Inflammatory bowel disease (IBD)-associated colorectal cancer has become less common in the modern era, but dysplasia remains the key precursor lesion and strongest marker of future neoplastic risk. Management has therefore shifted from population-level cancer prevention towards precision surveillance, risk stratification and organ-preserving endoscopic therapy. This narrative review summarises contemporary principles for the detection, reporting and management of neoplasia in IBD, with emphasis on practical decision-making. Cumulative inflammatory burden is a major modifiable driver of colorectal neoplasia, supporting durable mucosal healing as a central cancer-prevention strategy. High-quality surveillance should be performed in quiescent disease using high-definition imaging, with dye-spray or virtual chromoendoscopy and targeted biopsies where appropriate. Random biopsies have a limited routine role but remain useful in selected high-risk settings, including primary sclerosing cholangitis, prior dysplasia, invisible dysplasia or inadequate mucosal visualisation. Visible dysplasia should be described systematically using site, size, morphology, surface pattern and surrounding mucosa, and assessed for endoscopic resectability. Well-demarcated lesions without features of deep invasion should be considered for complete endoscopic resection, preferably en bloc where feasible. Larger, non-polypoid, fibrotic or poorly lifting lesions should be referred early to centres with IBD and advanced endoscopy expertise. Invisible dysplasia requires expert pathological confirmation, optimisation of inflammation and repeat high-quality chromoendoscopic reassessment. Persistent high-grade, multifocal or unresectable dysplasia should prompt multidisciplinary discussion regarding colectomy. Post-resection surveillance remains essential because of local recurrence and metachronous neoplasia risk. Shared decision-making should integrate dysplasia grade, visibility, resectability, inflammatory burden, comorbidity, local expertise and patient preference.
Bharath et al. (Mon,) studied this question.