Systematic review with meta-analysis: IBD-associated colonic dysplasia prognosis in the videoendoscopic era (1990 to present).

2020 
INTRODUCTION: The prognosis of dysplasia in patients with IBD is largely determined from observational studies from the pre-videoendoscopic era (pre-1990s) that does not reflect recent advances in endoscopic imaging and resection. AIMS: To better understand the risk of synchronous colorectal cancer and metachronous advanced neoplasia (ie high-grade dysplasia or cancer) associated with dysplasia diagnosed in the videoendoscopic era, and to stratify risk according to a lesion's morphology, endoscopic resection status or whether it was incidentally detected on biopsy of macroscopically normal colonic mucosa (ie invisible). METHODS: A systematic search of original articles published between 1990 and February 2020 was performed. Eligible studies reported on incidence of advanced neoplasia at follow-up colectomy or colonoscopy for IBD-dysplasia patients. Quantitative and qualitative analyses were performed. RESULTS: Thirty-three studies were eligible for qualitative analysis (five for the meta-analysis). Pooled estimated proportions of incidental synchronous cancers found at colectomy performed for a pre-operative diagnosis of visible high-grade dysplasia, invisible high-grade dysplasia, visible low-grade dysplasia and invisible low-grade dysplasia were 13.7% (95% CI 0.0-54.1), 11.4% (95% CI 4.6-20.3), 2.7% (95% CI 0.0-7.1) and 2.4% (95% CI 0.0-8.5) respectively. The lowest incidences of metachronous advanced neoplasia, for dysplasia not managed with immediate colectomy but followed up with surveillance, tended to be reported by the studies where high definition imaging and/or chromoendoscopy was used and endoscopic resection of visible dysplasia was histologically confirmed. CONCLUSIONS: The prognosis of IBD-dysplasia diagnosed in the videoendoscopic era appears to have been improved but the quality of evidence remains low. Larger, prospective studies are needed to guide management. PROSPERO registration no: CRD42019105736.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    62
    References
    7
    Citations
    NaN
    KQI
    []