Sigmoid cancer versus chronic diverticular disease: differentiating features at CT colonography.

PURPOSE: To retrospectively identify morphologic findings at computed tomographic (CT) colonography that are the most reliable in the differentiation of masslike chronic diverticular disease from sigmoid carcinoma in a large patient cohort. MATERIALS AND METHODS: This study was approved by the institutional review boards. The need for signed consent was waived for this retrospective study. The cohort consisted of 212 patients (mean age, 68 years; 113 women, 99 men) with focal masslike findings in the sigmoid colon at CT colonography, representing chronic diverticular disease (n = 97) or sigmoid carcinoma (n = 115). CT colonography studies were scored according to presence or absence of potential discriminators by a panel of four readers in consensus. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated, and multivariate analysis was performed. RESULTS: Absence of diverticula in the affected segment showed high NPV and PPV (0.95 and 0.93, respectively). Also, shoulder phenomenon showed a high NPV (0.92) and PPV (0.75). Segment length of 10 cm or less (NPV, 0.85; PPV, 0.61) and destroyed mucosal folds (NPV, 1.00; PPV, 0.62) had a high NPV but a low PPV. Although segments affected by carcinoma often showed straightened and eccentric growth patterns, no thick fascia sign, and more and larger local-regional lymph nodes (all P < .05), NPV was insufficient for discrimination (NPV ≤ 0.66). Combination of absence of diverticula and presence of shouldering showed a high diagnostic certainty (93%). CONCLUSION: Carcinoma is best differentiated from masslike diverticular disease by the absence of diverticula in the affected segment and the presence of shoulder phenomenon.

[CT colonography as first-line diagnostic procedure in patients with bowel symptoms]. / CT-colografie als eerstelijnsdiagnosticum bij patiënten met darmklachten.

OBJECTIVE: To investigate in how many patients with bowel or abdominal complaints, referred by the primary care physician (PCP) for exclusion of colorectal carcinoma (CRC), the more invasive colonoscopy could be avoided on the basis of the findings of CT colonography. DESIGN: Retrospective, descriptive. METHODS: All consecutive patients who underwent CT colonography in our centre on the request of their PCP from December 2006 to June 2009 were included. Demographic and referral data were collected. CT colonography results were described according to the 'CT Colonography Reporting and Data System'. We also investigated how many patients had to undergo colonoscopy in the 6 months following CT colonography. RESULTS: 398 patients (154 men and 244 women) with a median age of 61 years (range: 22-91) were included. Follow-up colonoscopy was indicated by CT colonography in 30 patients (7.5%) for suspected colorectal carcinoma, polyps > 10 mm, or 3 or more polyps 6-9 mm in size. In 33 patients (8.3%) follow-up colonoscopy or CT colonography was indicated for 1 or 2 polyps 6-9 mm in size, or suspicious lesions. 11 of these patients (2.8%) underwent colonoscopy. In 335 patients (84.2%) polyps > 6 mm or malignancies could be excluded. 18 of these patients (4.5%) still had a colonoscopy. In total, colonoscopy was spared in 341 patients (85.7%). Significant or potentially significant extra-colonic pathological abnormalities were found in 63 patients (15.8%). CONCLUSION: Our results support the theory that in the vast majority of patients with low or moderate suspicion of CRC referred by their PCP, invasive colonoscopy could be avoided, because CRC and polyps could be excluded by CT colonography. CT colonography could be a valuable additional diagnostic tool in primary care.