Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P252 Is the time for chromoendoscopy spreading in dysplasia surveillance of inflammatory bowel diseases? A systematic review with meta-analysis

A. Iannone*, G. Losurdo, M. Barone, E. Ierardi, A. Di Leo, M. PRINCIPI

University of Bari, DETO, Bari, Italy

Background

A previous meta-analysis identified chromoendoscopy (CE) as technique of choice for dysplasia surveillance in inflammatory bowel diseases (IBD). Nevertheless, successive studies reported a similar performance of CE and conventional white light endoscopy (WLE) in this setting. Our aim was to perform an updated meta-analysis comparing CE and WLE efficacy in the detection of dysplastic lesions amongst IBD patients undergoing surveillance colonoscopy.

Methods

We searched in MEDLINE, EMBASE, PubMed, Scopus, and Web-of-Science electronic databases for all references including the terms inflammatory bowel disease, Crohn’s, or ulcerative colitis, and chromoendoscopy. We evaluated the following outcomes: number of patients diagnosed with dysplasia, detection of any dysplastic lesion, detection of flat dysplastic lesions, and duration of the procedure. The random-effects model was used because it represents a conservative method that considers different sources of variation amongst studies. The heterogeneity was tested with Q and I2 statistics. Sensitivity analysis was performed by excluding 1 study at a time. Publication bias was assessed using Egger’s and trim and fill tests.

Results

Twelve studies involving 3 954 patients were included in the meta-analysis. CE showed higher efficacy in identifying patients with dysplasia (OR = 2.41; 95% CI 1.77–3.28; p < 0.001). Additionally, CE demonstrated higher performance in detecting both any type of dysplasia (RR = 2; 95% CI 1.48–2.71; p < 0.001) and flat dysplastic lesions (RR = 4.01; 95% CI 2.24–7.18; p < 0.001). Conversely, WLE showed a shorter time of procedure (Cohen’s d = -0.98; 95% CI -1.56, -0.41; p = 0.001). This effect was negatively affected by pancolitis percentages in examined samples. Sensitivity analyses did not show any significant change in overall effect sizes. No publication bias was observed.

Conclusion

Our meta-analysis, performed in a greater sample of IBD patients, supports the matter of higher efficacy of CE over WLE for dysplasia detection. Moreover, the time gap between procedures may be affected by pancolitis.