Background: Endoscopic polypectomy significantly reduces the incidence of colorectal cancer, but recurrence rates are high, especially for adenomas with advanced histology. The present guidelines recommend re-colonoscopy 3 to 5 years later. Due to limited resources, more precise predictions of adenoma recurrence are required.
Design: Lesions from 109 patients with colorectal adenomas recruited into a randomized, placebo-controlled chemoprevention trial with mesalazine were included. Formalin-fixed paraffin-embedded tissue sections were stained for ß-catenin, cyclooxygenase-2 (Cox-2), and p53 and scored. Adenoma recurrence rates were recorded after 3 years and associated with clinical and immunohistochemical parameters by contingency table analysis.
Results: After 3 years, adenomas recurred in 51.4% of patients. Out of 109 adenomas, 95 met at least one criterion of advanced adenoma (size >1 cm, villous histology, high-grade intraepithelial neoplasia). There was no influence of age, sex, size or villous histology on adenoma reappearance, whilst the number of adenomas at baseline was positively associated with recurrence (p = 0.003). In contrast, ß-catenin nuclear localisation, Cox-2 expression and p53 nuclear expression were significantly associated with adenoma recurrence after 3 years (ß-catenin: p = 0.002; Cox-2: p = 0.001; p53: p = 0.001). Combining these three markers led to a negative predictive value of 88.5% and a sensitivity of 94.6%. (OR = 13.54) CONCLUSIONS: Scoring each single parameter and, more strongly, the combination of all three parameters of the expression of ß-catenin, Cox-2 and p53 in colorectal adenoma tissue may be a useful negative predictor for adenoma recurrence in patients with advanced colorectal adenomas.