RESUMO
BACKGROUND AND AIMS: Low grade dysplasia (LGD) is believed to predispose to colorectal cancer (CRC), and proctocolectomy has been advocated when this is identified. Between 1978 and 1990, 160 patients with longstanding extensive ulcerative colitis (UC) were recruited for annual colonoscopic surveillance and 40 developed LGD at some stage. We report the outcome of this cohort 10 years after the original study ended. METHODS: Retrospective cohort study and histopathological review of the original diagnoses of LGD. The outcome of 158/160 (98.8%) patients was established in 2000. RESULTS: Of the 128 patients still alive and with an intact colon at the end of 1990, two were not traceable, 29 had LGD, and 97 had no dysplasia (controls). After 10 years, high grade dysplasia (HGD) or CRC developed in 3/29 LGD (10%) and in 4/97 controls (4.0%). Kaplan-Meier analysis from 1991 to death or colectomy did not show a statistically significant difference between the two groups (log rank test p=0.63). Histopathological review demonstrated the unreliability of LGD diagnosis. Agreement between pathologists was uniformly poor: kappa <0.4 for all comparisons. CONCLUSION: LGD diagnosis is not sufficiently reliable to justify prophylactic colectomy. Conservative management of established LGD cases should not be ruled out.