Accepted Manuscript

Title: Simultaneous quantification of preactivated ifosfamidederivatives and of 4-hydroxyifosfamide by high performanceliquid chromatography–tandem mass spectrometry in mouseplasma and its application to a pharmacokinetic study

Author: Alain Deroussent Charles Skarbek Adeline MauryHubert Chapuis Estelle Daudigeos-Dubus Ludivine Le DretSylvere Durand Patrick Couvreur Didier Desmaele AngeloPaci

PII: S1570-0232(15)00228-7DOI: http://dx.doi.org/doi:10.1016/j.jchromb.2015.04.025Reference: CHROMB 19416

To appear in: Journal of Chromatography B

Received date: 17-1-2015Revised date: 9-4-2015Accepted date: 11-4-2015

Please cite this article as: <doi>http://dx.doi.org/10.1016/j.jchromb.2015.04.025</doi>

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 25

Accep

ted

Man

uscr

ipt

1

Highlights

A LC-MS/MS assay for quantification of ifosfamide conjugates and hydroxy metabolite.

The hydroxyifosfamide is stabilized as semicarbazone derivative in plasma.

The sensitivity for hydroxyifosfamide is improved five-fold compared to analogues.

The assay is validated and applied to a comparative pharmacokinetic study in mice.

The active metabolite is more produced by some of antitumor ifosfamide conjugates.

Page 2 of 25

Accep

ted

Man

uscr

ipt

2

Simultaneous quantification of preactivated ifosfamide derivatives

and of 4-hydroxyifosfamide by high performance liquid chromatography–

tandem mass spectrometry in mouse plasma and its application to a

pharmacokinetic study.

Alain Deroussenta(*), Charles Skarbeka, Adeline Maurya, Hubert Chapuisb, Estelle Daudigeos-

Dubusa, Ludivine Le Dreta, Sylvère Durandc, Patrick Couvreurb, Didier Desmaëleb and

Angelo Pacia,d

a CNRS UMR 8203, Laboratory of Vectorology and Anticancer Therapeutics, Paris-Sud

University, Gustave Roussy Cancer Campus Grand Paris, Villejuif, France,

b CNRS UMR 8612, Laboratory of New Anticancer Targeting Strategies, Paris-Sud

University, Galien Institute, Châtenay-Malabry, France,

c Mass Spectrometry Metabolomics Facility, Paris-Sud University, Gustave Roussy Cancer

Campus Grand Paris, Villejuif, France,

d Pharmacology and Drug analysis Department, Gustave Roussy Cancer Campus Grand

Paris, Villejuif, France.

* Corresponding author, Tel: +33 142114695.

E-mail address: alain.deroussent@gustaveroussy.fr (A. Deroussent).

Page 3 of 25

Accep

ted

Man

uscr

ipt

3

ABSTRACT

The antitumor drug, ifosfamide (IFO), requires activation by cytochrome P450 (CYP)

to form the active metabolite, 4-hydroxyisfosfamide (4-OHIFO), leading to toxic by-products

at high dose. In order to overcome these drawbacks, preactivated ifosfamide derivatives

(RXIFO) were designed to release 4-OHIFO without CYP involvement. A high performance

liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method was developed

for the simultaneous quantification of 4-OHIFO, IFO and four derivatives RXIFO in mouse

plasma using multiple reaction monitoring. Because of its instability in plasma, 4-OHIFO was

immediately converted to the semi-carbazone derivative, 4-OHIFO-SCZ. For the six analytes,

the calibration curves were linear from 20 to 5000 ng/mL in 50 µL plasma and the lower limit

of quantitation was determined at 20 ng/mL with accuracies within ± 10% of nominal and

precisions less than 12%. Their recoveries ranged from 62 to 96% by using liquid-liquid

extraction. With an improved assay sensitivity compared to analogues, the derivative 4-

OHIFO-SCZ was stable in plasma at 4°C for 24h and at -20°C for three months. For all

compounds, the assay was validated with accuracies within ±13% and precisions less than

15%. This method was applied to a comparative pharmacokinetic study of 4-OHIFO from

IFO and three derivatives RXIFO in mice. This active metabolite was produced by some of

the novel conjugates with good pharmacokinetic properties.

Keywords: Preactivated ifosfamide; 4-Hydroxyifosfamide; HPLC-MS/MS; Mouse plasma;

Pharmacokinetics.

Page 4 of 25

Accep

ted

Man

uscr

ipt

4

1. Introduction

The prodrug ifosfamide (IFO) requires activation by hepatic cytochrome P450 (CYP)

to produce the active metabolite, 4-hydroxyisfosfamide (4-OHIFO), leading to the cytotoxic

isophosphoramide mustard. Nevertheless, after IFO administration at high dose, the metabolic

reactions released the by-products, acrolein and chloroacetaldehyde, responsible of urotoxic

and nephrotoxic effects, respectively [1]. To overcome these drawbacks, several novel

preactivated ifosfamide derivatives (RXIFO) were designed to generate directly 4-OHIFO

without CYP involvement, through the covalent binding of several O- and S-poly-isoprenoid

groups at the C-4 position of the oxazaphosphorine ring (Fig. 1A), as reported recently [2, 3].

Firstly, the metabolite 4-OHIFO has been indirectly quantified using high performance

liquid chromatography (HPLC) with fluorimetric detection of acrolein [4]. Then, because of

its instability in plasma, 4-OHIFO was stabilized after blood collection by derivatization

using semi-carbazide [5] or dinitrophenylhydrazine [6] and analyzed by HPLC-UV. But, all

these assays need a large plasma volume (>0.5 mL) and cannot meet the requirement of a

pharmacokinetic study in mice at a low dose. Recently, the analogue 4-

hydroxycyclophosphamide (4-OHCPM) was derivatized and quantified by high performance

liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) [7-9] with better

sensitivity. However, for one of these assays using 100 µL of human plasma [9], the HPLC

conditions were not suitable for stability and sensitivity of the studied compounds.

The aim was to develop and to validate a simultaneous HPLC-MS/MS assay of the

compounds, RXIFO, IFO and 4-OHIFO, as semi-carbazone derivative (4-OHIFO-SCZ), in 50

µL mouse plasma. This method was successfully applied to the comparative pharmacokinetic

study of 4-OHIFO from IFO and three conjugates RXIFO in mice.

Page 5 of 25

Accep

ted

Man

uscr

ipt

5

2. Experimental

2.1 Chemicals

Ifosfamide (Holoxan®, IFO) was obtained with 99.5% purity from Baxter Healthcare

SA (Maurepas, France). 4-hydroperoxyifosfamide (4-OOHIFO) was provided by Niomech

(Bielefeld, Germany) with 99% purity. 4-OHIFO was then prepared by reduction of 4-

HOOIFO. Five pre-activated IFO derivatives (RXIFO), 4-methyloxyifosfamide (MeIFO), 4-

geranyloxyifosfamide (GerIFO), 13,17-dimethyloctanyloxyifosfamide (DMOIFO), 4-

thiosqualenifosfamide (TSQIFO) and 4-squalenyloxyifosfamide (SQIFO) were synthesized

by our group [2]. Their structure was ascertained by 1H and 13C nuclear magnetic resonance

spectroscopy and by high resolution mass spectrometry [2]. All compounds showed 99%

purity by HPLC-MS/MS [2]. Hydrochloride semi-carbazide (SCZ), LC-MS grade acetonitrile,

ammonium thiosulfate, ammonium acetate, hexamethylphosphoramide (HMP) and methyl

tert-butyl ether (MTBE) were supplied by Sigma-Aldrich (St Louis, MO, USA). Deionised

water was prepared using a Milli-Q® (Millipore, Guyancourt, France). Drug-free heparinised

mouse plasma was provided by GeneTex Inc (Irvine, CA, USA).

2.2. Stock solutions

Independent stock solutions of the five compounds RXIFO were prepared at 1.00

mg/mL in acetonitrile. As described previously [5, 7], 4-OOHIFO solution (1.12 mg/mL in

50mM ammonium acetate buffer) was reduced into 4-OHIFO (1.00 mg/mL) with 20mM

ammonium thiosulfate at 20°C for 30 min, using ultra-sonication. Working solutions

containing the mixture of 4-OHIFO, IFO, MeIFO, GerIFO, TSQIFO and SQIFO were

prepared at 200 and 10 µg/mL in acetonitrile. Using 1.0 mg/mL solutions of HMP and

DMOIFO, as internal standards (IS), a solution of IS mixture (50:50, v/v) was prepared at 100

ng/mL in acetonitrile. All solutions were stored in glass tubes at -20°C and were stable up to

Page 6 of 25

Accep

ted

Man

uscr

ipt

6

at least three months. The derivatization reagent SCZ was prepared at 2M in pH 7.4 buffer (10

mM ammonium bicarbonate adjusted with ammonia 25%).

2.3 Calibration, quality control and sample preparation

Eight calibration samples (20, 50, 100, 250, 500, 1000, 2500 and 5000 ng/mL) and

four quality control samples (QCs) (20, 400, 2000 and 4000 ng/mL) were prepared daily in

475 µL mouse plasma by adding a 25 µL volume of appropriate working solutions containing

the six analytes 4-OHIFO, IFO, MeIFO, GerIFO, TSQIFO and SQIFO. Then, 50 µL of 2M

SCZ was added immediately and mixed to each calibration and QC sample (500 µL). Thus, 4-

OHIFO was converted into 4-OHIFO-SCZ at 4°C after 30 min.

Then, 50 µL of IS solution (100 ng/mL) was added in a 1.5 mL polypropylene tube

containing 55 µL aliquot of each calibration, QC or unknown plasma sample (containing 10%

2M SCZ). Extraction was performed by adding 1 mL MTBE before vortex-mixing. After

centrifugation at 4°C at 3000 x g for 5 min, 0.95 mL of the upper organic layer was

transferred and evaporated under vacuum centrifugation at 30°C. The dry extract was

dissolved in 200 µL of eluent until HPLC-MS/MS analysis of 20 µL.

2.4 Liquid chromatography-tandem mass spectrometry

Samples analysis was performed using a 1100 series HPLC system (Agilent

Technologies, Waldbronn, Germany) fitted with an Uptisphere® C18 column (100mm x

2.1mm i.d., particle size 5µm) from Interchim (Montluçon, France). Isocratic elution was

achieved with a flow rate of 0.25 mL/min using the mobile phase (acetonitrile/aqueous buffer

5mM ammonium acetate (pH 5.5), 95:5, v/v). The total run time was 16.0 min. Before each

injection, the auto-sampler was washed with acetonitrile. Detection of analytes was performed

with a QuattroLC® triple quadrupole mass spectrometer (Waters, Manchester, UK), operating

in positive electrospray mode and multiple reaction monitoring (MRM) with the following

Page 7 of 25

Accep

ted

Man

uscr

ipt

7

transitions: m/z 334 → 221 for 4-OHIFO-SCZ, m/z 291 → 259 for MeIFO, m/z 261 → m/z

154 for IFO, m/z 180 → 135 for HMP (IS1), m/z 413 → 259 for GerIFO, m/z 417 → 259 for

DMOIFO (IS2), m/z 718 → 259 for ThioSQIFO and m/z 645.5 → 259 for SQIFO. Using 125

ms dwell times, collision energy was optimized at 15, 20, 25, 25, 20, 20, 12 and 18 eV,

respectively. Data were processed using MasslynxTM software. Parameters setting were:

capillary and cone voltage at 3500 and 25 V, source and nitrogen temperature at 90 and

250°C, respectively. The drying gas was set at 700 L/h. Collision gas (Argon) pressure was

1.2 10-3 mbar. Q1 and Q3 quadrupoles were set at low and at unit mass resolution, respectively.

2.5 Assay validation

The assay was validated in terms of selectivity, extraction recovery, matrix effect,

linearity, sensitivity, accuracy, precision and stability, in accordance to the US FDA and

EMA guidelines on bioanalytical method validation [10, 11].

2.5.1 Selectivity

Control blank samples prepared with drug-free mice plasma were analyzed for testing

absence of matrix interference. Selectivity was performed by analysis of six plasma samples

containing all analytes at the lowest concentration.

2.5.2 Recovery and matrix effect

Recoveries for the six analytes in mouse plasma were determined at the four QCs

levels and for IS. They were calculated as the ratio of analyte peak area from extracted QC

plasma to mean peak area from extracted blank plasma spiked with the neat solutions. The

matrix effect variation was evaluated using extracted blank mouse plasma samples spiked

with analytes at three QCs concentrations (20, 400 and 4000 ng/mL). Matrix effect was

Page 8 of 25

Accep

ted

Man

uscr

ipt

8

determined as the ratio of analyte peak area from extracted blank plasma spiked with the

analyte solutions to the mean peak area of the analyte solutions at the same concentration

prepared in the HPLC eluent.

2.5.3 Linearity and sensitivity

Calibration curves were performed daily in duplicate over three different days over the

concentration range and fitted by plotting the peak area ratio of each analyte and to IS against

the nominal concentration (x) of each analyte using 1/x² least-squares linear regression. The

lower limit of quantitation (LLOQ) was the lowest concentration of the analyte with a

deviation less than ±20% of the nominal concentration. For other calibration samples, the

maximal deviation was ±15%. The lower limit of detection (LLOD) was calculated with a

signal to noise ratio of 3, using the ratio of the standard deviation of the intercept and the

mean slope from three calibration curves.

2.5.4 Accuracy and precision

Accuracies and precisions were assessed for the four QCs of mouse plasma with six

replicates for each run (within-day) and over three days (between-day). Accuracy was

expressed as the bias or relative deviation between the measured and nominal concentration.

Precision was defined by the coefficient of variation (CV) or the standard deviation divided

by the mean concentration. Acceptance criteria should be bias within ± 15% and CV ≤ 15%.

2.5.5 Stability

The stability of the extracted samples containing IFO, four compounds RXIFO and 4-

OHIFO-SCZ was evaluated in the eluent at room temperature. Their stability was performed

at three concentrations (20, 400 and 4000 ng/mL) in mouse plasma. QC samples were stored

Page 9 of 25

Accep

ted

Man

uscr

ipt

9

at 4°C during 4h, at -20°C for 4 weeks and after three freeze-thaw cycles. In addition, storage

of 4-OHIFO-SCZ was extended in mouse plasma at 4°C for 24h and at -20°C for 14 weeks.

2.6 Pharmacokinetic study in mice

Female Swiss nude mice (average body weight of 24 g) were used for this study. All

animal experiments were carried out according to the European Union guidelines. Mice were

divided into four groups before treatment with IFO and three available drugs RXIFO. Mice

were receiving an equivalent IFO dose of 50 mg/kg by intra-peritoneal injection of 60 µL

from solutions prepared in dimethylsulfoxide at 20.0 mg/mL for IFO, 23.4 mg/mL for

MeIFO, 31.7 mg/mL for GerIFO and 50.0 mg/mL for SQIFO. During sampling of mice,

under isoflurane anesthesia, intracardiac puncture allowed to draw about 0.5 mL blood in

heparinised polypropylene tubes cooled in ice-water before and after drug administration at

the following times: 5, 15, 30, 60, 120 and 180 min (three mice at each time). Blood samples

were immediately centrifuged at 4°C for 15 min at 1600 x g. Then, a 50 µL volume of plasma

was immediately collected in a 1.5 mL tube containing 5 µL of 2M SCZ reagent, vortex-

mixed for 1 min and kept at 4°C for 30 min. Samples were frozen at -20°C until preparation

(section 2.3).

3. Results and discussion

3.1 Optimization and selectivity of the HPLC-MS/MS assay

The weak acidic conditions of the HPLC solvent containing 95% acetonitrile were

suitable for the solubility and the chemical stability of all compounds RXIFO. Isocratic

elution was chosen rather than a gradient run and allowed good detection with an acceptable

run time of 16 min. For MRM transitions of compounds RXIFO, all precursor ions gave the

Page 10 of 25

Accep

ted

Man

uscr

ipt

10

same product ion m/z 259, corresponding to IFO iminium, as shown on the mass spectrum of

GerIFO (Fig. 1A). After an optimized time of 30 min, 4-OHIFO was fully derivatized into 4-

OHIFO-SCZ as shown by its mass spectrum (Fig. 1B). Reproducible 99% yields of 4-OHIFO

derivatization in plasma were obtained at 4°C rather than 25°C to avoid any degradation, as

previously reported [5],

In blank plasma samples, no co-eluting interference peaks were detected at the

retention times of IS and of analytes. The assay was selective as shown on Fig. 2 by

representative MRM chromatograms of the six analytes in mouse plasma at LLOQ of 20

ng/mL (mean bias ± 8% and CV <12%). Using isocratic elution, three compounds, 4-HOIFO-

SCZ, MeIFO and IFO, were separated thanks to their different MRM transitions. Moreover,

the mobility of compounds RXIFO depended on the size of the alkyl group as shown by their

capacity factor: 1.0 (MeIFO), 1.9 (GerIFO), 2.7 (DMOIFO), 9.2 (TSQIFO) and 23 (SQIFO).

3.2 Recovery and matrix effect

Mean extraction recoveries from mouse plasma were 71 ± 9%, 96 ± 9%, 72 ± 10%, 72

± 9%, 62 ± 10%, 62 ± 14%, 75 ± 10% and 72 ± 10% for 4-OHIFO-SCZ, IFO, MeIFO,

GerIFO, TSQIFO, SQIFO, DMOIFO and HMP, respectively. Mean values of matrix effects

were 90 ± 14%, 106 ± 5%, 90 ± 5%, 105 ± 5%, 91 ± 2%, 101 ± 2%, 94 ± 6% and 102 ± 2%,

for the eight analytes, respectively. No significant ion suppression was observed for all

analytes from extracted mouse plasma using efficient liquid-liquid extraction with MTBE.

3.3 Linearity and sensitivity

4-OHIFO-SCZ and IFO were quantified with HMP as internal standard (IS1) and the other

compounds RXIFO with the DMOIFO analogue as internal standard (IS2). Calibration curves

were linear using 1/x² weighted least-squares linear regression over the concentration range of

Page 11 of 25

Accep

ted

Man

uscr

ipt

11

20-5000 ng/mL for the six analytes with determination coefficients (r2) from 0.980 to 0.993.

The LLOQ was determined at 20 ng/mL for all analytes with accuracies within ± 10% of

nominal and precisions less than 12%. The LLODs (three- to six-fold lower than LLOQ) were

calculated at 3, 4, 6, 5, 6 and 5 ng/mL for 4-OHIFO, IFO, MeIFO, GerIFO, TSQIFO and

SQIFO, respectively. Thus, using only 50 µL plasma, this HPLC-MS/MS method is hundred-

fold more sensitive than HPLC-UV [5] and five-fold for 4-OHIFO than for the analogue 4-

OHCPM [9]. The three lipidic compounds RXIFO have a similar sensitivity with

gemcitabine-squalene [13].

3.4 Accuracy and precision

Accuracies (bias) and precisions (CV) for the six compounds in plasma QCs were

summarized in Table 1. Within–day and between-day bias ranged from -12.0 to 12.9% and

from -10.8 to 7.5%, respectively. Within–day and between-day CV were less than 13.9 and

15.0%, respectively. The HPLC-MS/MS assay was accurate and precise for the quantitation

of six compounds, 4-OHIFO, IFO and its conjugates RXIFO, over the range 20-5000 ng/mL

in plasma. It was then necessary to dilute ten-fold some mice samples to quantify

concentrations above the upper LOQ. Thus, QC spiked with analytes at 10,000 ng/mL were

diluted with blank plasma and provided bias between -4.1 and 9.7% indicating no dilution

effect with acceptable precision (CV <14%).

3.5 Stability

The compounds RXIFO, IFO and 4-OHIFO-SCZ, were stable in the HPLC solvent at

room temperature for 48h. They showed no significant degradation in QCs of mouse plasma

at 4°C during 4h preparation (bias ±6%) and at – 20°C for one month or after three freeze-

thaw cycles (bias less than ± 10%), which was suitable to a pharmacokinetic study in mice. In

Page 12 of 25

Accep

ted

Man

uscr

ipt

12

agreement with stability experiments in human plasma [5], the derivative 4-OHIFO-SCZ was

stable in mouse plasma at 4°C for 24h and at –20°C for at least three months (deviation less

than ±5.5%) as shown in Table 2. Thus, plasma samples can be kept at -20°C for a long-term

safe storage, in case of further in vivo studies.

3.6 Pharmacokinetic study in mice

This assay was applied to a comparative pharmacokinetic study of three conjugates

RXIFO versus IFO in mice, receiving an IFO dose at 50 mg/kg or 192 µmol/kg. Kinetic

profiles of 4-OHIFO concentrations from IFO, MeIFO, GerIFO and SQIFO were displayed

on Fig. 3. Mean values of pharmacokinetic parameters for each drug and the metabolite 4-

OHIFO were calculated using compartmental analysis with the program PK [14] and were

summarized in Table 3. On one hand, with a short maximum time Tmax (4-OHIFO) of 13 min,

MeIFO and GerIFO indicated faster absorption than IFO and SQIFO. On another hand, the

elimination half-time (t1/2) of 4-OHIFO was similar for IFO and MeIFO, but was higher for

GerIFO and SQIFO. Despite a large variability, 4-OHIFO, generated from GerIFO and

MeIFO, had a maximum concentration (Cmax) twice and four-fold higher than that from IFO,

and the area under the curve (AUC) half and three-fold, respectively. In addition, the AUC

quotients of metabolite and RXIFO were higher compared to IFO. Thus, 4-OHIFO was more

produced from MeIFO than IFO. Moreover, AUC values of metabolite 4-OHIFO were IFO

dose-dependent as shown between 50 mg/kg and higher doses [15, 16]. Taking into account

these preclinical data, some of the novel ifosfamide conjugates could be good candidates for

further efficacy investigations using human tumor xenograft models in mice, based on the in

vitro cytotoxic activities in human tumor cell lines, such as rhabdomyosarcoma [2].

Page 13 of 25

Accep

ted

Man

uscr

ipt

13

4. Conclusion

The HPLC-MS/MS method is validated for the simultaneous quantification of 4-

hydroxyifosfamide, IFO and conjugates RXIFO ranging from 20 to 5000 ng/mL in plasma. It

was applied to a pharmacokinetic study of 4-OHIFO from IFO and three derivatives RXIFO

in mice. Two novel ifosfamide conjugates showed good pharmacokinetic properties with a

higher AUC or Cmax of 4-OHIFO. This assay needs a small plasma volume and could be

useful for therapeutic dose monitoring of the active 4-OHIFO metabolite in pediatrics.

Acknowledgments

This study was part of the NanoSqualOnc research project (Program P2N Grant N°

Nano-00301) and has received funding by the French Research Agency.

References

[1] B. Giraud, G. Hebert, A. Deroussent, G.J. Veal, G. Vassal, A. Paci, New therapeutic

strategies for an old class of drugs, Expert Opin. Drug Metab. Toxicol. 6 (2010) 919–938.

[2] C. Skarbek, L.L. Lesueur, H. Chapuis, A. Deroussent, C. Pioche-Durieu, A. Davillé, J.

Caron, M. Rivard, T. Martens, J-R. Bertrand, E. Le Cam, G. Vassal, P. Couvreur, D.

Desmaële, A. Paci, Preactivated oxazaphosphorines designed for isophosphoramide mustard

delivery as bulk form or nanoassemblies: synthesis and proof of concept, J. Med. Chem. 58

(2015) 705-717.

[3] C. Skarbek, H. Chapuis, L. Lesueur, M. Rivard, T. Martens, D. Desmaële, E. Daudigeos-

Dubus, J-R. Bertrand, A. Deroussent, G. Vassal, P. Couvreur, A. Paci, Enhanced activity of

pre-activated oxazaphosphorines prodrugs designed for drug delivery strategy: influence of

the length of the engrafted group, Am. Ass. Cancer. Res., 104th Annual Meeting (2013),

abstract 5671.

Page 14 of 25

Accep

ted

Man

uscr

ipt

14

[4] G.P. Kaijser, P.G. Ter Riet, J. de Kraker, A. Bult, J.H. Beijnen, W.J. Underberg,

Determination of 4-hydroxyifosfamide in biological matrices by high-performance liquid

chromatography, J. Pharm. Biomed. Anal. 6 (1997) 773-81.

[5] T. Kerbusch, A.D. Huitema, J.J. Kettenes-van den Bosch, H.J. Keizer, J. Ouwerkerk, J. de

Kraker, J.H. Beijnen, High-performance liquid chromatographic determination of stabilized 4-

hydroxyifosfamide in human plasma and erythrocytes, J. Chromatogr. B 716 (1998) 275-284.

[6] R. Preiss, F. Baumann, D. Stefanovic, U. Niemeyer, W. Pönisch, D. Niederwieser,

Investigations on the pharmacokinetics of trofosfamide and its metabolites-first report of 4-

hydroxy-trofosfamide kinetics in humans, Cancer Chemother. Pharmacol. 53 (2004) 496-502.

[7] N. Sadagopan, L. Cohen, B. Roberts, W. Collard, C. Omer, Liquid chromatography-

tandem mass spectrometric quantitation of cyclophosphamide and its hydroxy metabolite in

plasma and tissue for determination of tissue distribution, J. Chromatogr. B 759 (2001) 277-

284.

[8] T.F. Kalhorn, W.N. Howald, S. Cole, B. Phillips, J. Wang, J.T. Slattery, J.S. McCune,

Rapid quantitation of cyclophosphamide metabolites in plasma by liquid chromatography-

mass spectrometry, J. Chromatogr. B 835 (2006) 105-113.

[9] C. Ekhart, A. Gebretensae, H. Rosing, S. Rodenhuis, J.H. Beijnen, A.D. Huitema,

Simultaneous quantification of cyclophosphamide and its active metabolite 4-

hydroxycyclophosphamide in human plasma by high-performance liquid chromatography

coupled with electrospray ionization tandem mass spectrometry (LC-MS/MS), J. Chromatogr.

B 854 (2007) 345-349.

[10] U.S. Department of Health and Human Services, Food and Drug Administration, Center

for Drug Evaluation and Research. Guidance for Industry: Bioanalytical Method Validation,

2001.

Page 15 of 25

Accep

ted

Man

uscr

ipt

15

[11] European Medicines Agency, Committee for Medicinal Products for Human Use,

Guideline on bioanalytical method Validation, 2011.

[12] A. Deroussent, S. Rodriguez, S. Martelli S, A. Seck, E. Dubus-Daudigeos, D. Desmaële,

G. Vassal, A. Paci, Quantification of dimethyl-ifosfamide and its N-deschloropropylated

metabolites in mouse plasma by liquid chromatography-tandem mass spectrometry, J.

Chromatogr. B 879 (2011) 743-750.

[13] H. Khoury, A. Deroussent, L.H. Reddy, P. Couvreur, G. Vassal, A. Paci,

Simultaneous determination of gemcitabine and gemcitabine-squalene by liquid

chromatography-tandem mass spectrometry in human plasma, J. Chromatogr. B 858 (2007)

71-78.

[14] J. Usansky, A. Desai, D. Tang-Liu, Department of Pharmacokinetics and Drug

Metabolism, Allergan, Irvine CA 92612, PK functions for Microsoft Excel,

http://www.pharmpk.com/soft.html.

[15] V. Kurowski, T. Cerny, A. Küpfer, T. Wagner, Metabolism and pharmacokinetics of oral

and intravenous ifosfamide, J. Cancer Res. Clin. Oncol. 117 (1991) 148-153.

[16] V. Kurowski, T. Wagner, Comparative pharmacokinetics of ifosfamide, 4-

hydroxyifosfamide, chloroacetaldehyde, and 2- and 3-dechloroethylifosfamide in patients on

fractionated intravenous ifosfamide therapy, Cancer Chemother. Pharmacol. 33 (1993) 36-42.

Page 16 of 25

Accep

ted

Man

uscr

ipt

16

Figures captions.

Fig. 1. Product ion mass spectra of (A) 4-geranyloxyifosfamide, representative of RXIFO

compounds, and (B) 4-hydroxyifosfamide semi-carbazone derivative (4-OHIFO-SCZ).

Fig. 2. MRM chromatograms of mouse plasma spiked with 20 ng/mL (LLOQ) of 4-OHIFO-

SCZ (a), MeIFO (b), IFO (c), GerIFO (d), TSQIFO (g), SQIFO (h) and with 100 ng/mL of

HMP as IS1 (e) and DMOIFO as IS2 (f).

Fig. 3. Plasma semi-logarithmic concentration-time curves of 4-OHIFO from IFO and three

compounds derivatives RXIFO in mice after intra-peritoneal administration (192 µmol/kg).

Tables captions.

Table 1. Accuracy (bias) and precision (C.V.) of the six analytes in mouse plasma.

c, concentration; C.V., coefficient of variation (Footnote).

Table 2. Stability of the derivative 4-OHIFO-SCZ in mouse plasma QC samples (n=3).

c, concentration; C.V., coefficient of variation (Footnote).

Table 3. Pharmacokinetic parameters of drugs, IFO, MeIFO, GerIFO, SQIFO, and of the

metabolite 4-OHIFO in mice (mean ± S.D., n=3).

RM/D Cmax , quotient of Cmax (M) and Cmax (D); RM/D AUC, quotient of AUC (M) and AUC

(D) (Footnote).

Page 17 of 25

Accep

ted

Man

uscr

ipt

17

Within-day (n=6) Between-day (n=18)

AnalyteNominal c (ng/mL)

Mean c (ng/mL)

Bias (%) C.V. (%)Mean c (ng/mL)

Bias (%) C.V. (%)

4-OHIFO 20.0 20.5 2.5 8.0 21.2 6.0 10.1

400.0 358.3 -10.3 6.0 383.7 -4.1 10.8

2000 1735 -12.2 2.5 1783 -10.8 9.7

4000 3727 -6.8 2.7 4009 0.2 8.6

MeIFO 20.0 21.6 8.0 9.9 20.8 4.0 10.1

400.0 426.2 6.5 12.3 405.2 1.3 12.4

2000 2218 10.9 10.5 1940 -3.0 10.4

4000 3965 -0.9 13.9 3722 -7.0 12.8

IFO 20.0 20.6 3.0 8.2 21.5 7.5 9.4

400.0 448.3 12.1 7.4 426.5 6.6 12.4

2000 2065 3.3 3.7 2069 3.5 7.2

4000 3591 -10.2 8.8 3597 -10.1 8.6

GerIFO 20.0 20.7 3.5 6.0 19.7 -1.5 12.7

400.0 451.6 12.9 4.9 417.9 4.5 7.5

2000 2064 3.2 2.8 1783 -10.8 10.7

4000 3652 -8.7 3.0 3652 -8.7 3.0

TSQIFO 20.0 19.0 -5.0 9.8 19.3 -3.5 11.1

400.0 389.2 -2.7 12.2 386.4 -3.4 12.2

2000 1786 -10.7 9.1 1817 -9.2 9.4

4000 3447 -12.0 9.3 3790 -5.2 7.8

SQIFO 20.0 17.6 -8.0 11.8 18.9 -5.5 15.0

400.0 424.9 6.2 13.2 407.1 1.8 13.1

2000 2048 2.4 9.5 1860 -7.0 15.0

4000 3973 -0.7 5.5 3621 -9.5 11.2

Page 18 of 25

Accep

ted

Man

uscr

ipt

18

Conditions Nominal c (ng/mL) Mean c (ng/mL) Bias (%) C.V. (%)

20.0 21.0 5.0 6.0

400 386 -3.5 4.24°C, 24h

4000 3764 -5.9 3.9

20.0 21.9 9.5 6.3

400 362 -9.8 3.13 freeze (-20°C) / thaw cycles

4000 3862 3.5 4.7

20.0 20.7 3.5 7.9

400 378 -5.5 3.7-20°C, 4 weeks

4000 3785 -4.8 3.3

20.0 18.9 -5.5 7.6

400 394 -1.5 6.8-20°C, 14 weeks

4000 3820 -4.5 5.5

Page 19 of 25

Accep

ted

Man

uscr

ipt

19

Drug (D) and Metabolite (M)

Tmax (h) t1/2 (h) Cmax

(µmol/L)RM/D Cmax

AUC(µmol.h/L)

RM/D AUC

IFO 0.19 ± 0.10 0.30 ± 0.03 239.7 ± 92.7 115.5 ± 21.5

4-OHIFO 0.33± 0.14 0.33 ± 0.05 16.0 ± 5.0 0.07± 0.02 16.3 ± 2.2 0.14± 0.02

MeIFO 0.08 ± 0.00 0.26 ± 0.02 96.6 ± 64.4 24.4 ± 8.2

4-OHIFO 0.22 ± 0.24 0.31 ± 0.05 67.6 ± 45.6 0.70± 0.47 44.2 ± 23.9 1.77± 0.98

GerIFO 0.08 ± 0.00 0.37 ± 0.11 23.1 ± 1.6 4.8 ± 0.8

4-OHIFO 0.22 ± 0.24 0.49 ± 0.25 32.2 ± 39.0 1.39± 1.69 8.6 ± 4.9 1.78± 1.01

SQIFO 1.00 ± 0.00 1.38 ± 0.39 1.2 ± 0.7 1.4 ± 0.6

4-OHIFO 0.50 ± 0.43 0.88 ± 0.11 5.0 ± 1.6 4.14± 1.29 6.7 ± 2.4 4.79± 1.68

Page 20 of 25

Accep

ted

Man

uscr

ipt

100 150 200 250 300 350 400 450 500 550 600 650 700m/z0

100

%

anrad264 53 (0.899) Daughters of 413ES+ 6.93e5259

221118

413

100 150 200 250 300 350 400 450 500 550 600 650 700m/z0

100

%

anrad429 59 (1.001) Daughters of 334ES+ 4.49e5x54221

114

334O

PHNN NH

ClCl

O

HN

ONH

2m/z 221

m/z 114[M2+H]+

(*)

Abu

nd

ance

(%

)

O

PN NH

ClCl

O+

(*)

O

PNXR NH

ClCl

O

[M1+H]+

A

bundan

ce (

%)

(B)

(A)

Figure 1_JCB-15-57R1

Page 21 of 25

Accep

ted

Man

uscr

ipt

0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00Time0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

0

100

%

adifox681 MRM of 7 Channels ES+ 334 > 221

1.85e41.16

adifox1393 MRM of 5 Channels ES+ 291 > 259

2.33e31.18

adifox681 MRM of 7 Channels ES+ 261 > 154

2.31e41.20

adifox681 MRM of 7 Channels ES+ 413 > 259

2.53e41.67

adifox681 MRM of 7 Channels ES+ 180 > 135

1.65e51.93

adifox681 MRM of 7 Channels ES+ 417 > 259

1.13e52.22

adifox681 MRM of 7 Channels ES+ 718 > 259

2626.07

adifox681 MRM of 7 Channels ES+ 645 > 259

5.30e314.41

Rel

ativ

e in

ten

sity

(%

)

Time (min)

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Figure 2_JCB-15-57R1

Page 22 of 25

Accep

ted

Man

uscr

ipt

10

100

1000

10000

0 30 60 90 120 150 180

4-O

HIF

O c

once

ntr

atio

n

(ng/m

L)

Time (min) after dosing

GerIFO SQIFO IFO MeIFO

Figure 3_JCB-15-57R1

Page 23 of 25

Accep

ted

Man

uscr

ipt

AnalyteNominal c

(ng/mL)

Mean c

(ng/mL)Bias (%) C.V. (%)

Mean c

(ng/mL)Bias (%) C.V. (%)

4-OHIFO 20.0 20.5 2.5 8.0 21.2 6.0 10.1

400.0 358.3 -10.3 6.0 383.7 -4.1 10.8

2000 1735 -12.2 2.5 1783 -10.8 9.7

4000 3727 -6.8 2.7 4009 0.2 8.6

MeIFO 20.0 21.6 8.0 9.9 20.8 4.0 10.1

400.0 426.2 6.5 12.3 405.2 1.3 12.4

2000 2218 10.9 10.5 1940 -3.0 10.4

4000 3965 -0.9 13.9 3722 -7.0 12.8

IFO 20.0 20.6 3.0 8.2 21.5 7.5 9.4

400.0 448.3 12.1 7.4 426.5 6.6 12.4

2000 2065 3.3 3.7 2069 3.5 7.2

4000 3591 -10.2 8.8 3597 -10.1 8.6

GerIFO 20.0 20.7 3.5 6.0 19.7 -1.5 12.7

400.0 451.6 12.9 4.9 417.9 4.5 7.5

2000 2064 3.2 2.8 1783 -10.8 10.7

4000 3652 -8.7 3.0 3652 -8.7 3.0

TSQIFO 20.0 19.0 -5.0 9.8 19.3 -3.5 11.1

400.0 389.2 -2.7 12.2 386.4 -3.4 12.2

2000 1786 -10.7 9.1 1817 -9.2 9.4

4000 3447 -12.0 9.3 3790 -5.2 7.8

SQIFO 20.0 17.6 -8.0 11.8 18.9 -5.5 15.0

400.0 424.9 6.2 13.2 407.1 1.8 13.1

2000 2048 2.4 9.5 1860 -7.0 15.0

4000 3973 -0.7 5.5 3621 -9.5 11.2

Within-day (n=6) Between-day (n=18)

Table 1_JCB-15-57R1

Page 24 of 25

Accep

ted

Man

uscr

ipt

Conditions Nominal c (ng/mL) Mean c (ng/mL) Bias (%) C.V. (%)

20.0 21.0 5.0 6.0

400 386 -3.5 4.2

4000 3764 -5.9 3.9

20.0 21.9 9.5 6.3

400 362 -9.8 3.1

4000 3862 3.5 4.7

20.0 20.7 3.5 7.9

400 378 -5.5 3.7

4000 3785 -4.8 3.3

20.0 18.9 -5.5 7.6

400 394 -1.5 6.8

4000 3820 -4.5 5.5

3 freeze (-20°C) /

thaw cycles

4°C, 24h

-20°C, 4 weeks

-20°C, 14 weeks

Table 2_JCB-15-57R1

Page 25 of 25

Accep

ted

Man

uscr

ipt

Drug (D) and

Metabolite (M) Tmax (h) t1/2 (h)

Cmax

(µmol/L)RM/D Cmax

AUC

(µmol.h/L)RM/D AUC

IFO 0.19 ± 0.10 0.30 ± 0.03 239.7 ± 92.7 115.5 ± 21.5

4-OHIFO 0.33± 0.14 0.33 ± 0.05 16.0 ± 5.0 0.07± 0.02 16.3 ± 2.2 0.14± 0.02

MeIFO 0.08 ± 0.00 0.26 ± 0.02 96.6 ± 64.4 24.4 ± 8.2

4-OHIFO 0.22 ± 0.24 0.31 ± 0.05 67.6 ± 45.6 0.70± 0.47 44.2 ± 23.9 1.77± 0.98

GerIFO 0.08 ± 0.00 0.37 ± 0.11 23.1 ± 1.6 4.8 ± 0.8

4-OHIFO 0.22 ± 0.24 0.49 ± 0.25 32.2 ± 39.0 1.39± 1.69 8.6 ± 4.9 1.78± 1.01

SQIFO 1.00 ± 0.00 1.38 ± 0.39 1.2 ± 0.7 1.4 ± 0.6

4-OHIFO 0.50 ± 0.43 0.88 ± 0.11 5.0 ± 1.6 4.14± 1.29 6.7 ± 2.4 4.79± 1.68

Table 3_JCB-15-57R1