Cancer risk in endoscopically unresectable colon polyps

Background: The purpose of the current study was to define the rate of underlying malignancy in endoscopically unresectable polyps. Methods: An institutional review board-approved review identified all patients undergoing colectomy for radiologically identified or endoscopically unresectable polyps between 1997 and 2006. Patients were included if the endoscopic impression and biopsy findings suggested an adenomatous polyp without invasive cancer. Patient information was abstracted. Results: Eighty procedures in 79 patients were identified. Median patient age was 66 (range 38 to 85) years, and patients were predominately male (98%) and Caucasian (73%). The median endoscopic size of polyps was 3.0 cm (range 0.8 to 10 cm). Polyps were most frequently proximal to the splenic flexure (72%). Biopsy histology included 36 (51%) tubulovillous and villous adenomas. Invasive cancer was identified in 13 of 80 (16%) subsequent surgical specimens, but only 3 of 71 (4%) were lymph node-positive. Surgical morbidity was 37% and in-hospital mortality 3%. Size of polyp (P = .81) and histologic type (P = .34) were not significantly associated with invasive cancer. Compared with polyps proximal to the splenic flexure, polyps located distally were more likely to harbor malignancy (rate; P < .02), by both univariate and multivariate analysis (odds ratio [OR] 1.38 [95% confidence interval 1.07 to 1.8]). Conclusion: The cancer risk in polyps deemed inappropriate for endoscopic resection was lower than previously reported. Neither polyp size nor histologic type appeared to be significantly associated with invasive cancer. Location of an endoscopically unresectable polyp distal to the splenic flexure confers an increased risk for occult malignancy.