INT . J . RADIAT . BIOL .,
1992,
VOL .
62,
NO .
2, 22 1 -227
Pharmacokinetics of fluorinated 2-nitroimidazole hypoxic cell radiosensitizers in murine peripheral nervous tissue K. SASAI*t, H . IWAI$, T. YOSHIZAWA$, S . NISHIMOTO§, Y . SHIBAMOTOt, Y. KITAKABUt, N . OYAt, M . TAKAHASHI¶ and M . ABEL
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(Received 6 September 1991 ; revision received 17 February 1992; accepted 29 February 1992)
Abstract. We have previously reported that KU-2285, a 2nitroimidazole with a fluorinated Nl-substituent (-CH 2CF2 CONH(CH 2 )„OH, n=2), was a promising hypoxic cell radiosensitizer. In this study the pharmacokinetics of KU-2285 and its related compounds (n=3 and n=4) were compared with those of entanidazole (a 2-nitroimidazole with an Nlsubstituent of -CH 2 CONH(CH 2 )„OH, n=2) and its related compounds (n=3 and n=4) to assess the effects of incorporation of a CF 2 group . The lipophilicity of the fluorinated compounds was higher than that of etanidazole, as measured by the octanol/water partition coefficient . As the number of CH 2 groups increased, the lipophilicity of the compounds in both the KU-2285 and etanidazole series increased . The brain tissue levels of the fluorinated compounds were as low as those of the etanidazole derivatives, while the biological half-lives of the fluorinated compounds in peripheral nervous tissues were shorter than those of related non-fluorinated compounds .
1. Introduction We have previously reported the pharmacokinetics of various 2-nitroimidazole hypoxic cell sensitizers in the sciatic nerve tissue of C3H/He mice (Sasai et al. 1990) . The apparent biological half-life of each compound in the peripheral nervous tissue was found to correlate with its lipophilicity, with the more lipophilic compounds having a shorter biological half-life . We have also demonstrated the KU2285, a fluorinated 2-nitroimidazole derivative with an N1 substituent (-CH 2CF2CONH(CH2 )„OH, n=2), was an effective radiosensitizer both in vitro and in vivo (Shibamoto et al. 1989, Sasai et al. 1991 a) . The radiosensitizing activity in vivo of KU-2285 was found to be similar to that of etanidazole when assayed by an in vivo-in vitro assay, a growth delay *Correspondence to : Keisuke Sasai, MD, Department of Radiology, Faculty of Medicine, Kyoto University, Sakyo-Ku, Kyoto 606, Japan . tDepartment of Radiology, Faculty of Medicine ; §Department of Hydrocarbon Chemistry, Faulty of Engineering; ¶Chest Disease Research Institute, Kyoto University, Kyoto 606 ; and I Chemical Division, Daikin Industries, Ltd, Settsu 566, Japan .
assay, and a tumour control assay using the SCC VII tumour or a transplanted mammary tumour in C3H/He mice . Although the radiosensitizing activity of etanidazole was reduced when it was administered orally, there was no significant difference in the radiosensitization produced by intravenous, intraperitoneal, or oral KU-2285 (Sasai et al . 1991a) . Furthermore, KU-2285 was found to sensitize two different murine tumours when given in combination with radiation dose fractionation (Sasai et al. 1991 b) . The acute toxicity of KU-2285 for 8-week-old female C3H/He mice was found to be higher than that of etanidazole, although it was lower than that of misonidazole . The lipophilicity of KU-2285 was similar to that of misonidazole, but our previous pharmacokinetic study demonstrated that the brain concentration of the former compound was surprisingly low (Sasai et al . 1990) . In this study the acute toxicity and the pharmacokinetics of KU-2285 and its related compounds (n=3 and n=4) were compared with those of etanidazole (a 2-nitroimidazole with an NI-substituent of -CH 2 CONH(CH 2 )„OH, n=2) and its related compounds (n=3 and n=4), to assess the effect of incorporating a CF 2 group into the side-chains of these radiosensitizers . 2. Materials and methods 2.1 .
Compounds
The structures of both etanidazole series and KU2285 series are shown in Figure 1 and Table 1 . All of the compounds used in this experiment were synthesized by our group. KU-2285, KU-3202 and KU3207 were fluorinated 2-initroimdazole derivatives with similar structures, while KU-3205 and KU3206 were compounds related to etanidazole . The only difference of the compounds in each group was the number of CH2 groups in the side-chain . For the phamacokinetic studies, all drugs were prepared in
0020-7616/92 $3 .00 © 1992 Taylor & Francis Ltd
222
K . Sasai
lecting blood from the orbital vein . Brains were removed and the sciatic nerves were dissected free of skin, connective tissue, and muscle. Because the sample volume of the sciatic nerves from one mouse was too small to assay (about 10 mg), four to eight nerves from two to four mice were used . Serum and tissue samples were weighed and placed into plastic dishes, after which they were immediately frozen and stored at -20°C prior to analysis . Just before analysis, the tissue samples were homogenized in 10 volumes of distilled water, while serum was vortexmixed after the addition of 10 volumes of distilled water . Forty volumes of methanol was then added to precipitate protein and the sample was mixed and centrifuged .
N
-
,,J NO2
Etanidazole Group
CHKCONH(CH2)nOH
N i
KU-2285 Group
2CF2CONH(CH 2)nOH
CH
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
Figure 1 .
Structural formulae of compounds of etanidazole and KU-2285 series .
physiological saline at a concentration of 10 mg/ml immediately prior to use . The octanol/water partition coefficients were measured in buffer at pH 7.4 according to the method of Fujita et al . (1964) . The octanol/water partition coefficients for KU-2285 and etanidazole differ from those of our previous reports (Shibamoto et al. 1989, Sasai et al . 1990), because of a slight modification of our experimental method .
High-performance liquid chromatography . The supernatants were analysed by reverse-phase highperformance liquid chromatography (HPLC) using a GL Science Inertsil ODS-2 column (4 . 6 x 150 mm, C 18, particle size 5µm) . The flow rate was 0 .5 ml/min . The eluents were as follows: CH 3 CN :H 2 O (10 :60), 0 .01 mol dm-3 NaH 2 PO 4, H 3 PO 4 (to pH 2 .4) for KU-2285 and KU-3202 ; CH 3 CN : H2 O (10 :40), 0 . 01 mol dm-3 NaH 2 PO 4 , H 3 PO 4 (to pH2 .4) for KU-3207 ; CH 3 CN :H 2 O (10 :60), 0.01moldm -3 NaH 2 PO 4 , H 3 PO 4 (to pH2.4) for KU-3205 ; and CH 3 CN :H 2 O (10 :120), 0.01 mol dm-3 NaH 2 PO4, H 3 PO 4 (to pH 2 .4) for KU-3206 . Peaks were detected by a Shimadzu Variable Wavelength UV Monitor SPD-6A at 330 nm ; the peak areas were determined using a Shimadzu Integrator C-R 5A. The concentrations of the com-, pounds were determined using an external standard . The efficiency of recovery of all compounds from fetal bovine serum after protein precipitation was more than 97% . The retention time of each compound was : KU-2285, 7 .3 min ; KU-3202 and its metabolite, 9 .0 and 10.0 min respectively ; KU-3207 and its metabolite, 9 .55 and 10 .4 min respectively ; KU-3205, 5 .2 min ; KU-3206 and its metabolite, 9.97 and 10 .3 min respectively. The lower limit of detection for the assay was about 0 .5µg/g (µg/ml) . The data used here for etanidazole were from our previous experiments (Sasai et al . 1990) .
2 .2. Animals In all experiments, 8-9-week-old female C3H/He mice (Crea Japan Inc ., Osaka; Japan SLC, Hamamatsu) were used . All animal experiments were performed under the guidelines for animal experiments of Kyoto University . 2.3 . Acute toxicity in mice The drug necessary to kill 50% of mice within 7 days (LD 50J7 ) was determined using 35 mice per compound) . Drugs were dissolved in 0 .5% Tween80 solution . All compounds were administered intravenously.
2 .4.
et al .
Pharmacokinetic studies
Sample preparation . All compounds were administered intravenously at a dose of 200 µg/g . At 10, 20, 40, 60, 120 and 240 min after injection, the mice were killed by decapitation immediately after col-
Table 1 . Structures of the compounds used . Value of n Etanidazole group -CH 2 CONH(CH 2 )„OH KU-2285 group -CH2 CF Z CONH(CH 2 )„OH
2
3
4
Etanidazole
KU-3205
KU-3206
KU-2285
KU3202
KU-3207
Pharmacokinetics
offluorinated
2-nitroimidazole
Data analysis . The area under the concentrationtime curve (AUC) for the period of 0-240 min was estimated using the trapezoidal rule (Rowland and Tozer 1980) . The apparent biological half-life in tissue was calculated from the best-fit lines, determined by least-squares linear regression analysis, for the exponential part of the decline phase of the drug concentration-time curve .
toogo 10-1
•
TL
3. Results •
3 . 1 . Octanol/water partition coe cient Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
223
60 120 180 240 Time After Injection (min)
Figure 2 .
Table 2 shows the octanol/water partition coefficient of each compound. The lipophilicity of the fluorinated compounds, as measured by the octanol/ water partition coefficient, was much higher than that of the etanidazole series . As the number of CH 2 groups in the side-chain increased, the lipophilicity of the compounds in each series increased .
3.2 . Distribution of the radiosensitizers Figures 2-7 show the concentrations of the six compounds in serum, brain, and peripheral nervous tissue . Metabolites of KU-3206, KU-3202, and KU3207 were also investigated and their concentrations are demonstrated in Figures 4a, 6b, and 7b. (The
Figure 3 .
Table 2 . Octanol/water partition coefficients and apparent biological half-life in various tissues of six 2-nitroimidazole hypoxic cell radiosensitizers . Apparent biological half-life (min) Octanol/water partition coefficient Etanidazole
0.040
KU-3205
0. 058
KU-3206
0 . 18
KU-2285
0. 25
KU-3202
0. 32
KU-3207
0 . 51
Serum
Peripheral nerve
66 (56-90) b 57 (32-274) 14 (11-18) 27 (20-44) 10 (8 . 2-12) 5 (3 . 2-7 . 7)
256 (150-885) 172 (92-1309) 101 ( 76-151) 167 (104-327) 45 ( 39-57) 49 ( 34-87)
AUC(0-240 min)-' (µg/g x min)
Serum
Peripheral nerve
Brain
LD50% (g/kg)
11400
6200
280
4
7600
4000
70
more than 4
4000 (3300) d 8400
2600 (970) 6400
110 (30) 370
not confirmed'
1500 (3100) 740 (3200)
1100 (930) 870 (2000)
50 (0) 20 (0)
more than 2 .5
2.3
3 .2
'AUC(0-240 min) : The area under the concentration-time curve for the period of 0-240 min of parent compound plus its metabolite if observed . b Figures in parentheses are the 95% confidence interval . `Figures in parentheses are the AUC of the metabolite of each compound .
224
K . Sasai 1000
et al .
-
E 00 z 1007 0 m c Y 0 0 0 U
01
to-
a O 0 J
1
I 60
0
I 120
I 180
240
Time After Injection (min)
(b)
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(a)
Figure 4 .
1
1000-Z
= =1
E
4.90
I
1 0
60
120
180
I 240
Time After Injection (min)
Figure 5 .
concentrations of these metabolites were normalized to the weight of the parent compounds .) The pharmacokinetics of KU-2285 injected intravenously reported here was very similar to the intraperitoneal pharmacokinetics of KU-2285 reported previously (Sasai et al . 1990) . The brain concentration of each compound was very low during the observation period . The drug concentration in the mouse sciatic nerve varied among the compounds . Low lipophilicity compounds such as those of the etanidazole group, and KU-2285, showed a relatively high concentration in peripheral nervous tissue even 240 min after intravenous injection . However, the more lipophilic fluorinated 2-nitroimidazoles (KU-3202 and KU-3207) rapidly disappeared from the peripheral nerves, and were impossible to detect 240 min after injection .
1000-
T a Z U' z " 100c 0
Y
C U
SID
10 -
D
0
J
1 1 0
60
120
180
Time After Injection
(b)
(a)
Figure 6 .
240 (min)
Pharmacokinetics offluorinated 2-nitroimidazole
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(a)
225
(b)
Figure 7 . Figures 2-7 . Concentrations of 2-nitroimidazoles in serum ( • ), brain (0), and sciatic nerve (0) in C3H/He mice as a function of time after a single intravenous injection . Etanidazole (Figure 2), KU-3205 (Figure 3), and KU-3206 (Figure 4) are nonfluorinated 2-nitroimidazoles, while KU2285 (Figure 5), KU-3202 (Figure 6) and KU-3207 (Figure 7) are fluorinated compounds . Each drug was injected as a single dose of 200,ug/g . Figures 4B, 6B and 7B show the concentrations of metabolites after i .v. injection of the parent compounds . The error bars represent 1 SD for the mean of more than four samples of more than two independent experiments . Figure 2 was drawn from the results of previous examinations (Sasai et al . 1990) .
3 .3. Apparent biological half-life Figure 8 shows the apparent biological half-lives of the six compounds in peripheral nervous tissue as a function of the octanol/water partition coefficient . The apparent half-life of etanidazole was the longest for these six compounds, and both groups of drugs showed a negative correlation between the octanol/ water partition coefficient and the apparent biological half-life in peripheral nervous tissue . Table 2 also shows the apparent biological half-lives of these compounds in serum . A negative correlation
•
3001
0
200-
•
2
0
The drug exposure to the peripheral nerve
The degree of exposure of the sciatic nerves was calculated as the AUC of the drug (plus its metabolite if observed) for the period from 0 to 240 min ; the data are summarized in Table 2 . Exposure of the peripheral nerve to drugs was found to be the greatest for etanidazole and KU-2285, while it was low for the more lipophilic fluorinated compounds such as KU-3202 and 3207 . 3.5. Acute toxicity
3 0 50 80
a
3 .4 .
40
100-
•
(r=0.96) between the octanol/water partition coefficient and the apparent biological half-life in serum was observed . In the brain, the half-lives could not be calculated because the concentrations were too low .
I
I
1 .0 0.01 0.1 Octanol/Water Partition Coefficient
Figure 8 . Plot of the apparent biological half-lvies of 2nitroimidazoles in the sciatic nervous tissue of C3H/He mice versus the partition coefficient of each drug . The correlation coefficient is 0.87 . (0) : Etanidazole group, (i) : KU-2285 group . (1) etanidazole, (2) KU-3205, (3) KU-3206, (4) KU-2285, (5) KU-3202, and (6) KU-3207 .
The acute toxicity of each compound after intravenous injection measured as the LD5017 is shown in Table 2 . The acute toxicity of the drugs of the KU2285 series was similar to or slightly higher than that of the etanidazole series of radiosensitizers . 4. Discussion After the elegant pharmacokinetic studies on 2nitroimidazoles with a wide range of octanol/water
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
226
K . Sasai
partition coefficients performed by Brown and Workman (1980), considerable effort has been devoted to the development of new drugs with a low lipophilicity which should have less peripheral neurotoxicity . Such investigations are based on the hypothesis that peripheral nerves are surrounded by a barrier which is similar to the blood-brain barrier (Brown and Workman 1980, White et al . 1980) . However, we have previously found that there is a big difference between the drug concentrations in central and peripheral nervous tissue (Sasai et al. 1990) . It is well known that the important factors determining the uptake of a drug from the blood into the central nervous system are its molecular weight, lipid solubility, and ionization (Bradbury, 1989) . The brain concentration of all the compounds tested in this study was surprisingly low, even when the lipophilicity of the compound was fairly high . Although the reason for this is unclear, in practical terms it means that the octanol/water partition coefficient (i .e. the lipophilicity) of the compounds in both groups tested here was not sufficiently varied to affect their concentrations in the central nervous system . The acute toxicity of the fluorinated compounds was similar to that of non-fluorinated 2nitroimidazoles, probably because of the low concentrations reached in brain tissue . We have previously calculated the apparent biological half-life of radiosensitizers in the mouse sciatic nerve as an indicator of their elimination from the tissue, and we found that it correlated strongly with the octanol/water partition coefficient (Sasai et al. 1990) . Thus the higher the lipophilicity of a drug, the more rapidly it disappeared from the peripheral nerves . We have also found that assessment of the extent of exposure of the rodent peripheral nervous system to a 2-nitroimidazole, rather than that of the brain, is a better indicator for estimating the risk of clinical peripheral neuropathy (Sasai et al . 1990) . Exposure of the peripheral nervous system to the fluorinated 2-nitroimidazole derivatives tested here was very low, and therefore these compounds should have a low peripheral neutrotoxicity . We could not detect any metabolites of KU-2285, etanidazole, or KU-3205 . However, with an increase in the number of CH 2 groups in the side-chain, some metabolites of other compounds (both nonfluorinated and fluorinated) were observed in the serum and tissues . Although further research into these metabolites is necessary, including toxicity studies, their pharmacokinetic properties were very similar to those of the parent compounds (Figures 4b, 6b, 7b) . The pharmacokinetic properties of the fluorinated
et al . compounds seemed superior to those of the nonfluorinated agents, especially in peripheral nervous tissue, and this suggests that the peripheral neurotoxicity of the compounds related to KU-2285 may be lower than that of the compounds related to etanidazole .
5. Conclusions The lipophilicity of the fluorinated compounds was much higher than that of the etanidazole derivatives . As the number of CH 2 groups increased, the lipophilicity of the compounds in each group increased . The brain concentration of all the compounds tested in this study was very low, even when their lipophilicity was fairly high . The exposure of the sciatic nerve to fluorinated 2-nitroimidazole derivatives was very low because the biological halflives of these compounds were short . Although we did not detect any metabolites in the case of KU2285 or etanidazole, an increase in the number of CH2 groups produced some metabolites of compounds of both the non-fluorinated and fluorinated 2-nitroimidazole derivatives in the serum and tissues .
References BRADBURY, N . H ., 1989, Transport across the blood-brain barrier . In : Implications of the Blood-Brain Barrier and its Manipulation. Edited by E . A. Neuwelt (Plenum, New York), Vol . 1, pp . 119-136 . BROWN, J . M . and WORKMAN, P., 1980, Partition coefficient as a guide to the development of radiosensitizers which are less toxic than misonidazole . Radiation Research, 82, 171-190 . FuJITA, T., IWASA, J . and HANNSCH, C ., 1964, A new substitute, n, derived from partition coefficients . Journal of the American Chemical Society, 86, 5175-5180 . ROWLAND, M . and TOZER, T. N ., 1980, Clinical Pharmacokinetics : Concepts and Applications (Lea & Febiger, Philadelphia), pp . 288-290 . SASAI, K ., SHIBAMOTO, Y., TAKAHASHI, M ., ITO, T ., NISHIMOTO, S . and ABE, M . 1990, Pharmacokinetics of 2nitroimidazole hypoxic cell radiosensitizers in rodent peripheral nervous tissue . International Journal of Radiation Biology, 57, 971-980 . SASAI, K ., NISHIMOTO, S ., SHIMOKAWA, Y ., HISANAGA, Y ., KITAKABU, Y., SHIBAMOTO, Y ., ZHOU, L ., WANG, J ., TAKAHASHI, M ., KAGIYA, T . and ABE, M ., 1991a, A fluorinated 2-nitroimidazole, KU-2285, as a new hypoxic cell radiosensitizer. International Journal of Radiation Oncology, Biology and Physics, 20, 1249-1254 .
Pharmacokinetics
SASAI,
K .,
offluorinated
FUSHIKI, M ., YUKAWA, Y., SUYAMA, S ., IWAI, H.,
SHIBAMOTO, Y., NISHIMOTO, S ., TAKAHASHI, M . and ABE, M ., 1991b, In vivo radiosensitizing activity of a new
fluorinated hypoxic cell radiosensitizer, KU-2285, in combination with radiation dose fractionation . Inter-
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
national Journal of Radiation Oncology, Biology, and Physics, 21, 1231-1234 . SHIBAMOTO, Y., NISHIMOTO, S ., SHIMOKAWA, K ., HISANAGA, Y ., ZHOU, L ., WANG, J ., SASAI, K ., TAKAHASHI, M ., ABE, M .
2-nitroimidazole
227
and KAGIYA, T., 1989, Characteristics of fluorinated nitroazoles as hypoxic cell radiosensitizers . International Journal of Radiation Oncology, Biology and Physics,
16,
1045-1048 . WHITE, R. A . S ., WORKMAN, P. and BROWN, J . M., 1980, The pharmacokinetics and tumour and neural tissue penetrating properties of SR-2508 and SR-2555 in the dog-hydrophilic radiosensitizers potentially less toxic than misonidazole . Radiation Research, 84, 542-561 .
1992,
VOL .
62,
NO .
2, 22 1 -227
Pharmacokinetics of fluorinated 2-nitroimidazole hypoxic cell radiosensitizers in murine peripheral nervous tissue K. SASAI*t, H . IWAI$, T. YOSHIZAWA$, S . NISHIMOTO§, Y . SHIBAMOTOt, Y. KITAKABUt, N . OYAt, M . TAKAHASHI¶ and M . ABEL
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(Received 6 September 1991 ; revision received 17 February 1992; accepted 29 February 1992)
Abstract. We have previously reported that KU-2285, a 2nitroimidazole with a fluorinated Nl-substituent (-CH 2CF2 CONH(CH 2 )„OH, n=2), was a promising hypoxic cell radiosensitizer. In this study the pharmacokinetics of KU-2285 and its related compounds (n=3 and n=4) were compared with those of entanidazole (a 2-nitroimidazole with an Nlsubstituent of -CH 2 CONH(CH 2 )„OH, n=2) and its related compounds (n=3 and n=4) to assess the effects of incorporation of a CF 2 group . The lipophilicity of the fluorinated compounds was higher than that of etanidazole, as measured by the octanol/water partition coefficient . As the number of CH 2 groups increased, the lipophilicity of the compounds in both the KU-2285 and etanidazole series increased . The brain tissue levels of the fluorinated compounds were as low as those of the etanidazole derivatives, while the biological half-lives of the fluorinated compounds in peripheral nervous tissues were shorter than those of related non-fluorinated compounds .
1. Introduction We have previously reported the pharmacokinetics of various 2-nitroimidazole hypoxic cell sensitizers in the sciatic nerve tissue of C3H/He mice (Sasai et al. 1990) . The apparent biological half-life of each compound in the peripheral nervous tissue was found to correlate with its lipophilicity, with the more lipophilic compounds having a shorter biological half-life . We have also demonstrated the KU2285, a fluorinated 2-nitroimidazole derivative with an N1 substituent (-CH 2CF2CONH(CH2 )„OH, n=2), was an effective radiosensitizer both in vitro and in vivo (Shibamoto et al. 1989, Sasai et al. 1991 a) . The radiosensitizing activity in vivo of KU-2285 was found to be similar to that of etanidazole when assayed by an in vivo-in vitro assay, a growth delay *Correspondence to : Keisuke Sasai, MD, Department of Radiology, Faculty of Medicine, Kyoto University, Sakyo-Ku, Kyoto 606, Japan . tDepartment of Radiology, Faculty of Medicine ; §Department of Hydrocarbon Chemistry, Faulty of Engineering; ¶Chest Disease Research Institute, Kyoto University, Kyoto 606 ; and I Chemical Division, Daikin Industries, Ltd, Settsu 566, Japan .
assay, and a tumour control assay using the SCC VII tumour or a transplanted mammary tumour in C3H/He mice . Although the radiosensitizing activity of etanidazole was reduced when it was administered orally, there was no significant difference in the radiosensitization produced by intravenous, intraperitoneal, or oral KU-2285 (Sasai et al . 1991a) . Furthermore, KU-2285 was found to sensitize two different murine tumours when given in combination with radiation dose fractionation (Sasai et al. 1991 b) . The acute toxicity of KU-2285 for 8-week-old female C3H/He mice was found to be higher than that of etanidazole, although it was lower than that of misonidazole . The lipophilicity of KU-2285 was similar to that of misonidazole, but our previous pharmacokinetic study demonstrated that the brain concentration of the former compound was surprisingly low (Sasai et al . 1990) . In this study the acute toxicity and the pharmacokinetics of KU-2285 and its related compounds (n=3 and n=4) were compared with those of etanidazole (a 2-nitroimidazole with an NI-substituent of -CH 2 CONH(CH 2 )„OH, n=2) and its related compounds (n=3 and n=4), to assess the effect of incorporating a CF 2 group into the side-chains of these radiosensitizers . 2. Materials and methods 2.1 .
Compounds
The structures of both etanidazole series and KU2285 series are shown in Figure 1 and Table 1 . All of the compounds used in this experiment were synthesized by our group. KU-2285, KU-3202 and KU3207 were fluorinated 2-initroimdazole derivatives with similar structures, while KU-3205 and KU3206 were compounds related to etanidazole . The only difference of the compounds in each group was the number of CH2 groups in the side-chain . For the phamacokinetic studies, all drugs were prepared in
0020-7616/92 $3 .00 © 1992 Taylor & Francis Ltd
222
K . Sasai
lecting blood from the orbital vein . Brains were removed and the sciatic nerves were dissected free of skin, connective tissue, and muscle. Because the sample volume of the sciatic nerves from one mouse was too small to assay (about 10 mg), four to eight nerves from two to four mice were used . Serum and tissue samples were weighed and placed into plastic dishes, after which they were immediately frozen and stored at -20°C prior to analysis . Just before analysis, the tissue samples were homogenized in 10 volumes of distilled water, while serum was vortexmixed after the addition of 10 volumes of distilled water . Forty volumes of methanol was then added to precipitate protein and the sample was mixed and centrifuged .
N
-
,,J NO2
Etanidazole Group
CHKCONH(CH2)nOH
N i
KU-2285 Group
2CF2CONH(CH 2)nOH
CH
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
Figure 1 .
Structural formulae of compounds of etanidazole and KU-2285 series .
physiological saline at a concentration of 10 mg/ml immediately prior to use . The octanol/water partition coefficients were measured in buffer at pH 7.4 according to the method of Fujita et al . (1964) . The octanol/water partition coefficients for KU-2285 and etanidazole differ from those of our previous reports (Shibamoto et al. 1989, Sasai et al . 1990), because of a slight modification of our experimental method .
High-performance liquid chromatography . The supernatants were analysed by reverse-phase highperformance liquid chromatography (HPLC) using a GL Science Inertsil ODS-2 column (4 . 6 x 150 mm, C 18, particle size 5µm) . The flow rate was 0 .5 ml/min . The eluents were as follows: CH 3 CN :H 2 O (10 :60), 0 .01 mol dm-3 NaH 2 PO 4, H 3 PO 4 (to pH 2 .4) for KU-2285 and KU-3202 ; CH 3 CN : H2 O (10 :40), 0 . 01 mol dm-3 NaH 2 PO 4 , H 3 PO 4 (to pH2 .4) for KU-3207 ; CH 3 CN :H 2 O (10 :60), 0.01moldm -3 NaH 2 PO 4 , H 3 PO 4 (to pH2.4) for KU-3205 ; and CH 3 CN :H 2 O (10 :120), 0.01 mol dm-3 NaH 2 PO4, H 3 PO 4 (to pH 2 .4) for KU-3206 . Peaks were detected by a Shimadzu Variable Wavelength UV Monitor SPD-6A at 330 nm ; the peak areas were determined using a Shimadzu Integrator C-R 5A. The concentrations of the com-, pounds were determined using an external standard . The efficiency of recovery of all compounds from fetal bovine serum after protein precipitation was more than 97% . The retention time of each compound was : KU-2285, 7 .3 min ; KU-3202 and its metabolite, 9 .0 and 10.0 min respectively ; KU-3207 and its metabolite, 9 .55 and 10 .4 min respectively ; KU-3205, 5 .2 min ; KU-3206 and its metabolite, 9.97 and 10 .3 min respectively. The lower limit of detection for the assay was about 0 .5µg/g (µg/ml) . The data used here for etanidazole were from our previous experiments (Sasai et al . 1990) .
2 .2. Animals In all experiments, 8-9-week-old female C3H/He mice (Crea Japan Inc ., Osaka; Japan SLC, Hamamatsu) were used . All animal experiments were performed under the guidelines for animal experiments of Kyoto University . 2.3 . Acute toxicity in mice The drug necessary to kill 50% of mice within 7 days (LD 50J7 ) was determined using 35 mice per compound) . Drugs were dissolved in 0 .5% Tween80 solution . All compounds were administered intravenously.
2 .4.
et al .
Pharmacokinetic studies
Sample preparation . All compounds were administered intravenously at a dose of 200 µg/g . At 10, 20, 40, 60, 120 and 240 min after injection, the mice were killed by decapitation immediately after col-
Table 1 . Structures of the compounds used . Value of n Etanidazole group -CH 2 CONH(CH 2 )„OH KU-2285 group -CH2 CF Z CONH(CH 2 )„OH
2
3
4
Etanidazole
KU-3205
KU-3206
KU-2285
KU3202
KU-3207
Pharmacokinetics
offluorinated
2-nitroimidazole
Data analysis . The area under the concentrationtime curve (AUC) for the period of 0-240 min was estimated using the trapezoidal rule (Rowland and Tozer 1980) . The apparent biological half-life in tissue was calculated from the best-fit lines, determined by least-squares linear regression analysis, for the exponential part of the decline phase of the drug concentration-time curve .
toogo 10-1
•
TL
3. Results •
3 . 1 . Octanol/water partition coe cient Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
223
60 120 180 240 Time After Injection (min)
Figure 2 .
Table 2 shows the octanol/water partition coefficient of each compound. The lipophilicity of the fluorinated compounds, as measured by the octanol/ water partition coefficient, was much higher than that of the etanidazole series . As the number of CH 2 groups in the side-chain increased, the lipophilicity of the compounds in each series increased .
3.2 . Distribution of the radiosensitizers Figures 2-7 show the concentrations of the six compounds in serum, brain, and peripheral nervous tissue . Metabolites of KU-3206, KU-3202, and KU3207 were also investigated and their concentrations are demonstrated in Figures 4a, 6b, and 7b. (The
Figure 3 .
Table 2 . Octanol/water partition coefficients and apparent biological half-life in various tissues of six 2-nitroimidazole hypoxic cell radiosensitizers . Apparent biological half-life (min) Octanol/water partition coefficient Etanidazole
0.040
KU-3205
0. 058
KU-3206
0 . 18
KU-2285
0. 25
KU-3202
0. 32
KU-3207
0 . 51
Serum
Peripheral nerve
66 (56-90) b 57 (32-274) 14 (11-18) 27 (20-44) 10 (8 . 2-12) 5 (3 . 2-7 . 7)
256 (150-885) 172 (92-1309) 101 ( 76-151) 167 (104-327) 45 ( 39-57) 49 ( 34-87)
AUC(0-240 min)-' (µg/g x min)
Serum
Peripheral nerve
Brain
LD50% (g/kg)
11400
6200
280
4
7600
4000
70
more than 4
4000 (3300) d 8400
2600 (970) 6400
110 (30) 370
not confirmed'
1500 (3100) 740 (3200)
1100 (930) 870 (2000)
50 (0) 20 (0)
more than 2 .5
2.3
3 .2
'AUC(0-240 min) : The area under the concentration-time curve for the period of 0-240 min of parent compound plus its metabolite if observed . b Figures in parentheses are the 95% confidence interval . `Figures in parentheses are the AUC of the metabolite of each compound .
224
K . Sasai 1000
et al .
-
E 00 z 1007 0 m c Y 0 0 0 U
01
to-
a O 0 J
1
I 60
0
I 120
I 180
240
Time After Injection (min)
(b)
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(a)
Figure 4 .
1
1000-Z
= =1
E
4.90
I
1 0
60
120
180
I 240
Time After Injection (min)
Figure 5 .
concentrations of these metabolites were normalized to the weight of the parent compounds .) The pharmacokinetics of KU-2285 injected intravenously reported here was very similar to the intraperitoneal pharmacokinetics of KU-2285 reported previously (Sasai et al . 1990) . The brain concentration of each compound was very low during the observation period . The drug concentration in the mouse sciatic nerve varied among the compounds . Low lipophilicity compounds such as those of the etanidazole group, and KU-2285, showed a relatively high concentration in peripheral nervous tissue even 240 min after intravenous injection . However, the more lipophilic fluorinated 2-nitroimidazoles (KU-3202 and KU-3207) rapidly disappeared from the peripheral nerves, and were impossible to detect 240 min after injection .
1000-
T a Z U' z " 100c 0
Y
C U
SID
10 -
D
0
J
1 1 0
60
120
180
Time After Injection
(b)
(a)
Figure 6 .
240 (min)
Pharmacokinetics offluorinated 2-nitroimidazole
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
(a)
225
(b)
Figure 7 . Figures 2-7 . Concentrations of 2-nitroimidazoles in serum ( • ), brain (0), and sciatic nerve (0) in C3H/He mice as a function of time after a single intravenous injection . Etanidazole (Figure 2), KU-3205 (Figure 3), and KU-3206 (Figure 4) are nonfluorinated 2-nitroimidazoles, while KU2285 (Figure 5), KU-3202 (Figure 6) and KU-3207 (Figure 7) are fluorinated compounds . Each drug was injected as a single dose of 200,ug/g . Figures 4B, 6B and 7B show the concentrations of metabolites after i .v. injection of the parent compounds . The error bars represent 1 SD for the mean of more than four samples of more than two independent experiments . Figure 2 was drawn from the results of previous examinations (Sasai et al . 1990) .
3 .3. Apparent biological half-life Figure 8 shows the apparent biological half-lives of the six compounds in peripheral nervous tissue as a function of the octanol/water partition coefficient . The apparent half-life of etanidazole was the longest for these six compounds, and both groups of drugs showed a negative correlation between the octanol/ water partition coefficient and the apparent biological half-life in peripheral nervous tissue . Table 2 also shows the apparent biological half-lives of these compounds in serum . A negative correlation
•
3001
0
200-
•
2
0
The drug exposure to the peripheral nerve
The degree of exposure of the sciatic nerves was calculated as the AUC of the drug (plus its metabolite if observed) for the period from 0 to 240 min ; the data are summarized in Table 2 . Exposure of the peripheral nerve to drugs was found to be the greatest for etanidazole and KU-2285, while it was low for the more lipophilic fluorinated compounds such as KU-3202 and 3207 . 3.5. Acute toxicity
3 0 50 80
a
3 .4 .
40
100-
•
(r=0.96) between the octanol/water partition coefficient and the apparent biological half-life in serum was observed . In the brain, the half-lives could not be calculated because the concentrations were too low .
I
I
1 .0 0.01 0.1 Octanol/Water Partition Coefficient
Figure 8 . Plot of the apparent biological half-lvies of 2nitroimidazoles in the sciatic nervous tissue of C3H/He mice versus the partition coefficient of each drug . The correlation coefficient is 0.87 . (0) : Etanidazole group, (i) : KU-2285 group . (1) etanidazole, (2) KU-3205, (3) KU-3206, (4) KU-2285, (5) KU-3202, and (6) KU-3207 .
The acute toxicity of each compound after intravenous injection measured as the LD5017 is shown in Table 2 . The acute toxicity of the drugs of the KU2285 series was similar to or slightly higher than that of the etanidazole series of radiosensitizers . 4. Discussion After the elegant pharmacokinetic studies on 2nitroimidazoles with a wide range of octanol/water
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
226
K . Sasai
partition coefficients performed by Brown and Workman (1980), considerable effort has been devoted to the development of new drugs with a low lipophilicity which should have less peripheral neurotoxicity . Such investigations are based on the hypothesis that peripheral nerves are surrounded by a barrier which is similar to the blood-brain barrier (Brown and Workman 1980, White et al . 1980) . However, we have previously found that there is a big difference between the drug concentrations in central and peripheral nervous tissue (Sasai et al. 1990) . It is well known that the important factors determining the uptake of a drug from the blood into the central nervous system are its molecular weight, lipid solubility, and ionization (Bradbury, 1989) . The brain concentration of all the compounds tested in this study was surprisingly low, even when the lipophilicity of the compound was fairly high . Although the reason for this is unclear, in practical terms it means that the octanol/water partition coefficient (i .e. the lipophilicity) of the compounds in both groups tested here was not sufficiently varied to affect their concentrations in the central nervous system . The acute toxicity of the fluorinated compounds was similar to that of non-fluorinated 2nitroimidazoles, probably because of the low concentrations reached in brain tissue . We have previously calculated the apparent biological half-life of radiosensitizers in the mouse sciatic nerve as an indicator of their elimination from the tissue, and we found that it correlated strongly with the octanol/water partition coefficient (Sasai et al. 1990) . Thus the higher the lipophilicity of a drug, the more rapidly it disappeared from the peripheral nerves . We have also found that assessment of the extent of exposure of the rodent peripheral nervous system to a 2-nitroimidazole, rather than that of the brain, is a better indicator for estimating the risk of clinical peripheral neuropathy (Sasai et al . 1990) . Exposure of the peripheral nervous system to the fluorinated 2-nitroimidazole derivatives tested here was very low, and therefore these compounds should have a low peripheral neutrotoxicity . We could not detect any metabolites of KU-2285, etanidazole, or KU-3205 . However, with an increase in the number of CH 2 groups in the side-chain, some metabolites of other compounds (both nonfluorinated and fluorinated) were observed in the serum and tissues . Although further research into these metabolites is necessary, including toxicity studies, their pharmacokinetic properties were very similar to those of the parent compounds (Figures 4b, 6b, 7b) . The pharmacokinetic properties of the fluorinated
et al . compounds seemed superior to those of the nonfluorinated agents, especially in peripheral nervous tissue, and this suggests that the peripheral neurotoxicity of the compounds related to KU-2285 may be lower than that of the compounds related to etanidazole .
5. Conclusions The lipophilicity of the fluorinated compounds was much higher than that of the etanidazole derivatives . As the number of CH 2 groups increased, the lipophilicity of the compounds in each group increased . The brain concentration of all the compounds tested in this study was very low, even when their lipophilicity was fairly high . The exposure of the sciatic nerve to fluorinated 2-nitroimidazole derivatives was very low because the biological halflives of these compounds were short . Although we did not detect any metabolites in the case of KU2285 or etanidazole, an increase in the number of CH2 groups produced some metabolites of compounds of both the non-fluorinated and fluorinated 2-nitroimidazole derivatives in the serum and tissues .
References BRADBURY, N . H ., 1989, Transport across the blood-brain barrier . In : Implications of the Blood-Brain Barrier and its Manipulation. Edited by E . A. Neuwelt (Plenum, New York), Vol . 1, pp . 119-136 . BROWN, J . M . and WORKMAN, P., 1980, Partition coefficient as a guide to the development of radiosensitizers which are less toxic than misonidazole . Radiation Research, 82, 171-190 . FuJITA, T., IWASA, J . and HANNSCH, C ., 1964, A new substitute, n, derived from partition coefficients . Journal of the American Chemical Society, 86, 5175-5180 . ROWLAND, M . and TOZER, T. N ., 1980, Clinical Pharmacokinetics : Concepts and Applications (Lea & Febiger, Philadelphia), pp . 288-290 . SASAI, K ., SHIBAMOTO, Y., TAKAHASHI, M ., ITO, T ., NISHIMOTO, S . and ABE, M . 1990, Pharmacokinetics of 2nitroimidazole hypoxic cell radiosensitizers in rodent peripheral nervous tissue . International Journal of Radiation Biology, 57, 971-980 . SASAI, K ., NISHIMOTO, S ., SHIMOKAWA, Y ., HISANAGA, Y ., KITAKABU, Y., SHIBAMOTO, Y ., ZHOU, L ., WANG, J ., TAKAHASHI, M ., KAGIYA, T . and ABE, M ., 1991a, A fluorinated 2-nitroimidazole, KU-2285, as a new hypoxic cell radiosensitizer. International Journal of Radiation Oncology, Biology and Physics, 20, 1249-1254 .
Pharmacokinetics
SASAI,
K .,
offluorinated
FUSHIKI, M ., YUKAWA, Y., SUYAMA, S ., IWAI, H.,
SHIBAMOTO, Y., NISHIMOTO, S ., TAKAHASHI, M . and ABE, M ., 1991b, In vivo radiosensitizing activity of a new
fluorinated hypoxic cell radiosensitizer, KU-2285, in combination with radiation dose fractionation . Inter-
Int J Radiat Biol Downloaded from informahealthcare.com by UB Kiel on 11/08/14 For personal use only.
national Journal of Radiation Oncology, Biology, and Physics, 21, 1231-1234 . SHIBAMOTO, Y., NISHIMOTO, S ., SHIMOKAWA, K ., HISANAGA, Y ., ZHOU, L ., WANG, J ., SASAI, K ., TAKAHASHI, M ., ABE, M .
2-nitroimidazole
227
and KAGIYA, T., 1989, Characteristics of fluorinated nitroazoles as hypoxic cell radiosensitizers . International Journal of Radiation Oncology, Biology and Physics,
16,
1045-1048 . WHITE, R. A . S ., WORKMAN, P. and BROWN, J . M., 1980, The pharmacokinetics and tumour and neural tissue penetrating properties of SR-2508 and SR-2555 in the dog-hydrophilic radiosensitizers potentially less toxic than misonidazole . Radiation Research, 84, 542-561 .
