Preparation and pharmacokinetics of 11C labeled stavudine (d4T)

Nuclear Medicine and Biology 31 (2004) 613– 621

Preparation and pharmacokinetics of

11

www.elsevier.com/locate/nucmedbio

C labeled stavudine (d4T)

Eli Livnia, Mark Berkera, Shawn Hilliera, Stephen C. Wallerc, Marc D. Oganc, Robert P. Discordiac, J. Kent Rienhartc, Robert H. Rubina,b, Alan J. Fischmana,b,* a

Division of Nuclear Medicine of the Department of Radiology, Massachusetts General Hospital, 55 Fruit St., Boston, MA 02114, USA, and the Department of Radiology, Harvard Medical School, Boston, MA,, USA b Center for Experimental Pharmacology and Therapeutics, Harvard–Massachusetts Institute of Technology, Division of Health Sciences and Technology, Boston, MA, USA c Bristol-Meyers Squibb Inc., Princeton, NJ, USA Received 15 August 2003; received in revised form 3 October 2003; accepted 27 November 2003

Abstract Stavudine, a potent antiviral agent for treating human immunodeficiency virus (HIV) infections, was radiolabeled with 11C by methylation of a specifically designed precursor, 5⬘-O-(2-tetrahydropyranyl)-5-bromo-2⬘,3⬘-didehydro-3⬘-deoxythymidine, with 11C H3I. The radiolabeled drug was isolated by reverse phase HPLC. A total time of approximately 45 minutes was required for synthesis, purification and isolation of 11C stavudine with chemical and radiochemical purities of greater than 98%. 11C stavudine was combined with unlabeled drug (2.0 mg/kg) and used to study its pharmacokinetics in rats by measurement of radioactivity in excised tissues. In this species, there was rapid accumulation of drug in all tissue. In all tissues, with the exceptions of testis and brain, highest concentrations of drug were detected at 5 minutes after injection and decreased monotonically thereafter. The peak concentration (␮g/g) of stavudine in blood was 1.78 ⫾ 0.16 and similar levels were achieved in most other tissues; heart 1.66 ⫾ 0.11, lung 1.60 ⫾ 0.15, liver 2.13 ⫾ 0.17, spleen 1.61 ⫾ 0.15, adrenal 1.47 ⫾ 0.20, stomach 1.40 ⫾ 0.11, GI tract 1.44 ⫾ 0.14, skeletal muscle 1.38 ⫾ 0.15 and bone 1.30 ⫾ 0.16. Much higher peak concentrations were achieved in kidney; 7.23 ⫾ 0.57 ␮g/g. Concentrations in testis were lower and remained relatively constant over 1 hour; peak 0.62 ⫾ 0.14 ␮g/g at 15 min Brain concentrations were low but increased monotonically over time; peak 0.26 ⫾ 0.02 ␮g/g at 60 min. Future PET studies with this radiopharmaceutical will allow in vivo measurements of the pharmacokinetics of stavudine in both animal models and human subjects. © 2004 Elsevier Inc. All rights reserved. Keywords: Stavudine; d4T; Pharmacokinetics; Radiolabeling; HIV; AIDS

1. Introduction Stavudine, 2⬘,3⬘ didehydro-3⬘-deoxythymidine (d4T), is a synthetic thymidine nucleoside analog that is effective in the treatment of human immunodeficiency virus (HIV) infection [1– 4]. Stavudine enters cells rapidly by nonfacilitated diffusion, with the rate of influx being linear with respect to concentration [5]. The drug is phosphorylated by cellular kinases to stavudine triphosphate, which inhibits HIV replication by at least two mechanisms: inhibition of the HIV reverse transcriptase; and inhibition of viral DNA synthesis through DNA chain termination [6,7]. The effects of stavudine on HIV reverse transcriptase are far more potent than its effects on cellular DNA polymerases and

* Corresponding author. Tel.: (617) 726-8353; fax: (617) 726-6265. E-mail address: [email protected] (A.J. Fischman). 0969-8051/04/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.nucmedbio.2003.11.006

mitochondrial DNA synthesis, thus permitting its use therapeutically as part of highly active antiretroviral therapy (HAART) regimens [8,9]. Although the clinical pharmacology of stavudine has been extensively studied, the delivery of the drug to key tissue sites is still incompletely understood. In particular, given the importance of central nervous system infection in the AIDS patient, as well as the importance of peripheral neuropathy as a side effect of stavudine, detailed knowledge of the tissue pharmacokinetics of this drug in the central nervous system would be useful. Available pharmacokinetic information concerning stavudine has demonstrated the following: When administered orally, bioavailability approaches 100%, with maximal concentrations being reached at 1 hour, and serum levels increasing linearly with dose. Serum half-life is approximately 1 hour, and there is no accumulation of the drug in the plasma with dosing 2– 4

614

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

times daily. Stavudine distributes into total body water and appears to enter cells by nonfacilitated diffusion, with clearance occurring by both renal and nonrenal mechanisms [10 –12]. There is significant penetration into the cerebrospinal fluid (CSF) with levels that are approximately 40% of serum values [13–16]. The chemical structure of stavudine, with the presence of a methyl group, offers the possibility for labeling the molecule with carbon-11, creating a labeled form of the native drug of interest rather than an analogue. If a rapid method for synthesizing 11C-stavudine could be devised, positron emission tomography (PET) would be useful for determining the tissue delivery of stavudine in humans. This report describes a method for preparing 11C-stavudine and initial pharmacokinetic studies in rats, as a prelude to human studies.

2. Methods and materials 2.1. Materials Acetyl bromide (99%), (-)-5-bromouridine (99⫹%), nbutyl lithium (1.6 mol/L in hexanes), 1,2-dibromoethane, (99⫹%), 3,4-dihydro-2H-pyran (97%), Dowex (50Wx8200), hydrogen bromide in acetic acid (30 wt%), methanesulfonyl chloride (99.5⫹%), pyridine (99.8%, anhydrous), sodium methoxide (25 wt% in methanol), tetrahydrofuran (anhyd, spectrophotometric grade, 99.5%), toluenesulfonic acid monohydrate (98.5%), and zinc dust (⬍10 ␮m, 98%⫹) were obtained from Aldrich Chemical Co. Methyl iodide and sodium benzoate were obtained from Mallinckrodt Chemical Co. Triethylamine (sequenal grade) was obtained from Pierce Chemical Co. Methanol, HPLC grade, was obtained from Burdick & Jackson, Inc. Hydrochloric acid (37%), methanol (Omnisolv) and tetrahydrofuran (anhydrous) were obtained from EM Science Inc. Sodium benzoate was milled in an electric grinder to obtain a fine dust before use. Tetrahydrofuran was distilled from Na, benzophenone. HPLC grade H2O was obtained by deionization and reverse osmosis (Milli-Q). All other reagents were used without further processing. 2.2. Synthesis of the precursor for preparing stavudine (11C d4T)

11

C labeled

2.2.1. General methods Reaction mixtures were stirred magnetically under nitrogen. TLC samples were chromatographed on glass-backed silica gel 60 F-254 plates (0.25 mm thick, 5 cm in length; EM Science) and were viewed using a 254 nm UV lamp and iodine. Flash column chromatography was performed on silica gel 60 (230-400 mesh; EM Science). Proton NMR spectra of intermediates were obtained using a Bruker AC-300 spectrometer in either CDCl3 or CD3OD (CD3OH line referenced at 4.78 ppm). Isotopic compositions were determined by positive ion electrospray on a Sciex API III mass spectrometer.

2.2.2. HPLC of intermediates HPLC was performed with a Rainin Dynamax system using two model SD-200 pumps and a model UV-1 UVVisible detector. HPLC retention times and purities were determined using a YMC ODS-A column (6.0 ⫻ 150 mm, 3 ␮m) at flow rate of 1.8 mL/min and detection at 254 nm. The Solvents and gradient conditions were as follows: Solvents A: (5% Acetonitrile/Water) 0.01 mol/L Ammonium Acetate; B: (50% Acetonitrile/Water) 0.01 mol/L Ammonium Acetate; Gradient 0-60% B at 0-10 minutes; 60-100% B at 15-20 minutes; hold at 100% B for 25 minutes. 2.2.3. Preparative HPLC Preparative HPLC was performed using a YMC ODS-A column (20 ⫻ 150 mm, 5 ␮m) at flow rate of 8 mL/min and UV 254 nm. HPLC solutions were prepared using Milli-Q water and were filtered through a 0.45 ␮m filter. Solvent and gradient conditions were as follows: solvents A (10% acetonitrile/water) and B (90% acetonitrile/water); gradient 20% B at 0-20 minutes; 20-60%% B at 20-21 minutes; hold at 60% B until 21-35 minutes 2.2.4. HPLC of d4T A Hitachi system L-6200 pump, L– 4000 Photo Detector and D–2000 Chromato–Integrator were used. Analytical HPLC used a YMC-Pack AQ column (4.6 ⫻ 150 mm, 3 ␮m, YMC Inc.). The flow rate was 1.0 mL/min. UV detection wavelength was 279 nm. Mobile phase consisted of MeOH/H2O (8:92). Samples of 20 ␮g (20 ␮L of a 1.0 mg/mL solution in mobile phase) were injected using a Rheodyne model 7125 injector with a 20 ␮L injection loop. 5-Bromo-1-(2⬘,3⬘,5⬘-tri-O-methanesulfonyl-␤-D-ribofuranosyl)uracil, (2). A solution of (-)-5-bromouridine (1) (10.0 g, 30.9 mmol), pyridine (120 mL), and methanesulfonyl chloride (10.8 mL, 15.9 g, 139 mmol) was stirred for 30 min at 0°C. After 3 h at room temperature, water (350 mL) was added. The mixture was extracted with EtOAc (3 ⫻ 100 mL). The combined organic extracts were washed consecutively with water (100 mL), 1.5 mol/L H2SO4 (2 ⫻ 125 mL), and brine (100 mL), dried (Na2SO4), filtered, and evaporated to give 2 (18.11 g, 32.5 mmol, 105% yield) as a light tan solid. TLC: Rf0.52 (EtOAc). 1-(2,2⬘-Anhydro-5⬘-O-benzoyl-3⬘-O-methanesulfonyl-␤D-ribofuranosyl)-5-bromo-uracil, (3). Sodium benzoate (4.54 g, 31.5 mmol) was added to a solution of 1-(2⬘,3⬘,5⬘tri-O-methanesulfonyl-␤-D-ribofuranosyl)-5-bromouracil (2) (7.28 g, 13.1 mmol) and anhyd DMF (100 mL) at 90°C. This mixture was stirred at 90°C for 40 hours. After cooling to room temperature, water (300 mL) was added. The resulting precipitate was vacuum filtered, washed with waster (2 ⫻ 50 mL), and dried at 0.1 mm Hg to give 3 (5.13 g, 10.9 mmol, 83% yield) as a tan solid. HPLC: RT 13.6 min, AP 72.9%. 1-(5⬘-O-Benzoyl–2⬘-bromo-3⬘-O-methanesulfonyl-␤-Dribofuranosyl)-5-bromo-uracil, (4). Acetyl bromide (1.97 mL, 26.6 mmol) and HBr (30 wt% in HOAc, 4.40 mL, 22.1 mmol) were individually added to a solution of 3 (8.14 g,

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

17.3 mmol) and anhyd DMF (200 mL) at 80 °C. After 5 hours at 80°C, the reaction was found to be complete via HPLC. The reaction was allowed to cool to room temperature and was used immediately in the next reaction without work-up. 4: HPLC: RT 18.4 minutes. 1-(5⬘-O-Benzoyl-2⬘,3⬘-dideoxy-␤-D-ribofuranosyl)-5-bromouracil, (5). Zinc dust (2.11 g, 32.3 g.atm) was added to the previous reaction mixture in two portions over 1 hour. After an additional 1 h at room temperature, 1,2dibromoethane (1.50 mL, 17.4 mmol) was added, and this mixture was allowed to stir 16 h at room temperature. Water (100 mL) was added, and the mixture was vacuum filtered. Methanol (100 mL) was added to the oily brown solid, and this solution was concentrated to 20 mL, cooled to –20 °C for 1 h, and vacuum filtered while cold. Water (20 mL) was added to the filtrate, giving a brown solid (1.63 g), which was found to be mostly an impurity by HPLC (RT 22.1 min, AP 67.6%). The golden-yellow filtrate was diluted with water (600 mL). The resulting mixture was filtered, and the solid was dried at 0.01 mm Hg to give 5 (2.45 g, 6.22 mmol, 36% yield from 3) as a tan solid. HPLC: RT 14.8 min, AP 90.4%. 5-Bromo-1-(2⬘,3⬘-dideoxy-␤-D-ribofuranosyl)uracil, (6). Sodium methoxide (25 wt% in methanol, 5.60 mL, 24.5 mmol) was added to a solution of 5 (2.40 g, 6.09 mmol) in methanol (300 mL). After 1.5 h at room temperature, THF (100 mL) was added, and the pH was adjusted to 5 with the addition of Dowex 50Wx8-200 acidic ion-exchange resin (52.4 g) in small portions. This mixture was vacuum filtered, and the filtrate was concentrated and dried at 0.1 mm Hg for 30 min The resulting tan solid was triturated with hexanes (2 ⫻ 25 mL) and dried at 0.1 mm Hg to give 6 as a tan powder, which was used immediately in the next reaction to avoid decomposition. HPLC: RT 5.1 min, AP 94.7%. 5-Bromo-1-[2⬘,3⬘-dideoxy-5⬘-O-(tetrahydro-2H-pyran2-yl)-␤-D-ribofuranosyl]-uracil, (7). A solution of 6, THF (400 mL), dihydropyran (2.20 mL, 24.1 mmol), and toluenesulfonic acid monohydrate (61 mg, 0.32 mmol) was allowed to stir for 18 h at room temperature. Triethylamine (178 ␮L, 1.28 mmol) was added, and the reaction mixture was concentrated to give a brown oil (4.03 g). Flash chromatography on silica gel (100 g) eluted with 20% acetone/ hexanes (750 mL) and then 50% acetone/hexanes gave the product as an off-white solid (1.68 g), AP 94%. After two recrystallizations from acetone/hexanes, the product was obtained as a white solid, AP 96.1%. Semi-preparative HPLC using method 2 afforded 7 (1.21 g, 3.24 mmol, 53% yield from 5) as a white solid. HPLC: RT 15.8-16.7, AP 99.8%. MS (ESI): 373 (M⫹H ⫹ ). Anal Cal⬘d for C14H17BrN2O5: C, 45.06; H, 4.59; N, 7.51; Br, 21.41. Found: C, 44.88; H, 4.59; N, 7.46; Br, 21.63. H2O (Karl Fischer): 0.28%. Proton NMR was consistent with a 1:1 mixture of two diastereomers of the expected compound. 1H NMR (CDCl3) ␦ 8.43 (br signal, 1 H), [8.05 (s), 8.03 (s), 1 H], [7.00 (m), 6.97 (m), 1 H], [6.38 (m), 6.33 (m), 1 H], [5.87 (m), 5.85 (m), 1 H], [5.04 (m), 5.01 (m), 1 H], [4.63 (m), 4.58 (m), 1 H], [4.09 (dd, J ⫽ 2.8, 11.9 Hz), 4.03 (dd,

615

J ⫽ 2.2, 11.5 Hz), 1 H], [3.87 (m), 3.81 (m), 1 H], [3.71 (dd, J ⫽ 2.2, 11.5 Hz), 3.65 (dd, J ⫽ 2.8, 11.9 Hz), 1 H], 3.57-3.50 (m, 1 H), 1.87-1.52 (m, 6 H). 2⬘,3⬘-Didehydro-3⬘-deoxythymidine (stavudine, d4T). A 1 mL Reacti-Vial with magnetic stir bar was oven dried, fitted with a Tuf-Bond septa and cooled under argon via a needle inlet. The vial was charged with a solution (50 mg/mL) of 5⬘-O-(2-tetrahydropyranyl)-5-bromo-2⬘,3⬘-didehydro-3⬘-deoxythymidine (50 ␮L, 2.5 mg, 6.7 ␮mol) in anhydrous THF and cooled to –78°C in a dry ice/isopropanol bath. A solution of 1.2 mol/L n-BuLi in hexanes (14 ␮L, 16.8 ␮mol) was added all at once via syringe. The reaction was stirred at –78°C for 90 sec, then a solution (97 mg/mL) of CH3I (10 ␮L, 0.1 mg, 6.8 ␮mol) in THF was added. The cooling bath was removed and the solution was stirred an additional 2 minutes. The reaction was acidified by addition of 0.4 mol/L HCl in MeOH (70 ␮L, 28 ␮mol). TLC (EtOAc) of the reaction versus standards showed the presence of THP-d4T and THP-d4U with trace amount of THPBr-d4U. The reaction was placed in an oil bath maintained at 60°C and stirred under argon for 5 min The reaction was made basic by addition of Et3N (10 ␮L). TLC (CHCl3/ MeOH 8:2) showed complete deprotection. The solvent was evaporated under a stream of nitrogen. The residue was dissolved in MeOH (50 ␮L), evaporated under a stream of nitrogen and dissolved in 150 ␮L of MeOH/H2O (2:8). The solution of crude product was purified on a YMC-Pack AQ semi-preparative column (10 ⫻ 250 mm; flow rate 4.0 mL/min; UV 278 nm) eluted with premixed, degassed MeOH/H2O (2:8). The desired d4T product eluted at 10.4 min and was isolated in a single fraction (total volume 2.2 mL). Analytical HPLC showed a single compound eluting at 20.3 minutes which co-eluted with authentic d4T. Integrated peak area compared to a d4T standards curve indicated a total of 0.38 mg (25% yield) of d4T was contained in this isolated fraction. The overall scheme for synthesis of the precursor that was used for preparing 11C labeled stavudine and its conversion to unlabeled stavudine is summarized in Fig. 1. 2.2. Preparation of

11

C labeled stavudine

A 5 ml Reacti-Vial containing 2.5 mg the precursor for C stavudine, 5⬘-O-(2-tetrahydropyranyl)-5-bromo-2⬘,3⬘didehydro-3⬘-deoxythymidine, (1, Fig. 2) in 0.1 mL THF was cooled at ⫺78°C in a dry ice isopropanol bath. A solution of 1.5 mol/L n-BuLi in hexanes (20 ␮L) was added and the vial was shaken lightly for 2 minutes. 11CO2 was produced by proton irradiation of 1% oxygen in nitrogen and trapped in 0.15 mL LAH (0.25 mol/L) in THF. The THF was brought to dryness under a stream of nitrogen at 120°C, HI (0.8 mL) was added and the resulting 11CH3I which was transferred to the Reacti-Vial over 5 minutes. The reaction mixture was then vortexed for 2 minutes, acidified by addition of 0.4 mol/L HCl in methanol (70 ␮L) 11

616

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

Fig. 1. Scheme for synthesis of the precursor that was used for preparing

and heated at 60°C for 3 minutes. A solution of triethylamine in methanol (14% v/v, 100 ␮L) was added and the solvent evaporated under a stream of nitrogen. A solution of methanol : water,17:83 (0.6 mL) was added and the mixture was purified on a C-18 Luna semi preparative HPLC column (10 ⫻ 250 mm). The column was eluted with methanol: water, 17:83 at a flow rate of 7 mL/min. 11C stavudine (3, Fig. 2) was eluted at 5.5 min. The solvent was evaporated and the product redissolved in saline. The resulting solution was filtered using a 0.22 ␮m Millex GV filter. Radiochemical, chemical purity and specific activity were determined by HPLC on a C-18 column 250 ⫻ 3.9 mm (Prodigy, Phenomenex) eluted with acetonitrile:water, 8:92 at a flow rate of 2 mL/min. 11C stavudine was identified by coinjection with a standard sample of authentic drug.

11

C labeled stavudine.

2.3. Pharmacokinetic studies Twelve male Sprague-Dawley rats weighing approximately 150 g (Charles River Breeding Laboratories, Boston, MA) were anesthetized with ether and injected, via tail vein, with approximately 0.5 mCi of 11C labeled stavudine plus unlabeled drug (2.0 mg/kg). At 5, 15, 30, and 60 minutes after injection, groups of three animals were sacrificed with an overdose of sodium pentobarbital and biodistribution was determined. Samples of blood, heart, lung, liver, spleen, kidney, adrenal, stomach, GI tract, testes, skeletal muscle, bone and brain were weighed and radioactivity was measured with a well type gamma counter (LKB model 1282, Wallac Oy, Finland). To correct for radioactive decay and permit calculation of the concentration of radioactivity in

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

Fig. 2. Scheme for preparing

11

C labeled stavudine.

each organ as a fraction of the administered dose, aliquots of the injected doses were counted simultaneously. The results were expressed as ␮g/g and ␮g/organ. The procedure was repeated with a second group of twelve animals to achieve a total of six animals per time point. 2.4. Statistical methods The results of the pharmacokinetic studies were evaluated by two-way analysis of variance (ANOVA) with a linear model in which tissue and time were the classification variables. Post-hoc comparisons of drug concentrations were performed by Duncan’s new multiple range test [17]. All results were expressed as mean ⫾ SEM. 3. Results 3.1. Preparation of

11

C stavudine

The seven-step procedure employed to synthesize the precursor (5⬘-O-(2-tetrahydropyranyl)-5-bromo-2⬘,3⬘-didehydro-3⬘-deoxythymidine) for preparing 11C stavudine afforded a pure product in excellent yield. The radiolabeling procedure provided 11C stavudine in adequate yield and excellent purity. Radiochemical and purity and specific activity of 5 synthetic runs analyzed by HPLC using a C-18 column (Prodigy, Phenomenex, 250 ⫻ 3.9 mm) eluted with acetonitrile:water, 8:92 at a flow rate of 2 mL/min. demonstrated that radiochemical purity was 98.08 ⫾ 0.26%, chemical purity was ⬎96% and specific activity was 185.0 ⫾ 58.5 mCi/␮mole [EOS].The average yield was 10.3 ⫾ 3.1 mCi of [EOS]. A representative HPLC of 11C Stavudine is shown in Fig. 3. 3.2. Pharmacokinetics of

11

C stavudine

When expressed a percent injected dose per gram, Fig. 4 and Fig. 6, two-way analysis of variance demonstrated

Fig. 3. Representative HPLC chromatogram of Absorbance was measured at 280 nm.

617

11

C labeled stavudine.

highly significant main effects of tissue (F12,260 ⫽ 142.42; P ⬍ 0.0001), time after injection (F12,260 ⫽ 105.37; P ⬍ 0.0001) and tissue by time interaction (F12,260 ⫽ 9.11; P ⬍ 0.0001). Independent of time after injection, the tissues can be classified into five groups in order of decreasing concentrations of stavudine. Highest concentrations of drug were detected in kidney (P ⬍ 0.001 compared with all other tissues). The second highest concentrations were detected in liver (P ⬍ 0.001 vs testis and brain, P ⬍ 0.05 vs bone). The third highest concentrations were measured in blood, heart, lung, spleen, adrenal, stomach, GI tract, and muscle. The fourth highest concentrations were detected in testis (P ⬍ 0.001 vs blood, stomach and GI tract, P ⬍ 0.05 vs heart and lung). Lowest concentrations were measured in brain (P ⬍ 0.001 vs all other tissues). With the exceptions of testis and brain, tissue concentrations of stavudine decreased monotonically with time. When expressed a percent injected dose / organ, Fig. 5 and Fig. 6, two-way ANOVA demonstrated highly significant main effects of tissue (F12,260 ⫽ 179.95; P ⬍ 0.0001), time after injection (F12,260 ⫽ 36.85; P ⬍ 0.0001) and tissue by time interaction (F12,260 ⫽ 11.09; P ⬍ 0.0001). Independent of time after injection, the tissues can be classified into 7 groups in order of decreasing concentrations of stavudine. Highest concentrations of drug were detected in skeletal muscle (P ⬍ 0.001 compared with all other tissues). The second highest levels were detected in blood and bone (P ⬍ 0.001 vs heart, lung, spleen, adrenal, stomach, GI tract, testis and brain, P ⬍ 0.005 vs kidney). The third highest levels were detected in liver and kidney (P ⬍ 0.001 vs heart, lung, spleen, adrenal, stomach, GI tract, testis, brain, P ⬍ 0.05 vs GI tract). The fourth highest levels were measured in the GI tract (P ⬍ 0.001 vs heart, lung, spleen, adrenal, stomach, testis and brain). The fifth highest levels were measured in lung, stomach and testis (P ⬍ 0.001 vs spleen, adrenal, and brain, P ⬍ 0.05 vs heart). The sixth highest levels were measured in heart (P ⬍ 0.001 vs adrenal and brain). Lowest levels were measured in brain (P ⬍ 0.001 vs all other tissues). With the exceptions of testis and brain, tissue concentrations of stavudine decreased monotonically with time.

618

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

Fig. 4. Pharmacokinetics of 11C labeled stavudine in of blood, heart, lung, liver, spleen, kidney, adrenal, stomach, GI tract, testes, skeletal muscle, and bone expressed as ␮g/g. Each value is the mean ⫾ SEM for six animals. Blood a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.05 vs 15 min, d: P ⬍ 0.001 vs 15 min, e: P ⬍ 0.001 vs 30 min; Heart a: P ⬍ 0.001 vs 5 min, b: P ⬍ 0.05 vs 15 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.001 vs 30 min; Lung a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.05 vs 30 min; Liver a: P ⬍ 0.001 vs 5 min, b: P ⬍ 0.001 vs 15 min, c: P ⬍ 0.001 vs 30 min; Spleen a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.005 vs 5 min, c: P ⬍ 0.001 vs 5 min, d: P ⬍ 0.05 vs 15 min e: P ⬍ 0.001 vs 15 min, f: P ⬍ 0.005 vs 30 min; Adrenal a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.05 vs 30 min; Stomach a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.001 vs 30 min; GI tract a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.01 vs 15 min, d: P ⬍ 0.05 vs 30 min; Skeletal muscle a: P ⬍ 0.001 vs 5 min, b: P ⬍ 0.01 vs 15 min, c: P ⬍ 0.005 vs 30 min; Bone a: P ⬍ 0.01 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.05 vs 15 min, d: P ⬍ 0.001 vs 15 min, e: P ⬍ 0.001 vs 30 min; Kidney a: P ⬍ 0.001 vs 5 min, b: P ⬍ 0.001 vs 15 min, c: P ⬍ 0.001 vs 30 min.

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

619

Fig. 5. Pharmacokinetics of 11C labeled stavudine in of blood, heart, lung, liver, spleen, kidney, adrenal, stomach, GI tract, testes, skeletal muscle, and bone expressed as ␮g/organ. Each value is the mean ⫾ SEM. for six animals. Blood a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.005 vs 5 min, c: P ⬍ 0.001 vs 5 min, d: P ⬍ 0.001 vs 15 min, e: P ⬍ 0.001 vs 30 min; Heart a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.001 vs 30 min; Lung a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.05 vs 30 min; Liver a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.001 vs 30 min; Spleen a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.005 vs 5 min, c: P ⬍ 0.001 vs 5 min, d: P ⬍ 0.05 vs 15 min e: P ⬍ 0.001 vs 15 min, f: P ⬍ 0.005 vs 30 min; Adrenal a: P ⬍ 0.01 vs 5 min, b: P ⬍ 0.05 vs 30 min; Stomach a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.001 vs 15 min, d: P ⬍ 0.001 vs 30 min; GI tract a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.005 vs 15 min, d: P ⬍ 0.05 vs 30 min; Skeletal muscle a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.05 vs 15 min, d: P ⬍ 0.001 vs 30 min; Bone a: P ⬍ 0.01 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.005 vs 15 min, d: P ⬍ 0.01 vs 30 min; Kidney a: P ⬍ 0.005 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.005 vs 15 min, d: P ⬍ 0.001 vs 30 min.

620

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621

Fig. 6. Pharmacokinetics of 11C labeled stavudine in rat brain expressed as ␮g/g and ␮g/organ. Each value is the mean ⫾ SEM for six animals. ␮g/g a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.005 vs 5 min, c: P ⬍ 0.05 vs 15 min; ␮g/g a: P ⬍ 0.05 vs 5 min, b: P ⬍ 0.001 vs 5 min, c: P ⬍ 0.05 vs 15 min, d: P ⬍ 0.05 vs 30 min.

4. Discussion The standard method for the commercial preparation of stavudine introduces the ring methyl group at early stages of the synthesis and is thus not applicable to the time scale required for the production of 11C-labeled drug. In previous studies with other drugs, we prepared specifically designed precursors that could be labeled with 11C or 18F at the final step of the synthesis [18 –22]. Thus, a similar approach was used to prepare 11C stavudine. Using the reaction scheme illustrated in Fig. 2 we demonstrated that a product of high chemical and radiochemical purity can be easily prepared within 30 minutes. Although this procedure results in a low specific activity product, this is not a problem for pharmacokinetic studies, since in these investigations, a large amount of unlabeled drug is usually coinjected with the tracer. With the procedure developed here, adequate amounts of radiolabeled drug for both animal and human studies can be easily prepared when ⬎ 1,000 mCi of 11CH3I are available for synthesis. The pattern of distribution of 11C stavudine in rats is consistent with the results of whole body qualitative autoradiographic studies in mice at one time point after injection of 14C-labeled stavudine [23]. Based on the pharamcokinetic data in Fig. 1 and total mass of drug injected (tracer ⫹ unlabeled stavudine), the peak concentration of stavudine in brain was 1.16 ␮mol/L, which is within the range of EC50 reported for HIV-1, 0.009 – 4.1 ␮mol/L [24 –32]. For all other tissues higher peak concentrations were achieved. The use of 11C stavudine to study pharmacokinetics has the advantage that measurements can be performed with PET. The use of PET to study pharmacokinetics has many clear

advantages over more conventional methodologies. Although concentrations of 14C and 3H-labeled drugs can be determined by radioactivity measurements of excised tissues or quantitative autoradiography using many animals, only PET allows multiple measurements in the same animal at different times and under varying physiological or pathological conditions. A major benefit of PET techniques is the quantitative nature of the measurement, permitting direct tissue assay of drug concentrations. Detailed HPLC analysis of 14C labeled stavudine has demonstrated that ⬃80% of intravenously injected drug circulates as intact stavudine at 1 hour after administration [33]. This suggests that measurements of tissue radioactivity largely reflect concentrations of intact drug. Based on the biodistribution data, preliminary MIRDOSE calculations indicate that approximately 20 mCi of 11C stavudine can be administered without delivering a radiation burden in excess of 20 mGy to any organ (unpublished results). In conclusion, this study describes the preparation of 11C stavudine. This radiopharmaceutical is chemically identical to the unlabeled drug and has a biodistribution similar to that of 14 C-labeled material. Future PET studies with this agent will allow in vivo measurements of the pharmacokinetics of this important drug in both animal models and human subjects. References [1] Spruance SL, Pavia AT, Mellors JW, Murphy R, Gathe J, Stool E, Jemsek JG, Dellamonica P, Cross A, Dunkle L, Gathe J. Clinical efficacy of monotherapy with stavudine compared with zidovudine in HIV-infected, zidovudine-experienced patients. A randomized, double-blind, controlled trial. Bristol-Myers Squibb Stavudine/019 Study Group. Ann Intern Med 1997;126:355– 63.

E. Livni et al. / Nuclear Medicine and Biology 31 (2004) 613– 621 [2] Foudraine NA, de Jong JJ, Jan Weverling G, van Benthem BH, Maas J, Keet IP, Jurriaans S, Roos MT, Vandermeulen K, de Wolf F, Lange JM. An open randomized controlled trial of zidovudine plus lamivudine versus stavudine plus lamivudine. AIDS 1998;12:1513–9. [3] Kline MW, Van Dyke RB, Lindsey JC, Gwynne M, Culnane M, McKinney RE, Nichols S, Mitchell WG, Yogev R, Hutcheon N, McKinney RE. A randomized comparative trial of stavudine (d4T) versus zidovudine (ZDV, AZT) in children with human immunodeficiency virus infection. AIDS Clinical Trials Group 240 Team. Pediatrics 1998;101:214 –20. [4] Clumeck N. Stavudine plus a non-thymidine nucleoside reverse transcriptase inhibitor as a backbone for highly active antiretroviral therappy. Antivir Ther 1998;(suppl 4):39 – 43. [5] August EM, Birks EM, Prusoff WH. 3⬘-Deoxythymidin-2⬘-ene permeation of lymphocyte H9 cells by non-facilitated diffusion. Mol Pharmacol 1991;39:2246 –9. [6] Ho H-T, Hitchcock MJM. Cellular pharmacology of 2⬘,3⬘-dideoxy-2⬘3⬘didehyrothymidine, a nuclceoside analog active against human immunodeficiency virus. Antimicrob Agents Chemother 1989;33:844–9. [7] Dyatkina N, Minassian S, Kukhanova M, Krayevsky A, von JantaLipinsky M, Chidgeavadze Z, Beabealashvilli R. Properties of 2⬘,3⬘dideoxy-2⬘,3⬘-dehyrothymidine 5⬘-triphosphate in terminating DNA synthesis catalyzed by several different DNA polymerases. FEBS Lett 1987;219:151–5. [8] Ono K, Nakane H, Herdewijn P, Balzarini J, De Clercq E. Differential inhibitory effects of several pyrimidine 2⬘,3⬘-dideoxynucleoside 5⬘triphosphates on the activities of reverse transcriptase and various cellular DNA polymerases. Mol Pharmacol 1989;35:578 – 83. [9] Martin JL, Brown CE, Matthews-Davis N, Reardon JE. Effects of antiviral nucleoside analogs on human DNA polymerases and mitochondrial DNA synthesis. Antimicrob Agents Chemother 1994;38: 2743–9. [10] Skowron G. Biologic effects and safety of stavudine; overview of phase I and II clinical trials. J Infect Dis 1995;171:S113–7. [11] Browne MJ, Mayer KH, Chafee SB, Dudley MN, Posner MR, Steinberg SM, Graham KK, Geletko SM, Zinner SH, Denman SL. 2⬘-3⬘ Didehydro- 3⬘-deoxythymidine (d4T) in patients with AIDS or AIDSrelated complex: a phase I trial. J Infect Dis 1993;167:21–9. [12] Stavudine Investigator Brochure; Bristol-Myers Squibb Pharmaceutical Research Institute; 3 June, 1999. [13] Yang Z, Brundage RC, Barbhaiya RH, Sawchuk RJ. Microdialysis studies of the distribution of stavudine into the central nervous system in the freely-moving rat. Pharm Res 1997;14:865–72. [14] Straube E, Kuhlmann BK, Weimann MW, Hoetelman RH, Von Wussow P. Stavudine penetrates the blood-brain barrier. Int Conf AIDS 1998;12:84 (abstract 12354). [15] Brady K, Aldrich J, Broston R, MacGregor RR. Stavudine entry into cerebrospinal fluid of HIV-infected subjects after single and multiple doses. Int Conf AIDS 1998;12:84 (abstract 12355). [16] Haworth SJ, Christofalo B, Anderson RD, Dunkle LM. A single-dose study to assess the penetration of stavudine into human cerebrospinal fluid in adults. J Acquir Immune Defic Syndr Hum Retrovirol 1998; 17:235– 8. [17] Duncan DB. Multiple range tests and multiple F-tests. Biometrics 1955;11:1– 42. [18] Livni E, Fischman AJ, Ray S, Sinclair I, Elmaleh DR, Alpert NM, Weiss S, Correia JA, Webb D, Dahl R, Robeson W, Margouleff D, Liss R, Strauss HW, Rubin RH. Synthesis of 18F-labeled fluconazole and positron emission tomography studies in rabbits. Int J Nucl Med Biol Part B 1992;19:191–9.

621

[19] Livni E, Babich J, Alpert NM, Liu YY, Thom E, Cleeland R, Prosser BL, Correia JA, Strauss HW, Rubin RH, Fischman AJ. Synthesis and biodistribution of 18F-labeled fleroxacin. Int J Nucl Med Biol Part B 1993;20:81–7. [20] Livni E, Satterlee W, Robey RL, Alt CA, Van Meter EE, Babich JW, Wheeler WJ, O’Bannon DD, Thrall JH, Fischman AJ. Synthesis of 11C dapoxetine HCl, a serotonin re-uptake inhibitor: Biodistribution and preliminary PET imaging in the monkey. Nucl Med Biol 1994;21:669–75. [21] Fischman AJ, Alpert NM, Livni E, Ray S, Sinclair I, Callahan RJ, Correia JA, Webb D, Strauss HW, Rubin RH. Pharmacokinetics of 18 F-labeled fluconazole in normal human subjects studied with positron emission tomography. Antimicrob Agents Chemother 1993; 37:1270 –1277. [22] Fischman AJ, Livni E, Babich J, Alpert NM, Liu YY, Thom E, Cleeland R, Prosser BL, Correia JA, Strauss HW, Rubin RH. Pharmacokinetics of [18F] fleroxacin in normal human subjects studied with positron emission tomography. Antimicrob Agents Chemother 1993;37:2144 –52. [23] Stavudine Investigator Brochure; Bristol-Myers Squibb Pharmaceutical Research Institute; 3 June, 1999. [24] Inoue T, Tsushita K, Itoh T, Ogura M, Hotta T, Saneyoshi M, Yoshida S, Saitoh H, Tomoda Y, Nagai Y. In vitro bone marrow toxicity of nucleoside analogs against human immunodeficiency virus. Antimicrob Agents Chemother 1989;33:576 –9. [25] Schreiber SL, Ikemoto N. Synthesis of chemically reactive analogs of AZT and their biological evaluation against HIV. Tetrahedron Lett 1988;29:3211– 4. [26] Balzarini J, Van Aerschot A, Herdewijn P, De Clercq E. 5-Chlorosubstituted derivatives of 2⬘,3⬘-didehydro-2⬘,3⬘-dideoxyuridine, 5⬘fluoro-2⬘,3⬘dideoxyuridine and 3⬘-azido-2⬘,3⬘-dideoxyuridine as antiHIV agents. Biochem Pharmacol 1989;38:869 –74. [27] Baba M, Pauwels R, Herdewijn P, De Clercq E, Desmyter J, Vandeputte M. Both 2⬘,3⬘-didexythymidine and its 2⬘,3⬘-unsaturated derivative (2⬘,3⬘-dideoxythymidinene) are potent and selective inhibitors of human immunodeficiency virus in vitro. Biochem Biophys Res Commun 1987;142:128 –34. [28] Hamamoto Y, Nakashima H, Matsui T, Matsuada A, Ueda T, Yamamoto N. Inhibitory effect of 2⬘,3⬘-didehydro-2⬘,3⬘dideoxynucleosides on infectivity, cytopathic effects, and replication of human immunodeficiency virus. Antimicrob Agents Chemother 1987;31:907–10. [29] Balzarini J, Kang GJ, Dalal M, Herdewijn P, De Clercq E, Broder S, Johns DG. The anti-HTLV-III (anti-HIV) and cytotoxic activity of 2⬘,3⬘didehydro-2⬘,3⬘-dideoxyribonucleosides: a comparison with their parental 2⬘,3⬘-dideoxyribonucleosides. Mol Pharm 1987;32:162–7. [30] Herdewijn P, Balzarini J, DeClercq E, Pauwels R, Baba M, Broder S, Vanderhaeghe H. 3⬘-substituted 2⬘,3⬘-dideoxynucleoside analogs as potential anti-HIV (HTLV-III/LAV) agents. J Med Chem 1987;30:1270–8. [31] Mansuri MM, Starrett JE, Ghazzouli I, Hitchcock MJ, Sterzycki RZ, Brankovan V, Lin TS, August EM, Prusoff WH, Sommadossi JP, Starrett JE. 1-(2,3-Dideoxy-D-glycero- pent-2-enofuranosyl) thymidine. A highly potent and selective anti-HIV agent. J Med Chem 1989;32:461– 6. [32] Lin TS, Schinazi RF, Prusoff WH. Potent and selective in vitro activity of 3⬘-deoxythymidine-2⬘-ene (3⬘-deoxy-2,3⬘-didehydrothymidine) against human immunodeficiency virus. Biochem Pharmacol 1987;36:2713– 8. [33] Cretton EM, Zhou Z, Kidd LB, McClure HM, Kaul S, Hitchcock MJM, Sommadossi JP. In vitro and in vivo disposition and metabolism of 3⬘-deoxy-2⬘, 3⬘didehydrothymidine. Antimicrob Agents Chemother 1993;37:1816 –25.