Long-Term Follow-up of the Italian Flexible Sigmoidoscopy Screening Trial

Background: Recent reports showed that the protective effect of flexible sigmoidoscopy (FS) screening was maintained up to17 years, although differences were reported by sex. Objective: To assess long-term reduction of colorectal cancer (CRC) incidence and mortality after a single FS screening. Design: Parallel randomized controlled trial. (ISRCTN registry number: 27814061) Setting: 6 centers in Italy. Participants: Persons aged 55 to 64 years expressing interest in having FS screening if invited, recruited from 1995 to 1999 and followed until 2012 (incidence) and 2014 to 2016 (mortality). Intervention: Eligible persons were randomly assigned (1:1 ratio) to either the once-only FS screening group or control (usual care) group. Measurements: Incidence and mortality rate ratios (RRs) and rate differences. Results: A total of 34 272 persons (17 136 in each group) were included in the analysis; 9911 participants had screening in the intervention group. Median follow-up was 15.4 years for incidence and 18.8 years for mortality. Incidence of CRC was reduced by 19% (RR, 0.81 [95% CI, 0.71 to 0.93]) in the intention-to-treat (ITT) analysis, comparing the intervention with the control group, and by 33% (RR, 0.67 [CI, 0.56 to 0.81]) in the per protocol (PP) analysis, comparing participants screened in the intervention group with the control persons. Colorectal cancer mortality was reduced by 22% (RR, 0.78 [CI, 0.61 to 0.98]) in the ITT analysis and by 39% (RR, 0.61 [CI, 0.44 to 0.84]) in the PP analysis. Incidence of CRC was statistically significantly reduced among both men and women. Colorectal cancer mortality was statistically significantly reduced among men (ITT RR, 0.73 [CI, 0.54 to 0.97]) but not among women (ITT RR, 0.90 [CI, 0.59 to 1.37]). Limitation: Self-selection of volunteers from the general population sample targeted for recruitment may limit generalizability. Conclusion: The strong protective effect of a single FS screening for CRC incidence and mortality was maintained up to 15 and 19 years, respectively.