Potent and non-specific inhibition of cytochrome P450 by JM216, a new oral platinum agent.

Bis-acetato-ammine-dichloro-cyclohexylamine-platinum (IV), JM216, is the first antineoplastic platinum compound that can be given to patients orally. Several phase II clinical trials of JM216 monotherapy have already been reported. However, no information on the potential drug interactions caused by JM216 is available. In this study, the capacity of JM216 to inhibit cytochrome P450 (CYP) in human liver microsomes was investigated by measuring the inhibition potential (IC50 and Ki) on prototype reactions. Specific substrates of CYP included testosterone (catalysed by CYP3A4), paclitaxel (CYP2C8), 7-ethoxyresorufin (CYP1A1, CYP1A2), coumarin (CYP2A6), aniline (CYP2E1) and (+/-)-bufuralol (CYP2D6). JM216 inhibited the catalytic activities of CYP isozymes. The IC50 values were between 0.3 microM and 10 microM, indicating strong and non-specific inhibitory effects of JM216. The inhibition occurred in a non-competitive manner, and the Ki value was 1.0 and 0.9 microM for metabolite formation of testosterone and paclitaxel respectively. Therefore, some in vivo studies should be conducted to determine whether or not there is a correlation between in vivo and in vitro results.

Platinum anti-tumour agents. such as cisplatin and carboplatin. have been widely used in combination chemotherapy for many cancers. especially for ovarian and lung cancers (Fukuoka et al. 1991: McGuire et al. 1996. These agents available today are. however, generally administered intravenously. The development of an oral platinum drug has been desired to improve the quality of life of patients receiving cancer chemotherapy in terms of easy administration. Bis-acetato-ammine-dichloro-cvclohexylamineplatinum (IV). JM2 16. is the first oral antineoplastic platinum agent currently under development. In preclinical studies. JM' 16 exhibited in vitro and in vivo anti-tumour efficacy comparable with cisplatin and carboplatin. and an activity against cell lines that were resistant to cisplatin (Kelland et al. 1993). Several phase H clinical trials of JM216 monotherapy have already been performed in the United States and Europe. and a phase I study has finished in Japan (Groen et al. 1996: Fujii et al. 1997: Peereboom et al. 1997). Further, some combination regimens of JM216 with other anti-tumour agents. such as taxans. can be expected. but no information on the potential dru2 interactions beteen JM2 16 and other drugs has been reported.
The metabolic pathway of the drug is complicated and has not been well understood (Raynaud et al. 1996). However, there are some implications that metabolism of JM216 might affect drugmetabolizing enzymes in the liver. First. at least six metabolites were detected in plasma samples from patients who received JM216 (Raynaud et al. 1996). Four of them were also obtained by in mvitro incubations with fresh human plasma. whereas the remaining two metabolites were detected orly in in vivo studies.
Second. according to the results of organ distribution in mice platinum accumulated to a high level in the liver and the level was retained steadily for several days (McKeage et al. 1994).
Besides the metabolic pathway of JM216 itself. it is important to evaluate the effects of the drug on the metabolism of other drugs. This study was undertaken to investigate whether JM216 would interact with drugs being metabolized by cytochrome P450 (CYP).  Concentrations of platinum compounds were 0.3, 1, 3 and 10 gm for JM216 (open symbols), and 10 gm for cisplatin (closed symbols). The rate of the metabolite formation without JM216 was 870 (HL12) and 240 (HL48) for 6"hydroxylation of testosterone, 23 (HL15) and 2.8 (HL51) for O-deethylation of 7ethoxyresorufin, 260 (HL12) and 80 (HL48) for 7-hydroxylatin of coumann, 380 (HL12) and 410 (HL48) for p-hydroxylation of aniline and 63 (HL12) and 120 (HL48) for 1'-hydroxylation of (±)-bufuralo (pmol min-' mg-protein). Liver microsomes from human subject HL1 2 ( A), HL1 5 (T. V), HL47 (> *), HL48 (7 U) and HL51 ( .0) were used. Each plot represents the mean of duplicate determinations Human liver microsomes Human liver microsomes were prepared from autopsy samples with informed consent in writing from each guardian. The use of human liver for the study had been approved by the Institutional Committee of Hokkaido University. Liver tissues were stored at -800C. Microsomes were prepared as described previously (Kamataki and Kitagawa. 1973). and were stored at -80°C until use. Protein concentration was measured according to the method of Lowry et al ( 195 1 ).

Analytical procedures
Inhibition by JM216 of CYP in human liver microsomes was examined by measungn their inhibition potential (IC, and K) on prototype reactions. Specific substrates and the reactions measured in this study included testosterone 6f-hydroxylation (catalysed by CYP3A4) (Waxman et al. 1988). paclitaxel 6ahydroxylation (CYP2C8) (Cresteil et al. 1994: Rahman et al. 1994. 7-ethoxyresorufin O-deethylation (CYPlAl. CYPIA2) (Guengerich et al. 1982). coumarin 7-hydroxylation (CYP2A6) (Pearce et al. 1992). aniline p-hydroxylation (CYP2EI) (Ryan et al. 1985) and (±)-bufuralol l'-hydroxylation (CYP2D6) (Nakamura et al. 1996). Substrates were incubated alone or together with JM216 (0.3-10 g-t) to estimate the concentration of JM2 16 yielding 50% inhibition of the metabolism (ICio) Values of IC, were evaluated directly from the plots. Detailed kinetic studies were performed to determine the apparent inhibition constant (K!). and to clarify the mechanism(s) involved in the inhibition using testosterone and paclitaxel as substrates. All reactions were initiated by addition of each substrate after a 5-min preincubation at 37°C in a shaking water bath. JM216 or cisplatin were preincubated in the reaction mixture before substrate addition. In studies with testosterone. pacitaxel and (±)bufuralol as substrates. the determinations of metabolites were performed by the high-performance liquid chromatoraphy (HPLC) system. The system included a Hitachi model D-7000 (Hitachi. Tokyo. Japan) equipped with an L-7100 pump. a L-7200 autosampler and a L-7400 detector. and a Capcell Pak C18 (5 gm) 4.6 x 250 mm column (Shiseido. Tokyo. Japan). Determinations were performed in duplicate and the representative results were shown.
The assay of testosterone 6$-hydroxylation was performed as described by Arlotto et al (1991). A reaction mrixture consisted of 100 mtir potassium-phosphate buffer (pH 7.4). 50 -iM EDTA. an  I  I  I  I  I  I  I  I  I  I  I  I  I  I   I   I  I  I  I  I  I  I  I  I  I   I  I  I   I  I -1 0 1 JM216 (mu) Figure 2 Representative Lineweaver-Burk plots of testosterone 6lhydroxylation by liver microsomes from a human subject HL12 (A) and the secondary plots showing the K value of 1.0 pu (B). The concentrations of JM216 were 0 gm (as a control, 0), 1 iu (A) and 3 gm (U). The concentratons of testosterone were 18.8, 37.5, 75 and 150 gm. Each plot represents the mean of duplicate determinations NADPH generating system (0.5 m-nt NADP+. 5 inst magnesium chloride. 5 insm glucose 6-phosphate and 1 U mll lucose-6phosphate dehydrogenase). a desired concentration of JM216 or cisplatin. and 0.2to 0.4-mg microsomes in a final volume of 1 ml. The final testosterone concentration was 18.8-160 gm. After a 15nun incubation. the reaction was terminated by addition of 5 ml of diethylether followed by addition of 1 nmol of 11 ptestosterone as an internal standard. The sample was mixed vigorously. and the organic phase was separated by centrifuging. After the extract was Figure 3 Representative Lineweaver-Burk plots of paditaxel metabolism by liver microsomes from a human subject HL12 (A) and the secondary plots showing the K value of 0.9 gu (B). The concentrations of JM216 were 0 gm (as a control, 0), 1 gu (A) and 3 iu (U). The concentrations of pacitaxel were 2.5, 5, 10, 20 m. Each plot represents the mean of duplicate determinations evaporated to dryness bv centrifugal evaporator Hitachi CE1D (Hitachi Koki. Tokyo. Japan). the residue was dissolved in 200 l of a solvent used as an initial HPLC mobile phase and the solution applied to HPLC. The mobile phase was a mixture of methanol. water and acetonitrile at 39:60:1 (v/v. solvent A) and at 80:18:2 (vx/v. solvent B). The separation was accomplished at 40°C using a 30-min linear gradient from 98% (v/v) solvent A (0 min) to 20% (v/v) solvent A (30 min) at a flow rate of 1 ml mmn-'. Absorbance was monitored at 254 nm. The formation of 6-testosterone was British Joumal of Cancer (1998) 78 (9) acetonitrile-water before HPLC analysis. Under the conditions described above, baccatin ml 6a-hydroxypaclitaxel and pacitaxel were eluted with retention times of 18.2 min, 24.5 min and 26.7 min respectively, which were similar to those reported by Harris et al (1994). As an authentic reference standard of the metabolite of paclitaxel was not available, we assumed that the metabolite eluted with a retention time of 24.5 min as 6a-hydroxypaclitaxel and expressed the velocity of biotransformation as the peak height ratio of the metabolite to the intemal standard. 7-Ethoxyresorufin O-deethylation and coumarin 7-hydroxylation were measured by determination of metabolites using a Hitachi F-2000 fluorescence spectrophotometer (Hitachi, Tokyo. Japan: Lake, 1987;Pearce et al, 1992). Aniline p-hydroxylation was assayed colorimetrically with a Hitachi U-1000 spectrophotometer (Hitachi; Imai et al. 1966). The l'-hydroxylated metabolite of (±)-bufuralol was determined by HPLC as reported previously (Nakamura et al, 1996). Incubation times were 10 min for 7-ethoxyresorufin (with a final concentration of 2 PIM), 15 min for coumarin (50 gM), 15 min for aniline (4 mM) and 30 min for (±)-bufuralol (20 giM) oxidations.

RESULTS
Effects of JM216 on the 6-hydroxylation of testosterone Clear inhibition by JM216 of testosterone 6$-hydroxylation was seen. At the 160 gM concentration of testosterone, an ICvalue was estimated to be between 0.3 ILM and 1 JM, suggesting a strong inhibitory effect of JM216 on CYP3A (Figure 1). Lineweaver-Burk plots showed that the inhibition occurred in a non-competitive manner. and the K value derived from the secondary plots was evaluated to be 1.0 JLM (Figure 2). The hydroxylase also seemed to be inhibited by cisplatin, but the inhibition was rather weak. The inhibition was only 15% at 10 Jim concentration of cisplatin (Figure 1).

Effects of JM216 on the metabolism of paclitaxel
The hydroxylation of pactitaxel was inhibited with an IC-, value between 1 JiM and 3 gM at a paclitaxel concentration of 10 Jim ( Figure 1). Fonnation of the metabolite, possibly 6a-hydroxypaclitaxel, followed Michaelis-Menten kinetics as demonstrated by linear Lineweaver-Burk plots (Figure 3). Apparent Km value was 17 gm, which was consistent with that measured as the formation of 6a-hydroxypaclitaxel in previous reports (Cresteil et al. 1994;Harris et al, 1994). The inhibition also occurred in a noncompetitive manner with the K, value of 0.9 jim (Figure 3).

DISCUSSION
Several drugs, such as SKF-525A (Buening and Franklin. 1974). metyrapone (Testa andJenner. 1981), cimetidine (Winzor et al. 1986) and ketoconazole (Pasanen et al, 1988), have been known to inhibit CYP non-specifically. JM2 16 would be another example of a non-specific inhibitor of CYP with high inhibition potential. Thus, more detailed mechanism(s) responsible for the inhibition should be investigated.
Further, it remains to be examined whether phamicokinetics of drugs being metabolized mainly by CYP would be altered by JM216. As in vitro results do not always translate to the in vivo situation, and as very little or no JM216 is found in the systemic circulation after oral administration in human (Raynaud et al. 1996), we cannot be sure exactly how much, if any, of the compound actually reaches the liver through the portal vein. The in vitro inhibition of CYP by JM216 found in this study, however, agrees with the results of combination chemotherapy involving etoposide in vivo (Rose. 1997). When etoposide was given orally to mice in combination with JM216, the maximum tolerated dose was reduced to 25% of that seen with etoposide alone. Although no pharmaokinetic data were reported. it might be possible that JM216 inhibited the metabolism of etoposide. a substrate of CYP3A (Relling et al, 1994).
This report suggests that careful attention should be paid to interactions of drugs metabolized mainly by CYP, including many antineoplastic agents, when treating cancer patients with JM216. Additionally, if the in vitro/in vivo correlations are demonstrated. we can propose an advantageous use of JM216 as a potential suppresser of drug metabolism in combination cancer chemotherapy. In other words, JM216 can be used to reduce the necessary dose for treatment of combined agents that are detoxified by CYPs, i.e. paclitaxel (CYP2C8. CYP3A4: Cresteil et al. 1994;Harris et al, 1994;Rahman et al, 1994). docetaxel (CYP3A4; Marre et al, 1996), etoposide (CYP3A4: Relling et al, 1994) and vinca alkaloids (CYP3A4; Zhou et al, 1993). With this kind of intervention, the inhibition of cyclosporin or etoposide metabolism by ketoconazole has already been used intentionally to reduce the cost of cyclosporin treatment and to improve the bioavailability of oral etoposide (First et al, 1989;Kobayashi et al. 1996). Schwartz et al (1995) have successfully used fluconazole to reverse the accelerated trans-retinoic acid clearance in patients with acute promyelocytic leukaemia On the other hand. as cyclophosphamide and ifosfamide are activated by CYP2B and CYP3A respectively (Chang et al. 1993). combination use of JM2 16 may decrease the anti-tumour effects of these prodrugs.
This in vitro study revealed that JM2 16 inhibited multiple forms of CYP. Therefore, some in vivo studies should be conducted to determine whether or not there is a correlation between in vivo and in vitro results.