Clinical-endoscopic relevance of incidental colorectal lesions detected by PET-CT

to determine the proportion of incidental colon lesions detected by PET-CT and their correlation with the endoscopic and histological findings. In addition, to determine the maximum standardized uptake value (SUVmax) that can discriminate between benign and malignant lesions in our series of cases.

this was a retrospective study of 3,000 patients evaluated by PET-CT for staging or response to treatment of primary neoplasms, between 2011 and 2015. Patients with incidental uptake in the colon were included in the study. Exclusion criteria included an incomplete, poorly prepared or abandoned colonoscopy, inflammatory bowel disease and treatment with metformin.

the study cohort comprised 71 patients evaluated by PET-CT and subsequently analyzed by endoscopy; 69% were male with a mean age of 65.77 ± 11.2. The rate of incidental colon lesions found by PET-CT was 1.73%, with 52 incidental colonic uptakes reported in 50 patients. The location of the uptake was the rectum (19.23%), sigmoid colon (34.62%), descending colon (13.46%), transverse colon (1.9%), ascending colon (19.23%), cecum (9.62%) and ileocolic anastomosis (1.92%). Thirty-five pathological colonoscopies (71.15%) were identified: the findings included five neoplasms (13.51%), two inflammatory lesions (5.4%) and 30 adenomatous polyps (81.1%). Significant differences were found between neoplastic SUVmax (11.7 g/ml; p = 0.03) and polyps (9.26 g/ml; p = 0.04) in relation to inflammatory lesions and normal endoscopies (6.05 g/ml). There were no differences in terms of the size of the polyps, nor the presence or absence of high grade dysplasia (p = 0.12 and 0.33). Both PET-CT and endoscopy proved consistent for locating lesions (k 0.90; CI 95% 0.86-0.93).

there is a good correlation between the findings identified by PET-CT and the endoscopic study. In our study, a SUVmax > 11 g/ml suggests a malignant pathology, which aids the prioritization of an endoscopic study.