The plasma and cerebrospinal fluid pharmacokinetics of cyclopentenyl cytosine (CPE-C) were studied following i.v. bolus and continuous i.v. infusion in male rhesus monkeys. Following an i.v. bolus dose of 100 mg/m2 plasma elimination of CPE-C was biexponential with a mean t1/2 alpha of 8.4 min, a mean t1/2 beta of 36 min, and a total clearance (CLTB) of 662 ml/min/m2, which is 5- to 10-fold higher than clearance rates in rodents and dogs. Less than 20% of the total dose of CPE-C was excreted unchanged in the urine. The remainder was excreted as the inactive deamination product cyclopentenyl uridine (CPE-U). The ratio of the areas under the plasma concentration versus time curves of CPE-U to CPE-C was 7.0 +/- 2.4 following i.v. bolus CPE-C. The cerebrospinal fluid:plasma ratios of CPE-C and CPE-U were 0.08 and 0.30, respectively. Continuous i.v. infusion of CPE-C was compared to continuous infusion of 1-beta-D-arabinofuranosylcytosine in two monkeys. Steady state plasma concentrations, normalized to a dose of 12.5 mg/m2/h of CPE-C and an equimolar dose of 1-beta-D-arabinofuranosylcytosine, were 2.1 and 0.53 microM, respectively. The steady state concentrations of their corresponding uridine metabolites (CPE-U and 1-beta-D-arabinofuranosyluridine) were 8.2 and 15.5 microM. The rapid elimination of CPE-C by deamination in the primate resulted in a much higher CLTB and considerably lower total drug exposure than in rodents and dogs that clear CPE-C at a much lower rate by renal excretion. These significant interspecies differences in the disposition of CPE-C should be considered in the selection of a starting dose and schedule for human trials and suggest that a pharmacologically directed dose escalation scheme should be used in the planned phase I studies.