Background UK and USA surveillance guidelines recommend 3-yearly surveillance for intermediate and higher-risk groups, respectively. To date, no study has examined surveillance needs in this group, which comprises nearly half of patients with adenomas.
Aims and objective(s) To identify the optimum frequency of surveillance and assess whether there is substantial heterogeneity in risk; to examine the risks of cancer and advanced adenomas (AA).
Design and Setting Retrospective, multi-centre cohort study, involving a hospitals dataset drawn from 17 UK NHS hospitals (n = 11,944), and three pooled screening cohorts (n = 2,353).
Subjects Patients with intermediate-grade adenoma (s) defined as having 3–4 small adenomas (<10 mm), or 1–2 adenomas, at least one of which is large (≥10 mm).
Primary outcomes AA and colorectal cancer (CRC) detected at the first and second follow-up visits, and CRC incidence after baseline and first follow-up.
Results Among 4,608 patients with follow-up in the hospital dataset, an increase in interval length was associated with a significant increased odds of AA and CRC at the first follow-up (p < 0.001). Of 1,635 patients attending a second follow-up, a significant association was also found between interval and odds of advanced neoplasia at the second follow-up (p = 0.026).
Among 11,944 patients, 168 CRCs occurred during 81,442 person-years of observation time after baseline (206 per 100,000 pyrs, 95% CI 177–240). A single surveillance visit conferred a considerable reduction in risk of CRC after baseline (p = 0.0001). Other independent predictors of CRC were used to devise higher (HIR) and lower (LIR) intermediate risk subgroups, between which there was substantial heterogeneity in risk. A single surveillance exam lowered risk in the HIR subgroup (n = 9,265); however the benefit of surveillance in the LIR (n = 2,679) was unclear.
1,828 intermediate risk patients with at least one follow-up in the pooled screening cohorts were younger, on average, than the hospital cohort. No association was found between findings at follow-up and interval, however, there was evidence of the benefit of surveillance and the LIR and HIR subgroups derived from the hospital dataset were discriminant of CRC risk in the screening participants.
Conclusion A surveillance interval of three to four years seems suitable for the majority of intermediate-risk patients. Surveillance lowers future risk of CRC in intermediate risk patients; however there was heterogeneity in risk and surveillance needs, which suggested that a single follow-up may suffice in certain intermediate-risk patients.
Disclosure of interest None Declared.