age-dependent tetrahydrothiophenium ion formation in …[cancer research 57. 5509-5516, december 15....

[CANCER RESEARCH 57. 5509-5516, December 15. 1997)

Age-dependent Tetrahydrothiophenium Ion Formation in Young Children andAdults Receiving High-Dose Busulfan1

John P. Gibbs, Georgia Murray, Linda Risler, Jenny Y. Chien, Raj Dev, and John T. Slattery2

Department of Pharmaceutics, University of Washington. Seattle. Washington 98195 Â¡J.P. C.. J. Y. C.. R. D., J. T. S.], and Fred Hutchinson Cancer Research Center, Seattle,Washington 98109 Â¡G.M.. L. R.. J. T. S.}

ABSTRACT

Busulfan, a bifunctional alkylating agent, is a mainstay of myeloabla-

tive preparative regimens before hematopoietic stem cell transplantation.The apparent oral clearance of busulfan expressed relative to body surface area is 2â€”3-foldhigher in children 1-4 years old than it is in adults.The first step in busulfan elimination is the formation of a tetrahydro-thiophenium ion (IMI') in a glutathione 5-transferase-catalyzed reac

tion. We present computer simulations that demonstrate that the ratio ofthe AUC of THT+ to that of busulfan over 6 h [(AUCTH^/AUCm,),,^] ishighly correlated (r2 = 0.805) with the determinants of THT+ formation

and is virtually independent of the determinants of its elimination(r2 = 0.0201). We compared (AUCTHT+/AUCBi,)0_^ determined in 14

children (0.5-4 years) to that of 11 adults (12-54 years) and found a1.5-fold elevation in the area ratio I/' = 0.0098) and a similarly significant

increase in busulfan apparent oral clearance expressed relative to bodysurface area (/' = 0.042). The only common explanation for the elevated

busulfan apparent oral clearance and (AUCTHT+/AUCBU)0_<, is an en

hanced ability of children to metabolize busulfan through glutathioneconjugation.

INTRODUCTION

High-dose busulfan with allogeneic bone marrow transplantation is

a potentially curative treatment for hematopoietic cancers and certaingenetic diseases (1). Busulfan is a bifunctional electrophilic cytotoxinthat acts putatively through DNA alkylation. In the bone marrowtransplantation setting, excessively high busulfan plasma concentrations are associated with hepatic veno-occlusive disease, whereas low

levels allow relapse of chronic myelogenous leukemia, and evenlower levels allow rejection of grafted marrow (2-5).

In pharmacokinetic studies, the CL/F3 of busulfan expressed rela

tive to BSA (in nr) is 2-3-fold higher in children (1-4 years old) than

it is in adults (4, 6, 7). CL/F is a pharmacokinetic term defined as theratio of clearance (a specific measure of efficiency of elimination):thefraction of the dose that transits from the site of administration (in thiscase, the gastrointestinal tract) to the systemic circulation (8). BecauseCL/F is a function of both drug absorption and metabolism, busulfanapparently is either absorbed less extensively in children or metabolized more efficiently by them. Currently, there is no approved i.v.formulation of busulfan available to distinguish between these possibilities.

The primary elimination pathway for busulfan involves GSH conjugation, resulting in the formation of THT+. The THT"1"species that

is formed directly on conjugation of busulfan with GSH is y-glutamyl-ÃŸ-(5-THT+)-alanyl-glycine, which has been identified in rat

Received 4/7/97; accepted 10/15/97.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 indicate this fact.

1Supported in part by NIH Grants CAI8029 and GM07750.2 To whom requests for reprints should be addressed, at Fred Hutchinson Cancer

Research Center, 1100 Fairview Avenue North, AB 122. P.O. Box 19024, Seattle,WA 98109-1024. Phone: (206) 667-3372; Fax: (206) 667-6115; E-mail: jts@u.

washington.edu.3 The abbreviations used are: CL/F, apparent oral clearance; GSH. glutathione; THT.

tetrahydrothiophene; THT*, tetrahydrothiophenium ion: GST, glutathione S-transferase;

GC-MS. gas chromatography-mass spectrometry; AUC, area under the plasma concentration-time curve: BSA, body surface area.

bile after i.v. infusion (9) and in the rat liver perfusion model (10).THT+ accounted for 38% of busulfan supplied to the liver after a 4-h

isolated rat liver perfusion experiment. Ethacrynic acid coadministra-tion reduced THT+ formation to 12% of the dose after 4 h, presum

ably by inhibiting GST or depleting GSH (10). In humans, thebusulfan-GSH conjugate has never been measured. However, THT

has been detected after base treatment of urine (11). Recently, wedemonstrated that busulfan conjugation with GSH is catalyzed byhuman liver cytosolic GST (12), and that GSTA1-1 is the predomi

nant isoform responsible for busulfan conjugation (13).We have now characterized the formation of THT * as reflected in

the plasma concentrations of the ion and parent drug after oralbusulfan administration to patients being prepared for hematopoieticstem cell transplantation. We hypothesized that if children have anenhanced ability to metabolize busulfan, then they should have elevated AUCTHT- relative to busulfan AUC (AUCBtl) in comparison to

adults. We present mass balance equations to show that the ratio ofAUCTHT7AUCBl, is independent of the fraction of the dose absorbedand computer simulations to support interpretation of these data whenthe AUCs are measured for 6 h after the first dose of busulfan(AUC,,.^). The computer simulations show that for these case,(AUCrHT'/AUCgi,),,^, yields selective insight into the ability toform THT"1".The data indicate that young children have an enhanced

ability to form the GSH conjugate of busulfan.

PATIENTS AND METHODS

Reagents and Chemicals. GSH. monobromobimane, THT. 1-bromopen-

tane, and busulfan were purchased from Sigma Chemical Co. (St. Louis. MO).GC-MS-grade hexane was purchased from Burdick and Jackson (Muskegon.

MI). All other solvents were analytical grade or higher.Patients. Blood for busulfan and THT+ analysis was obtained from 25

patients receiving busulfan in preparation for allogeneic bone marrow transplantation. All patients received a phenytoin loading dose of 5 mg/kg administered three times and a 300-mg daily maintenance dose for seizure prophylaxis. Patients received busulfan tablets at a dose of 1-2 mg/kg body weightwith the exception of three children who received a 0.5-mg/kg dose (Table 1)as required to achieve steady-state busulfan concentrations required for ther

apy. Because the pharmacokinetics of busulfan are independent of dose (14),differences in busulfan dose did not influence the results of the study. Foryoung children incapable of swallowing tablets, a suspension was administered, otherwise busulfan was administered as 2-mg tablets (Glaxo-Wellcome,

Research Triangle Park. NC). Plasma GSH concentration was determined in asecond group of patients in whom THT+ levels were not measured.

For measurement of busulfan and THT+ concentrations in plasma, serial

blood samples (2-3 ml) were collected in heparinized tubes after the admin

istration of the first dose of busulfan. For adult patients, samples were collected30, 60, 90, 120, 180. 240. 300, and 360 min after the dose. For children(age < 5 years), an additional sample at 15 min was taken to better characterizethe absorption phase, which is more rapid in children receiving the suspensionthan in adults receiving tablets. For patients at the Fred Hutchinson Cancer

Research Center, plasma was separated within l h of collection and storedat â€”¿�20Â°Cuntil analysis. For patients at other institutions, plasma samples were

frozen and shipped on dry ice to the Fred Hutchinson Cancer Research Centeron the day of collection and were kept frozen until analysis. All samples wereassayed within 24 h of blood collection. Blood samples drawn for GSHanalysis (Fred Hutchinson Cancer Research Center patients only) were ob-

5509

Research. on January 31, 2021. © 1997 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

AGE-DEPENDENT BUSULFAN-GSH CONJUGATION

Table I Characteristics of husulfan/THT^ patients

UPNChildren11300100020003(XX)400070001082281908218104071041292165(XX)7828Adults

andadolescentsUDII10371104501016011)101959597119824993399366000SexMMPFMMMMP1-MMMFMFMFMFFFMMM(yr)0.50.61.5223233334441217253436363845465354Weight

(kg)6.37.41212121517151117191721143456727270615972689595BSA(m2)0.290.370.540.500.570.610.680.600.450.680.710.640.780.57.18.65.92.88.82.67.63.75.742.022.13Busulfan

dosemg133.752761824181813.71781810143458727270605872608484mg/kg2.060.512.230.52.50.60.04.22.290.980.431.060.480.971.011.041.011.001.000.990.991.000.880.880.88mg/m244.210.150.011.931.639.326.430.330.425.211.328.312.924.728.835.237.538.338.535.935.641.134.541.639.4

tained on the schedule used for busulfan after the first busulfan dose and hourlyafter the busulfan dose when dosing intervals other than the first were evaluated. For GSH analysis, a portion of the blood collected was derivatized withmonobromobimane at the bedside and handled as described below.

Busulfan Analysis. Busulfan was extracted from plasma (0.2 ml)with ethyl acetate and derivatized with iodine at 70Â°C, and the resulting

di-iodobutane was assayed with reference to a deuterated di-iodobutane internal standard by GC-MS as described previously (4).

THT+ Analysis. Plasma (0.2-0.4 ml) was added to glass test tubes con

taining I(X) m.M (pH 7.4) potassium phosphate buffer with 1 mM EDTA tomake a final volume of I ml. MÃ©thylÃ¨nechloride (5.5 ml) was added, and themixture was sealed with a Teflon cap. shaken for 5 min. and centrifuged at1.(XX)X g for 5 min. An aliquot (0.7 ml) of the aqueous layer was transferredto a clean test tube, and hexane (0.20 ml). 1-bromopentane (10 ng). and l N

NaOH (0.10 ml) were added. The mixture was recapped, and after standing atroom temperature for 5 min. it was shaken for 5 min and centrifuged atI .(MX)x fl for an additional 5 min. After placing samples on dry ice, the hexanelayer was removed via plastic syringe to autosampler vials for GC-MS analysis

according to the previously published method (12). A standard curve was

prepared daily by spiking 0.7 ml of a 20:80 (v/v) plasma/phosphate buffermixture (which had been extracted with 5.5 ml of mÃ©thylÃ¨nechloride) withTHT (dissolved in acetonitrile) to make final concentrations of 25-250pmol/ml plasma. The intra- and inter-day coefficients of variation were 4.55%(n = 10) and 2.43% (n = 3), respectively, determined with pooled patientplasma (190 pmol/ml THT+ ). The half-life of the THT ' in plasma at -20Â°C

was 100 h. Because this assay is indirect and would detect large molecularweight thiophenium ions from spontaneous alkylation of proteins, mÃ©thylÃ¨nechloride-extracted plasma from patients with busulfan concentrations rangingfrom 352-871 ng/ml was subjected to ultrafiltration with a 10.000 Mr cutoff(Alltech Associates, Inc.. Deerfield. IL). The THT' concentration after ultra-

filtration was 97.1 Â±1.4% of the THT+ concentration before ultrafiltration.

(Ã¬SHAnalysis. Blood samples drawn for busulfan pharmacokinetics (therapeutic targeting) were split for plasma GSH analysis by the method ofCotgreave and Moldeus (15). Whole blood (200 /xl) was added to a lithium-

heparin microcentrifuge tube that contained 200 \Â¡\of monobromobimanederivatizing solution (15 mg of monobromobimane dissolved in 0.5 ml ofacetonitrile and 4.5 ml of 124 mM phosphate buffer (pH 8.0)1 within 1 min ofcollection. The derivatized sample was centrifuged at 2000 X ^ for 2 minwithin 15 min of collection. The plasma layer was transferred to a 1.5-ml

plastic conical tube, and the derivatization reaction was allowed to proceed at4Â°Cuntil complete. Duplicate aliquots ( 150 /Â¿Ieach) were transferred to 1.5-ml

plastic conical tubes, and 30 \i\ of 50% trichloroacetic acid were added. Thetubes were vortexed and centrifuged for 2 min. The supernatant was transferredto a clean 1.5-ml conical tube for storage at -20Â°C until high-performance

liquid chromatography analysis. The high-performance liquid chromatographysystem consisted of a Hewlett-Packard 1050 series system with a quaternary

pump, an autosampler. and a 1046A programmable fluorescence detector. Theexcitation wavelength was 394 nm. and the emission wavelength was 484 nm.Mobile phase A was 87.5 mM acetate buffer with 6% acetonitrile (v/v; pH 3.8),whereas mobile phase B was water with 40% acetonitrile (v/v). Initially, themobile phase was 97.2% A and 2.8% B. As soon as the sample was injected,B was linearly increased to 4.8% over the first 7.26 min and then abruptlyincreased to 100% B to wash the column for 3 min. at which time the mobilephase was returned to the initial condition, and the column was reequilibratedfor 11 min before the injection of the next sample. The flow rate was 0.8ml/min. Analytes were separated using a 4.6 mm x 10 -cm, 3-/nm Ranin

microsorb C18 column. The injection volume was 10 fil. and tne mn 'Â¡mewas21 min. A time-averaged GSH concentration was calculated for each dosing

interval by calculating the AUC and dividing by 6 h.Liver Size. A direct measurement of liver size was performed in the course

of autopsy of cadavers at Harborview Medical Center (Seattle, WA). Allspecimens were obtained from individuals who had died within 2 days of headtrauma. Organ weights were measured in 38 subjects between the ages of 1week and 4 years ( 1.91 Â±1.42 years) and 49 subjects between the ages of 14and 36 years (27 Â±6 years). Liver weight was expressed relative to either bodyweight (in kg) or BSA (in nr). BSA was calculated as BSA = log"'

[(0.425 x log weight (kg)) + (0.725 x log height (cm)) - 2.144].

Pharmacokinetic Analysis. Busulfan CL/F was determined by dividingbusulfan dose by AUCBU determined by fitting the concentration time curveafter the first dose of busulfan to a one-compartment model with first-order

absorption and elimination with the RSTRIP program (MicroMath. Salt LakeCity, UT). The trapezoidal rule was used to calculate the AUC,,.^ (AUC for0-6 h after the first busulfan dose) for busulfan and THT+.

Statistical Analysis. All statistical comparisons between children andadults were made with the unpaired Student's t test, assuming nonuniform

variance. Data are presented as the mean Â±SD of three or more determinations. In computer simulations. Pearson's correlation coefficients reported are

statistically significant even when weak (P < 0.001). due to the number of

simulated subjects.Mass Balance Relationships. It has previously been shown that the ratio

of metaboliteiparent drug AUC after i.v. administration is equal to CL,/CLm.in which CL, is the metabolite formation clearance, and CLm is the metaboliteelimination clearance (16). We have applied the same principles to the busulfan situation in which the drug is given p.o. and may be metabolized in boththe enterocyte and the liver. To consider a drug with the characteristics ofbusulfan and its THT' metabolite, we considered the elimination of THT+

into the gut either through bile or directly from the enterocyte without transitto the systemic circulation. FB is the fraction of THT+ formed in the enterocyte

or the hepatocyte that escapes to the systemic circulation. The metabolite:parent drug area ratio that was derived from the mass balance description of the

model is:

AUCm

AUCD

CL(1 -

CL,

- FGFH)\I

F0FH l\(A)

in which F0 is the fraction of the dose that exits the lumen of the intestinal tractand crosses the enterocyte intact. FH is the fraction of drug entering the liverthat escapes metabolism on a single pass through the liver, and CL is theclearance of the parent drug. Because blood flow to the absorptive surface ofthe small intestine is low relative to that of the liver, gut as a site of eliminationof busulfan that had reached the systemic circulation was assumed to benegligible.

Computer Simulations. Because therapeutic requirements precluded determination of AUCTnr* for more than 6 h after the first dose, computer

simulations were conducted to support the interpretation of the AUC ratiousing the programs Mathematica (Wolfram Research. Inc.. Champagne. IL)and Excel (Microsoft, Redmond. WA). (AUCm/AUCp)0_ and (AUC,/AUCm),,_Â«(AUC,,_Â« is AUC from 0 to infinite time after the first dose) werecalculated using the integrated form of the parent drug and metabolite concentration-time equations after oral administration:

5510



ACiE-DEPENDKNT BUSULFAN-GSH CONJUCÃŒATION

Table 2 Descriptive statistics Â¡orthe Â¡ihunmu-okinelic pitnimeier vaines (n = 685)

used in lite computer simulations Â»/(AUC,,/AUCp)l>_^>

ParameterFGFHFBka,

minV(ml/kg)CL(ml/mm/kg)CL/F(ml/min/kg)fmVm

(ml/kg)CLm(ml/min/kg)CLf

(ml/min/kg)Mean0.9050.8950.4990.1106563.003.750.7546541.622.26SD0.0580.0330.1290.0271660.761.070.1461580.410.71CV

(%)6.413.6925.924.525.325.328.519.424.225.331.4

VIK.-1-(B)

vâ€ž,ÃŒ\ v-((-CUV li)

CL CL\v/VKj~VJ

Ka( l - FG)

(V,â€ž)K. -CLâ€ž

RESULTS

Plasma busulfan and THT+ concentrations over the first dosing

interval in a representative child (UPN 11300) and adult (UPN 10450)are shown in Fig. I. In both adults and children. THT ' concentrations

rose over the 6-h sampling period to a broad peak from which a

decline was not evident during the dosing interval. A second dose ofbusulfan administered at 6 h as required for therapy prevented estimation of an elimination rate constant for THT ' from the decline ofTHT ' concentration and, therefore. AUC0_,V of the metabolite.

To estimate the half-life of THTf for computer simulations conducted to assess the utility of AUC()_^,, the accumulation of THT+ tosteady state was evaluated by assaying THT ' concentration over the

6-h dosing interval after doses 1, 5, and 9 of a 16-dose regimen given

over 4 days (every 6 h). These measurements were carried out in twoadult patients with a constant busulfan dose. Mean (AUCTHTÂ»)0_^was 27.2. 104. and 102 pmol x min/ml after the first, fifth, and ninthdoses of busulfan. There was a 3.8-fold increase in (AUCTHTÂ»)0_^

between the first and fifth dose of busulfan (P < 0.001 ). and there wasno change between the fifth and ninth dose (P = 0.31). Thus, THT'

attained steady-state levels after the first 24 h of busulfan adminis

tration, from which it was evident that the apparent eliminationhalf-life for THTf is ~5 h.

The results of the computer simulations in which plasma concentration of parent drug and metabolite were calculated according to thetime-integrated forms of equations B and C are shown in Fig. 2 for

AUC,,_.,. The numerical identity of the area ratio (AUCnl/AUCp)0_400and the determinants of the ratio as given on the right side of equationA, FB(CLf/Clm + (CL/CLm)(l - FciFH)/FciFH), validates the simu-

KaFG(l - FH)FBDpo((e"-CL"/v-") - '-"

CLâ€ž

Vm

(C)

The ratio of AUCI11/AUC|, over both time intervals was evaluated by randomlyconstructing 685 individuals from Gaussian distributions of pharmacokineticparameter values corresponding to measured or estimated values for busulfanand THT^ using the random number generator in Excel. The AUCs of parent

and metabolite were calculated using the time integral of the concentration-

time equations (equations B and C) as supplied by Mathematica and implemented as a spread sheet on Excel. Mean and SD of parameter values used aregiven in Table 2. Parameters for which values were not available in theliterature were estimated from available data (17-19) or from the characteris

tics of the compound. FG and FH have not been measured directly, but anapparent hepatic extraction ratio for children with a CL/F of 176 ml/min/nr(the mean value for children in this study) is 20%. Because elimination cantake place in both liver and gut. the loss of the drug was divided between thetwo sites. FH was calculated as ( 1 - CL,/QH) in which QH is the hepatic blood

flow (21.4 ml/min/kg). FB was estimated to be 50%. based on the isolated ratliver perfusion data (10). The mean traction of a dose of busulfan convened toTHT4 (the only known metabolic route for elimination of busulfan; fm) was

taken to be 0.75, based on the excretion of radioactivity in urine after theadministration of labeled busulfan (18. 19) and the observation that busulfanreacts relatively slowly with sulfhydryls in relation to its enzyme-catalyzedclearance (12. 14). Busulfan lost to the formation of THT+ most likely

spontaneously alkylates endogenous molecules. The volume of distribution ofTHT+ (Vm) was estimated to be equal to that of busulfan. because it isexpected to accumulate where formed. The clearance of THT4 (CLm) wasestimated from Vm, and the apparent half-life of THT4 was estimated from theaccumulation of THT4 between doses. Because the AUC,,,/AUCp ratio is

independent of the fraction of the dose of parent drug that is absorbed from thelumen of the gastrointestinal tract under any conditions in which a metaboliteis formed only after the drug is absorbed (the value is the same for parent drugand metabolite and cancels when the ratio of equations B and C is takenl. thefraction of the dose absorbed was set to 1.

10-,

1-

0.1-

*- O.OlJ03

0OCou

1-

0.1-

0.01-1

O

Busulfan

THT4

100 200 300 400

Busulfan

THT4

100 200 300 400

Time, minFig. 1. Representative time course of plasma busulfan and THT' concentrations in a

child (upper panel; UPN I13WI and an adult (Itm-rr panel: UPN 104501.

5511



ACÃŒh-DKPENDHNTBUSULFAN-C.SH CONJUGATION

Fig. 2. The relationship between (AUC,,,/AUCp)0-_Â«and selected parameters from equationsA-C as determined by computer simulation.

4-1

|J< 2-

IÂ«f'1

r* = 0.999

1/CLm

2CC Â¿coeco eco ieco i2co

o 00

r* * 0.287

246

CUF, mUmln/Vg

0.339

2 3O.i, rri/ninAg

lations. The lack of correlation with either V or Vm also validates thesimulations, because (AUCp),,^^ and (AUCm)0^K are theoreticallyindependent of volumes of distribution. Because (AUCm/AUCp)0_^ = CL,yCLm for a drug administered i.v. and eliminated

only in the liver, the correlation between (AUCm/AUCp)0_^ wasevaluated, and r was found to be 0.692. This result is expected,because in the busulfan-THT ' model. CL, does not completely ac

count for the formation of the metabolite (the terms FH and FGdescribe loss of drug to formation of metabolite on first pass). CL/Fand CLf were weakly related to (AUCm/AUCp)0^, with r2 values of

0.287 and 0.339, respectively. To evaluate the correlation between allparameters that account for formation of the busulfan-GSH conjugateand its elimination as they appear in the model,4 correlations between

(AUCJAUCpJo^ and FB(CL, + CL(1 - FGFH)/F0FH) and l/CLm,

respectively, were evaluated and found to be 0.644 and 0.303.The analogous simulations for (AUCm/AUCp)0_^, are shown in

Fig. 3. In this case, the relationships between the area ratio and the

4 Because the parameters that determine the value of AUC appear in the model in a

specific mathematical construct, and because the correlation analysis tests for a linearrelationship between the * and v variables, it is only valid lo evaluate the correlations ofthe AUC ratio with its determinants included in the form that the variables determine thevalue of AUC.

determinants of the formation of metabolite on the right side ofequation A are no longer numerically identical (i2 = 0.655), because

the areas are incomplete. When the determinants of the formation andthe elimination of the busulfan-GSH conjugate were evaluated asdeterminants of (AUCm/AUCp)()j6, FB(CL, + CL(1 - Pa^H^oFÂ»)was found to be a very strong determinant (r = 0.805), whereasl/CLm was found to be weak (r = 0.0201 ). This result indicates that

(AUCm/AUCp),)_^ is a function of the formation of the metaboliteand has little dependence on the parameter describing its elimination,CLm. Fig. 3 also shows that the area ratio is weakly related to Vm(r = 0.127) but is essentially independent of V (r = 0.006). This

result is expected, because in theory, AUC,,^^ is independent ofvolume of distribution, and (AUCp)0_Ki approximates (AUCp)0_Â«â€ž,whereas (AUCm)0_>6 is only a fraction of (AUCJ0_â€ž. CL,/CLm,CL/F, and CLf are all relatively weak determinants of (AUCm/AUCpJo^, with r of 0.311, 0.363, and 0.396, respectively.

The ratios of (AUCTHTÂ»/AUCB1J)0^6for children and adults areshown in Table 3. (AUCj.^^!.^,^,^, was 1.5-fold higher inchildren (P = 0.0098). Busulfan CL/F declined with age in a mannerconsistent with previous reports. Children had a busulfan CL/F 2.3-

fold higher than that of adults when CL/F was expressed relative tobody weight (P = 0.0012) and 1.5-fold higher when expressed rela-

5512



AGE-DEPENDENT BUSULFAN-GSH CONJUGATION

Fig. 3. The relationship between (AUCm/AUC,,),,.^ and selected parameters from equationsA-C as determined by computer simulation.

2.0

Â§0.5

0.0

2.0 n

i LO

0.5 -I

0.0

2.0-

i:l0-5:

c.o-

r' = 0.655

CL,

I3 o 0.805

r* = 0.0201

J^cpffo

1/CLâ€ž

2CO Â¿CO 600 eco l CCO 12CO

Vm.

0.0

r* = 0.311

r1= 0.363

246

CUT, ml_/mln/kg

o l* = 0.396

*0.0005

2CO Â¿co eco eco icco 1200V. mLAg

live to BSA (P = 0.0423). Although busulfan CL/F was significantlyelevated in children, there was no difference in the terminal half-life

of busulfan elimination between children and adults. Correlationsbetween (AUCTHT-/AUCBU),,_X, and busulfan CL/F expressed relative to body weight or BSA were observed (r = 0.738 and P < 0.001and f = 0.711 and P < 0.001. respectively; Fig. 4). Excluding the

one outlier in these relationships, a child with a busulfan CL/F of 18ml/min/kg (430 ml/min/rrr) had no effect on the estimated slope andintercept; however, the r values were 0.480 (P < 0.001) and 0.405(P < 0.001). These results suggest that children are more proficient in

forming the GSH conjugate of busulfan than are adults.To gain some insight into the apparent difference between children

Table 3 Busulfan disposition in children timi adults after a single oral dose

GroupChildrenMeanSDAge(yr)31CL/F(ml/min/kg)7.323.71CL/F(ml/min/m2)17689'1/2(min)11824.5(AUCTHT+/u0.06310.0237Adults

andadolescentsMeanSDP"The(AUCTHT-36143.180.750.0012120270.0423)oâ€”

Â»6divided by (AUCB(j)<)_^) after the14348.00.1380.04210.01200.0098firstdose of busulfan.

and adults in the ability to form GSH conjugates, we measured plasmatime-averaged GSH levels in young children and in older children andadults. Serial samples (n = 6) were obtained from patients in the 6-h

period after doses 1, 5, 9, and 13 of busulfan. A plot of time-averaged

plasma GSH concentration versus age showed no relationship (Fig. 5,upper panel). The mean Â±SD GSH concentration was 3.20 Â±1.23/UMin children <4 years old and 4.19 Â±2.06 /UMin older children andadults (P = 0.27). In addition, we examined time-averaged plasma

GSH concentrations in 3 patients <4 years old and 5 patients >4years old after doses 1. 5. 9, and 13 of busulfan (Fig. 5, lower panel).Time-averaged plasma GSH concentrations did not change over the

course of busulfan therapy.Because the GSH conjugate of busulfan is formed most actively by

GSTA1-1. which is the predominant GST in liver, we examined the

relationship between liver size and body size in children and adults atautopsy (Fig. 6). Liver weight expressed relative to body weight wasapproximately 1.7-fold higher in children ages 1 week to 4 years

(0.0345 Â±0.0083 g/kg) than it was in older children and adults(0.0205 Â±0.0050 g/kg; P < 0.001). The ratio of liver weight ex

pressed relative to BSA was not different between young children(0.792 Â±0.222 g/nr) and adolescents and adults (0.851 Â±0.208g/m2; P = 0.21).

5513



AGE-DEPENDENT BUSULFAN-OSH CONJUGATION

SmOu

U

0.14-,

0.12-

0.10-

0.08-

0.06-

0.04-

0.02-

0.00O

mL/mln/kg

10 15 20

0.14-,0.12-0.10-0.08-0.06-0.04-0.02-0.00-/"x/XDmL/mln/m2

100 200 300 400 500

others using indirect measures of liver size or small numbers atautopsy (23-25), whereas the livenBSA ratio is not different between

children and adults. Thus, if children and adults had the same intrinsicclearance per gram of liver weight, children would seem to have ahigher clearance when clearance was expressed per kilogram of bodyweight, but the difference would vanish when clearance was expressed relative to BSA. The same would be true for liver blood flow,assuming it is proportional to liver mass. The observed age-dependent

clearances for theophylline, caffeine, carbamazepine, and valproicacid seem to reflect liver size to body weight differences rather thandifferences in intrinsic clearance per gram of liver weight. This doesnot seem to be the case with busulfan.

There are no examples of drugs that have age dependence in theextent of the dose absorbed, although there are age-dependent changes

in the physiology of the gastrointestinal tract (20). Changes in gastricpH and gastric acid secretion occur during development. Lowering ofthe gastric pH and increased gastric acid secretion would not alterbusulfan absorption, because busulfan is very lipophilic and has noionizable groups. The motility of the gut and the gastric emptying timeare diminished in newborns; however, these changes tend to alter thetime course of drug absorption rather than the extent of absorption.

Busulfan seems to be quite different from other drugs with regardto age-dependent clearance. Previous pharmacokinetic studies inyoung children (1-4 years old) have demonstrated an elevated busul

fan CL/F relative to adults whether busulfan CL/F is expressed

Busulfan CL/FFig. 4. The relationship between (AUCTHT-/AUCBl.)(,_<, and busulfan CL/F in pa

tients undergoing stem cell transplantation. The stippled regression line is fit to all of thedata, whereas the soliti lint1 excludes the outlier, a child with a busulfan CL/F of 18ml/min/kg or 430 ml/min/nv. Upper panel, clearance relative to body weight; r2 ~ 0.738with the outlier, and r = 0.480 without the outlier. P < 0.001. Lower panel, clearancerelative to BSA; r = 0.711 with the outlier, and r = 0.405 without the outlier.P < 0.001.

DISCUSSION

We have shown that: (a) children <4 years of age have an elevatedratio of (AUCTHT-/AUCBU)0_^, in comparison to older children and

adults; (b) the increase in this area ratio indicates that children are ableto form the busulfan-GSH conjugate more efficiently than adults; (c)

children do not have elevated GSH levels in plasma in comparison toolder children and adults: and (d) the expression of CL/F relative toBSA appropriately compensates for differences in liver size relative tobody size when making comparisons of clearance across age. Theonly common factor between the increase (AUCTHTÂ»/AUCBU)0_6and the increase in CL/F relative to surface area is an enhanced abilityto form the GSH conjugate of busulfan in young children.

Age dependencies in drug clearance have been reported for anumber of drugs that are primarily eliminated by hepatic metabolismincluding theophylline, caffeine, carbamazepine, and valproic acid(20). In all of these examples, clearance expressed relative to bodyweight is about 2-fold higher when comparing the youngest children(1-3 years old) to adolescents (>15 years old). Generally, a steady

decline in clearance is reported until adult clearance values are attained around puberty. Because these drugs have a low hepatic extraction ratio, clearance is a function of protein binding and intrinsicclearance (according to the well-stirred model of hepatic drug clear

ance). For both valproic acid and carbamazepine. the unbound drugconcentration does not vary with age (21, 22). Intrinsic clearance is ameasure of drug-metabolizing enzyme activity. We found that theliverbody weight ratio in young children (0-4 years old) is 1.7-fold

greater than that of adults, which agrees with conclusions made by

10-,

ti

4 -

2-

TI

TAA

TL

TiTT

0COEIOÂ«o

10-0

10 20 30 4050Age,

years60

8-

6-

4 -

2-

0-1

Busulfan

8

dose

12 16

numberFig. 5. Upper panel, relationship between plasma lime-averaged GSH levels and age

in patients administered busulfan (l mg/kg) p.o. Points and error bars represent themean Â±SD for 3 or 4 determinations measured after doses I. 5. 9. and 13 of busulfan.Only the mean is shown for those individuals in whom only one or two time-averagedplasma GSH determinations were made. Lower panel, plasma time-averaged GSH levelsplotted as a function of busulfan dose number. Busulfan was given every 6 h for a totalof 16 doses. Plasma time-averaged GSH levels were measured after doses 1, 5. 9. and 13of busulfan. /I. children <4 years old; *. older children and adults.

5514



AGE-DUPI-NDI-NT Bl'SlÃ®LFAN-CKSHCONJUGATION

u.us-0.06-sCOss

0> 0.04-â€¢¿�o Â¿eO â€”¿�â€¢¿�CDÂ°>2

0.02-S81

0JVvEv-

^V^W^J7^V vv V yVVn

1n >ii in .1

2 -

1.5-

0.5-

o-i10 20 30 40

bolic factors and displays age dependence similar to busulfan CL/Fwhen expressed relative to BSA. Furthermore, the only commonfactor between CL/F and (AUCTHT*/AUCBl,)()_^) is the formation ofthe busulfan-GSH conjugate. Taken together, these data strongly

suggest that the elevated busulfan CL/F in children is a consequenceof enhanced ability to form the busulfan-GSH conjugate.

To further investigate the mechanistic basis for enhanced busulfanmetabolism in children, we measured plasma time-averaged GSHlevels in children and adults receiving high-dose busulfan. The plasma

GSH values we report are comparable to those observed by others (26,27). In rats, plasma GSH concentrations were reflective of hepaticGSH depletion after oxidative stress induced by acetaminophen metabolism (28). There was not a significant difference in plasma GSHlevels comparing young children and adults. In addition, there havebeen no reports of age dependency in GSH levels in young children.It seems unlikely that the apparently enhanced ability of youngchildren to form the busulfan-GSH conjugate is due to enhanced

availability of GSH.In conclusion, our findings indicate that young children are more

efficient metabolizers of busulfan than adults. Because busulfan elimination is solely mediated by GST, it seems likely that children havea higher level of GST expression than do adults. The eliminationhalf-life of busulfan does not seem to be different between childrenand adults, which implicates the gut in a first-pass effect. GST-a is

expressed to an appreciable extent in the small intestinal mucosa (29)and could participate in first-pass busulfan metabolism. Future studies

must focus on the potential metabolic differences in tissues collectedfrom children and adults.

Age, yearsFig. 6. The relationship between age and liver weight expressed relative to body weight

(upper panel) and BSA (lower panel).

relative to body weight or BSA, an observation confirmed in thisstudy. A single study has examined busulfan bioavailability after oraland i.v. dosing in children and adults and found no difference (17).However, in this study, no elevation in busulfan CL/F was observedin the children (5.2 Â±2.1 and 3.6 Â±1.3 ml/min/kg, children versusadults; P > 0.10), which was probably due to the age of the childrenstudied. Only five of eight children were less than 4 years old; theoldest was 6 years old.

In the present study, we have measured plasma THT+ levels inchildren and adults after the first dose of busulfan.5 We were not able

to completely characterize AUCTHT* due to the short busulfan dosinginterval. Using computer simulations, we found that (AUCTHT*/

AUCBU)0^6 is strongly determined by parameters describing theformation of the busulfan-GSH conjugate and is essentially independ

ent of CLm, which describes the elimination of the conjugate. Thisresult is consistent with intuition in that (AUCTHT-/AUCBU)0_^ is

comprised of a nearly complete (AUCBLJ)0_â€žbut contains only afraction of the (AUCTHT+)o^Å“, because the half-life of THT+ is

considerably longer than that of busulfan.We found that young children have a (AUCTHT-/AUCBU),,_^, that

is 50% higher than that of adults. The magnitude of this elevationcorresponds to the increase in busulfan CL/F observed in youngchildren. It is important to note that (AUCTHT-/AUCBU)0_^, is not a

function of the fraction of the dose absorbed. As indicated by thesimulations, (AUCTHT*/AUCBU)0^6 is solely dependent on meta-

REFERENCES

1.

io.

14.5 An indirect assay was used to measure THT* that did not distinguish the busulfan-

GSH conjugate from other possible busulfan-derived thiophenium ions such as busulfan- 15.cysteine or busulfan-W-acetylcysteine. neither of which has been reported to be present in

plasma.

5515

Forman, S. J., Blume. K. G., and Thomas. E. D. Bone Marrow Transplantation.Cambridge. MA: Blackwell Scientific Publications. 1994.Grochow. I.. B. Busulfan disposition: the role of therapeutic drug monitoring in bonemarrow transplantation induction regimens. Semin. Oncol., 20: 18-25. 1993.

Slaltery. J. T.. Clift. R. A.. Buckner. C. D.. Radich. J.. Slorer. B.. Bensinger, W. I.,Soil. E., Anasetti. C.. Bowden, R., Bryant, E., Chauncey, T., Deeg. H. J.. Doney.K. C., Rowers. M., Gooley, T., Hansen. J. A.. Martin, P. J.. MacDonald. G. B., Nash,R.. Petersdorf, E. W., Sanders. J. E.. Schoch, G.. Stewart. P.. Storb, R.. Sullivan.K. M., Thomas, E. D.. Witherspoon. R. P.. and Appelbaum. F. R. Marrow transplantation for chronic myeloid leukemia: the influence ol plasma busulfan levels on theoutcome of transplantation. Blood. 89: 3055-3060. 1997.Slattery, J. T., Sanders, J. E.. Buckner. C. D.. Schafler. R. 1... Lambert. K. W.. Langer.F. P.. Anasetti. C.. Bensinger. W. I.. Fisher. L. D.. Appelbaum. F. R.. and Hansen.J. A. Graft rejection and toxicily following bone marrow transplantation in relation tobusulfan pharmacokinetics. Bone Marrow Transplant.. 16: 31-42. 1995.

Dix. S. P.. Wingard. J. R.. Mullins, R. E.. Jerkunica, I.. Davidson. T. G.. Gilmore.C. E., York, R. C., Lin, L. S., Devine, S. M., Geller. R. B.. Heffner. L. T., Hillyer.C. D.. Holland. H. K.. Winton. E. F.. and Sarai. R. Association of busulfan area underthe curve with veno-occlusive disease following bone marrow transplantation. BoneMarrow Transplant.. 17: 225-230, 1996.Grochow. L. B., Krivil. W., Whitely. C. B.. and Blazar. B. Busulfan disposition inchildren. Blood, 75: 1723-1727. 1990.Regazzi. M. B., Locateli!. L.. Buggia. !.. Bonetti. F.. Zecca. M.. Pregnolato. M., andQuaglini. S. Disposition of high-dose busulfan in pediatrie patients undergoing bonemarrow transplantation. Clin. Pharmacol. Ther., 54: 45-52, 1993.

Gibaldi. M., and Ferner. D. Pharmacokinetics. New York: Marcel Dekker. Inc.. 1982.Marchand. D. H.. Remmel, R. P.. and Abdel-Monem. M. M. Biliary excretion of uglutalhione conjugate of busulfan and 1.4-dibromobutane in the rat. Drug Metah.Dispos.. 16: 85-92 1988.Hassan. M., and Ehrsson. H. Metabolism of l4C-busulfan in isolated perfused rat

liver. Eur. J. Drug Metab. Pharmacokinet., 12: 71-76. 1987.Hassan. M.. Oberg. G.. Ehrsson. H.. Ehmebo, M.. Wallin. I.. Smedmyr. B.. Toller-

man, T., Eksborg, S., and Simonsson. B. Pharmacokinetic and metabolic studies ofhigh-dose busulfan in adults. Eur. J. Clin. Pharmacol.. 36: 525-530, 1989.Gibbs. J. P., Czerwinski. M., and Slattery, J. T. Busulfan-glutathione conjugationcatalyzed by human liver cytosolic glutathione 5-transferases. Cancer Res., 56:3678-3681. 1996.Czerwinski. M.. Gibbs. J. P., and Slattery, J. T. Busulfan conjugation by glutalhioneS-transferases a. n, and IT. Drug Metab. Dispos.. 24: 1015-1019, 1996.

Ehrsson. H.. Hassan, M.. Ehrnebo. M.. and Beran. M. Busulfan kinetics. Clin.Pharmacol. Ther., 34: 86-89. 1983.Cotgreave, I. A., and Moldeus. P. Methodologies for the application of monobromo-

bimane to the simultaneous analysis of soluble and protein thiol components ofbiological systems. J. Biochem. Biophys. Methods. 13: 231-249. 1986.



A<;iDI'ITM>iM m sru AS-(isHCONJIÃœAIKIN

16. Lane, E. A., and Levy. R. H. Prediction of steady-state behavior of metabolite fromdosing parent drug. J. Pharm. Sci.. 69: 610-612. 1980.

17. Hassan. M.. Ljungman. P.. Ringden. O.. Syruckova. Z.. Bekassy. A., Stary. J.. Wallin.I., and Kallberg, N. Busulfan bioavailability. Blood. K2: 2144-2150. 1995.

18. \.ikk.ii m. M. V.. Trams. E. G.. and Smith. P. K. Preliminary studies on the distributionand fate of TKM, TEPA. and myleran in the human. Cancer Res.. IV: 713-718. 1959.

19. Vodopick. H.. Hamilton. H. E.. Jackson, H. L.. Peng. C. T.. and Sheets, R. F.Metabolic fate of tritialed busulfan in man. J. Lab. Clin. Med.. 73: 266-276. 1969.

20. Rane. A. Drug disposition and action in infants and children, in: Pediatrie Pharmacology. Therapeutic Principles in Practice. Philadelphia: W. B. Saunders Co.. 1992.

21. Cloyd. J. C.. Fischer. J. H.. Kriel. R. L.. and Kraus. D. M. Valproic acid pharmaco-kinetics in children. IV. Effects of age and antiepileptic drugs on protein binding andintrinsic clearance. Clin. Pharmacol. Ther.. 53: 22-29, 1993.

22. Kodama, Y., Tsulsumi. K.. Kuranari. M.. Kodama. H.. Fujii, !.. and Takeyama. M. In vÃ¬vobinding characteristics of carbama/epine and carbamazepine-10.11-epoxide to serumproteins in paediatric patients with epilepsy. Eur. J. Clin. Pharmacol., 44: 291-293, 1993.

23. Grygiel. J. J., Ward. H.. Ogborne. M., Goldin, A., and Birkett. D. J. Relationshipsbetween plasma theophylline clearance, liver volume, and body weight in childrenand adults. Eur. J. Clin. Pharmacol.. 24: 529-532. 1983.

24. Murray. D. J.. Crom. W. R., Reddick. W. E.. Bhargava. R.. and Evans, W. E. Livervolume as a determinant of drug clearance in children and adolescents. Drug Metab.Dispos., 23: 1110-1116. 1995.

25. Urata. K., Kawasaki. S.. Matsunami. H.. Hashikura. Y.. Ikegami. T.. Ishizone. S..Momose. Y.. Komiyama. A., and Makuuchi. M. Calculation of child and adultstandard liver volume for liver transplantation. Hepatology. 21: 1317-1321. 1995.

26. Michelel. F.. Gueguen. R.. Leroy. P.. Wellman. M.. Nicolas. A., and Siest. G. Bloodand plasma glutathione measured in healthy subjects by HPLC: relation to sex, aging,biological variables, and life habils. Clin. Chem., 41: 1509-1517, 1995.

27. Loguercio, C., Taranto, D.. Vitale. L. M.. Beneduce, F.. and Del Vecchio Bianco. C.Effect of liver cirrhosis and age on the glutathione concentration in the plasma,erythrocytes. and gastric mucosa of man. Free Radical Biol. Med.. 20: 483-488.

1996.28. Adams. J. D.. Lauterburg. B. H., and Mitchell, J. R. Plasma glutathione and gluta

thione disulfide in the rat: regulation and response to oxidative stress. J. Pharmacol.Exp. Ther., 227: 749-754. 1983.

29. Peters. W. H. M.. Roelofs. H. M. J.. Nagengast. F. M., and Van Tongeren. J. M.Human intestinal glutathione S-transferases. Biochem. J.. 257; 471-476. 1989.

5516



1997;57:5509-5516. Cancer Res John P. Gibbs, Georgia Murray, Linda Risler, et al. Children and Adults Receiving High-Dose BusulfanAge-dependent Tetrahydrothiophenium Ion Formation in Young

Updated version

http://cancerres.aacrjournals.org/content/57/24/5509

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

[email protected] 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/57/24/5509To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



age-dependent tetrahydrothiophenium ion formation in …[cancer research 57. 5509-5516, december 15....

Documents