metabolism, excretion and pharmacokinetics of [ 14 c]crizotinib...

15

Click here to load reader

Upload: bill-j

Post on 05-Apr-2017

216 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

http://informahealthcare.com/xenISSN: 0049-8254 (print), 1366-5928 (electronic)

Xenobiotica, Early Online: 1–15! 2014 Informa UK Ltd. DOI: 10.3109/00498254.2014.941964

RESEARCH ARTICLE

Metabolism, excretion and pharmacokinetics of [14C]crizotinib followingoral administration to healthy subjects

Theodore R. Johnson1, Weiwei Tan2, Lance Goulet1*, Evan B. Smith1*, Shinji Yamazaki1, Gregory S. Walker3,Melissa T. O’Gorman4, Gabriella Bedarida5, Helen Y. Zou6, James G. Christensen6*, Leslie N. Nguyen1*,Zhongzhou Shen1*, Deepak Dalvie1, Akintunde Bello7, and Bill J. Smith1

1Department of Pharmacokinetics, Dynamics and Metabolism, Pfizer Inc., San Diego, CA, USA, 2Department of Clinical Pharmacology,

Pfizer Inc., San Diego, CA, USA, 3Department of Pharmacokinetics, Dynamics and Metabolism, Pfizer Inc., Groton, CT, USA, 4Department of

Clinical Pharmacology, Pfizer Inc., Groton, CT, USA, 5Clinical Research Unit, Pfizer Inc., New Haven, CT, USA, 6Oncology Research, Pfizer Inc.,

San Diego, CA, USA, and 7Department of Clinical Pharmacology, Pfizer Inc., New York, NY, USA

Abstract

1. Crizotinib (XALKORI�), an oral inhibitor of anaplastic lymphoma kinase (ALK) andmesenchymal-epithelial transition factor kinase (c-Met), is currently approved for thetreatment of patients with non-small cell lung cancer that is ALK-positive.

2. The metabolism, excretion and pharmacokinetics of crizotinib were investigated followingadministration of a single oral dose of 250 mg/100 mCi [14C]crizotinib to six healthy malesubjects.

3. Mean recovery of [14C]crizotinib-related radioactivity in excreta samples was 85% of the dose(63% in feces and 22% in urine).

4. Crizotinib and its metabolite, crizotinib lactam, were the major components circulating inplasma, accounting for 33% and 10%, respectively, of the 0–96 h plasma radioactivity.Unchanged crizotinib was the major excreted component in feces (�53% of the dose).In urine, crizotinib and O-desalkyl crizotinib lactam accounted for �2% and 5% of the dose,respectively. Collectively, these data indicate that the primary clearance pathway forcrizotinib in humans is oxidative metabolism/hepatic elimination.

5. Based on plasma exposure in healthy subjects following a single dose of crizotinib andin vitro potency against ALK and c-Met, the crizotinib lactam diastereomers are notanticipated to contribute significantly to in vivo activity; however, additional assessment incancer patients is warranted.

Keywords

Crizotinib, excretion, human, metabolism,oncology, pharmacokinetics

History

Received 9 May 2014Revised 1 July 2014Accepted 2 July 2014Published online 18 July 2014

Introduction

Lung cancer is the most common and lethal cancer worldwide

(Janku et al., 2010), and the majority of lung cancers (85%)

are non-small cell lung cancer (NSCLC) (Tyczynski et al.,

2003). Recently, advances in the understanding and treatment

of NSCLC have been made based on the identification of

molecular alterations specific to tumor cells (Herbst et al.,

2008). Rearrangements of anaplastic lymphoma kinase

(ALK), a receptor tyrosine kinase (RTK), in NSCLC were

reported in 2007, primarily as fusions to echinoderm micro-

tubule-like protein 4 (EML4) (Soda et al., 2007). The potent

oncogenic activity of the EML4-ALK fusion kinase was

confirmed when expressed in the lungs of transgenic mice

(Soda et al., 2008). These fusion proteins are found in �3–7%

of NSCLC patients overall, frequently in younger patients and

in never or light smokers (Gandhi & Janne, 2012).

Collectively, these findings suggested that a pharmacotherapy

specifically targeting ALK may prove beneficial in treatment

of patients with advanced NSCLC.

Crizotinib (XALKORI�, PF-02341066) is a potent small-

molecule inhibitor of ALK and its oncogenic variants

(i.e. ALK fusion events and selected ALK mutations)

(Figure 1) (Cui et al., 2011). It is also a potent inhibitor of

the RTK mesenchymal-epithelial transition (c-Met), also

known as hepatocyte growth factor receptor, as well as

ROS1 (c-ros) and Recepteur d’Origine Nantais RTKs

(Cui et al., 2008; Zou et al., 2007). In mice bearing NSCLC

*Present addresses: Lance Goulet, PharmAkea Therapeutics, San Diego,CA, USA. Evan B. Smith, Neurocrine Pharmaceuticals, San Diego, CA,USA. James G. Christensen, Mirati Therapeutics, San Diego, CA, USA.Leslie N. Nguyen, Johnson & Johnson Pharmaceutical Research &Development, San Diego, CA, USA. Zhongzhou Shen, DartNeuroScience, San Diego, CA, USA.

Address for correspondence: Theodore R. Johnson, Ph.D.,Pharmacokinetics, Dynamics and Metabolism, La Jolla Laboratories,Pfizer Worldwide Research and Development, 10646 Science CenterDrive, San Diego, CA 92121, USA. Tel: +858-622-7988.Fax: +484-323-8332. E-mail: [email protected]

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 2: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

tumor xenografts that expressed ALK fusion protein,

crizotinib demonstrated marked, dose-dependent antitumor

efficacy, which correlated to pharmacodynamic inhibition of

phosphorylation of ALK fusion proteins (including

EML4-ALK) in tumors in vivo (Yamazaki et al., 2012; Zou

et al., 2011). Crizotinib is approved globally for the treatment

of patients with locally advanced or metastatic NSCLC that is

ALK-positive. The recommended dose schedule of crizotinib

is 250 mg taken orally twice daily. Results from a phase III

trial demonstrated that crizotinib was superior to standard

chemotherapy (pemetrexed or docetaxel) in patients with

previously treated, advanced ALK-positive NSCLC (Shaw

et al., 2013). In 347 patients who had previously been treated

with platinum-based chemotherapy, crizotinib prolonged

progression-free survival to a median of 7.7 months compared

with 3.0 months in the chemotherapy group. Likewise,

objective response rate was significantly higher in those

treated with crizotinib (65% versus 20%).

The disposition of crizotinib has been evaluated in rats and

dogs to support safety assessment and the pharmacokinetics

(PK) of crizotinib have been characterized in healthy subjects

and cancer patients in a number of clinical studies. The major

metabolic pathways in rats and dogs were oxidation of the

piperidine ring, with direct sulfation and O-dealkylation

also observed in rat (Smith et al., 2011; Zhong et al., 2010).

In nonclinical mass balance studies, recovery of total

radioactivity was essentially complete (>89%), and fecal/

biliary excretion was the major route of elimination.

Crizotinib is primarily metabolized by CYP3A and is also a

time-dependent inhibitor and inducer of CYP3A in vitro

(Johnson et al., 2011; Mao et al., 2013). In vivo, crizotinib is a

moderate CYP3A inhibitor, with a 3.7-fold increase in the

area under the plasma concentration–time curve (AUC) of

oral midazolam observed following 28 days of crizotinib

dosing at 250 mg BID (Tan et al., 2010). In cancer patients,

peak plasma concentrations of crizotinib were reached 4 h

after a single 250-mg oral dose and declined with an apparent

terminal half-life of 42 h (Li et al., 2011; Tan et al., 2010).

Oral bioavailability of crizotinib in healthy volunteers was

43% following a single 250-mg oral dose (Xu et al., 2011a).

Crizotinib exhibits time-dependent PK, as reflected by a

40% decrease in oral clearance with repeat-dose adminis-

tration, likely due to auto-inhibition/inactivation of CYP3A

(Li et al., 2011).

Given the complex pharmacokinetic profile of crizotinib in

humans, a more thorough understanding of its disposition was

warranted. Herein, we report the findings from a study of the

metabolism, excretion and PK of [14C]crizotinib in

healthy human subjects following a single 250-mg oral dose

as well as exploratory assessments of the PK and pharmaco-

logical activity of the primary oxidative metabolites of

crizotinib.

Materials and methods

Chemicals and reference compounds

Crizotinib (PF-02341066) was synthesized by Medicinal

Chemistry and Pharmaceutical Sciences, Pfizer Worldwide

Research and Development (San Diego, CA and Sandwich,

UK, respectively) (Cui et al., 2011). Synthetic standards of

crizotinib metabolites, crizotinib lactam (PF-06260182), the

constituent diastereomers of crizotinib lactam (PF-06270079

and PF-06270080), O-desalkyl crizotinib (PF-03255243) and

O-desalkyl crizotinib lactam (PF-06268935), were synthe-

sized by Medicinal Chemistry, Pfizer Worldwide Research

and Development (San Diego, CA). The absolute configur-

ation of the crizotinib lactam constituent diastereomers,

PF-06270079 and PF-06270080, were not determined.

[14C]Crizotinib (specific activity 0.407mCi/mg; radiochem-

ical purity >99%) and [2H5]crizotinib (PF-03623192) were

synthesized by Radiochemistry, Pfizer Worldwide Research

and Development (Groton, CT). Perma Fluor, Carbo-Sorb and

Ultima Gold liquid scintillation fluid was obtained from

PerkinElmer (Waltham, MA). Reagents and solvents of

analytical or high pressure liquid chromatography (HPLC)

grade were purchased from commercial manufacturers.

Study subjects

Subjects were healthy males between the ages of 18 and 55

years with body mass index of 17.5 to 30.5 kg/m2 and a total

body weight of >50 kg (110 lbs). All subjects signed an

informed consent document before screening. Exclusion

criteria included, but was not limited to: evidence or history

of clinical significant diseases, any condition possibly affect-

ing drug absorption, a positive drug screen, history of excess

alcohol consumption, use of tobacco- or nicotine-containing

products, QTc >450 ms at screening, use of drugs and dietary

supplement within 7 days or 5 half-lives prior to the start of

study drug treatment, radionucleotide study or radiotherapy

within 12 months prior to screening or such that total

radioactivity would have exceeded acceptable dosimetry

(i.e. occupational exposure of 5 rem per year), blood donation

of �500 mL within 56 days prior to dosing and a positive

serology for hepatitis B and C.

Study design

This was an open label, single dose, single center study

(Study A8081009) to evaluate the mass balance and PK of

crizotinib in six healthy male subjects, conducted at the Pfizer

Clinical Research Unit (New Haven, CT). Based on previous

clinical experiences where 250 mg twice daily dosing was

determined to be the maximum tolerated dose in cancer

patients (Kwak et al., 2010), a single 250-mg dose was

selected as the dose for this study. Dosimetry estimations

allow the use of the 100 mCi radiolabel dose based on a

[14C]crizotinib study in rats.

ON

NN

NH

ClF

ClNH2

* *

Figure 1. Chemical structure of crizotinib and position of [14C] label(asterisk).

2 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 3: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

Study conduct and sample collection

The study was conducted in compliance with the ethical

principles originating in or derived from the Declaration of

Helsinki and in compliance with all International Conference

on Harmonization Good Practice Guidelines. The study

protocol and informed consent documentation were approved

by an institutional review board, and all local regulatory

requirements were followed, in particular, those affording

greater protection to the safety of study participants.

Following an 8-h fast, each subject received a single oral

250-mg crizotinib dose, containing �100 mCi of

[14C]crizotinib, suspended in 100 mL of 0.1% (w/v) methyl-

cellulose, followed by an additional 140 mL of water (total

aqueous volume administered, 240 mL). Blood was collected

by an in-dwelling catheter or venipuncture into K2EDTA

vacutainer tubes pre-dose and at 1, 2, 3, 4, 6, 8, 10, 12, 16, 24,

36, 48, 72, 96, 120, 144, 168, 192, 216, 240, 264, 288 and

312 h post-dose in all subjects and 336, 360, 384 and 408 h in

subjects where recovery of radioactivity in excreta was

incomplete (vide infra). Plasma was prepared from blood

samples by centrifuging for 10 min at 1700 g and at �4 �C.

Urine was collected pre-dose, from 0 to 4, 4 to 8, 8 to 16

and 16 to 24 h post-dose, and at intervals of 24 h up to 480 h

post-dose. The collected urine from each void was stored

at 4 �C throughout the collection interval after the addition

of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesul-

fonate (2% weight/volume). Feces were collected pre-dose

and as passed, at intervals of 24 h up to 480 h post-dose. Daily

sample collections continued beyond 312 h until one of the

following conditions were met: the amount of radioactivity

recovered in excreta was �90% of administered radioactivity,

or51% was recovered in excreta from two consecutive days.

Total weight of urine and fecal output was recorded for each

collection interval. All samples were stored at ��20 �C or

below until the time of analysis.

Determination of total radioactivity

Radioactivity in blood, plasma, urine and feces was measured

using PerkinElmer Tri-Carb Model 2900TR liquid scintilla-

tion counters (Waltham, MA). All sample combustions were

performed in a PerkinElmer Model 307 Sample Oxidizer

(Waltham, MA) and the resulting 14CO2 was trapped in a

mixture of Perma Fluor and Carbo-Sorb scintillation fluids.

Oxidation efficiency was evaluated on each day of sample

combustion by analyzing a commercial radiolabeled standard

both directly in scintillation cocktail and by oxidation. Liquid

scintillation counting (LSC) data (cpm) were automatically

corrected for counting efficiency using the external standard-

ization technique and an instrument-stored quench curve

generated from a series of sealed quenched standards to

obtain dpm. Samples were counted for at least 5 min or

100 000 counts for urine and feces and for at least 30 min or

100 000 counts for plasma.

Blood samples were mixed and single-weighed aliquots

(�0.4 g) were combusted and analyzed by LSC. Plasma

samples were mixed, and single-weighed aliquots (�1 g) were

analyzed directly by LSC. Specific gravity values of 1.05 and

1.02 g/mL for blood and plasma, respectively, were used to

calculate radioactivity concentration as nanogram-equivalents

per milliliter (Trudnowski & Rico, 1974). Total radioactivity

concentrations in red blood cells (RBCs) were calculated for

each time point (where possible) as [Cb � Cp�(1 � H)]/H,

where Cb, Cp and H were the whole blood radioactivity

concentration, plasma radioactivity concentration and the

hematocrit, respectively (average of Day 1 and Day 14 H

values). Negative RBC concentration (CRBC) values were

reported as zero. Urine samples were mixed and triplicate-

weighed aliquots (�0.2 g) were analyzed directly by LSC.

Fecal samples were homogenized with water (1:2, w/w, feces/

water) and triplicate-weighed aliquots (�0.2 g) were com-

busted and analyzed by LSC. The limits of quantitation for

blood and plasma were 59.6 and 19.8 ng-equivalents/mL,

respectively.

Quantification of crizotinib in plasma and urine

Concentrations of crizotinib in plasma and urine samples

were determined with validated liquid chromatography–

tandem mass spectrometry (LC–MS/MS) analytical methods

in compliance with Pfizer standard operating procedures

at Covance Bioanalytical Services, LLC (Indianapolis, IN).

Briefly, isotopically labeled internal standard ([2H5]crizotinib,

PF-03623192) was added to 200 mL of plasma or 100 mL of

urine. For plasma samples, crizotinib was extracted by protein

precipitation with acetonitrile. Samples were centrifuged, and

a 500 mL aliquot was evaporated to dryness. Plasma extracts

were reconstituted in 200 mL water:formic acid (10:1, v/v),

mixed and centrifuged. For urine samples, urine (100 mL) was

treated with internal standard and diluted with 100 mL

water:formic acid (10:1, v/v). The diluted urine sample was

mixed and centrifuged. Aliquots (10 mL) of plasma extracts or

diluted urine were injected onto a Phenomenex Onyx

Monolith HPLC column (50� 4.6 mm; Torrance, CA) main-

tained at 40 �C. Analytes were eluted with mobile phases

containing mixtures of water, formic acid and methanol using

a step gradient and flow rate of 1.5 mL/min. Detection was

performed by positive ion MS/MS on a SciEx API 4000 mass

spectrometer (Applied Biosystems, Foster City, CA) with

multiple reaction monitoring (m/z 450 to 260 for crizotinib

and m/z 455 to 265 for internal standard). The LLOQ for

crizotinib measurement was 0.200 ng/mL. Calibration stand-

ard responses were linear over the range of 0.200 ng/mL to

200 ng/mL, using a weighted (l/concentration2) linear least

squares regression. For plasma samples, the between-day

assay accuracy, expressed as percent relative error for quality

control (QC) concentrations, ranged from 1.3% to 3.0% for

the low (0.600 ng/mL), medium (70.0 ng/mL) and high

(150.0 ng/mL) QC samples. Assay precision, expressed as

the between-day percent coefficients of variation (% CV) of

the mean estimated concentrations of QC samples was

�4.1%. For urine QC samples, the between-day assay

accuracy ranged from �4.7% to �2.4%, and assay precision

was �5.2%.

Pharmacokinetic analysis

PK parameters for total radioactivity (plasma, whole blood

and RBC), crizotinib (plasma), and crizotinib lactam

diastereomers (plasma) were calculated for each subject

using noncompartmental analysis of concentration–time data

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 3

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 4: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

using Pfizer internal software eNCA (version 2.2.2). Actual

sample collection times were used for the PK analysis. The

maximum plasma concentration (Cmax) and the time at which

Cmax was achieved (tmax) were estimated directly from the

concentration data. The area under the plasma concentration–

time curve (AUClast) from time 0 to the last time (tlast) with a

quantifiable concentration (Clast) was estimated using linear-

log trapezoidal approximation. AUC extrapolated to infinity

(AUC1) was calculated using the equation AUClast +

(Clast/�), where � was the elimination rate constant calculated

by a linear regression of the log-linear concentration–time

curve. Apparent terminal elimination half-life (t1/2) was

calculated as ln2/�. A well-characterized terminal phase was

defined as one with at least three data points, a goodness-of-

fit statistic for the log-linear regression (r2) �0.9, a span ratio

�2 (duration over which t1/2 was assessed, divided by the t1/2

estimate) and percent of AUC1 due to forward extrapolation

(AUCextrap%) �20%.

The amount of crizotinib excreted in urine during each

collection interval was calculated by multiplying the concen-

tration of crizotinib by the volume of urine collected over that

interval. The amount of unchanged crizotinib excreted in the

urine expressed as percent of dose was calculated as the

cumulative amount excreted over all the collection periods

divided by the dose. Renal clearance was calculated as the

cumulative amount of crizotinib excreted divided by AUC1.

Recovery of total radioactivity in urine and feces as a

percentage of dose was calculated using a similar method.

Preparation of samples for radioactivity profiling

Plasma

Time-normalized pooled plasma samples representing 0–96 h

post-dose were prepared for each subject (Hamilton et al.,

1981). A global pool was then prepared by mixing equal

volumes of each individual pool. The time interval selected

for sample pooling was limited due to low levels of circulating

radioactivity in all subjects. Two samples (�2.5 g each) of

pooled plasma were extracted three times with four volumes

of 80:20 acetonitrile:methanol. Each extraction was carried

out by vortex mixing for �1 min, sonication for �5 min and

centrifugation for 10 min at 800 g. The supernatants from each

extraction were combined, and 100mL of dimethyl sulfoxide

was added. Solvent was removed with heat under vacuum.

The concentrated extracts were reconstituted in 250 mL of

70:30 water:acetonitrile containing 0.1% formic acid and

centrifuged at 17 000 g for 5 min.

Urine

Urine samples were prepared for each subject such that the

amount of radioactivity contained in the pool collectively

represented 90% or greater of the total radioactivity recovered

in urine for a given subject (�0–120 h post-dose). The amount

of sample added for each time point was normalized by

weight of sample within each given subject. Samples (1 mL)

of pooled urine from each subject were combined with 1 mL

acetonitrile, vortex-mixed for 15 min and centrifuged for

10 min at 3000 g. Supernatants were evaporated to dryness

with heat under vacuum and then reconstituted in 200 mL of

80:20 water:acetonitrile containing 0.1% formic acid. The

reconstituted extracts were sonicated for 5 min, vortex-mixed

for 10 min and centrifuged for 10 min at 2000 g.

Feces

Weight-normalized pooled fecal homogenate samples from

each individual were generated in the same manner as used

for urine (representing �0–120 h post-dose). Samples

(�0.5 g) of pooled fecal homogenate from each subject

were extracted by solid phase using Sep Pak C18 6 cc

cartridges with a 1 g bed volume (Waters, Milford, MA).

Cartridges were conditioned with 15 mL acetonitrile/water

(98:2, v/v) containing 0.2% formic acid, and then equilibrated

with acetonitrile/water (2:98, v/v) containing 0.2% formic

acid. Fecal homogenates were loaded and cartridges were

washed with 10 mL acetonitrile/water (2:98, v/v) containing

0.2% formic acid. Analytes were eluted with 10 mL aceto-

nitrile/water (1:1, v/v) containing 0.2% formic acid into tubes

containing 50 mL dimethyl sulfoxide. The solutions resulting

from the elution step were concentrated with heat under

vacuum and reconstituted in 200 mL of acetonitrile/water

(2:8, v/v) containing 0.1% formic acid. The reconstituted

extracts were sonicated for 5 min, vortex-mixed for 10 min

and centrifuged for 10 min at 2000 g.

Supernatants of the reconstituted plasma, urine or fecal

extracts were transferred to pre-tared HPLC vials and the total

weights of the samples were recorded. Aliquots were weighed

into scintillation vials, and radioactivity was determined

(LSC) to estimate recovery. Aliquots (200mL for plasma and

100 mL for urine and feces) of the concentrated extracts were

injected into an HPLC system for analysis.

Metabolite profiling

Metabolic profiling was performed on an Agilent 1100 series

HPLC equipped with binary pumps, an autosampler and a

diode array detector (Agilent Technologies, Santa Clara, CA).

Two HPLC methods were used for generation of metabolic

profiles. The first method was used for initial metabolite

profile generation and the second method was used to

separate early eluting (55 min), unresolved components in

the initial profiles.

HPLC method 1

A Kromasil C4 column (4.6� 150 mm, 5 m) (AkzoNobel,

Brewster, NY) was used for metabolite separation. HPLC

analyses were performed at 1 mL/min flow rate and ambient

temperature. The HPLC mobile phase consisted of water with

0.1% formic acid (A) and acetonitrile with 0.1% formic acid

(B). The following linear gradient was used: 0 min, 2% B;

5 min, 2% B; 10 min, 10% B; 50 min, 30% B; 56 min, 95% B;

58 min, 95% B; and 60 min, 2% B. The column was re-

equilibrated at 2% B for 10 min prior to the next injection.

HPLC method 2

A Synergi Polar-RP column (2.0� 250 mm, 4 m)

(Phenomenex, Torrance, CA)was utilized for metabolite

separation at flow rate of 0.2 mL/min flow and ambient

temperature. The HPLC mobile phase consisted of water with

4 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 5: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

10 mM ammonium formate (A) and acetonitrile (B). The

following linear gradient was used: 0 min, 2% B; 7 min, 2% B;

15 min, 10% B; 35 min, 20% B; 45 min, 40% B; 62 min, 95%

B; 64 min, 95% B; and 65 min, 2% B. The column was re-

equilibrated at 2% B for 25 min prior to the next injection.

For the plasma profiles, fraction collection of the HPLC

column effluent with subsequent off-line determination of

radioactivity by LSC was used. Fractions were collected at

0.5 min intervals using a Gilson FC 204 Fraction Collector

(Middleton, WI). Aliquots of each fraction were removed for

analysis by LC/MS, and the radioactivity in the remaining

fractions was determined by LSC using a 30 min count time.

Radiochromatographic metabolite profiles were constructed

by plotting the dpm values from the LSC of the HPLC

fractions versus time post-injection using Microsoft Excel

(Microsoft, Redmond, WA). The relative abundance of drug-

related materials observed in the chromatogram (expressed as

a percent of total radioactivity in the chromatogram) were

determined by dividing the sum of the radioactive content of

fractions deemed to contribute to a particular peak by the sum

of the radioactive content of all fractions used to construct the

radio chromatogram. To assign the relative abundance of

drug-related materials in plasma resolved using HPLC

method 2, the percent of total radioactivity corresponding to

each resolved peak was normalized by multiplying by the

relative abundance of radioactivity corresponding to the early

eluting fractions (55 min) from the profile generated using

HPLC method 1 (i.e. 24%).

For urine and feces samples, the radioactivity in the HPLC

column effluent was measured using an IN/US System Model

3 b-ram radio-HPLC detector equipped with a 600 mL flow cell

operated in stop-flow modes using an AIM research Company

model C StopFlow System controller. The fraction time was

8.64 s, the scintillation cocktail was StopFlow AD (AIM

Research Company, Hockessin, DE), and the HPLC effluent:

scintillation cocktail ratio was 1:1.5. The count time was 120 s

for urine and 20 s for feces. Urinary and fecal metabolite

abundance (expressed as percent of administered dose) was

calculated by multiplying the fractional contribution of

radioactive response for the peak in the radio-HPLC chro-

matogram by the percent of administered dose recovered in the

matrix.

Fractionation of samples for metabolite identification

Plasma

Fractions generated from radio-HPLC profiling containing

radioactivity, as described previously, were analyzed directly

by LC/MS to identify metabolites.

Urine

Pooled urine from two subjects (2.5 mL from each) was

extracted as described previously. The evaporated extract was

reconstituted in 2 mL of acetonitrile/water (2:8) containing

0.1% formic acid. Repeated injections (10� 100 mL) of the

urine extract were performed using HPLC method 1 with the

effluent from each run collected into a single set of

scintillation vials such that the same time interval for all

runs was collected in the same vial. Fractions were collected

at 0.33 min intervals from 0 to 18 min. Aliquots of each

fraction were removed for analysis by LC/MS and the

radioactivity in the remaining fractions was determined by

LSC using a 5 min count time. Fractions eluting before 5 min

that contained significant radioactivity were combined,

evaporated to dryness and reconstituted in 600 mL of aceto-

nitrile/water (2:8) containing 0.1% formic acid. Repeated

injections (3� 200 mL) were performed using HPLC method

2 with the effluent from each run collected into a single set of

scintillation vials. Fractions were collected at 0.25 min

intervals from 0 to 44 min. Aliquots of each fraction were

removed for analysis by LC/MS and the radioactivity in the

remaining fractions was determined by LSC using a 15 min

count time. Fractions containing radioactivity from the two

chromatographic methods were analyzed by LC/MS to

identify metabolites.

Feces

Samples of pooled feces were extracted as described previ-

ously. The evaporated extract was reconstituted in 0.8 mL of

acetonitrile/water (2:8) containing 0.1% formic acid.

Repeated injections (4� 200 mL) of the fecal extract were

performed using HPLC method 1 with the effluent from each

run collected into a single set of scintillation vials. Fractions

were collected at 0.5 min intervals from 0 to 50 min. Aliquots

of each fraction were removed for analysis by LC/MS and the

radioactivity in the remaining fractions was determined by

LSC using a 5 min count time. Fractions eluting before 5 min

that contained significant radioactivity were combined,

evaporated to dryness and reconstituted in 600 mL of aceto-

nitrile/water (2:8) containing 0.1% formic acid. These fecal

extracts were fractioned by HPLC method 2 as described for

the urine samples. Fractions containing radioactivity from the

two chromatographic methods were analyzed by LC/MS to

identify metabolites.

Metabolite identification by LC/MS

Metabolite identification by mass spectrometry was per-

formed using a hybrid LC/MS system consisting of two

Shimadzu pumps, a CTC Analytics autosampler, an Agilent

1100 series diode array detector and an API 4000 QTRAP

mass spectrometer (Applied Biosystems, Foster City, CA).

All components of the system were controlled using Analyst

1.4.1 (Build 1200) software (Applied Biosystems). For all

analyses, the mass spectrometer was operated in

TurboIonSpray mode with conditions adjusted as necessary.

Scan modes utilized included enhanced MS, multiple reaction

monitoring, neutral loss, precursor ion and MS2. Fractions

were analyzed by direct infusion and/or LC/MS. Infusions

were conducted with a Harvard Apparatus Pump 11 syringe

pump. For LC/MS analyses a Synergi Polar-RP column

(2.0� 250 mm, 4 m) was utilized for metabolite separation at

flow rate of 0.2 mL/min flow and ambient temperature. The

HPLC mobile phase and gradient was the same as described

for HPLC method 2.

Quantification of the diastereomers of crizotiniblactam (M10) in plasma

As there was insufficient plasma volume remaining from the

study, the quantification of diastereomers of crizotinib lactam

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 5

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 6: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

was conducted by using plasma samples from a sub-set of

subjects (n¼ 4) receiving a single oral 250-mg dose of

crizotinib from a relative bioavailability study (A8081008;

ClinicalTrials.gov identifier NCT00939731). Blood samples

were collected by an indwelling catheter or venipuncture into

K2EDTA vacutainer tubes pre-dose and at 0.5, 1, 2, 3, 4, 6, 8,

10, 12, 16, 24, 36, 48, 60, 72, 84 and 96 h post-dose. Plasma

was prepared from blood samples as described previously, and

samples were stored at or below�20 �C until the time of

analysis.

Concentrations of the constituent diastereomers of crizo-

tinib lactam (M10, PF-06270079 and PF-06270080) in

plasma samples were determined with a non-validated

LC–MS/MS analytical method. Briefly, internal standard

([2H5]crizotinib, PF-03623192) was added to 100mL of

plasma followed by addition of 100mL of sodium bicarbonate

buffer, pH 7.4. Two hundred mL of each sample was added

into an Isolute solid–liquid extraction (SLE)+ 200 mg plate,

and a light vacuum was applied to the SLE block. Samples

were eluted with 1.2 mL of methyl tert-butyl ether and

evaporated under nitrogen. Samples were reconstituted with

100mL of water/acetonitrile (1:1, v/v). Aliquots (5 mL) of

plasma extracts were injected onto ChiralPAK, RH-AS

column (5mm, 150� 4.6 mm; Chiral Technologies, Inc.,

West Chester, PA) maintained at ambient temperature.

Analytes were eluted using an isocratic mobile phase of

water/acetonitrile (48:52, v/v) containing 10 mM ammonium

acetate and a flow rate of 0.4 mL/min. Detection was

performed by positive ion MS/MS on a SciEx API 4000

mass spectrometer (Applied Biosystems, Foster City, CA)

with multiple reaction monitoring (m/z 464 to 274 for

PF-06270079 and PF-06270080 and m/z 455 to 265 for

internal standard). The LLOQ for PF-06270079 and

PF-06270080 measurement was 0.1 ng/mL. Calibration stand-

ard responses were linear over the range of 0.1–125 ng/mL,

using a weighted (l/concentration2) linear least squares

regression. Calibration standards and QC samples (prepared

in triplicate at 0.75, 15 and 100 ng/mL) were within ± 20% of

the nominal value.

Cellular pharmacodynamic assays

Crizotinib and its metabolites, crizotinib lactam

(PF-06260182), the constituent diastereomers of crizotinib

lactam (PF-06270079 and PF-06270080), O-desalkyl crizoti-

nib (PF-03255243) and O-desalkyl crizotinib lactam

(PF-06268935), were evaluated for their ability to inhibit

the phosphorylation of ALK fusion variants and c-Met in cells

lines expressing these targets. Inhibition of ALK was

evaluated in NCI-H3122 and NCI-H2228 cells, two human

lung adenocarcinoma cell lines harboring chromosomal

inversion events resulting in the expression of EML4-ALK

fusion protein variant 1 (V1) and variant 3 (V3), respectively

(Koivunen et al., 2008). Inhibition of c-Met phosphorylation

was evaluated in human A549 lung adenocarcinoma cells.

Briefly, cells were seeded in 96-well plates in medium

supplemented with 10% fetal bovine serum and transferred to

serum-free medium (with 0.04% bovine serum albumin) after

being cultured overnight. After incubation of cells with an

inhibitor for 1 h, cells were washed once with Hank’s

balanced salt solution supplemented with 1 mM Na3VO4

and protein lysates were generated from cells. In experiments

investigating ligand-dependent RTK phosphorylation, corres-

ponding growth factors were added for up to 20 min.

Subsequently, phosphorylation of selected protein kinases

was assessed by sandwich enzyme-linked immunosorbent

assays using specific capture antibodies used to coat 96-well

plates and a detection antibody specific for phosphorylated

tyrosine residues as described previously (Cui et al., 2011;

Zou et al., 2007). EC50 values were calculated by concentra-

tion–response curve fitting using a Microsoft Excel-based

four-parameter method. Statistical significance of differences

in mean EC50 values was determined by an unpaired two-

sample t-test using Microsoft Excel.

Pharmacological activity index

The pharmacological activity index (PAI) (Leclercq et al.,

2009) was used as a framework within which the potential

contribution of metabolites to pharmacological activity could

be evaluated. The PAI index was calculated as:

PAI ¼ AUCm

AUCp

� fum

fup

� EC50, p

EC50, m

where, AUCm and AUCp are the area under the plasma

concentration–time curves for metabolite and parent, respect-

ively; fum and fup are the unbound fraction values in plasma

for metabolite and parent, respectively; and EC50,p and EC50,m

are the half maximal inhibitory concentrations for parent and

metabolite, respectively, in the cellular pharmacodynamic

assays.

Safety assessments

Safety evaluations included clinical monitoring, collection of

subject-reported adverse events, including serious adverse

events, vital signs, 12-lead electrocardiograms and clinical

laboratory tests.

Results

Subjects

Six male subjects (three African-American, two White and

one Hispanic) enrolled and completed the study. The mean

(SD) age of the subjects was 43.2 (1.8) years, and the mean

body weight was 85.2 (9.5) kg. In the relative bioavailability

study, samples from four male subjects were evaluated for

metabolite PK (three African-American, one other). The

mean age of these subjects was 29.5 (11.1) years, and the

mean body weight was 68.8 (9.0) kg.

Excretion and mass balance

The mean cumulative recovery of radioactivity in urine and

feces from 0 to 480 h post-dose in six healthy male subjects

administered a single 250-mg (100mCi) oral dose of

[14C]crizotinib is presented in Figure 2 and Table 1. The

overall mean recovery of drug-derived radioactivity was

85.2% of the dose with recovery in individual subjects ranging

from 68.6% to 91.3%. Radioactivity was recovered predom-

inantly in feces, with a mean (±SD) cumulative amount of the

dose recovered of 63.1 ± 5.9%. The mean cumulative amount

6 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 7: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

of the dose recovered in urine was 22.2 ± 4.6%. A majority of

the administered radioactivity was recovered in the first 120 h

post-dose (77.9 ± 9.4%).

Pharmacokinetics

The mean concentration–time profiles and pharmacokinetic

parameters of unchanged crizotinib (plasma) and total

radioactivity (plasma, whole blood and RBCs) in healthy

male subjects administered a single 250-mg (100 mCi) oral

dose of [14C]crizotinib are presented in Figure 3(A, B) and

Table 2, respectively. The peak concentrations of crizotinib

occurred between 2.0 and 6.0 h after oral dosing and the mean

Cmax was 109 ng/mL. The mean apparent t1/2 value for

crizotinib in plasma was 94 h and was longer than determin-

ations in other studies (29–51 h), likely due to the extended

sampling period of up to 408 h post-dose (Li et al., 2011;

Tan et al., 2010; Xu et al., 2011a). The variability, evaluated

by the CV, was 46 and 38% for crizotinib Cmax and AUC1,

respectively. Total radioactivity measured in plasma peaked

between 3.0 and 6.0 h after dosing, similar to the tmax for

crizotinib. Plasma exposure of crizotinib was considerably

lower than that of the total radioactivity. Unchanged drug

represented �12% of drug-related material in plasma based

on AUClast values, indicating the presence of circulating

metabolic products in plasma. The apparent t1/2 of total

radioactivity in plasma was prolonged relative to unchanged

drug (Figure 3A). Total radioactivity in whole blood paral-

leled the profile of total radioactivity in plasma up to 48 h, the

last quantifiable concentration of radioactivity in blood, with

peak radioactivity occurring at similar times (3.0–6.0 h after

dosing; Figure 3B). Mean RBC concentrations of total

Figure 3. Mean (±SD) concentration versus time profiles for total radioactivity and crizotinib in plasma (A) and total radioactivity (to 48 h) in plasma,blood and RBCs (B) after a single 250-mg oral dose of [14C]crizotinib to male volunteers (n¼ 6).

Table 2. Pharmacokinetic parameters of crizotinib in plasma and totalradioactivity in plasma, whole blood and RBC after a single 250-mg oraldose of [14C]crizotinib to six healthy male subjects.

Crizotinib Total radioactivityb

Parametera Plasma Plasma Whole Blood RBC

Cmax (ng/mL) 109 (46) 436 (19) 312 (20) 175 (26)tmax (h) 3 (2–6) 5 (3–6) 4 (3–6) 5 (3–8)AUClast

(ng�h/mL)2686 (40)c 22 830 (11)d 7032 (18)e 3641 (23)e

AUC1(ng�h/mL)

2777 (38) 29 000, 29 600f nc nc

t1/2 (h) 94.0 (15) 134, 178f nc nc

nc, not calculated.aGeometric mean (%CV) for Cmax and AUC; arithmetic mean (%CV) for

t1/2; median (range) for tmax; n¼ 6.bTotal radioactivity expressed in units of nanogram-equivalents per

milliliter (Cmax) or nanogram-equivalents�hour per milliliter (AUC).cTime of last quantifiable concentration was 312 h in three subjects,

336 h in one subject, 360 h in one subject and 408 h in one subject.dTime of last quantifiable concentration was 312 h in three subjects,

336 h in one subject and 360 h in two subjects.eTime of last quantifiable concentration was 36 h in one subject and 48 h

in five subjects.fIndividual values for the two evaluable subjects. For the remaining four

subjects, AUC1 and t1/2 could not be estimated.

Figure 2. Mean (±SD) cumulative excretion of total radioactivity inurine, feces and overall after a single 250-mg oral dose of [14C]crizotinibto male volunteers (n¼ 6).

Table 1. Cumulative excretion of total radioactivity in urine and fecesafter a single 250-mg oral dose of [14C]crizotinib to six healthy malesubjects.

Percent of radioactive dose recovered by subjecta

Matrix 1001 1002 1003 1004 1005 1006 Mean (SD)

Urine 22.6 28.8 15.1 23.0 24.3 19.2 22.2 (4.6)Feces 63.2 58.9 53.5 68.2 66.0 68.7 63.1 (5.9)Total 85.8 87.7 68.6 91.3 90.2 87.8 85.2 (8.4)

aTime period of excretion was �480 h.

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 7

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 8: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

radioactivity were 29–45% of plasma radioactivity concen-

trations at corresponding time points, indicating that crizoti-

nib or its metabolites do not preferentially distribute to RBCs.

A mean of 2.3% of the dose was excreted in urine as

unchanged crizotinib, based on a specific LC-MS/MS

assay for crizotinib, with a mean renal clearance estimate of

2.5 L/h.

Metabolite profiling

The metabolic profile in a pooled (0–96 h) plasma sample is

presented in Figure 4(A) and the distribution of metabolites is

summarized in Table 3. The recovery of radioactivity

following acetonitrile/methanol extraction of this pooled

plasma sample was �60%. Unchanged crizotinib was the

major circulating component, accounting for 33% of radio-

activity in the profile. M10 represented 10% of circulating

radioactivity and early eluting components (55 min)

accounted for 24% of circulating radioactivity. A second

pooled plasma sample was analyzed with HPLC conditions

optimized to retain and resolve these early eluting compo-

nents and the resulting metabolic profile is shown in

Figure 4(B). The early eluting peaks were composed of

multiple components (M1, M2, M3 and M8), each accounting

for510% of the recovered radioactivity individually.

Representative radio-HPLC profiles of pooled feces

samples and pooled urine samples are shown in Figure 5(A

and B), respectively. Unchanged crizotinib was the major

excreted component in feces, accounting for an average of

53% of total administered dose (Figure 5A). In urine, an

average of 2.4% of the dose was excreted as unchanged

crizotinib, based on radio-HPLC profiling. M8 was the major

excreted component in urine, accounting for an average of

4.5% of dose (Figure 5B). No other metabolites accounted for

�1% of the total administered dose in excreta.

Metabolite identification

Nine metabolites were identified in this study. Metabolite

structures were elucidated by LC–MS analysis (full scan

and MS2), and in the case of M2 (PF-06268935), M4

(PF-03255243) and M10 (PF-06260182), by comparison to

Figure 4. HPLC radiochromatograms of pooled human plasma (0–96 h)after a single 250-mg oral dose of [14C]crizotinib to male volunteers(n¼ 6). Primary full profile (A) and secondary profile resolving earlyeluting (55 min) components in the primary profile (B).

Figure 5. Representative HPLC radiochromatograms of individualsubject-pooled (0–120 h) human feces (A) and urine (B) after a single250-mg oral dose of [14C]crizotinib to male volunteers.

Table 3. Relative percent distribution of radioactivemetabolites in 0–96 h-pooled plasma after a single250-mg oral dose of [14C]crizotinib to six healthy malesubjects.

ComponentPercent of radioactivity

(0–96 h)a

M1 3.0M2 1.1M3 2.5M8 0.8M10 10.2Crizotinib 33.1Total 50.7

aPercent of total radioactivity in 0–96 h plasma AUC.

8 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 9: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

retention time and MS fragmentation of authentic metabolite

standards. The MS fragmentation patterns of crizotinib and

metabolites are presented in Table 4. The proposed metabolic

pathways of crizotinib in humans is presented in Figure 6 and

identification of metabolites is described in order of primary

metabolites, followed by secondary products, as outlined in

this scheme.

The fragmentation pattern of crizotinib was studied to

facilitate interpretation of mass spectra of its metabolites. The

protonated molecular ion for crizotinib was at m/z 450.

The collision-induced dissociation MS2 spectrum showed

fragment ions at m/z 367, 260, 177 and 84. The fragment ion

at m/z 367 resulted from a loss of the piperidine ring from the

molecule, whereas fragment ions at m/z 260 and 177 were

Table 4. Mass spectral data and proposed structures of crizotinib metabolites.

Metabolite ID/MS data Proposed structure Metabolite ID/MS data Proposed structure

Crizotinib [M + H]+: 450Product ions (m/z): 367, 260, 177, 84

OF

N N

NCl

Cl

NH

84367

260

+2H

+2H+2H

+H

+H 177

NH2

M5 [M + H]+: 452Product ions (m/z): 276

N N

NHO

HO

OHO

HOHOOC

OH

NH

276+H

NH2

M1 [M + H]+: 436Product ions (m/z): 353, 260, 177

O

N N

NHO

HO353 HO

HOHOOC

NH

177

260

+2H

+H

+H

NH2

M6 [M + H]+: 290Product ions (m/z): 177

N N

NHO

OOHNH

177

+2H

NH2

M2 (PF-06268935) [M + H]+: 274Product ions (m/z): 177, 98

ON N

NHO

NH

98

177

+2H

+H

NH2

M8 [M + H]+: 354Product ions (m/z): 274, 177, 98

N N

NHO

OO

OSHO

NH98

274

177

+2H

+H

+H

NH2

M3 [M + H]+: 340Product ions (m/z): 260, 177

O

OSHO

N N

NHO

NH

260

177

+2H

+H

NH2

M9 [M + H]+: 393Product ions (m/z): 296, 279, 235, 209, 98

N N

NHO

O

O

SHO

NH

98

279

209235

296

+2H

+2H

+3H+2H

+H

NH2

NH2

M4 (PF-03255243) [M + H]+: 260Product ions (m/z): 177, 84

N N

NHO

NH

84

177

+2H

+H

NH2

M10 (PF-06260182) [M + H]+: 464Product ions (m/z): 367, 274, 177, 98

NH2

ON

NN

NH

ClF

Cl

O367

274

98

177

+2H

+2H

+H

+2H

+H

PF-number, synthetic standard available.

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 9

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 10: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

products of cleavage of the dichloro-fluorobenzyl group and

subsequent loss of the piperidine ring, respectively. The

fragment ion at m/z 84 was indicative of the piperidine

moiety.

Metabolite M10

M10 (PF-06260182) was detected in plasma and feces

(MS only). The protonated molecular ion of M10 was at

m/z 464 and indicated the addition of 14 amu to the

molecular ion of crizotinib. The mass spectrum showed

fragment ions at m/z 367, 274, 177 and 98. The ions at

m/z 367 and 177 were characteristic of ions present in

the mass spectrum of crizotinib, as described previously.

The ions at m/z 274 and 98 showed additions of 14 amu

to characteristic fragment ions of m/z 260 and 84 in the

spectrum of crizotinib, suggesting oxidation of the

piperidine ring.

Metabolite M4.

M4 (PF-03255243) was detected in feces (early eluting

component; radiochromatographic data not shown). The

protonated molecular ion of M4 was at m/z 260, indicative

of an O-desalkyl metabolite of crizotinib. Fragments at

177 and 84 were consistent with the fragmentation spectrum

of crizotinib, as described previously.

Metabolite M1

M1 was detected in plasma, urine and feces (early eluting

component in excreta; radiochromatographic data not shown).

The protonated molecular ion of M1 was at m/z 436, 176 amu

greater than M4 (O-desalkyl crizotinib), suggesting that M1

was a glucuronide conjugate of M4. The mass spectrum of the

metabolite showed fragment ions at m/z 353, 260 and 177.

The ions at m/z 260 and 177 were similar to those observed in

the mass spectrum of crizotinib. The ion at m/z 353 suggested a

loss of the piperidine ring from the molecule and indicated

that the glucuronide was not conjugated to the piperidine ring

and was possibly attached to the hydroxyaminopyridine

moiety.

Metabolite M3

M3 was detected in plasma, urine and feces (early eluting

component in excreta; radiochromatographic data not shown).

The protonated molecular ion of M3 was at m/z 340, 80 amu

greater than M4 (O-desalkyl crizotinib). This molecular ion

was indicative of a sulfate conjugate of M4. The mass

ON

NN

NH

ClF

Cl NH2 NH2

Crizotinib

HON

NN

NH

NH2

O

HON

NN

NH

NH2

O

O N

NN

NH

ClF

Cl

O

SHOO

O

HON

NN

NH

NH2

O

HO

HOOCHO

HO

M8

M2

M10

M1

HON

NN

NH

NH2

O

M9

HON

NN

NH

NH2

M4

HON

NN

NH

NH2

SHOO

O

M3

HON

NN

NH

NH2

O

M6

OH

HON

NN

NH

NH2

M5

O

HO

HOOCHO

HOOH

SHOOC

NH2

Figure 6. Proposed biotransformation pathways of crizotinib based on metabolites identified in plasma, urine and fecal samples from male volunteersadministered a single 250-mg oral dose of [14C]crizotinib.

10 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 11: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

spectrum showed fragment ions at m/z 260 and 177, which

were consistent with those observed in the mass spectrum of

crizotinib, as described previously.

Metabolite M5

M5 was detected in urine (early eluting component; radio-

chromatographic data not shown). The protonated molecular

ion of M5 was at m/z 452. The molecular ion was 192 amu

greater than M4 (O-desalkyl crizotinib), indicative of a

hydroxylation and glucuronidation of M4. A fragment ion at

m/z 276 was detected, indicative of the characteristic neutral

loss of 176 amu from a glucuronide conjugate. Moreover, this

fragment (m/z 276) was 16 amu greater than the molecular ion

of M4 (m/z 260), indicating a hydroxylation of O-desalkyl

crizotinib.

Metabolite M2

M2 (PF-06268935) was detected in plasma, urine and feces

(early eluting component in excreta; radiochromatographic

data not shown). The protonated molecular ion of M2 was at

m/z 274. This molecular ion was 14 amu greater than the ion

of M4 (O-desalkyl crizotinib), indicative of a lactam metab-

olite of M4. The absence of a fragment ion at m/z 84 and the

presence of an ion at m/z 98 suggested that the piperidine was

the site of oxidation.

Metabolite M6

M6 was detected in feces (early eluting component; radio-

chromatographic data not shown). The protonated molecular

ion of M6 was at m/z 290. The molecular ion was 16 amu

greater than M2 (O-desalkyl crizotinib lactam), indicative of a

hydroxylation of M2. The fragment at m/z 177 was consistent

with the fragmentation of crizotinib and indicated that the

oxidation occurred on the piperidine ring.

Metabolite M8

M8 was detected in plasma and urine. The protonated

molecular ion of M8 was at m/z 354. The molecular ion

was 80 amu greater than M2 (O-desalkyl crizotinib lactam),

indicative of a sulfate conjugate of M2. The mass spectrum of

the metabolite showed fragment ions at m/z 274, 177 and 98.

The fragment ion at m/z 274 results from the loss of the

sulfate moiety. The fragment ion at m/z 177 was consistent

with the fragmentation pattern of crizotinib, whereas, the ion

at m/z 98 represented the oxopiperidine moiety.

Metabolite M9

M9 was a trace metabolite detected in urine (MS only). The

protonated molecular ion of M9 was m/z 393. This molecular

ion was 119 amu greater than M2 (O-desalkyl crizotinib

lactam), indicative of a cysteine conjugate of M2. The

fragment at m/z 98 represented the oxopiperidine moiety,

and the fragment at m/z 296 resulted from the loss of

this moiety. This fragmentation pattern indicated that the

cysteine was not conjugated to the oxopiperidine ring and was

possibly attached to the hydroxyaminopyridine moiety.

Additional fragment ions at m/z 279, 235 and 209 were

attributed to loss of the oxopiperidine moiety and cleavages of

the cysteine. M9 is proposed as a downstream metabolite of a

glutathione adduct of M2. A proposed mechanism of

nucleophilic attack by glutathione via a quinone imine

intermediate leading to the formation of M9 is presented in

Figure 7.

Plasma pharmacokinetics of the diastereomers ofcrizotinib lactam (M10)

The mean plasma concentration–time profiles and pharma-

cokinetic parameters of crizotinib and the diastereomers of

crizotinib lactam (M10), PF-06270079 and PF-06270080,

from an exploratory assessment of four healthy male subjects

administered a single 250-mg oral dose of crizotinib as part of

a relative bioavailability study are presented in Figure 8 and

Table 5, respectively. Plasma exposure (as assessed by Cmax

and AUC1) of PF-06270079 was greater than PF-06270080,

with a mean AUC1 ratio of 1.66 (range: 1.59–1.76). The

mean metabolite:parent AUC1 ratios (%) for PF-006270079

and PF-06270080 were 14% and 8%, respectively. Values of

tmax for the lactam metabolites (range: 3–6 h) were similar to

those observed for crizotinib. The mean apparent t1/2 values

for PF-006270079 and PF-06270080 were similar (20.3 and

20.2 h, respectively).

Activities of crizotinib and its primary metabolitesagainst ALK and c-Met kinases

The inhibitory activity of crizotinib and its primary metab-

olites M2 (O-desalkyl crizotinib lactam, PF-06268935), M4

(O-desalkyl crizotinib, PF-03255243) and M10 (crizotinib

lactam, constituent diastereomers PF-06270079 and

PF-06270080) against phosphorylation of ALK fusion vari-

ants and c-Met was evaluated in cell lines expressing these

targets, as summarized in Table 6. Crizotinib was a potent

inhibitor of these target kinases in vitro, with mean EC50

values of 63 nM, 74 nM and 5 nM in cells lines expressing

EML4-ALK V1, EML4-ALK V3 and c-Met, respectively.

M10 (crizotinib lactam) was �2.5- to 7.7-fold less potent than

crizotinib against ALK and 2.5- to 4-fold less potent against

c-Met. One of the constituent diastereomers of M10

(PF-06270080) was more potent compared with the other

diastereomer (PF-06270079). The two O-desalkyl crizotinib

metabolites, M2 and M4, did not inhibit EML4-ALK or

c-Met activity in cell-based assays at concentrations up to

10 mM.

Safety summary

There were no serious adverse events, severe adverse events

or deaths during the clinical studies. In addition, there were no

permanent or temporary discontinuations due to adverse

events. No laboratory test results, vital signs or electrocar-

diogram abnormalities met the criteria for being clinically

relevant, and therefore, none were recorded as adverse events.

In the human absorption, distribution, metabolism and

excretion (ADME) study, all subjects experienced treatment-

related gastrointestinal adverse events; however, all adverse

events observed during the study were considered mild in

severity.

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 11

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 12: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

ON

NN

NH

HN

H

H

ON

NN

NH

NH

ON

NN

NH

NH2

H

GS

quinone-imine

HON

NN

NH

NH2

SHOOC

NH2

O

M9

ON

NN

NH

ClF

ClNH2

O O

OO

ON

NN

NH

NH2

O

SGH

HON

NN

NH

NH2

O

M9

SCOOH

NH2

M10 M2

ON

NN

NH

ClF

ClNH2

Crizotinib

GSH

(or)

−H2

GSH

[O] [O]

[O]

GSH

GSH

Figure 7. Proposed glutathione conjugation mechanism via a quinone imine intermediate leading to the formation of the cysteine conjugatemetabolite (M9).

Figure 8. Mean ( ± SD) concentration versus time profiles for crizotiniband the diastereomers of the lactam metabolite (M10), PF-06270079 andPF-06270080, in plasma after a single 250-mg oral dose of crizotinib tomale volunteers (n¼ 4).

Table 5. Pharmacokinetic parameters of crizotinib and the diastereomersof the lactam metabolite (M10), PF-06270079 and PF-06270080, inplasma after a single 250-mg oral dose of crizotinib to four healthy malesubjects.

Parameter Crizotinib PF-06270079 PF-06270080

Cmax (ng/mL) 124 (55) 32.5 (51) 20.7 (55)tmax (h) 5 (3–6) 5 (3–6) 5 (4–6)AUClast (ng�h/mL) 2948 (53) 443 (66) 272 (69)AUC1 (ng�h/mL) 3000 (53) 451 (66) 277 (69)t1/2 (h) 31.7 (15) 20.3 (11) 20.2 (35)MRAUC (%)a na 14 (18) 8 (22)

Arithmetic mean (percent CV); median (range) for tmax; n¼ 4.na¼ not applicable.aMetabolite-to-parent ratio based on AUC1, expressed as a percentage.

AUC1 values were corrected for the molecular weight of crizotnib andlactam metabolite diastereomers (i.e. ng�h/mL converted to nM�h).

12 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 13: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

Discussion

The utility of conducting human ADME studies with

radiolabeled drugs has been extensively reviewed (Beumer

et al., 2006; Penner et al., 2009; Roffey et al., 2007). The

major objectives of these studies include determination of:

(1) mass balance and routes of elimination, (2) circulatory and

excretory metabolites, (3) exposure of parent drug and

metabolites in circulation, (4) the potential contribution of

metabolites to the pharmacological/toxicological effects,

(5) clearance mechanisms and potential contributors to

interpatient variability and drug–drug interactions (DDIs)

and (6) the validity of the animal species used in toxicological

evaluation. All of these objectives were achieved for the

described investigation with crizotinib.

Following a single 250-mg/100 mCi dose of [14C]crizotinib

to healthy human subjects, the overall mean recovery of the

radioactive dose in the 20-day collection period was 85%,

with 78% recovering in the first 5 days. In a retrospective

analysis of human ADME studies with 27 radiolabeled

compounds, the overall mean (±SD) total recovery value

was 87 ± 10% (Roffey et al., 2007). Therefore, the extent of

recovery in the present study was well within this historical

range, suggesting that mass balance was achieved.

[14C]Crizotinib-related radioactivity was primarily recovered

in the feces after oral administration (63% of dose), consistent

with findings in nonclinical mass balance studies (Smith

et al., 2011).

Crizotinib represented �12% (AUClast) to 25% (Cmax) of

drug-related material in plasma. The radioactive profile of a

0–96 h Hamilton pool of plasma from all six subjects revealed

that unchanged crizotinib was the major circulating compo-

nent, representing 33% of radioactivity and crizotinib lactam

(M10) accounted for 10% of circulating radioactivity. No

other single circulating component accounted for >10% of

radioactivity in plasma over the time interval profiled. Other

minor metabolites included O-desalkyl metabolites and

related conjugates. The difference in the percentage of

drug-related material attributed to crizotinib using the two

orthogonal methods (i.e. 12% versus 33%) could potentially

be due to: (1) different time-intervals assessed in the two

methods (i.e. �312 h for AUClast versus 96 h for metabolite

profiling) and/or (2) non-extractable radioactivity in plasma

impacting the value obtained via metabolite profiling. In total,

�51% of the radioactivity in human plasma was accounted for

by crizotinib and identified metabolites. The remaining

radioactivity was attributed to multiple minor metabolites.

The low levels of radioactivity in plasma in the 0–96 h

Hamilton pool (�132 dpm/g) and the wide range of

metabolite polarity made identifying components with

510% abundance extremely challenging. This illustrates one

limitation of the Hamilton pooling technique when working

with low levels of radioactivity, because the sample is diluted

making low abundance metabolites even more difficult to

quantify and/or identify without resorting to high sensitivity

methods such as accelerator mass spectrometry.

Unchanged crizotinib was the major component in feces

(53% of dose). This significant recovery of unchanged drug in

the feces likely represents unabsorbed parent drug. Although

biliary excretion of intact crizotinib in humans cannot be

conclusively ruled out, unchanged crizotinib was not observed

in bile collected from rats following oral administration of

[14C]crizotinib (Smith et al., 2011). Therefore, biliary excre-

tion of unchanged crizotinib is not anticipated to be a

significant clearance pathway in humans, as rats are typically

considered to excrete xenobiotics into bile more readily than

humans (Kwon, 2001). Moreover, intestinal deconjugation of

direct glucuronide metabolites of crizotinib and/or reduction

of crizotinib N-oxide is not anticipated to play a significant

role in crizotinib clearance, as these metabolites were not

identified in the present study or in rats or dogs (Smith et al.,

2011). In urine, an average of �2% of the dose was excreted

as unchanged crizotinib based on radiochromatographic or

LC–MS/MS quantification methods. The major excreted

component in urine (4.5% of dose) was the sulfate conjugate

of O-desalkyl crizotinib lactam (M8). The remaining radio-

activity in urine was attributed to multiple minor metabolites,

each accounting for51% of the administered dose.

Upon oxidation of crizotinib to crizotinib lactam (M10), a

new chiral center is introduced, resulting in the formation of

two possible diastereomers of M10, PF-06270079 and

PF-06270080. In plasma samples from a subset of subjects

(n¼ 4) administered a single oral 250-mg dose of crizotinib in

a relative bioavailability study, PF-06270079 exposure

Table 6. In vitro pharmacodynamic inhibition of target tyrosine kinases and fusion variants by crizotinib and the diastereomers of the lactam metabolite(M10), PF-06270079 and PF-06270080 and corresponding PAIs.

Cellular kinase assay Compound EC50 (nM)a nb EC50 M:P ratioc,d p Valued PAI (%)e

EML4-ALK V1 phosphorylation Crizotinib 63 ± 31 20PF-06270079 284 ± 182 7 3.7 50.0001 2PF-06270080 194 ± 118 6 2.5 50.0001 2

EML4-ALK V3 phosphorylation Crizotinib 74 ± 23 6PF-06270079 554 ± 102 3 7.5 50.0001 1PF-06270080 355 ± 66 3 5.0 0.0026 1

c-Met phosphorylation Crizotinib 5.0 ± 3.8 164PF-06270079 18.5 ± 7.3 3 4.0 0.0065 2PF-06270080 11.8 ± 2.7 3 2.5 0.0308 2

aMean ± SD.bNumber of replicate kinase inhibition experiments.cMetabolite-to-parent ratio for EC50.dRatios and p values were derived by comparing EC50 values of crizotinib versus metabolites within the same experiment.ePAI expressed as a percentage. For calculation of unbound exposures, fraction unbound in human plasma was 0.093 for crizotinib (Yamazaki et al.,

2011). For PF-06270079 and PF-06270080, fraction unbound values in human plasma were 0.055 and 0.059, respectively (unpublished data).

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 13

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 14: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

exceeded that of PF-06270080 (mean AUC ratio of 1.66;

range: 1.59–1.76). The AUC values of PF-06270079 and PF-

06270080 represented 14% and 8%, respectively, of the AUC

of crizotinib. Therefore, the overall ratio of M10 to crizotinib

based on this exploratory analysis (22%) was in general

agreement with the ratio of derived from the radioactive

plasma profile (30%). In vitro, PF-06270079 and

PF-06270080 were �2.5- to 7.7-fold less potent than

crizotinib in ALK and c-Met cellular pharmacodynamic

assays. O-Desalkyl crizotinib (M4, PF 03255243) and

O-desalkyl crizotinib lactam (M2, PF 06268935) were

inactive against both ALK and c-Met kinases

(EC50> 10 mM). This finding is consistent with the known

SAR for crizotinib, where the 3-benzyloxy group is known to

be a critical structural feature for binding to the hydrophobic

pocket of the c-Met kinase (Cui et al., 2011). It has been

proposed that a PAI, vide supra, of >25% can be used as

a criterion to assess the significance of contribution of a

metabolite to the pharmacology of the parent drug (Leclercq

et al., 2009). The PAI values for both PF-06270079 and

PF-06270080 are estimated to be �2% (Table 6). Based on

this preliminary assessment, neither metabolite is anticipated

to contribute significantly to the pharmacological activity of

crizotinib. However, the limitations of this exploratory

assessment include data obtained from a low number of

healthy subjects following a single 250-mg crizotinib dose.

Given that crizotinib exhibits time-dependent PK, likely due

to auto-inhibition/inactivation of CYP3A, and is also primar-

ily metabolized by CYP3A, subsequent studies have included

the measurement of plasma M10 concentrations following

multiple doses to better evaluate its potential contribution to

in vivo activity at steady state.

The long-lived radioactivity in human plasma, relative to

parent drug and the incomplete extraction recovery of total

radioactivity from plasma (�60%) may suggest the potential

for a proportion of crizotinib-derived radioactivity to bind to

plasma proteins and/or tissue macromolecules. In addition,

the observation of a trace cysteine conjugate metabolite (M9),

a possible downstream metabolite of a glutathione adduct of

M2, in urine suggests the potential for a low level of

bioactivation of crizotinib in vivo (Figure 7). To date, the

nature of this apparent binding has not been fully character-

ized. In this study, the mean recovery of total radioactivity

was 85% of the dose, indicating that any apparent binding to

macromolecules would comprise only a small fraction of the

total crizotinib dose (Roffey et al., 2007).

Based on the excretion profile, where a majority of the

radioactive dose was recovered in feces with limited urinary

recovery of unchanged crizotinib, and the identity of metab-

olites, the primary elimination pathway for crizotinib in

humans appears to be via hepatic clearance, specifically,

oxidative metabolism. In vitro studies demonstrated that

crizotinib was predominantly metabolized by CYP3A, which

also mediated the formation of the major oxidative lactam

(M10) and O-desalkyl (M4) metabolites to a significant extent

(Figure 6) (Johnson et al., 2011). Based on these in vitro data,

the fraction of systemically available crizotinib metabolized

by CYP3A (fm,CYP3A) was estimated to be �0.8. In humans, a

major role for CYP3A in the clearance of crizotinib was

confirmed in clinical studies with the potent CYP3A inhibitor

ketoconazole and inducer rifampin (Xu et al., 2011b,c). Using

the fm,CYP3A value derived in vitro, the observed magnitude of

the interaction between crizotinib and ketoconazole (AUC

ratio¼ 3.2) or rifampin (AUC ratio¼ 0.18) was reasonably

predicted using mechanistic static models (Ito et al., 1998;

Johnson et al., 2011; Shou et al., 2008). Thus, the under-

standing of the major clearance mechanism and metabolic

pathways derived from the present study were consistent with

findings from clinical DDI studies.

Circulating metabolites observed in humans were also

observed in at least one of the nonclinical safety species, rats

and dogs, confirming that the choice of these species for

nonclinical safety assessment was acceptable (Smith et al.,

2011). Qualitatively, the metabolism of crizotinib in rats was

representative of that observed in humans. Quantitatively,

plasma exposure of the M10 diastereomers in the rat was

similar to or exceeded human exposure (unpublished data).

Conclusion

The disposition of crizotinib has been characterized in healthy

subjects to augment understanding of its complex PK. After a

single oral dose of [14C]crizotinib, mass balance was

achieved, with a majority of drug-related radioactivity

recovered in the feces. The primary clearance pathway for

crizotinib is hepatic elimination/oxidative metabolism, and

the major metabolic pathways in humans were oxidation of

the piperidine ring to crizotinib lactam (M10) and

O-dealkylation, with subsequent conjugation of O-desalkyl

metabolites. Unchanged crizotinib and M10 were the princi-

pal drug-related components in circulation. Exploratory

assessments suggest that M10 is unlikely to be a major

contributor to crizotinib’s pharmacological activity.

Definitive assessments of the clinical relevance of M10 are

ongoing.

Acknowledgements

We thank personnel from Covance Laboratories Inc.

(Madison, WI and Indianapolis, IN) for radioanalysis support

of the human ADME study and for quantitation of crizotinib

in human plasma and urine samples; Yao (Grace) Ni for

facilitating clinical study conduct; Chunze Li for preliminary

pharmacokinetic analyses; Nicoletta Brega for clinical sup-

port; and Klaas Schildknegt for preparation of radiolabeled

crizotinib. The support of these colleagues is gratefully

acknowledged.

Declaration of interest

The authors report no conflicts of interest. The authors alone

are responsible for the writing and content of this article.

These studies were sponsored by Pfizer Inc. All authors are

employees of Pfizer Inc. or were at the time of conducting

these studies.

References

Beumer JH, Beijnen JH, Schellens JHM. (2006). Mass balance studies,with a focus on anticancer drugs. Clin Pharmacokinet 45:33–58.

Cui JJ, Botrous I, Shen H, et al. (2008). Structure based drug design forthe discovery of clinical candidate PF-2341066 as potent and highly

14 T. R. Johnson et al. Xenobiotica, Early Online: 1–15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.

Page 15: Metabolism, excretion and pharmacokinetics of [               14               C]crizotinib following oral administration to healthy subjects

selective c-Met inhibitor. Abstracts of Papers, 235th ACS NationalMeeting 2008 April 6–10; New Orleans, LA. Washington, DC:American Chemical Society, Abstract MEDI 177.

Cui JJ, Tran-Dube M, Shen H, et al. (2011). Structure based drug designof crizotinib (PF-02341066), a potent and selective dual inhibitor ofmesenchymal-epithelial transition factor (c-MET) kinase and anaplas-tic lymphoma kinase (ALK). J Med Chem 54:6342–63.

Gandhi L, Janne PA. (2012). Crizotinib for ALK-rearranged non-smallcell lung cancer: a new targeted therapy for a new target. Clin CancerRes 18:3737–42.

Hamilton RA, Garnett WR, Kline BJ. (1981). Determination of meanvalproic acid serum level by assay of a single pooled sample.Clin Pharmacol Ther 29:408–13.

Herbst RS, Heymach JV, Lippman SM. (2008). Lung cancer. N Engl JMed 359:1367–80.

Ito K, Iwatsubo T, Kanamitsu S, et al. (1998). Prediction ofpharmacokinetic alterations caused by drug-drug interactions: meta-bolic interaction in the liver. Pharmacol Rev 50:387–411.

Janku F, Stewart DJ, Kurzrock R. (2010). Targeted therapy in non-small-cell lung cancer – is it becoming a reality? Nat Rev Clin Oncol7:401–14.

Johnson T, Zhou S, Lee C, et al. (2011). Identification of the humanenzymes involved in the oxidative metabolism of crizotinib. DrugMetab Rev 43:160.

Koivunen JP, Mermel C, Zejnullahu K, et al. (2008). EML4-ALK fusiongene and efficacy of an ALK kinase inhibitor in lung cancer. ClinCancer Res 14:4275–83.

Kwak EL, Bang Y-J, Camidge R, et al. (2010). Anaplastic lymphomakinase inhibition in non-small-cell lung cancer. N Engl J Med 363:1693–703.

Kwon Y. (2001). Handbook of essential pharmacokinetics, pharmaco-dynamics, and drug metabolism for industrial scientists. New York:Kluwer Academic/Plenum Publishers.

Leclercq L, Cuyckens F, Mannens GSJ, et al. (2009). Which humanmetabolites have we MIST? Retrospective analysis, practical aspects,and perspectives for metabolite identification and quantification inpharmaceutical development. Chem Res Toxicol 22:280–93.

Li C, Alvey C, Bello A, et al. (2011). Pharmacokinetics (PK) ofcrizotinib (PF-02341066) in patients with advanced non-small celllung cancer (NSCLC) and other solid tumors. J Clin Oncol 29(15suppl):e13065.

Mao J, Johnson TR, Shen Z, Yamazaki S. (2013). Prediction ofcrizotinib-midazolam interaction using the Simcyp population-basedsimulator: comparison of CYP3A time-dependent inhibition betweenhuman liver microsomes versus hepatocytes. Drug Metab Dispos 41:343–52.

Penner N, Klunk LJ, Prakash C. (2009). Human radiolabeled massbalance studies: objectives, utilities and limitations. Biopharm DrugDispos 30:185–203.

Roffey SJ, Obach RS, Gedge JI, Smith DA. (2007). What is the objectiveof the mass balance study? A retrospective analysis of data in animaland human excretion studies employing radiolabeled drugs. DrugMetab Rev 39:17–43.

Shaw AT, Kim D-W, Nakagawa K, et al. (2013). Crizotinib versuschemotherapy in advanced ALK-positive lung cancer. N Engl J Med368:2385–94.

Shou M, Hayashi M, Pan Y, et al. (2008). Modeling, prediction, andin vitro in vivo correlation of CYP3A4 induction. Drug Metab Dispos36:2355–70.

Smith E, Goulet L, Walker G, et al. (2011). Metabolism and excretion of[14C]crizotinib in rats and dogs. Drug Metab Rev 43:76–7.

Soda M, Choi YL, Enomoto M, et al. (2007). Identification of thetransforming EML4-ALK fusion gene in non-small-cell lung cancer.Nature 448:561–6.

Soda M, Takada S, Takeuchi K, et al. (2008). A mouse model forEML4-ALK-positive lung cancer. Proc Natl Acad Sci USA 105:19893–7.

Tan W, Wilner KD, Bang Y, et al. (2010). Pharmacokinetics (PK) ofPF-02341066, a dual ALK/MET inhibitor, after multiple oral doses toadvanced cancer patients. J Clin Oncol 28(15 suppl):2596.

Trudnowski RJ, Rico RC. (1974). Specific gravity of blood and plasma at4 and 37 degrees C. Clin Chem 20:615–16.

Tyczynski JE, Bray F, Maxwell Parkin D. (2003). Lung cancer in Europein 2000: epidemiology, prevention, and early detection. Lancet Oncol4:45–55.

Xu H, O’Gorman M, Tan W, et al. (2011a). The absolute oralbioavailability of crizotinib, a dual ALK/MET inhibitor, in healthysubjects. Meeting Abstracts, AAPS Annual Meeting and Exposition;2011 October 23–27; Washington, DC. Arlington, VA: AmericanAssociation of Pharmaceutical Scientists, Abstract T2370.

Xu H, O’Gorman M, Tan W, et al. (2011b). The effect of apotent CYP3A inhibitor ketoconazole on the pharmacokinetics ofcrizotinib, a dual ALK/MET inhibitor, in healthy subjects. MeetingAbstracts, AAPS Annual Meeting and Exposition; 2011 October 23–27; Washington, DC. Arlington, VA: American Association ofPharmaceutical Scientists, Abstract W5393.

Xu H, O’Gorman M, Tan W, et al. (2011c). The effect of a strongCYP3A inducer rifampin on the pharmacokinetics of crizotinib, a dualALK/MET inhibitor, in healthy subjects. Meeting Abstracts, AAPSAnnual Meeting and Exposition; 2011 October 23–27; Washington,DC. Arlington, VA: American Association of PharmaceuticalScientists, Abstract W5392.

Yamazaki S, Skaptason J, Romero D, et al. (2011). Prediction of oralpharmacokinetics of cMet kinase inhibitors in humans: physiologic-ally based pharmacokinetic model versus traditional one-compartmentmodel. Drug Metab Dispos 39:383–93.

Yamazaki S, Vicini P, Shen Z, et al. (2012). Pharmacokinetic/pharmacodynamic modeling of crizotinib for anaplastic lymphomakinase inhibition and antitumor efficacy in human tumor xenograftmouse models. J Pharmacol Exp Ther 340:549–57.

Zhong W-Z, Zhan J, Kang P, Yamazaki S. (2010). Gender specific drugmetabolism of PF-02341066 in rats – role of sulfoconjugation. CurrDrug Metab 11:296–306.

Zou HY, Li Q, Lee J, et al. (2011). Antitumor efficacy of crizotinib(PF-02341066), a potent and selective ALK and c-Met RTK inhibitor,in EML4-ALK driven NSCLC tumors in vitro and in vivo.Proceedings of the AACR 102nd Annual Meeting; 2011 April 2–6;Orlando, FL. Philadelphia, PA: American Association for CancerResearch, Abstract LB-390.

Zou HY, Li Q, Lee JH, et al. (2007). An orally available small-moleculeinhibitor of c-Met, PF-2341066, exhibits cytoreductive antitumorefficacy through antiproliferative and antiangiogenic mechanisms.Cancer Res 67:4408–17.

DOI: 10.3109/00498254.2014.941964 Crizotinib disposition in humans 15

Xen

obio

tica

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Uni

vers

ity o

f W

ashi

ngto

n on

12/

02/1

4Fo

r pe

rson

al u

se o

nly.