The disposition of the new arabinosylcytosine derivative—5′-chloro-5′-deoxy-arabinosylcytosine—in rats

Pergamon

0306-3623(94)00275-4

Gen. Pharmac.Vol. 26, No. 5, pp. 1101-1105,1995 Copyright © 1995ElsevierScienceLtd Printed in Great Britain.All rights reserved 0306-3623/95$9.50+ 0.00

The Disposition of the New Arabinosylcytosine Derivative 5'-Chloro-5'-Deoxy-Arabinosylcytosine Rats

in

L A D I S L A V N O V O T N Y , l* V I E R A R E I C H E L O V A , I IVO J A N K I ~ ) P E T E R R A U K O I and VILIAM U J H A Z Y I 1Department of Experimental Therapy, Cancer Research Institute, Slovak Academy of Sciences, ,~pit,ilska 21, 812 32 Bratislava, Slovak Republic and :Institute of Pharmacology, Academy of Sciences of Czech Republic, Videtiskd 1083, 14000 Prague 4--Kr~, Czech Republic (Received 1 September 1994)

Abstract--1. Pharmacokinetic properties of a new derivative of the widely used and very potent antileukemic agent arabinosylcytosine (araC)--5'-chloro-5'-deoxy-arabinosylcytosine (5'-Cl-araC)--were investigated after intraperitoneal (i.p.) and oral routes of administration in rats and compared with the equimolar dose of araC administered orally. 2. It was found that substitution of the hydroxyl group at position 5' resulted in a change of pharmacokinetic parameters. 3. There is a large difference in average serum concentrations of 5'-Cl-araC administered by the i.p. and oral routes; the average serum concentration obtained after i.p. injection being several times higher in comparison to those after oral administration. 4. However, the latter are, at the same time, lower than the average serum concentrations of araC administered by the same route in an equimolar dose. 5. On the other hand, the apparent volume of distribution is much larger, and the area under the curve of serum concentration of 5'-Cl-araC is smaller, after oral as compared to the i.p. route of administration indicating more extensive tissue distribution together with higher tissue binding of 5'-Cl-araC when compared to the parental drug araC.

Key Words:AraC, araC derivative, pharmacokinetics, 5'-chloro-araC, HPLC

INTRODUCTION

(Novotn~ et al., 1984) and also when incubated with isolated deaminase (Kfira et al., 1982). This obserA new synthetic nucleoside analogue of naturally vation was supported by a detailed physico-chemical occurring cytidine--5'chloro-5'-deoxy-l-fl-o-arabinostudy (Birnbaum et al., 1988). Moreover, it has a furanosyicytosine (5'-Cl-araC, Fig. l) (H~ebabeck~ higher transport rate than araC (Novotn~ et al., et al., 1980) is a non-polar derivative of araC, the 1984). Out of the 5'-deoxy-derivatives of araC, most potent anti-leukemic agent widely used in the 5'-CbaraC exhibits a high inhibition of DNA syntreatment of leukemia in man (Chabner, 1982). thesis (Ber~inek and Acton, 1984). In conjunction The studies on structure-activity relationships of with the mechanism of action, it was found that nucleoside analogues (Novotn~ et al., 1984; Berfinek, 1986) indicate that 5"-deoxy (including 5'-halogeno) 5'-Cl-araC could be directly phosphorylated to araC derivatives of araC represent a new potentially 5'-triophosphate, the biologically active form of araC interesting group of biologically active compounds. (Novotn~, et al., 1991). The results stimulate further 5'-Cl-araC was found to be very stable towards investigation of the mechanism of action and the metabolic alterations during intestinal transport biological activity in in vitro experiments. The aim of the pharmacological study of 5'-Cl-araC was to *To whom all correspondence should be addressed. improve the efficiency and to overcome the problems 1101

1102

Ladislav Novotn~, et al. NH2

NH2

0

W

HO

0

HO

araC 5' - chloro-araC Fig. 1. Chemical structure of araC and its new analogue 5'-Cl-araC. of tumor cell resistance and lack of therapeutic selectivity of araC through biochemical modulation (Grant, 1990). In this paper we report on the disposition of 5'-Cl-araC in rats compared with araC after intraperitoneal (i.p.) and peroral (p.o.) administration. MATERIALS AND METHODS Chemicals

Arabinosylcytosine (araC) (NSC 63878) and 5'-CIarabinosylcytosine (5'-Cl-araC) (NSC 318799) were prepared as described elsewhere (H~ebabeck~ et al., 1980). Dihydrogen potassium phosphate was obtained from Merck (Darmstadt, Germany), methanol and potassium hydroxide were of the highest purity available from Lachema (Brno, Czech Republic). Double-distilled deionized water was purified using Elgastadt Spectrum SC 20, Elga Ltd (Lane End, Wycombe, England).

samples were collected after decapitation of six animals at the following times using both methods of administration: 5, 10, 15, 30, 60 and 120rain after i.p. administration of 5'-Cl-araC and p.o. administration of araC and at 240 and 360 min after p.o. administration of 5'-Cl-araC. H P L C assay

A reversed-phase isocratic HPLC chromatography was used for the determination of 5'-CI-araC in the rat serum after i.p. and p.o. administration of the drug and for the araC determination after p.o. administration. The blood serum was collected and centrifuged at 2800 rpm for 15 min at 4°C. Ice-cold methanol was added to the blood serum (1:i, v/v) and the mixture was centrifuged at 17,000 rpm at 4°C for 15 min. The supernatant was injected onto the column. We used reversed-phase columns Separon SGX C-18 (150 x 3 mm i.d., 7/~m). The mobile phase for Cl-araC determination consisted of 0.01 M potassium dihydrogen phosphate with 8% of methanol, pH5.8. Detection was performed at 275 nm and using a flow-rate of 0.45 ml/min. AraC was assayed at the same conditions as 5'-Cl-araC with the exception of the mobile phase. A mobile phase of 0.01 M potassium dihydrogen phosphate with 3% of methanol, pH5.8 was used for the araC determination. Data analysis

The relationship between drug concentration and time was described using the one compartmental pharmacokinetic model of drug disposition:

H P L C instrumentation

The HPLC equipment consisted of a Pye Unicam PU 4003 gradient elution high performance liquid chromatograph (HPLC) equipped with a Rheodyne 7125 model injector (20#1 sample loop), PU 4030 controller, PU 4031 column oven and PU 4020 u.v. detector with a variable wavelength (Philips Scientific, Cambridge, England). A dual channel SP 4200 computing integrator (Spectra Physics, Darmstadt, Germany) was used to record detector signals. Animal studies

Male Wistar rats (150-170g body wt) were obtained from Velaz (Prague, Czech Republic) and fed on a pellet diet with water ad libitum. The animals were divided into three parts. The first part of the animals were given an i.p. dose of 54mg/kg (equimolar dose to 50 mg/kg of araC) and the second part was administered 54 mg/kg of 5'-Cl-araC perorally. The last part of the animals got 50 mg of araC per body wt kg perorally. The blood

C(t)=

FD Vd

ka --(exp(-kj)-exp(-kat)) k~-ke

(1)

where F is the fraction of dose D reaching the systemic circulation, I'd and distribution volume of the drug, k~ and ke represent the first order rate constants of absorption and elimination, respectively. The fitting of individual concentration values to equation (1) and estimation of the parameters ka, ke and V J F (the apparent volume of distribution) was executed on a desk top Hewlett-Packard 86B computer using an iteration procedure based on the Levenberg-Marquardt modification of the GaussNewton minimization algorithm. Weighting concentration data by the reciprocal of the predicted value contributed to a more random scatter of the residuals around zero as well as to an increased precision of parameter estimates. The estimates of first order rate constants k a and ke were converted to corresponding half-lifetimes (tt/2)

New arabinosylcytosine derivative in rats of absorption and elimination, respectively, using a common formula: In 2 tl/2 = k

(AUC) could be calculated as: SD.s SD e = P . -

(2)

where t j~2 is the half-lifetime and k the corresponding first order rate constant of absorption or elimination. The apparent total drug clearance of the drug from the body was then calculated as: C L / F = k e • Vd/F

where n and S D n are the estimate and asymptotic standard deviation of the original parameter (rate constant or apparent total clearance) and P and SDp

(A) 200

(3)

and the area under the curve of Concentration in serum according to the formula: FD A UC = - - . CL

1103

"~

120

~

4o

(4)

Precision o f parameter estimates

iL 13

Asymptotic standard deviations for directly estimated parameter values were calculated as the square root of the corresponding diagonal element of the variance-covariance matrix A defined as:

15

t

I

I

30

60

120

Time (rain) 14

-- ( S )

12 A = S~es " (FTF) -1

(5)

where S~es is the residual variance around the predicted time-concentration curve, F represents the matrix of partial derivatives of the predicted concentration values at times of measurement with respect to directly estimated parameters of the model, F T the transpose of the same matrix.

lO o :=L

a

i

U

r~'

4

Calculation o f standard deviations f o r derived pharmacokinetic parameters

The asymptotic standard deviation of the total apparent clearance was estimated using the formula:

ill l

I

I

I

I

30

70

120

240

360

0

Time (min)

SDcL, r = (V2 / F • SD2(ke) + k2¢ . S D 2 ( V J F ) + 2" C L / F ( c o v ( k e , Vd/F)) 1/2

(6)

50

t (C)

,o

which is a special case of the standard Taylor expansion given by Boxenbaum et al. (1974) adapted for the equation (3). In equation (6) SDcL/r is the asymptotic standard deviation of the apparent total clearance, SD(ke) and S D ( V d / F ) the asymptotic standard deviations of k¢ and Vd/F respectively, and cov(ke, V J F ) the covariance of the two latter parameters given by the corresponding element of the variance-covariance matrix. Adaptation of the general relationship of Boxenbaum et al. (1974) to equations (2) and (3) showed that the asymptotic standard deviations of the half-lifetimes of absorption and elimination as well as of the area under the drug concentration

lO

0

15

I

I

I

30

60

120

Time (rain) Fig. 2. Serum concentrations in rats after: (A) i.p. administration of 5'-Cl-araC; (B) peroral administration of 5'-ClaraC; (C) peroral administration of araC. (Each point represents an average from the determinations of serum concentrations performed in the group of six animals + SD.)

1104

Ladislav Novotn~, et al.

the estimates and asymptotic standard deviations of derived parameter (half-lifetime and AUC). F o r testing the statistical significance of differences between parameter estimates from studies of different compounds the t was calculated according to the following equation (Boxenbaum et al., 1974): t =

P~i Pik (SD~ + SD~,) ''2 -

(7)

-

where P~i and SD~j are the estimate and asymptotic standard deviation of the parameter i of c o m p o u n d j and Pik and SDik the estimate and asymptotic standard deviation of the same parameter of the compound k. The value of the t statistic was then compared with the tabulated value for Nj + N k - 2Np d.f. where N / a n d Nk are the number of observations in the studies i and j, respectively, whereas Np represents the number of directly estimated parameters of the pharmacokinetic model. These tests were performed to test the statistical significance of differences between estimates of pharmacokinetic parameters of 5'-Cl-araC after both routes of administration on one hand and the differences in values of pharmacokinetic parameters of 5'-Cl-araC after oral administration as compared to those found for orally administered araC on the other hand. Statistical significance was tested at P = 0.05 and 0.01 probability levels. RESULTS Time-course o f serum concentrations The time--courses of average serum concentrations of 5'-Cl-araC after i.p. and p.o. administrations to groups of rats as compared to average serum concentrations obtained after oral administration of equimolar dose of araC are depicted in Fig. 2. It can be seen that there is a large difference in average serum concentrations of 5'-Cl-araC between each route of administration, the average serum concentration obtained after i.p. injection being several times higher compared to those after oral administration. How-

ever, the latter are at the same time lower than the average serum concentrations of araC a d m i n i s t e r e d by the same route in an equimolar dose. Pharmacokinetie parameters The estimates of pharmacokinetic p a r a m e t e r s which were obtained by a n a l y s i s of individual s e r u m concentration data are summarized in Table 1. When oral administration of 5'-Cl-araC is compared with its i.p. injection it is evident that there is a slower absorption of the compound from the gastrointestinal tract than from the peritoneal cavity (P < 0.01) as shown by the absorption half-lifetimes. However, the values of the elimination half-lifetimes demonstrate that there is, at the same time, also slower elimination of 5'-Cl-araC (P < 0.05) when given orally. The apparent volume of distribution is much larger (P < 0.01) and the area under the curve (AUC) of serum concentration of 5'-Cl-araC, on the other hand, smaller (P < 0.05) after oral as compared to the i.p. route of administration. The only statistically insignificant difference between both routes is in their total serum clearances. When comparison is made between pharmacokinetic parameters of 5'-Cl-araC and araC then the values of the absorption half-life show that only oral absorption is slower (P < 0.05) for 5'-Cl-araC, the difference in the elimination half-lifetimes as well as in total serum clearances being statistically insignificant. On the other hand, the apparent volume of distribution is larger (P < 0.01) for 5'-Cl-araC as compared to araC whereas the opposite is true for the areas under the curves of serum concentrations (P < 0.05). DISCUSSION 5'-Cl-araC is a new derivative of the anti-leukemic agent araC. Its advantages when compared to araC consist of the better membrane transport (Novotn2~ et aL, 1984) and high stability towards deamination

Table 1. Pharmacokineticparameters of 5'-CI-araCafter i.p. and oral administration in rats as compared to araC administered orally 5"-Cl-araC 5'-Cl-araC araC i.p. C.V. (%) p.o. C.V. (%) p.o. C.V. (%) tma (min) 4.57 _+1.06 23.27 14.84 + 3.07 20.70 6.58 + 1.84 27.89 a,d

it/2a (min)

33.77 + 3.05

9.04

480.65+ 197.19

41.03

291.96_+82.84

28.38

12.60

6.23 _+0.81

13.08

30.53 30.53

14.72 _+2.87 13.97 +_2.72

19.49 19.49

c

Vd (I/kg)

0.96 + 0.10

10.56

22.33 + 2.81 b,d

CL (ml/min/kg) ,4UC (mM rain)

19.50 + 1.07 10.55 + 0.58

5.48 5.48

32.03 + 9.78 6.42 + 1.96 a,c

*Significantly different from bSignificantly different from CSignificantly different from aSignificantly different from

araC administered p.o. (P < 0.05). araC administered p.o. (P < 0.01). 5'-CI-araC administered i.p. (P < 0.05). 5'-Cl-araC administered i.p. (P < 0.01).

New arabinosylcytosine derivative in rats by the main araC degrading enzyme--cytidine deaminase (K/ira et al., 1982). Because it possesses an anti-leukemic activity (Novotn~ et al., 1991) and because some physico-chemical properties have been changed (i.e. lipophilicity, Novotn# et al., 1984, 1991), it is important to investigate how the structural change of the araC molecule (substitution of -OH group at the position 5' by the chlorine atom) influenced its pharmacokinetic properties using various ways of administration. When comparing the pharmacokinetic parameters of 5'-Cl-araC after i.p. and oral routes of administration it is not surprising that the oral absorption half-life is longer compared to that obtained after the i.p. injection because it is well known that absorption of main compounds from the peritoneal cavity is usually very rapid. More remarkable is the very large difference in the volumes of distribution. Although one cannot exclude that this difference is at least partly due to metabolic transformation of the compound during intestinal absorption or during the initial passage through the liver portal system, such a difference might indicate more extensive penetration of 5'-Cl-araC into tissues associated with higher tissue binding. The latter could offer an explanation for the statistically slower elimination of the compound when given orally. Similar reasons may explain the approx. 60% availability of 5'-Cl-araC in systemic circulation after oral administration as compared to i.p. injection. Although it has been shown previously (Kfira et al., 1982; Novotn~ et al., 1984) that 5'-Cl-araC is more stable in vivo than araC, the results of our pharmacokinetic study do not support this expectation when availability of both compounds in the systemic circulation is considered by comparing the areas under the curves of concentration of both compounds because the relative availability of 5'-Cl-araC represents only approx. 46% of that of araC. However, when higher lipophilicity of 5'-Cl-araC (Novotn~, et al., 1984) is

GP 26/~D

1105

taken into consideration it is clear that not only presystemic elimination, but also more extensive tissue distribution, together with higher tissue binding, might be responsible for lower serum concentrations of 5'-Cl-araC in comparison with araC. Acknowledgements--This work was supported by the grant

of the Slovak Academy of Sciences No. 1331 and by the Central Tax Ofl%e of Slovak Republic, Bansk/t Bystrica (Mr Vladimir Beni~ek and Mr ~tefan Jfi.gersk#). The authors are grateful to Ms Gabriela Ga~ajovfi for the excellent technical assistance. REFERENCES Ber~nek J. (1986) A study on structure-activity relationships of nucleoside analogues. Drugs exptl Clin. Res. 12, 355 367. Ber/mek J. and Acton E. M. (1984) Inhibition of nucleicacid synthesis in L210 cells by nucleoside analogs. Collect. Czech. Chem. Commun. 49, 2551-2555. Birnbaum G. I., Bud~ginsky M. and Ber/mek J. (1988) Structure and conformation of 5'-chloro-5'-deoxyarabinosylcytosine: X-ray, IH and 13C-NMR analyses. Can. J. Chem. 66, 1203-1208. Boxenbaum G., Riegelman S. and Elashoff R. M. (1974) Statistical estimation in pharmacokinetics. J. Pharmacokin. Biopharmac. 2, 123-148. Chabner B. A. (1982) Pharmacologic Principles of Cancer Treatment (Edited by Chabner B. A.), pp. 387-401. Saunders, Pa. Grant S. (1990) Biochemical modulation of cytosine arabinoside. Pharmac. Ther. 48, 29-44. Hfebabeck~ H., Brokeg J. and Berfinek J. (1980) Reaction of nucleosides with thionyl chloride, preparation of the deoxy derivatives of cytidine and adenosine. Collect. Czech. Chem. Commun. 45, 599~05. K~ra J., B~rtovfi M., Ryba M., Hfebabeck# H., Broke~ J., Novotn~, L. and Ber~tnek J. (1982) Sensitivity of some arabinosylcytosine derivatives to enzymatic deamination by cytidine deaminase from mouse kidney. Collect. Czech. Chem. Commun. 47, 2824-2830. Novotn~, L., Farghali H., Ryba M., Jankfi I. and Berfinek J. (1984) Structure-intestinal transport and structuremetabolism correlations of some potential cancerostatic pyrmidine nucleosides in isolated rat jejunum. Cancer Chemother. Pharmac. 13, 195-199. Novotn~, L., Reichelovfi V., Bal~.2ov~iE., Scheithauer W., Benesovfi M. and Ujhfizy V. (1991) The biotransformation and cytotoxic activity of arabinosylcytosine 5'-chloro-5'-deoxy analog. Anti-Cancer Drugs 2, 495-502.