get more info later

1. Imatinib Plasma Levels Are Correlated With Clinical Benefit
in Patients With Unresectable/Metastatic Gastrointestinal
Stromal Tumors
George D. Demetri, Yanfeng Wang, Elisabeth Wehrle, Amy Racine, Zariana Nikolova, Charles D. Blanke,
Heikki Joensuu, and Margaret von Mehren
From the Ludwig Center, Dana-Farber/
Harvard Cancer Center, and Harvard
Medical School, Boston, MA; Oncology
Business Unit, Novartis Pharmaceuticals
Corp, Florham Park, NJ; Oncology Busi-
ness Unit and Biostatistics, Novartis
Pharmaceuticals AG, Basel, Switzerland;
British Columbia Cancer Agency, Univer-
sity of British Columbia, Vancouver, Brit-
ish Columbia, Canada; Department of
Oncology, Helsinki University Central
Hospital, Helsinki, Finland; and Fox Chase
Cancer Center, Philadelphia, PA.
Submitted October 30, 2008; accepted
January 23, 2009; published online
ahead of print at www.jco.org on May
18, 2009.
Supported in part by Novartis Pharma-
ceuticals Corp, Oncology Business Unit,
Florham Park, NJ. Additional philan-
thropic support for this work has been
provided by the Virginia and Daniel K.
Ludwig Trust for Cancer Research and
the Stutman Gastrointestinal Stromal
Tumor Cancer Research Fund (G.D.D.).
Presented in part at the 44th Annual
Meeting of the American Society of
Clinical Oncology, May 30-June 3,
2008, Chicago, IL, as well as the
Gastrointestinal Cancer Symposium of
the American Society of Clinical Oncol-
ogy, the American Gastroenterological
Association, the American Society for
Therapeutic Radiology and Oncology,
and the Society of Surgical Oncology,
January 25-27, 2008, Orlando, FL.
Authors’ disclosures of potential con-
flicts of interest and author contribu-
tions are found at the end of this
article.
Corresponding author: George D.
Demetri, MD, Ludwig Center, Dana-
Farber/Harvard Cancer Center, Dana-
Farber Cancer Institute, D1212, 44
Binney St, Boston MA 02115; e-mail:
gdemetri@partners.org.
The Acknowledgment and Appendix
are included in the full-text version
of this article; they are available
online at www.jco.org. They are
not included in the PDF version
(via Adobe® Reader®).
© 2009 by American Society of Clinical
Oncology
0732-183X/09/2719-3141/$20.00
DOI: 10.1200/JCO.2008.20.4818
A B S T R A C T
Purpose
To study the pharmacokinetics (PK) of imatinib (IM) in patients with advanced GI stromal tumors
(GISTs) treated in a randomized phase II study and to explore the potential relationship between
IM plasma levels and long-term clinical outcomes.
Patients and Methods
Patients were randomly assigned to receive IM at 400 mg versus 600 mg daily. IM plasma levels
were analyzed in a subset of patients (n ϭ 73) for whom PK data on day 1 and at steady-state (SS,
day 29) were available. IM PK was evaluated using a population PK approach. The relationship
between IM plasma exposure and clinical outcome was explored by grouping patients into
quartiles according to IM trough concentration (Cmin). The clinical outcome parameters evaluated
include overall objective benefit rate (OOBR; complete response plus partial response plus stable
disease) time to progression (TTP), and KIT genotyping.
Results
IM PK exposure showed a high inter-patient variability, and clinical outcomes were correlated with
IM trough levels at SS. The median TTP was 11.3 months for patients in the lowest Cmin quartile
(Q1, Ͻ 1,110 ng/mL) compared with more than 30 months for Q2 to Q4 (P ϭ .0029). OOBR was
also inferior in Q1 patients. In patients with GIST with KIT exon 11 mutations (n ϭ 39), the OOBR
was 67% for Q1 patients versus 100% for all others (P ϭ .001).
Conclusion
In patients with advanced GIST, IM trough levels at SS were associated with clinical benefit.
Patients with IM Cmin below 1,100 ng/mL showed a shorter TTP and lower rate of clinical benefit
(OOBR). Further studies are justified to test whether monitoring IM plasma levels might optimize
clinical outcomes for patients with GIST.
J Clin Oncol 27:3141-3147. © 2009 by American Society of Clinical Oncology
INTRODUCTION
Imatinib mesylate (Glivec/Gleevec; Novartis Oncol-
ogy, East Hanover, NJ) is a tyrosine kinase inhibitor
with proven activity in the treatment of GI stromal
tumor (GIST), the most common mesenchymal
tumor of the GI tract, which typically harbors
activating mutations of the KIT or PDGFRA
gene.1-7
Imatinib targets the adenosine triphos-
phate–binding site of the tyrosine-kinase region
in KIT, BCR-ABL, platelet-derived growth factor
receptor (PDGFR) ␣, and PDGFR␤, preventing
phosphorylation and interrupting proliferation and
survival signaling pathways within the cell.8-10
Pa-
tients with different activating KIT mutations have
different clinical outcomes; for example, presence
of exon 9 mutations is an adverse prognostic fac-
tor for response to imatinib in advanced GIST,
compared with KIT exon 11 mutants.9-12
Clinical
characteristics and biologic mechanisms prognostic
for response and survival have been described.11,12
We report here the pharmacokinetics (PK) of ima-
tinib in patients with advanced metastatic and/or
unresectable GIST from a randomized phase II trial
of imatinib (study B2222) and exploratory analyses
of correlations between drug exposure and long-
term clinical benefits in these patients.6,13
Previous PK studies showed that imatinib has
favorable PK characteristics, including rapid, near-
complete (98%) oral bioavailability and a propor-
tional dose-exposure relationship.14,15
Imatinib is
metabolized through the cytochrome P450 system,
primarily via CYP3A4, although other CYPs (1A2,
2D6,2C9,2C19)contributesomewhat.14
Themajor
JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T
VOLUME 27 ⅐ NUMBER 19 ⅐ JULY 1 2009
© 2009 by American Society of Clinical Oncology 3141
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

2. metabolite formed, CGP74588, has similar biologic activity to ima-
tinib and represents approximately 20% to 25% of the parent drug
levelatsteady-state(SS)inpatientswithGIST.16
Asaresultofintrinsic
variability of CYP enzyme activity,17,18
high inter-patient variability
was reported for imatinib exposure in patients with chronic myelog-
enous leukemia (CML) and GIST. In patients with CML, cytogenetic
and molecular responses are highly correlated with imatinib trough
plasma exposure.19,20
The relationship between imatinib exposure
and clinical benefit rates in patients with GIST has not previously
been reported.
PATIENTS AND METHODS
The study (B2222) on which this PK substudy is based was a randomized,
open-label, multicenter, phase II clinical trial evaluating imatinib in patients
with unresectable or metastatic CD117-positive GIST. All patients gave writ-
ten informed consent to participate in the clinical trial and the PK substudy.
B2222 study design, inclusion criteria, and exclusion criteria have been
described.6,13
Briefly, 147 patients were randomly assigned in a 1:1 ratio to
receive imatinib 400 mg or 600 mg/d. Dose escalation was allowed if there was
disease progression from 400 mg once daily to 600 mg once daily or from 600
mg once daily to 400 mg twice a day. The estimated proportion of patients
surviving 60 months is 54%, regardless of initial dose level. PK blood samples
were collected from the first half of the patients enrolled onto the study (73 of
147 patients): samples for PK analysis were collected on days 1 and 29 (SS), at
predose, and at 1 to 3 hours, 6 to 9 hours, and approximately 24 hours (before
the morning dose on the next day).
PK blood samples (8 mL each) were collected into heparinized polypro-
pylene tubes, centrifuged at 1,000 ϫ g for 10 minutes for plasma separation,
andkeptbelowϪ20°Cuntilanalysis.PlasmaimatinibandCGP74588concen-
trations were determined using a validated liquid chromatography–tandem
mass spectrometry assay.21
The lower limit of quantification was 4 ng/mL for
both imatinib and CGP74588.
Data Analysis
PK data analysis. The PK of imatinib in humans has been studied using
a population-PK (pop-PK) approach.16,22-24
Here, a similar approach was
used in patients with advanced GIST; the calculated PK parameters from the
pop-PK model for imatinib were used for correlation analysis with B2222
long-term follow-up data on clinical outcomes. PK data for CGP74588
were not included in the PK modeling and the correlation analysis because
(1) its level was highly correlated with imatinib exposure, and (2) it repre-
sented a relatively small fraction of the parent drug level. A one-compartment
linear disposition model with zero order absorption was used for the analysis
as reported.23
The clinical and demographic covariates considered in pop-PK analysis
were age, sex, body weight, serum creatinine, creatinine clearance, WBC
count,AST,ALT,albumin,totalbilirubin,andhemoglobin,aswellaspresence
of liver metastases, edema, and disease progression. Except edema and liver
metastasis, covariate values were taken on the day of PK assessment or just
before in both occasions day 1 and day 29. Covariates were incorporated into
the model as in the analysis of PK data in patients with CML.23
Additional
information on the population PK model is described (Appendix, on-
line only).
The area under the plasma concentration-time curve (AUC), peak
concentration (Cmax), trough concentration (Cmin), and effective half-life
(t1/2 ϭ 0.693 ⅐ V/clearance[CL]) after the first dose and at SS were derived for
eachpatientoneachPKassessmentday,usingpatient’sactualdosingregimen,
individual conditional estimates of oral drug clearance (CL/F) and oral drug
volume of distribution (V/F), and population duration of absorption. These
individual PK parameters at SS were used for correlation analysis with the
clinical response data.
Response data and PK/pharmacodynamic data analysis. The pharmaco-
dynamic (PD) parameter was the primary efficacy parameter, best overall
tumor response, based on Southwest Oncology Group criteria, over a 5-year
follow-upperiodasdescribed.13
Tumormeasurementswereobtainedatbase-
line, 1 and 3 months, and every 3 to 6 months. A responder was defined as any
patientwhosebestconfirmedresponsewascompleteresponse(CR,definedas
disappearanceofallmeasurableandnonmeasurablebutassessablediseaseand
no new lesions), partial response (PR, defined as Ն 50% decrease below
baseline in sum of products of perpendicular diameters of all measurable
lesions), or stable disease (SD) as defined.6
Secondary efficacy parameters
included time to progression (TTP) and overall survival (OS). TTP was de-
fined as the earliest occurrence of progression, death from any cause, or
discontinuation from the trial for any reason other than “condition no longer
requires treatment.” The PK parameters from both 400 mg and 600 mg doses
were pooled for the PK/response correlation analysis.
The correlations among imatinib PK parameters, AUC, Cmax, and Cmin,
were examined by simple linear regression; Cmin was selected for correlation
analysis with response data. To assess the impact of Cmin values on response,
Cmin was arbitrarily grouped into four quartiles. The lower quartile (Q1)
included the 25% of patients with the lowest imatinib plasma exposure; Q2
and Q3 extended 25% below and above the median concentration, respec-
tively; Q4 included the 25% of patients with the highest imatinib plasma
exposure. The central 50% of the data were combined for all analyses and are
referred to as Q2-Q3. These three PK categories (Q1, Q2-Q3, and Q4) were
used for stratification as appropriate. The TTP was estimated using the
Kaplan-Meier method, and strata were compared using the log-rank test. The
TTP was censored for patients who did not experience disease progression at
thelastavailabletumorassessment.Correlationofclinicalbenefitandimatinib
Cmin levelswasalsoexaminedforsubsetsofpatientswithspecificmutationsin
KIT exon 11 or exon 9.
RESULTS
Demographics and Imatinib PK
In 73 patients with assessable imatinib PK data (41 men, 32
women), most were white (n ϭ 69; 94.5%), with median age of 51
years(range,25to79years)andmedianbodyweightof75.5kg(range,
44.7 to 139.7 kg). In general, the pop-PK model predicted reasonably
well the concentration observations, despite a relatively large residual
error. There is no obvious trend or bias in weighted residuals (Appen-
dix Fig A1, online only). The final population model parameters at SS
are shown (Appendix Table A1, online only). The typical values for
apparent clearance (TVCL) and volume of distribution (TVV) are
predicted by the following equations:
TVCL ϭ 8.18 ⅐ (albumin/38.3)1.66
⅐
(WBC/[7 ⅐ 109
])Ϫ0.418
([L/h]) (1)
TVV ϭ (168 ϩ 58.5 on day 28) ⅐ (albumin/38)1.66
⅐
(WBC/[7 ⅐ 109
])Ϫ0.418
[L] (2)
Imatinib was absorbed quickly (mean duration of absorption, 1.69
hours).Amongclinical,laboratory,anddemographicvariablesexam-
ined, only albumin and WBC count were significant covariates for
imatinib CL and volume of distribution (Fig 1). The PK parameters
determinedbythepopulationmodelandthealbuminandWBClevels
measured on days 1 and 29 are shown (Table 1). The imatinib PK
parameters at SS, AUC, Cmax, and Cmin values were highly correlated
to each other, with correlation coefficients of 0.896, 0.994, and 0.961
between AUC and Cmin, Cmax and Cmin, and AUC and Cmax, respec-
tively (Pearson correlation test).
Demetri et al
3142 © 2009 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

3. Correlation of Imatinib Exposure With Demographics,
Sex, and Body Weight
Although sex was not identified as a significant covariate for
imatinibCLwithinthetotalPKpatientpopulation,theaveragetrough
level in women (1,770 ng/mL; n ϭ 32) appeared to be approximately
25% higher than in men (1,420 ng/mL; n ϭ 41), consistent with body
weight differences between men and women, 83.8 Ϯ 17.2 kg (n ϭ 41)
and 69.6 Ϯ 15.2 kg (n ϭ 32), respectively. Of 41 male patients, 14
(34%) were distributed in the Q1 group, compared with four (12.5%)
of32femalepatientsintheQ1group.Theoverallclinicalresponserate
(CR,PR,orSD)was26(81%)of32femalepatientsand31(76%)of41
male patients (P ϭ .56; ␹2
test).
Correlation of Imatinib Exposure With Clinical
Responses and Progression
The SS Cmin parameter was selected for correlation analysis
with response data (Fig 2). Of 73 patients with assessable Cmin data,
57 (78%) achieved CR, PR, or SD. The median Cmin value for
responders (CR, PR, or SD) was higher than that of nonresponders
(disease progression or not assessable): 1,446 ng/mL (range, 414 to
3,336 ng/mL; 25th and 75th percentiles, 1,204 and 2,062 ng/mL;
n ϭ 57) and 1,155 ng/mL (range, 545 to 4,182 ng/mL; 25th and
75th percentiles, 1,041 and 1,562 ng/mL; n ϭ 16), respectively. No
statistically significant difference was achieved (P ϭ .25, Wilcoxon
test). Of the 57 patients who achieved a CR, PR, or SD, 12 (67%) of
18 patients were in Q1, 29 (81%) of 36 patients were in Q2-Q3, and
16 (84%) of 19 patients were in Q4, or 82% overall in Q2-Q4
(P ϭ .177 between Q1 and upper quartiles Q2-Q4, ␹2
test; Table 2).
If the best objective response was defined as CR and PR (without
SD), the response rate would be 63% (n ϭ 46), with eight (44.4%)
of 18 patients in Q1, 24 (66.7%) of 36 patients in Q2-Q3, and 14
(73.3%) of 19 patients in Q4 (P ϭ .0601 between Q1 and upper
quartiles Q2-Q4, ␹2
test). Most CR and PR responders were ob-
served at 3 months after treatment.
Patients in Q1 group showed a faster TTP (median 11.3 months)
thanupperquartiles:30.6monthsforQ2-Q3,and33.1monthsforQ4
(overall P ϭ .0105, and P ϭ .0029 between Q1 and Q2-Q4; hazard
ratio ϭ 0.418 with 95% CI, 0.231 to 0.756; Fig 3). No significant
difference in median OS was observed between different imatinib
Cmin groups (P ϭ .16).
20 25 3530 4540 20 25 3530 4540
2 5 118 1714 2 5 118 1714
Effects of change in ALB levels on subject’s
predicted exposure: AUC day 1
0
20
40
60
80
100
120
140
160
180
ALB (g/L)
AUC(μg/mL•h)
Effects of change in ALB levels on subject’s
predicted exposure: AUC day 28 or more
0
20
40
60
80
100
120
140
160
ALB (g/L)
AUCtau(μg/mL•h)
Effects of change in WBC count on subject’s
predicted exposure: AUC day 1
0
20
40
60
80
100
120
140
160
180
WBC (E9/L)
AUC(μg/mL•h)
Effects of change in WBC count on subject’s predicted
exposure: AUC tau day 28 or more
0
20
40
60
80
100
120
140
160
180
WBC (E9/L)
AUCtau(μg/mL•h)
Fig 1. Effects of albumin (ALB) and WBC on model predicted area under the plasma concentration-time curve (AUC)ϱ day 1 and single-dose AUC24 at steady-state
(day 29).
Imatinib Plasma Levels in Patients With GIST
www.jco.org © 2009 by American Society of Clinical Oncology 3143
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

4. Correlation of Imatinib Exposure With Clinical Benefit
for Patients With KIT Mutations in Exon 11 or 9
Of73patientswithGIST(53.4%),39hadKITexon11mutations
and12(16.4%)hadmutationsinexon9.Theremainder(nϭ22)had
different tumor genotypes (more rare KIT mutations, PDGFRA mu-
tation, or wild-type KIT).
The 39 patients with KIT exon 11 mutations were distributed as
follows: nine (50%) of 18 in Q1, 17 (47.2%) of 36 in Q2-Q3, and 13
(68.4%) of 19 in Q4. The overall objective benefit rate (OOBR; CR ϩ
PR ϩ SD) was six (67%) of nine patients in Q1, 17 (100%) of 17
patients in Q2-Q3, and 13 (100%) of 13 patients in Q4 (P ϭ .0001
betweenQ1andQ2-Q4,␹2
test).PatientswithKITexon11mutations
Table 1. Summary of Imatinib Pharmacokinetic Parameters at 400-mg and 600-mg Once-Daily Doses in Patients With GIST
Parameter
400 mg Once Daily (n ϭ 36) 600 mg Once Daily (n ϭ 37)
Mean
Standard
Deviation CV% Mean
Standard
Deviation CV %
Day 1†
AUC0-inf, ␮g‫ء‬
hr/mL 69.7 41.8 60 73.5 25.9 35.3
Cmax, ng/mL 3,251 1,822 56.1 3,444 1,153 33.5
Cmin,24hr, ng/mL 1,093 660 60.4 1,152 409 35.5
CL/F, L/h 8.01 4.88 60.9 9.27 3.46 37.3
V/F, L 157 87.6 55.8 186.8 64.8 34.7
t1/2, hours 14.0 1.5 108 14.2 1.5 10.9
Albumin, g/L‫ء‬
35.0 6.5 18.5 38.0 3.66 9.6
WBC, counts/L ϫ 109‫ء‬
7.76 2.91 37.5 6.72 2.74 40.9
Day 29†
AUC␶, ␮g‫ء‬
hr/mL 56.4 23.9 42.5 64.3 27.1 42.2
Cmax, ng/mL 3,405 1,434 42.1 3,868 1,574 40.7
Cmin, ng/mL 1,530 666 43.6 1,752 794 45.3
CL/F, L/h 8.84 4.87 55.1 10.9 4.43 40.6
V/F, L 241 117 48.6 300 107 35.6
t1/2, hours 19.3 2.2 11.5 19.5 2.8 14.3
Albumin, g/L‫ء‬
35.5 5.0 14.1 38.1 3.8 10.0
WBC, counts/L ϫ 109‫ء‬
5.59 1.86 33.3 4.60 1.34 29.2
Abbreviations: GIST, GI stromal tumor; CV, coefficient of variation; AUC, area under the plasma concentration-time curve; Cmax, peak concentration; Cmin, trough
concentration; CL/F, oral drug clearance; V/F, oral drug volume of distribution; t1/2, terminal half-life.
‫ء‬
Measured biologic laboratory values.
†On day 1, n ϭ 32 and n ϭ 34 for 400 mg and 600 mg, respectively. On day 29, n ϭ 31 and n ϭ 32 for 400 mg and 600 mg, respectively. Total sample size with
pharmacokinetic data was 73.
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73
Patient Rank by Cmin
Cmin(ng/mL)
0
1,000
2,000
3,000
4,000
Q4Q1 Q2 to Q3
400-mg dose
600-mg dose
Fig 2. Distribution of imatinib trough con-
centration (Cmin) at steady-state (day 29) for
400-mg and 600-mg daily doses combined.
For the 36 patients in the 400-mg dose
group, 12 patients were in quartile 1 (Q1),
17 were in Q2-Q3, and seven were in Q4,
whereas for the 37 patients in the 600-mg
dose group, six were in Q1, 19 were in
Q2-Q3, and 12 were in Q4. The vertical
lines represent 25% and 75% percentiles
(ie, 1,100 and 2,040 ng/mL), respectively,
dividing the population into three groups:
Q1 (414 to 1,110 ng/mL; n ϭ 18), Q2-Q3
(1,110 to 2,040 ng/mL; n ϭ 36), and Q4
(2,041 to 4,182 ng/mL; n ϭ 19).
Demetri et al
3144 © 2009 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

5. seemed to have improved outcomes at imatinib levels greater than
1,100 ng/mL, the lower cut point for Q2-Q4 group combined. For the
12 patients with KIT exon 9 mutations, the OOBR was two of three
patients in Q1, five of eight patients in Q2-Q3, and one patient in Q4
with SD. The imatinib median (mean Ϯ standard deviation) Cmin in
patients with KIT exon 9 mutations was 1,270 ng/mL (1340 Ϯ 420
ng/mL, n ϭ 12), which was only slightly lower than that in patients
with KIT exon 11 mutations, 1,540 ng/mL (1630 Ϯ 660 ng/mL,
n ϭ 39; P ϭ .15 by Wilcoxon test). The plasma clearance of imatinib
(defined as dose/AUC) was 10.8 Ϯ 3.9L/h in the patients with exon 9
mutations and 9.2 Ϯ 4.5L/h in the patients with exon 11 mutations
(P ϭ .11 by Wilcoxon test). Further studies are needed to confirm
these PK findings and to assess the PK/outcome relationships for
patients with exon 11 and exon 9 mutations.
Correlation of Imatinib Trough Levels With
Patient Disposition
Of the 73 total patients, 21 patients (28.8%) remain on study and
52 patients (71.2%) discontinued. Disease progression (“unsatisfac-
tory therapeutic effect”) was the most frequently cited reason for
discontinuation(nϭ35,47.9%);nodifferencewasobservedbetween
different imatinib Cmin groups with the limited PK data available.
Other reasons for withdrawal included adverse events (5.5%), abnor-
mal laboratory values (2.7%), consent withdrawal (8.2%), death
(4.1%), or protocol violation (1.4%). Dose escalation of imatinib was
allowed after disease progression. Seven patients’ imatinib dose in-
creasedfrom600mgto800mgdailyafteramedianof846days(range,
689 to 1,338 days); 21 patients had dose increased from 400 to 600 mg
after a median of 363 days (range, 29 to 1,512 days), with four of these
21 patients undergoing further dose increases to 800 mg daily at a
median of 790 days (range, 735 to 965 days). By PK quartiles, the
frequency of dose escalation was nine (50%) of 18 in Q1, 15 (41.7) of
36 in Q2-Q3, and four (21%) of 19 in Q4 (P ϭ .165, ␹2
test). Detailed
information on dose escalation and corresponding clinical outcomes
for subsets of patients with different KIT mutations is available (Ap-
pendix Table A2, online only).
DISCUSSION
Thepop-PKofimatinibinpatientswithGISThasbeendescribed,16,22
but the relationship between imatinib PK, drug exposure, and clinical
outcomes has not been reported. In our study of patients with ad-
vanced GIST, imatinib trough plasma levels seemed to be correlated
with clinical benefits, including TTP and objective response.
Reported pop-PK analyses suggest that plasma ␣-1 acid gly-
coprotein (AGP), WBC, granulocyte, albumin, hemoglobin, or
patient demographics could be identified as significant covariates
for CL/F and V/F, depending on patient population, sample size,
and availability of these parameters for analysis.16,22-24
Some of
these parameters are correlated with each other.22,25-27
Thus as a
result of improvement of disease conditions after imatinib treat-
ment, plasma AGP, albumin, and WBC count may change with
time (normalization), which in return could affect the CL/F and
exposure of imatinib as observed in our study. From day 1 to 29,
the albumin level remained similar, whereas WBC counts de-
creased significantly, by approximately 28.0% and 31.5%, for the
400 mg and 600 mg doses, respectively (Table 1). Concurrently, the
CL/F increased slightly by 10.4% and 17.7%, whereas V/F in-
creased significantly by 53.8% and 60.5% from day 1 to 29 for the
400 mg and 600 mg doses, respectively. A more pronounced in-
crease in V/F might be a reflection of normalization of plasma AGP
levels,16
in addition to WBC and albumin levels. As a result of a
Table 2. Rates of Best Objective Benefit (CR ϩ PR ϩ SD) or Objective Response per SWOG Criteria (CR ϩ PR) for All PK Population and for Patients With KIT
Exon 11 and Exon 9 Mutations by Imatinib Cmin Quartiles
PK Cmin Group
Q1 Q2-Q3 Q4
No. % No. % No. %
All patients, n ϭ 73 18 36 19
CR ϩ PR ϩ SD 12 67 29 81 16 84
CR ϩ PR 8 44 24 67 14 74
Exon 11 mutations, n ϭ 39 9 17 13
CR ϩ PR ϩ SD 6 67 17 100 13 100
CR ϩ PR 5 56 16 94 12 92
Exon 9 mutations, n ϭ 12 3 8 1
CR ϩ PR ϩ SD 2 67 5 62.5 1 100
CR ϩ PR 0 0 4 50 0 0
Abbreviations: CR, complete response; PR, partial response; SD, stable disease; SWOG, Southwest Oncology Group; PK, pharmacokinetics; Cmin, trough
concentration; Q, quartile.
1.0
0.8
0.6
0.4
0.2
0.9
0.7
0.5
0.3
0.1
0 6 12 18 24 30 36 42 48 54 60
Progression-FreeSurvival(proportion)
Time Since Start of Treatment (months)
Q1
Q2-Q3
Q4
Cmin quartile
Fig 3. Time to progression by imatinib day 29 trough level (Cmin) quartile (Q).
Imatinib Plasma Levels in Patients With GIST
www.jco.org © 2009 by American Society of Clinical Oncology 3145
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

6. more significant increase in V/F than CL/F, imatinib half-life was
prolonged from 14 hours on day 1 to 19 hours at SS. Preliminary,
long-term (from Ͼ 6 months to 2.5 years) sparse PK data from 30
patients showed that imatinib plasma CL (estimated using the
same population PK model) and biologic parameters such as WBC
and albumin levels remained stable over time (data not shown).
Imbalanced albumin levels and WBC counts between different
dose groups could impact the apparent dose-exposure relationship.
Patients in the 400-mg dose group seemed to have a slightly lower
albumin level and higher WBC count than those in the 600-mg dose
group.ThisresultedinanartificiallyhigherthanexpectedAUCforthe
400-mg dose or lower than expected AUC for the 600-mg dose. The
average body weight of patients in the two dose groups was similar,
with 79.6 Ϯ 19.3 kg and 75.7 Ϯ 16.1 kg for 400 mg/d and 600 mg/d,
respectively (P ϭ .33, Wilcoxon test). In patients with CML,15
a linear
dose-exposure relationship was observed at doses up to 1,000 mg
daily. The imatinib CL, 8 to 10 L/h (Table 1), was similar to the CL (8
to 10 L/h) reported for patients with GIST,16,22
which, on average,
was somewhat lower than that in patients with CML (10 to 14
L/h).14,15,23,24
A large inter-patient variability in imatinib plasma ex-
posure was observed in patients with GIST and CML, underlining a
potential impact of imatinib variability on clinical outcomes for both
patient populations.
Certain KIT or PDGFRA mutations are predictive factors for
clinical response to imatinib.11-13
Normal albumin level, neutrophil
level, and female sex were also shown as significant covariates for
clinical outcomes,13
perhaps through their correlation with imatinib
exposure. The response in women seemed to be slightly better than
that of men. Here, we found that patients with a low plasma exposure
(Cmin less than 1,100 ng/mL) showed a trend of low rate of objective
response and rapid evolution of resistance (short TTP). These results
suggest that a minimal plasma threshold may be necessary to achieve
and maintain clinical response. This hypothesis is best supported in
patientswithKITexon11mutations,whoexhibitedimprovedclinical
outcomes with imatinib trough levels more than 1,110 ng/mL. Too
few patients with KIT exon 9 mutation were available to draw any
conclusions. Although in vitro data suggest that imatinib is equally
effective against KIT exon 9 and exon 11 mutants,11,12
cumulative
clinicaldatasuggestpooreroutcomesforpatientswithKITmutations
in exon 9 versus exon 11, and data from large randomized trials have
supported the hypothesis that a higher starting dose of imatinib (800
mg/d) improves TTP for patients with exon 9 mutation as compared
with the lower conventional dose (400 mg/d).7,11-13
Additional data
are needed to confirm whether there are any PK differences between
subsets of patients with different KIT mutations and to define the
minimum effective plasma concentration of imatinib (with and with-
out considering the contribution of CGP74588) for different KIT
mutation subpopulations.
The large inter-patient variability in imatinib plasma exposure
seemed to be correlated with patients’ albumin level and concurrent
WBC count. Imatinib plasma CL increased minimally with time, by
10% to 17%, after 1 month of treatment, which was probably a
reflection of an improvement of patients’ disease conditions with
concurrentnormalizationofWBCcount.Becauseimatinibhasahigh
binding affinity to plasma AGP proteins, a concurrent normalization
of AGP level over time could also result in a decrease in imatinib total
plasma exposure. This hypothesis, however, has not been confirmed,
because the AGP level was not determined in the present study. Pa-
tients with lower imatinib trough plasma levels (Ͻ 1,100 ng/mL)
showed a trend of lower rate of clinical response and faster TTP. The
medianimatinibtroughlevelforrespondersseemedtobehigherthan
that of nonresponders. Further studies are warranted to define the
effective Cmin levels of imatinib for patients with different KIT muta-
tions. Monitoring imatinib plasma trough level may be advisable in
patients with suspected drug-drug interaction, unusual side effects,
lackofexpectedclinicalbenefit,orcomplianceconcerns.Thisprelim-
inary analysis suggests that a low SS plasma level of imatinib (ie,
Ͻ 1,100 ng/mL) might contribute to drug failure in patients with
advanced GIST.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
Although all authors completed the disclosure declaration, the following
author(s) indicated a financial or other interest that is relevant to the subject
matter under consideration in this article. Certain relationships marked
with a “U” are those for which no compensation was received; those
relationships marked with a “C” were compensated. For a detailed
description of the disclosure categories, or for more information about
ASCO’s conflict of interest policy, please refer to the Author Disclosure
Declaration and the Disclosures of Potential Conflicts of Interest section in
Information for Contributors.
Employment or Leadership Position: Yanfeng Wang, Novartis (C);
Elisabeth Wehrle, Novartis (C); Amy Racine, Novartis (C); Zariana
Nikolova, Novartis (C) Consultant or Advisory Role: George D.
Demetri, Novartis (C), Pfizer (C), Bayer Pharmaceuticals (C), Infinity
Pharmaceuticals (C); Charles D. Blanke, Novartis (C); Heikki Joensuu,
Novartis (C); Margaret von Mehren, Novartis (C) Stock Ownership:
Yanfeng Wang, Novartis; Elisabeth Wehrle, Novartis; Zariana Nikolova,
Novartis Honoraria: George D. Demetri, Novartis, Pfizer; Charles D.
Blanke, Novartis; Heikki Joensuu, Novartis Research Funding: George
D. Demetri, Novartis, Pfizer, Bristol-Myers Squibb Infinity; Charles D.
Blanke, Novartis; Margaret von Mehren, Novartis Expert Testimony:
George D. Demetri, Novartis (U), Pfizer (U), Infinity (U) Other
Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: George D. Demetri, Yanfeng Wang, Charles D.
Blanke, Heikki Joensuu, Margaret von Mehren
Financial support: George D. Demetri
Administrative support: George D. Demetri
Provision of study materials or patients: George D. Demetri, Charles D.
Blanke, Heikki Joensuu, Margaret von Mehren
Collection and assembly of data: George D. Demetri, Yanfeng Wang,
Amy Racine, Zariana Nikolova, Charles D. Blanke, Heikki Joensuu,
Margaret von Mehren
Data analysis and interpretation: George D. Demetri, Yanfeng Wang,
Elisabeth Wehrle, Amy Racine, Zariana Nikolova, Charles D. Blanke,
Heikki Joensuu, Margaret von Mehren
Manuscript writing: George D. Demetri, Yanfeng Wang, Elisabeth
Wehrle, Amy Racine, Zariana Nikolova, Charles D. Blanke, Heikki
Joensuu, Margaret von Mehren
Final approval of manuscript: George D. Demetri, Yanfeng Wang, Amy
Racine, Zariana Nikolova, Charles D. Blanke, Heikki Joensuu,
Margaret von Mehren
Demetri et al
3146 © 2009 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13

7. REFERENCES
1. Miettinen M, Majidi M, Lasota J: Pathology
and diagnostic criteria of gastrointestinal stromal
tumors (GISTs): A review. Eur J Cancer 5:S39-S51,
2002 (suppl)
2. Fletcher CD, Berman JJ, Corless C, et al:
Diagnosis of gastrointestinal stromal tumors: A con-
sensus approach. Hum Pathol 33:459-465, 2002
3. Corless CL, Fletcher JA, Heinrich MC: Biology
of gastrointestinal stromal tumors. J Clin Oncol
22:3813-3825, 2004
4. van Oosterom AT, Judson I, Verweij J, et al:
Safety and efficacy of imatinib (STI571) in metastatic
gastrointestinal stromal tumours: A phase I study.
Lancet 358:1421-1423, 2001
5. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et
al: Effect of the tyrosine kinase inhibitor STI571 in a
patient with a metastatic gastrointestinal stromal
tumor. N Engl J Med 344:1052-1056, 2001
6. Demetri GD, von Mehren M, Blanke CD, et al:
Efficacy and safety of imatinib mesylate in advanced
gastrointestinal stromal tumors. N Engl J Med 347:
472-480, 2002
7. Verweij J, Casali PG, Zalcberg J, et al:
Progression-free survival in gastrointestinal stromal
tumours with high-dose imatinib: Randomised trial.
Lancet 364:1127-1134, 2004
8. Savage DG, Antman KH: Imatinib mesylate: A
new oral targeted therapy. N Engl J Med 346:683-
693, 2002
9. Heinrich MC, Corless CL, Demetri GD, et al:
Kinase mutations and imatinib response in patients
with metastatic gastrointestinal stromal tumor.
J Clin Oncol 21:4342-4349, 2003
10. Miettinen M, Lasota J: Gastrointestinal stro-
mal tumors: Review on morphology, molecular pa-
thology, prognosis, and differential diagnosis. Arch
Pathol Lab Med 130:1466-1478, 2006
11. Heinrich MC, Corless CL, Blanke CD, et al:
Molecular correlates of imatinib resistance in gastro-
intestinal stromal tumors. J Clin Oncol 24:4764-
4774, 2006
12. Debiec-Rychter M, Sciot R, Le Cesne A, et al:
KIT mutations and dose selection for imatinib in
patients with advanced gastrointestinal stromal tu-
mours. Eur J Cancer 42:1093-1103, 2006
13. Blanke CD, Demetri GD, Mehren M, et al:
Long-term results from a randomized phase II trial of
standard versus higher-dose imatinib mesylate for
patients with unresectable or metastatic gastroin-
testinal stromal tumors expressing KIT. J Clin Oncol
26:620-625, 2008
14. Peng B, Lloyd P, Schran H: Clinical pharmaco-
kinetics of imatinib Clin Pharmacokinet 444:879-
894, 2005
15. Peng B, Hayes M, Resta D, et al: Pharmaco-
kinetics and pharmacodynamics of imatinib in a
phase I trial with chronic myeloid leukemia patients.
J Clin Oncol 22:935-942, 2004
16. Delbaldo C, Chatelut E, Re´ M, et al:
Pharmacokinetic-pharmacodynamic relationships of
imatinib and its main metabolite in patients with
advanced gastrointestinal stromal tumors. Clin Can-
cer Res 12:6073-6078, 2006
17. Wilkinson GR: Cytochrome P4503A (CYP3A)
metabolism: Prediction of in vivo activity in humans.
J Pharmacokinet Biopharm 24:475-490, 1996
18. Wojnowski L: Genetics of the variable expres-
sion of CYP3A in humans. Ther Drug Monit 26:192-
199, 2004
19. Picard S, Titier K, Etienne G, et al: Trough
imatinib plasma levels are associated with both
cytogenetic and molecular responses to standard-
dose imatinib in chronic myeloid leukemis. Blood
109:3496-3499, 2007
20. Larson RA, Druker BJ, Guilhot F, et al: Ima-
tinib pharmacokinetics and its correlation with re-
sponse and safety in chronic phase chronic myeloid
leukemia: A sub-analysis of the IRIS Study. Blood
111:4022-4028, 2008
21. Bakhtiar R, Lohne J, Ra mos L, et al: High-
throughput quantification of the anti-leukemia drug
STI571 (Gleevec) and its main metabolite (CGP74588)
in human plasma using liquid chromatography-tandem
mass spectrometry. J Chromatogr B Analyt Technol
Biomed Life Sci 768:325-340, 2002
22. Judson I, Ma P, Peng B, et al: Imatinib pharma-
cokinetics in patients with gastrointestinal stromal
tumour: A retrospective population pharmacokinetic
study over time—EORTC Soft Tissue and Bone Sar-
coma Group. Cancer Chemother Pharmacol 55:379-
386, 2005
23. Schmidli H, Peng BM, Riviere GJ, et al: Pop-
ulation pharmacokinetics of imatinib mesylate in
patients with chronic-phase chronic myeloid leukae-
mia: Results of a phase III study. Br J Clin Pharmacol
60:35-44, 2005
24. Widmer N, Decosterd LA, Csajka C, et al:
Population pharmacokinetics of imatinib and the role
of alpha-acid glycoprotein. Br J Clin Pharmacol 62:
97-112, 2006
25. le Coutre P, Kreuzer KA, Na IK, et al: Deter-
mination of alpha-1 acid glycoprotein in patients with
Phϩ chronic myeloid leukemia during the first 13
weeks of therapy with STI571. Blood Cells Mol Dis
28:75-85, 2002
26. de Silva CM, Reid R: Gastrointestinal stromal
tumors (GIST): C-kit mutations, CD117 expression,
differential diagnosis and targeted cancer therapy
with Imatinib. Pathol Oncol Res 9:13-19, 2003
27. Slaviero KA, Clarke SJ, Rivory LP: Inflammatory
response: An unrecognised source of variability in the
pharmacokinetics and pharmacodynamics of cancer
chemotherapy. Lancet Oncol 4:224-232, 2003
■ ■ ■
Imatinib Plasma Levels in Patients With GIST
www.jco.org © 2009 by American Society of Clinical Oncology 3147
Copyright © 2009 by the American Society of Clinical Oncology. All rights reserved.
Downloaded from jco.ascopubs.org by Salvatore Salamone on July 20, 2009 from 12.47.26.56.
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13
confidential
Kevin Harter
Saladax Biomedical
Sep 02, 2015 09:13