Abstract
Cytarabine (Ara-C) is a prodrug requiring conversion to its 5′-triphosphate (ara-CTP) form for antileukemic effect. Alteration of deoxycytidine kinase (DCK), a rate-limiting enzyme in this conversion, is associated with ara-C resistance. We sought to identify and determine the functional and clinical significance of DCK genetic variants. Sixty-four genetic polymorphisms were identified, including 3 coding changes (Ile24Val, Ala119Gly, and Pro122Ser amino acid substitutions), in subjects of European or African ancestry. The activity of recombinant DCK-24Val, DCK-119Gly, and DCK-122Ser was 85±5%, 66±3%, and 43±4%, respectively, that of wild-type (WT) DCK. These 3 DCK mutants also demonstrated altered substrate kinetics. Lymphoblast cell lines derived from subjects heterozygous for these coding changes had DCK activity significantly lower than that of homozygous-WT lymphoblasts (176.7±15 vs. 340.8±19 pmol CdA-MP/min/mg protein; p=0.0004). Moreover, for a 3′UTR 35708T>C SNP, the C allele was significantly associated with lower DCK mRNA expression in lymphoblast cell lines (p=0.02). To determine the clinical implications of DCK SNPs, we screened pediatric AML patients treated on the St Jude AML 97 protocol (n=55) for the 3 coding variants and the 3UTR SNP. Patients were randomly assigned to receive ara-C as either a short daily infusion (500 mg/m2/dose intravenously over 2 hrs daily for 5 days) or a continuous infusion (500 mg/m2/day as a continuous infusion over 5 days). Bone marrow was collected at the end of the ara-C infusion on day1 for patients receiving the short daily infusion (n=27), and at 10 hours after the start of the infusion for those receiving the continuous infusion (n=28). Ara-CTP levels in leukemia cells were analyzed by HPLC. Intracellular accumulation of ara-CTP was significantly higher in patients given a short daily infusion than in those given a continuous infusion (0.56±0.5 vs. 0.33±0.31 nmol ara-CTP/2x107 leukemia cells; p = 0.014). The inter-patient variability of blast ara-CTP concentration was 40-fold (0.06–2.43 nmol/2x107 cells) when ara-C was administered as short infusion and 101-fold (0.012–1.22 nmol/2x107 cells) when given as a continuous infusion. Karyotype, sex, age, and ethnicity were not significantly associated with leukemia-cell ara-CTP level. An Ala119Gly or Pro122Ser non-synonymous polymorphism was present in 1 patient each; the patient heterozygous for Ala119Gly (continuous infusion arm) had the lowest intracellular concentration of ara-CTP (0.012 nmol/2x107 cells; Figure). Both of these patients had relapses, despite the presence of favorable cytogenetic abnormalities [inv16 and t(8;21)]. The 3′UTR 35708 T>C SNP was significantly associated with lower intracellular ara-CTP concentration in patients receiving ara-C as a continuous infusion (35708T allele 0.5±0.3 vs. 35708C allele 0.22±0.17 nmol/2x107 cells; p= 0.04; Figure). These results suggest that genetic polymorphism of DCK influences its activity and expression and may underlie interpatient variability in intracellular ara-CTP concentration, which in turn can influence treatment response.
Influence of DCK SNPs on leukemic blast ara-CTP levels in AML patients receiving ara C-as continuous infusion
Influence of DCK SNPs on leukemic blast ara-CTP levels in AML patients receiving ara C-as continuous infusion
