A randomised controlled trial of eicosapentaenoic acid (EPA) and/or aspirin for colorectal adenoma (or polyp) prevention during colonoscopic surveillance in the NHS Bowel Cancer Screening Programme: The seAFOod polyp prevention trial

Research design: A randomised, double-blind, placebo-controlled 2 x 2 factorial trial integrated into the screening and surveillance phases of the National Bowel Cancer Screening Programme (BCSP). Study population: BCSP patients stratified as 'high risk' (>/=5 small adenomas or >3 adenomas with at least one >10 mm in diameter) after adenoma clearance at screening colonoscopy. Exclusion criteria to include regular aspirin use, aspirin/fish hypersensitivity, peptic ulcer disease and colorectal resection. Planned interventions: Enteric-coated eicosapentaenoic acid (EPA) free fatty acid 2 g daily po or placebo (medium-chain triglyceride Migyol 810™) AND enteric-coated aspirin 300 mg daily po or placebo, for 12 months until scheduled BCSP ‘high risk’ surveillance colonoscopy. Proposed outcome measures: Primary efficacy end-point; the number of patients with at least one adenoma detected at BCSP surveillance colonoscopy at 12 months. Secondary end-points; the number of patients with at least one ‘advanced’ (>10 mm diameter, high-grade dysplasia or villous component) adenoma; the total number of adenomas per patient; the number of patients re-classified as 'intermediate risk' (for subsequent three year BCSP surveillance colonoscopy); adverse events (AE) including clinically significant bleeding episodes; erythrocyte and mucosal PUFA incorporation; production of EPA- and aspirin-dependent eicosanoids including prostaglandin E3, 18R-HEPE and resolvin (Rv) E1; analysis of resected adenoma tissue. Assessment and follow up: Compliance and AE monitoring, as well as blood/tissue sampling, all integrated into the BCSP with only one extra visit at 6 months. Colonoscopy and polypectomy complications monitored by the BCSP. Colonoscopic outcomes post-intervention available via the BCSP database. Proposed sample size: 678 individuals randomised equally to the four treatment arms for 80% power, with 5% two-sided significance, to detect a minimum 18% relative reduction in adenoma recurrence (from 60% to 49%) in each two-arm comparison. This is less than the EPA effect observed in a familial adenomatous polyposis (FAP) RCT and below the reduction in polyp number at one year (38%) in previous aspirin polyp prevention trials. Assuming 40% ineligibility (including 20% aspirin use) and a 15% drop-out rate, 1400 'high risk' individuals need to be identified at screening colonoscopy. Statistical analysis: ITT analysis using an ‘at the margins’ approach and direct estimation of relative risk by Poisson regression (with robust standard errors), including treatment arm as an explanatory variable. Predictive value of EPA and eicosanoid biomarkers for adenoma outcomes determined by multiple logistic regression with known clinical factors (eg. adenoma characteristics) as independent variables. Project timetable: 50 'high risk' patients will be identified in each of 15 BCSP Centres per year. Accrual to be completed in 2 years with full primary outcome data available 3 years from the start of randomisation. Six months set-up and six months for clinical and biomarker data analysis are integrated into the four year project.