measured at colonoscopy and postsurgical resection by using an open biopsy forceps and ruler, respectively. Subsequent colonoscopic and surgical records to June 2003 were evaluated to determine the rate of repeat colonoscopy and subsequent findings and the PPV of CT colonography. All decisions were made in consensus.

By using subsequent colonoscopic and postcolonoscopic surgical findings (in 45 patients) as the reference standard, we determined whether lesions previously seen at CT colonography represented true-positive or false-positive findings. A true-positive lesion at CT colonography was defined as a lesion that was also confirmed at repeat colonoscopy or during surgery (which was performed after repeat colonoscopy in six patients), a lesion that was in the same or an adjacent colonic segment, and a lesion for which the size correlated within 50% of the diameter. A false-positive lesion was defined as such if the lesion reported at CT colonography was not detected at follow-up colonoscopy, the lesion was not in the same or an adjacent colonic segment, or there was more than a 50% discrepancy in the lesion diameter. The end point for each endoscopically nonvisualized endoluminal lesion of 6 mm or larger in diameter was repeat colonoscopy (in 45 patients) and postcolonoscopic surgery (in six patients who also underwent repeat colonoscopy before the surgery). In six patients who underwent surgery, the lesions were actually measured at surgery; in all the other patients, the lesions were measured at colonoscopy. For each of the 27 patients in whom no repeat colonoscopy or postcolonoscopic surgery was performed, we reviewed the medical chart for follow-up visits or hospitalization to June 2003. This data collection was performed by two radiologists (L.C. and J.S.), who also produced a consensus opinion. No new relevant information was revealed during the follow-up period.

All discordant CT colonographic studies (false-positive CT colonographic results, negative findings at repeat colonoscopy) from the retrospective review were also evaluated at a workstation (Advantage Windows, version 4.1; GE Medical Systems) by two experienced radiologists in consensus (M.M.M. and J.S., who had previously interpreted scans from 720 and 250 CT colonographic studies, respectively) for misinterpretation. Evaluation included review of the transverse images in a rapid cine mode, multiplanar reformations, and volume-rendered images of navigation through the lumen of the air-distended colon. The evaluation was performed with direct comparison of the CT colonographic and colonoscopic findings. Discordant studies were categorized in one of three groups: polypoid, if there was high confidence among the readers that a mass or a polypoid lesion was present, despite that the lesion was not found at repeat colonoscopy; perceptive, if no mass or polypoid lesion was found retrospectively; and technical, if a collapsed colonic segment or excessive residual stool was noted.

Statistical Analysis
Interobserver variability was not assessed in this study. Descriptive statistics were used to calculate the PPV of CT colonography in comparison with colonoscopy for masses (≥20 mm), large (≥10-mm) polyps, and medium (6–9-mm) polyps (Advanced Models, version 12.0; SPSS, Chicago, Ill).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
In 72 (13.2%) patients, CT colonography depicted 88 endoscopically nonvisualized lesions of 6 mm or larger. At CT colonography, more than one lesion of 6 mm or larger per patient was depicted in 11 patients. Eleven patients were reported to have 12 masses (≥20 mm) that they were suspected of having, 18 patients had 23 large (10–19-mm) polyps that they were suspected of having, and 47 patients had 53 medium (6–9-mm) polyps that they were suspected of having. Ninety-six percent (27 of 28) underwent repeat colonoscopy if CT colonography showed endoscopically nonvisualized lesions of 10 mm or larger compared with 45% (21 of 47) of patients in whom CT colonography showed endoscopically nonvisualized lesions of 6–9 mm in size. At a median follow-up of 31 months (range, 6–42 months), 45 (63%) of 72 patients underwent follow-up colonoscopy because of their CT colonographic findings.

All repeat colonoscopic examinations were performed within a mean of 42 days after CT colonography (range, 7 days to 1 year). All the segments of the colon with endoluminal CT colonographic findings were reached at repeat colonography. However, no segmental look back ("segmental unblinding") was performed during colonoscopy. No new endoluminal lesions were found at repeat colonoscopy that were not seen at the CT colonographic examination that was performed after the original incomplete colonoscopy. To our knowledge, no patient whose CT colonographic findings were reported as normal underwent repeat colonoscopy during the follow-up period (range, 6–42 months) at Beth Israel Deaconess Medical Center, Boston, Mass.

Endoscopically Nonvisualized Masses of 20 mm or Larger
Twelve masses in 11 patients who were suspected of having them were identified at CT colonography. Of these, CT colonography correctly depicted 11 masses in the proximal part of the colon that was not visualized at colonoscopy in 10 patients (1.8% of the cohort of patients) (Fig 1). The 11 masses that were reported correctly as such at CT colonography comprised seven adenocarcinomas (in six [1.1%] of the cohort patients), three ileocecal lipomas, and one case of sigmoid thickening presumed to be secondary to previous surgery. One patient with a sigmoid adenocarcinoma that was not reached by the colonoscope had an additional synchronous adenocarcinoma in the transverse colon. Both masses were correctly identified at CT colonography. None of these patients had endoscopically visualized masses or adenomatous polyps at initial colonoscopy, and all 10 patients underwent repeat colonoscopy because of their CT colonographic findings—that is, the repeat colonoscopy rate for masses was 100%. In the 11th patient, a 30-mm cecal lesion that was reported at CT colonography was not confirmed at repeat colonoscopy (false-positive mass). The per-patient and per-lesion PPVs of CT colonography for masses were 90.9% and 91.7%, respectively.


Figure 1A
View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Cecal carcinoma in an 80-year-old man with abdominal pain and iron-deficiency anemia. The colonoscope could not be advanced beyond the hepatic flexure because of an extremely tortuous colon. (a) Transverse CT image obtained with the patient in the supine position shows a thickened fold with a mass-like appearance (arrows) in the cecum. (b) Coronal multiplanar reformation shows a concentric thickening of a fold (arrows) in the cecum. (c) Mass within a fold covered by fresh blood (arrows) in the cecum seen at repeat colonoscopy performed after CT colonography. Adenocarcinoma was confirmed with colonic biopsy.

 

Figure 1B
View larger version (111K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Cecal carcinoma in an 80-year-old man with abdominal pain and iron-deficiency anemia. The colonoscope could not be advanced beyond the hepatic flexure because of an extremely tortuous colon. (a) Transverse CT image obtained with the patient in the supine position shows a thickened fold with a mass-like appearance (arrows) in the cecum. (b) Coronal multiplanar reformation shows a concentric thickening of a fold (arrows) in the cecum. (c) Mass within a fold covered by fresh blood (arrows) in the cecum seen at repeat colonoscopy performed after CT colonography. Adenocarcinoma was confirmed with colonic biopsy.

 

Figure 1C
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: Cecal carcinoma in an 80-year-old man with abdominal pain and iron-deficiency anemia. The colonoscope could not be advanced beyond the hepatic flexure because of an extremely tortuous colon. (a) Transverse CT image obtained with the patient in the supine position shows a thickened fold with a mass-like appearance (arrows) in the cecum. (b) Coronal multiplanar reformation shows a concentric thickening of a fold (arrows) in the cecum. (c) Mass within a fold covered by fresh blood (arrows) in the cecum seen at repeat colonoscopy performed after CT colonography. Adenocarcinoma was confirmed with colonic biopsy.

 
Endoscopically Nonvisualized Large Polyps of 10–19 mm
CT colonography depicted 23 large polyps in the proximal part of the colon that was not visualized at colonoscopy in 18 patients (3.3% of the cohort of patients) (Fig 2). Seventeen (94%) of 18 patients underwent repeat colonoscopy because of their CT colonographic findings; one patient in whom CT colonography depicted three polypoid lesions of 10–15 mm in diameter was lost to follow-up. Eleven of the patients who underwent repeat colonoscopy had 14 true-positive large polyps confirmed: nine large adenomatous polyps (one with dysplasia), two large hyperplastic polyps, two large polyps of undetermined histologic type proximal to a sigmoid carcinoma in one patient, and one appendix stump impression. Six patients had normal findings at repeat colonoscopy—that is, six false-positive polyps at CT colonography. At reevaluation of the CT colonographic images, we determined that three (50%) of six false-positive large polyps were miscalled as polypoid (ie, both readers determined that these lesions represented polyps), two were miscalled as perceptive (ie, both readers determined that these lesions were not present), and one was miscalled as technical due to poor colonic distention. The per-patient and per-lesion PPVs of CT colonography for large polyps were 64.7% and 70%, respectively.


Figure 2A
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Adenomatous polyp of 10 mm of the middle transverse colon in a 57-year-old patient who underwent routine screening colonoscopy that was stopped at the distal transverse colon because of severe discomfort. (a) Transverse CT image obtained with the patient in the supine position shows 10-mm polypoid lesion (arrow) in the middle transverse colon. (b) Coronal multiplanar reformation shows the polypoid lesion (arrow) in the middle transverse colon lying on a fold. (c) Polyp (arrows) lying on a fold in the transverse colon at repeat colonoscopy performed after the CT colonography. Histologic evaluation revealed an adenomatous polyp.

 

Figure 2B
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Adenomatous polyp of 10 mm of the middle transverse colon in a 57-year-old patient who underwent routine screening colonoscopy that was stopped at the distal transverse colon because of severe discomfort. (a) Transverse CT image obtained with the patient in the supine position shows 10-mm polypoid lesion (arrow) in the middle transverse colon. (b) Coronal multiplanar reformation shows the polypoid lesion (arrow) in the middle transverse colon lying on a fold. (c) Polyp (arrows) lying on a fold in the transverse colon at repeat colonoscopy performed after the CT colonography. Histologic evaluation revealed an adenomatous polyp.

 

Figure 2C
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Adenomatous polyp of 10 mm of the middle transverse colon in a 57-year-old patient who underwent routine screening colonoscopy that was stopped at the distal transverse colon because of severe discomfort. (a) Transverse CT image obtained with the patient in the supine position shows 10-mm polypoid lesion (arrow) in the middle transverse colon. (b) Coronal multiplanar reformation shows the polypoid lesion (arrow) in the middle transverse colon lying on a fold. (c) Polyp (arrows) lying on a fold in the transverse colon at repeat colonoscopy performed after the CT colonography. Histologic evaluation revealed an adenomatous polyp.

 
Endoscopically Nonvisualized Medium Polyps of 6–9 mm
CT colonography depicted 53 medium polyps in the proximal part of the colon that was not visualized at colonoscopy in 47 patients (8.6% of the cohort of patients) (Table 3). Twenty-one patients (45%) underwent repeat colonoscopy because of their CT colonographic findings; of these, seven had positive correlation with repeat colonoscopic findings. These seven patients had four medium adenomatous polyps and three medium hyperplastic polyps. The per-patient and per-lesion PPVs of CT colonography for medium polyps were 33.3% and 30.4%, respectively. This group was further divided into two subgroups, one of polyps of 8–9 mm and the other of polyps of 6–7 mm. The follow-up rates were 55.6% in the first subgroup and 37.9% in the latter subgroup. The per-patient PPVs were 50% in the 8–9-mm subgroup but only 18.2% in the 6–7-mm subgroup (Table 3). Reevaluation of the CT colonographic images indicated that 13 (81%) of 16 false-positive medium polyps were miscalled as polypoid, and the remaining three (19%) (all 7-mm polypoid lesions) were miscalled as perceptive.


View this table:
[in this window]
[in a new window]

  		Table 3. Endoscopically Nonvisualized Medium Polyps

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Although the rate of incomplete colonoscopy is operator dependent, 2%–10% of colonoscopic examinations performed by experienced endoscopists are incomplete (4,5,26). In our series, 3.5% of colonoscopic examinations performed by experienced endoscopists were incomplete for reasons other than poor bowel preparation.

Although the capability of CT colonography to depict polyps is both operator and technique dependent, this modality has a relatively high specificity (14,16–19). Our study findings indicate that our clinicians appear to be comfortable in accepting the negative results of a CT colonographic examination—to the best of our knowledge, none of the 474 patients (86.8% of the cohort) in our study whose CT colonographic findings were reported as normal underwent repeat colonoscopy for examination of the unvisualized colon during the follow-up period. By contrast, 96% of patients underwent repeat colonoscopy if CT colonography depicted endoscopically nonvisualized lesions of 10 mm or larger in diameter.

CT colonography correctly depicted 11 endoscopically nonvisualized masses that were larger than 20 mm in diameter in 10 patients. In six patients, seven colonic carcinomas in the endoscopically nonvisualized part of the colon were correctly identified at CT colonography (1.1% of the study cohort). Although all three lesions on the ileocecal valve were diagnostic for lipoma at imaging, repeat colonoscopy was performed and the diagnosis was confirmed. Our per-lesion PPV of 91.7% for masses is similar to that reported by others (16,23). Only one 30-mm cecal lesion reported at CT colonography was not confirmed at repeat colonoscopy (false-positive CT colonographic findings). In this case, the ascending colon was evaluated twice; however, no mass was noticed in the cecum even after repeat endoscopic inspection. At reevaluation of the CT colonographic images, we determined that this false-positive mass was miscalled as polypoid and was indistinguishable from a mass. The presumed explanation for this discrepancy was that a large but normal fold in the cecum was mistaken for a mass.

In our study, repeat colonoscopy demonstrated that CT colonography correctly depicted 70% of large 10–19-mm polyps in the endoscopically nonvisualized colon. Although our per-patient PPV of 64.7% for 10–19-mm polyps is lower than the 90%–100% per-patient PPV reported in other series (15–17,23), most of the other studies included masses or lesions of 20 mm or larger in diameter among their large-polyp group.

The follow-up rates for medium polyps were low and were related to size: 55.6% in the 8–9-mm group and 37.9% in the 6–7-mm group. This may be explained in part by the tendency of the referring gastroenterologist to defer repeat colonoscopy in patients who are at a very small risk of neoplasia (7,27,28) who had already undergone colonoscopy and CT colonography. Although it is known that polyp size is directly related to the potential for malignancy of the polyp, there remains no consensus as to what constitutes a clinically important polyp (3). In regard to polyps that are smaller than 10 mm, only 5% have high-grade dysplasia (29), and 1% of these polyps may contain invasive carcinoma (27).

With respect to lesions that are between 6 and 9 mm in diameter, we also found large differences in the PPV when this group was divided into two subgroups with a cutoff of 8 mm. The per-patient PPV was 50% in the 8–9-mm subgroup but declined steeply to only 18.2% in the 6–7-mm subgroup. Therefore, as suggested by others (19), it might be prudent to perform immediate repeat colonoscopy in the 8–9-mm subgroup, whereas patients with 6–7-mm lesions should undergo a short-term follow-up examination in 3–5 years (30,31).

In our study, 19 patients had 23 polypoid lesions depicted at CT colonography that were subsequently proved to be false-positive at repeat colonoscopy. Seven (30%) of these lesions were 10 mm or larger. Interestingly, reevaluation of the CT colonographic images by two experienced readers who knew that no polyp was found at repeat colonoscopy revealed that the majority of these false-positive polyps (17 of 23 [74%]) would have been called polypoid lesions. Although the gastroenterologists who performed the repeat colonoscopy were aware of the exact locations of the lesions that were found at CT colonography, no segmental look back (segmental unblinding) was performed at repeat colonoscopy as part of our study. Therefore, one must allow for the possibility that the endoscopist who performed the repeat colonoscopy was not always able to accurately identify the location of the polypoid lesion. Consequently, it is possible that a small number of true polyps reported at CT colonography were not actually detected at repeat colonoscopy (false-negative colonoscopic findings). Although conventional colonoscopy remains the reference standard for visualization of the colon, CT colonography and tandem colonoscopic studies have shown that colonoscopy, even in experienced hands, can lead to the missing of 6%–12% of adenomatous polyps larger than 10 mm in diameter (19,32). Interestingly, no technical failure was noticed in the completion of the second colonoscopic examination after findings at CT colonography suggested the presence of an endoluminal lesion. This might be explained by increased motivation and insistence of the gastroenterologist to look for the findings even at the possible cost of discomfort to the patient. Further studies are suggested to evaluate the characteristics of second colonoscopic examinations compared with the first diagnostic one.

Our study had limitations. The most important one was the low rate of follow-up for medium polyps, a decision largely dictated by the referring gastroenterologist. We suggest that, when a year has passed and no colonoscopy has been performed for medium polyps that are suspected, it might be reasonable to send a reminder to the referring physician about the findings, especially in patients suspected of having 8–9-mm polyps. This might increase follow-up rates in future years. We cannot address the rate of false-negative CT colonographic findings after incomplete colonoscopy because only patients with positive results at CT colonography were reexamined. Some limitations are inherent in this study because of its retrospective nature: There was a selection bias (most patients admitted were at high risk), and there was a highly variable time between initial examination and attempted repeat colonoscopy. We did not assign grades to the percentage of nonvisualized segments of the colon for adequate cleansing or distention, and the use of intravenous contrast material in 228 (41.8%) patients might account for better results than those seen with the current standard nonenhanced study (33,34). In contrast to other studies (19) and because of the retrospective nature of this study, no segmental look back (segmental unblinding) was performed at repeat colonoscopy as part of our study. We also performed only a single reading of scans by an experienced reader, with augmentation by an abdominal imaging fellow. Prospective studies tend to use double reading (14–16), which is obviously preferable (14); however, in the context of clinical practice, this technique is unlikely to be widely performed even in an academic setting. In 11 patients, CT colonography depicted more than one lesion of 6 mm or larger in diameter per patient. Since data dependency or clustering is possible, the PPV was calculated not only per lesion but also per patient.

In conclusion, our study findings indicate that CT colonography has the potential to become an accepted technique for evaluation of the nonvisualized part of the colon after incomplete colonoscopy and that it can increase the diagnostic yield of masses and clinically important polyps in this part of the colon. Further, 96% of patients underwent repeat colonoscopy if CT colonography depicted endoscopically nonvisualized lesions of 10 mm or larger in diameter. CT colonography after incomplete colonoscopy has per-patient and per-lesion PPVs of 90.9% and 91.7%, respectively, for masses and 64.7% and 70%, respectively, for polyps 10 mm or larger but led to overcalls for one of three large (≥10-mm) polyps and two of three medium (6–9-mm) polyps in the endoscopically nonvisualized colon.


   ADVANCE IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE
 References
 


   IMPLICATIONS FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
References
 


   FOOTNOTES
 

Abbreviations: PPV = positive predictive value

Authors stated no financial relationship to disclose.

Author contributions:Guarantors of integrity of entire study, L.C., J.S., M.M.M.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, L.C., J.S., R.J.F., M.M.M.; clinical studies, L.C., J.S., J.B.K., V.R., M.M.M.; statistical analysis, L.C., J.S., R.J.F., M.M.M.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 

  1. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med 1993;329:1977–1981.
  2. Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. Minnesota Colon Cancer Control Study. N Engl J Med 1993;328:1365–1371.
  3. Lieberman DA, Weiss DG, Bond JH, et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med 2000;343:162–168.
  4. Anderson ML, Heigh RI, McCoy GA, et al. Accuracy of assessment of the extent of examination by experienced colonoscopists. Gastrointest Endosc 1992;38:560–563.[Medline]
  5. Marshall JB, Barthel JS. The frequency of total colonoscopy and terminal ileal intubation in the 1990s. Gastrointest Endosc 1993;39:518–520.[Medline]
  6. Dafnis G, Blomqvist P, Pahlman L, et al. The introduction and development of colonoscopy within a defined population in Sweden. Scand J Gastroenterol 2000;35:765–771.[CrossRef][Medline]
  7. Winawer SJ, Fletcher RH, Miller L, et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology 1997;112:594–642.[CrossRef][Medline]
  8. Chong A, Shah JN, Levine MS, et al. Diagnostic yield of barium enema examination after incomplete colonoscopy. Radiology 2002;223:620–624.[Abstract/Free Full Text]
  9. Lappas JC, Maglinte DD, Chernish SM, et al. Discomfort during double contrast barium enema examination: a placebo controlled double-blind evaluation of the effect of glucagon and diazepam. Radiology 1995;197:95–99.[Abstract/Free Full Text]
  10. Eckardt VF, Kanzier G, Willems D, et al. Colonoscopy without premedication versus barium enema: a comparison of patient discomfort. Gastrointest Endosc 1996;44:177–180.[CrossRef][Medline]
  11. Bova JG, Jurdi RA, Bennett WF. Antispasmodic drugs to reduce discomfort and colonic spasm during barium enemas: comparison of oral hyoscyamine, i.v. glucagon and no drug. AJR Am J Roentgenol 1993;161:965–968.
  12. Winawer SJ, Stewart ET, Zauber AG, et al. A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. National Polyp Study Work Group. N Engl J Med 2000;342:1766–1772.
  13. Rockey DC, Paulson E, Niedzwiecki D, et al. Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: prospective comparison. Lancet 2005;365:305–311.[Medline]
  14. Johnson CD, MacCarty RL, Welch TJ, et al. Comparison of the relative sensitivity of CT colonography and double-contrast barium enema for screen detection of colorectal polyps. Clin Gastroenterol Hepatol 2004;2:314–321.[CrossRef][Medline]
  15. Yee J, Akerkar GA, Hung RK, et al. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219:685–692.[Abstract/Free Full Text]
  16. Fenlon HM, Nunes DP, Schroy PC 3rd, et al. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 1999;341:1496–1503.[Abstract/Free Full Text]
  17. Macari M, Bini EJ, Xue X, et al. Colorectal neoplasms: prospective comparison of thin-section low-dose multi-detector row CT colonography and conventional colonoscopy for detection. Radiology 2002;224:383–392.[Abstract/Free Full Text]
  18. Sosna J, Morrin MM, Kruskal JB, et al. CT colonography of colorectal polyps: a metaanalysis. AJR Am J Roentgenol 2003;181:1593–1598.[Abstract/Free Full Text]
  19. 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]
  20. Rex DK. Is virtual colonoscopy ready for widespread application? Gastroenterology 2003;125:608–610.
  21. Fenlon HM, McAneny DB, Nunes DP, et al. Occlusive colon carcinoma: virtual colonoscopy in the preoperative evaluation of the proximal colon. Radiology 1999;210:423–428.[Abstract/Free Full Text]
  22. Morrin MM, Kruskal JB, Farrell RJ, et al. Endoluminal CT colonography after an incomplete endoscopic colonoscopy. AJR Am J Roentgenol 1999;172:913–918.[Abstract/Free Full Text]
  23. Neri E, Giusti P, Battolla L, et al. Colorectal cancer: role of CT colonography in preoperative evaluation after incomplete colonoscopy. Radiology 2002;223:615–619.[Abstract/Free Full Text]
  24. Macari M, Berman P, Dicker M, et al. Usefulness of CT colonography in patients with incomplete colonoscopy. AJR Am J Roentgenol 1999;173:561–564.[Abstract/Free Full Text]
  25. Britton I, Dover S, Vallance R. Immediate CT pneumocolon for failed colonoscopy: comparison with routine pneumocolon. Clin Radiol 2001;56:89–93.[CrossRef][Medline]
  26. Shumaker DA, Zaman A, Katon RM. Use of a variable-stiffness colonoscope allows completion of colonoscopy after failure with the standard adult colonoscope. Endoscopy 2002;34:711–714.[Medline]
  27. Muto T, Bussey HJ, Morson BC. The evolution of cancer of the colon and rectum. Cancer 1975;36:2251–2270.[Medline]
  28. Hofstad B, Vatn MH, Andersen SN, et al. Growth of colorectal polyps: redetection and evaluation of unresected polyps for a period of 3 years. Gut 1996;39:449–456.[Abstract/Free Full Text]
  29. O'Brien MJ, Winawer SJ, Zauber AG, et al. The National Polyp Study. Patient and polyp characteristics associated with high-grade dysplasia in colorectal adenomas. Gastroenterology 1990;98:371–379.
  30. Pickhardt PJ, Taylor AJ, Johnson GL, et al. Building a CT colonography program: necessary ingredients for reimbursement and clinical success. Radiology 2005;235:17–20.[Free Full Text]
  31. Zalis ME, Barish MA, Choi JR, et al. CT colonography reporting and data system: a consensus proposal. Radiology 2005;236:3–9.[Free Full Text]
  32. Rex DK, Cutler CS, Lemmel GT, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997;112:24–28.[CrossRef][Medline]
  33. Sosna J, Morrin MM, Kruskal JB, et al. Colorectal neoplasms: role of intravenous contrast-enhanced CT colonography. Radiology 2003;228:152–156.[Abstract/Free Full Text]
  34. Neri E, Vagli P, Picchietti S, et al. CT colonography: contrast enhancement of benign and malignant colorectal lesions versus fecal residuals. Abdom Imaging 2005;30:694–697.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
D. Shaham, E. Libson, and M. Graif
Radiology in Israel: What Is Happening?
Am. J. Roentgenol., October 1, 2008; 191(4): 951 - 953.
[Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
D. M. Hough, M. A. Kuntz, J. L. Fidler, C. D. Johnson, B. T. Petersen, J. M. Kofler, and J. G. Fletcher
Detection of Occult Colonic Perforation Before CT Colonography After Incomplete Colonoscopy: Perforation Rate and Use of a Low-Dose Diagnostic Scan Before CO2 Insufflation
Am. J. Roentgenol., October 1, 2008; 191(4): 1077 - 1081.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
S. An, K. H. Lee, Y. H. Kim, S. H. Park, H. Y. Kim, S. H. Kim, and N. Kim
Screening CT Colonography in an Asymptomatic Average-Risk Asian Population: A 2-Year Experience in a Single Institution
Am. J. Roentgenol., September 1, 2008; 191(3): W100 - W106.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Copel, L.
Right arrow Articles by Morrin, M. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Copel, L.
Right arrow Articles by Morrin, M. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE