Glutathione S-transferase P1 Ile105Val polymorphism and breast cancer risk: convergence and divergence of the two recent meta-analyses

Abstract

We read with great interest the recent meta-analysis by Lu et al. [1], which has reached important conclusions about the association between glutathione S-transferase P1 (GSTP1) Ile105Val polymorphism and breast cancer risk. Nevertheless, some methodological issues need to be addressed concerning the meta-analysis by Lu et al. [1]. Importantly, three sizeable eligible studies [2–4] (2,204 cases and 2,530 controls) have not been included in the metaanalysis, even though they satisfied the search criteria. In addition, close inspection of the data provided by the authors (Table 1) revealed some issues that are worth discussing, so as to provide the scientific audience with an accurate presentation of the underlying data. Specifically, the data reported by Lu et al. [1] for the studies by Maugard et al. [5], Krajinovic et al. [6], Unlu et al. [7] and Torresan et al. [8] do not seem in line with the data provided in the original publications. The genotype frequencies for Ile/Ile, Ile/Val, Val/Val in cases and controls should read: 99-101-20-81-90-25 for the study by Maugard et al. [5], 69-55-5-91-73-13 for the study by Krajinovic et al. [6], 2826-11-51-37-20 for the study by Unlu et al. [7] and 54-3513-77-22-3 for the study by Torresan et al. [8] (italics denote the discrepancies between Lu et al. [1] and the original publications). The above may imply that the original odds ratios for the aforementioned studies may significantly differ from those calculated by Lu et al. [1]. It can be also secondarily noted that the selection of controls in the study by Lee et al. [9] is not hospital-based, given that Lee et al. [9] clearly states that ‘‘controls were randomly selected from the general population of Shanghai’’. As a result, the studies by Lee et al. [9] and Egan et al. [10] seem to have been performed on mutually overlapping populations; consequently the smaller study [10] should have been excluded from the analysis. Despite the above, the results reported by Lu et al. [1] confirm our previous meta-analysis [11] and essentially expand them among the subset of hospital-based studies. Discussing methodological issues of meta-analyses [12, 13] may well elaborate and substantiate their original results.