[CANCER RESEARCH 5.1, 3758-3764. August 15. 1993]

Pharmacokinetics and Metabolism of Ifosfamide Administered as a ContinuousInfusion in Children1

A. V. Boddy,2 S. M. Yule, R. Wyllie, L. Price, A. D. J. Pearson, and J. R. Idle

Departments n[ Pharmacological Sciences /A. V. B., J. R. /./ ami Child Health /S. M. Y., R. W., L. P., A. D. J. P.], The Medical School, University of Newcastle upon Tyne, NE24HH, Great Britain

ABSTRACT

The pharmacokinetics and metabolism of ifosfamide was investigatedin a group of 16 pediatrie patients (5 girls) aged 1-17 years. Each receiveda dose of 3 g/m2/day for up to 3 days by continuous infusion. Plasma and

urine were collected, and concentrations of ifosfamide and its principalmetabolites were determined by a quantitative high-performance thin

layer chromatography method. During 3 days of continuous infusion, theplasma concentrations of parent drug decreased. This was accompaniedby a continuous increase in dechloroethylated products in plasma but notin urine. Estimated pharmacokinetic parameters (clearance, volume ofdistribution, and half-life) were dependent on body size and age but not

any other patient variable. Renal clearance was a relatively minor route ofelimination for parent drug and corresponded to <25% of glomerularfiltration rate.

Metabolite data from plasma and urine indicated a high degree ofinterindividual variation in metabolism. Comparison of metabolite recoveries in urine indicated a positive correlation between activation andinactivation routes of metabolism. Prior exposure to ifosfamide was associated with a higher recovery in urine of dechloroethylated metabolites.The severity of hematological toxicity was inversely correlated with glomerular filtration rate but not to parameters of ifosfamide metabolism.There was marked variation in levels of the carboxy metabolite, whichcould not be detected in the plasma of 5 subjects. However, evidence for apolymorphism in metabolism to this metabolite was weaker than that seenwith the isomerie oxazaphosphorine cyclophosphamide. There appearedto be a higher clearance of ifosfamide in pediatrie patients compared toadults. The significance of this, and of the variation in metabolism ofifosfamide, for clinical outcome remains to be established, but the increasein the dechloroethylation route of metabolism may be associated with anincreased risk of toxicity.

INTRODUCTION

The alkylating agent ifosfamide was introduced into clinical trialsin 1970, but its early use was limited by severe hemorrhagic cystitis.Further research led to the development of Mesna as a safe andeffective means of regional uroprotection (1). Following this discovery, phase II studies in children demonstrated activity against a widerange of tumor types (2-4). Ifosfamide is presently included in com

bination chemotherapy for several tumors, including Ewings sarcoma,osteosarcoma, rhabdomyosarcoma, and other soft tissue sarcomas.Adverse effects of ifosfamide include myelosuppression, nausea andvomiting, alopecia, and urotoxicity (5, 6). Urotoxicity is minimalwhen Mesna is administered concurrently with ifosfamide. Encepha-

lopathy is a serious consequence of therapy but is seen much morefrequently in adults than in children (7, 8). The major chronic toxicity in pediatrie patients treated with ifosfamide is nephrotoxicity.leading to rickets and growth retardation due to renal tubular acido-

sis (9, 10). These side effects are unpredictable and are often severeenough to restrict treatment. Moreover, just as host toxicity varies

Received 2/1/93; accepted 6/5/93.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicale this faci.

1This work was supported by grants from the North of England Cancer ResearchCampaign. North of England Children's Cancer Research Fund, and Asta Medica AG,

Frankfurt. Germany.2 To whom requests for reprints should be addressed.

among individuals, some tumors are chemosensitive and curable,whereas others remain resistant or recur following an initial response.

Ifosfamide itself possesses little cytotoxic effect. It is a prodrugwhich is metabolised in vivo to produce a variety of therapeuticallyactive and potentially toxic metabolites (11, 12). Thus, for a givenindividual, variation in metabolism between host and tumor tissuemay result in variability in toxicity and in differences in chemosen-

sitivity of the tumor. The initial activation reaction in the metabolismof ifosfamide is thought to be mediated by a hepatic cytochrome P450enzyme (Fig. 1) (13, 14). Hydroxylation at the carbon-4 position ofthe oxazaphosphorine ring produces 4-hydroxyifosfamide, which exists in equilibrium with its tautomerie form, aldo-ifosfamide. Thelatter form may then either be oxidized by an ALDH-1 enzyme (15, 16)

to CX (an inactive metabolite) or spontaneously decompose to formIPM. The mustard is thought to be the primary alkylating agent (17).Acrolein is formed as a by-product of the latter reaction and is be

lieved to be responsible for the urotoxic effects of ifosfamide. Up to50% of a dose of ifosfamide undergoes a separate oxidative A'-dealky-

lation reaction, resulting in the loss of one or other of the chloroethylside chains to produce either 2-DCI or 3-DCI (18-20). An equimolar

quantity of chloroacetaldehyde is formed in each of these reactions,and this toxic metabolite has been implicated in the neurotoxicitywhich may accompany ifosfamide therapy (21) and may also beassociated with nephrotoxicity (10). Other inactive metabolites include KETO. which is thought to result from oxidation of 4-hydrox

yifosfamide (11).Large interpatient differences in ifosfamide metabolism have been

reported in adults (12), including wide variation in CX excretion.Studies of the urinary metabolites of cyclophosphamide, an isomer ofifosfamide, indicate that certain individuals may be totally deficient inthe excretion of the corresponding metabolite (22, 23). It has beensuggested that this may be the result of phenotypic variation in ALDHactivity, which would be expected to apply equally to ifosfamidemetabolism, assuming that both oxazaphosphorines are metabolizedby the same ALDH enzyme. Patients with low levels of ALDH activity may deactivate ifosfamide less efficiently and may be at increased risk of toxicity, possibly accompanied by apparently greatertumor sensitivity. Conversely, tumor inactivation of ifosfamide byALDH may result in chemoresistance.

Similarly, the activation of cyclophosphamide and of ifosfamidemay be subject to great interindividual variability (23, 24). The humancytochrome P450 enzymes responsible for the initial hydroxylationreaction and inactivating dechloroethylation reactions remain uncertain. Environmental and genetic factors lead to large interindividualdifferences in activities of various members of the P450 superfami-

lies, which would, in turn, have an influence on the balance of activation and inactivation of a dose of ifosfamide in an individual.

Early studies of ifosfamide demonstrated that fractionated dosing ofifosfamide produced better therapeutic activity and was better toler-

'The abbreviations used are: ALDH. aldehyde dehydrogenase; IPM. isophosphora-mide mustard. CX. carboxyifosfamide: DC1, dechloroethylifosfamide; KETO, 4-ketoifos-famide; Cl. clearance; TLC. thin-layer Chromatograph; AUC. area under the plasmaconcentration-time curve; NBP. 4-nitrobenzylpyridinc; GFR. glomerular nitration rate;

Vß.volume of distribution: IFO. ifosfamide.

3758

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

O Ov°\y\A" .

IFOSFAMIDE

o o

Va\/\ Atí* H'

2- AND 3-DECHLOROETHYMFO+ chloroacetaldehyde

ALDO IFOSFAMIDE

/O

COOH

CARBOXYIFOSFAMIDE

CHO

ISOPHOSPHORAMIDE MUSTARD ACROLEIN

Fig. I. Metabolism of ifosfamide.

ated compared with single doses (25, 26). Subsequent trials have usedlower doses administered over 3-5 days, hut at present the optimum

dosage schedule remains uncertain. To investigate the variation inmetabolism of ifosfamide and the rationale of fractionated dosing inchildren, we studied the pharmacokinetics and metabolism of ifosfamide administered as a continuous infusion over 72 h. As yet, theclinical consequences of interpatient variation in ifosfamide metabolism arc unknown and are not the major focus of this investigation.

MATERIALS AND METHODS

Ifosfamide and its metabolites were obtained from Asia Medica (Frankfurt,Germany). Cyclophosphamide and NBP were purchased from Sigma (Poole.

United Kingdom). All other reagents were of appropriate analytical grade.Sixteen patients (5 girls) were being treated with ifosfamide regimens for

sarcomas. Ages ranged from 1-17 years (median, 4 years). Patients received

successive doses of ifosfamide every 3 weeks as a continuous infusion (GeminiPC-2 volumetric infusion pump; Imed. San Diego, CA) at a dose of 3 g/irr

each day for 3 days (14 patients) or 2 days (2 patients). This was accompaniedby 3 liters/m2 of hydration each day and Mesna (3 g/rrr/day), infused during

and for 12 h after ifosfamide administration. Other chemotherapy is listed in

Table 1. Antiemetics (ondansetron. metoclopramide, or dexamethasone) and

prophylactic antibiotics (cotrimoxazole) were also administered. For each patient, clinical status, renal function (GFR by EDTA clearance or estimated fromplasma crcatininc), liver function (alanine transaminase, bilirubin. and albumin), and hematological toxicity were monitored throughout the treatmentperiod. Thirteen patients are disease free at 6-26 (median, 12 months) follow

ing diagnosis. Ten have completed therapy. Two patients relapsed at 21 months,and one child died from overwhelming infection during therapy. The study wasapproved by the Ethical Committees of the Medical School of the Universityof Newcastle upon Tyne and the Royal Victoria Infirmary, Newcastle.

Blood samples (3-5 ml depending on the size of the child) were collected

immediately before, at 3. 6, 12, 18, 24, 36, 48, and 60 h after the start of the

infusion, at the end of the infusion, and at 1.2, 4, 6, 12, 18, and 24 h after theend of the infusion. Blood was anticoagulated with EDTA, and plasma wasseparated and frozen immediately at -20°C prior to analysis. Urine was col

lected at 6-h intervals throughout the infusion and for 24 h alter. Each passageof urine was stored at 5°Cuntil the end of the collection period. The volumeof each urine collection was measured, and an aliquot was frozen at -20°C for

subsequent analysis.Concentrations of ifosfamide, isophosphoramide mustard, carboxyifosfa-

mide, 2- and 3-dechlorocthylifosfamide. and 4-ketoifosfamide were determined in urine and plasma using a quantitative thin-layer chromatography-

photography densitomctry technique (27). Briefly. 1 ml of each urine sampleand 50 JA|of internal standard (500 fig-mh' cyclophosphamide in methanol)

was applied to an XAD-2 Spe-Ed solid phase extraction cartridge (500 mg/3

ml; Laboratory Impex Ltd., Teddington, United Kingdom). The cartridge waswashed with 3 ml water and dried. Drug and metabolites were cluted with

methanol. which was evaporated to dryness.Plasma (750 jul) was added to 750 n\ cold acelonitrile and 50 f¿linternal

standard. After vortex mixing and centrifugation, the clear supernatant wasevaporated to dryness.

Dry residues from the above procedures were reconstituted in methanol (70fji\). and 30 /il were applied to silica gel TLC plates (E. Merck, Darmstadt.

Table 1 l'alieni anil stuify ¡Iftails

PatientDiagnosisI

Rhabdo"2

Rhahdo3Ewings4Rhahdo5Rhahdo6

RhahdoNRsarcoma8Rhahdo9

RhahdoW1Ostcosarcoma11JOsteosarcoma12

NRsarcoma13Rhabdo14Rhabdo15Rhabdo16

NRsarcoma"Rhahdo.rhahdomyosarcoma;hConcommitantchemotherapyon

days 1, 2, and 3 ofifosfamideonday 1 of ifosfamidetreatment;c

GFR corrected to 1.73 m2.Concomitantmedication''V,

AElopD20.

VV.AElopEtopEtopV.

AEtopD25D25EtopEtopEtopEtopEtopNR.

non-rhabdomyosarcoma.;Etop, etoposidc (2(XImg/m2) as aCoursestudied17538323221132221SexMMMFMMMMFFMFFMMM2-to 4-h infusion on days 1.treatment;

D25, doxoruhicin (25 mg/m2) asaA.actinomycin D (1.5 mg/m") asa bolus onAge(yr)4.26.710831.91243.917111i.431Bodv

surfacearea(m2)0.60.91.0l.n0.9OJ1.10.80.71.31.10.50.80.80.60.4Weight(kg)14.324.227.224.815.911.43(1.7m.o17.144.029.08.421.320.414.89.12.

and 3 of ifosfamide treatment; D2I1.doxoruhicin (2(1mg/m2);4-to 6-h infusion on days 1. 2.and 3 of ifosfamide treatment;V. vincristine ( 1GFRC(ml/min)80184104124164115128%IS')5512(111211717(i100155is

a 4- to 6-hinfusion.5mg/m") as abolusday

1 of ifosfamidetreatment.11

Patients received infusion for only 2 days.

3759

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

Germany) which had been preeluted with methanol and dried at 150°Cfor 10

min. An automated. Linomat IV, TLC sample applicator was used (CAMAG,Berne, Switzerland). Chromatography was performed in glass TLC tanks,saturated with solvent. The mobile phase was dichloromethane-dimethylfor-mamide-glacial acetic acid (90:8:1, v/v/v), which was allowed to rise to a

height of at least 9.5 cm. After the plates were dried, they were run again in asecond mobile phase of chloroform-methanol-glacial acetic acid (90:60:1) to a

height of 2 cm. The plates were dried again and sprayed for at least 10 s with5% NBP in acetone-0.2 M acetate buffer, pH 4.6 (8:2, v/v), dried, and thenrespraycd for 10 s. Plates were heated in an oven at 150°Cfor 10 min and left

to cool.Plates were dipped in 3% methanolic potassium hydroxide to reveal the blue

spots formed from alkylated NBP. The plates were photographed within 10 sof dipping because of the unstable nature of the chromophore. To ensure

uniform exposure and printing, a Kodak standard gray scale was photographedwith each plate. The negative was enlarged to the exact size of the originalplate. Uniform exposure was ensured by comparison of the Kodak gray scalewith the original.

The photographs of the plates were scanned with a CAMAG Scanner II

densitometer using the program CATS3 (CAMAG) to integrate the areas underthe chromatogram peaks. The peak areas for ifosfamide and metabolites weredivided by the area under the internal standard (cyclophosphamide) peak, andthe peak area ratio was used for calibration. Each plate contained samples andat least 6 tracks derived from spiked urine or plasma containing known concentrations of authentic standards (2-50 fig/ml). Calibration curves were ob

tained for ifosfamide and each of the metabolites and used to determine theconcentrations in patient urine and plasma samples.

A noncompartmental approach was used to estimate Cl from concentrationsof ifosfamide in plasma. A monoexponential equation was fitted to the postin-

fusion data to estimate /., and Vp for each subject. Exposure of each patient toifosfamide and each of its metabolites was expressed as the area under theAUC for that species. Recoveries of ifosfamide and metabolites in urine wereexpressed as a percentage of the administered dose. Both AUC and percentageof dose were corrected for molecular weight, and a dose-correction factor of 3/2

was used for those patients treated for only 2 days. Renal clearance of ifosfamide was determined from the product of Cl, the fraction of the doserecovered unchanged in the urine, and from amount excreted in each collectioninterval divided by the corresponding AUC.

Comparisons among patient groups were made using the Mann-Whitney U

test. Correlations of pharmacokinetic, metabolite, and patient variables wereanalyzed using normal linear regression or Spearmans rank correlation whereappropriate.

RESULTS

Ifosfamide and up to five of its metabolites could be detected inplasma and urine from all subjects. Typical plasma concentration andurine excretion profiles are shown in Figs. 2 and 3. The concentrationof ifosfamide in plasma reached an apparent steady state after 24 h butsubsequently declined throughout the administration period despite aconstant rate of infusion. This indicates that the clearance is increasingduring the infusion, and the Cl value calculated as the ratio of the doseto the total AUC is a measure of the average Cl during and after theinfusion period. Thus, the t^ value determined from the exponentialdecline of the postinfusion plasma concentrations reflects the higherCl after 3 days of drug administration. Similarly, Vp calculated fromthe average Cl and the reduced /,/. is likely to be an underestimate ofthe true volume of distribution. Plasma concentrations of parent drugdecreased rapidly after the end of the infusion, but measurable concentrations of IPM, CX, and dechloroethylated metabolites persistedbeyond the time when ifosfamide could be detected. In some patients,the concentration of dechloroethylated metabolites continued to increase long after ifosfamide concentrations reached a maximum andstarted to decline. Plasma concentrations of parent drug decreased inall patients, with a median decrease of 37% (range, 19-76%), while

concentrations of dechloroethylated metabolites increased by a median of 64% (range, -7-390%). This was not accompanied by a

significant increase in recovery of dechloroethylation metabolites inurine, although some patients did show an increase (Fig. 3), or by aconsistent change in plasma or urine concentrations of CX or IPM(data not shown).

Pharmacokinetic parameters for ifosfamide are given in Table 2. Itshould be remembered that these parameter values were estimatedsubject to the caveats resulting from time-dependent clearance. If not

1201 Patient 6IPM

Fig. 2. Concenlralion-time profile of ifosfamide

and its metabolites in plasma during and after a3-day continuous infusion.

72 84 96

Time3760

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

Patient 7

Fig. 3. Time profile of excretion of ifosfamideand its metabolites in urine during and after a 3-day

continuous infusion.

100

Table 2 Pharmaeoktnetic parameters for ifosfamide administered as a continuousinfusion over 3 days

Patient123456789HIII1213141516Means

±SDCl(liters/h/m2)3.275.805.332.964.234.903.193.395.925.593.193.606.438.187.507.195.04±1.66Vß(lilers/kg)0.322(1.62(10.4420.7600.7560.4010.4080.3870.590LOCH0.3760.4380.3130.7421.1030.7520.5KK±0.236Half-life(h)1.001.991.564.412.191.292.481.951.814.202.251.581.591.602.511.572.12±0.92Renal

clearance(litcrs/h/m2)0.6761.0380.5240.1830.5250.4821.5290.1400.7060.8230.663

±0.386

corrected for body size, clearance, Vp, and f,/, all increased withincreasing age, body surface area, and weight (P < 0.05). Therefore,Cl was corrected for body surface area, and Vßwas corrected forweight (28). Half-life and clearance increased in the same direction

compared to body size due to a greater increase in V¡¡and the difference between the time-averaged Cl calculated and the higher clearance when the half-life was estimated. There was no significant cor

relation of pharmacokinetic parameters with course, dose, GFR, orother patient parameters, including concomitant medication. Renalclearance was a relatively minor pathway for elimination of ifosfamide, represented by between 11.6 and 24.0% of GFR, and did notchange during the study period.

Fig. 4 is a stacked bar graph of plasma AUCs for ifosfamide and itsmetabolites in the 16 subjects studied. Exact values are given in Table3. Recoveries of ifosfamide and metabolites are shown in Fig. 5 andTable 4. Because of the young age of this patient group, reliable urinecollection was possible in only 10 of the 16 subjects. The major

metabolites in plasma were IPM or 3-DCI. KETO was detectable in

the plasma of only 3 patients and represented only a small percentageof the dose recovered in urine (median, 0.60% of dose). Carboxyi-

fosfamide concentrations in plasma varied greatly among individuals.In 5 subjects, this metabolite was undetectable in plasma (patients 4,7, 10, and 16) or present only at very low concentrations (patient 12).However, it was detectable in the urine in the 2 of these 5 from whomurine was available (patient 4, 1.95%; patient 7, 12.67% of dose).

There was no association between the AUCs of the metabolites andpatient variables, other than that boys had higher AUC values for theCX metabolite than did girls (P = 0.037). There were no correlations

among the AUCs of parent drug and the individual metabolites. In theurine, there was a positive correlation between the recovery of theactive metabolite IPM and the dechloroethylated products of the com-

25000

20000-

9 10 11 12 13 14 15 16

PatientFig. 4. Areas under concentration-time curves in plasma for ifosfamide and its metab

olites.

3761

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

Table 3 Areas under concentration-time curves (mM/h) of ifosfamide and metabolitesin plasma following continuous administration over 3 days

Patient123456789III"11"121314a16MedianRangeIfosfamide13.947.848.5315.357.169.2614.2610.175.996.1910.879.636.365.005.454.767.544.13-15.35IPM3.780.243.781.651.161.662.272.382.761.373.080.541.761.791.980.791.780.24-3.78CX0.130.860.86ND"0.670.97ND0.230.11ND0.300.010.580.900.31ND0.220-0.973-DCI2.631.580.842.172.944.084.312.311.396.752.931.941.622.843.251.972.240.84-^1.502-DCI0.671.420.381.650.901.700.301.650.571.700.820.390.190.190.190.300.560.19-1.70

" ND, not detectable.b AUCs for have been multiplied by 1.5 to standardize for 3 days of administration.

peting inaclivation reaction (P < 0.05). Recovery of IPM also correlated positively with albumin levels (P = 0.007). The recovery of

dechloroethylated metabolites increased with the course of administration (P < 0.05). These relationships observed ¡nurine were notmatched by corresponding correlations among AUCs and treatmentcourse. Urinary recoveries were not related to renal function as determined by GFR, although renal clearance of ifosfamide was proportional to GFR (r = 0.750, P = 0.013). There was no correlation

between AUC and urine recovery values individually for ifosfamideand each of the metabolites.

There was no correlation of measures of hematological or hepatictoxicity with AUCs of ifosfamide or of its metabolites. Hematologicaltoxicity, assessed by World Health Organization grading, was inversely related to GFR (P < 0.01 for thrombocytopenia, P < 0.05 forneutropenia). Patients experiencing more severe thrombocytopenia(World Health Organization grade 2 and above) had a lower GFR thanthose with no or mild toxicity (106 ±3 versus 151 ±32 ml/min/1.7m2, P = 0.013). Hematological toxicity was not independently cor

related with any other patient variable.

DISCUSSION

The complex metabolism of the oxazaphosphorines has led to intense interest in the influence of activation and inactivation pathwayson the efficacy and toxicity of these drugs and on the phenomenon oftumor resistance (11, 29-31). While host metabolism is probably the

most important determinant of tumor response, metabolism by thetumor itself may be important in some instances. Thus, it has beenproposed that variation in the initial activation of ifosfamide andcyclophosphamide may underlie interindividual variation in tumorresponse (11). The generation of chloroacetaldehyde has been associated with the neurotoxicity which sometimes accompanies administration of ifosfamide, but not cyclophosphamide (21). Cell lineswhich have been made to be resistant to oxazaphosphorines haveincreased ALDH activity (29, 31). Sensitivity can be restored byaddition of inhibitors of this enzyme (32-34), and selective tissue

toxicity of oxazaphosphorines has been associated with distribution ofALDH (30). The clinical implications of interindividual variation inthe metabolism of these drugs are far from clear. Likewise, the role oftumor-specific activation of these drugs has not been investigated.

Results of the present study indicate a wide degree of variability inthe pharmacokinetics and metabolism of ifosfamide in pediatrie patients. In addition to complications due to multiple drug therapy, the

study of the pharmacokinetics of oxazaphosphorines is confounded byinduction of their own metabolism (13, 35^0), which in the presentstudy produced up to 3-fold increase in clearance during 3 days of

administration. Thus, any comparison of the pharmacokinetics of cyclophosphamide or ifosfamide is conditional on the administrationregimen used and on previous treatment. The underlying cause of theapparent increase in clearance of ifosfamide during the infusion periodhas not been previously determined. There was no significant changein renal clearance during this period, and this route of eliminationplays only a minor role in total ifosfamide clearance. Increased elimination of ifosfamide by metabolism to dechloroethylated products didappear to occur. 3-Dechloroethylifosfamide accumulated in the

plasma of some patients, and, although there was no significantchange in the renal excretion of this metabolite during the infusion,recovery of dechloroethylated metabolites also tended to increase withincrease in prior ifosfamide treatment. The decline in plasma concentration of these metabolites postinfusion did not indicate eliminationrate-limited kinetics (41). Under time-independent formation kinetics,

concentrations of the metabolites should change in parallel with thoseof the parent drug. It appears that the decline in ifosfamide concentrations during continuous infusion may be partially explained by anincrease in metabolism to dechloroethylated products. Thus, the au-

toinduction of ifosfamide metabolism following continuous or repeated exposure may result in an increase in metabolism to inactivedechloroethylated metabolites, presumably accompanied by an increase in the formation of the toxic metabolite chloroacetaldehyde.

Although some individuals have been reported to be deficient in theexcretion of the carboxy metabolite of cyclophosphamide (22, 23), asimilar deficiency has not been found in studies with ifosfamide (12).In the present study, the variability in AUC of CX (0-0.97 mM/h) wasgreater than that for the other metabolites including IPM (0.24—3.78

min/h). In 5 subjects, CX was undetectable or present at negligibleconcentrations in plasma. This suggests that these individuals are notable to form this metabolite systematically, an observation similar tothat found in studies with the carboxy metabolite of cyclophosphamide in urine (22, 23). However, in 2 of these 5 patients, analysis ofurine samples showed that CX accounted for an appreciable fractionof the dose administered. It is possible that CX is formed fromaldoifosfamide in the kidney and excreted without returning to thesystemic circulation or is present in plasma at concentrations belowthe limit of detection of the assay. Thus, patients may be deficient inthe ALDH enzyme needed to inactivate the activated of oxazaphos-

phorine intermediates systemically but still possess ALDH activity in

1 2 3 4 7 8 9 11 13 14

PatientFig. 5. Total recovery of ifosfamide and its metabolites in urine.

3762

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

Table 4 Recoveries of drug and metabolites in urine(%Patient1234784II1314Mean±

SDIfosfamide2(1.7(117.91)9.846.1916.4414.2125.824.3810.9910.0713.6516.67IPM12.2(111.306.047.213.582.9110.512.737.308.037.1813.44CX10.907.4617.501.9512.6710.1210.115.157.455.358.8714.403-DCI29.2027.208.043.9413.687.146.141.946.337.3211.0919.52of

dose)2-DCI18.8019.305.192.604.134.4')2.901.364.895.466.9116.52KETO3.350.160.490.030.600.120.600.260.410.250.631

0.98Total95.1583.3247.1021.9251.1039.0056.11815.8237.3836.4848.33124.92

the kidney. A positive correlation of IPM recovery with albumin levelsis interesting, given a report that this plasma protein can catalyze therelease of phosphoramide mustard from 4-hydroxycyclophosphamide

(42). Although the ultimate alkylating species is thought to be IPM,it has been suggested that this is not active unless formed intracellu-larly and that circulating concentrations of 4-hydroxyifosfamide/

aldoifosfamide are the most likely predictors of pharmacological effect (24). Thus, measurement of IPM in plasma may not be the mostuseful predictor of pharmacological response.

This study was not designed to investigate systematically the relationship between therapeutic outcome and ifosfamide pharmacokinet-

ics and metabolism, which will be confounded by the effects of otherchemotherapeutic agents. The correlation of platelet nadir and GFRmay reflect the ability to eliminate toxic metabolites from the body orthe toxic effect of etoposide, which depends on renal function. Renalfunction has previously been shown to be an important risk factor forthe development of both central nervous system (43) and renal (9, 44)toxicity with ifosfamide.

Comparison of pharmacokinetic parameters derived from thepresent study and those previously reported is complicated by differences in patient populations and administration protocols. Very fewstudies have been performed on children. Ninane et al. (45) examinedthe pharmacokinetics of ifosfamide following the administration 3g/m2 on 2 consecutive days in a group of patients similar to ours.

However, the analytical method used in that paper is very questionable, i.e., alkylating activity was converted directly to ifosfamideconcentrations in plasma and urine. Boos et al. (19) examined theexcretion of the enantiomers of ifosfamide and its dechloroethylatedmetabolites in urine in a group of slightly older children. Their resultsare consistent with ours, 3-DCI representing the major dechloroethylated metabolite and the total recovery of unchanged drug, 2- and3-DCI being approximately 40%. Comparison with other reports of

the pharmacokinetics of ifosfamide in other patient groups is difficultbecause of the autoinduction seen after 3 days of continuous administration. Lind e/ al. (39) found that the half-life of ifosfamide in adultsranged from 2.4-5.0 h after 3 consecutive days of 1.5 g/m2/day

administered as a short infusion. These authors saw an increase in Clfrom a median of 69.2-94.0 ml/min during the first 3 days of a 5-day

regimen, with a relatively constant median Vßof 401. This comparesto a mean ±SD in the present study of 2.12 ±0.92 h, 84.0 ±21.1ml/min/rrr, and 0.588 ±0.236 liters/kg for half-life, clearance and

volume of distribution, respectively. Converting to an average adultbody size (1.7 m2 and 70 kg), ifosfamide Cl is greater in the group of

children studied here than in adults, while volume of distribution issimilar. Lind et al. also reported a median value for renal clearance ofifosfamide of 15.1 ml/min, which did not change with continuedtreatment. Renal clearance in the present study was 8.93 ±4.67ml/min and was determined almost entirely by GFR, not age or bodysize. Pearcey et al. (47) studied the pharmacokinetics of ifosfamide inadults receiving 2 g/m2 over 72 h or 1 g/m2 over 24 h. These authors

reported a clearance of 85 ±23 ml/min and a volume of distribution

of 25.6 ±13.1 liters. Again, correcting for body size, it would appearthat clearance is greater in our pediatrie study group.

Apart from the study by Boos et al. (19), there have been fewstudies of the recovery of ifosfamide and its metabolites in urine andnone of a pediatrie patient group. Allen et al. (13) reported recoveriesof unchanged drug of 20-50%, which is consistent with our data. Lind

et al. (12) reported much lower recoveries of the individual metabolites using a quantitative TLC method similar to that used in thepresent study. However, modifications to the Chromatographie technique have allowed us to separate and more reliably quantify both thedechloroethylated metabolites and IPM. Unlike Lind étal.,we saw noevidence for an increase in the excretion of dechloroethylated metabolites during continuous administration, although an increase inplasma concentrations of these metabolites occurred in some patients.This increase in dechloroethylated metabolites in plasma may explainin part the increase in ifosfamide elimination during continuous administration.

The ultimate fate of up to 70% of the dose of ifosfamide administered i.v. remains unknown. Metabolism to some species not detectedby the current techniques remains a possibility, but studies usingradiolabeled material and 3'P-nuclear magnetic resonance have failed

to account for a significantly greater fraction of the administered dose(11, 20). Other alternatives are conjugation of parent drug or reactivemetabolites to Mesna or endogenous species, such as glutathione,possibly followed by biliary excretion and removal in the feces. In astudy with radiolabeled cyclophosphamide, only 6.5% of the dose wasrecovered as total radioactivity in bile (46).

In conclusion, we have examined the pharmacokinetics and metabolism of ifosfamide in a pediatrie population. Although the dose andmode of administration was constant, there was marked interindivid-

ual variability in the pharmacokinetics of the parent drug, dependenton age and previous exposure to ifosfamide, even with a 3-week breakbetween courses. Metabolism also showed a wide degree of interin-

dividual variation, with some patients apparently unable to form theinactive carboxy metabolite systemically. Clearance of ¡fosfamideincreased during continuous administration, with an apparent increasein dechloroethylated metabolites. The clinical significance of this isnot yet known. The long-term influence of ifosfamide metabolism on

tumor response and toxicity in these patients continues to be monitored.

ACKNOWLEDGMENTS

We would like lo thank Professor A. W. Craft for his support.

REFERENCES

1. Brock, N., Pohl. J.. and Stekar. J. Studies on the uroloxicity of oxazaphosphorinecytostatics and its prevention. HI. Profile of action of sodium 2-mercaptocthanesulfonate (mcsna). Eur. J. Cancer Clin. Oncol., 18: 1377-1387, 1982.

2. De Kraker. J. Ifosfamide in paediatric oncology. Anticancer Drugs. 2: 339-341. 1991.3. Pinkerton. C. R., and Pritchard. J. A phase II study of ifosfamide in paediatric solid

tumours. Cancer Chemother. Pharmacol.. 24 (Suppl.): 13-15. 19S9.

3763

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

IFOSFAMIDE METABOLISM IN PEDIATRICS

4. Pratt. C. B., Douglass, E. C, and Etcuhanas, E. L. Ifosfamide in paedialric malignan!solid tumours. Cancer Chemolhcr. Pharmacol., 24 (Suppl.): 24-27, 1989.

5. Zalupski, M.. and Baker, L. H. Ifosfamide. J. Nati. Cancer Inst., 80: 556, 1988.6. Schoenike, S. E., and Dana, W. J. Ifosfamide and mesna. Clin. Pharmacol., 9:

179-191, 1990.7. Pratt, C. B., Green, A. A., Horowitz, M. E., Meyer, W. H., Etcubanas, E.. Douglass,

E., Hayes, F. A., Thompson, E., Wilimas, J., Igarashi, M., and Kovnar, E. Centralnervous system toxicity following the treatment of pediatrie patients with ifosfamide/mesna. J. Clin. Oncol., 4: 1253-1261, 1986.

8. Vcrdeguer, A., Castel, V., and Esquembre, C. Fatal encephalopathy with ifosfamide/mesna. Pedialr. Hematol. Oncol., 6: 383-385, 1989.

9. Burk, C. D.. Restaino, I., Kaplan, B. S., and Meadows, A. T. Ifosfamide-induced renaltubular dysfunction and rickets in children with Wilms tumor. J. Pediatr., 117: 331-

335, 1990.10. Skinner. R., Pearson, A. D. J., Price. L., Coulthard, M. G., and Craft, A. W. Neph-

rotoxicity after ¡fosfamide.Arch. Dis. Child., 65: 732-738, 1990.11. Sladek. N. Metabolism of oxazaphosphorines. Pharmacol. Ther., 37: 301-355, 1988.

12. Lind, M. J., Roberts, H. L., Thatcher, N., and Idle, J. R. The effect of route ofadministration and fractionation of dose on the metabolism of ifosfamide. CancerChemothcr. Pharmacol., 26: 105-111, 1990.

13. Allen. L. M.. Creaven. P. J., and Nelson, R. L. Studies on the human pharmacokineticsof isophophamide (NSC 109724). Cancer Treat. Rep.. 60: 451^*58, 1976.

14. Clarke, L., and Waxman, D. J. Oxidative metabolism of cyclophosphamide: identification of the hepatic monooxygenase catalysts of drug activation. Cancer Res., 49:2344-2350, 1989.

15. Manthcy, C. L., Landkamer. G. J., and Sladek. N. E. Identification of the mousealdehyde dehydrogenases important in aldophosphamide detoxification. Cancer Res.,50: 4991-5002, 1990.

16. Dockham, P. A., Lee, M-O.. and Sladek. N. E. Identification of human liver aldehydedehydrogenases that catalyze the oxidation of aldophosphamide and retinaldehyde.Biochem. Pharmacol., 43: 2453-2469, 1992.

17. Hipkens, J. H., Struck, R. F.. and Gurtoo. H. L. Role of aldehyde dehydrogenase inthe metabolism-dependent biological activity of cyclophosphamide. Cancer Res., 41:3571-3583, 1981.

18. Norpoth, K., Addicks, H. W., Witting, U.. Muller, G., and Raidt, H. Quantitativedetermination of cyclophosphamide, ifosfamide and trofosfamide and their stablemetabolites on TLC plates wilh the aid of 4-pyridine-aldehyde-2-benzothiazolyl-hydrazone (PBH). Arzneim. Forsch., 25: 1331-1336, 1975.

19. Boos. J., Welslau, U., Ritter. J.. Blaschke. G., and Schellong, G. Urinary excretion ofthe enantiomers of ifosfamide and its inactive metabolites in children. CancerChemother. Pharmacol.. 2«:455-460, 1991.

20. Martino, R., Crasnier. F.. Chouini-Lalanne, N.. Guard, V., Niemeyer, U., Forni, M. D..and Malet-Martino, M-C. A new approach to the study of ¡fosfamidemetabolism bythe analysis of human body fluids with "P nuclear magnetic resonance spectroscopy.J. Pharmacol. Exp. Then. 260: 1133-1144, 1992.

21. Goren, M. P., Wright, R. K., Pratt, C. B., and E. P. F. Dechloroethylation of ifosfamideand neurotoxicity. Lancet, 2: 1219-1220, 1986.

22. Hadidi, A-H. F. A., Coulter, C. E. A., and Idle, J. R. Phenotypically deficient urinaryelimination of carboxyphosphamide after cyclophosphamide administration to cancerpatients. Cancer Res., 48: 5167-5171, 1988.

23. Boddy, A. V., Furtun, Y., Sardas, S., Sardas. O., and Idle, J. R. Individual variation inthe activation and inactivation metabolic pathways of cyclophosphamide. J. Nati.Cancer Inst., 84: 1744-1748, 1992.

24. Sladek, N. E.. Doeden, D., Powers, J. F, and Krivit, W. Plasma concentrations of4-hydroxycyclophosphamide and phosphoramide mustard in patients repeatedlygiven high doses of cyclophosphamide in preparation for bone marrow transplantation. Cancer Treat. Rep., 68: 1247-1254, 1984.

25. Klein, H. O., Wickramanayake. P. D., Christian, E., and Coerper, C. Therapeuticeffects of single-push or fractionated injections or continuous infusion of oxazaphosphorines (cyclophosphamide. ifosfamide and Asta Z 7557). Cancer (Phila.), 54:1193-1203. 1984.

26. Brock. N. Oxazaphosphorinc cytostalics: past-present-future. Cancer Res., 49: 1-7,

1989.

27. Boddy, A. V., and Idle, J. R. Combined thin-layer chromatography-photography-

densitometry for the quantification of ifosfamide and its principal metabolites inurine, cerebrospinal fluid and plasma. J. Chromatog. Biomed. Appi., 575: 137-142,

1992.28. Lind, M. J., Margison, J. M., Cerny, T., Thatcher, N., and Wilkinson, P. M. Prolon

gation of ifosfamide elimination half-life in obese patients due to altered drug distribution. Cancer Chemother. Pharmacol., 25: 139-142, 1989.

29. Hilton, J. Role of aldehyde dehydrogenase in cyclophosphamide-resistant L12IOleukemia. Cancer Res., 44: 5156-5160, 1984.

30. Colvin, M., Russo, J. E., Hilton, J., Dulik, D. M., and Fenselau, C. Enzymaticmechanisms of resistance to alkylating agents in tumor cells and normal tissues. Adv.Enzyme Regul., 27: 211-221, 1988.

31. Parsons, P. G., Lean, J., Kable, E. P. W., Favier, D., Khoo, S. K., Hurst, T., Holmes.R. S., and Bellett, A. J. D. Relationships between resistance to cross-linking agents,

glutathione metabolism, aldehyde dehydrogenase isozymes and adenovirus replication in human tumour cell lines. Biochem. Pharmacol.. 40: 2641-2649, 1990.

32. Kohn, F. R., Lankamer, G. J.. Manthey, C. L., Ramsay, N. K. C., and Sladek, N. E.Effect of aldehyde dehydrogenase inhibitors on the ex vivo sensitivity of humanmultipotent, committed hematopoietic progenitor cells and malignant blood cells tooxazaphosphorines. Cancer Res., 47: 3180-3185, 1987.

33. Kohn, F. R., and Sladek, N. E. Effects of aldehyde dehydrogenase inhibitors on the exvivo sensitivity of murine late spleen colony-forming cells (day-12 CFU-S) and

hematopoietic repopulating cells to mafosfamide (ASTA Z 7557). Biochem. Pharmacol., 36: 2805-2811. 1987.

34. Sahovic, E. A.. Colvin, M., Hilton, J., and Ogawa, M. Role for aldehyde dehydrogenase in survival of progenitors for murine blast cell colonies after treatment with4-hydroperoxycyclophosphamide. Cancer Res., 48: 1223-1226, 1988.

35. Sladek, N. E.. Priest. J.. Doeden, D., Mirocha, J., Pathre, S., and Krivil, W. Plasmahalf-life and urinary excretion of cyclophosphamide in children. Cancer Treat. Rep..64: 1061-1066, 1980.

36. D'Inclaci, M., Bolis, G., Facchinetti. T.. Mangioni, C., Morasca, L., Morazzoni, P.,

and Salmona. M. Decreased half life of cyclophosphamidc in patients under continualtreatment. Eur. J. Cancer, 13: 7-10, 1979.

37. Graham, M. I., Shaw, I. C., Souhami, R. L.. Sidau. B.. Harper. P. G., and McLean, A.E. M. Decreased plasma half-life of cyclophosphamide during repeated high-doseadministration. Cancer Chemother. Pharmacol., 75: 192-193. 1983.

38. Moore, M. J.. Hardy, R. W. Thiessen, J. J., Soldin, S. J., and Erlichman, C. Rapiddevelopment of enhanced clearance after high-dose cyclophosphamide. Clin. Pharmacol. Ther., 44: 622-628, 1988.

39. Lind, M. J., Margison, J. M., Cemy, T, Thatcher, N., and Wilkinson, P. M. Comparative pharmacokinetics and alkylating activity of fractionated intravenous and oralifosfamide in patients with hronchogenic carcinoma. Cancer Res., 49: 753-757, 1989.

40. Lewis, L. D., Fitzgerald, D. L.. Mohan, P., Thatcher. N.. Harper, P. G., and Rogers, H.J. The pharmacokinetics of ifosfamide given as short and long intravenous infusionsin cancer patients. Br. J. Clin. Pharmacol., 31: 77-82, 1991.

41. Houston. J. B. Drug metabolite kinetics. Pharmacol. Ther., 51: 521-552, 1981.42. Kwon, C-H., Borch, R. F., Engel, J.. and Niemeyer. U. Activation mechanisms of

mafosfamide and the role of thiols in cyclophosphamidc metabolism. J. Med. Chcm.,30: 395-399, 1987.

43. Meanwell, C. A.. Blake. A. E., Kelly, K. A., Honigsberger, L., and Blackledge, G.Prediction of ifosfamide and mesna associated encephalopathy. Eur. J. Cancer Clin.Oncol., 22: 815-819, 1986.

44. Skinner, R., Sharkey, I. M.. Pearson, A. D. J., and Craft. A. W. Ifosfamide, Mesna andnephrotoxicity in children. J. Clin. Oncol., //: 173-190, 1993.

45. Ninane. J., Baurain. R., Kraker, J. D., Ferster, A., Trouet, A., and Cornu, G. Alkylalingactivity in serum, urine and CSF following high-dose ifosfamide in children. CancerChemother. Pharmacol.. 24 (Suppl.): S2-S6, 1989.

46. Shaw, I. C., and Graham. M. I. Excretion of cyclophosphamide metabolites in bile.Biochem. Pharmacol., 33: 2535-2538, 1984.

47. Pearcey. T., Calven, R., and Mehla, R. Disposition of ifosfamide in patients receiving¡fosfamideinfusion therapy for the treatment of cervical carcinoma. Cancer Chemother. Pharmacol.. 22: 353-355. 1988.

3764

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1993;53:3758-3764. Cancer Res   A. V. Boddy, S. M. Yule, R. Wyllie, et al.   as a Continuous Infusion in ChildrenPharmacokinetics and Metabolism of Ifosfamide Administered

  Updated version

  http://cancerres.aacrjournals.org/content/53/16/3758

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/53/16/3758To request permission to re-use all or part of this article, use this link

on June 17, 2018. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from