Background: Early detection of ulcerative colitis-associated neoplasia (UCAN) is often difficult. The aim of this study was to clarify the morphology of initial UCAN.
Methods: White-light colonoscopy images obtained within the 2 years before UCAN diagnosis were retrospectively reviewed. The primary endpoint was the frequency of visible or invisible neoplasia on the endoscopic images before UCAN diagnosis. The secondary endpoints were comparisons of (1) visible or invisible neoplasia on initial endoscopic images of early-stage and advanced cancers, (2) the clinical backgrounds of patients in whom neoplasia was visible or invisible on initial endoscopic images, and (3) the clinical backgrounds of patients with distinct and indistinct UCAN borders.
Results: Of the 27 UCAN lesions (11 early-stage; 16 advanced-stage), 25.9% (n = 7) were initially visible and 74.1% (n = 20) were invisible. The mean interval between the last surveillance colonoscopy and UCAN diagnosis was 14.5 ± 6.7 months. Of early-stage cancers, 18.2% (n = 2) were visible and 81.8% (n = 9) were invisible. Of advanced-stage cancers, 31.3% (n = 5) were visible and 68.8% (n = 11) were invisible. Invisible lesions were significantly more common in the rectum (p = 0.011) and tended to be more common in patients with inflammation and left-sided colitis (p = 0.084, p = 0.068, respectively). Patients with indistinct UCAN borders were significantly more likely to present with inflammation than those with distinct UCAN borders (p = 0.021).
Conclusion: More careful surveillance is needed because rectum lesions and inflammation are difficult to identify as neoplasia even within the 2 years before a UCAN diagnosis.
Keywords: colorectal cancer; interval cancer; surveillance; ulcerative colitis; ulcerative colitis-associated neoplasia.