Two papers have recently been published in Gut by Boparai et al.1 and Edelstein et al.2 regarding the risks of hyperplastic polyposis syndrome (HPS), also called serrated polyposis. Both suggest a substantial cancer risk and continued development of serrated polyps associated with the syndrome. Previous reports have suggested that the prevalence of hyperplastic polyposis syndrome is relatively low in the general population (1:3000; 0.033%)3 with recent correspondence in Gut from Orlowska4 on this issue suggesting even lower levels (1:100 000). However, we noticed a number of cases during bowel cancer screening colonoscopy based on guaiac faecal occult blood testing (gFOBT) and, therefore, formally assessed this within our screening cohort in an audit as part of service development. Currently, the NHS Bowel Cancer Screening Programme (BCSP) does not offer any surveillance for serrated polyps in line with British Society of Gastroenterology 2010 Polyp Surveillance Guidelines.

We reviewed endoscopic and pathology records for all patients presenting for Bowel Cancer Screening in Oxfordshire between April 2010 (programme start) and January 2012. Three BCSP-accredited endoscopists performed the procedures. Narrow-band imaging or chromoendoscopy was not used routinely. Patients were defined as having HPS if they met either of the two main WHO criteria for HPS: either more than 20 hyperplastic polyps throughout the colon, or five hyperplastic polyps in the proximal colon with two ≥10 mm in size. Patients who were first-degree relatives of HPS patients were not considered as these patients would normally receive screening separately in line with recommendations from clinical genetics services. In total, 755 patients attended for index screening colonoscopy. Five patients met WHO criteria for HPS, of whom three had a synchronous advanced adenoma (see table). The mean age of HPS patients was 65.4 years (range 62–69 years); three male, two female. The prevalence of HPS in our BCSP population was 0.66% (95% CI 0.24% to 1.52%), a 20-fold increase compared with the estimated rate in the general population.

This high prevalence of HPS in our population was unexpected. This may reflect a number of issues that may be specific to bowel cancer screening, specifically: the use of accredited endoscopists; better bowel preparation due to younger mean age, fitness and pre-colonoscopy counselling compared with symptomatic colonoscopy service; improved colonoscope optics with the advent of high-definition instruments (available in our units); greater awareness of colonoscopists about the significance of serrated polyps and hyperplastic polyposis syndrome; and the high likelihood of advanced adenomas in patient with positive gFOBT which are known to be associated with proximal serrated polyps.5 The previous estimate of 1:3000 was based on the result of flexible sigmoidoscopy and may, therefore, be an underestimate as there was no opportunity to systematically examine the right colon. Although the number of cases we report is relatively small, even the lower boundary of our 95% CI would represent a 7-fold increase on previous estimates. As gFOBT-based bowel cancer screening has a ∼2% positivity rate, even if we had detected all cases in our screening population (highly unlikely), our population rate would be at least 1:7550, at least 10-fold greater than that suggested by Orlowska.4

We conclude that HPS appears to be relatively common (1:151) in BCSP patients, and is often associated with advanced neoplasia as per previous reports.5 Detection of a large serrated polyp, or multiple hyperplastic polyps, should alert BCSP colonoscopists to the possibility of HPS where they may wish to augment detection with dye-spray or advanced imaging techniques given the high risk of further polyps or cancer if HPS is diagnosed. BCSP surveillance guidelines for large proximal serrated polyps may need to be reviewed to ensure such patients are not overlooked.⇓