1. ¹Division for Cancer Prevention, National Cancer Institute, 6130 Executive Blvd., Bethesda, MD 20892, USA
2. ²Center for Cancer Research, National Cancer Institute, Frederick, MD 21702, USA
3. ³Advanced Biomedical Computing Center, SAIC, National Cancer Institute, Frederick, MD 21702, USA

Correspondence: Dr IU Ali, Division of Cancer Prevention, National Cancer Institute, 6130 Executive Blvd., Bethesda, MD 20892-7315, USA. E-mail: alii@mail.nih.gov

Received 4 May 2005; Revised 18 August 2005; Accepted 26 August 2005; Published online 4 October 2005.

Abstract

Cyclooxygenase 2 (Cox-2) is upregulated in colorectal adenomas and carcinomas. Polymorphisms in the Cox-2 gene may influence its function and/or its expression and may modify the protective effect of nonsteroidal anti-inflammatory drugs (NSAIDs), thereby impacting individuals' risk of developing colorectal cancer and response to prevention/intervention strategies. In a nested case–control study, four polymorphisms in the Cox-2 gene (two in the promoter, -663 insertion/deletion, GT/(GT) and -798 A/G; one in intron 5-5229, T/G; one in 3'untranslated region (UTR)-8494, T/C) were genotyped in 726 cases of colorectal adenomas and 729 age- and gender-matched controls in the prostate, lung, colorectal, and ovarian (PLCO) cancer screening trial. There was no significant association between the Cox-2 polymorphisms and adenoma development in the overall population. However, in males, the relatively rare heterozygous genotype GT/(GT) at -663 in the promoter and the variant homozygous genotype G/G at intron 5-5229 appeared to have inverse associations (odds ratio (OR)=0.59, confidence interval (CI): 0.34–1.02 and OR=0.48, CI: 0.24–0.99, respectively), whereas the heterozygous genotype T/C at 3'UTR-8494 had a positive association (OR=1.31, CI: 1.01–1.71) with adenoma development. Furthermore, the haplotype carrying the risk-conferring 3'UTR-8494 variant was associated with a 35% increase in the odds for adenoma incidence in males (OR=1.35, CI: 1.07–1.70), but the one with a risk allele at 3'UTR-8494 and a protective allele at intron 5-5229 had no effect on adenoma development (OR=0.85, CI: 0.66–1.09). Gender-related differences in adenoma risk were also noted with tobacco usage and protective effects of NSAIDs. Our analysis underscores the significance of the overall allelic architecture of Cox-2 as an important determinant for risk assessment.