Breast cancer is a disease with unique phenotypic manifestations in different racial/population groups. It has been hypothesized that these population-specific characteristics of breast cancer may be partly due to differences in genetic susceptibility to the disease arising from variation in the frequency of different polymorphic alleles of key candidate genes involved in estrogen and xenobiotic metabolism in different populations. Our study was designed to evaluate the hypothesis that the CYP1B1 Val432Leu polymorphisms in the CYP1B1 gene, a key candidate gene involved in phase I hydroxylation of estrogens (17β-estradiol and estrone) to 4-hydroxy catechols might contribute to breast cancer risk in Nigerian women.
Comparison of our data with reports from other populations indicates wide variation in the distribution of the CYP1B1 codon 432 Val → Leu polymorphism across different populations groups. The frequency of the CYP1B1 (Val) allele among control subjects in our study (0.89) is closer to the frequency in African-Americans (0.70)  but much higher than the figures reported in Caucasians (0.42) , Asians in China (0.46) , Japan (0.15) , and Korea (0.11) .
Our results suggest that harboring one CYP1B1 (Leu) allele was significantly associated with breast cancer (OR = 1.59, 95% CI 1.01–2.52). Subgroup analysis based on menopausal status showed that the risk conferred by this polymorphism was essentially restricted to premenopausal women in whom the combination of CYP1B1 (Val/Leu) and CYP1B1 (Leu/Leu) genotypes was associated with over 2-fold increased risk of breast cancer (OR = 2.04, 95% CI 1.10–3.78). This association was not confirmed in postmenopausal women. The associations were not significantly modified by the adjustment of the data for waist/hip ratio (a surrogate measure of etiologically relevant obesity) in either premenopausal or postmenopausal women, despite the finding of preferential 4-hydroxylation of estrogens in obese women on high fat diet .
Several other investigators have evaluated the relationship between CYP1B1 polymorphisms and breast cancer in different populations. There appear to be no significant overall association between the CYP1B1 Val/Leu variant and breast cancer risk in Caucasian populations in a recent meta-analysis . However, a pooled analysis suggests a possible association of both the Val/Leu and Val/Val genotypes with breast cancer in Caucasians but no significant effect was observed in Asians or African-American subjects . Among the seven studies in Caucasians, three reported no association between the CYP1B1 Val432Leu polymorphism and breast cancer risk [4, 10, 14], while three reported a risk effect for the valine allele [9, 11, 13]; one study reported an inverse association between the valine allele and breast cancer . Statistical significance for the association between breast cancer and the CYP1B1 Val/Val polymorphism was reached in two studies [11, 13]. In one of these studies, Listgarten et al.  found that harboring the heterozygous Val/Leu genotype was associated with a 2.15-fold increased risk of breast cancer (95% CI 1.31–3.52) while the homozygous mutant (Leu/Leu) conferred a 3.30-fold increased risk (95% CI 1.76–6.19). In the second study of 84 cases and 103 controls among the Turkish population, Kobacas et al.  reported an overall association between carriers of at least one Val allele and breast cancer risk among women with body mass index (BMI) > 24 kg/m2 (OR = 2.81, 95% CI 1.38–3.74). Although Bailey et al.  failed to demonstrate a significant association between the CYP1B1 Val432Leu polymorphism and breast cancer risk, they noted that Caucasian patients with the Val/Val genotype had a significantly higher percentage of breast cancer that were positive for estrogen receptors (ERs) or progesterone receptors (PRs), suggesting that this polymorphism may be functionally important for the expression of these steroid receptors.
A small hospital-based case-control study involving 59 African-American women found no statistically significant association between the CYP1B1 Val432Leu polymorphism and breast cancer risk . Of three studies in mixed U.S. populations [25–27], one  reported no significant association, while two [25, 27] reported an inverse association between the Val/Val genotype and breast cancer (OR = 0.4, 95% CI 0.1–1.0 and OR = 0.7, 95% CI 0.6–0.9, respectively) and for the Val/Val and Val/Leu genotypes combined (OR = 0.4, 95% CI 0.1–1.0 and OR = 0.8, 95% CI 0.7–0.9, respectively). Meta-analysis of these studies in African-American and mixed populations showed no overall significant risk associated with the CYP1B1 Val432Leu polymorphism in breast cancer susceptibility in these populations .
Meta-analysis of studies on the association between CYP1B1 Val432Leu polymorphism and breast cancer risk failed to demonstrate an overall significant association in Asian women . Of the four available studies, three [15, 16, 26] found no significant association. Only one study in the Chinese population  reported that, compared with those with the Val/Val genotype, women with the Leu/Leu genotype had a 2.3-fold (95% CI 1.2–4.5) elevated risk of breast cancer after adjusting for confounding variables, the positive association between the Leu/Leu genotype and breast cancer was more pronounced in postmenopausal women (OR = 3.1, 95% CI 1.0–9.1) than in premenopausal women (OR = 1.9, 95% CI 0.8–4.3).
The differences in breast cancer risk associated with the CYP1B1 Val432Leu polymorphism in different populations may be due to several reasons including differences in frequency of Val and Leu alleles in different populations, the relatively small sample size of some studies [4, 11, 13, 25, 26], as well as differences in study designs.
The mechanisms through which polymorphisms in CYP1B1 might influence breast cancer risk are not completely known. CYP1B1 is expressed constitutively in extrahepatic tissues including lung and mammary tissue . Although other cytochrome P450 enzymes, such as CYP1A2 and CYP3A4, are involved in hepatic and extrahepatic estrogen hydroxylation, CYP1A1 and CYP1B1 display the highest levels of expression in breast tissue ; CYP1B1 exceeds CYP1A1 in its catalytic efficiency as an estradiol (E2) hydroxylase, and differs from CYP1A1 in its main site of catalysis [5, 29, 30]. CYP1B1 has its primary activity at the C-4 position of estradiol (E2), whereas CYP1A1 has its primary activity at the C-2 position. These two metabolites differ greatly in their carcinogenicity. Treatment with 4-OH-E2, but not 2-OH-E2, induced renal cancer in Syrian hamster [7, 31]. Analysis of renal DNA demonstrated that 4-OH-E2 significantly increased 8-hydroxydeoxyguanosine levels, whereas 2-OH-E2 did not cause oxidative DNA damage . Similarly, 4-OH-E2 induced DNA single-strand breaks whereas 2-OH-E2 had a negligible effect . Comparison of the corresponding catechol estrogen quinones showed that E2-3,4-quinone produced two to three orders of magnitude higher levels of depurinating adducts than E2-2,3-quinone. Significantly higher 4-OH-E2/2-OH-E2 ratios were observed in breast tumor tissue than in adjacent normal breast tissue . These findings suggest a causative role of 4-OH catechol estrogens in carcinogenesis and implicate CYP1B1 as a key player in the process.
To the best of our knowledge, ours is the largest case-control study on CYP1B1 Val432Leu polymorphism and breast cancer risk conducted in African populations.
The epidemiologic literature on determinants of breast cancer in sub-Saharan African populations is scanty despite the evidence that breast cancer is already a public health problem in these developing countries [35, 36] and the burden of the disease is likely to increase as women in these populations adopt Western diets and sedentary lifestyles. The use of hospital controls instead of population-based controls might be a source of systematic bias; poor research infrastructure including the absence of a population-based cancer registry and lack of functional communication facilities limited our choice of recruitment for the control subjects. Recruitment of incident and prevalent cases of breast cancer may also be a source of systematic bias as women with rapidly progressive forms of breast cancer may have died early, leaving us with a subpopulation of less aggressive prevalent cases. This is particularly important given the report of an interaction between the CYP1B1 Val → Leu polymorphism and hormone receptor status noted in Caucasian women . The observed preponderance of hormone receptor positive tumors in individuals harboring the Val/Val genotype would result in selective survival advantage of this genotype. However, such an interaction if present in our study population would result in underestimation of the breast cancer risk associated with the presence of the CYP1B1 (Leu) allele. The non-availability of facilities for hormone receptor assays during the recruitment of study participants in the Nigerian study sites limited our ability to actually evaluate the presence of such interaction if any.