Data in a recent meta-analysis about TNF-alpha polymorphisms and breast cancer susceptibility need for clarification

Abstract

We have read with great interest the article ‘‘TNF-alpha polymorphisms and breast cancer’’ published in April 2011 issue of ‘‘Breast Cancer Research and Treatment’’. Yang et al. [1] identified a total of 17 studies with 44,442 breast cancer patients and 49,926 controls involved, and performed a meta-analysis to examine the association between tumor necrosis factor-alpha (TNF-alpha) gene polymorphisms and breast cancer susceptibility. The results indicated that the TNF-alpha-308, -238, -857, -863, -1031, and -1210 polymorphisms were not associated with the risk for BC in overall populations, whereas a negative association was found between TNF-alpha-308 polymorphism and BC in Caucasian populations. Nevertheless, some methodological issues need to be addressed concerning the data provided in the meta-analysis by Yang et al. [1]. We would like to comment and raise some concerns. First, the ethnicity reported by Yang et al. [1] for the studies Skerrett et al. [2] and Erdel et al. [3] do not seem in line with the ethnicity provided in their original publication. The research by Skerrett et al. was a case–control study, patients were obtained from tumor bank, the ethnicity for cases was as follows: 5.7% unknown, 65.5% White, 18.4% Hispanic, 8.0% Black, and 2.3% Asian. Controls were obtained from cord blood maternal donor samples randomly selected from the DNA banked, the ethnicity for cord bank were as follows: 10% African-American, 7% Asian-Pacific Island, 19% Hispanic, 5% other or mixed, and 61% Caucasian. Similarly, the study population of Erdel et al. contained 2.5% African-American and 2.5% Native American in both cases and controls. These data suggested that the ethnicity of the two studies were mixed population, not Caucasian. Furthermore, the recent meta-analysis that mainly regarding TNF-alpha-308, -238 polymorphisms and BC by Shen et al. [4] did not include these two studies. To address the reason, we have carefully read the original publications, which did not supply the genotype distributions of both cases and controls, this may be the possible explanation for these discrepancies. Yang et al. should indicate the source of the data or rule out these two studies. Second, the study by Mestiri et al. [5] was from Tunisia, Yang et al. [1] categorized the source of population from Tunisia into Arabs. It is generally accepted that both of Tunisia and Iran [6] belong to Arab world. We believed that the source of population from Iran goes more appropriate for the Arabs, rather than the Asian population. M.-B. Chen Department of Medical Oncology, Kunshan First People’s Hospital Affiliated to Jiangsu University, 91 Qianjin Road, Kunshan 215300, Jiangsu, China