P22 Impact of adenoma detection rates at flexible sigmoidoscopy on long-term colorectal cancer incidence and mortality

Introduction Detection and removal of adenomas reduces colorectal cancer (CRC) risk. However, the effects of variable adenoma detection rates (ADRs) on long-term CRC incidence and mortality are not known. We investigated this using data from the UK Flexible Sigmoidoscopy Screening Trial (UKFSST). Methods We analysed data from 167,882 UKFSST participants, of whom 111,503 were in the control arm and 56,379 in the intervention arm. The control arm was not contacted while the intervention arm was offered a single flexible sigmoidoscopy screen. In total, 40,085 participants underwent flexible sigmoidoscopy screening at 13 trial centres. Median follow-up was 17 years. At each centre, a single endoscopist performed nearly all flexible sigmoidoscopies. We used multivariable logistic regression to classify centres into high-, intermediate-, and low-detector ranking groups based on the ADR of their main endoscopist. We calculated CRC incidence and mortality rates, and estimated hazard ratios (HRs) with 95% confidence intervals (CIs) using Cox regression. Results Five centres were classified into the high-detector group, four into the intermediate-detector group, and four into the low-detector group. Average ADRs in the high-, intermediate-, and low-detector groups were 15%, 12%, and 9%, respectively. In all three groups, all-site CRC incidence and mortality were reduced among screened participants, compared to the control arm, and although the heterogeneity was not statistically significant, a larger effect was seen in the high-detector group (incidence: HR=0.58, 95%CI 0.50–0.67; mortality: HR=0.52, 0.39–0.69) than in the low-detector group (incidence: HR=0.72, 0.61–0.85; mortality: HR=0.68, 0.51–0.92). For distal CRC, incidence and mortality were reduced among screened participants, compared to the control arm, in all three groups and there was significant heterogeneity by detector ranking, with a substantially larger effect in the high-detector group (incidence: HR=0.34, 0.27–0.42; mortality: HR=0.22, 0.13–0.37) than in the low-detector group (incidence: HR=0.55, 0.44–0.68; mortality: HR=0.54, 0.34–0.86). Conclusions Higher ADRs at screening flexible sigmoidoscopy result in greater long-term protection against CRC incidence and mortality.