OBJECTIVES: Induction of cyclooxygenase-2 (COX-2) by inflammatory mediators, oncogenes, and carcinogens has been demonstrated in preclinical models. However, there are limited clinical data regarding COX-2 induction by chemotherapy or radiation. Experimental data suggest cross-talk between the EGFR and COX-2 pathways. The aim of this study was to analyze the expression of COX-2 before and after chemoradiation (CRT) and correlate the same with tumor (T) down-staging and survival. Similar data were obtained for EGFR expression before and after chemoradiation. METHODS: Archival paraffin-embedded tumor specimens from patients undergoing CRT between 1995 and 2001 were analyzed. COX-2 expression was measured by immunohistochemistry (IHC), using the 160112 COX-2 mouse monoclonal antibody. For EGFR, we used mouse monoclonal Ab-10. Standard immunoperoxidase technique was used to detect the avidin- biotin peroxidase complex. Staining in tumor tissue was visually scored and confirmed by an image analyzer (ACIS; ChromaVision Medical Systems, Inc, San Juan Capistrano, CA). RESULTS: Twenty pretreatment biopsy samples from rectal cancer patients and their paired, post-CRT surgical specimens (n = 17) were analyzed. Three cases had no primary tumor after CRT. COX-2 expression was noted in 19 of 20 pretreatment samples and 17 of 17 surgical specimens. EGFR expression was noted in 10 cases pretreatment. Six patients with weakly positive COX-2 expression pretreatment had increased COX-2 expression after CRT, whereas in 1 patient the expression decreased after CRT. No EGFR induction was noted. There was no statistical association between EGFR and COX-2 expression in this data set. Median survival for the entire cohort was 38.9 months. There was no difference in survival between the COX-2 induced and noninduced groups. CONCLUSIONS: COX-2 induction was seen with CRT in this population of rectal cancer patients. Prognostic significance of this induction remains to be defined in a larger cohort.