Pilot Study of a National Screening Programme for Bowel Cancer in Norway

• colorectal cancer mortality [ Time Frame: 10 years ] [ Designated as safety issue: No ]
  colorectal cancer mortality after 10 years of follow-up, possibly extending to 15 years of follow-up
  
  

• Colorectal cancer incidence [ Time Frame: 10 years ] [ Designated as safety issue: No ]
  Colorectal cancer incidence after 10 years of follow-up, possibly extending to 15 years of follow-up
  
  
• Complications and quality assurance [ Time Frame: 4 years ] [ Designated as safety issue: Yes ]
  Both screening arms will be subject to continuous registration of complications and quality assurance measures of screening itself and work-up of screening positives
  
  
• Psychological effects of screening [ Time Frame: 5 years ] [ Designated as safety issue: Yes ]
  The level and duration of psychological reactions to screening (anxiety, quality of life) and influence on lifestyle (smoking, physical exercise, dietary habits) will be assessed during the first 4-5 active screening years of the study
  
  

There are several candidate screening modalities - fecal occult blood (FOBT), flexible sigmoidoscopy, colonoscopy, CT and MRI colonography and a range of molecular markers. Of these, only FOBT and FS have been subjected to long-term follow-up in randomised trials (RCTs). These two modalities will be tested in a head-to-head comparison by 1:1 randomisation. Previous studies have suggested that the attendance for FS may be lower than for FOBT. However, participation has been shown to decline with repetitive rounds required for FOBT, while infrequent or once-only screening may suffice for FS. A better test performance for FS makes it uncertain which method may be most beneficial in a public health perspective. This is the first time a national screening programme is designed as a platform for comparative effectiveness studies.

The pilot study will be carried out in two hospital catchment areas in South-East Norway - each with a target population of 70,000 men and women at 50-74 years of age - altogether 140,000 individuals to be randomised 1:1 between screening with an immunochemical test for faecal occult blood (iFOBT) biennially or FS once only. The primary endpoint is colorectal mortality reduction after 10 years. Attendance for FS is expected to be 50% and 60% for iFOBT. Expected CRC mortality reduction is 30% (286 CRC deaths) in the FS arm and 15% (143 CRC deaths) in the iFOBT arm (intention-to-treat). In a 1:1 randomisation with 80% statistical power and a significance level of 5% it will require 65,000 individuals in each arm to disclose a statistically significant difference between FS and iFOBT screening in an intention-to-treat model. We expect 5% in the iFOBTs group to test positive and require colonoscopy work-up. A positive FS is defined as 'any advanced neoplasia' (CRC, adenoma >10mm, adenoma with high-grade dysplasia or villous components). A finding of advanced neoplasia is expected in 5% of FS requiring full colonoscopy.

Study entry-date: First round screening of the iFOBT arm (70,000 invitees) has to be finished in years 1 and 2 of the trial while the flexible sigmoidoscopy arm (70,000 invitees) requires years 1-4 to be completed. Randomization of altogether 140,000 invitees was performed in year 1 of the trial - thus rendering the flexible sigmoidoscopy arm prone to more relevant time-dependent events between randomization and time of screening actually being offered. Therefore, primary entry-date was defined as day of real (for those attending) or suggested appointment (for non-attenders) in the screening arm as all letters of invitation stated a suggested day for appointment chosen by randomization. Primary entry date in the iFOBT arm was similarly defined as date of iFOBT sampling for those attending and date of mailing iFOBT-kits plus one week for non-compliant invitees. Randomization date was chosen as a secondary study entry date to allow comparative analysis of effects of choosing the two entry date definitions.