Most patients experience severe hematological toxicities during treatment with gemcitabine; thus, preventing such toxicities would improve the treatment effects and patients’ quality of life. We analyzed 13 polymorphisms in the transporters, metabolizing enzymes, targets and genes involved in DNA damage and the folate pathway among 132 patients treated with gemcitabine and studied their association with the severity of the hematological toxicities. Single-locus analysis showed that the SNPs RRM1 rs12806698 and rs11031918 and DCTD rs7663494 were significantly associated with severe neutropenia; hENT1 rs760370, hCNT3 rs7867504 and rs4877831 were associated with severe leucopenia; CDA rs2072671, DCTD rs7663494 and WEE1 rs3910384 were associated with severe anemia; and MTHFR rs1801133 was associated with severe thrombocytopenia after stringent Bonferroni correction (P<0.0038). The gene-gene interaction analysis identified the overall best models, including a two-way interaction model (hCNT3 rs7867504 and dCK rs12648166) for severe leucopenia (P = 0.0022) and a three-locus model (CDA rs207671, DCTD rs7663494 and WEE1 rs3910384) for severe anemia with a strong synergistic effect (P = 0.0001). The association with hematological toxicity was further strengthened by the results of a haplotype analysis, wherein the homozygous genotype combination of rs3910384 CC, rs2072671 AA, rs12648166 GG, rs7867504 CC and rs7663494 TT conferred high genetic susceptibility to severe thrombocytopenia. Our results suggested that the gene-gene interaction of gemcitabine metabolic pathway genes and WEE1 contributes to susceptibility to gemcitabine-induced hematological toxicity. Moreover, we proposed a promising data mining analysis approach (generalized multifactor dimensionality reduction, GMDR) to detect and characterize gene-gene interactions. This article is protected by copyright. All rights reserved.
This article is protected by copyright. All rights reserved.

Author