Narrow-band imaging and chromoendoscopy for the detecion of colonic dysplasia in inflammatory bowel disease: a prospective and randomized study

Abstract

Introduction: Patients with inflammatory bowel disease (IBD) are underincreased risk of colonic dysplasia and neoplasia, approximately, 8 years after diagnosis.1 The development of techniques that improve the diagnostic ability todetect those dysplasias has scientific, economic and practical impact.MATERIALS AND Methods: The NBI (narrow band image) has been described as avaluable method comparable to chromoendoscopy for the detection of many cancersof the upper digestive and respiratory systems.2 The two techniques werecompared in this study in patients with IBD after at least 8 years from diagnosis. Results: 29 patients were randomized (13 for chromoendoscopy and 16 for NBI).61.5% and 68.75% were female, mean age of 50.3 and 49.5 years, in chromoendoscopyand NBI groups, respectively. The mean disease duration was 15.3 (DP 6.5 years 2) and 15.9 years (DP 9.0 years 2) for chromoendoscopy and NBI, respectively. 53.8% of patients in the chromoendoscopy group and 56.3% in theNBI had Crohn's disease (CD). None of those epidemiological data, extension andbehavior of CD and Ulcerative Colitis, use of medications, endoscopic grade ofdisease activity and symptoms at the time of the exam disclosed statistical significance.The average time of examination was 43.6 minutes for the chromoendoscopygroup, versus 34.1 minutes for the NBI group.Regarding the presence of dysplasia, 30% of patients in the chromoendoscopy groupshowed some dysplastic lesions on histological examination (all biopsies directed tomucosal lesions), while no patients in the NBI group had such lesions (chi-square = 4.13; Rcritical> 3.841, considering an error of 5%). We found two adenomas and onedysplastic lesions of the type DALM (dysplasia-associated lesion or mass), typical ofIBD. When we look at correcting by means of the Yates correction test for small samples, we observed R = 2, 0 (Rcritical > 3.841, considering an error of 5%). Conclusion: Those preliminary data have shown no statistical differencebetween the endoscopic techniques (NBI and chromoendoscopy), but theyrevealed a strong statistical tendency of superiority of chromoendoscopy comparedto NBI. Increasing the sample size will certainly help determine in a moreprecise way whether there is equality or difference between both methods.