Macrophage migration inhibitory factor (MIF) was the first cytokine identified nearly 50 years ago. At that time its described activity was as a soluble factor produced by T-lymphocytes that inhibited the directed migration of macrophages [1]. Since then, MIF is documented to be expressed by a broad variety of cells and tissues, including immune and non-immune cells [2, 3]. Presently, MIF is considered a pleiotropic cytokine that is a central regulator of innate immunity and a regulator of inflammation since it is constitutively expressed and acts as an upstream regulator of many other inflammatory cytokines [4]. Experimental evidence has shown that MIF plays a critical role in immune and inflammatory diseases including sepsis [5], rheumatoid arthritis [6], delayed-type hypersensitivity [7], Crohn's disease [8] and gastric ulcer formation [9].

Because MIF is an important regulator of inflammation and given the established link between chronic inflammation and cancer [10], recent studies have examined the link between Mif mRNA expression and/or MIF protein levels and cancer [11]. Increased Mif mRNA expression in prostate tissue [12] and increased serum MIF protein levels [13] were first documented for prostate cancer patients. Other investigations corroborated these findings for prostate cancer [14, 15] and extended the association between increased Mif mRNA and/or protein expression to other types of cancer [16–19]. Therefore, strong evidence has accumulated suggesting that MIF is an important link between inflammation and cancer [11]. Two regions within the Mif promoter contain functional polymorphisms that result in variable Mif mRNA amounts and subsequent protein production and high expressing promoter polymorphism genotypes are associated with increased risk of inflammatory disease [20, 21]. A -173 G/C transversion results in a single nucleotide polymorphism (SNP) that was first described with higher frequency in arthritis patients [22, 23]. Another functional promoter polymorphism was identified at position -794, where increasing copies of a CATT tetranucleotide repeat (5 to 8 copies, -794 CATT(5-8)) were also associated with increased disease severity in arthritis [20]. The association between these polymorphisms and disease has been extended to other inflammatory conditions including ulcerative colitis [24], peptic ulcer disease [25] and increased morbidity due to sepsis [26]. These studies thus suggest that an association exists between Mif promoter genotypes that result in increased MIF protein production and an increased risk of inflammatory disease. Similarly, studies examining the association between Mif promoter polymorphisms and cancer have reported an increased risk of cancer for patients with higher MIF protein producing promoter genotypes [3, 27–30] while others have reported a lack of association [31]. The aim of this study is to conduct a meta-analysis of all available studies that examine the association between Mif promoter polymorphism and the incidence of cancer.

Macrophage migration inhibitory factor is a pro-inflammatory cytokine shown to be upstream of and able to upregulate the production of other inflammatory cytokines [4]. Consequently, MIF is a central modulator of inflammation and a possible link between chronic inflammation and cancer [10]. Several studies have examined the relationship between increased Mif mRNA expression and cancer [12–19]. Mif promoter polymorphisms that increase Mif mRNA expression also result in increased MIF protein production [21, 27] and recent studies have established an association between increased MIF serum levels and increased risk for cancer [13, 14, 32]. The association between high expressing functional Mif promoter polymorphisms and prostate cancer appears biologically plausible in view of the documented association between elevated serum MIF protein levels and prostate cancer. MIF's ability to override p53 tumor suppressor activity, sustain MAP (ERK-1/2) kinase phosphorylation and induce the formation of other proinflammatory mediators within the tumor microenvironment [33–36] likely contribute to tumor promotion. The present meta-analysis study is the first, to our knowledge, to critically review published studies on the association between Mif promoter polymorphisms and the incidence of cancer. Our meta-analysis focused on the (-173 G/C) Mif promoter region polymorphism and cancer since very few studies have examined the association between the number of -794 CATT repeats and cancer. From the 5 studies included in our analysis, our results showed that there is a significant association between having any C allele at the -173 site within the Mif promoter and cancer. Because of large heterogeneity in the results, further sub-group analysis showed that there is a strong association between Mif promoter genotypes containing a C allele and the development of solid tumors [OR = 2.67 (1.26-5.65), p = 0.0105], while no association was observed in non-solid tumors [OR = 1.21 (0.95-1.55), p = 0.122]. Lastly, analysis of studies dealing only with prostate cancer patients showed the strongest association between cancer incidence and the C allele Mif promoter genotype [OR = 3.72 (2.55-5.41)]. Thus, Mif promoter genotype may serve as a useful prostate cancer risk marker.

Some limitations of this meta-analysis study should be considered. First, the number of studies included was low, reflecting the small number of published studies. Second, heterogeneity was large in the global analysis, which justified subgroup analysis but made the number of studies in each subgroup even smaller. Given these two limitations, the link between the C allele Mif promoter genotype and prostate cancer awaits validation by future studies. Similarly, whether this link also extends to other solid tumors needs to assessed by future studies. Third, the analysis was based on individual unadjusted OR from each of the studies included, while a better estimate may be obtained by adjusting for potentially contributing factors. For example, Arisawa et al [28] reported an association between the -173C Mif allele and gastric cancer only in patients older than 60 years. Therefore, the inclusion of other factors (such as age) may yield a more accurate estimate of the association. Finally, the focus of the current analysis was limited to the Mif -173 G/C promoter and did not include -794 CATT repeat polymorphisms due to the very small numbers of studies. Two recent studies described an association between higher MIF producing genotypes (greater number of CATT repeats) and gastric [28] and prostate cancer [27] respectively. These observations should be confirmed by additional studies and the relationship between the two functional Mif promoters and carcinogenesis remains to be investigated.

The current meta-analysis results suggest that the -173C Mif promoter polymorphism is associated with an increase in the risk of solid tumor cancer, particularly for prostate cancer. More studies that include a larger variety of cancers are needed to confirm the association of high expressing Mif promoter polymorphisms and prostate cancer.