Purpose. GTI 2040, a 20-mer phosphorothioate oligonucleotide targeting a R2 subunit of ribonucleotide reductase (RNR) mRNA, is currently in a phase I trial in combination with high-dose cytarabine in treatment of patients with refractory or relapsed acute myeloid leukemia (AML). The purpose of this study was to characterize the plasma pharmacokinetics and metabolism of GTI 2040 in patients and determine its levels in red blood cells (RBC), blood (PBMC) and bone marrow (BM) mononuclear cells. Methods. A validated ultra-sensitive and specific hybridization-ligation ELISA method previously developed in our laboratory was used. Plasma, urine, PBMC, and BM samples were obtained from 11 patients receiving continuous infusion (CIVI) of GTI 2040 with 5mg/kg/day for 6 days under an institutional approved protocol. Metabolites were characterized by electrospray LC-mass spectrometry. Results. Steady-state concentrations (Css) in plasma were achieved rapidly with the mean value of 396.1 ± 200.1 nM, which are in the range found active in down-regulation of RNR mRNA in vitro. Thereafter, the plasma concentrations decayed bi-exponentially with a short half life of 0.65 ± 0.24 h and a long half life of 25.0 ± 10.3 h. RBC uptake of GTI2040 was <5%, when the mean Css level was compared with that in plasma, and after end of infusion, RBC GTI 2040 levels declined essentially in parallel to those of plasma. The RBC drug levels remained significantly lower than those in plasma throughout. Drug levels in PBMC were found to be 16.2 ± 10.6, 29.6 ± 23.2 and 26.1 ± 20.8 nM on days 1, 2, and 3 during infusion, and sustained at 20.9 ± 14.9, 14.4 ± 10.6 nM on days 7 and 8 after removal of dose. Css ratio between PBMC and plasma increased 14-fold between day-1 and day-7 and from 4 hr post-infusion, PBMC drug levels remained 2.7-fold higher than that in plasma. Bone marrow uptake of GTI 2040 was evidenced with drug levels of 249 ± 203 nM on day 1 and 335 ± 156 nM on day 5. Urinary excretion was low with <0.1% dose recovered in 24 h. Plasma protein binding of GTI 2040 in human was high, >99.9% at concentration range below 1 μM. Five chain-shortened metabolites in patients plasma samples were identified as 3’ N-1, 3’N-2, 3’N-3, 3’N-4 and 3’N-5 GTI 2040 based on their molecular weights and MS/MS (MS2) spectra. No metabolite was found in urine up to 250x enrichment. Conclusion. Pharmacokinetics of GTI 2040 in plasma was well described by a 2 compartment infusion model. Sustained drug levels in BM higher than those in plasma were achievable in AML patients. Chain shortened metabolites, whose pharmacological roles yet to be determined, were detected in plasma but not in urine. Pharmacokinetics-pharmacodynamics (PK-PD) correlation will be sort in the future. Supported by Grant R21 CA105879 from NCI.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]