In Response: We appreciate the opportunity to respond to the concern raised about the GSTM1 and GSTT1 genotyping methods we used for our recent publication on the relationship between genes in the catechol estrogen metabolism pathway and breast cancer risk. Specifically, Dr. Parl questioned our use of an assay that did not allow for delineation between wild-type homozygotes (+/+) and heterozygotes (+/−). Although we cannot completely rule out the possibility of an increased risk of breast cancer in heterozygotes, it seems unlikely given that previous studies have observed an identical risk (for GSTT1) or nearly identical risk (for GSTM1) between women with the homozygote variant (−/−) and other women (1, 2), a dichotomy (based on the relationship between genotype and phenotype; refs. 3-6) is the one for which a difference in risk would be the most plausible.
It is conceivable, however, that the risk of the −/− genotype more closely approximates the risk of the heterozygote, and that combining the +/− and +/+ genotypes could obscure potential differences between the +/+ and −/− genotypes. In the two reports cited by Parl as examples of the advantage of more refined genotyping, one study observed that colorectal adenoma risk associated with the +/− genotype was identical to that of the −/− genotype (7), whereas the other study observed that breast cancer risk associated with the −/− genotype more closely resembled that of the +/+ genotype, (2) neither of which provided support for the hypothesis of a gene-dosage relationship. Nonetheless, we acknowledge that epidemiologic studies that can distinguish the heterozygotes from the wild-type homozygotes do offer the ability to uncover patterns of association that could differ from those identified to date through the many studies, including ours, which have not had this level of detail.