LAUSR

Background Emerging data suggest a negative role of cyclooxygenase-2 (COX-2) in colorectal carcinomas (CRC). Investigating this in developing communities such as ours helps to contribute to existing understanding of these lesions. Methods and findings Formalin-fixed paraffin-embedded CRC colectomy tissues and their corresponding non-tumour margins of resected tissues were sectioned and stained with COX-2 antibody. Adenomatous polyp tissues from non-cancer bearing individuals were similarly processed for comparison. COX-2 expression was scored for percentage (< 5% = 0; 6%-25% = 1; 26%-50% = 2; 51%-75% = 3; 76%-100% = 4) and intensity (no staining = 0; yellow = 2; yellowish-brown = 3, brown = 4). Total immunoscore (percentage + intensity score) ≥ 2 was regarded as positive COX-2 expression. Outcome was statistically evaluated with clinicopathological data to determine COX-2 expression-associated and predictor variables. Ninety-five CRC cases and 27 matched non-tumour tissues as well as 31 adenomatous polyps met the inclusion criteria. Individuals with CRC had a mean age of 56.1 ± 12.6 years while those with adenomatous polyps had a median age of 65 years (range 43–88). COX-2 was differentially overexpressed in CRCs (69/95; 72.6%) and in adenomatous polyps (17/31; 54.8%) than in non-tumour tissues 5/27 (18.5%); p < 0.001). The difference in COX-2 expression between CRC and polyps was non-significant (p > 0.065). Tumour grade, advanced pT-stage, tumour-infiltrating lymphocytes, and dirty necrosis were also significantly associated with COX-2 expression (p < 0.035; 0.043, 0.035 and 0.004, respectively). Only dirty necrosis and Crohns-like lymphocytic aggregates predicted COX-2 expression (p < 0.05). Conclusion This study showed a progressive increase in COX-2 expression from normal to adenomatous polyp and CRC tissues, this being associated with poorer prognostic indicators. Although COX-2 appears early in CRC, it may play a secondary role in promoting tumour growth and invasiveness.