PLASMA AND CEREBROSPINAL-FLUID PHARMACOKINETICS OF O-6-BENZYLGUANINE AND TIME-COURSE OF PERIPHERAL-BLOOD MONONUCLEAR CELL O-6-METHYLGUANINE-DNA METHYLTRANSFERASE INHIBITION IN THE NONHUMAN PRIMATE

O-6-Benzylguanine (O(6)BG) enhances the cytotoxicity of the nitrosoure as by irreversibly binding and inhibiting the DNA repair enzyme O-6-me thylguanine-DNA methyltransferase (MGMT). The plasma and cerebrospinal fluid (CSF) pharmacokinetics of O(6)BG and its active metabolite, O-6 -benzyl-8-oxoguanine, were studied in a nonhuman primate model after 2 00 mg/m(2) had been injected i.v. The parent drug and the metabolite w ere measured with a reverse-phase HPLC assay, A pharmacokinetic model incorporating separate compartments for O(6)BG and the O-6-benzyl-8-ox oguanine metabolite, first-order conversion of O(6)BG to the metabolit e, and additional first-order elimination rate constants for each comp ound, was simultaneously fitted to the parent drug and metabolite plas ma concentration time data. Elimination of O(6)BG from plasma was rapi d; it had a half-life of 1.6 h and a clearance of 68 ml/min/m(2). On t he basis of the pharmacokinetic model, essentially all of the O(6)BG w as converted to O-6-benzyl-8-osoguanine. The plasma pharmacokinetic pr ofile of the metabolite differed considerably from that the parent dru g. The half-life (14 h) was 10-fold longer and the area under the curv e (2420 mu M/h) was 11-fold higher than that of O(6)BG (212 mu M/h). T he clearance rate of O-6-benzyl-8-oxoguanine was 6.4 ml/min/m(2). The CSF:plasma ratio was 4.3% for O(6)BG and 36% for O-6-benzyl-8-oxoguani ne, and the metabolite area under the curve was 90-fold higher than th at of O(6)BG in CSF. The excellent CSF penetration of the active metab olite provides a rationale for the use of O(6)BG as a chemosensitizing agent for brain tumors, We also studied the duration of MGMT inhibiti on in peripheral blood mononuclear cells. By 2 h after a 200 mg/m(2) d ose of O(6)BG, >98% of MGMT activity was suppressed, and >95% suppress ion of enzyme activity persisted at 18 and 48 h after the dose. By 2 w eeks after the treatment, MGMT levels had returned to baseline, Persis tent high concentrations of the active metabolite appear to provide a pharmacological explanation for the prolonged suppression of MGMT acti vity.