Surgical rather than endoscopic resection of early-stage colorectal cancers promotes excessive imaging surveillance.

Background And Aims: Imaging surveillance after curative resection of colorectal cancer (CRC) is debated, particularly in cases of early-stage CRC. The aim of this study was to retrospectively analyze whether and how patients with screened stage 0 and stage 1 CRC were monitored by imaging.

Methods: A cohort of patients with stage 0 (intramucosal) or stage 1 (T1N0) CRC detected from 2003 to 2015 through the French national screening programme was included. All imaging findings were recorded. Statistical analyses were performed for the entire cohort (n = 450) and separately for the two groups (stage 0 n = 268, stage 1 n = 182). Factors associated with imaging surveillance, including the patient's referring gastroenterologist, were determined by logistic regression.

Results: A total of 450 patients were followed up for 6.6 ± 3.9 years. Imaging surveillance was performed for 159 (35.3%), more often for those with stage 1 (66.5%) than stage 0 (14.2%) tumours (p < 0.0001). Within the stage 1 group, 17 of the 47 patients (36.2%) treated by local (endoscopic or surgical transanal) resection alone were followed up by imaging monitoring. Factors significantly associated with surveillance in the entire cohort were the gastroenterologist assigned to the patient (p < 0.0001) and surgical vs endoscopic resection (OR = 39.0, p < 0.0001). The histological risk of lymph node metastasis was not significantly associated with imaging monitoring for stage 1 patients. Of the 5 patients who developed distant metastasis during follow-up, one was diagnosed through imaging surveillance.

Conclusion: This study demonstrates excessive imaging surveillance for early-stage cancers. The use of surgical over endoscopic tumour resection could promote unnecessary surveillance.