3440 A prospective single centre study comparing ct pneumocolon against conventional colonoscopy for detecting colorectal neoplasms.

BACKGROUND and AIMS: To directly compare CT pneumocolon against conventional colonoscopy for detecting colorectal neoplasms. METHODS: 201 consecutive patients (male:female ratio 83:118, mean age 71 years, range 55-91) with colorectal symptoms or needing surveillance for colorectal cancers or polyps, underwent CT pneumocolon prior to conventional colonoscopy. RESULTS: Conventional colonoscopy found 62 patients with colorectal neoplasms; 13 of these patients had invasive cancers, 14 had significant or ≥1cm polyps, 25 had 6-9mm polyps, and 79 had ≤5mm polyps. CT pneumocolon detected all 13 invasive cancers but also reported 2 false positive cancers. CT pneumocolon detected 7 patients with ≥1cm polyps, 4 with 6-9mm polyps and 9 with ≤5mm polyps. For invasive cancers, sensitivity was 100%, specificity 99%, positive predictive value (PPV) 87%, and negative predictive value (NPV) 100%. For ≥1cm polyps, sensitivity was 50%, specificity 96%, PPV 47%, and NPV 96%. 31 patients had incomplete colonoscopies. 3 of these patients had invasive cancers detected on CT pneumocolon, subsequently confirmed on repeat colonoscopy (2) or operation (1). CT pneumocolon also detected 7 extra-colonic cancers (3 pancreatic, 1 renal and 3 metastatic adenocarcinomas). CONCLUSIONS: CT pneumocolon has a high sensitivity and specificity for detecting invasive colorectal cancers, but it is not useful for detecting colorectal polyps. In the elderly patient, in whom a colorectal malignancy is suspected, it maybe a useful first-line investigation.