Ex Parte Miller

Board of Patent Appeals and Interferences

Jul 8, 2010

10318659 (B.P.A.I. Jul. 8, 2010)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
10/318,659 12/12/2002 Richard A. Miller 25922-716.201 8345
21971 7590 07/08/2010
WILSON, SONSINI, GOODRICH & ROSATI
650 PAGE MILL ROAD
PALO ALTO, CA 94304-1050
EXAMINER
GEMBEH, SHIRLEY V
ART UNIT PAPER NUMBER
1618
MAIL DATE DELIVERY MODE
07/08/2010 PAPER
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte RICHARD A. MILLER
____________

Appeal 2009-011751
Application 10/318,659
Technology Center 1600
____________

Before ERIC GRIMES, RICHARD M. LEBOVITZ, and
JEFFREY N. FREDMAN, Administrative Patent Judges.

LEBOVITZ, Administrative Patent Judge.

DECISION ON APPEAL1

1 The two-month period for filing an appeal or commencing a civil action, as
recited in 37 C.F.R. § 1.304, or for filing a request for rehearing, as recited
in 37 C.F.R. § 41.52, begins to run from the "MAIL DATE" (paper delivery
mode) or the "NOTIFICATION DATE" (electronic delivery mode) shown
on the PTOL-90A cover letter attached to this decision.

Appeal 2009-011751
Application 10/318,659

2
This is a decision on the appeal under 35 U.S.C. § 134 by the Patent
Applicant from the Patent Examiner’s rejection of claims 1-25 as obvious
under 35 U.S.C. § 103(a). The Board’s jurisdiction for this appeal is under
35 U.S.C. § 6(b). We reverse the Examiner’s rejection, but enter a new
ground of rejection of all the claims as obvious under 35 U.S.C. § 103(a).

STATEMENT OF THE CASE
The claims are drawn to methods for delaying neurologic progression
in humans afflicted with lung cancer comprising administering a Formula I
compound. The Specification states the inventor discovered that the
Formula I compound significantly improved neurological function in lung
cancer patients with brain metastases, but not in other systemic cancers
accompanied by brain metastases (Spec. 2:11-17).
Claims 1-25 are pending and stand rejected under 35 U.S.C. § 103(a)
as obvious in view of Roder,2 Young,3 and Viala4 (Ans. 3). Claim 1 is
representative and reads as follows:
1. A method for delaying neurologic progression in a human
afflicted with lung cancer comprising administering to the

2 Roder et al., US 6,013,646, issued Jan. 11, 2000.
3 Young et al., Gadolinium (III) texaphyrin: A tumor selective radiation
sensitizer that is detectable by MRI, 93 PROC. NATL. ACAD. SCI. USA 6610-
6615 (1996).
4 Viala et al., Phases IB and II Multidose Trial of Gadolinium Texaphyrin, a
Radiation Sensitizer Detectable at MR Imaging: Preliminary Results in
Brain Metastases, 212 RADIOLOGY 755-759 (1999).
Appeal 2009-011751
Application 10/318,659

3
human about 5.0 mg/kg to about 10 mg/kg of a compound of
Formula I [as reproduced on page 22 of the Appeal Brief].

RODER, YOUNG, & VIALA
The Examiner did not meet the burden of showing that it would have
been obvious to persons of ordinary skill in the art to delay neurologic
progression in a human with lung cancer by administering “about 5.0 mg/kg
to about 10 mg/kg of a compound of Formula I.” The Examiner asserted
that Roder taught inhibiting dementia with gadolinium texaphyrin (Ans. 3),
which is the claimed Formula I compound. However, this was a factual
error. Roder taught indolocarbozole compounds for treating dementia (col.
1, ll. 5-12; col. 5, ll. 44-47). Such compounds do not include the claimed
Formula I compound. Accordingly, because the Examiner erred at least with
this finding, we reverse the rejection of claims 1-25 as obvious in view of
Roder, Young, and Viala.

NEW GROUND OF REJECTION
Pursuant to 37 C.F.R. § 41.50(b), we enter the following new ground
of rejection:
Claims 1-25 are rejected under 35 U.S.C. § 103(a) as obvious in view
of Viala.
PRINCIPLES OF LAW
Prima facie obviousness has been found when the claimed range and
the prior art range do not overlap, but are “are so close that prima facie one
Appeal 2009-011751
Application 10/318,659

4
skilled in the art would have expected them to have the same properties,”
shifting the burden to the applicant to show they are different. Titanium
Metals Corp. v. Banner, 778 F.2d 775, 783 (Fed. Cir. 1985).
“[W]hen the difference between the claimed invention and the prior
art is the range or value of a particular variable,” then a patent should not
issue if “the difference in range or value is minor.” Haynes Int'l v. Jessop
Steel Co., 8 F.3d 1573, 1577 n.3 (Fed. Cir. 1993)(quoted in Iron Grip
Barbell Co., Inc. v. USA Sports, Inc., 392 F.3d 1317, 1321 (Fed. Cir. 2004)).
Where, as here, the claimed and prior art products are
identical or substantially identical, or are produced by identical
or substantially identical processes, the PTO can require an
applicant to prove that the prior art products do not necessarily
or inherently possess the characteristics of his claimed product.
See In re Ludtke, supra. Whether the rejection is based on
“inherency” under 35 U.S.C. § 102, on “prima facie
obviousness” under 35 U.S.C. § 103, jointly or alternatively,
the burden of proof is the same, and its fairness is evidenced by
the PTO's inability to manufacture products or to obtain and
compare prior art products. See In re Brown, 459 F.2d 531, 59
CCPA 1036, 173 USPQ 685 (1972).
In re Best, 562 F.2d 1252, 1255 (CCPA 1977).
Finally, Baxter argues that the unexpected hemolysis-
suppression quality of DEHP rebuts any prima facie showing of
obviousness. However, when unexpected results are used as
evidence of nonobviousness, the results must be shown to be
unexpected compared with the closest prior art. In re De
Blauwe, 736 F.2d 699, 705 . . . (Fed. Cir. 1984). Here, the
closest prior art was the Becker system, utilizing a DEHP
primary bag. Mere recognition of latent properties in the prior
art does not render nonobvious an otherwise known invention.
Appeal 2009-011751
Application 10/318,659

5
In re Prindle, 297 F.2d 251, 254 . . . (CCPA 1962). Since the
prior art bags plasticized with DEHP were inherently
suppressing hemolysis, albeit unknown at the time of the
Becker document, this hemolysis-suppressing function is not a
basis for rebutting a prima facie finding of obviousness.
In re Baxter Travenol Labs., 952 F.2d 388, 392 (Fed. Cir. 1991).

FINDINGS OF FACT (FF)
1. Viala describes administering gadolinium texaphyrin during whole brain
radiation therapy to cancer patients with brain metastases (Viala, at p. 755-
56; App. Br. 11-12).
2. Appellants do not dispute that the gadolinium texaphyrin described by
Viala is the Formula I compound recited in claim 1 (App. Br. 12).
3. Patients admitted to Viala’s study had brain metastases (id. at 756, first
column, lines 5-12). “Eleven patients . . . were enrolled in this study. Ten
patients had already undergone treatment for a tumor: three for breast
cancer, five for lung cancer, one for thyroid cancer and one for skin cancer
(melanoma).” (Id. at 756, third column.)
4. “One patient had brain metastases, which revealed a lung cancer.” (Id.)
5. Doses of gadolinium texaphyrin were from 0.5 µmol/kg to 4.10 µmol/kg
(id. at 756-757).
6. Patients were treated with whole-brain radiation therapy (id. at 756, third
column.)
7. Viala described the preliminary results as showing that gadolinium
texaphyrin was “tumor selective, since enhancement of signal intensity was
Appeal 2009-011751
Application 10/318,659

6
detected only in target lesions and never in normal brain tissue.” (Id. at 759,
second column; see also “Results of Brain MR Examination” at 758-59.)
8. “Response was evaluated according to World Health Organization
criteria. Seven patients achieved a partial remission . . . One patient
achieved a minor response . . . and in three patients the disease stabilized,
with tumor regression . . . One patient had new brain metastases” (id. at 759,
first column).
9. The Formula I compound is also known as motexafin gadolinium
(Appeal Br. 2).
10. The molecular weight of motexafin gadolinium is 1148.40 gms/mole
(Boswell5 at p. 282, Fig. 1).
11. The dose of 4.10 µmol/kg of the Formula I compound administered by
Viala is equivalent to 4.70844 mg/kg calculated as follows:
1148.40 gms/mole x 4.10 µmol/kg = 4.70844 mg/kg.
12. The Specification states that the “amount of the [Formula I] compound
. . . required for use in treatment will vary” depending on several factors,
including “the nature of the condition being treated and the age and
condition of the patient” (Spec. 5:3-10).

DISCUSSION

5 Boswell, G.W., Miles, D.R., Thiemann, P.A., & Mesfin, M. Population
pharmacokinetics and bioavailability of motexafin gadolinium (Xcytrin®) in
CD1 mice following intravenous and intraperitoneal injection. 24
INVESTIGATIONAL NEW DRUGS 2810289 (2006).
Appeal 2009-011751
Application 10/318,659

7
Viala administered gadolinium texaphyrin, the claimed Formula I
compound, to patients with cancer, including five patients with lung cancer
(FF3-4). Viala reported that several patients who received the gadolinium
texaphyrin, along with whole brain radiation therapy, experienced cancer
remission (FF8). Viala’s step of administering gadolinium texaphyrin to
lung cancer patients is the same manipulative step and the same compound
recited in all the claims on appeal, administered to the same recited class of
patients. The only differences between the claims and Viala are that Viala
does not teach that its method is “for delaying neurological progression” and
Viala recites its dosages in µmoles/kg not mg/kg as in the claims.
As far as the amounts of the Formula I compound, Viala administered
a dose as high as 4.70844 mg/kg (FF11), which is the same or close to the
lowest recited value in the claimed range of “about 5.0 mg/kg to about 10
mg/kg.” Because the claimed dosage and Viala’s are so close in value, there
is sound basis, as explained below, to believe that Viala’s method would
have achieved the claimed purpose of delaying “neurologic progression,”
shifting the burden to Appellant to show that it did not. Titanium Metals
Corp. v. Banner, 778 F.2d at 783; In re Best, 562 F.2d at 1255.
The Specification reported that the Formula I compound improved
neurological function selectively in patients with lung cancer where the lung
cancer had metastasized to the brain (Spec. 2:12-17), but did not state that
the claimed dosage was responsible for this. Thus, there is no evidence that
the dosage “about 5.0 mg/kg to about 10 mg/kg” would achieve a delay in
progression while a marginally lower dosage would not. In fact, the
Appeal 2009-011751
Application 10/318,659

8
Specification expressly states that the “amount of the compound . . . required
for use in treatment will vary” depending on several factors, including “the
nature of the condition being treated and the age and condition of the
patient” (FF12). In other words, it appears that the step of administering the
Formula I compound to a lung cancer patient with brain metastases is
sufficient to achieve a delay in neurological progression, the same step
accomplished by Viala (FF1).
In any event, the claimed value of “about 5.0 mg/kg” is so close to
Viala’s of 4.70844 mg/kg (FF11), it is reasonable to believe that Viala’s
dosage would have the same effect as claimed when administered to lung
cancer patients with brain metastases.
In Exhibit 2 provided with the Appeal Brief,6 Appellant had argued
that Viala “does not point to the selective effectiveness and utility of a
compound of Formula I in patients with lung cancer and brain metastases”
(Exhibit 2, at p. 7). Appellant contended that Viala related to the tolerability
of the compound at various dosage levels (id.).
This argument is not persuasive. While Viala did not describe the
compound as selective for lung cancer, Viala explicitly disclosed that the
compound in combination with whole-brain radiation was associated with
varying degrees of brain cancer remission (FF7 & FF8). Six of the patients
in the study group had lung cancer with brain metastases (FF3-4). As the
same compound and manipulative step were performed by Viala, it is logical

6 Exhibit 2 is the Response to Office Action filed by Appellant on January
23, 2004.
Appeal 2009-011751
Application 10/318,659

9
that neurological progression was delayed, as well. “Mere recognition of
latent properties in the prior art does not render nonobvious an otherwise
known invention.” Baxter, 952 F.2d at 392.
Appellant provided evidence, unchallenged by the Examiner, that
differences in tumor size as evaluated by MRI were not “reliable indicators
of clinical benefit determined by neurologic progression or time to survival”
(App. Br. 14).
We have considered this evidence, but do not find it sufficient to rebut
the rejection. The fundamental reason we are making this rejection is that
there was a reasonable basis to believe that Viala inherently accomplished a
delay in neurologic progression when the Formula I compound was
administered to lung cancer patients with brain metastases. This basis was
reasonable because Viala administered the same compound to the same class
of patients recited in the claims and at about the same dosage. There is no
evidence that the claimed dosages achieved a different effect than the
dosages used or suggested by Viala.
Appellant’s evidence goes to the issue of whether persons of ordinary
skill in the art would have predicted that Viala’s regime produced a delay in
neurological progression. This question is different from the one at issue in
this rejection: whether there is a reasonable basis to require Appellant to
prove that Viala’s method, and dosages suggested by it, did not inherently
accomplish a delay in neurologic progression. Best, 562 F.2d at 1255.
Because the PTO does not have the facilities to compare the claimed process
with Viala’s, and because the process steps of Viala are substantially
Appeal 2009-011751
Application 10/318,659

10
identical to those claimed, the burden is properly shifted to Appellant to
demonstrate that administration of 4.70844 mg/kg of a compound of
Formula I to patients with lung cancer would not inherently result in the
neurological benefits described in the claim preambles. Id. Appellant has
provided no evidence to demonstrate that the result is not inherent.7 While
the ordinary skilled worker might not have recognized the radiologic (MRI)
evaluations in Viala to predict clinical neurologic benefit (App. Br. 14), the
rejection does not require knowledge of this fact.
To extent there are claim limitations in any of the pending claims that
require further analysis, we leave it the Examiner to address them should
prosecution be re-opened.

TIME PERIOD FOR RESPONSE
This decision contains a new ground of rejection pursuant to 37
C.F.R. § 41.50(b) (effective September 13, 2004, 69 Fed. Reg. 49960
(August 12, 2004), 1286 Off. Gaz. Pat. Office 21 (September 7, 2004)). 37
C.F.R. § 41.50(b) provides “[a] new ground of rejection pursuant to this
paragraph shall not be considered final for judicial review.”
37 C.F.R. § 41.50(b) also provides that the appellant, WITHIN TWO
MONTHS FROM THE DATE OF THE DECISION, must exercise one of

7 We have considered the Miller Declaration filed 9/14/2005, but this
Declaration does not address whether Viala’s treatment with 4.70844 mg/kg
would inherently result in the effects recited in the claim preambles in
treated lung cancer patients relative to treatment with “about 5.0 mg/kg” as
recited by the claims.
Appeal 2009-011751
Application 10/318,659

11
the following two options with respect to the new ground of rejection to
avoid termination of the appeal as to the rejected claims:
(1) Reopen prosecution. Submit an appropriate amendment of
the claims so rejected or new evidence relating to the claims so
rejected, or both, and have the matter reconsidered by the
examiner, in which event the proceeding will be remanded to
the examiner. . . .
(2) Request rehearing. Request that the proceeding be reheard
under § 41.52 by the Board upon the same record. . . .

REVERSED; 37 C.F.R. § 41.50(b)

cdc

WILSON, SONSINI, GOODRICH & ROSATI
650 PAGE MILL ROAD
PALO ALTO CA 94304-1050

Appeal 2009-011751
Application 10/318,659

12

Application/Control No.

10/318,659
Applicant(s)/Patent Under
Reexamination
Miller, Richard A.
Notice of References Cited
Examiner

Gembeh, Shirley
Art Unit

1618

Page 1 of 1

U.S. PATENT DOCUMENTS

* Document Number Country Code-Number-Kind Code Date MM-YYYY

Name

Classification

A US-

B US-

C US-

D US-

E US-

F US-

G US-

H US-

I US-

J US-

K US-

L US-

M US-
FOREIGN PATENT DOCUMENTS

* Document Number Country Code-Number-Kind Code Date MM-YYYY

Country

Name

Classification

N

O

P

Q

R

S

T
NON-PATENT DOCUMENTS
* Include as applicable: Author, Title Date, Publisher, Edition or Volume, Pertinent Pages)

U

Boswell et al., Population pharmacokinetics and bioavailability of motexafin gadolinium (Xcytrin) in CD1 mice following
intravenous and intraperitoneal injection, Investigational New Drugs 24:281-289 (2006).

V

W

X

*A copy of this reference is not being furnished with this Office action. (See MPEP § 707.05(a).)
Dates in MM-YYYY format are publication dates. Classifications may be US or foreign.

U.S. Patent and Trademark Office
PTO-892 (Rev. 01-2001) Notice of References Cited Part of Paper No.
Delete Last PagelAdd A Page
Investigational New Drugs 24: 281–289, 2006. 281
C© 2006 Springer Science + Business Media, LLC. Manufactured in The United States.
DOI: 10.1007/s10637-006-5383-1
Population pharmacokinetics and bioavailability of motexafin gadolinium
(Xcytrin
r©
) in CD1 mice following intravenous and intraperitoneal injection
Boswell GW1, Miles DR2, Thiemann PA3 and Mesfin M4
1Pharmacyclics, Inc 995 E. Aruques Ave Sunnyvale, CA; 2Pharmacyclics, Inc 995 E. Aruques Ave Sunnyvale, CA;
3Pharmacyclics, Inc 995 E. Aruques Ave Sunnyvale, CA; 4Pharmacyclics, Inc 995 E. Aruques Ave Sunnyvale, CA
Published online: 9 March 2006
Keywords Motexafin Gadolinium, Population pharmacokinetics, Mouse pharmacokinetics, Intraperitoneal dose, Intravenous
dose, Bioavailability
Abstract
Motexafin gadolinium (Xcytrin
R©
) is an expanded porphyrin macrocyclic compound under development for the treatment of
several types of cancer. Currently clinical trials and non-clinical pharmacology and toxicology studies are ongoing. The goals
of this open label, four arm, non-crossover bioavailability study were to explore motexafin gadolinium pharmacokinetics,
determine the IP bioavailability, and define a pharmacokinetic model suitable for descriptive and predictive use. Mice
received one or seven daily IV or IP injections (40 mg/kg) then blood samples were collected and analyzed. Plasma
concentration data were modelled using population pharmacokinetic methods and a two compartment model was the most
appropriate model. The stability and predictive performance of the model were evaluated using bootstrap procedures. The
accuracy of the predicted concentrations was 8.3%. Motexafin gadolinium was rapidly cleared from the plasma and although
T1/2β was 12.9 h there was no accumulation following seven doses. The IP bioavailability was 87.4% and higher plasma
concentrations were sustainable for a longer period with IP dosing. Vc was larger than the blood volume and the tissue
compartment volume was 38% of Vc, suggesting motexafin gadolinium was not widely distributed into less well perfused
tissues. The pharmacokinetic profile in this study was similar to that in oncology patients administered multiple doses of
motexafin gadolinium. The unbiased model yields reliable parameter estimates and insight into the pharmacokinetics of
motexafin gadolinium in mice, is suitable for both descriptive and predictive purposes, and is a valuable tool in the planning,
analysis, and interpretation of pharmacology and toxicology studies in mice.
Introduction
Motexafin gadolinium (Xcytrin
R©
) is a new compound un-
der investigation for the treatment of several types of tumors
[1–4]. Motexafin gadolinium is a texaphyrin (Figure 1), a
class of expanded porphyrin macrocyclic compounds that
can form stable pentadentate complexes with large cations
[5]. Currently several clinical trials (Phase 1, 2, and 3) are
being conducted with motexafin gadolinium [6–10]. Addi-
tionally, non-clinical pharmacology and toxicology studies
are ongoing using mouse and rat xenograft tumor models.
Studies have shown that motexafin gadolinium can local-
ize in tumors following IV administration. A previous study
characterized the tissue disposition of motexafin gadolinium
in SMT-F tumor bearing DBA/2N mice following IV dos-
ing [11]. Mice were dosed with 9.9 µmol/kg14C-motexafin
gadolinium then tissue and plasma sample radioactivity con-
centrations were quantified by liquid scintillation counting.
Motexafin gadolinium levels in tumor tissue were more than
double those in the plasma by 5 h post dose and remained
elevated out to 48 h post dose. Although limited plasma
concentration data was collected in this study no pharma-
cokinetic (PK) analysis of these data was reported.
Motexafin gadolinium is administered clinically by short
(usually less than 1 h) IV infusions as it is not suitable for
oral dosing. In the mouse model, long term repeated IV
administration can be problematic and therefore intraperi-
toneal (IP) administration is sometimes used. With IP
dosing, systemic exposure to the administered compound
is partially a function of the amount of drug absorbed
into systemic circulation (bioavailability, F). Therefore
proper interpretation of toxicology, pharmacology, or
pharmacokinetic study data requires knowledge of the
IP bioavailability. The current study was undertaken to
characterize the pharmacokinetics and bioavailability of
motexafin gadolinium when administered by IP injection
compared to IV dosing. An additional goal was to develop a
PK model that could be used for predictive purposes to aid
in the design and intepretation of future studies in mouse
models.
282
Figure 1 The structure of motexafin gadolinium.
Methods and materials
Motexafin gadolinium was supplied (Pharmacyclics, Inc,
Sunnyvale, CA) as a sterile, non-pyrogenic aqueous 5%
mannitol solution containing 2.5 mg/mL motexafin gadolin-
ium. It was stored under refrigeration prior to use and pro-
tected from light. Doses were administered without further
dilution. All chemicals used for bioanalytical analysis were
analytical or HPLC grade. The organic solvents were ob-
tained from Burdick and Jackson (Muskegon, MI, USA).
Glacial acetic acid, ammonium acetate, 30% ammonium hy-
droxide, and zinc sulfate heptahydrate were obtained from
Mallinckrodt Baker (Phillipsburg, NJ, USA). The water
used for mobile phase and solution preparation was ob-
tained from a Milli-Q Academic A10 unit (Megohms-com,
Billerica, MA, USA), and had a resistivity ≥18 megohms-
cm. Blank K3EDTA plasma was obtained from Biochemed
Pharmacologicals (Winchester, VA, USA).
Animal methods
This study was conducted in accordance with the National
Institute of Health “The Guide for the Care and Use of Lab-
oratory Animals” and the Animal Welfare Act Regulations
(Title 9, CFR, Chapter 1). The protocol was reviewed and
approved by the Institutional Animal Care and Use Commit-
tee (IACUC). Male and female HSD:ICR(CD-1) mice, age
46–60 days, were obtained from Harlan Laboratories (In-
dianapolis, IN) and were allowed to acclimatize for seven
days prior to study treatment. Animals were housed three
per cage and were maintained in a temperature and humid-
ity controlled environment with a 12 h light and dark cycle
and were allowed free access to food and water. Clinical
observations were recorded at 30 min post dose and daily
until sacrifice. Body weights were obtained on dosing day
1 and dosing day 7. Animals were randomly assigned by
sex to one of four treatment groups: single dose IP (Group
1, 27 per sex per group), single dose IV (Group 2, 30 per
sex per group), multiple dose IP (Group 3, 33 per sex per
group) or multiple dose IV (Group 4, 30 per sex per group).
Doses of 40 mg/kg motexafin gadolinium were delivered as
a bolus (less than 30 s). Group 1 received a single IP injec-
tion, Group 2 received a single IV (via tail vein) injection,
Group 3 received seven consecutive daily IP injections, and
Group 4 received seven consecutive daily IV (via tail vein)
injections.
Sample collection
A single blood sample (∼500 µL) for motexafin gadolin-
ium concentration determination was collected from each
animal via cardiac puncture (under isoflurane anesthesia)
prior to euthanasia. Alternatively, some animals were sam-
pled from the tail vein using a small incision at the base of
the tail immediately prior to euthanasia. All blood samples
were collected into K3EDTA. Following sample collection,
the animals were euthanized using an IACUC approved
procedure (CO2 inhalation). Generally 3 mice per sex were
used for each blood collection time point. For Groups 1 and
3 (IP administration) blood samples were collected pre-dose
(0) and at 0.5, 1.0, 1.33, 1.67, 2, 4, 9 and 12 h following the
last dose. Additional samples were collected at 16 and 24 h
post dose on dosing day 7 for Group 3. For Groups 2 and
4 (IV administration) samples were collected pre-dose (0)
and at 0.083, 0.5, 1.0, 2, 4, 8, 12, 16 or 24 h following the
last dose. Blood samples were immediately placed on ice
then the plasma was separated by centrifugation at 4◦C and
immediately frozen at −80◦ ± 10◦C until analyzed.
Analytical methods
Mouse plasma samples were assayed for motexafin gadolin-
ium concentrations using an HPLC method with fluores-
cence detection. Ten microliters of 1.5 µM PCI-0350 (a
texaphyrin analog) were added to each sample (100 µL) as
the internal standard, followed by 10 µL of 40% zinc sulfate
solution in water. Proteins were precipitated by the addition
of 100 µL of a 0.16 M solution of acetic acid in 50/50
v/v methanol/acetonitrile. Samples were vortex mixed vig-
orously for at least 8 seconds, centrifuged at 16000 × g
for 30 min, and the supernatant loaded into polypropylene
HPLC vials for analysis. The HPLC system consisted of an
Agilent 1100 Integrated HPLC System (Hewlett-Packard,
Agilent Technologies, Inc., Palo Alto, CA, USA) using a
Zorbax Eclipse XDB C-18 (3.0 × 150 mm, 3.5 µm particle
size) column (Agilent Technologies, Agilent Technologies,
Inc., Palo Alto, CA, USA). The mobile phase was 79/21 v/v
100 mM ammonium acetate (pH adjusted to 2.8 with glacial
acetic acid)/acetonitrile run isocratically at a flow rate of
0.82 mL/min and a temperature of 50◦C. Analyte and inter-
nal standard fluorescence were monitored using a FL-750
fluorescence detector (McPherson, Chelmsford, MA, USA)
configured with a 200 W Mercury/Xenon lamp, 40 µL flow
283
cell, and an FL-750-031 high sensitivity accessory contain-
ing a R3896 photomultiplier tube (Hamamatsu, Hamamatsu
City, Shizuoka Pref., Japan). The excitation wavelength was
436 nm (mercury line) and emission was detected using a
750 nm bandpass filter (Andover Corporation, Salem, New
Hampshire, USA). The lower limit of quantification for this
method was 0.005 µM (0.0057 µg/mL) with a linear range
of 0.005–1.0 µM (0.0057–1.15 µg/mL).
Pharmacokinetic model development
The plasma concentration data were tabulated by dose route
(IP or IV), dose schedule (single or multiple dose) and sex.
Mean concentration values and standard deviations (SD)
were calculated for each subgroup at each time point. Values
that appeared extreme were excluded from the data for that
time point and the mean and SD were recalculated. Data that
were outside the range of the recalculated mean ±3 times
the SD were excluded from the pharmacokinetic dataset.
The remaining data were included in the pharmacokinetic
data analysis. The concentration data were transformed to
their natural logarithm values for analysis.
Pharmacokinetic data were analyzed using NONMEM
V5, Level 1.1 (Globomax LLC, Hanover, MD) with a Visual
Fortran V 6.6 compiler. Preliminary data analysis and prior
study data indicated that a one or two compartment model
with first order output was the most appropriate structural
model. Thus ADVAN2 TRANS 2 and ADVAN4 TRANS4
using the first order conditional estimation method (FOCE)
were used to analyze these data. Because only a single
sample per animal was available the inter-animal and intra-
animal variability estimates could not be independently ob-
tained [12]. Therefore the value of the inter-animal variabil-
ity term (ETA) was fixed to 0 and the total combined error
was modeled as the intra-individual (residual) error.
Once the structural model was established potential co-
variates (body weight and sex) were evaluated for inclusion
in the model. Differences in the NONMEM objective func-
tion values, in addition to an assessment of goodness-of-fit
plots, were used to select among tested models. For nested
models (full model compared to single parameter reduced
model) a difference in the minimum value of the objective
function (OFV) of 3.84 is equivalent to a p-value of 0.05,
assuming a chi squared distribution [13]. This difference in
the OFV was used in conjunction with the goodness-of-fit
plots as criteria for accepting the full compared to the re-
duced model. Non-nested models were compared using the
Akaike information criteria (AIC) as previously described
[14] in addition to evaluation of the goodness of fit diag-
nostic plots. Using the final model, posterior conditional
(POSTHOC) estimates of individual PK parameters were
calculated.
Model validation
The stability and predictive performance of the final model
were evaluated using non-parametric bootstrap procedures
[15]. Bootstrap datasets were generated by repeated sam-
pling (with replacement) from the original dataset [16]. A
minimum of 1000 datasets were generated containing the
same number of animal subjects as the original dataset.
These bootstrap datasets were analyzed using the final PK
model to calculate the PK parameters of interest. Param-
eters calculated for each dataset were tabulated then rank
ordered. The 2.5th and 97.5th percentile values were de-
termined which then constituted the 95% confidence in-
terval (CI) for that parameter. The tabulated values were
re-ordered for the next parameter and the 95% CI deter-
mined. This process was repeated for all the parameters of
interest.
Both the predictive accuracy and bias of the final PK
model were evaluated using POSTHOC predicted motex-
afin gadolinium plasma concentrations. A prediction error
(PEoi) was determined for each ith concentration estimate
(predicted – observed) in the original dataset. Then a mean
prediction error for the original dataset (MPEo) was calcu-
lated ( PEoi/n). The MPEo was the metric used to assess
model bias [17]. The accuracy of the model predictions was
assessed using the mean absolute prediction error (MAPEo)
calculated as the mean of the the absolute values of each
PEoi.
As observed by Ette et al, calculation of a prediction
error from data previously used to define the PK model
may lead to overly optimistic values [15]. These authors
recommended calculation of an adjustment factor (OPT) to
provide a more realistic metric. OPT was calculated as fol-
lows. First, 200 bootstrap data sets (BDS1–BDS200) were
prepared as described above. Datasets BDS1–BDS200 were
analyzed using the final PK model as above resulting in
an additional 200 PK models (M1–M200) whose structure
was the same as that of final PK model but with differ-
ent parameter coefficients. POSTHOC predicted concentra-
tions were obtained for individual animals in each bootstrap
dataset. A mean prediction error (MPEb) and the absolute
value of the mean prediction error (MAPEb) were calcu-
lated for each bootstrap dataset as above. Next, compara-
ble metrics (MPEm and MAPEm) were calculated based on
models M1–M200 applied to the original dataset. Both the
parameter coefficients and random effect parameters of each
model were fixed then the original dataset was analyzed to
obtain POSTHOC motexafin gadolinium plasma concen-
tration estimates. OPT was calculated for both the mean
prediction error and absolute value of the mean prediction
error.
OPT1 =
200∑
k=1
(MPEmk − MPEbk)/n (1)
OPT2 =
200∑
k=1
(MAPEmk − MAPEbk)/n (2)
284
Revised MPER and MAPER were calculated by adding
the OPT value to the original metric.
MPER = MPEo + OPT1 (3)
MAPER = MAPEo + OPT2 (4)
Results
Animal Observations
During the study there were no significant clinical signs
or symptoms noted for any treatment group. All animals
survived to the scheduled euthanasia time. Comparing body
weights on the day of sample collection using a Student’s t
test, the mean body weights for the single dose groups (31.0
and 31.2 g, IP and IV respectively) were not significantly
different (p = 0.94). The mean body weight for the multiple
dose IV group (28.7 g) was significantly less (p = 0.003)
than for the IP group (31.1 g). However, the mean body
weight at the start of the multiple dose treatments was also
significantly different (p = 0.02) for the IP and IV dose
groups (31.0 and 29.1 g, respectively).
From a total of 257 treated animals, usable pharmacoki-
netic data was obtained from 230 mice. Samples lost for data
analysis included those due to experimental errors (dosing,
sample collection, or analytical errors, 5 samples), samples
with concentrations below the quantifiable limits of the an-
alytical method (21 samples), and samples excluded by the
outlier test described above (1 sample). Extra animals re-
ceived from the vendor were randomly assigned to treatment
groups and time points within treatment groups resulting in
some time points having more than 3 mice/sex/time point.
Bioanalytical results
A total of 15 analytical runs were used to assay the study
samples. Each run included quality control samples pre-
pared at low (0.015 µM), mid (0.075 µM) and high
(0.75 µM) concentration levels, with a minimum of n =
3 replicate determinations at each level. Based on a single
representative run, the within-run accuracy was determined
to be 105.7%, 104.7%, and 102.9% of nominal, and the
within-run precision (expressed as the percent coefficient
of variation) was determined to be 1.1%, 0.9%, and 0.7% at
the low, mid and high concentration levels, respectively. The
between-run accuracy of the method was 106.3%, 104.0%,
and 102.3% of nominal, and the between-run precision was
3.8%, 3.7% and 3.1% at the low, mid, and high levels,
respectively. Therefore the accuracy and the precision of
the bioanalytical method were adequate for the analysis of
pharmacokinetic samples.
Pharmacokinetic model selection and validation
Plasma motexafin gadolinium concentrations dropped
rapidly following 40 mg/kg doses, reaching less than 0.6%
of the observed maximum concentration by eight hours post
dose (Figure 2). These data were evaluated using both a
one-and two-compartment model with first order elimina-
tion. Based on the minimum value of the AIC and graphical
diagnostic plots, a two compartment model was selected as
the appropriate structural model. Two relevant covariates
were available for evaluation: animal weight (WT) and sex
(SEX). With only two covariates the effect of each covari-
ate was investigated individually and then in combination
on each structural model parameter. Inclusion of the covari-
ate WT on the volume of the central compartment (Vc) and
SEX on Vc or clearance (CL) resulted in a significant re-
duction in the OFV. Inclusion of SEX on Vc resulted in the
greatest decrease (13.8) in the OFV relative to the model
without covariates. Combining SEX on Vc and WT on CL
further reduced the OFV by 3.84. Other combinations did
not meet the reduction in OFV criteria or provided unstable
models. In the final model CL and Vc were estimated by
equations 5 and 6.
CL = 1 +
(
WT
30
)7
(5)
Vc = 2 + SEX ∗ 8 (6)
Several error models were explored and an additive model
(Equation 7) was selected as it provided the most improved
fit to the data.
Yi = Ci + EPS (1) (7)
where Yi is the observed concentration in the ith animal, Ci
is the estimated concentration in the ith animal and EPS(1)
is the residual variability.
The calculated MPEo was −0.000876 units and OPT1
was 0.00253 units resulting in an MPER of 0.00166 units.
Compared to the expected value of zero for an unbi-
ased model, the final model provided predicted plasma
motexafin gadolinium concentrations with a small posi-
tive bias. The MAPEo was 0.434 units and OPT2 was
0.0214 units therefore MAPER was 0.455 units. This value
is equal to a non-transformed concentration of approxi-
mately 1.58 µg/mL which is 8.3% of the mean non-trans-
formed plasma concentration of motexafin gadolinium.
The model parameter values along with their mean boot-
strap estimates and non-parametric 95% confidence inter-
vals are presented in Table 1. There was good concordance
between the final model parameters and their mean non-
parametric estimates, with less that 7% difference between
them, except for WT with a difference of 31.9%. As noted
by Ette et al [15], if the model parameters calculated using
the bootstrap method are within ± 15% of those from the
285
Figure 2 Plasma Motexafin gadolinium Concentrations. Natural logarithm of mean motexafin gadolinium plasma concentrations following single and
multiple IV and IP doses of 40 mg/kg for male (
) and female () mice. Dashed line represents the population estimated concentrations for males and
dotted line represents the estimated population concentrations for females.
final model then the final model may be considered reliable.
The combined inter- and intra-animal variability (σ ) was
estimated as 0.57 µg/mL.
Selected two compartment model pharmacokinetic pa-
rameters calculated from the model population parameter
values are shown in Table 2. Absorption from the peritoneal
cavity was relatively complete with an absolute bioavail-
Table 1 Population Model Pharmacokinetic Parameters.
Parameter
Nonmem
estimate
Mean bootstrap
estimate
Non-parametric 95%
confidence interval
θCL (mL/h) 16.4 16.3 14.5–18.4
θVc (mL) 9.05 9.19 7.85–11.2
θVp (mL) 3.47 3.59 2.61–4.96
θQ (mL/h) 0.188 0.189 0.145–0.250
θKa (h−1) 0.895 0.918 0.737–1.28
θF 0.874 0.883 0.693–1.08
θWT 4.88 3.70 −6.24–7.32
θSEX 3.29 3.08 0.765–4.64
ERRAdd 0.567 0.554 0.485–0.621
ability of 87.4%. The absorption rate (Ka) of 0.895 hr−1
was approximately 50% of rate of elimination from the
central compartment (K10) of 1.81 hr−1, indicating that
elimination from the central compartment was twice as
fast as the absorption into this compartment. Compared to
K10, β (the terminal elimination rate constant) was much
smaller, leading to a terminal half life of 12.9 h. Transfer
of motexafin gadolinium to the tissue compartment (K12
= 0.0208 hr−1) was about 38% as fast as transfer back
into the central compartment (K21 = 0.0542 hr−1) which
shows that loss of motexafin gadolinium from the central
compartment was predominately determined by K10. The
volume of the central compartment (Vc = 9.05 mL) was
much larger than the blood volume (∼2 mL). The volume
of the tissue compartment was only 38% as great (Vp =
3.47 mL) as Vc. Therefore motexafin gadolinium was not
widely distributed into less well perfused tissues. Total body
clearance (CL) of motexafin gadolinium (0.27 mL/min) was
approximately equal to the GFR in the mouse (0.28 mL/min)
[18].
The observed concentration values versus time grouped
by dose schedule and dose route are shown in Figure 2. Data
for males and females are presented separately since SEX
286
Table 2 Typical Population Pharmacokinetic Parameters.
CL Vc Vp α β T1/2β K10 K12 K21 Ka F
(mL/hr) (mL) (mL) (hr−1) (hr−1) (hr) (hr−1) (hr−1) (hr−1) (hr−1) (%)
16.4 9.05 3.47 1.834 0.054 12.9 1.813 0.021 0.054 0.895 87.4
Figure 3 Mean Observed Versus Population Predicted Motexafin Gadolinium Concentrations. Individual natural logarithm observed versus population
predicted motexafin gadolinium concentrations (•). Solid line is the line of identity.
was a significant covariate in the model. Not unexpectedly,
given the disposition and elimination characteristics of mo-
texafin gadolinium in this model, the concentration versus
time profiles were very similar for single and multiple ad-
ministration for both IP and IV routes. Although there was
inter-individual variability in the data at each time point,
predicted population concentrations were reflective of the
observed mean data. Note that the estimates of total vari-
ability were not used in the calculation of the predicted
population concentrations.
The model predicted population concentrations versus
the observed concentrations are shown in Figure 3. These
concentrations cluster around the line of identity demon-
strating that the model population concentration estimates
were in good agreement with the observed concentrations.
A plot of the weighted residuals versus predicted con-
centrations for the final model is presented in Figure 4.
This diagnostic plot shows that there was no systematic
bias in the model and that generally the residuals fell
within ±3 units of the zero line and all were within ±5
units. From this we concluded that there were no influ-
ential outliers remaining in the dataset and no obvious
bias.
Discussion
In this study we evaluated the plasma pharmacokinetics of
motexafin gadolinium in mice following single and mul-
tiple dose administration. The animals tolerated doses of
40 mg/kg well. Data were collected at sufficient time points
to well characterize the PK profile following both IP and
IV dosing. Even though the plasma concentrations declined
quite rapidly, the HPLC method employed to quantify mo-
texafin gadolinium was sufficiently sensitive and specific to
provide reliable data for up to 24 h post dose. This allowed a
good estimate of the terminal elimination half life (Table 2).
A two compartment pharmacokinetic model with first
order input and first order output proved to be the most ap-
propriate model to describe the data in this study. A one
compartment model significantly under predicted the con-
centrations at early and late time points. As seen in Fig-
ures 2 and 3 the final model performed satisfactorily with
these data with both IP and IV administration. Using a
natural logarithm transformation of the concentration data
provided a more Gaussian distribution to the residual er-
rors and significantly improved the fit of the model. The
population model parameters were well estimated with the
287
Figure 4 Natural Logarithm Transformed Plasma Concentration versus Weighted Residuals.
covariate WT being the least well estimated. This may re-
flect the fact that WT minimally met the OFV criteria for
inclusion in the model. However, the inclusion of WT in
the model improved the overall fit and also parsed out the
effect of weight from other sex effects since interchang-
ing SEX and WT on the model parameters degraded the
model.
Based on the bootstrap evaluations the final model proved
to be reliable for predictive use. The small over prediction
(MPER) bias seen is likely to be of little practical signif-
icance in the utilization of this model. The accuracy of
the predicted concentrations relative to the observed con-
centrations was good (8.3% error). Because no estimates
of inter-animal variability were available, model predicted
individual animal estimates were not calculated; only pop-
ulation estimates were calculated. However, when the total
variability was included in the POSTHOC estimates for
individual animal concentrations, a plot very similar to Fig-
ure 2 was observed albeit with somewhat greater dispersion
about the line of identity. Since such a plot included not
only inter-and intra-animal variability but residual variabil-
ity, the predictability of the model was better represented
by the population estimates.
Following IP administration of motexafin gadolinium, the
predicted maximum plasma concentration was reached at
approximately 45 min post dose and was similar to the con-
centration following IV administration at that time. How-
ever, IP plasma concentrations were greater than IV concen-
trations from that time forward because IP absorption was
not complete until approximately 3 h post dose. Thus higher
concentrations of motexafin gadolinium are more sustain-
able following IP administration. With an F of 87.4%, IP
dosing would seem to be a viable way to provide longer
sustained plasma concentrations.
As seen in Figure 2, motexafin gadolinium is rapidly
eliminated from the central (sampling) compartment. The
distribution phase (α) for motexafin gadolinium was es-
sentially complete by 1.5 h post dose at which time the
plasma concentration following IV dosing was only 6.4%
of the maximum concentration (Cmax). Because of this, even
though the biologic (terminal) half life was approximately
12.9 h there was no significant plasma accumulation of mo-
texafin gadolinium with dosing to steady state. Following
IP administration motexafin gadolinium plasma concentra-
tions did not decline to that same level until 5 h post dose
because of the IP absorption profile. However, there was
similarly no plasma accumulation upon multiple IP dosing.
Based on Vp, motexafin gadolinium was not widely dis-
tributed into peripheral tissues. Motexafin gadolinium ap-
peared to localize more in the blood and highly perfused
tissues which may explain a CL of 16.4 mL/h. The appar-
ent biodistribution of motexafin gadolinium in this study
was similar to that previously reported [11]. Although CL
was approximately the same as GFR in the mouse, we have
shown that motexafin gadolinium is readily metabolized in
vitro in human and rat liver fractions [19]. Additionally,
in a human mass balance study, the majority of motexafin
gadolinium was metabolized and excreted in the feces with
only 14% of parent motexafin gadolinium eliminated in the
urine (unpublished data). It is therefore probable that in
mice motexafin gadolinium is also primarily metabolized
and excreted in the feces. However, the route of clearance
of motexafin gadolinium was not addressed in the current
study.
288
Motexafin gadolinium pharmacokinetics has been eval-
uated in 243 oncology patients administered 4–5.3 mg/kg
motexafin gadolinium for 2–6.5 weeks. [20]. Using popula-
tion pharmacokinetic methods, a three compartment model
was found to be most the appropriate to describe the plasma
concentration profile. Similar to the current study, in these
patients motexafin gadolinium rapidly distributed out of the
plasma following IV administration and had a prolonged
terminal half life (80.8 h). Plasma clearance and the volume
of distribution for the central compartment were greater in
males compared to females. In the current study, plasma
clearance was not a significant function of the sex of the
animal although the volume of distribution for the central
compartment was 36.4% greater in males than in females.
However, since animal body weight was a significant co-
variate for plasma clearance and body weight is highly cor-
related with the sex of the animal (males are typically larger
than females at the same age), it is possible that a body
weight effect would also include an effect related to the sex
of the animals. Thus the pharmacokinetic profile found in
mice in this study is generally reflective of the motexafin
gadolinium pharmacokinetics seen in oncology patients that
have been studied.
The pharmacokinetic model developed in this study is
currently being utilized in the development of new dose
regimens and new indications for motexafin gadolinium.
With an understanding of the bioavailability of motex-
afin gadolinium it is now possible to evaluate extended
dosing schemes (≥ 30 days) using IP dosing and com-
pare these data to the shorter term IV doses previously
studied. Based on the clearance and half life of motex-
afin gadolinium, targeted plasma concentrations can now
be achieved. Additionally, using this model more defini-
tive measures of exposure (area under the plasma con-
centration versus time curve) are now available to aid in
the toxicological assessment of these extended dosing regi-
mens. By simulating various dosing schemes and calculat-
ing the resulting exposures, currently available data can be
meaningfully extrapolated to other scenarios to guide future
studies.
In this study we have characterized the pharmacoki-
netic profile of motexafin gadolinium in male and female
CD-1 mice following IV and IP administration of single
and multiple doses at a single dose level. The pharma-
cokinetic parameters calculated from these data indicate
that motexafin gadolinium was rapidly distributed out of
the central compartment with a longer terminal elimina-
tion half life but there was a lack of accumulation upon
multiple dosing. An unbiased and reliable two compart-
ment population pharmacokinetic model was developed
and validated using bootstraping methods. This model is
suitable for both descriptive and predictive purposes and
can serve as a valuable tool in the planning, analysis, and
interpretation of pharmacology and toxicology studies in
mice.
References
1. Miller RA, Woodburn KW, Fan Q, Lee I, Miles D, Duran G, Sikic B,
Magda D: Motexafin gadolinium: a redox active drug that enhances the
efficacy of bleomycin and doxorubicin. Clin Cancer Res 7:3215–3221,
2001
2. Miller RA, Woodburn K, Fan Q, Renschler MF, Sessler JL, Koutcher
JA: In vivo animal studies with gadolinium (III) texaphyrin as a radi-
ation enhancer. Int J Radiat Oncol 45:981–989, 1999
3. Xu S, Zakian K, Thaler H, Matei C, Alfieri A, Chen Y, Koutcher JA:
Effects of motexafin gadolinium on tumor metabolism and radiation
sensitivity. Int J Radiat Oncol 9:1381–1390, 2001
4. Young SW, Qing F, Harriman A, Sessler JL, Dow WC, Mody TD,
Hemmi GW, Hao Y, Miller RA: Gadolinium(III) texaphyrin: a tumor
selective radiation sensitizer that is detectable by MRI. [published
correction appears in Proc Natl Acad Sci USA. 96:2569, 1999]. Proc
Natl Acad Sci USA 93:6610–6615, 1996
5. Mody TD, Fu L, Sessler JL: Texaphyrins: synthesis and development
of a novel class of therapeutic agents, in: Karlin KD (ed): Progress in
Inorganic Chemistry. New York, NY, John Wiley and Sons Inc, pp.
551–598, 2001
6. Evens AMl: Motexafin gadolinium: a redox-active tumor selective
agent for the treatment of cancer, Curr Opin Oncol 16:576–580, 2004
7. Carde P, Timmerman R, Mehta MP, Koprowski CD, Ford J, Tishler
RB, Miles D, Miller RA, Renschler MF: Multicenter Phase Ib/II Trial
of the Radiation Enhancer Motexafin Gadolinium in Patients With
Brain Metastases, J Clin Oncol 19(7):2074–2083, 2001
8. Mehta MP, Rodrigus P, Terhaard C, Rao A, Suh J, Roa W, Souhami
L, Bezjak A, Leibenhaut M, Komaki R, Schultz C, Timmerman R,
Curran W, Smith J, Phan S, Miller R, Renschler M: Survival and Neu-
rological Outcomes in a Randomized Trial of Motexafin Gadolinium
and Whole-Brain Radiation Therapy in Brain Metastases, J Clin Oncol
21(13):2529–2536, 2003
9. Meyers CA, Mehta MP, Rodrigus P, Dehnad H, Suh J, Roa W, Souhami
L, Bezjak A, Komaki R, Schultz C, Timmerman R, Curran WJ, Phan S,
Miller R, Renschler MF: Neurocognitive Function and Progression in
Patients With Brain Metastases Treated With Whole-Brain Radiation
and Motexafin Gadolinium: Results of a Randomized Phase III Trial,
J Clin Oncol 22:157–165, 2004
10. Jac J, Hernandez J, Phan S-C, Amato RJ: Phase II trial of motexafin
gadolinium (MGd) for treatment of metastatic renal cell carcinoma
(MRCC). ASCO Abstract # 4724 J Clin Oncol 23(16S):433s, 2005
11. Miller RA, Woodburn K, Qing F, Renschler MF, Sessler JL, Koutcher
JA: In vivo animal studies with gadolinium (III) texaphyrin as a radi-
ation enhancer. Int J Radiat Oncol 45(4):981–989, 1999
12. Ette EI, Kelman AW, Howie CA, Whiting B: Analysis of An-
imal Pharmacokinetic Data: performance of the One Point Per
Animal Design. J Pharmacokinet and Biopharm 23(6):551–566,
1995
13. Boeckmann AJ, Sheiner LB, Beal SB: NONMEM User Guide Vols
1-5, San Francisco:NONMEM Project Group, 1994
14. Imbimbo BP, Martinelli P, Rocchetti M, Ferrari G, Bassotti G, Im-
bimbo E: Efficiency of different criteria for selecting pharmacoki-
netic multiexponential equations. Biopharm Drug Dispos 12:139–147,
1991
15. Ette EI, Williams PJ, Kim YH, Lane JR, Liu MJ, Capparelli EV:
Model appropriateness and population pharmacokinetic modeling. J
Clin Pharmacol 43(6):610–623, 2003
16. Parke J, Holford NHG, Charles BC: A procedure for generating boot-
strap samples for the validation of nonlinear mixed-effect population
models. Comput Meth Prog Bio 59:19–29, 1999
17. Sheiner LB, Beal SL: Same Suggestions for Measuring Predictive
Performance. J Pharmacokin Biopharm 9(4):503–512, 1981
18. Davies B, Morris T: Physiological Parameters in Laboratory Animals
and Humans. Pharm Res 10(7):1093–1095, 1993
289
19. Mani C, Upadhyay S, Lacy S, Boswell GW, Miles DR: Reductase-
mediated metabolism of motexafin gadolinium (Xcytrin
R©
) in rat and
human liver subcellular fractions and purified enzyme preparations. J
Pharm Sci 94(3):559–570, 2005
20. Miles DR, Smith JA, Phan S, Hutchinson SJ, Renschler MF, Ford JF,
Boswell GW: Population Pharmacokinetics of Motexafin Gadolinium
in Adults With Brain Metastases or Glioblastoma Multiforme. J of
Clin Pharmacol 45(3):299–312, 2005
Address for offprints: Boswell GW, Pharmacyclics, Inc 995 E. Aruques
Ave Sunnyvale, CA 95085. Tel.: 408–328-3635, Fax: 408-328-3689; E-
mail: gboswell@pcyc.com