Pharmacokinetics of intracellular zidovudine and its phosphorylated anabolites in the absence and presence of stavudine using an in vitro human peripheral blood mononuclear cell (PBMC) model

I\I,H\\no\\I

,Ol H\\I

II,

Antimicrobial Agents ELSEVIER

International Journal of Antimicrobial Agents 9 (1997) 131- 135

Pharmacokinetics of intracellular zidovudine and its phosphorylated anabolites in th.e absence and presence of stavudine using an in vitro human peripheral blood mononuclear cell (PBMC) model Suzanne

R. Brody a,b, Francesca

T. Aweeka b,*

a Department of Clinical Pharmacy, University of California, San Francisco, CA 94143, USA b Division of Clinical Pharmacology, University of California San Francisco, CA 94143, USA

Accepted 8 August 1997

Abstract Both zidovudine (ZDV) and stavudine (D4T) must be intracellularly converted to their respective active triphosphate anabolites (ZDV-TP and D4T-TP). It is hypothesized that the combination of ZDV and D4T may lead to altered formation of phosphorylated anabolites for either drug. The objective of this study was to investigate the effect of D4T on intracellular ZDV phosphorylation. Human PBMCs were incubated with [3H]ZDV in the presence and absence of D4T. Cells were harvested at several time points over 12 h to determine area under the intracellular concentration versus time curve (AUC) of ZDV and its phosphorylated anabolites. Radiolabled ZDV and anabolites were quantified using HPLC and LS. The AUC for ZDV-TP was 0.53 and 0.52 pmol.h/lO’j PBMC in the absence and presence of D4T, respectively. The AUC for ZDV monophosphate was 157.45 and 172.44 pmol.h/106 PBMC pre and post D4T. D4T does not appear to affect the formation of intracellular ZDV phosphates in human PBMCs under the conditions studied. 0 1997 Elsevier Science B.V. Keywords:

Zidovudine;

Stavudine;

Intracellular;

Drug interaction

1. Introduction Zidovudine is the first nucleoside analog to be approved by the FDA for the treatment of Human Immunodeficiency Virus (HIV) disease in humans and continues to be the most commonly prescribed antiretroviral for patients infected with HIV. Clinically, therapy with zidovudine in combination with other antiretrovirals, including other nucleoside analogs, is part of the currently recommended management of HIV disease [l]. Zidovudine works by inhibiting the HIV reverse transcriptase, but it first must be activated by intracellular kinases to its triphosphate anabolite.

* Corresponding author. Present address: UCSF Drug Research Unit, University of California, San Francisco, 521 Parnassus Avenue, Room C-152, Box 0622, San Francisco, CA 94143-0622, USA. Tel.: + 1 415 4760339; fax: + 1 415 4760307; e-mail: [email protected] 0924-8579/97/$32.00 0 1997 Elsevier Science B.V. All rights reserved. PIISO924-8579(97)00041-l

The mechanism of phosphorylation is well established [2,3] but the factors which may alter or influence this process are less understood. Enhancement or inhibition of phosphorylation may result in an increased or decreased antiviral effect of zidovudine. It is generally believed that both zidovudine and stavudine are phosphorylated intracellularly to their mono-, di-, and triphosphate anabolites by the same enzymatic pathways, namely thymidine kinase, thymidilate kinase and nucleotide diphosphate kinase respectively, [4] although some evidence exists to the contrary [5]. The AIDS Clinical Trials Group (ACTG) has embarked on a study to investigate the clinical effect of combination nucleoside therapy in zidovudine experienced patients, including the combination of zidovudine and stavudine [6]. During the course of the study, the zidovudine/stavudine combination arm was prematurely discontinued because CD4 cell counts decreased

132

S.R. Brody, F.T. Aweeka /International

Journal of Antimicrobial Agents 9 (1997) 131-135

in patients receiving this combination; the drop in CD4 cells was approximately 70 cells/mm3 lower in patients receiving the combination as compared to the stavudine monotherapy arm (Diane Havilr, M.D., protocol Chair, ACTG 290, personal communication). Other investigators have found that, in vitro, the combination of zidovudine and stavudine may be antagonistic [7,8]. One potential explanation for these recent findings is that zidovudine and stavudine have altered pharmacokinetic disposition characteristics during concomitant therapy. In light of the recent evidence for intracellular antagonism between zidovudine and stavudine previously mentioned, it is believed that the intracellular interaction between these two drugs should be investigated further. This study was conducted to determine the pharmacokinetics of intracellular zidovudine and its phosphorylated anabolites in the absence and presence of stavudine in human PBMCs.

2. Methods 2.1. Reagents

Ficoll-Paque (Pharmacia Biotechnology, Piscataway, NJ), phosphate buffered saline and phytohemagglutinin (PBS and PHA, Sigma, St. Louis, MO), fetal bovine serum (FBS, Gemini Bio-Products, Calabasas, CA), RPMI-1640 with r_-glutamine (Irvine Scientific, Santa Ana, CA) and IL-2 natural 10 kUj50 ml (Boehringer Mannheim, Indianapolis, IN), were used as received. Radiolabled (3H) zidovudine (ZDV) and ZDV mono-, di-, and triphosphates were purchased from Moravek Biochemicals (Brea, CA). Nonlabled stavudine (D4T), ZDV and ZDV mono-, di-, and triphosphates were purchased from Sigma (St. Louis, MO). Drug free human peripheral blood mononuclear cells (PBMC) were obtained from healthy volunteers (Irwin Memorial Blood Center, San Francisco, CA). All other solvents and chemicals were of high performance liquid chromatography (HPLC) grade and obtained from Fisher Scientific (Fair Lawn, NJ). 2.2. Experimental procedure PBMCs were isolated by Ficoll-Paque centrifugation from buffy coat obtained from the blood of healthy volunteers. The Buffy coat was diluted 1: 1 with PBS (warmed to 37°C) and 20 ml was layered onto 15 ml of Ficoll-Paque in a 50 ml conical Falcon@ tube (Fisher Scientific, Fair Lawn, NJ) and centrifuged at 500 x g for 30 min. Cells were washed twice, counted and resuspended in stimulation media which consisted of RPMI-1640 supplemented with 10

pug/ml PHA, 10 U/ml IL-2, 0.05 mg/ml gentamicin and 20% FBS. The PBMCs were incubated with PHA for 48 h to activate them, then washed, counted and resuspended in RPM1 growth media supplemented with 20% FBS. 5 U/ml IL-2 and 0.05 mg/ml gentamicin at a cell concentration of 4 x lo6 cells/ml. The cells were separated into two groups: a control group (ZDV alone) and an experimental group (ZDV plus D4T). The control group was incubated overnight in the growth media alone and the experimental group was pre-incubated overnight with D4T 4 PM. The following morning, cells in both groups were counted and placed onto 6 well plates ( = 20 x lo6 cells per well) in duplicate for each time-point (0.5, 2, 5, 7, 10 and 12 h after adding ZDV) for which cells would be harvested. Each well was spiked with radiolabled [3H]ZDV (12 FCi per well, final concentration ZDV 4 PM) and the time spiked was recorded. At each timepoint (0.5, 2, 5, 7, 10 and 12 h) following the addition of ZDV, the cells in each well were harvested with ice cold PBS and centrifuged at 1000 x g for 15 min, washed twice and the pellet was subsequently resuspended in 2 ml cold PBS. The cells were counted and then 3 cc 100% methanol was added (to make up a 60% methanol solution for extracting the intracellular ZDV and anabolites). The cells were extracted overnight at - 20°C. The extracts were then concentrated under nitrogen, transferred into a 1.5 ml EppendorP tube and dried in a Speed Vaca (Savant Instruments, Farmingdale, NY) centrifuge. The dried extracts were stored at - 80°C to await subsequent HPLC analysis. 2.3. Determination of ZDV and phosphate anabolites The cell extract residues were reconstituted in 150 ~1 of HPLC mobile phase and 130 ~1 was injected onto the HPLC column. The separation was performed as described previously by Peter et al. utilizing identical HPLC equipment and mobile phase [9]. After the four fractions (ZDV and mono-, di- and triphosphate) were collected from the HPLC, they were dried under nitrogen and reconstituted in 0.5 ml phosphatase buffer, dephosphorylated with alkaline phosphatase at 37°C for 20 h and then underwent solid phase extraction as previously described [9]. The samples from solid phase extraction were dried under nitrogen and the residues were reconstituted in 500 ~1 distilled water. Following the reconstitution, 450 ~1 was added to 5 ml of the scintillation cocktail (ScintiVerse II, Fisher Scientific, Fair Lawn, NJ) in 7-ml vials. The samples were then counted in the liquid scintillation counter (Beckman LS 3801, Beckman Instruments, Irvine, CA) to quantify the intracellular ZDV and anabolites.

S.R. Brady, F.T. Aweeka /International Journal of Antimicrobial Agents 9 (1997) 131-135 Table 1 Intracellular concentrations and area under the intracellular concentration and presence of stavudine Zidovudine phosphate anabolite

‘Time (h) following ZDV incubation

133

versustimecurves(AUC) of zidovudine and phosphates in the absence

Intracellular concentration lion PBMC

pmol/mil-

AUC pmol.h/million

Without D4T

With D4T

Without D4T

PBMC

With D4T

Zidovudine (ZDV)

0.5 2 4 7 IO 12

1.06 0.82 0.8 0.98 1.02 1.34

1.05 0.96 0.87 1.03 1.16 1.28

11.26

12.14

ZDV-monophosphate

0.5 2 4 7 IO .12

2.19 1.66 10.49 16.34 17.88 21.49

2.39 8.86 12.89 14.71 20.89 30.70

157.45

172.44

ZDV-diphosphate

0.5 2 4 I 10 12

0.014 0.029 0.037 0.049 0.055 0.092

0.012 0.032 0.039 0.047 0.060 0.100

0.52

0.54

ZDV-triphosphate

0.5 2 4 I 10 12

0.026 0.044 0.036 0.046 0.054 0.069 -

0.021 0.039 0.041 0.044 0.054 0.069

0.53

0.52

3. Results

Our work with this in vitro PBMC model has demonstrated that human PBMCs derived from the buffy coat of healthy volunteers maintain greater than 90% viability under the study conditions described based on cell counting with trypan blue (data not shown). The intracellular concentrations (in pmol/million PBMC) of zidovudine and each phosphorylated anabolite for both the control (without D4T) and experimental (with D4T) groups at each timepoint, as well as the area under the intracellular concentration versus time curve (AUC) 0- 12 h in pmol.h/million PEIMC for zidovudine and each anabolite (calculated using the linear-log trapezoidal rule [lo]) are shown in Table 1. All results are based on the average concentration from two measurements. There was no observable difference between the control or experimental group in the intracellular concentrations at each timepoint or AUCs for either zidovudine or any of the phosphorylated anabolites as illustrated in the graphical representation of the data (Fig. 1). 4. Discussion Results for this current study indicate that, using an

-

in vitro human PBMC model, formation of zidovudine phosphorylated anabolites (as determined by unextrapolated AUC, O-12 h) is not altered in the presence of stavudine. This data is consistent with previous work. Most recently, Hoggard et al. [l l] reported that total zidovudine phosphate formation after co-incubation of zidovudine and stavudine for 5 h was not changed. This earlier study is limited by the fact that total phosphates were reported rather than individual phosphates; it is important to quantify individual phosphate formation since it is not possible to estimate the active triphosphate concentration by measuring the total phosphorylated compounds. Additional work also suggests that the formation of zidovudine phosphates is not changed by the presence of stavudine under similar conditions in MOLT-4 cells [12] and CEM cells [13]. However, these previous investigations of the potential intracellular interaction between zidovudine and stavudine evaluated ZDV and anabolite concentrations at a single point in time. The study by Ahluwalia et al. utilized a potentially suboptimal concentration of zidovudine (plasma levels of 0.5 PM are not clinically observed) and the study by Ho et al. did not provide quantitative information on the formation of intracellular zidovudine phosphates, even at the single timepoint sampled.

S.R. Brady, F.T. Aweeka /International Journal of Antimicrobial Agents 9 (1997) 131-135

134

--o-t-

5

Time (hr)

10

* --t

o.ooi;- - . .

I

5

.

.

15

5

nmo

ZDV-MP ZDV-MPID4T

IO

ZDV-DP ZDV-DPlD4T

.

Tlme (hr)

-

I

.

.

15

(hr)

e --c .

1

ZDV-TP ZDV-TP/U4T

,

”

5

Tlme (hr)

10

Fig. 1. Intracellular zidovudine and zidovudine mono-, di-, and triphosphate concentration vs. time curves (graphs A, B, C and D. respectively) for both zidovudine alone (0) and zidovudine in combination with D4T (0).

More definitive information is obtained by determining the intracellular concentration of each phosphorylated zidovudine anabolite over time (as done in this present work), thereby determining the AUC, a more complete measure of zidovudine phosphate formation. In addition, we preincubated the PBMCs with stavudine in order to determine if the pre-formation of stavudine phosphates would alter the formation of zidovudine phosphates. This was done because the rate limiting step in the phosphorylation pathway for stavudine is the formation of the monophosphate by thymidine kinase [13], and once formed may compete with zidovudine monophosphate for thymidilate kinase and for subsequent formation of zidovudine diphosphate (the rate limiting step for zidovudine phosphorylation). The plasma pharmacokinetics of zidovudine and stavudine in combination have not been investigated, however, drug interaction studies of nucleoside analogs at the plasma level are of limited clinical value since the active anabolites are not measured at the site of pharmacologic activity. Furthermore, it has

been shown by Peter et al. that there is no direct linear correlation between plasma zidovudine levels and intracellular zidovudine phosphate levels [ 141. Therefore, it is important to investigate the intracellular pharmacokinetics of zidovudine in the presence of agents that are likely to be used in combination with it clinically. The results of this study do not explain the apparent negative clinical effects observed in the ACTG 290 trial. However, the ACTG 290 clinical findings may be explained by the effect of zidovudine on the phosphorylation of stavudine. As stavudine has a 600-fold lower affinity for thymidine kinase than zidovudine [13], it is possible that zidovudine successfully competes with stavudine for this initial phosphorylation step, thereby lowering intracellular stavndine triphosphate concentrations. It has been shown in single timepoint in vitro studies that the intracellular levels of stavudine phosphates are decreased in the presence of zidovudine [ 11 - 131. In this study, we have provided complete in vitro findings indicating that stavudine has no apparent ef-

S.R. Brady, F.T. Aweeka /International Journal of Antimicrobial Agents 9 (1997) 131-135

feet on the intracellular phosphorylation of zidovudine. To explain the observed clinical results associated with zidovudine and stavudine (combination therapy, further in vivo and in vitro studies investigating a potential intracellular interaction bletween these two drugs are warranted.

Acknowledgements Support for this project was provided entirely by the UCSF Drug Research Unit at San Francisco General Hospital, San Francisco, CA.

References [1] International AIDS Society, USA. Antiretroviral therapy for HIV Infection in 1996; recommendations of an international panel. J Am Med Assoc 1996;276:146-154. [2] Furman PA, Fyfe JA, St. Clair MH, et al. Phosphorylation of 3’-azido-3’-deoxythymidine and selective interaction of the 5’triphosphate with human irnmunodeficiency virus reverse transcriptase. Proc Natl Acad Sci USA 1986;83:8333-7. [3] Tomevik Y, Jacobsson B, Britton S, Eriksson S. Intracellular metabolism of 3’azidothymidine in isolated human peripheral blood mononuclear cells. AIDS Res Hum Retroviruses 1991;7:751-9. [4] Sommadossi J. Comparison of metabolism and in vitro antiviral activity of stavudine versus other 2’,3’-dideoxynucleoside analogues. J Infect Dis 1995;171 Suppl 2:%X8-92. [5] Balzarini J, Herdewijn P, De Clercq E. Differential patterns of 2’,3’-didehydro-2:3’metabolism of intracellular

13.5

dideoxythymidine and 3’-azido-2’,3’-dideoxythymidine, two potent anti-human immunodeficiency virus compounds. J Biol Chem 1989;264:6127-33. 161Pollard RB. D4T nucleoside combinations for HIV. Antiviral Res 1996;29:101-4. [71 Merrill DP, Moonis M, Chou T-C, Hirsch MS. Lamivudine or stavudine in two- and three-drug combinations against human immunodeficiency virus type 1 replication in vitro. J Infect Dis 1996;173:355-64. PI Zhu Q-Y, Scarborough A, Polsky B, Chou T-C. Drug combinations and effect parameters of zidovudine, stavudine, and nevirapine in standardized drug-sensitive and resistant HIV type 1 strains. AIDS Res Hum Retroviruses 1996;12:507-17. [91 Peter K, Lalezari JP, Gambertoglio JG. Quantification of zidovudine and individual zidovudine phosphates in peripheral blood mononuclear cells by a combined isocratic high performance liquid chromatography radioimmunoassay method. J Pharm Biomed Anal 1996;14:491-9. VOI Gibaldi M. Noncompartmental pharmacokinetics. In: Biopharmaceutics and clinical pharmacokinetics, 3rd ed. Philadelphia: Lea and Febiger, 1984:1728. Pll Hoggard P, Khoo S, Barry M, Back D. Intracellular metabolism of zidovudine and stavudine in combination. J Infect Dis 1996;174:671-2. [12] Ahluwalia GS, Gao W, Mitsuya H, Johns DG. 2’3’-didehydro3’deoxythymidine: Regulation of its metabolic activation by modulators of thymidine-5’-triphosphate biosynthesis. Mol Pharmacol 1996;50:160-5. [13] Ho H, Hitchcock MJM. Cellular pharmacology of 2’,3’dideoxy-2’,3’-didehydrothymidine, a nucleoside analog active against human immunodeficiency virus. Antimicrob Agents Chemother 1989;33:844-9. [14] Peter K, Gambertoglio JG. Zidovudine phosphorylation after short-term and long-term therapy with zidovudine in patients infected with the human immunodeficiency virus. Clin Pharmaco1 Ther 1996;60:168-76.