Search in the Abstract Database

Search Abstracts 2011

* = Presenting author

P091. Longstanding ulcerative colitis related dysplasia – COX-2 and ultrastructural correlations

O. Fratila1, C. Craciun2, O. Straciuc1, T. Ilias1

1University of Oradea, Oradea, Romania; 2Electron Microscopy Center, “Babes-Bolyai” University, Cluj Napoca, Romania

Aim: Because longstanding ulcerative colitis (LUC) is associated with increased risk of developing colonic cancer through a dysplasia-carcinoma sequence and dysplasia usually precedes colorectal carcinoma it is very important to identify patients at risk, by obtaining clear, earlier evidence of dysplastic mucosal changes. Therefore the aim of our paper was to evaluate the usefulness of COX-2 immunohistochemical marker for routine surveillance of the development of dysplasia using ultrastructural control.

Methods: We studied by ultrastructural and immunohistochemical examination 75 patients with LUC. During surveillance colonoscopy (performed with an Olympus Exera CLE 145 videoendoscope – Olympus, Japan), multiple random biopsies from patients with longstanding ulcerative colitis were taken in all 4 quadrants from each 10 cm of the macroscopically affected colon. The biopsies were contrasted with acetate-uranyl and lead-citrate and studied with a a TESLA BS500 transmission electron microscope. We used monoclonal antibody to COX-2 (Dako, Carpinteria, USA) to perform immunohistochemical staining by a standard avidin-biotin method with 3,39-diaminobenzidine as the chromogen and nickel chloride enhancement. The degree of staining was assessed using criteria previously described, namely a scoring system with staining intensity graded as 1-weak, 2-moderate or 3-strong, and the positively stained area as 0, 1, 2, and 3. Total scores of ≥3 were defined as positive and those <3 as negative. A single observer assessed all sections blindly.

Results: COX-2 was positive (total score ≥3) in 73.3% of the cases, revealing no consistent localization to crypt base, mid-crypt, or surface. The staining intensity showed moderate to strong cytoplasmic overexpression in all of the dysplastic biopsies, regardless of grade. In regenerative atypia cases COX-2 expression was negative and as the pathologic process progressed towards dysplasia/malignant transformation, COX-2 expression became positive with a progressive increase of the total score. Ultrastructural parameters for mucosal remodeling correlate well with UC duration, indicating accumulation of structural alterations. We observed that the main ultrastructural changes seeming to indicate a degeneration process which leads to premalignant lesions were the enlarged, irregularly shaped nuclei and changes in the relationship of adjacent cells. In these specimens the immunohistochemical stain showed a COX-2 expression present in most of the cells (>90%) in LGD and HGD. Irrespective of the intensity of the inflamatory process all regenerative atypia were COX-2 negative.

Conclusions: In our study we observed that COX-2 expression correlates well with ultrastructural changes of dysplasia suggesting that this technique could be trusted for the evaluation of colorectal mucosa in order to improve diagnostic accuracy and to appreciate the risk of malignant transformation. A combination of enhanced colonoscopic surveillance and IHC markers that are more sensitive for dysplasia might be the optimal way to manage the increased CRC risk in these patients. Further studies to find additional prognostic parameters will provide valuable insights into the behavior of LUC.