Previous studies investigating the associations between NQO1 Pro187Ser polymorphism and breast cancer presented inconsistent results, and most of those studies involved no more than a few hundred breast cancer cases, which is too few to assess any genetic effects reliably. Meta-analysis has been recognized as an important tool to more precisely define the effect of selected genetic polymorphisms on the risk for disease and to identify potential important sources of between-study heterogeneity [27–29]. Hence, we performed this meta-analysis including all available studies to provide the most comprehensive assessment of the associations between the NQO1 Pro187Ser polymorphism and breast cancer risk. Our results showed that the NQO1 is a candidate gene for breast cancer susceptibility. The NQO1 Pro187Ser polymorphism was associated with an increased breast cancer risk among Caucasians (Ser/Pro vs. Pro/Pro: OR = 1.145, 95% CI = 1.008–1.301, P = 0.038; Ser/Ser + Ser/Pro vs. Pro/Pro: OR = 1.177, 95% CI = 1.041–1.331, P = 0.009). Our result is consistent with the previous meta-analysis performed by Yuan et al. .
Given the biochemical properties of NQO1 in protecting cells from oxidative damage and tumor development, this result may be biologically plausible. NQO1 is a key enzyme which catalyzes the two-electron reduction of quinoid compounds into hydroquinones, which reduces and detoxifies quinines and thus protects cells against redox cycling and oxidative stress [30, 31]. Previous study suggested that some variants, especially those in the promoter regions of genes, may affect either the expression or activity levels of enzymes and therefore may be mechanistically associated with cancer risk . The NQO1 Pro187Ser polymorphism is a C-to-T transition in the NQO1 gene which leads to a proline to serine amino acid substitution at codon 187 in the protein [33, 34]. The NQO1 Pro187Ser polymorphism has been found correlated with decreased enzymatic activity of NQO1 and may affect host’s susceptibility to cancer by changing the enzymatic activity of NQO1 [33, 34]. More importantly, the NQO1 Pro187Ser polymorphism has been shown associated with increased risk for many different types of cancers, including colorectal cancer , lung cancer , esophageal cancer , and hepatocellular carcinoma . Our results add new evidence that the NQO1 Pro187Ser polymorphism contributes to cancer susceptibility.
In subgroup analysis stratified by ethnicity, the NQO1 Pro187Ser polymorphism presented a risk factor for breast cancer in Caucasian populations, but not in Asian and Arab subjects. The inconsistent data among the different ethnicities may indicate different effects of the NQO1 Pro187Ser polymorphism on breast risk in different ethnic genetic and environmental backgrounds. Studies reported that NQO1 enzyme not only detoxify carcinogenic compounds  but also bioactivate several kinds of procarcinogen ; thus, decreased activity of NQO1 enzyme may have dual effect on carcinogenesis. There may be some alternative ways in the Asian and Arab populations to detoxify carcinogenic compounds which can effectively compensate for the loss of NQO1 enzyme activity. Nevertheless, owing to the limited number of relevant studies among Asian and Arab populations included in the meta-analysis, the observed negative association between NQO1 Pro187Ser polymorphism and breast cancer risk in Asians and Arabs is likely to be caused by chance because study with small sample sizes may have insufficient statistical power to detect a slight effect or may have generated a fluctuated risk estimate. Currently there are only 3 studies in Asian population and 1 in Arab population for NQO1 Pro187Ser polymorphism and breast cancer risk. Therefore, the negative results of the Asain and Arab populations should be interpreted with caution.
In subgroup analysis according to the source of control, statistical significant increased breast cancer risk was found in population-based studies but not in hospital-based studies. The reason may be that the hospital-based studies have inherent selection biases due to the fact that such controls may not be representative of the study population or the general population, particularly when the genotypes under investigation were associated with the disease-related conditions that hospital-based controls may have. Thus, the use of proper and representative population-based control participants is of great importance in reducing biases in such genotype association studies.
Heterogeneity is a potential problem when interpreting the results of a meta-analysis, and finding the sources of heterogeneity is one of the most important goals of meta-analysis [40, 41]. In the present study, statistical significant between-study heterogeneity among studies was observed when all studies were pooled into the meta-analysis (Ser/Ser vs. Pro/Pro: P
= 0.004; Ser/Pro vs. Pro/Pro: P
= 0.033; Ser/Ser + Ser/Pro vs. Pro/Pro: P
= 0.023; Ser/Ser vs. Ser/Pro + Pro/Pro: P
= 0.007; Table 3). To explore the sources of heterogeneity, we performed metaregression and subgroup analyses. Metaregression analysis of data showed that the genotyping methods, ethnicity, source of controls, quality scores, HWE status, and breast cancer ascertainment were not effect modifiers of heterogeneity. Subgroup analyses stratified by ethnicity, menopausal status, quality score, source of control, and HWE status showed that heterogeneity still existed among Caucasians, low quality studies, population-based studies, and studies consistent with HWE (all P
values < 0.10). Subsequently, we performed Galbraith plots analysis to further explore the source of heterogeneity. Galbraith plots analysis showed that the study Menzel et al.  was the outlier in all genetic models in the overall populations. When excluding the study Menzel et al. , the heterogeneity decreased obviously and all P
values were greater than 0.10 in all genetic comparison models in overall populations, Caucasians, population-based studies, and studies consistent with HWE. However, the summary ORs in the overall populations, Caucasians, population-based studies, and studies consistent with HWE were not material changed by omitting this study, indicating that our results were robust and reliable. The results indicated that the study Menzel et al.  was the major source of the heterogeneity in the meta-analysis.
However, there are still some limitations in this meta-analysis. First, in subgroup analysis by ethnicity, the included studies regarded only Caucasians, Asians, and Arabs for NQO1 Pro187Ser polymorphism. Data concerning other ethnicities such as Africans were not found. Thus, additional studies are warranted to evaluate the effect of this functional polymorphism on breast cancer risk in different ethnicities, especially in Africans. Second, our results were based on unadjusted estimates. We did not perform analysis adjusted for other covariates such as age, obesity, drinking and smoking status, use of contraceptives, environment factors, and so on, because of the unavailable original data of the eligible studies. Third, subgroup analyses were based on studies with relevant information available. Owing to the lack of detailed information in most studies, the subgroup analysis for menopausal status consisted of only two studies for each subgroup, which might not be sufficient to reach a reliable conclusion.