Effect of Ifosfamide Treatment on Glutathione and Glutathione Conjugation Activity in Patients with Advanced Cancers’

Vol. 1, /525-1536, Dee,nher 1995 Clinical Cancer Research 1525

Titia M T. Mulders,2 H. Jan Keizer, ever, whether a 35 % decrease is sufficient to increase tumor Jan Ouwerkerk, E. A. van der Velde, sensitivity toward (other) cytostatics remains uncertain. Douwe D Breimer, and Gerard J. Mulder3 INTRODUCTION Divisions of Toxicology and Pharmacology, LeidenlAmsterdam Center for Drug Research IT. M. T. M.. D. D. B.. G. J. M.]. and Ifosfamide is an alkylating oxazaphosphorine nitrogen Division of Medical Statistics IE. A. v. d. V.1, Leiden University, and mustard with activity against a wide range of tumors both in Division of Clinical Oncology, University Hospital H J. K.. J. 0.1, adults and children. It is a prodrug that requires biotransforma- Leiden, the Netherlands tion by cytochrome P450 isoenzyrnes to become cytotoxic. This metabolic process results in the formation of the active bifunc- ABSTRACT tional alkylating metabolite, isophosphoramide mustard, and Several studies have suggested that the glutathione/ acrobein, a metabolite believed to contribute to the hemorrhagic glutathione S-transferase (GSH/GST) system is involved in cystitis associated with oxazaphosphorine treatment. Ifosfamide resistance of tumors toward ifosfamide and other cytostatic can also undergo side chain dealkylation, resulting in the for- agents. Besides, ifosfamide metabolites (in vitro) as well as mation of 2- and 3-dechboroethylifosfarnide and equimolar ifosfamide treatment (in vivo) have been shown to decrease amounts of the highly reactive chloroacetaldehyde ( 1-4). High cellular GSH availability. In the present study, the in vivo plasma levels of chboroacetaldehyde have been suggested to be effects of three different ifosfamide treatment schedules on associated with ifosfamide-related neurotoxicity (5. 6). the GSH/GST system were studied in patients with ad- Resistance to cytostatic therapy is a common clinical prob- vanced cancers (n 24): continuous i.v. infusions of 1304) lem in the treatment of cancer. A wide variety of mechanisms mg/m2 daily for 10 days and 5000 mg/m2/day for 24 h, as has been implicated in the aetiobogy of drug resistance, includ- well as a 4-h infusion of 3000 mg/m2 daily for 3 days. The ing detoxification by the GSH/GST4 system. The relationship GSHJGST system was characterized by administering bro- between drug resistance and this system has been studied ex- misoval, a probe drug to assess GSH conjugation activity in tensively in human cancer cell lines and tumor tissues (e.g., vivo, as well as by daily monitoring of GSH concentrations in 7-13). However. evidence for involvement of the GSH/GST blood cells and plasma. Bromisoval pharmacokinetics was system in drug resistance in patients in vito is limited. Such assessed before and at the end of the ifosfamide treatment. studies are generally restricted to the assessment of GSH content Blood cell GSH levels decreased significantly (P < 0.05) (tumors, blood cells) on GST profile/activity of tumors of pa- during the 3- and 10-day ifosfamide treatment schedules; tients (14). the 24-h treatment had no effect. The ifosfamide treatment So far, several in vitro studies have provided evidence for schedules had only minimal effects on bromisoval pharma- the involvement of GSH and GSTs in the detoxification of cokinetics. Assuming that the kinetics of the probe drug metabolites derived from ifosfarnide and its structural analogue provide an accurate reflection of enzyme activity, this sug- cycbophospharnide. The ifosfarnide-derived 4-hydroxy interme- gests that GST activity remains unchanged. Because GSH diate and especially chloroacetaldehyde depleted intracellular conjugation of bromisoval enantiomers requires both GST GSH and inhibited the GST-catalyzed conjugation of b-chloro- activity and GSH availability, these results also indicate 2,4-dinitrobenzene (15). Depletion of cellular GSH levels was that, despite the 35% decrease in GSH in blood cells of two also demonstrated for the 4-hydroxy metabolite of cycbophos- patient groups, the GSH availability of the cancer patients phamide (16). Furthermore, GSH conjugation of 4-hydroxycy- was not rate-limiting for GSH conjugation of bromisoval cbophospharnide and phosphoramide mustard was shown to be enantiomers. If GSH levels in blood cells reflect those in catalyzed by cytosolic GSTs, in particular GSTA1-1 ( 17. 18). tumors/other tissues, the present results indicate that ifosf- Acrolein, a metabolite derived from both ifosfamide and cycbo- amide may be used clinically to decrease GSH levels. How- phosphamide. could be inactivated by human class a. p.. and in

particular class ii GST isoenzymes. either by conjugation with GSH or by covalent binding to the enzymes in the absence of GSH ( 19). In vivo. Lind et a!. ( 1 5) demonstrated that an infusion Received 7/10/95; accepted 8/2/95.

I Supported in part by ASTA Medica Aktiengesellschaft (Frankfurt am Main, Germany) and the Dutch ‘Praeventiefonds” (Project 28-1962, Den Haag. the Netherlands).

2 Present address: NV Organon, Medical Research and Development

Unit, Molenstnaat 1 10, P. 0. Box 20, 5340 BH Oss, the Netherlands. 4 The abbreviations used are: GSH, glutathione: GSSG, glutathione 3 To whom requests for reprints should be addressed. at Division of disulfide (oxidized glutathione): GST, glutathione S-transferase: GSTM, Toxicology, Leiden/Amstendam Center for Drug Research. Leiden Uni- class p. glutathione S-transfenases: Hb, hemoglobin: HPLC, high pen- versity. Wassenaarseweg 72, P. 0. Box 9503, 2300 RA Leiden. the fonmance liquid chromatography: Cl11. clearance of formation of men- Netherlands. capturate: AUC, area under the curve: Cl/F, clearance/bioavailability.

of ifosfamide in a patient resulted in a 70% decrease of GSH in ment period, mesna infusion was continued at a lower dose for lymphocytes. 1 day. For the prolonged ifosfamide (and mesna) infusions, Based on the above-mentioned in vitro and in vivo data, it external infusion pumps (CADD-I pump) were used, which was hypothesized that changes in the GSH/GST system may were connected to a Port-A-Cath (Pharmacia Deltec Inc., Woer- affect tumor resistance toward ifosfamide. Furthermore, GSH den, the Netherlands) iv. device. The ifosfarnide and mesna conjugation of ifosfamide (metabolites) may also lead to con- solutions were supplied by the Leiden University Hospital Phar- sumption of GSH, thereby making ifosfamide a potentially macy. GSH-lowering cytostatic agent, which may be used in combi- nation with (an)other cytostatic agent(s)-for which involve- Characterization of GSH Conjugation ment of the GSHJGST system in resistance is a major contrib- Characterization of the GSHJGST system was performed uting factor-with the aim to improve the efficacy of the by administering the racemic drug bromisoval as well as by cytostatic treatment. daily assessment of GSH (plasma and blood cells) and 0550 In the present study, the in vivo effects of ifosfamide (blood cells) concentrations. In addition, the GSTM 1 phenotype treatment on the GSHJGST system were studied in patients with of all patients was determined. advanced cancers. Three different ifosfamide treatment sched- Bromisoval. Brornisoval pharmacokinetics (plasma ules were studied, i.e., a continuous iv. infusion of 1 300 mg/rn2 pharmacokinetics of the bromisoval enantiomers and urinary daily for 10 days or 5000 mg/m2/day for 24 h and a 4-h infusion excretion of their corresponding mercapturates) was assessed on of 3000 mg/rn2 daily for 3 days. The GSH/GST system of the day prior to the start of the ifosfamide infusion (day 0), the patients in vivo was characterized by administering 2-bro- last day of the 3 (3000 mg/rn2 daily)-. and 1 0 ( 1 300 mg/rn2 moisovalerylurea (bromisoval), a probe drug to assess GSH daily)-day ifosfamide treatment period, and the day after the conjugation activity in vivo (20, 21), as well as by daily moni- 24-h (5000 mg/m2/day) ifosfamide infusion. On both bromi- toring of GSH (plasma and blood cells) and GSSG (blood cells) soval study days, patients received a capsule containing 600 rng concentrations. (2.69 mmol) racemic brornisoval after an overnight fast. The brornisoval (Dutch Pharmacopoeia) capsules were supplied by PATIENTS AND METHODS the Leiden University Hospital Pharmacy. Patients refrained The study protocol was approved by the Ethics Review from all food and drinks (except plain mineral water ad libitum) Board of the Leiden University Hospital (Leiden, the Nether- until 3 h after brornisoval administration. To ensure sufficient lands). The study was an open trial. Twenty-four patients par- urine production, the patients were asked to drink regularly. ticipated after each gave oral informed consent before study Patient 102 (24-h ifosfamide treatment) did not receive bromi- entry. soval because this patient refused to participate in the brornisoval study on the ifosfamide pretreatment day. Patient 304 Patients (3-day ifosfamide treatment) did not receive bromisoval because Patients of either sex, 75 years old or less, with histologi- this patient had a jejunal fistula. cally proven advanced cancer and for whom no standard therapy Blood and Urine Sampling. On the ifosfamide pretreat- was available were eligible. Furthermore, patients had to have ment day (day 0), before the start of the first brornisoval study normal renal function, no major disturbance of liver function period, an iv. cannula (Venflon, BOC Ohmeda AB, Sweden) (bilirubin 30 p.rnollliter), adequate bone marrow function was inserted into a suitable forearm vein. The cannula was kept

(WBCs 4.0 10/l, platelets 100 l0/l), and a WHO per- patent by intermittent flushing with saline solution containing formance 2. Patients with postsurgical complications, an heparin. All blood samples were collected in 10-mb lithium active infection, or central nervous system, cardiac, or pulmo- heparin tubes. nary disease, as well as those with a second malignant disease, On both brornisoval study days, brornisoval blood samples were ineligible. (10 mI/sample) were collected before and at 0.5, 1, 2, 3, 4, 5, 6, All patients, except patient 301, had received prior treat- 8, 12, and 24 h after bromisoval administration. Plasma was ment consisting of radiotherapy, surgery, and/or treatment with obtained after centnifugation (6 mm at 3000 X g) immediately other cytostatic agents or hormones. None of the patients, except after blood collection. The bromisoval plasma samples were patient 102, had received ifosfamide therapy on a previous stored at -20#{176}Cuntil analysis. Urine was collected quantita- occasion (Table 1). tively in fractions up to 24 h after brornisoval intake. The end of the collection periods was 3, 4, 5, 6, 7, 8, 9, 10.5, 12, 15, and 24 Ifosfamide Treatment Schedules h after administration of racernic bromisoval. Of each urine Three different ifosfamide infusion schedules were studied, fraction, the volume was measured and an aliquot ( 10 ml) was i.e., continuous infusions of 1300 mg/rn2 daily for 10 days (12 stored at -20#{176}Cuntil analysis. patients with in general gastrointestinal primary tumors) and Blood samples (10 rnllsample) for the assessment of GSH 5000 mg/m2/day for 24 h (6 patients with various primary (plasma and blood cells) and 0550 (blood cells) levels were tumors) as well as a 4-h infusion of 3000 mg/rn2 daily for 3 days collected daily after an overnight fast at approximately 9.00 am. (6 patients with in general head/neck primary tumors; Table I). from the day before until the last day of (day after for the 24-h Ifosfamide infusions were administered along with the uropro- ifosfarnide infusion) the ifosfarnide treatment period. However, tector mesna. The infused solution always consisted of a mixture in patients 1-6 (Table 1A; 10-day ifosfamide treatment), blood of mesna and ifosfamide. After the end of the ifosfamide treat- for the OSH assay was collected only on the pretreatment day

I- I I ++ I + I + + I + I+1+1+ ++ I ++

C Vi V C. -1 a fl: h 1 . . C C C l-c -. Cl c C. r- c C C ‘ C C\ r- C .C ri C n-- C +1 C C 00 C ri +1 1 r N (- C’, Cl (N (N - (- r C C r. C .C Cl

‘C V V Vi a ‘C C. ‘-i C . C C r- C C C C C Cl ‘1- C’ C C c-I o Cl 3C r- C .,: ,-; r t- C Cl C’1 OC ri t ‘r - C; i - C’. OC +1 ,: +1 a

r-i V (“l V

‘C

--- C C - Cl -- Cl Cl - Cl Cl Cl -- C C. V C E , ‘C E C V ‘C C E . C : C ‘6 . C .- Th-a C a V C C C a d ‘C _) . C a ‘C V V - C :E.E Vi V a C V ‘C a E C C. C V C - a U L) . & t) V a ‘C #{149}E a ‘C V . C V C V V V Vi C a V V > C C a . a ZC E E C C CCC , C ? a a V VVE V VVC V V B a E . E- E” ECE - EE CC..CC.C Vi Vi CCCVi ‘C ‘C ‘C V V V ‘C E e ‘C VC C ‘C.’CC CC’ .C .C . . . . > C . . : C .C.CC . ,#{149},4.C.C .C - - &)&<< C. a Vi C V C V E C L) V E V C a = ‘C

a ‘“C D i . C’5’’ ‘ F-. . E 2EE E”- C. ‘CCE =JQ--- ‘;-J-- ‘C->-aC> .aaaVa aVVCCC2C’CCVCC aCaCVa’C rCVVVL..V ‘CL2&UC.F- ‘C -. C C E E C a V _i C\ , V ‘C ‘C Vi E E-r-Cl-occ--.r--+l- ErcrCl-r+jr V V E ‘C a C ‘C E C - Vi C C C C C ‘C ‘C C - C. (N t- ‘1- N C’ r’ Qt’tCl-OCC C .i Vi OCCNir-IriiCNCr-.C+IN . N- C f C r- C +1 C V . r’, ‘ ri r r +1 ri - Cl - 00 C C- C fl . + C V C V .9 .J : 00 : #{149}.C E00ooC-r.c-oo-C- +11-i E’rr’ot-N fl - ‘l’ r- -CC +1 C ‘ fl I- C fl ‘ C C i ‘1- t C t C v’ t a +1 .- C. V V (N >5 V t. a. a. a. -‘ -. 00 C C-. +IC . . c_) i ‘C i a a C .C a V C C. C a L) r - r’ ri C a cccccc V ccc VV < - Cl r f C C- 00 C - - - C. U r’r’r’r’rr

(day 0), day 7, and the last day (day 10) of the infusion period. eluent consisted of buffer (0. 1 msi EDTA disodium salt, 0. 1 M The decision to extend the OSH sampling scheme was made sodium nitrate, 0.01 si potassium bromide, and 2.5 msi sodium after the results of the first six patients were available; daily octane-l-sulfonic acid) mixed with I 1.8 M acetic acid and meth- sampling at a relatively fixed time point would give more anol (1 75: 1:4). The pH of the eluent was adjusted to 2.4. specific information on the time course of the effects of the three After centrifugation of the pretreated plasma GSH samples, different ifosfamide treatment schedules on GSH levels. To get the supernatant was directly injected on the HPLC. For the an impression on the effects of the first course of ifosfamide assessment of blood GSH concentrations, pretreated blood sam- treatment on a subsequent course, OSH blood cell levels of pies were 25-fold diluted with mobile phase and centrifugated: patient 306 (3-day ifosfamide treatment) were studied during the supernatant was injected on the HPLC. The additional dilu- both the first and second course of the ifosfamide therapy. (The tion with mobile phase was not necessary for the 0550 analy- second course started approximately 5 weeks after the end of the sis. All blood (GSH and 0550) and plasma (GSH) analyses first course.) In addition, in both patients 306 (first and second were performed in duplicate. Quantification was performed course of ifosfarnide treatment) and 308, blood for the assess- using external standards [OSH and 0550 solutions in water ment of GSH blood cell levels was collected more frequently, diluted with perchloric acid for plasma, mobile phase for GSH i.e., prior to and at 2, 4, 6, and 8 h after the start of each blood, and EDTA/perchloric acid ( I : 1 ) for 0550 blood analy- ifosfamide infusion. The batter scheme would, in addition to the sis] which were freshly prepared prior to the analysis of a daily GSH sampling, also give an impression on the effects of sample (in duplicate). GSH and GSSG levels are expressed as ifosfamide treatment within the separate treatment days. After jimol/g Hb. Because RBCs are the most predominant blood collection, the blood samples were immediately pretreated and cells, the GSH and 0550 blood cell levels presented in this

stored at - 80#{176}Cuntil analysis (see below). A separate blood article will mainly reflect those in RBCs. The expression of sample was always collected along with the daily GSH sample GSH/GSSG levels in i.mol/g Hb was considered to be necessary for the assessment of Hb and hernatocrit values, to enable the because: (a) the largest amount of GSH in blood is present in calculation of OSH and 0550 blood cell concentrations (see blood cells (mM versus i.M range for cells and plasma, respec- below). Blood (approximately 500 pA) for the assessment of the tively) and (b) all three ifosfarnide treatment schedules caused a GSTM1 phenotype was collected only on the ifosfamide pre- statistically significant decrease in Hb and hematocnit values treatment day (day 0) (collected along with the first OSH (see below). sample). The GSTM1 blood sample was stored at -20#{176}Cuntil GSTM1 Polymorphism. The GSTM I phenotype was analysis. determined with an ELISA (MUKIT: Medlabs, Dublin, Ireland) which required 200 p.1 whole heparinized blood. Analytical Methods (R)- and (S)-Bromisoval Enantiomers (Plasma) and Bromisoval Pharmacokinetic Analysis Their Corresponding Mercapturates (Urine). (R)- and (S)- The pharmacokinetic analysis was performed with Siphan bromisoval enantiomers in plasma were separated by normal (Release 3.3; Simed, Cr#{233}teil,France) software (the interactive phase HPLC on a chiral column (Chiralcel OD column; Daicel program for exponentional models and urinary data analysis). Industries, Tokyo, Japan) and detected spectrophotometrically Pharmacokinetic parameters were calculated separately accord- at 210 nm (22). ing to standard procedures (24) for both bromisoval enantiomers The diastereorneric bromisoval mercapturates were as- in plasma and both bromisovab mercapturates in urine. All sayed by direct injection (after dilution) on a reversed phase calculations were performed under the assumption of complete HPLC using a bromine generation cell and electrochemical absorption of the parent compound. The Cl of the mercaptur- detection (23). Because human urine, compared with rat urine, ates of (R)- and (S)-brornisoval was calculated as the ratio of the contains compounds which interfere with the chromatography cumulative amount of mencapturate excreted in urine and the of the mercapturates, the eluent was slightly changed. It con- AUC of the corresponding bromisoval enantiomer in plasma. sisted of buffer (0. 1 M sodium nitrate, 0.01 M potassium bro- The (S):(R) ratios in plasma were calculated by dividing the mide, 0.01 M citric acid, and 1 1 pM sodium decane-l-sulfonic AUC values of both bromisoval enantiomers (either the AUC acid) mixed with methanol and acetonitribe (180:9:9). extrapolated to infinity or the AUC values up to 3 h after GSH (Plasma and Blood Cells) and GSSG (Blood administration). The (R):(S) ratios in urine were calculated by Cells). Immediately after collection of the blood sample, 100 dividing the cumulative amount of the (R)- and (S)-brornisoval p.1 blood were added to 1 .9 ml 1 mM EDTA, after which 2 ml 0.8 mercapturates in urine (cumulative amount excreted in 24 h). M perchloric acid were added to the bboodIEDTA mixture. After

centrifugation (6 mm, 3000 X g) of the remaining blood sample, Statistical Analysis plasma was also diluted with 0.8 M perchboric acid ( 1 : 1). All The statistical analysis was performed with SPSSIPCT pretreated samples were frozen in liquid nitrogen as soon as (Release 5.0; SPSS Inc., Chicago, IL) statistical software. Pa- possible (within 30 mm after sample collection) and stored at rarnetens are reported as mean ± SD. For each ifosfamide -80#{176}C until analysis (within 1 month after collection of the treatment schedule, the bromisoval pharmacokinetics of the sample). ifosfamide pretreatment day were compared with those of the The OSH (plasma and blood) and GSSG (blood) concen- last ifosfamide treatment day (day after for the 24-h ifosfamide trations were assayed on a reversed phase HPLC using a bro- treatment schedule) by paired t tests. The significance bevel was

mine generation system and electrochemical detection (23). The set at a = 0.05.

GSH plasma GSH plasma

25 25

20 20 0 E E 15 1s a E E a lO I#J%%_TT 10 00 00 #{231}

0 #{149} 0 7 10 10T r”r:i’ I day day GSH blood cells GSH blood cells

20 20

.0 .0

1S is C C E E

(I = 10 . 10 V V ---i___i_I -k .

8 ,0 .0 S I 00 T 00

10 0 1 2 3 4 5 6 7 $ 9 10 day day

Fig. 1 GSH levels (mean ± SD) in plasma (upperpanels) and blood cells (lowerpanels) ofthe patients receiving a continuous ifosfamide infusion

of 1300 mg/m2 daily for 10 days (a = II ). GSH levels of patient 10 are shown separately: the ifosfamide infusion in this patient was stopped on day 7 (i.e.. after 6 days of treatment) because of neunotoxicity. Left panels. average GSH concentrations of all patients on the pretreatment day (day 0). day 7, and the last ifosfamide treatment day (day 10). Daily monitoring of GSH levels during the ifosfamide treatment was started from patient 7 onward (right panels). , P < 0.05. Left panels. , mean (a 1 1 ): , patient 10. Rig/it panels, #{149}.mean (‘i = 5): A. patient 10.

The effect of each ifosfamide treatment schedule on GSH for the 24-h, and 8% for the 3-day ifosfarnide treatment and 0550 concentrations in blood cells and plasma were ana- schedule. lyzed by two-way ANOVA using subjects as a random and days In patient 10 (1300 mg/rn2 ifosfamide daily for 10 days), as a fixed factor. The significance level was set at a = 0.05. To the ifosfarnide infusion was stopped on day 7 of the ifosfarnide maintain the probability of type I error at 0.05, a subsequent infusion period (i.e., after 6 days of ifosfarnide treatment) be- analysis, comparison of each day to the pretreatment day (day cause of neurotoxicity, which was probably related to the ifos- 0), was only performed if the analysis of variance showed famide treatment. Neurotoxicity symptoms included confusion, significant differences between days. disorientation, inability to concentrate, visual hallucinations, and tremors in the hands and legs. Furthermore, patient 301 RESULTS (3000 mg/rn2 ifosfamide daily for 3 days) was suspected of Clinical Effects neurotoxicity. However, the neurotoxicity symptoms of this All ifosfamide treatment schedules were reasonably well patient, consisting of disorientation and confusion, became man- tolerated by all patients. Remarkably. all three ifosfamide treat- ifest after the end of the ifosfamide treatment period.

ment schedules caused a significant decrease in Hb values (P < The concomitant medication of the patients during the

0.001 for the 10- and 24-h treatment and P < 0.05 for the 3-day ifosfamide treatment period consisted in general of antiernetic, ifosfamide treatment): the decreases in Hb were significant analgetic (e.g., acetarninophen), and anxiolytic (e.g.. benzodi- from day 2 (1300 mg/rn2 for 10 days), day I (5000 mg/rn2 X azepines) drugs. The antiernetic drugs used were dexarnethasone 24 h), and day 3 (3000 mg/rn2 for 3 days) onward. The plus granisetron (Kytril) in the 24-h and 3-day ifosfarnide treat- average decreases in Hb levels (expressed as percentage of ment groups and metocboprarnide (Primpenan) or dompenidom the pretreatment level) were about 22% for the 10-day, 17% (Motilium) in the 10-day ifosfamide treatment group.

GSH plasma GSH plasma

25 -

20 20

C E E 15 a 15 a E E 10 0. 10

00 00 S 1 J S

0 i I I 0 F’Ii 0 1 2 0 2 3 day day

GSH blood cells GSH blood cells

,0 .0 16 0 C E E 12

V (I,

#{149}0 I 8 .0 4 Cl)

0 0 1 2 2 day day fig. .1 GSH levels (mean SD) in plasma (upper panel) and blood Fig. 2 GSH levels (mean ± SD) in plasma (upper panel) and blood cells (lower panel) of the patients receiving a continuous ifosfamide cells (lowerpanel) of the patients receiving a 4-h ifosfamide infusion of 3000 mg/m2 daily for 3 days (ii = 6). GSH levels were measured daily infusion of 5000 mg/m2/day for 24 h (n = 6). Day I is the ifosfamide from the pretreatment day (day 0) up to the last ifosfamide treatment day treatment day. (day 3). *, < 0.05 compared to pretreatment.

A 600-mg p.o. dose of racernic brornisoval caused slight sedation in some patients. None of the patients experienced any

bromisoval-related adverse events. ‘ ‘Patients and Methods’ ‘ ), blood cell GSH levels gradually decreased on each ifosfarnide treatment day: the highest levels Characterization of the GSHIGST System were measured on day 1 and the lowest levels on day 3, i.e., the Glutathione (Blood Cells and Plasma) and the GSTMJ last day of the ifosfarnide treatment. In addition, during the Phenotype second course of ifosfamide treatment in patient 306, the height Both the 10- and the 3-day ifosfarnide treatment schedules of the blood cell GSH levels and the profile of its decrease on caused a statistically significant decrease in blood cell GSH each ifosfamide treatment day were more or less identical to

levels (P = 0.03 and P < 0.001, respectively), whereas the 24-h those measured during the first course of ifosfamide treatment. ifosfamide treatment schedule did not have a statistically sig- The three ifosfamide treatment schedules did not have a nificant effect on blood cell GSH levels (lower panels of Figs. statistically significant effect on plasma GSH and blood cell 1-3 and Table 2). Compared with the levels measured on the 0550 levels. A tendency for plasma GSH levels to decrease pretreatment day, blood cell GSH levels were significantly during the ifosfarnide treatment period seemed to be present in

lowered (P < 0.05) from day 2 of the 3-day and day 5 of the the 3-day ifosfamide treatment group (upper panels of Figs. 10-day ifosfamide infusion schedule onward. The average de- 1-3, and Table 2). crease in blood cell GSH levels for both treatment schedules Of all patients studied, 46% were deficient in GSTM 1 was 35%. In patients 306 and 308, receiving the 3-day ifosf- isoenzymes (50% of the patients receiving the 10-day and 24-h amide treatment and in whom blood cell OSH levels were treatment and 33% of the patients receiving the 3-day ifosf- measured more frequently on each ifosfamide study day (see amide treatment; Table I).

Table 2 GSH levels lmean ± SD (nange)b in plasma and blood cells Urinary Excretion. The urinary excretion of brornisoval of pauents receivingifosfamide therapy mercapturates on the pretreatment day was not statistically

Ifosfamide GSHb GSSGb significantly different from that on the last study day (day 10) in treatment Patients GSHp” (p.mol/g (pmol/g the patients receiving the 10-day ifosfamide treatment (Table (mg/m2 X time) (a) Day Hb) Hb) 4A). For the 24-h and 3-day ifosfamide treatments, the cumu- 1300 X 10 days ll 0 6.8 ± 3.2 8.8 ± 3.0 0.19 ± 0.12 lative amount excreted of the mercapturate derived from the (2.9-10.9) (3.1-12.5) (0.04-0.37) (S)-brornisoval was significantly lower on the last day compared 10 9.3 ± 5.1 6.4 ± 1.7’ 0.15 ± 0.12 (2.4-20. 1 ) (3.6-8.8) (0.05-0.32) to the first bromisoval study day (Table 4. B and C). In addition, 5000 X 24 h 6 0 5.4 ± 4.4 8.2 ± 2.6 0.18 ± 0.05 the (R):(S) ratio of the patients receiving the 24-hour ifosfarnide (0.4-1 1.8) (5.7-1 1.7) (0.13-0.27) treatment was significantly higher on the last bromisoval study 2 6.2 ± 9.9 8.9 ± 2.7 0.18 ± 0.10 day (Table 4B). (0.5-26.2) (6.1-13.9) (0.10-0.38) 3000 x 3 days 6 0 5.9 ± 3.2 1 1.9 ± 2.6 0.19 ± 0.13 For all ifosfamide treatment schedules, similar tendencies (2.9-10.0) (9.8-16.9) (0.09-0.44) could be observed in most of the individual urinary excretion 3 3.1 ± 2.2 7.8 ± 3.1’ 0.15 ± 0.10 profiles of the bromisoval mercapturates: the cumulative (0.8-5.6) (5.8-13.6) (0.04-0.32) -- ----------______-______amount excreted of both brornisoval mercaptunates as well as 1 GSHp. GSH in plasma: GSHb and GSSGb, GSH and GSSG in the total amount of mercapturates excreted in 24 h tended to be blood cells, respectively. lower on the last day compared to the first bromisoval study day. I, Without patient 10. dropout because of neurotoxicity. The ifosfamide infusion in this patient was stopped on day 7. i.e., after 6 days of Furthermore, the (R):(S) ratios tended to be increased on the last treatment. bromisoval study day. It should be mentioned that, like for the ‘ Both the 10- and 3-day ifosfamide treatment schedules caused a plasma pharrnacokinetic parameters. all urinary excretion pa- statistically significant decrease in GSH blood cell levels (P 0.03 and rameters of the mencapturates derived from both bromisoval P < 0.001, respectively). Compared with the levels on the pretreatment day. the decrease was significant from day 2 of the 3-day and day 5 of enantiomers were highly variable on both bromisoval study days the 10-day ifosfamide infusion onward. (coefficient of variation values in general <40%, except the values for the clearance of formation), although this variation seemed to be less than that observed in the plasma pharmacokinetic parameters.

Bromisoval Pharmacokinetics Broinisoval Pharmacokinetics and GSH Levels in the In all patients, irrespective of the ifosfamide treatment Neurotoxicity Patient schedule and bromisoval study day. the pharmacokinetics of In patient 10. receiving the 10-day ifosfamide treatment bromisoval enantiomers was more or less similar to those ob- schedule, the ifosfamide infusion was stopped after 6 days (i.e., served in healthy male volunteers (2 1 ): the plasma elimination on day 7) of treatment because of neurotoxicity. Therefore, the and urinary excretion of (R)-bromisoval and its mencapturate, bromisoval pharmacokinetics could only be assessed on the respectively, were always higher than those of (S)-bromisoval pretreatment day (day 0). Blood cell/plasma GSH and 0550 and its mencapturate. levels were measured from day 0 until day 9. i.e., up to 3 days Plasma Pharmacokinetics. In the 24-hour and 3-day after the end of the ifosfamide treatment. ifosfamide treatment groups, the brornisoval plasma pharmaco- On day 0, the plasma elimination pharmacokinetics of kinetics on day 0, i.e., the ifosfamide pretreatment day, were not bromisoval enantiomers and the urinary excretion of their con- statistically significant different from those on day 2 (day after responding mercapturates were not different from those of the the 24-h ifosfamide treatment) and day 3 (last day of the 3-day other patients. On the other hand. both plasma and blood cell ifosfamide treatment). respectively (Table 3. B and C). Signif- OSH levels decreased: plasma levels from 16 p.M (day 0) to 3 icant (P < 0.05) differences between the plasma pharmacoki- P.M (day 9) and blood cell levels from 12 p.mol/g Hb (day 0) to netics on the pretreatment day and last day of the ifosfarnide p.mol/g Hb (day 9; Fig. I ). The blood cell 0550 levels of treatment only occurred for the 10-day ifosfarnide treatment: the this patient were not affected by the ifosfamide treatment, and C,x of both (R)- and (S)-bromisoval and the AUC(00) of they were also not different from those observed in the other (S)-bromisovab were significantly decreased, whereas the p.o. patients (Table 2). Patient 10 was not deficient in GSTMI Cl/F ofthe (S)-enantiomer was significantly increased on day 10 isoenzymes (Table IA). compared to day 0 (Table 3A). Similar tendencies could be observed in the individual DISCUSSION bromisoval plasma pharmacokinetic profiles of both enanti- GSH Concentrations in Blood Cells. In the present omens for, in particular, the 3- and 10-day ifosfamide treatment study. the effects of ifosfamide treatment on the GSH/GST schedules: Cuax and AUC(x) values as well as (S):(R)- ratios system in cancer patients were studied. Statistically significant [AUC(00) and (AUC,18)1 tended to be decreased, whereas p.o. decreases of about 35% were observed for blood cell GSH Cl/F values tended to be increased on the last day compared to levels in the patients receiving the 3- and 10-day ifosfamide the first bromisoval study day. However, the plasma pharmaco- treatments. In the patients receiving the 24-h ifosfamide treat- kinetic parameters of both bromisoval enantiomers were highly ment, blood cell GSH levels were hardly affected. Therefore, variable on both bromisoval study days: the coefficient of van- ifosfamide treatment of longer duration with a lower daily dose ation of all pharmacokinetic parameters generally ranged be- is associated with a more pronounced decrease in blood cell tween 30 and 70%. GSH levels than a relatively short 24-h treatment with a higher

Table 3 Pharmacokinetic parameters (mean ± SD) of the bromiso val enantiomens in p lasma of patients neceiving ifosfam ide therapy

Cmax tmaxb AUC(c) Cl/F ‘I/2 MRT Day (p.mol/liter) (mm) (mmollliter . mm) (liter/mm) (mm) (mm) A. Continuous ifosfamide infusion of 1300 mg/m2 daily for 10 days” (R)-bromisoval 0 1.4 ± 0.7 91 ± 99 0.16 ± 0.10 1 1.5 ± 6.5 60 ± 35 134 ± 73 10 0.8 ± 0.5’ 142 ± 75 0.12 ± 0.12 20.6 ± 14.9 45 ± 31 153 ± 74 (S)-bromisoval 0 11.4±5.2 118±91 2.28±0.78 0.66±0.26 121 t44 247± 108 10 6.2 ± 4.1’ 154 ± 91 1.05 ± 0.41’ 1.53 ± 0.74’ 1 12 ± 81 271 ± 134 (S):(R) ratio AUC(00) AUC(,!SO) 0 17.6 ± 5.7 12.6 ± 6.2 10 15.1 ± 9.7 9.1 ± 7.4

B. Continuous ifosfamide infusion of 5000 mg/m2/day for 24 h” (R)-bromisoval 0 3.0 ± 4.0 96 ± 78 0.22 ± 0.23 12.5 ± 9.5 41 ± 6 124 ± 66 2 2.2 ± 1.0 90 ± 60 0.20 ± 0.04 6.9 ± 1.5 53 ± 28 128 ± 67 (S)-bromisovab 0 11.7±6.6 114±68 2.57± 1.23 0.66±0.37 145±53 271 ± Ill 2 9.6 ± 4.7 1 14 ± 80 2.14 ± 0.98 0.86 ± 0.68 130 ± 18 246 ± 59 (S):(R) ratio AUC(00) AUC(.ISOI 0 17.0 ± 6.2 8.6 ± 3.9 2 10.6 ± 4.5 6.4 ± 2.6 C. 4-h ifosfamide infusion of 3000 mg/m2 daily for 3 days” (R)-bromisoval 0 1.6 ± 0.8 95 ± 62 0.27 ± 0.10 5.5 ± 1.6 85 ± 37 178 ± 88 3 1.3 ± 0.7 150 ± 99 0.21 ± 0.09 7.4 ± 3.5 76 ± 56 209 ± 125 (S)-bromisoval 0 8. 1 ± 4.7 1 24 ± 43 1 .96 ± I .25 0.94 ± 0.52 99 ± 20 227 ± I 8 3 5.8 ± 2.3 162 ± 66 1.07 ± 0.40 1.37 ± 0.38 85 ± 22 238 ± 70 (S):(R) ratio AUC(oo) AUC.1SO) 0 8.2 ± 5.6 5.9 ± 3.2 3 5.5 ± 2.5 5.1 ± 3.5

‘I Data of patient 2 (day 0, (R)- and (S)-bromisoval) and patient 9 (day 0, (R)-bromisoval, except and tnix) are missing (due to analytical problems and very low concentrations of (R)-bromisoval, respectively). Data of patient 10 [(R)- and (S)-bromisoval] were not included in the data analysis because the ifosfamide infusion was stopped on day 7 (neurotoxicity). S tmax, time to reach Cmax; MRT, mean residence time.

C p < 0.05, significantly different compared to day 0, i.e., the ifosfamide pretreatment day.

d , patients 102 and 304 did not receive bromisoval.

ifosfarnide dose. Because blood cell GSH levels measured daily assay differences may have contributed to this observed discrep- for 1 1 days in two healthy adults appeared to be stable (mean ± ancy: Lind et al. (15) used the rnonochborobirnane method to SD for the whole 1 1-day study period: 7.6 ± 0.6 and 5.8 ± 1.1 measure cellular OSH levels. However, GSH quantification by p.moVg Hb, respectively),5 the decrease observed in the patients this assay is dependent on the GST isoenzyme composition of receiving the 3- or 10-day ifosfamide treatment is likely to be cells/tissues (25) and may therefore not accurately estimate the

caused by ifosfamide and/or its metabolites. ‘ ‘real’ ‘ OSH levels of lymphocytes. The data obtained in patients 306 and 308 (3-day ifos- Plasma GSH and blood cell 0550 levels were not signif- famide treatment) support that the decrease in blood cell GSH icantly affected by any of the three ifosfamide infusion sched- levels is related to ifosfamide treatment: blood cell GSH levels ules. This may be partially related to the large variation ob- of these two patients decreased gradually on each ifosfarnide served in these levels: the coefficients of variation on each study treatment day, and the height of and decrease in blood cell GSH day (irrespective of the treatment) were about 60% for both levels were very similar during the first and second ifosfamide plasma GSH and blood cell 0550 levels compared to 30% for treatment course (patient 306 only). The latter also indicates that blood cell GSH levels. Because OSH concentrations in plasma blood cell OSH levels were recovered from a decrease of about are much lower than those in blood cells (p.M s’ersus mM range, 40% in the period (i.e., about 5 weeks) between both ifosfamide respectively), a small extent of hemolysis may have a pro- treatment courses. nounced effect on the plasma GSH levels. Recently, it was even The absence of effects caused by the 24-h ifosfarnide demonstrated that hemolysis during sample collection is inevi-

treatment on blood cell OSH levels seems to be in contrast to table and contributed on average 25% to ‘ ‘plasma’ ‘ GSH (26). results obtained by Lind et a!. (15). They reported a decrease of Information about tissue or organ OSH availability in pa- GSH to about 30% of its original value in lymphocytes isolated tients in s’ivo can in general only be obtained indirectly by from one patient undergoing an 8-h ifosfamide infusion of 5 monitoring of GSH bevels in blood (cells)/plasma. However, it g/m2 along with an equal dose of the uroprotector mesna. Apart still remains to be established whether GSH levels in blood cells from the fact that this observation was made in only one patient, are related to those in tumors or other tissues. So far, a corre- lation has only been described between GSH levels in plasma and liver of rats and liver disease patients (27, 28). Furthermore, several factors may contribute to the intra- and interpatient

5 T. M. T. Mulders, H. J. Keizer, J. Ouwerkerk, unpublished results. variability in blood OSH levels: these include the nutritional

Table 4 Pharmacokinetic parameters (mean ± SD) of the mercapturates derived from (R)- and (S)-bromisoval in urine of patients receiving ifosfamide therapy

Ae Ae(k Cl” Day t,,2 (mm) (i.mol) dose) (liter/mm) A. Continuous ifosfamide infusion of 1300 mg/lw daily for 10 days” Mencaptunate derived from (R)-bnomisovab 0 99 ± 26 762 ± 162 28.3 ± 6.0 6.56 ± 3.88 10 124 ± 57 700 ± 192 26.0 ± 7.1 9.35 ± 4.44 Mercapturate derived from (S)-bromisoval 0 122 ± 37 176 ± 42 6.56 ± 1.57 0.09 ± 0.04 10 168 ± 88 163 ± 125 6.07 ± 4.66 0.16 ± 0.10

Ae total (( dose) (R):(S) ratio 0 34.9 7.3 4.42 ± 0.80 10 32.1 ± 10.3 5.18 ± 1.61

B. Continuous ilosfaniide infusion of 5000 mg/m2/day for 24 h’ Mencapturate derived from (R)-bromisoval 0 138 ± 30 686 ± 239 25.5 ± 8.9 6.31 ± 4.10 2 109 ± 13 632 ± 145 23.5 ± 5.4 3.15 ± 0.58 Mercapturate derived from (S)-bromisoval 0 163 ± 15 20 I ± 45 7.46 ± 1.66 0. 10 ± 0.06 2 148 ± 30 132 ± 18” 4.89 ± 0.67” 0.08 0.05

Ae total (c/c dose) (R):(S) ratio 0 33.0 ± 10.3 3.37 ± 0.79 2 28.4 ± 5.0 4.95 ± 1.55”

C. 4-h ifosfamide infusion of 30()0 mg/m daily ftn 3 days’ Mercapturate derived from (R)-hromisoval 0 1 1 1 ± 20 700 ± 30 26.0 ± 1.1 2.86 ± 0.91 3 127 ± 42 675 ± 130 25.1 ± 4.8 3.61 ± 1.58 Mencapturate derived from (S)-bromisoval 0 166 ± 21 185 ± 37 6.88 ± 1.37 0.12 ± 0.05 3 170 ± 71 1 12 ± 47” 4.17 ± 1.74” 0.12 ± 0.07

Ae total (% dose) (R):(S) ratio 0 32.9 ± 2.5 3.89 ± 0.66 3 29.3 ± 5.7 7.08 ± 3.42

“ Data of patient 10 not included (dropout on day 7 because of neurotoxicity). ‘, Data of patients 2 (day 0, (R)- and (S)-bromisoval) and 9 (day 0. (R)-bnomisoval) were not included because of missing AUC data.

‘ ii = 5, patients 102 and 304 did not receive bromisoval. ‘I p < 0.05, significantly different compared to day 0. i.e.. the ifosfamide pretreatment day.

status of the patients un particular intake of ascorbate (29-3 1 )], enantiomer in healthy male volunteers (20, 2 1 ) and cancer age (32), the presence of (additional) disease states associated patients (present study), is likely to reflect the steneoselectivity with changes in GSH availability (e.g.. disturbances in liver and of specific human GST isoenzymes toward both bromisoval thyroid function; Refs. 28 and 33-35). intra- and interpatient enantiomers. So far. the tissues and GST isoenzymes involved differences in concomitant medication (e.g.. paracetamol). phys- in preferential conjugation of (R)-bromisoval in humans in s’ito ical activity (36), and interpatient differences in expression of could not (yet) be identified (38). -y-glutamybcysteine synthetase (the rate-limiting enzyme in GSH In all three treatment groups. i.e., irrespective of the ifos- biosynthesis; Ref. 37). famide infusion schedule, the plasma elimination pharmacoki- GSH Conjugation Activity as Assessed by Bromisoval netics of (R)- and (S)-bromisoval and urinary excretion of their Pharmacokinetics. In s’ii’o characterization of GSH conjuga- corresponding mercapturates were not significantly affected by tion activity in patients was performed by administering the the ifosfarnide treatment, although certain trends could be ob- probe drug bromisoval. The advantage of using a probe drug for served. The bromisoval pharmacokinetics of cancer patients on i?i ViVO characterization is that information can be obtained the ifosfarnide pretreatment day were even comparable to those about the activity and/or profile of GST isoenzymes as well as assessed in healthy male volunteers (2 1 ). which indicates that the OSH availability. Because specific activities of individual GSH conjugation of bromisoval enantiomers in these cancer GST isoenzyrnes toward a model substrate. i.e.. in the present patients in vito was also not affected by their disease state. study both bromisoval enantiomers. are likely to be different, Because bromisovab pharmacokinetics are not influenced by the the pharmacokinetics of the substrate and its corresponding presence or absence of GSTM 1 isoenzymes (2 1 ). differences in GSH conjugates and mercapturates (N-acetylcysteine conju- GSTM 1 phenotype will not have influenced the present results. gates) will selectively reflect the activities of the isoenzyrnes Assuming that the kinetics of the probe drug indeed provide an involved. In this respect. the stereoselectivity in brornisoval accurate reflection of GST enzyme activity, the bromisoval data pharmacokinetics. preferential GSH conjugation of the (R)- therefore suggest that OST activity remains unchanged and that

the GSH availability of the cancer patients was not rate-limiting etron (Kytril) as antiernetic therapy in almost all patients receiv- in GSH conjugation of bromisoval enantiomers. ing the 24-h and 3-day ifosfamide treatment schedules. Whether Comparison of the individual pharmacokinetic profiles of the administration of dexarnethasone had any effect on the the bromisoval enantiomers and their mercapturates on both therapeutic efficacy of ifosfamide is not clear. Nevertheless, the study days revealed similar tendencies for all patients, in par- administration of dexamethasone may have contributed, by in- ticular those receiving the 3- and 10-day ifosfamide infusion creasing the presence of ifosfamide metabolites, to the observed schedules: the Cmax and AUC values were lower and, corre- decrease in OSH blood cell levels in the patients receiving the spondingly, the oral Cl/F was higher on the last day compared 3-day ifosfamide treatment. Furthermore, the possible occur- to the first bromisoval study day. Assuming complete absorption rence of autoinduction of ifosfamide metabolism in the course of bromisoval enantiomers on both bromisoval study days, these of ifosfamide therapy (4, 43) may (partially) explain why only changes in bromisoval pharmacokinetics may be caused by the longer ifosfamide treatment schedules, i.e., the 3- and 10- (slightly) increased activity of specific GST isoenzymes. If day treatment regimens, were associated with a significant de- ifosfamide and/or its metabolites cause increases in the activity crease in blood cell OSH levels. of specific GST isoenzymes (and/or changes in GST profiles) Conclusions. Characterization of the GSHIGST system after long-term exposure, this may lead to development of drug in cancer patients in vivo by administering the probe drug resistance toward specific cytostatic agents (including ifos- bromisoval showed that the influence of all three ifosfamide farnide) for which the GSH/GST system is involved in bioinac- tivation. treatment schedules on the pharmacokinetics of both bromisoval For all three ifosfamide treatment schedules, the cumula- enantiomers was only limited. On the other hand, the two longer tive amount of mercapturates excreted tended to be lower on the treatment schedules, i.e., the 3- and 10-day ifosfarnide treat- last day compared to the first bromisoval study day, which does ments, were associated with a significant decrease in GSH blood not seem be to in agreement with the apparently higher p.o. Cl/F cell levels of approximately 35%. Since lowering of GSH levels of both brornisoval enantiomers. However, because the second may sensitize (resistant) tumors to cytostatic agents, these data bromisoval study day coincided with the last day of the ifosf- suggest that ifosfamide may potentially be used for this purpose amide treatment for the 3- and 10-day treatment schedules, an during cytostatic therapy. However, whether blood cell OSH influence of ifosfamide and its metabolites on the bromisoval levels reflect those in tissues and whether a decrease of 35% is pharmacokinetics on this study day cannot be excluded. In the sufficient for sensitization of (resistant) tumors need to be patients receiving the 24-h ifosfamide treatment, an interaction studied in greater detail. Furthermore, it should be mentioned between ifosfamide (metabolites) and bromisoval on the day that the complexity of ifosfarnide metabolism [dependence on after the ifosfamide treatment can also not be ruled out [plasma CYP3A isoenzyrnes for bioactivation, autoinduction of the he- ti,2 of ifosfamide after iv. administration of 5 g/m2/day for 24 patic oxygenase system, and stereoselectivity in ifosfamide me- h: 4.5 hours (39)]. tabolism (4, 44)] may make this GSH-rnodulating approach of Changes in blood cell GSH levels did not seem to affect combined administration of ifosfamide and (an)other cytostatic GSH conjugation of bromisoval enantiorners on the second agent(s) difficult. In addition, because resistance of tumors to bromisoval study day. This may be related to the presence of a cytostatic therapy [including ifosfamide (45)] is in general rnul- low Km of GST isoenzymes for GSH: in rats, 70-80% of tifactorial, modulation of GSH levels will only be effective for hepatic GSH must be depleted before the GSH availability those tumors and for those cytostatic treatments for which becomes rate limiting for in vivo GSH conjugation (40). A involvement of the GSHIGST system in resistance is a major decrease in blood cell GSH levels of approximately 70% was contributing factor (14). observed in the patient (patient 10) exhibiting severe neurotoxicity. However, the clinical condition of this patient did not allow bromisoval administration on the last day of the ifosfarnide treatment, and, consequently, the effects of this possibly ACKNOWLEDGMENTS rate-limiting decrease in GSH blood cell levels on the in vivo We thank the nurses of the Department of Clinical Oncology for OSH conjugation of bromisoval enantiomers could not be stud- their contribution to the practical part of this study, and V. Venizebos ied. Nevertheless, this large decrease in GSH blood cell (and and J. Schutrups for their technical assistance (bromisoval and GSHI plasma) levels may have contributed to the neurotoxicity expe- GSSG assays). The assessment of the GSTM 1 phenotype was pen- rienced by this patient. Renal dysfunction, as detected by ele- formed by J. J. P. Bogaards (TNO Nutrition and Food Research, De- vated serum creatinine levels (>200 p.mol/liter; range, 70-133 partment of Biological Toxicology, Zeist. the Netherlands). i.mol/liter), may also have been a contributing factor (6). Concomitant Medication and Autoinduction of Ifos- famide Metabolism. Ifosfamide requires biotransformation REFERENCES by cytochrome P450 isoenzymes to exert its alkylating action. 1. Zalupski, M., and Baker, L. H. Ifosfamide. J. NatI. Cancer Inst., 80: Recently, studies using human liver microsomes revealed the 556-566, 1988. involvement of CYP3A4 as the major cytochrome P450 respon- 2. Brock, N. Oxazaphosphonine cytostatics: past-present-future. Cancer sible for both ifosfamide 4-hydroxylation (bioactivation) and Res., 49: 1-7, 1989. N-dechloroethylation (a detoxification pathway; Refs. 41 and 3. Skinner, R., Sharkey, I. M., Pearson, A. D. J., and Craft, A. W. 42). In this respect, it is important to mention that the CYP3A Ifosfamide, mesna, and nephnotoxicity in children. J. Clin. Oncol., 11: inducer dexamethasone was administered along with granis- 173-190, 1993.

4. Kaijser, G. P.. Beijnen, J. H., Bult, A., and Undenbeng, W. J. M. by electrochemically generated bromine. J. Chromatogn.. 427: 67- lfosfamide metabolism and pharmacokinetics (Review). Anticancer 77. 1988. Res., 14: 517-532, 1994. 24. Gibaldi, M., and Pemer, D. Noncompartmental analysis based on 5. Lewis, L. D., and Meanwell. C. A. Ifosfamide pharmacokinetics and statistical moment theory. In: M. Gibaldi and D. Perrier (eds.), neurotoxicity. Lancet. 335: 175-176, 1990. Phanmacokinetics, Ed. 2, pp. 409-424. New York: Marcel Dekker, Inc.. 6. Miller, L. J., and Eaton, V. E. tfosfamide-induced neurotoxicity: a 1982. case report and review of the literature. Ann. Pharmacother.. 26: 183- 25. Ublacker, G. A., Johnson, J. A., Siegel, F. L., and Mulcahy, R. T. 187. 1992. Influence of glutathione S-transferases on cellular glutathione determi- 7. Waxman, D. J. Glutathione S-transfenases: role in alkylating agent nation by flow cytometry using monochlorobimane. Cancer Res., 51: resistance and possible target for modulation chemotherapy-a review. 1783-1788, 1991. Cancer Res., 50: 6449-6454. 1990. 26. Schofield, D.. Mei, G., and Bnaganza, J. M. Some pitfalls in the 8. Hayes. J. D.. Pickett, C. B.. and Mantle. T. J. (eds.) Glutathione measurement of blood glutathione. Clin. Sci.. 85: 213-218, 1993. S-Tnansfenases and Drug Resistance. London: Taylor and Francis, 1990. 27. Adams, J. D., Lautenburg, B. H., and Mitchell, J. R. Plasma gluta- 9. Ozols. R. F., O’Dwyer. P. J.. Hamilton, T. C.. and Young. R. C. The thione and glutathione disulfide in the rat: regulation and response to role of glutathione in drug resistance. Cancer Treat. Rev., 17: 45-50, oxidative stress. J. Pharmacol. Exp. Then., 227: 749-754, 1983. 1990. 28. Shigesawa. T., Sato, C., and Marumo, F. Significance of plasma 10. Mistry, P., and Hanrap, K. R. Historical aspects of gbutathione and glutathione determination in patients with alcoholic and non-alcoholic cancer chemotherapy. Pharmacol. Then., 49: 125-132, 1991. liver disease. J. Gastroenterob. Hepatotol., 7: 7-1 1, 1992.

1 1. Tsuchida, S., and Sato, K. Glutathione transferases and cancer. Cnit. 29. Meister, A. Depletion of glutathione in normal and malignant Rev. Biochem. Mol. Biol., 27: 337-384, 1992. human cells in vivo by L-buthionine sulfoximine: possible interaction with asconbate. J. Natl. Cancer Inst., 84: 1601-1602, 1992. 12. Ciaccio, P. J.. and Tew, K. D. Glutathione S-transferases. In: B. A. Teicher (ed), Drug Resistance in Oncology. pp. 351-374. New York: 30. Meisten, A. On the antioxidant effects of ascorbic acid and gluta- Marcel Dekker, Inc., 1993. thione. Biochem. Pharmacol.. 44: 1905-1915, 1992. 13. Schecter, R. L., Alaoui-Jamali, M. A.. Woo, A., FahI, W. E., and 31 . Johnston, C. S., Meyer, C. G., and Snilakshmi, J. C. Vitamin C Batist, G. Expression of a rat glutathione-S-tnansferase complementary elevates red blood cell glutathione in healthy adults. Am. J. Clin. Nutr., DNA in rat mammary carcinoma cells: impact upon alkylaton-induced 58: 103-105, 1993. toxicity. Cancer Res.. 53: 4900-4906. 1993. 32. Lang. C. A., Naryshkin, S., Schneider, D. L., Mills, B. J., and 14. Mulders. T. M. T.. Keizer, H. J., Breimer, D. D.. and Mulden, G. J. Lindeman, R. D. Low blood glutathione levels in healthy aging adults.

1,1 tOO characterization and modulation of the glutathione/glutathione J. Lab. Clin. Med., 120: 720-725, 1992. S-transfenase system in cancer patients. Drug. Metab. Rev.. 27: 191- 33. Israel, Y., Speisky, H., Lanca, A. J., Iwamura, S., Hirai, M., and 229, 1995. Vanghese. G. Metabolism of hepatic glutathione and its relevance in 15. Lind, M. J., McGown, A. T., Hadfield, J. A., Thatcher, N., alcohol induced liver damage. In: B. Clement and A. Guillouzo (eds.), Crowther. D., and Fox, B. W. The effect of ifosfamide and its metab- Cellular and Molecular Aspects of Cirrhosis, pp. 25-37. Colloque olites on intracellular glutathione levels in vitro and in vito. Biochem. Institut National de Ia Sante et de Ia Recherche M#{233}dicale/John Libbey Pharmacol.. 38: 1835-1840. 1989. Eurotext Ltd.. 1992. 16. Peters, R. H., Jollow, D. J., and Stuart, R. K. Role of glutathione in 34. Sir, T., Wolff, C., Soto, J. R., Perez, G., and Memo, R. A. the in i’itro synergism between 4-hydnopenoxycyclophosphamide and Relationship between hepatic levels of glutathione and sulphobromoph- cisplatin in leukemia cell lines. Cancer Res., 51: 2536-2541. 1991. thalein retention in hyperthynoidism. Clin. Sci., 73: 235-237. 1987. 17. Yuan, Z. M., Smith, P. B., Bnundnett, R. B.. Colvin, M., and 35. Fernandez, V., Simizu, K., Barros, S. B. M., Azzalis. L. A., Pimen- Fenselau, C. Glutathione conjugation with phosphoramide mustard and tel, R., Junqueina, V. B. C., and Videla, L. A. Effects of hyperthyroidism cyclophosphamide. A mechanistic study using tandem mass spectrom- on rat liven glutathione metabolism: related enzymes’ activities, efflux, etny. Drug Metab. Dispos., 19: 625-629, 1991. and turnover. Endocrinology, 129: 85-91, 1991. 18. Dinven, H. A. A. M., van Ommen, B., and van Bladeren, P. 36. Ji. L. L.. Katz, A.. Fu, R., Griffiths. M., and Spencer. M. Blood Involvement of human glutathione S-transferase isoenzymes in the glutathione status during exercise: effect of carbohydrate supplementa- conjugation of cycbophosphamide metabolites with glutathione. Cancer tion. J. Appl. Physiol., 74: 788-792, 1993. Res., 54: 6215-6220, 1994. 37. Yao, K. S., Godwin, A. K., Ozols, R. F., Hamilton, T. C., and 19. Berhane, K., and Mannervik, B. Inactivation of the genotoxic aIde- O’Dwyer. P. J. Variable baseline -y-glutamylcysteine synthetase mes- hyde acrolein by human glutathione transferases of classes Alpha. Mu senger RNA expression in peripheral mononuclear cells of cancer pa- and P1. Mol. Pharmacol., 37: 25 1-254, 1990. tients, and its induction by buthionine sulfoximine treatment. Cancer Res., 53: 3662-3666, 1993. 20. te Koppele. J. M., Schippen. C., Bneimer, D. D., and Mulder, G. J. Steneoselectivity in the urinary excretion of the mencapturates of (R-) 38. Molders, T. M. T.. van Ommen, B., van Bladenen, P. J., Breimer, and (S-) a-bnomoisovalenylurea in man. Br. J. Clin. Pharmacol., 28: D. D., and Mulder, G. J. Stereoselectivity of human liver and intestinal 151-154. 1989. cytosolic fractions as well as purified human glutathione S-transferase isoenzymes towards 2-bromoisovalerylurea enantiomers. Biochem. 21. Mulders, T. M. T., Venizelos, V., Schoemaker. R.. Cohen, A. F.. Pharmacol., 46: 1775-1780, 1993. Breimer, D. D., and Mulder. G. J. Characterization of glutathione conjugation in humans: Stereoselectivity in plasma elimination pharma- 39. Lewis, L. D., Fitzgerald, D. L., Mohan, P., Thatcher, N., Harper, cokinetics and urinary excretion of (R)- and (S)- 2-bnomoisovalerylunea P. G.. and Rogers, H. J. The pharmacokinetics of ifosfamide given as in healthy volunteers. Clin. Pharmacol. Then., 53: 49-58, 1993. short and long intravenous infusions in cancer patients. Br. J. Clin. Pharmacol., 3/: 77-82. 1991. 22. Venizelos, V., Irth, H., Tjaden, U. R., van den Greef, J., Breimer, D. D., Muldens, T. M. T.. and Mulder, G. J. Enantioselective detenmi- 40. Polhuijs, M., Lankhaar, G., and Mulden, G. J. Relationship between nation of R- and S- (a-bromoisovalenyl)-unea in plasma using high- glutathione content in the liver and glutathione conjugation rate in the performance liquid chromatography after solid-phase extraction. J. rat in vito. Biochem. J., 285: 401-404, 1992. Chromatogn., 573: 259-264, 1992. 41. Chang, T. K. H., Weber, G. F., Crespi, C. L., and Waxman, D. J. 23. te Koppele. J. M., van den Mark, E. J.. and Mulden, G. J. Liquid Differential activation of cyclophosphamide and ifosphamide by cyto- chromatographic determination of diastereomenic glutathione conjugates chromes P-450 2B and 3A in human liver microsomes. Cancer Res., 53: and further denivates of cx-bnomoisovalenylurea in rat bile and urine 5629-5637, 1993.

42. Walker, D., Flinois, J. P., Monkman, S. C., Beloc, C., Boddy, A. V., ifosfamide therapy. Cancer Chemothen. Phanmacol.. 33: 36-42. Cholerton, S., Daly, A. K.. Lind, M. J., Pearson, A. D. J., Beaune, P. H.. 1993. and Idle, J. R. Identification of the major human hepatic cytochrome 44. Wainer, I. W., Ducharme, J., Granvil, C. P., Trudeau, M., and P450 involved in activation and N-dechloroethylation of ifosfamide. Leyland-Jones, B. Ifosfamide stereoselective dechlorocthylation and Biochem. Pharmacol., 47: 1157-1163, 1994. neurotoxicity. Lancet, 343: 982-983, 1994. 43. Kurowski, V., and Wagner, T. Comparative pharmacokinetics of 45. Sladek, N. E. Oxazaphosphonine-specific acquired cellular resis- ifosfamide, 4-hydnoxyifosfamide, chloroacetaldehyde, and 2- and tance. In: B. A. Teicher (ed), Drug Resistance in Oncology. pp. 375- 3-dechboroethylifosfamide in patients on fractionated intravenous 411. New York: Marcel Dekker, Inc., 1993.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 1995 American Association for Cancer Research. Effect of ifosfamide treatment on glutathione and glutathione conjugation activity in patients with advanced cancers.

T M Mulders, Keizer, J Ouwerkerk, et al.

Clin Cancer Res 1995;1:1525-1536.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/1/12/1525

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/1/12/1525. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.