Outcome of 24 years national surveillance in different hereditary colorectal cancer subgroups leading to more individualised surveillance.

BACKGROUND: Individuals with hereditary non-polyposis colorectal cancer (HNPCC) have a high risk of colorectal cancer (CRC). The benefits of colonic surveillance in Lynch syndrome and Amsterdam-positive (familial CRC type X familial colorectal cancer type X (FCCTX)) families are clear; only the interval between colonoscopies is debated. The potential benefits for families not fulfilling the Amsterdam criteria are uncertain. The aim of this study was to compare the outcome of colonic surveillance in different hereditary subgroups and to evaluate the surveillance programmes.

METHODS: A prospective, observational study on the outcome of colonic surveillance in different hereditary subgroups based on 24 years of surveillance data from the national Danish HNPCC register.

RESULTS: We analysed 13 444 surveillance sessions, including 8768 incidence sessions and 20 450 years of follow-up. CRC was more incident in the Lynch subgroup (2.0%) than in any other subgroup (0.0-0.4%, p<0.0001), but the incidence of advanced adenoma did not differ between the Lynch (3.6%) and non-Lynch (2.3-3.9%, p=0.28) subgroups. Non-Lynch Amsterdam-positive and Amsterdam-negative families were similar in their CRC (0.1-0.4%, p=0.072), advanced adenoma (2.3-3.3%, p=0.32) and simple adenoma (8.4-9.9%, p=0.43) incidence. In moderate-risk families, no CRC and only one advanced adenoma was found.

CONCLUSIONS: The risk of CRC in Lynch families is considerable, despite biannual surveillance. We suggest less frequent and more individualised surveillance in non-Lynch families. Individuals from families with a strong history of CRC could be offered 5-year surveillance colonoscopies (unless findings at the preceding surveillance session indicate shorter interval) and individuals from moderate-risk families could be handled with the population-based screening programme for CRC after an initial surveillance colonoscopy.