B149

Background: Oral cancer development is a multifocal and multiclonal process of accumulation of genetic and epigenetic damage in the carcinogen-exposed oral mucosal field. Presence of OPL is a clinical manifestation of the process. Cyclooxygenase-2 (COX-2) is upregulated in many solid tumors and their premalignant precursors. In oral carcinogenesis, COX-2 is progressively overexpressed and is associated with many aspects of the malignant phenotype, including suppression of apoptosis, angiogenesis, inflammation, immune response and invasiveness. Abundant in vitro and in vivo data illustrate that specific inhibition of COX-2 suppresses tumorigenesis and represents a promising strategy for prevention of oral cancer in high-risk patients. Design: A randomized, placebo-controlled phase II study was conducted to evaluate the efficacy of the COX-2 inhibitor celecoxib, in subjects with OPLs by both clinical (reduction by 50% in bidimensional lesion measurements)and histological response (improvement in the degree of dysplasia or resolution of atypia), the safety of the drug, and its effects on biomarkers relevant to the head and neck tumorigenesis process. Subjects were stratified as having either early (presence of at least atypical cells or mild dysplasia) or advanced (moderate and/or severe dysplasia) OPLs at baseline. Subjects were randomized to either placebo, 100 mg of celecoxib bid or 200 mg of celecoxib bid for 12 weeks, and a cohort of 6 patients was treated with 400 mg bid in an unblinded extension of the study. Pre- and post-treatment biopsies after 12 weeks of celecoxib and a biopsy 3 months after completion of study drug intervention were obtained. The Cochran-Mantel-Haenszel test was used for comparison of responses in intervention arms with the placebo arm, p values of <0.05 were considered significant. Results: Fifty six subjects with OPLs were enrolled; 42 were evaluable based on at least 80% compliance with the assigned treatment and presence of both baseline and 12 week evaluation or progression before week 12. Clinical responses, were seen in 5 of 15 , 6 of 12 (p=0.5), 2 of 12 (p=0.3) and 2 of 3 subjects, while histologic responses were seen in 3 of 15, 4 of 12 (p=0.3), 2 of 12 (p=0.3) and 1 of 3 subjects in the placebo, 100 mg bid, 200 mg bid and 400 mg bid arms respectively. Celecoxib was well-tolerated and no major cardiovascular adverse events were observed. Conclusion: Although there were no statistically significant differences in the rate of response between placebo and active intervention, the responses (2/3) in the high-dose celecoxib arm suggest that increasing the dose might result in superior activity. Analysis of biomarkers including baseline levels of COX-2 in relation to response to therapy, might provide additional insights into the drug's activity. Supported by Pfizer.

[Fifth AACR International Conference on Frontiers in Cancer Prevention Research, Nov 12-15, 2006]