Background: This systematic review and meta-analysis aims to evaluate efficacy and safety of endoscopic treatment for the non-polypoid dysplasia in patients with long-standing IBD.
Methods: Medline, Embase, Cochrane, and clinicaltrials.gov registry were comprehensively searched. Pooled estimates of curative, R0, en-bloc resection rates, CRC, metachronous dysplasia, and local recurrence rates were calculated. Subgroup analysis according to areas, lesion size, endoscopic resection techniques, and grades of dysplasia were conducted. Data synthesis was completed in R using the package "meta".
Results: Of the 973 studies initially identified, 7 met the inclusion/exclusion criteria. These were all single-arm cohorts and included a total of 202 patients with IBD and non-polypoid dysplasia. The combined R0 and en-bloc resection rate were 0.70 (95% CI 0.55-0.81) and 0.86 (95% CI 0.65-0.95), respectively, with a recurrence rate of 0.08 (95% CI 0.05-0.13). CRC and metachronous dysplasia incidences were pooled as 32.53 (95% CI 12.21-86.67) and 90.24 (95% CI 44.91-181.33) per 1000 patient years.
Conclusions: Non-polypoid dysplasia associated with IBD can be resected endoscopically, especially by ESD. However, these patients have higher CRC and metachronous dysplasia incidence rates than patients with polypoid dysplasia, indicating a closer endoscopic surveillance.
Keywords: Endoscopic resection; Inflammatory bowel disease; Meta-analysis; Non-polypoid dysplasia.