Introduction In NHS BCSP, high and intermediate risk subjects with colorectal adenomas undergo surveillance colonoscopies. This guideline evidence was derived mostly from general population based studies. This study aims to evaluate the individual and adenoma specific characteristics detected at the index colonoscopy which can predict occurrence of advanced neoplasia during surveillance in a well-defined FOB screening population.

Methods The national BSCP database was interrogated to identify all subjects who participated during the period of June 2006 to July 2012 and completed their first surveillance. The subjects where all the adenomas were retrieved during screening colonoscopy were included. Multivariate analysis was performed to identify the factors which determine occurrence of CRC and advanced adenoma (AA= adenoma with size ≥10 mm/ >25% villous histology/ high grade dysplasia) at surveillance.

Results A total of 17694 high and intermediate risk subjects participated, and 7015 of them completed their first surveillance and were included for analysis. The adenoma specific factors evaluated were high grade dysplasia, villous histology ≥25%, and size ≥10 mm, number of adenomas and any proximal location of adenoma at screening. The individual characteristic evaluated was gender. The outcomes measured at surveillance were detection of CRC, AA, non-advanced adenoma (NAA) and normal finding. Any location proximal to splenic flexure was considered as proximal location for this study.

There were 43 (0.6%) subjects with CRC, 786 (11.2%) with AA, 5566 (79.3%) with NAA and 620 (8.8%) subjects with normal findings during first surveillance. The result of the multivariate analysis was summarised in the table below.

Table 1 showing significant result of multivariate analysis:

Conclusion In contrast with current guidelines, the size of adenomas failed to achieve statistical significance. The number of adenomas, male gender and any proximal location at screening were the important predictors of advanced adenoma during surveillance (table). Future adjustments in the risk stratification strategy for screening population could incorporate these predictors to identify high and low risk cohorts more accurately at screening.

Reference 1 Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002), Gut 2010;59:666–690

Disclosure of Interest None Declared.