Detecting dysplasia using white light endoscopy or chromoendoscopy in ulcerative colitis patients without primary sclerosing cholangitis: A systematic review and meta-analysis.

Background: Endoscopic examinations are a vital diagnostic tool for dysplasia. Establishing the precision of different modes of examination is essential due to the disparate pick-up rates of dysplasia.

Objective: The aim of this article was to establish the pick-up rates of dysplastic or cancerous lesions using white light endoscopy (WLE) and random/targeted biopsies, or chromoendoscopy (CE), in patients with ulcerative colitis (UC) without primary sclerosing (PSC) or Crohn's disease (CD).

Methods: A systematic review to identify all studies up to November 2017, without language restriction, was conducted from PubMed, the Cochrane Controlled Trials Register (1960-2017), MEDLINE, CINAHL and EMBASE (1981-2017). MeSH and text word terms used included "ulcerative colitis", "dysplasia", "random biopsy", "targeted biopsy", "colonoscopy", "white light", and "chromoendoscopy". Further searches were performed using the bibliographies of these articles. Methods: All studies reporting on colonoscopy detection rates of dysplasia and cancers in UC without involvement of PSC or CD were included. There was no age restriction to include patients. Methods: Outcome data were extracted by 2 authors independently using outcome measures defined a priori. Quality assessment was performed using the Newcastle-Ottawa scales.

Results: Data were extracted and analysed according to meta-analytical techniques using comprehensive meta-analysis. The pooled overall pick-up rate of dysplastic/cancerous lesions on WLE random biopsies was 5.6% [Event rate 0.06 (0.01, 0.23), df = 4, I2 = 94%]. Using a combined random and targeted approach with WLE the incidence was 5.1% [Event rate 0.05 (0.03, 0.09), df = 4, I2 = 96%]. One study reported on CE and found a 7% pick-up rate for dysplastic lesions.

Conclusions: Endoscopic examination of UC patients without PSC identifies dysplastic or cancerous lesions in 5-7% of cases. WLE and random biopsies may pick-up a similar number of lesions to targeted biopsies, however the number of biopsies may need to be greater to achieve this equivalence. CE has a slightly higher pick-up rate. Further comparative studies are required to strengthen the body of evidence.