Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

2'--Fluoro-2',3'-dideoxyadenosine (lodenosine, F-ddA)² is an experimental anti-AIDS drug that is currently undergoing adult and pediatric Phase I clinical trials at the National Cancer Institute. This synthetic nucleoside was rationally designed to have improved chemical and enzymatic stability compared with 2',3'-dideoxyadenosine (ddA), its parent compound, or ddA's major metabolite, 2',3'-dideoxyinosine (didanosine or ddI; Marquez et al., 1987). The addition of an electrophilic fluorine in the 2' position of the dideoxyribose ring (Fig. 1) yields a compound that is acid stable and has a greater than 90% reduction in adenosine deaminase (ADA)-catalyzed hydrolysis compared with ddA (Marquez et al., 1990). F-ddA is a reverse transcriptase inhibitor with a mechanism of action similar to other dideoxynucleoside analogs (didanosine, zidovudine, lamivudine, zalcitabine, and stavudine) currently approved by the Food and Drug Administration to treat HIV infection (Food and Drug Administration, 1998).

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Chemical structures of F-ddA, its major metabolite, F-ddI, and 2-chloroadenosine (2-Cl-A), the internal standard.

F-ddA has two potential advantages over didanosine for treating AIDS. Its acid stability makes it a much better candidate for oral administration. Dog studies have demonstrated better bioavailability compared with didanosine (Stoltz et al., 1989). Also, F-ddA is more lipophilic than didanosine (Barchi et al., 1991), which may increase its ability to cross the blood-brain barrier. Penetration into the central nervous system (CNS) is critical in the treatment of patients with AIDS-related dementia (Gallo et al., 1987).

F-ddA pharmacokinetics has been studied previously in other animals, and the CNS pharmacology of a variety of nucleoside analogs has been investigated in our nonhuman primate model, but F-ddA has not been studied in primates. In earlier studies, F-ddA, administered to rats as a 2-h i.v. infusion, was rapidly converted to 2'--fluoro-2',3'-dideoxyinosine (F-ddI) with a plasma-concentration time profile exhibiting biexponential elimination (Singhal et al., 1996). Other studies investigated the effects of halo-substitutions on the cerebrospinal fluid (CSF) penetration of 2',3'-dideoxyguanosine (ddG) in monkeys. Administration of these halo-substituted compounds produced higher CSF to plasma ratios of ddG exposure compared with ddG administration; however, the prodrugs themselves were present at very low concentrations in the CSF so the mechanism of the improved penetration is unclear (Hawkins et al., 1995).

The purpose of the present study was to determine the pharmacokinetics and urinary excretion of F-ddA and its major metabolite, F-ddI, in a nonhuman primate. In addition, the extent of penetration into the CNS was investigated using the ratio of CSF area under the curve (AUC) to plasma AUC as a measure of the relative exposure to these compounds. The results from these experiments are compared and contrasted with those of other nucleoside analogs as well as data from various animal models.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Reagents. F-ddA (NSC-613792, lodenosine) and F-ddI (NSC-616290) were supplied by the Pharmaceutical Resources Branch, National Cancer Institute (Bethesda, MD), and the ADA inhibitor, 2'-deoxycoformycin (NSC-218321) was obtained from the Drug Synthesis and Chemistry Branch, National Cancer Institute (Bethesda, MD). The internal standard, 2-chloroadenosine, was purchased from Sigma Chemical Co. (St. Louis, MO). HPLC grade methanol, acetonitrile, and water, as well as certified 1.00 N sodium hydroxide solution, were purchased from Fisher Scientific (Fairlawn, NJ). HPLC grade dimethyl sulfoxide (DMSO), used for preparation of stock solutions of analytes and internal standard, was procured from Aldrich Chemical Co. (Milwaukee, WI) and monobasic potassium phosphate used for buffer preparation was obtained from Mallinkrodt (St. Louis, MO). Phosphate-buffered saline (0.9% NaCl, pH 7.0) was purchased from Biofluids, Inc. (Rockville, MD).

Animals. Three adult male rhesus monkeys (Macaca mulatta) weighing from 4.5 to 10.3 kg were used for this study. Animals were fed National Institutes of Health Open Formula Extruded NonHuman Primate Diet ad libitum and were housed in groups in accordance with established guidelines (National Institutes of Health, 1996). The monkeys were dosed through either a saphenous catheter or jugular port. Blood was collected from a contralateral catheter or port. CSF was collected from a chronically indwelling s.c. Ommaya reservoir attached to a fourth ventricular Pudenz catheter (McCully et al., 1990). Urine was obtained by a catch collection sometime during the first 10 h after the experiment.

Experimental Design. F-ddA was dissolved in a small volume of DMSO by sonication and then sufficient saline solution was added to make the final dosing solution contain 25 to 33% DMSO. F-ddI was dissolved in normal saline. Both dosing solutions were filter-sterilized before administration to the animals. Three monkeys received F-ddA as a 20 mg/kg i.v. push that lasted 3 to 5 min. In addition, in a subsequent experiment, F-ddI was administered to one of the original animals as a 20 mg/kg 4-min i.v. push. Blood was collected at various times into heparinized Vacutainers (Becton Dickinson & Co., Franklin Lakes, NJ) and plasma was separated by centrifugation. CSF and urine were also collected. Plasma and urine had sufficient 2'-deoxycoformycin added to give a final concentration of 20 µM of the ADA inhibitor. Plasma, CSF, and urine samples were immediately frozen and stored until analysis. Monkeys were not sacrificed for tissue samples, but were returned to the monkey colony after the experiment was completed.

Analytical Methodology and Pharmacokinetic Analysis. Plasma concentrations of F-ddA and F-ddI were determined using a previously published method (Roth and Kelley, 1995). CSF samples were treated the same, except an initial 0.25-ml aliquot was used. Urine samples were sonicated for 10 min after thawing and then diluted by a factor of 200 before processing.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pharmacokinetics after F-ddA Dose. Plasma and CSF kinetics after a 20 mg/kg dose of F-ddA were determined in three male rhesus monkeys with Ommaya reservoirs. The plasma concentration versus time curves from these experiments are shown in Fig. 2. The F-ddA levels appeared to be described by a two-compartment model with a correlation coefficient from a fit to a biexponential decay better than 0.999 in each case. However, the F-ddA elimination half-lives could not be accurately determined because this portion of the curve was defined by only a few points near the limit of quantitation. Rather than include a parameter with so great an uncertainty, a noncompartmental value, the mean residence time (MRT), is reported instead. A summary of pharmacokinetic values from the three monkeys is shown in Table 1. F-ddA was rapidly metabolized to F-ddI with a total body clearance ranging from 106 to 156 ml/min/kg and MRTs ranging from 0.12 to 0.21 h for the three monkeys. F-ddI accounted for the majority of plasma exposure and represented 84 to 89% of the total AUC on a molar basis. F-ddA and F-ddI were both detected in the CSF with F-ddA accounting for a much higher percentage of dideoxynucleoside exposure than in the plasma, ranging from 12 to 36% of total AUC. The mean combined CSF-to-plasma ratio of F-ddA and F-ddI based on F-ddA equivalents was 7%.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Plasma concentration versus time profiles for F-ddA () and F-ddI () in three rhesus monkeys after 20 mg/kg i.v. push of F-ddA. Solid lines are computer fits of F-ddA data to biexponential decay and dashed lines are fits of final F-ddI points to a monoexponential decay.

Pharmacokinetics after F-ddI Dose. Plasma and CSF kinetics after a 20 mg/kg dose of F-ddI were also determined in one of the original rhesus monkeys used for the F-ddA experiment. The plasma time versus concentration curve fit well to a biexponential decay as can be seen in Fig. 3B. The elimination half-lives for F-ddI from plasma and CSF were similar to that obtained after the F-ddA dose. However, the CSF/plasma AUC ratio was much lower at only 1.7% instead of 6.0% for the same monkey after F-ddA administration (Table 2).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Plasma and CSF concentration versus time profiles in monkey 3 after 20 mg/kg i.v. push of F-ddA (A) or F-ddI (B). F-ddA in plasma (), F-ddI in plasma (), F-ddA in CSF (), and F-ddI in CSF (). Solid lines have been computer fit and dashed lines are point to point.

Urinary Disposition. Urine was collected for each of the monkeys after the F-ddA dose for several hours and analyzed for both F-ddA and F-ddI. About half of the total dose was accounted for by urinary excretion and the majority of the dose was excreted as F-ddI. A summary of the data appears in Table 3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

F-ddA pharmacokinetics in monkeys is similar to that of its nonfluorinated predecessors, ddA and didanosine, with the important exception of its much greater stability. A previous study in which rhesus monkeys were dosed with ddA showed such rapid disappearance of the drug that many parameters could not be accurately determined, with a plasma elimination half-life for one animal estimated at 24 s (Hawkins et al., 1995). By contrast, the MRT for F-ddA in this study was measured in minutes, due to its greater stability toward deamination by ADA. The average total body clearance for F-ddA reported here is also considerably slower at 126 ml/min/kg compared with the 500 ml/min/kg estimated for ddA. For both ddA and F-ddA, however, the majority of plasma exposure was from the deaminated metabolites, ddI or F-ddI, respectively. Because F-ddI is resistant to metabolism by purine nucleoside phosphorylase (PNP), more compound is available for transformation to the active form, F-ddATP, intracellularly (Hitchcock et al., 1990; Masood et al., 1990). Several studies of i.v. didanosine in pigtailed macaques (Ravasco et al., 1992; Odinecs et al., 1996) or rhesus monkeys (Hawkins et al., 1995) exhibited plasma profiles described by a biexponential decay with terminal half-lives between 1 and 2 h, which is comparable with the results for F-ddI. This implies that the PNP degradation is not a major factor in the plasma clearance of didanosine in nonhuman primates. It is further confirmed by the comparable clearance rates of 10.8 ml/min/kg for F-ddI versus 10 ml/min/kg reported previously for didanosine in Rhesus monkeys (Hawkins et al., 1995).

Pharmacokinetic modeling of new drugs in nonhuman species has gained importance as interspecies scaling has been shown useful in designing initial dosage regimens in human clinical trials (Obach et al., 1997). Interspecies scale-up has been used successfully for two other nucleoside reverse transcriptase inhibitors, AZT and ddC (Ibrahim and Boudinot, 1989; Patel et al., 1990). For F-ddA, cross-species pharmacokinetics in dogs and monkeys are quite comparable. For example, F-ddA pharmacokinetics in beagle dogs was determined for two animals each at 100, 250, and 500 mg/kg (Campbell et al., 1996). These experiments indicated that F-ddI was the major source of exposure in plasma and accounted for 40 to 70% of the dose as urinary excretion. Terminal half-lives for F-ddA ranged from 1.5 to 2.0 h and for F-ddI from 1.8 to 2.9 h. These results are similar to the data presented here for rhesus monkeys. In addition, results from rat studies have demonstrated that animals treated with 120-min i.v. infusions of either ddI, F-ddI, or F-ddA demonstrated much more rapid clearance of ddI relative to F-ddA or F-ddI, as would be expected. However, there was a much longer terminal half-life for F-ddA in rats than in dogs or monkeys (Singhal et al., 1997). It is unclear why the rat data differ from the other two animals, but earlier experiments had indicated that monkeys are a good model for didanosine in humans (Ravasco et al., 1992). It is interesting to note that the deamination kinetics of F-ddA were determined in fresh rat and monkey plasma with a resulting half-life for each of 4 to 5 h (Roth and Kelley, 1995). This indicates that the difference in in vivo results is probably not due to a simple difference in the plasma ADA levels.

Because the brain is a sanctuary for HIV and AIDS dementia is a devastating manifestation of HIV infection (Gallo et al., 1987), CNS penetration is an important consideration for anti-AIDS drugs. Although lipophilicity and plasma protein binding are generally predictive of CNS entry, a study in pyrimidine dideoxynucleosides indicated that these factors were not sufficient to explain the differences among several members of this group. It was suggested that an unknown carrier-mediated process was a major factor (Collins et al., 1988). Because entry into the CNS has been poor for other dideoxy-nucleosides, with studies on didanosine in rhesus monkeys yielding only a 4.8% CSF-plasma ratio (Hawkins et al., 1995), it was predicted the more lipophilic F-ddA would lead to improved penetration of the blood-brain barrier. In fact, previous investigations in rats had demonstrated improved CNS delivery of total dideoxynucleosides when animals were dosed with F-ddA compared with didanosine. AUC brain-plasma ratios increased from 3.8% for ddI to 12.1% for F-ddA and AUC CSF-plasma ratios increased from 1.4 to 9.1%, respectively (Singhal et al., 1997). However, the AUC CSF-plasma ratio for F-ddA in monkeys was similar to that of didanosine at 6.8 versus 4.8%. Perhaps this is due to an unknown carrier-mediated process similar to the case of the pyrimidine nucleosides. Also, in vitro studies using bovine brain tissue suggest that the levels of ADA and PNP in cerebral microvessels may play a role in determining the membrane permeability of a drug (Johnson and Anderson, 1996). Thus, varying levels of these enzymes in the cerebral microvessels of each species may cause some of the differences observed in CNS penetration.

The current recommendation for treating AIDS patients is combination therapy of three different anti-HIV drugs, generally two nucleoside analogs and a third agent from a different class of compounds-either a protease inhibitor or non-nucleoside reverse transcriptase inhibitor. Therapeutic failure of the first treatment regimen should be followed by immediate substitution of at least two and preferably all three of the original components (Moyle et al., 1998). Given that there are only eleven drugs currently approved by the FDA for the treatment of HIV infection, of which only didanosine is a purine dideoxynucleoside reverse transcriptase inhibitor (Food and Drug Administration, 1998), the need for new alternative drugs is apparent. Perhaps even more significant, lodenosine has shown activity against clinical isolates and cell lines that are resistant to AZT, ddI, and ddC (Driscoll et al., 1997), so this apparent lack of cross-resistance makes it a potentially valuable component as a substitute in multiple drug therapy. In addition, due to its acid stability, F-ddA may be administered orally without the use of antacids or buffers and still retain good bioavailability. Therefore, given that its penetration into the CNS in this monkey model is at least comparable with that of didanosine, F-ddA remains as a potential candidate for addition to the current arsenal of therapeutic agents in the treatment of AIDS.