Colonoscopy, tumors, and inflammatory bowel disease

Narrow−band imaging (NBI) is a new technique that allows us to distinguish neoplastic and non− neoplastic lesions without chromoendoscopy, and should be very useful in detecting small or flat lesions missed by standard endoscopy. A study recently examined 148 colorectal lesions, 16 hyperplastic polyps, 84 tubular adenomas, and 48 early carcinomas [1]. Lesions were ob− served first under NBI magnification and then under standard magnification with chromo− endoscopy. Correspondence between the two diagnostic methods was 88% for type II, 100% for type IIIs, 98% for type IIIl, 88% for type IV, 78 % for type Vi, and 100% (3/3) for type Vn pit patterns. NBI depicted a brownish change on the basis of surface capillaries in 6% of hyperplasia and 99% of tubular adenomas (P < 0.005) [1]. In a second, prospective study, 180 colorectal lesions were ob− served with conventional colonoscopy, under low− and high−magnification NBI, and chromo− endoscopy [2]. The diagnostic accuracy of NBI with low or high magnification was significantly higher than that of conventional colonoscopy (P = 0.0434 and P < 0.001), and was comparable to that of chromoendoscopy. Both low− and high−magnification NBI were capable of distin− guishing neoplastic from non−neoplastic colorec− tal lesions [2]. A randomized controlled trial [3] of colonoscopy withdrawal in white light versus NBI in 434 pa− tients aged 50 years was performed using high− definition, wide−angle (1708 field of view) colo− noscopes. There was no difference between groups in the percentage of patients with 3 1 ade− noma for the entire cohort (white light 67% vs. NBI 65%; P = 0.61). The high prevalence rates of adenomas were accounted for by detection of large numbers of adenomas, including flat ade− nomas, which were £ 5 mm. For differential diag− nosis of neoplastic (adenoma and adenocarcino− ma) and hyperplastic polyps, the sensitivity of the conventional colonoscope for diagnosis of neoplastic polyps was 82.9 %, specificity was 80.0 %, and diagnostic accuracy was 81.8 % [4], significantly lower than those achieved with the NBI system (sensitivity 95.7 %, specificity 87.5%, accuracy 92.7%), and chromoendoscopy (sensi− tivity 95.7%, specificity 87.5%, accuracy 92.7 %). Therefore, no significant difference existed be− tween the NBI system and chromoendoscopy during differential diagnosis of neoplastic and non−neoplastic polyps [4]. Moreover, pit patterns were not always identical with NBI and chromo− endoscopy, so, the Kudo classification may need to be modified and revalidated before it can be used with confidence with NBI [5]. Another study compared the accuracy of NBI with standard colonoscopy in a prospective, ran− domized, crossover study of 42 patients with longstanding ulcerative colitis [6]. With NBI, 52 suspicious lesions were detected in 17 patients, compared with 28 suspicious lesions in 13 pa− tients detected during conventional colonoscopy. In four patients the neoplasia was detected by both techniques, in four patients neoplasia was detected only by NBI, and in three patients neo− plasia was detected only by conventional colo− noscopy (P = 0.7). The sensitivity of the NBI sys− tem for the detection of patients with neoplasia seems to be comparable to conventional colonos− copy, although more suspicious lesions were found during NBI [6]. The value of combined chromoscopy and endo− microscopy was assessed for the diagnosis of in− traepithelial neoplasias in a randomized con− trolled trial of ulcerative colitis in clinical remis− sion [7]. In the conventional colonoscopic group (n = 73), random biopsy examinations and targe− ted biopsy examinations were performed. In the endomicroscopy group (n = 80), circumscribed mucosal lesions were identified by chromoscopy