studied 2 SNPs in the DHFR gene, taking 4 criteria into consideration in the selection procedure (as defined in the Patients and Methods section). These criteria were minor genotype frequency of Ο³10% or more, validated SNP, SNP causing nonsynonymous amino acid change, and indications for clinical relevance determined from previous publications. Since there were no nonsynonymous SNPs reported for the DHFR gene, we focused on the validation status, allele frequencies, and indications for clinical relevance of the SNPs. Our study population included only 205 patients, which made use of the minor genotype frequency to perform association analysis of even greater importance in order to increase power. Although for DHFR 829CΟΎT more is known about the functional nature of the SNP, the frequency of the homozygous mutant genotype is Ο³5%; thus, availability of a very small number of mutant allele carriers for performing association analysis with this SNP was to be expected. In addition, its functional effects were studied in vitro using leukemia cells of Japanese origin, which are distinct (3). Nevertheless, we agree with Dr. Ranganathan that an understanding of the functional nature of the variant is worthwhile and may improve the interpretation of our results. Although the functional consequences of the SNPs in the DHFR gene are unknown, their selection in our study is justified from a hypothesis-generating perspective, with an acceptable validation status of the SNP and a reasonable minor genotype frequency.
Second, MTX or its polyglutamated forms alter the intracellular balance of reduced folates through the inhibition of DHFR, TYMS, and ATIC. As a consequence, the inhibition of these enzymes results in the indirect inhibition of AMPD1 and ADA (4,5). We agree with Dr. Ranganathan that it is preferable to describe MTX as indirectly influencing MTHFR and RFC. Indeed, the activity of these enzymes is altered due to changes in the intracellular reduced folates.
Finally, in response to Dr. Ranganathan's questions regarding the numbers of patients included in the toxicity analyses, we provide more detail herein. Briefly, all patients, including those who were rechallenged, remained in the study. However, genotype comparisons were performed between patients with and those without adverse drug events. Therefore, the number of adverse drug events at each time point had no influence on the association.