Su1083 Predictors of Outcome At Surveillance Colonoscopy Following a Diagnosis of Colorectal Adenoma

Abstract

Background. Post polypectomy surveillance colonoscopies are a huge burden on healthcare systems. Improved risk assessment could result in reduced colonoscopy use. Our aim was to identify patient, polyp and program factors associated with incidence of new neoplasia at first surveillance colonoscopy following an initial diagnosis of colorectal adenomaMethods. We reviewed colorectal cancer (CRC) surveillance outcome data contained in a colonoscopy recall database. Findings at first surveillance colonoscopy in people who had been diagnosed with advanced (high risk, HRA) adenoma ( ≥ 3 polyps, size ≥10mm, villous component, serration, high grade dysplasia) at diagnostic colonoscopy were compared to those with non-advanced (low risk, LRA) adenomas at diagnosis. Data included patient age and sex, polyp features and location, and interval between colonoscopies. Outcomes were compared using bivariate (Fisher's exact) and multivariate (multinomial logistic regression) analyses. Results. Among 457 eligible patients, 266 (58%) were males, and mean age at diagnostic colonoscopy was 63 years. Mean time between diagnostic and surveillance colonoscopy were 43 and 31 months (median 40 and 35 months) for HRA and LRA respectively. There was a significant difference between groups for outcome at surveillance (Table, p=0.014). At the multivariate level, significant associations between outcome at surveillance and polyp features at diagnosis were limited. Controlling for age, gender and colonoscopy interval, those with LRA at diagnosis were more likely to have LRA at surveillance (RRR=1.91, p= 0.023, 95% CI 1.09-3.34). There was a non-significant trend towards increased risk for advanced neoplasia (HRA plus cancers) at surveillance in those originally diagnosed with HRA (RRR=1.31). Age and sex were significant predictors of adenoma incidence at surveillance. Relative risk ratios for HRA increased with each year of age for both males (RRR 1.04, p=0.007, 95% CI 1.01-1.07) and females (RRR 1.04, p=0.019, 95% CI 1.01-1.07), while the relative risk ratio for LRA increased significantly per year of age for males only (RRR 1.03, p=0.007, 95% CI 1.01-1.06). Conclusions: While features of polyps at diagnostic colonoscopies remain key predictors of findings at surveillance, age and sex are additional variables that predict worse outcome. Reduced risk of LRA at surveillance in patients previously diagnosed with HRA may reflect a more conscientious approach by proceduralists at diagnosis. Post polypectomy surveillance could focus more towards older patients. Table. Surveillance outcomes following a diagnosis of low or high risk adenoma.