INTRODUCTION. TP53 is a common tumor suppressor gene implicated in the pathogenesis of many cancers. The TP53 codon 72 polymorphism is suggested to play a crucial role in modulating apoptotic capacity. However, studies have not established whether variation in TP53 is associated with breast cancer risk. METHODS. We examined TP53 in relation to breast cancer risk using oral mucosal DNA collected from mouthwash samples donated by women aged 20-69 enrolled in a population-based case-control study in three U.S. states. A total of 1,652 invasive breast cancer cases and 1,492 population controls, all white (98% of the study population), were included in this analysis. We genotyped for all 11 TP53 single nucleotide polymorphisms (SNPs) with minor allele frequency >0.03 identified in a re-sequence project in 94 healthy Norwegian women (http://snp500cancer.nci.nih.gov). These included the TP53 codon 72 polymorphism (Arg72Pro; rs1042522), 7 intronic SNPs, and 3 SNPs residing in the 3’UTR. Multivariate logistic regression was used to obtain age- and state of residence-adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for individual SNPs. Haplotypes were estimated using the expectation-maximization algorithm, and the overall association with breast cancer risk was assessed using a global score test. RESULTS. For all women combined, none of the individual SNPs nor 8 common haplotypes (frequency >0.01) observed in this population were significantly related with breast cancer risk (P-value for global score test=0.79). While the global score tests among postmenopausal (P=0.79) and premenopausal women (P=0.16) were also not significant, two individual SNPs were related with premenopausal breast cancer risk. An inverse association for TP53 codon 72 Pro/Pro homozygotes was observed for risk of premenopausal (OR 0.53, CI 0.33-0.85 vs Arg/Arg) but not postmenopausal (OR 1.14, CI 0.76-1.70) breast cancer (P-interaction with menopausal status, P=0.05). A reduced risk of premenopausal breast cancer was also suggested for the homozygous minor allele in a noncoding SNP (rs1642785; OR 0.62, CI 0.40-0.97). This SNP was not related to postmenopausal breast cancer risk (OR 1.24, CI 0.83-1.84; P-interaction with menopausal status, P=0.07). CONCLUSION. Results suggest that the common haplotypes for TP53 estimated in this white U.S. population are not significantly associated with breast cancer risk, overall or among subgroups according to menopausal status. However, two SNPs including the functional Arg72Pro polymorphism deserve further evaluation among premenopausal women.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]