Local Drug Delivery for Treatment of
Brain Tumor Associated Edema
MASSACHUSETTS
OFTE
IITU TE
Y
by
NO! 1 3 2014
Qunya Ong
L_
___j
LIBRARIES
B.S.E. Biomedical Engineering
University of Michigan - Ann Arbor, 2007
Submitted to the Harvard-MIT Division of Health Sciences and Technology
in partial fulfillment of the requirements for the degree of
Doctorate of Philosophy in Medical Engineering and Medical Physics
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
September 2014
2014 Massachusetts Institute of Technology. All rights reserved.
Signature redacted
Signature of Author
Harvard-MIT Program in Health Sciences and Technology
August 18, 2014
Signature redacted
Certified by ................
Ce......edb.........................--
--
---------------------- ...
Michael J. Cima, PhD
David H. Koch Professor of Engineering
Faculty Director, Lemelson-MIT Program
Thesis Supervisor
-
-
-
-
Signature redacted
Accepted byEmery N. Brown, MD, PhD
Director, Harvard-MIT Program in Health Sciences and Technology
Professor of Computational Neuroscience and Health Sciences and Technology
Local Drug Delivery for Treatment of Brain Tumor Associated Edema
by
Qunya Ong
Submitted to the Harvard-MIT Division of Health Sciences and Technology
on August 18', 2014 in partial fulfillment of the requirements for the degree of
Doctorate of Philosophy in Medical Engineering and Medical Physics
Abstract
Brain tumor associated edema, a common feature of malignant brain neoplasms, is a
significant cause of morbidity from brain tumor. Systemic administration of corticosteroids, the
standard of care, is highly effective but can introduce serious systemic complications. Agents
that inhibit the vascular endothelial growth factor (VEGF) pathway, such as cediranib, are
promising alternatives, but are also associated with systemic toxicity as VEGF is essential for
normal physiological functions.
A miniature drug delivery device was developed for local drug delivery in rodents. It
comprises of a drug reservoir and a cap with orifice(s) through which drug is released. Drug
release kinetics is dependent on the payload, the drug solubility, and the surface area for
diffusion. Sustained releases of dexamethasone (DXM), dexamethasone sodium phosphate
(DSP), and solid dispersion of cediranib (AZD/PVP) were achieved. Employing the solid
dispersion technique to increase the solubility of cediranib was necessary to enhance its release.
Therapeutic efficacy and systemic toxicity of local drug administration via our devices
were examined in an intracranial 9L gliosarcoma rat model. Local delivery of DSP was effective
in reducing edema but led to DXM induced weight loss at high doses in a pilot study. DXM,
which is much less water-soluble than DSP, was used subsequently to reduce the dose delivered.
The use of DXM enabled long-term, sustained zero-order release and a higher payload than DSP.
Local deliveries of DXM and AZD/PVP were demonstrated to be as effective as systemic dosing
in alleviating edema. Edema reduction was associated with survival benefit, despite continuous
tumor progression. Animals treated with locally delivered DXM did not suffer from body weight
loss and corticosterone suppression, which are adverse effects induced by systemic DXM.
Local drug administration using our device is superior to traditional systemic
administration as it minimizes systemic toxicity and allows increased drug concentration in the
tumor by circumventing the blood brain barrier. A much lower dose can therefore be utilized to
achieve similar efficacy. Our drug delivery system can be used with other therapeutic agents
targeting brain tumor to achieve therapeutic efficacy without systemic toxicity.
Thesis Supervisor: Michael J. Cima, PhD
Title: David H. Koch Professor of Engineering
3
4
ACKNOWLEDGEMENTS
I have been very blessed to be surrounded with many great people.
I am very thankful for my thesis supervisor, Professor Cima. He has taught me many
lessons and made time for me despite his busy schedule. Thank you, Michael, for your
continuous guidance and unwavering support throughout my time in your lab.
It is a great privilege to have Professor Jain and Professor Langer on my thesis
committee. I am grateful to them for sharing their constructive insights and precious time with
me. I would also like to thank Dr. Hochberg for his encouragement and advice since the
beginning of this projecL My passion for the project is renewed after every discussion with him.
I would like to thank Professor Charest, Dr. Zhu, Dr. Betty Tyler and Dr. Chen for
sharing their expertise with me. I would also like to thank Dennis Ward, Dr. Scott Malstrom, Dr.
Alison Hayward, Wayne Au and many other staff at MIT facilities. This thesis work would not
have been possible without their help.
I am thankful for the amazing group of people at the Cima lab. Lenny, Barbara, Urvashi,
Chris, Byron, Alex, Yoda, Yibo, Syed, Jen, Hong Linh, Noel, Negar, Yuan, Ollie, Maple, Jay,
Kevin, Vincent, Greg, Laura, Matt, Katerina, Lina, Rachel, Joan and Nuria have made my lab
experience a great one. Lenny and Barbara are two great pillars of the Cima lab that have helped
me tremendously along the way. Yoda and Alex were instrumental in showing me the ropes
when I first joined the lab. Syed and Vincent helped me with MRI even on weekends and
holidays. Chris, Byron and Jay were always generous in sharing their knowledge and offering
their help. Urvashi is like a big sister to me. Her support and counsel in and out of lab were much
appreciated. Laura is one of the sweetest persons I have ever known and she is always so
generous in sharing her wonderful cupcakes. The camaraderie and the mentorship I have
received from many in this lab have profoundly influence me in many ways.
My friends have made this journey a lot more bearable. My EBC family, especially
Pastor Colin, Laurie, Helga, Shannon, Suzanne, Michelle, Katie, Jessica, Wade, Kelsea and MJ,
has been a great blessing to me. Their love and prayers had seen me through difficult phases in
my life. I am thankful for the friendship and company of Alvin, Chyan Ying, Teck Chuan,
Hongye, Wang Jia, Ninghan, Mingjuan, Richie, Shengyong, Sheng Rong, Yoong Keok, Yifeng,
Chia Min, and many others. I am especially grateful to Melissa and Xiaosai for supporting me
through tough times. All of them made Boston felt like a home away from home for me. Huilin
and Kaizhen are the greatest friends I could ever ask for. Thank you so much for always being
there for me for the last 17 years. I cannot imagine how my life would be without you two.
My family has been my pillar of support. My mummy and brother, Qunxiang would do
anything to help me graduate. Although my father is no longer with us, he has been an
inspiration to me whenever I feel like giving up. This thesis is dedicated to my dearest papa.
5
6
TABLE OF CONTENTS
List of Figures...............................................................................................................................
11
List of Tables ........................................................................................................................
15
CHAPTER 1. INTROD UCTION ..............................................................................................
17
1.1.
Significance....................................................................................................................17
1.2.
Brain Tumor Associated Edema ..................................................................................
17
1.2.1.
Pathogenesis ............................................................................................................
17
1.2.2.
Standard of Care - Corticosteroids......................................................................
19
1.2.3.
Novel Drugs........................................................................................................
24
1.3.
Drug delivery in the brain ...........................................................................................
1.3.1.
Traditional Systemic Delivery..................................27
27
1.3.2.
Local Delivery ........................................................................................................
29
Thesis Objective .............................................................................................................
37
1.4.
CHAPTER 2. CHARACTERIZATION OF BRAIN EDEMA ANIMAL MODELS............... 39
2.1. Introduction................... ..... -........ ................... ................................... 39
2.2. Materials and Methods..............................................................................................
41
2.2.1.
Anim al Care........................................................................................................
41
2.2.2.
Tum or Induction in Transgenic M ice .................................................................
41
2.2.3.
9L Tum or Implantation in Rats...........................................................................
42
2.2.4.
Biolum inescence Im aging-..........
42
2.2.5.
M agnetic Resonance Imaging.............................................................................
43
2.2.6.
Brain Water Content............................................................................................
43
2.2.7.
Histology ..............-
43
2.2.8.
Statistical Analysis ..................................................................................................
2.3.
-..
...
.............................
---------....... ..................
.................................
...............................................
Results and Discussion..............................................................................................
44
44
2.3.1.
Characterization of Transgenic Mouse Model...................................................
44
2.3.2.
Characterization of 9L Rat M odel.....................................................................
49
2.4.
Conclusions...................................................................................................................50
7
CHAPTER 3. LOCAL DELIVERY OF DEXAMETHASONE SODIUM PHOSPHATE ..... 51
3.1.
Introduction....................................................................................................................
51
3.2.
Materials and Methods................................................................................................
52
3.2.1.
Materials .................................................................................................................
52
3.2.2.
Device Fabrication and A ssembly ......................................................................
53
3.2.3.
In Vitro and In Vivo Drug Release......................................................................
53
3.2.4.
H PLC Analysis ...................................................................................................
54
3.2.5.
In Vivo Experim ent ..............................................................................................
55
3.2.6.
Brain W ater Content ............................................................................................
56
3.2.7.
Statistical Analysis..............................................................................................
56
3.3.
Results and Discussion...............................................................................................
57
3.3.1.
Drug Release Characteristics.............................................................................
57
3.3.2.
Brain Water Content ............................................................................................
62
3.3.3.
System ic Drug Exposure.....................................................................................
63
3.3.4.
Dexamethasone-Induced Weight Loss ..............................................................
64
3.4.
Conclusions....................................................................................................................
CHAPTER 4. LOCAL DELIVERY OF DEXAMETHASONE ..................................................
66
67
4.1.
Introduction....................................................................................................................
67
4.2.
M aterials and M ethods..............................................................................................
68
4.2.1.
M aterials ..............................................................................................................
68
4.2.2.
Device Fabrication and Assem bly ......................................................................
68
4.2.3.
In Vitro Drug Release .........................................................................................
69
4.2.4.
Drug Stability..................................................................................................
4.2.5.
HPLC Analysis ...................................................................................................
4.2.6.
Anim al Care.........................................................................................
4.2.7.
Drug Distribution in Rat Brain ...........................................................................
72
4.2.8.
Efficacy Study......................................................................................................
73
4.2.9.
M agnetic Resonance Im aging.............................................................................
74
.. 71
71
..... ... 71
4.2.10.
Brain Water Content........................................................................................
74
4.2.11.
Plasma Dexamethasone and Corticosterone ....................................................
74
8
4.2.12.
4.3.
Statistical Analysis.........................................................................................
Results...........---------.
.........................
..........................................................
75
76
4.3.1.
In Vitro Release....................................................................................................76
4.3.2.
Drug Stability.........................................................
4.3.3.
Drug Distribution in Rat Brain ..........................................................................
79
4.3.4.
Tumor and Edema Quantification from MRI.........................................................
80
4.3.5.
Brain Water Content............................................................................................
83
4.3.6.
Survival..................................................
84
4.3.7.
Dexamethasone-Induced Weight Loss..............................................................
85
4.3.8.
Plasma Dexamethasone and Corticosterone.......................................................
86
4.4.
4.5.
Discussion ..------------
..
...........................................
.. . ..... ...........
...................................................
Conclusions......................................................................................
CHAPTER 5. LOCAL DELIVERY OF CEDIRANIB............................................................
79
88
92
93
5.1.
Introduction ------------.-----...........................................................................
93
5.2.
Materials and Methods..............................................................................................
95
5.2.1.
Materials ...................
5.2.2.
............................ 95
Preparation of Solid Dispersion of Cediranib......................................................
95
5.2.3.
Device Fabrication and Assembly ......................................................................
5.2.4.
In Vitro Drug Release.......................................96
5.2.5.
Drug Stability.
5.2.6.
HPLC Analysis................
5.2.7.
In Vitro HUVEC Proliferation Assay .................................................................
97
5.2.8.
In ivo Experiment .............................................................................................
98
5.2.9.
Brain W ater Content.............................................................................................100
.............................................................................................
5.2.10.y
5.3.
............
.......................................................
......................................................................................
Results.-----------------
.................................................................................
5.3.1.
In Vitro Drug Release......................................101
5.3.2.
Drug Stability.
5.3.3.
Bioactivity...........-..
5.3.4.
In Vivo Efficacy Study ........................................................................................
.......................................................................................
----...........
..................................................................
9
96
96
.... 97
.. 100
101
102
103
103
5.3.5.
System ic Toxicity ................................................................ ................................. 108
5.4.
Discussion .................................................................................................................... 110
5.5.
Conclusions .................................................................................................................. 114
REFERENCES ............................................................................................................................ 115
10
LIST OF FIGURES
Figure 1.1 Illustration of (a) a normal blood-brain barrier (BBB) and (b) the formation of brain
tumor associated edema through a disrupted BBB in the setting of a brain tumor. (Adapted from
K aal et al. [16]).............................................................................................................................
19
Figure 1.2 Schematic representation of corticosteroid-induced suppression of the HPA axis.... 22
Figure 1.3 Intracavitary delivery via an Omniaya reservoir into the tumor resection cavity.
(Reproduced from Reulen [691)................................................
..........................................
31
Figure 2.1 (A) Schematic illustration of the activation of the mutated (vIII) human EGFR in
transgenic mice. A strong ubiquitous promoter (CAGGS) is positioned upstream of a
transcriptional stop cassette (STOP) which is flanked by 2 loxP sites (blue triangles), followed
by EGFRVm cDNA and a polyA signal sequence (purple rectangle). Localized expression of
EGFRVm is conditional to the removal of the stop cassette, which is mediated by Ad-Cre. (B)
Experimental design. Tumor growth was monitored longitudinally by BLI post tumor induction
by injection of Ad-Cre. When mice were determined to have bioluminescent signal output of
more than 106 p/ s/ cm2/ sr, MRI was performed to further assess tumor and its associated edema.
------------...... -----.........................................................................................................................
44
Figure 2.2 Longitudinal monitoring of tumor growth by BLI. (A) Representative bioluminescent
images for a mouse imaged 7, 14, 21, and 28 days after Ad-Cre injection. (B) Individual
measurements of bioluminescent outputs from mice. The solid lines represent outputs from mice
which had tumors, while the dotted lines represent outputs from mice which did not form tumor.
....................................................------------------.....
. --..................................................................
45
Figure 2.3 Kaplan-Meier survival curves of mice with tumor (n = 14) and mice with no tumor
despite Ad-Cre injection (n = 3) ................................................................................................
47
Figure 2.4 Enlarged spleen, a sign of lymphoma, in a typical mouse. ....................................
47
-
Figure 2.5 Tumor and edema assessment using MRI. (A) Representative consecutive T2
weighted coronal MR images of a mouse. Hydrocephalus and tumor with a necrotic core were
noted. (B) A T2-weighted image and its corresponding T2 map. Tumor region could be
distinguished from normal brain tissue on the T2-weighted image, but not with certainly on the
T2 map. The green highlighted regions of interest, normal cerebral cortex on the left and tumor
on the right, were used to make T2 measurements on the 12 map. 12 signal of tumor was higher
than that of cerebral cortex. ........--------------. ..
--....... ............ ............................................. 48
Figure 2.6 Water content of the ipsilateral cerebrum and contralateral cerebrum. The water
content of the ipsilateral cerebrum was significantly higher than that of the contralateral
cerebrum in tumor-bearing mice (*P = 0.016), while the water content of the ipsilateral cerebrum
11
and the contralateral cerebrum were lower and not significantly different from each other in
control animals. Data represent mean values S.E.M. (n = 3 for control, n = 14 for tumor)...... 48
Figure 2.7 Characteristics of 9L gliosarcoma rat model. (A) Representative T2-weighted coronal
MR images of a rat imaged 5, 7 and 9 days after intracranial tumor implantation. Tumor
(outlined by a white dotted line) and its associated edema (outlined by a black dotted line) can be
seen on day 9. (B) Water content of the ipsilateral cerebrum of tumor-bearing rats was
significantly higher than that of the control rats (****P < 0.0001). (C) Kaplan-Meier survival
curve of rats with intracranial 9L gliosarcoma. .......................................................................
49
Figure 3.1 Drug delivery device. (A) Illustration of the device. (B) Schematics of how the device
works. Some drug inside the reservoir dissolved in PBS with the influx of PBS after vacuum.
Drug then diffuses across the orifice down a concentration gradient into the external solution. (C)
Image of a laser-drilled orifice, 100 pm in diameter, taken under a laser scanning microscope.. 57
Figure 3.2 DSP release characteristics. (A) In vitro and in vivo release profiles of DSP from
devices with different drug payloads. In vitro release was performed in PBS while in vivo release
at day 7 was obtained indirectly by subtracting the amount of drug left in explanted device from
the original drug payload. Data represent mean values S.E.M. (n = 3 for each payload). (B)
Same in vitro release data from (A) plotted against the square root of time. Cumulative release
was demonstrated to vary linearly with square root of time for each payload. The lines represent
2
linear regression lines of best fit. (C) A linear relationship (R = 0.9) was observed between the
slopes of the lines of best fit and the drug payloads................................................................
60
Figure 3.3 (A) Schematic illustration of the Higuchi model. (B) Theoretical concentration profile
within the device reservoir where the drug transport is not fast. The device cap which has holes
through which the drug is released, acts as a rate-limiting membrane. The solid line in each
diagram represents the concentration gradient after time, t. The shift in concentration profile after
61
interval of At is shown as a dotted line....................................................................................
Figure 3.4 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals
treated with systemic or local administration of DSP had significantly lower water content in the
tumor and ipsilateral cerebrum than the untreated animals (P< 0.001). Animals that received
PBS device are not significantly different from the untreated animals. + 411 jg/day is the DSP IP
dose for a 250g rat. Data represent the mean S.E.M. (n = 5 per group)................................ 62
Figure 3.5 Scatter plot of the percentage weight gain from day 5 to day 12 post tumor
implantation, after 7 days of treatment. Dose-dependent weight loss was observed in animals
treated with local delivery of DSP. Systemic DSP dosing also led to significant weight loss. The
lines represent the mean S.E.M. (n = 5 per group). *** P < 0.001, **** P < 0.0001, relative to the
control group. '411 pg/day is the dose for a 250g rat. ...........................................................
65
Figure 4.1 Device fabrication and assembly. Fabrication of (A) device reservoir and (B) device
cap were performed separately. Schematics of the micro-machining processes were shown. (C)
Device assembly involved loading the drug into the reservoir, covering the reservoir with a cap,
70
and then sealing the cap to the reservoir with biomedical grade epoxy. ..................................
12
Figure 4.2 Illustration of the spatial concentration profile of the device. c2 , the drug
concentration inside the device, is assumed to be the saturated drug concentration in the early
phase of release while c1 , the drug concentration outside the device, is assumed to be zero in
perfect sink conditions. 8, the distance over which the concentration gradient forms, is assumed
to be the thickness of the cap.....----------------------............ ............. ............................................ 76
Figure 4.3 In vitro releases of DXM from drug delivery devices. (A) Cumulative release curves
of DXM from devices with lhole (DXM 1H Device), 5 holes (DXM 5H Device) and 9 holes
(DXM 9H Device). Releases of DXM from these devices were shown to obey zero-order
kinetics. The lines represent linear regression lines of best fit. (B) Plot of the mean release rate
against the number of orifices in the device. A linear relationship (R 2 = 0.998) was observed
between the mean release rates, determined from the slopes in (A), and the number of orifices.
Data represent mean values S.E.M. (n = 5) ... ................................
................................. 78
Figure 4.4 Stabilities of DXM in solid form and in PBS over the course of 10 days. Data were
expressed as the mean percentage of the initial amount of drug on day 0 S.E.M. (n = 3 for each
time point).....................................................................................................................................79
Figure 4.5 Distribution of DXM in rat brain resulting from systemic DXM dosing (DXM IP),
intracranial implantation of DXM-loaded devices withi hole (DXM 1H Device) and 9 holes
(DXM 9H Device). Data represent mean values S.E.M. (n = 3)....................
80
Figure 4.6 Tumor and edema assessment using MRI. (A) Representative T2-weighted coronal
MR images before (day 9) and after (day 12) treatment. (B) Tumor volumes measured on MR
images. There was no significant difference among all the groups on both day 9 and day 12. (C)
Relative change in tumor volume calculated from volumes in (B) confirmed that increase in
tumor was not significantly different. (D) Edema volumes measured on MR images. Edema on
day 12 was significantly reduced in groups treated with systemic or local DXM compared with
controls ( P < 0.001). (E) Relative change in edema volume calculated from volumes in (D)
confirmed that the decrease in edema was significant relative to control (****P < 0.0001). Data
are expressed as the mean values S.E.M. (n = 6 per group)..................................................
82
Figure 4.7 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals
treated with systemic or local administration of DXM had significantly lower water content in
the tumor and in the ipsilateral cerebrum than the control animals (*P < 0.05, **P< 0.01). Data
represent mean values S.E.M. (n = 3 per group)...................................................................
83
Figure 4.8 Kaplan-Meier survival curves. Systemic and local delivery of DXM led to significant
survival benefit (P< 0.05, n= 6 per group). ..........-- ..........-.......... .................................. 84
Figure 4.9 Relative body weight. (A) Plots of relative body weight versus time since the start of
treatment on day 9. All the groups had decreasing weight with time, with the DXM IP group
experiencing the most weight loss. (B) Comparison of the mean relative weights on day 12 and
day 16. The DXM IP group had significantly lower relative body weight than all other groups on
both days (* P < 0.05). Data represent mean values S.E.M. (n = 6 per group)....................... 85
13
Figure 4.10 Plasma levels of (A) DXM and (B) corticosterone at the time of sacrifice. Plasma
DXM level in the DXM IP group was 3 orders of magnitude higher than the plasma DXM levels
in the DXM 1H device and DXM 9H device groups. Plasma corticosterone level in the DXM IP
group was significantly lower than that in any other group (*P < 0.05). Data represent the mean
values
87
S.E.M . (n = 6 per group)..............................................................................................
Figure 5.1 (A) Illustration and (B) photograph of an assembled drug delivery device with (C) 9
101
holes, each 300 pm in diameter. .................................................................................
Figure 5.2 Cumulative release profiles of DXM and AZD/PVP. Data represent mean values
S.E.M . (n= 5 for each drug).......................................................................................................
101
Figure 5.3 Stabilities of AZD/PVP in solid form and in PBS for 28 days. Data were expressed as
the mean percentage of the initial amount of drug on day 0
S.E.M. (n = 3 for each time point).
.....................................................................................................................................................
102
Figure 5.4 Dose-response curves of AZD/PVP and cediranib (AZD). Both drugs inhibited
VEGF-induced HUVEC proliferation in similar dose-dependent fashion. Data represent mean
103
values S.E.M., performed on triplicate experiments. ..............................................................
Figure 5.5 Tumor and edema assessment using MRL (A) Representative T2 -weighted coronal
MR images before (day 9) and after (day 12) treatment. (B) Tumor volumes measured on MR
images. (C) Relative change in tumor volume calculated from the volumes in (B). Both (B) and
(C) show that there was no significant difference in tumor growth among all the groups. (D)
Edema volumes measured on MR images. Edema on day 12 was significantly reduced in groups
treated with DXM or cediranib compared with the control group (***P < 0.001). (E) Relative
change in edema volume calculated from the volumes in (D) confirmed that the decrease in
edema was significant relative to control (****p < 0.0001). Data represent mean values S.E.M.
(n = 7 per group).........................................................................................................................
105
Figure 5.6 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals
treated with systemic or local administration of DXM and cediranib had significantly lower
water content in the tumor and ipsilateral cerebrum than the control animals (*P < 0.05). Data
106
represent the mean S.E.M. (n = 4 per group)...........................................................................
Figure 5.7 Kaplan-Meier survival curves. Systemic and local delivery of both dexamethasone
and cediranib led to significant survival benefit (P< 0.05, n = 7 per group)............................. 107
Figure 5.8 Relative body weight as a measure of systemic toxicity. (A) Plots of body weight
versus time since start of treatment on day 9. All groups had decreasing weight with time, with
the DXM IP group experiencing the most weight loss. (B) Relative body weights of the groups
on day 12 and day 16. Mean relative weight of the DXM IP groupis significantly lower than that
of any other group on day 12 and day 16 (*P < 0.05, **P < 0.01, P < 0.001,
Data represent mean S.E.M. (n = 7 per group)..............................
14
P < 0.0001).
109
LIST OF TABLES
Table 1.1 Comparison of corticosteroids (Data from Kaal et al. [16])........................................
20
Table 1.2 Systemic side effects of corticosteroids (Adapted from Wen et al. and Kesari et aL [5,
36])................................................................................................................................................
23
Table 1.3 Anti-VEGF agents (Data from Norden et al. [431) .................................................
25
Table 1.4 Comparison of local delivery strategies...................................................................
30
Table 3.1 In vitro release data..................................................................................................
61
Table 3.2 Comparison of water content of tumor, ipsilateral cerebrum, contralateral cerebrum
across all the groups.
P < 0.001,
P < 0.0001, relative to the control group. Data represent
the mean S.E.M. (n = 5 per group). ' 411 g/day is the dose for a 250g rat......................... 63
Table 3.3 Concentration of DXM in plasma samples. Data represent the mean S.E.M. (n =5
per group). ' 411 g/day is the dose for a 250g rat...................................................................
64
Table 4.1 DXM dose comparison. 'Dose converted to DXM equivalent...............
92
Table 5.1 DXM and cediranib dose comparison. 'Dose converted to dexamethasone equivalent.
'Dose converted to cediranib equivalent...................................................................................
113
15
16
CHAPTER 1
INTRODUCTION
1.1. SIGNIFICANCE
Brain tumor associated edema, a common feature of malignant brain neoplasms such as
high-grade gliomas and metastases, is one of the major causes of morbidity from brain tumor [11.
It also correlates with poorer survival in patients with high-grade gliomas [2]. Malignant gliomas
account for approximately 70% of the 22,500 new cases of malignant primary brain tumors
diagnosed in the United States each year [3]. The median survival for patients with
glioblastomas, the most common malignant glioma, is only 12 to 15 months despite optimal
treatment [3]. Brain metastases develop in approximately 20% to 40% of cancer patients, and the
subsequent median survival of these patients is generally less than 6 months [4]. Prognosis is
grim in all these cases.
Corticosteroids are highly effective in treating brain tumor associated edema but can have
adverse effects such as steroid myopathy, hyperglycemia, and neuropsychiatric syndromes that
can impair quality of life [5]. Administration of corticosteroids has remained the mainstay of
treatment due to no better alternative.
1.2. BRAIN TUMOR ASSOCIATED EDEMA
1.2.1. PATHOGENESIS
Brain tumor associated edema is vasogenic in nature, characterized by the breakdown of
the blood-brain barrier (BBB). The BBB is composed of an intricate network of endothelial cells,
17
pericytes and astrocyte foot processes. It results from intact tight junctions, minimal endothelial
pinocytosis, and lack of fenestration, restricting transport between the circulating blood and the
central nervous system (CNS) [1].
The loss of integrity of the BBB in high-grade gliomas and brain metastases is attributed
to increased endothelial pinocytosis, endothelial fenestrations and in particular, the opening of
tight junctions [1]. Defective tight junctions are caused partly by the impaired ability of normal
astrocytes to produce factors required to maintain a normal BBB [1], and compounded by the
production of factors such as glutamate [6], leukotriene [7], and vascular endothelial growth
factor (VEGF)-A [8] by the tumor cells. VEGF-A, commonly known as VEGF, has been shown
to be upregulated in both malignant gliomas [9] and metastatic tumors [10], and its level in
human gliomas correlates directly with the degree of malignancy [11]. There is a high degree of
correlation between VEGF expression in high-grade gliomas and the occurrence of edema [12].
VEGF is secreted by both tumor and host stromal cells, and binds to its receptors, VEGF
receptor (VEGFR)-1 and VEGFR-2, which are found mainly on the surface of endothelial cells.
VEGFR-2 is the primary mediator of the downstream effects of VEGF in tumor angiogenesis.
Binding of VEGF to its receptor stimulates the formation of gaps, fragmentations, and
fenestrations in the brain endothelium, and leads to the degeneration of the basement membrane
[13]. These structural changes lead to leakage of plasma fluid and proteins into the extracellular
space of brain parenchyma, resulting in vasogenic edema (Figure 1.1). Edema is largely localized
in the extracellular space of the white matter rather than the gray matter [14]. Brain metastases
and primary brain tumor such as glioblastomas multiforme (GBM) produce up to 90 ml of edema
fluid per day [1], overwhelming the sink effect provided by the ventricles and subarachnoid
cerebrospinal fluid under normal physiological conditions [15].
18
a~m
extracellular
space
:
.
*
tight
Junction
...
*:
. *
S
endothelum
*.G..
capillary
VEGF
* @e
defective
tight
. .
.
fenestration
Figure 1.1 Illustration of (a) a normal blood-brain barrier (BBB) and (b) the formation of brain tumor
associated edema through a disrupted BBB in the setting of a brain tumor. (Adapted from Kaal et al. [16])
Brain tumor associated edema exerts additional mass effect, often exceeding that induced
by the tumor itself, and increases intracranial pressure within the rigid skull [15]. It also disrupts
tissue homeostasis and reduces local blood flow, leading to debilitating neurological symptoms
[15]. Severe brain edema can result in widespread ischemia, herniation, and ultimately, death.
1.2.2. STANDARD OF CARE - CORTICOSTEROIDS
Systemic corticosteroids are indicated in all patients with symptomatic brain tumor
associated edema [14]. The efficacy of corticosteroids is well established. Significant decrease in
tumor vascular permeability, reduction of intracranial pressure and improvement in neurologic
symptoms usually begins within 24 hours of treatment [17-19].
1.2.2.1. Mechanism of action
The mechanism of action of corticosteroids in controlling brain tumor associated edema
is not fully understood, although corticosteroids have been used clinically since the 1950s based
on empirical observations that they could significantly decrease mortality and morbidity in
patients with brain tumors [20]. It has been suggested that corticosteroids reduce edema by
19
reducing the permeability of tumor capillaries [1, 21]. Corticosteroids diffuse through the plasma
membrane and bind to the cytoplasmic receptor, allowing the steroid-receptor complex to move
to the nucleus where it affects gene transcription and interacts with other transcription factors,
mediating regulation of other signaling pathways [22]. Studies have shown that corticosteroids
reduce edema by downregulating VEGF while upregulating angiopoietin-1, a strong BBB
stabilizing factor [21, 23]. Corticosteroids can also cause dephosphorylation of the tight junction
component proteins, occludin and Z01, resulting in decrease in endothelial permeability [1].
1.2.2.2. Dose and schedule
Dexamethasone is the most widely used corticosteroid due to its long biological half-life
(36 to 54 hours) and low mineralocorticoid (sodium retaining) activity [16]. A comparison of
dexamethasone with other synthetic corticosteroids is shown in Table 1.1.
Table 1.1 Co
arison of corticosteroids (Data from Kaal et al. [16])
Dexamethasone
0.75
25-30
< 0.2
36-54
A typical dexamethasone dosing regimen for patients with severe symptoms secondary to
a brain tumor consists of a large 10 to 20 mg intravenous (IV) dose, followed by 4 mg four times
per day or 8 mg twice per day [24-27]. Maintenance dosing comprising of oral or IV
dexamethasone can range from 4 to 24 mg per day, in divided doses [28-30]. Patients with mild
symptoms will benefit from a starting dose of 4 to 8 mg/day, considering the side effects of
20
dexamethasone, while higher doses such as 16 mg/day are indicated in patients with more severe
symptoms [24, 28, 31].
Tapering of dexamethasone after successful treatment of edema is necessary because of
dexamethasone-induced suppression of the hypothalamic-pituitary-adrenocortical (HPA) axis
(Figure 1.2) [16]. The hypothalamus secretes corticotrophin-releasing hormone (CRH), which
acts on the anterior pituitary. CRH stimulates the secretion of adrenocorticotropic hormone
(ACTH), which in turn acts on the adrenal cortex. The adrenal cortex releases approximately 20
mg of cortisol per day in response to ACTH stimulation. Circulating cortisol will in turn inhibit
the hypothalamus and the pituitary gland through a negative feedback cycle. Typical doses of
corticosteroids used in treatment are supraphysiologic in nature, and will result in the inhibition
of this cycle and suppression of its activity. Long term suppression of the axis can cause it to
atrophy. Abrupt cessation or rapid tapering can lead to steroid withdrawal syndrome
characterized by adrenal insufficiency, arthralgia, myalgia, and general symptoms such as
headache and lethargy [16]. A regimen of twice daily dosing of dexamethasone, starting with 8
mg twice per day, tapering the dosage by 50% every 4 days and continuing with 2 mg twice per
day until the last day of radiotherapy has provided good clinical results [14, 32]. Tapering should
be slower, by 25% every 8 days in patients with extensive edema [16].
Chronic treatment with a low dose of 1 to 2 mg per day may be needed for some patients
to have an acceptable quality of life [16]. One prospective cohort study found that 71% of
malignant glioma patients were still on steroids at 3 months post radiotherapy [33]. A
retrospective study of 138 patients showed that primary brain tumor patients required steroids for
an average of 23 weeks post radiotherapy whereas secondary brain tumor patients required
steroids for an average of 7 weeks [25]. There is much variability in corticosteroid prescribing
21
practices. The dose and schedule of dexamethasone is often adjusted to the patient's tolerance
throughout the course of treatment in clinical practice.
CRH:
_Synthetic corticosteroids
(e.g. dexamethasone)
AUHU
ACTH
Cortisol
Figure 1.2 Schematic representation of corticosteroid-induced suppression of the HPA axis.
1.2.2.3. Complications
Corticosteroids are associated with numerous adverse effects, which can contribute to
overall treatment morbidity (Table 1.2). Gastrointestinal complications, steroid myopathy and
opportunistic infections such as Pneumocysisjeroveciipneumonia (PJP) are of particular
concern to brain tumor patients [34]. Patients on corticosteroid therapy are routinely treated with
histamine H2 antagonists to lower the risk of gastrointestinal complications although no
significant correlation between steroid usage and gastrointestinal complications was found [30].
Steroid myopathy is a common side effect with an estimated incidence between 2 and 20 percent,
and causes significant morbidity in brain tumor patients [34]. PJP is a life-threatening
opportunistic infection that occurs in immunocompromised hosts, such as patients receiving
corticosteroids. The severity of complications usually correlates with the treatment duration
(more than 3 weeks) and total dose [35]. Most complications resolve after cessation of therapy,
22
except for cataract and osteoporosis [5]. The frequency of these complications can be reduced by
using the lowest possible dose [5].
Table 1.2 Systemic side effects of corticosteroids (Adapted from Wen et al. and Kesari et aL [5, 36])
Common
Uncommon
General
Cushing syndrome
Weight gain
Night sweats
Infection
Candidiasis
Some opportunistic infections (e.g. PJP)
Neurologic
Behavioral changes
Insomnia
Myopathy
Tremor
Reduced taste and smell
Cerebral atrophy
Psychosis
Seizures
Dependence
Dementia
Hiccups
Hallucinations
Paraparesis due to epidural lipomatosis
Dermatologic
Fragile skin
Purpura
Striae
Acne
Hirsutism
Poor wound healing
Kaposi's sarcoma
Gastrointestinal Increased appetite
Bloating
Gastrointestinal bleeding
Pancreatitis
Liver hypertrophy
Perforation
Rheumatologic
Osteoporosis
Fracture
Growth retardation in children
Avascular necrosis
Tendinous rupture
Ophthalmologic
Visual blurring
Cataract
Glaucoma
Uveitis
Endocrine/
metabolic
Hyperglycemia
Hypokalemia
Hypernatremia
Hyperlipidemia
Fluid retention
Urogenital
Polyuria
Hematologic
Neutrophilia
Lymphopenia
23
1.2.3. NOVEL DRUGS
The multitude of complications associated with corticosteroid therapy has sparked the
interest in alternative treatments for brain tumor associated edema. Novel anti-edema agents that
have been developed include corticotrophin-releasing factor, cyclooxygenase-2 (COX-2)
inhibitors, and anti-VEGF agents.
1.2.3.1. Corticotrophin-releasing factor
Human corticotrophin-releasing factor (hCRF) is an endogenous neuropeptide that has
been reported to reduce brain edema by acting directly on tumor vasculature, independent of the
release of adrenal glucocorticoids [37]. It was effective in an animal model, but had limited
efficacy in a phase I clinical trial [37, 38]. Corticorelin acetate, a synthetic analog of hCRF, was
shown to significantly reduce steroid-related adverse events such as myopathy and Cushing
syndrome, although the primary endpoint of at least 50% reduction in dexamethasone with stable
or improved neurologic condition and Karnofsky performance was not achieved in a phase II
trial [39]. The full clinical potential of this drug as a corticosteroid-sparing agent remains to be
evaluated.
1.2.3.2. COX-2 inhibitors
Preclinical studies with COX-2 inhibitors suggest that they may be effective in treating
brain edema. Rofecoxib, a selective COX-2 inhibitor, had a similar effect as dexamethasone in
decreasing the diffusion of contrast agent into the brain parenchyma of glioma-bearing rats [40].
Another selective COX-2 inhibitor, SC-236, appeared to be as effective as dexamethasone in
prolonging survival in a rat brain tumor model [41]. The effects of COX-2 inhibitors on brain
tumor associated edema have yet to be examined in human clinical trials.
24
1.2.3.3. Anti-VEGF agents
Agents that target the VEGF signaling pathway are attractive therapeutic options for the
treatment of brain tumor associated edema since VEGF plays an important role in the
pathogenesis of brain tumor associated edema. They can normalize abnormal tumor vasculature,
thus restoring the integrity of the BBB and reducing edema formation [42]. Two major classes of
anti-VEGF agents are anti-VEGF ligands such as bevacizumab and aflibercept, and inhibitors of
VEGFR such as cediranib (Table 1.3).
Table 1.3 Anti-VEGF a ents (Data from Norden et al. [43])
Bevacizumab
VEGF-A
None
Pazopanib
VEGFR-2
All VEGFR subtypes, PDGFRa and , c-Kit
Sunitinib
VEGFR-2
PDGFRB, FLT3, c-Kit
Vatalanib
VEGFR-2
All VEGFR subtypes, PDGFRP, c-Kit
VEGFR inhibitors
Abbreviations: EGFR, epidermal growth factor receptor, FLT3, FL cytokine receptor, PDGFR, platelet-derived
growth factor receptor, PIGF, placental growth factor, VEGF, vascular endothelial growth factor, VEGFR, vascular
endothelial growth factor receptor.
Bevacizumab is a recombinant humanized monoclonal antibody against VEGF that
works by binding VEGF with high affinity and inhibiting coupling of VEGF to its receptors, thus
blocking the downstream effects of VEGF. It was granted approval by the Food and Drug
25
Administration (FDA) for treatment of recurrent GBM in May 2009 based on two phase II
clinical trials [44]. The majority of the patients using systemic corticosteroids at baseline in the
first trial experienced corticosteroid dose reduction while receiving bevacizumab, suggesting that
bevacizumab may have corticosteroid-sparing effect [45, 461. Edema was decreased in 50% of
the patients, and 58% of the patients receiving corticosteroids at the start of therapy were able to
reduce their corticosteroid doses by an average of 59% in the second trial [47]. The reduction in
edema was apparent as early as 18 days after initiating treatment with bevacizumab in an early
study [48].
Another anti-VEGF ligand is aflibercept, also known as VEGF Trap. It is a recombinant
fusion protein of the extracellular domains of VEGF fused to the constant region of
immunoglobulin G1. It inhibits tumor angiogenesis by binding with high affinity to VEGF and
placental growth factor (PIGF). A phase I of aflibercept showed minimal activity in patients
with recurrent malignant gliomas [49].
Inhibiting VEGFR by small molecule tyrosine kinase inhibitors (TKIs) is another
approach to anti-VEGF therapy beside neutralization of the VEGF ligand by anti-VEGF ligands.
Cediranib (AZD2171), an oral pan-VEGFR TKI, was demonstrated to induce vascular
normalization within 24 hours and alleviates vasogenic brain edema in glioblastomas patients in
a phase II clinical trial [50]. Corticosteroid usage was either reduced or discontinued in these
patients as a result [50, 51]. Vatalanib, another pan-VEGFR inhibitor, decreased vascular
permeability and cerebral blood volume in a phase I/II trial of patients with recurrent GBM [52].
Pazopanib, a TKI with multiple targets, reduced vasogenic edema in 32% of the patients in a
phase II trial [53]. XL-184 also demonstrated its anti-edema effect and its corticosteroid-sparing
26
effect with at least 50% reduction in corticosteroid dose in 5 out of 14 patients, in a separate
phase II study [54].
Treatment with anti-VEGF agents is associated with adverse effects, as with
corticosteroids. This is expected since VEGF is essential for normal physiological functions.
Common toxicities include hypertension, proteinuria, poor wound healing, and increased risk for
venous and arterial thromboembolic events [55].
1.3. DRUG DELIVERY IN THE BRAIN
Progress has been made in drug discovery and development for treatment of malignant
brain tumors, but clinical trials of new therapeutic agents have been disappointing, and prognosis
for high grade gliomas has remained dismal. A significant part of the problem is the poor
penetration of the systemically delivered drugs, which includes approximately 98% of all small
molecule drugs and nearly 100% of large molecule therapeutics, across the BBB to achieve
efficacious drug levels in the desired target [56]. Advancement in drug delivery may maximize
efficacy of currently available drugs and improve treatment outcome. Drug delivery methods
can be broadly classified into two categories - systemic delivery and local delivery. This section
describes both traditional systemic delivery and local delivery strategies, with an emphasis on the
latter.
1.3.1. TRADITIONAL SYSTEMIC DELIVERY
Drugs approved for treatment of brain tumor are usually administered systemically either
by mouth or by intravenous injection. Lomustine and temozolomide are taken orally while
bevacizumab is given intravenously. Systemic drug delivery to brain tumor is limited by multiple
27
factors, including the physical barrier posed by the BBB, active efflux transport mechanism,
plasma protein binding and tumor interstitial pressure.
The BBB selectively restricts transport between the peripheral circulation and the
cerebral parenchyma. The two major transport mechanisms across the BBB are passive diffusion
and endogenous carrier- or receptor-mediated transport. Only lipid soluble small molecules with
a molecular weight of less than 400 Da can readily penetrate the brain down their concentration
gradient via passive diffusion [57]. Almost all other essential nutrients and metabolic substrates,
such as glucose, amino acids, nucleosides, hormones, and growth factors, do not satisfy the
criteria for passive diffusion, and have to rely on the carrier- or receptor-mediated transport
systems to get into the brain [58]. After entering the BBB endothelium, molecules may be
actively pumped back into the bloodstream by active efflux transporters, such as P-glycoprotein
that is expressed in the luminal membrane of the endothelial cells [59]. These unique properties
of the BBB serve to protect the brain from toxic substances, and to provide rigorous control of
the brain microenvironment that is necessary for precise neural signaling, but also pose an
insurmountable obstacle for most therapeutic drugs to reach efficacious levels in the brain
parenchyma. Plasma protein binding and tumor interstitial pressure are other contributing factors
to poor drug delivery to brain tumors. Many chemotherapeutics such as chlorambucil,
vincristine, and pacitaxel bind more than 90% to plasma proteins, thus reducing the free fraction
of drug in plasma that is available to cross the BBB [60]. Tumor interstitial fluid pressures can be
more than 50 mmHg in peritumoral areas compared with 2 mmHg in a normal brain, generating
a high pressure gradient which reduces drug diffusion into tumor and enhances drug diffusional
loss into surrounding brain tissue [611.
28
The integrity of the BBB is often compromised in high-grade gliomas and brain
metastases, as suggested by gadolinium enhancement of tumor during magnetic resonance
imaging (MRI). However, vascular permeability in tumors has been shown to be heterogeneous.
Vascular permeability is the greatest at or near the tumor core where the BBB is disrupted and
sharply decreases at the tumor edge adjacent to normal brain parenchyma where the BBB
remains relatively intact [62]. Clinical studies have confirmed that most antitumor agents
accumulate to a greater extent and persist longer in the tumor core while drug levels are
markedly lower or undetectable in the peritumoral region [63, 641.
Only a small fraction of a drug delivered systemically reaches the brain tumor while the
bulk of it is lost to the rest of the body, which acts as a large sink [65]. Obtaining therapeutic
drug levels will require administering a dosage much higher than what systemic organs can
tolerate, resulting in systemic toxicity [66]. Maximum tolerated dose (MTD), rather than
therapeutic dose, is often given due to the dose-limiting side effects of chemotherapeutic agents
comprising of anemia, thrombocytopenia, leukopenia, stomatitis, nausea, and vomiting [67].
1.3.2. LOCAL DELIVERY
Local drug delivery in the brain can theoretically avoid the side effects of systemic
exposure, and allow increased local drug concentration and prolonged drug exposure in the
tumor by completely circumventing the BBB. High drug concentration in the tumor over a
prolonged period is generally desirable for anti-tumor activity, thus local drug delivery may be
more effective than traditional systemic approach. Some chemotherapeutic agents also only work
in specific phases of the cell cycle. Maintaining substantial concentration of these agents for a
prolonged period increases the chance of drug exposure during the sensitive phases of the cell
29
cycle [67]. Furthermore, local delivery may allow patients to maintain their quality of life by
substantially reducing or eliminating side effects that are usually associated with systemic
treatment.
Local drug delivery can also expand the drug arsenal against brain tumor. Agents that are
intrinsically excluded from the brain by the BBB and/ or unsuitable for systemic delivery (for
instance, due to strong plasma protein binding) can realize their anti-tumor function through
local drug delivery approaches. Although a wide range of drugs appears to be amenable to local
drug delivery, a major limiting factor on drug choice and maximum drug dose is neurotoxicity.
The main disadvantages of local drug delivery are the need for invasive brain procedures
and its inability to distribute drug to distant infiltrating tumors. However, local drug delivery
approaches can be performed at the time of surgical tumor resection, which is the standard of
care in cases of operable brain tumors, so no additional surgery is required. Local drug delivery
remains a relevant method of drug delivery despite its limited drug distribution profile, since
most brain tumors appear to remain localized to a single region of the brain and 90% of
recurrences occur within 2 cm of the margin of the original tumor [68].
Table 1.4 Comparison of local delivery strategies
Drug
Adesefct
Refillability
compliance
Patient
Ease of
ditribution
implement
stability
Adverse effects
Intracavitary
infusion
Limited
Yes
Additional
burden
Complex
Unstable
liquid
drug
Risk of infection/
edema
CED
Enhanced
but
unpredictable
Yes
Additional
burden
Complex
Unstable
liquid
drug
Risk of infection;
infusion-related
edema
Implantable
polymers
Limited
No
Easy
Simple
Stable
solid
drug
Risk of infection/
edema
Statgy
Strategy
Disrut
30
Some local drug delivery strategies used for treatment of brain tumor include
intracavitary infusion, convection-enhanced delivery (CED), and implantable controlled-release
polymer systems. Table 1.4 shows a comparison of these local delivery strategies.
1.3.2.1. Intracavitary infusion
The simplest approach to circumvent the BBB involves manual drug injection into the
resected tumor cavity through a catheter.
Figure 1.3 Intracavitary delivery via an Ommaya reservoir into the tumor resection cavity. (Reproduced
from Reulen [69])
The Ommaya reservoir is a device consisting of a fluid reservoir that is implanted under
the scalp and connected to an outlet catheter, which is positioned in the tumor bed for
intracavitary delivery (Figure 1.3). It offers the advantage of subsequent percutaneous access to
the drug reservoir, enabling the delivery of intermittent bolus injections of drug. Drugs are
injected percutaneously into the reservoir and delivered to the tumor by manual compression of
the reservoir through the scalp [67]. Drug distribution within the tumor cavity relies on passive
diffusion. The efficacy of a drug delivered by a diffusion based technique is dependent on its
31
diffusivity in the extracellular space of the brain, its uptake into tumor, and its rate of clearance
[67]. Poor drug diffusion through tumor and brain parenchyma relative to drug clearance will
result in treatment of only a small volume of tissue surrounding the drug source, leading to
limited efficacy.
The Ommaya reservoir has been extensively used in clinical trials for intracavitary
delivery of anti-tumor agents such as doxorubicin [70], mitoxantrone [71, 72], methotrexate [73],
bleomycin [74, 75], interleukin-2 [76, 77], lymphokine-activated cells [77] and monoclonal
antibodies labeled with cytotoxic radionuclides [72, 78]. Reservoir-related infection was rare.
Local delivery using the Ommaya reservoir was generally well tolerated with minimal systemic
and neurological toxicity, and no clinical deterioration; but its efficacy was controversial in the
clinical studies, therefore hindering its use in clinical practice.
The Ommaya reservoir can be modified to achieve continuous drug release by covering
the catheter with a semipermeable polyvinyl alcohol membrane [74]. Subcutaneously
implantable pumps, such as the Infusaid pump and the Medtronic SynchroMed system, which
are percutaneously refillable like the Ommaya reservoir, can also be used to deliver drug
continuously over an extended period of time and at variable rates [67].
1.3.2.2. Convection-enhanced delivery
CED uses the pressure gradient generated from an infusion pump to drive bulk flow of
infusate in the brain interstitium. A drug-containing fluid is delivered continuously to the
targeted region in the brain via a catheter that is connected to an infusion pump. Using
convection to supplement simple diffusion, CED can significantly enhance the spatial
distribution of drugs beyond what is possible with diffusion alone, while producing drug
32
concentration in the brain that is orders of magnitude higher than that can be achieved by
systemic administration [79, 80].
Convection relies mainly on the pressure gradient generated and is independent of the
molecular weight of the delivered agent. Delivery of both small and large molecules, such as
paclitaxel [81, 82] and targeted toxins [83-89], is therefore possible with convection. The drug
distribution profile is constant during infusion but drops abruptly when infusion ends [90]. Drug
diffusion, loss within capillaries, and drug metabolism determine the drug distribution over
longer time periods after infusion [80].
Several factors, such as brain regions, catheter placement, catheter size, catheter design,
rate of infusion and infusion volume, can affect the distribution of drug delivered by CED [90].
Regional differences in drug distribution exist between gray and white matter. Drug distribution
tends to be spherical in the isotopic gray matter, while distribution in the anisotropic white matter
preferentially occurs along the parallel fiber tracts [61]. Infusing into the white matter produces
edema that appears similar to brain tumor associated edema [801. Infusion-related edema, which
is often present at a greater extent than brain tumor associated edema, can significantly affects
the flow of the infused agent, and exacerbate morbidity [80]. Another major concern during
infusion is backflow as it can lead to diminished drug delivery to the target tissues and spreading
of the infused agent into unintended regions of the brain. Backflow along the insertion track into
the subarachnoid space, in the case of cortical infusions, can cause subsequent widespread
distribution of the agent via the circulating CSF, increasing the possibility of widespread
neurotoxicity [91]. Drug leakage into the CSF, due to bad catheter placement, was reported to be
responsible for chemical meningitis [81].
33
Additional challenges arise in the application of CED in the setting of brain tumors. The
complex anatomy and heterogeneous physical properties of brain tumor-infiltrated tissue,
coupled with the effect of tumor resection surgery and the presence of edema causes uneven and
unpredictable drug distribution [91]. Dramatically higher interstitial fluid pressure in brain tumor
compared to that in normal brain tissues creates a pressure gradient, which drives the infusate out
of the tumor toward the normal surrounding tissues [61]. Rapid efflux of drug from brain tumor
shortens drug residence time within tumor, leading to decreased drug penetration and drug
activity in tumors [61]. Inadequate drug delivery due to CED has resulted in limited efficacy in
clinical trials. Adverse effects, including surgical complications and significant neurological
deterioration, further hinder the progress of CED in the clinical setting.
Osmotic micro-pumps have been utilized extensively for CED of therapeutic agents, such
as angiogenesis inhibitors [92, 93], heat shock protein 90 (Hsp 90) inhibitor [94] and
dexamethasone [95], in preclinical studies of treatment of brain tumor.
1.3.2.3. Implantable polymer systems
Surgically implantable controlled-release polymer systems can be placed in the tumor
resection cavity for continuous local drug delivery. These systems can be matrix-based or
reservoir based. Drugs are dissolved or dispersed throughout a solid polymer phase in a matrix
system while drug is enclosed within a polymer in a reservoir system [96].
An example of a matrix-based polymer system is the Gliadel® wafer, the first and only
FDA approved local drug delivery therapy for malignant glioma [97]. It is a biodegradable
polyanhydride poly[bis(p-carboxyphenoxy-propane)-sebacic acid] (PCPP-SA) polymer
impregnated with carmustine (1,3-bis(2-chloroethyl)-1-nitrosourea or BCNU). Treatment with
34
Gliadel®wafers improved survival of newly diagnosed and recurrent malignant glioma patients
by 2 months on average, which is considered significant survival benefit for this disease [98, 991.
Drug is released by means of both drug diffusion and polymer degradation at the surface [100].
Drug diffusion is dependent on the permeability of the matrix and the diffusional properties of
the drug. Polymer degradation can be controlled to occur over days or years by altering the ratio
of the monomers in the polymer matrix. Constant rate of drug release is achieved since
breakdown of the polymer is limited to the surface in a process known as surface erosion [100].
No foreign body is left behind due to the biodegradable nature of the polymer, and the
breakdown products are not mutagenic, cytotoxic or teratogenic [1011. Any drug can
theoretically be incorporated into the polymer matrix as long as it does not react chemically with
the matrix backbone [67]. The PCPP-SA polymer matrix has been used to deliver other
chemotherapeutics, such as doxorubicin [102], taxol [103], and temozolomide [104], in
prechnical models.
Non-biodegradable, inert, ethylene vinyl acetate copolymer (EVAc) is another polymer
matrix that has been used for diffusion-controlled drug delivery. It has been used experimentally
to deliver carmustine [105] and dexamethasone [106] locally within brain tumor, resulting in
prolonged survival and reduction in brain tumor associated edema respectively.
Reservoir-based polymer systems offer the advantages of high drug payload and longterm, nearly zero-order drug release kinetics over matrix-based systems [96]. Local delivery of
temozolomide via a reservoir-based, polymeric microcapsule device has led to improvement in
survival in a rodent model [107]. Drug is released through the orifices in the device and any drug
can theoretically be used with this device without the need for further drug modification.
Silicone rubber capsules have also been used to deliver anti-tumor drugs for treatment of brain
35
tumors [108, 109]. The rate of drug release is governed by drug diffusivity within silicone
rubber, thus only drugs with certain diffusivity can be delivered satisfactorily with this approach.
The effect of drug delivered by polymer systems is restricted to residual tumor cells at the
resection margin due to limited drug distribution, which occurs mainly by diffusion. Drug
concentrations generally fall exponentially away from the source and drug penetration into brain
tissues is limited to a few millimeters [110]. The brain concentration of carmustine was found to
decrease 90% at a distance of around 1 mm removed from the polymer [111]. An important
consideration when translating polymer systems from preclinical to clinical studies is that
diffusion distance in the human brain is more prohibitive than that in the rat brain since the
weight of the human brain is a 1000-fold greater than that of the rat brain [112]. Other factors
like drug elimination rate, brain peculiar extracellular matrix and tumor local environment can
further hinder diffusion in brain parenchyma [113]. One approach to improve spatial drug
distribution is to implant multiple polymers at the tumor site. Up to 8 Gliadel® wafers may be
implanted in the tumor cavity depending on the size of the cavity [114].
Unreleased drug in solid form is usually protected from degradation in polymer systems
while drug is usually stored in a liquid reservoir at body temperature in catheter systems and
many drugs are not stable under these conditions [96]. Polymers are also not subject to clogging
and blockage by tissue debris, which are major issues with catheter technology [67]. Another
advantage of polymer systems is that they do not pose additional burden to the patient as no
further intervention is usually required after implantation. A disadvantage of polymer systems
delivery over catheter systems is that it is not refillable. Surgery is required if any changes in
dose is necessary in polymer-based delivery.
36
1.4. THESIS OBJECTIVES
The overall goal of this work is to develop a local drug delivery device that delivers antiedema agents in the brain for treatment of brain tumor associated edema with minimal systemic
complications. Systemic administration of corticosteroids, the current standard of care, is highly
effective but can cause serious adverse effects. Treatment outcome can generally be improved
with novel therapeutic agents and better drug delivery to the brain. VEGF inhibitors, such as
bevacizumab and cediranib, are novel anti-edema agents that have demonstrated their potential
as alternatives to corticosteroids in clinical trials [47, 50, 511. They can have systemic side
effects like corticosteroids, though it is not clear how their side effects compared to those of
corticosteroids. Local drug delivery in the brain is a promising approach that can avoid the side
effects of systemic exposure and increase local drug concentration and exposure in the tumor by
completely circumventing the BBB. Local administration of dexamethasone via an implantable
polymer matrix system [1061 and an osmotic pump [951 have shown efficacy in reducing edema.
However, these local delivery technologies have limitations and systemic effects have not been
thoroughly examined in these experimental studies. Systemic toxicity has been reported in both
preclinical and clinical studies when very high doses of carmustine are locally delivered in the
brain [115, 116]. Based on these findings, we hypothesize that optimizing local drug delivery can
minimize systemic side effects while being as effective as systemic dosing in treating brain
tumor associated edema. A reservoir-based, diffusion-controlled drug delivery device that can be
easily used with any drug without further chemical modification and be easily altered to give
varying drug release rates has been developed to test this hypothesis. Our specific objectives are
to:
37
1) Identify valid animal model(s) for study of brain tumor associated edema (Chapter 2)
2) Develop a dexamethasone-loaded intracranial drug delivery device whose drug release
rate can be easily optimized, characterize its release kinetics, and evaluate its efficacy and
toxicity in valid animal model(s) (Chapters 3 and 4)
3) Develop an intracranial drug delivery device that delivers a VEGF inhibitor, characterize
its release kinetics, evaluate its efficacy and toxicity in valid animal model(s) (Chapter 5)
38
CHAPTER 2
CHARACTERIZATION OF
BRAIN EDEMA ANIMAL MODELS
2.1.
INTRODUCTION
Accurate disease modeling is important in evaluating therapy efficacy. While no animal
model can perfectly emulate all aspects of the human glioma, valid models should fulfill the
following criteria: 1) clinically relevant glioma-like characteristics; 2) sufficient host survival
time following tumor formation to permit therapy evaluation; and 3) similar response to
conventional therapy as in human glioma [1171.
Brain tumor animal models that have been used for the study of brain tumor associated
edema are either xenografts or allografts. These tumors are easy to develop, highly reproducible,
highly penetrant and develop rapidly but they lack their native stroma and may be genetically
different from the original tumor due to selective pressure during cell culture [118, 119]. Human
glioma xenografts implanted in immune-deficient mice may retain many of the genetic mutations
but there is a lack of immunological interaction between tumor and host [118]. Allografts, on the
other hand, have an intact immune system but they do not fully recapitulate the genetics and
pathobiology of human glioma [118]. The 9L gliosarcoma, an allogeneic tumor model in
syngeneic Fischer rats, has been the most widely used experimental model for brain tumor
associated edema [21, 41, 106, 120, 121]. Other models that have been used include C6 glioma
in rats [122], RG2 rat glioma in cats [123], VX2 carcinoma in rabbits [95], Walker 256
carcinosarcoma in rats [124, 125], CNS-1 rat glioma and U87 human glioma in mice [126]. The
39
use of these models remains controversial as they have been criticized for not recapitulating the
features of human brain tumor.
Genetically engineered mouse (GEM) models with spontaneous tumor formation arising
from mutations in the genes characteristic of human malignancy are attractive as they can
recapitulate the complex tumor genetics, angiogenesis, tumor-host interactions and
pathophysiology of human GBM [1191. However, most GEM models are not practical for
therapy studies due to their low tumor penetrance, poor reproducibility, prolonged tumor
formation latency, and inherent difficulty in diagnosing the tumor in a timely fashion [127]. No
GEM model has been used for investigating brain tumor associated edema to date.
Zhu et al. has developed a GEM model which has a fully penetrant, rapid-onset highgrade malignant glioma phenotype with prominent pathological and molecular resemblance to
human GBM, when the mutant (vIII) epidermal growth factor receptor (EGFR) is activated along
with the loss of function of the tumor suppressor genes pl6Ink4a/pl9ARF and PTEN [94, 128].
More than 25% of all GBM patients are found with the combination of activated EGFR, loss of
p16Ink4a/p19ARF and PTEN [94]. An important characteristic of this model is the ability to
have spatiotemporal control over EGFRvrn expression using the Cre/loxP system. An adenovirus
transducing Cre recombinase (Ad-Cre), which is injected in the basal ganglia of adult mice,
mediates the removal of the floxed stop cassette and leads to localized somatic expression of
EGFRVII. Another appealing feature of this model is the presence of a Cre/loxP conditional
luciferase reporter transgene that allows longitudinal, noninvasive tumor monitoring via
bioluminescence imaging (BLI). Only cells exposed to Ad-Cre give rise to EGFR-dependent
gliomas and express the bioluminescent luciferase reporter [94]. BLI output has been
demonstrated to correlate linearly with glioma volume in this model [129].
40
The brain tumor, its associated edema and survival of this GEM model were
characterized to determine the validity of this mouse model for the study of brain tumor
associated edema. Tumor growth was longitudinally monitored using BLI and edema
measurement was attempted using MRI. Edema was also quantified at the end of the study using
wet/dry weight ratio. Survival curves were generated to evaluate if host survival time following
tumor formation was sufficient for therapy evaluation. The well-established intracranial 9L
gliosarcoma model was also characterized and compared with this GEM model.
2.2.
MATERIALS AND METHODS
2.2.1. ANIMAL CARE
All the animal procedures were conducted in accordance with the Massachusetts Institute
of Technology (MIT) Committee for Animal Care guidelines.
2.2.2. TUMOR INDUCTION IN TRANSGENIC MICE
Tumor induction procedures were adapted from Zhu et al. [94]. Adult compound
Clatm2(CAG-EGFR*)CharAm2(CAG-EGFR*)Char Cdkn2a" Rnlhun]Rdp;,~RIP
Cola
Petm1Hwunm1Hwu;
gCG
Pten
lwIt1)w;
Tg(CAG-
luc)C6Charconditional transgenic mice, at least 3 months old, were anesthetized with an
intraperitoneal (IP) injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). The scalp was
shaved and the surgical site was prepped in a sterile fashion with betadine/ethanol scrub. The
animals were mounted on a Stoelting stereotaxic frame with non-puncturing ear bars (Stoelting,
Wood Dale, IL, USA). A midline incision was made and a 1 mm burr hole was drilled at the
stereotactically defined location of the striatum (0.5 mm rostral to the bregma, 1.5 mm lateral to
the midline, 2 mm depth to the pia surface). 1 x 108 adeno-CMV-Cre viral particles (Gene
41
Transfer Vector Core, University of Iowa, Iowa City, IA, USA) in approximately 1 pL was
injected at a rate of 0.1 pIWmin using a NanoFil 10 pL syringe (World Precision Instruments,
Sarasota, FL, USA) mounted onto a Quintessential stereotaxic injector (Stoelting). After careful
retraction of the syringe, the skin was sutured and the animal was allowed to recover.
2.2.3. 9L TUMOR IMPLANTATION IN RATS
Fischer 344 (F344) rats were obtained from Charles River Laboratories (Wilmington,
MA, USA). 9L gliosarcoma was provided by Dr. Betty Tyler from Johns Hopkins University and
propagated in the flanks of carrier F344 rats. Tumor was harvested from a carrier rat and cut into
1 mm3 pieces for intracranial tumor implantation. Animals were anesthetized with IP injection of
ketamine (75mg/kg) and xylazine (7mg/kg). The scalp was shaved and the surgical site was
prepped in a sterile fashion with betadine/ethanol scrub. A midline incision was made and the
pericranium was carefully removed. A 3 mm burr hole was drilled over the left hemisphere with
its center 3 mm lateral to the sagittal suture and 5 mm posterior to the coronal suture. The
underlying cortex was gently aspirated to expose the sulcus, where the tumor piece was placed.
The skin was sutured and the animal was allowed to recover.
2.2.4. BIOLUMINESCENCE IMAGING
Bioluminescence, a reporter of tumor size, was monitored noninvasively using the IVIS
Spectrum imaging system (PerkinElmer, Waltham, MA). The mice were anesthetized with 2%
isoflurane and their heads were shaved. D-Luciferin (PerkinElmer) was administered via IP
injection at a dose of 165 mg/kg approximately 10 min prior to imaging. Signal intensity was
quantified as photon count rate per unit body area per unit solid angle subtended by the detector
42
(p/ s/ cm2/ sr) for defined regions of interest. Image analysis was performed using Living Image
software.
2.2.5. MAGNETIC RESONANCE IMAGING
MRI was performed on 7T Varian MRI scanner (Agilent Technologies). All the animals
were anesthetized with 2% isoflurane and placed prone in the cradle. The field of view used for
mice and rats were 25 mm x 25mm and 32 mm x 32 mm respectively. T2 weighted images (TR
= 4,000ms, TE = 40ms, 1 mm slice thickness, 128x128 matrix size) were acquired to assess the
development of tumor and its associated edema. Edema and tumor volumes were measured using
OsiriX. T2 relaxation maps was also generated in Osirix from multi-echo spin-echo images (TR
= 2,000ms, TE = 10 ms, 16 echoes, 2 mm slice thickness, 128x128 matrix size) acquired. T2
maps were used to confirm the regions of tumor and its associated edema.
2.2.6. BRAIN WATER CONTENT
Animals were euthanized by carbon dioxide asphyxiation and their brains were collected.
The brains were divided into ipsilateral and contralateral cerebrums. The cerebrums were
weighed immediately to obtain the wet weight, and then dried at 100*C for up to a week, until a
final dry weight was achieved. The tissue water content was calculated as water content = (wet
weight - dry weight)/ wet weight, and expressed as a percentage.
2.2.7. HISTOLOGY
Brains were excised after the animals were euthanized, rinsed in cold phosphate buffered
saline (PBS), and fixed in 10% neutral buffered formalin solution (Sigma-Aldrich, St Louis, MO,
USA) at 4TC overnight. Fixed tissues were embedded in paraffin, sectioned at 5 pm, and stained
43
with hematoxylin and eosin (H&E) by the Hope Babette Tang Histology Facility at MIT
for
histopathological analysis.
2.2.8. STATISTICAL ANALYSIS
All the data were expressed as mean
standard error of the mean (S.E.M.). Paired
Student's t test was performed for brain water content comparison. P < 0.05 was considered
statistically significant.
2.3. RESULTS AND DISCUSSION
2.3.1. CHARACTERIZATION OF TRANSGENIC MOUSE MODEL
A__
B
TumorB induction
by injection of Ad-Cre
Collal locus
output > 106
Longitudinal
p/s/cmlsr
tumor monitoring by BLI
*
Ad-Cre+
Tumor & edema
assessment by
MRI
Figure 2.1 (A) Schematic illustration of the activation of the mutated (vIII) human EGFR in transgenic
mice. A strong ubiquitous promoter (CAGGS) is positioned upstream of a transcriptional stop cassette
(STOP) which is flanked by 2 loxP sites (blue triangles), followed by EGFRym cDNA and a polyA
signal
sequence (purple rectangle). Localized expression of EGFR'I is conditional to the removal of the stop
cassette, which is mediated by Ad-Cre. (B) Experimental design. Tumor growth was monitored
longitudinally by BLI post tumor induction by injection of Ad-Cre. When mice were determined to have
bioluminescent signal output of more than 106 p/ s/ cm2/ sr, MRI was performed to further assess tumor
and its associated edema.
44
Glioma was induced in the transgenic mouse model by injecting Ad-Cre in the basal
ganglia. Ad-Cre mediates the removal of the floxed stop cassette, which is positioned between a
strong ubiquitous promoter (CAGGS) and the mutated (vIII) human EGFR cDNA, leading to
localized somatic expression of EGFR'I (Figure 2.1A). Tumor growth was monitored
periodically by BLI. MRI was carried out in animals with bioluminescent signal output of more
than 106 p/ s/ cm2 / sr to further assess tumor and its associated edema (Figure 2.1B).
A
Days post Ad-cr. Injection
7
14
21
28
108
107
0 ssctcm)Jr)
cekr SC*
i.SE+09
E 1.OE+Os
I.OE+07
1.OE+06C
.OE+05--
-
Z
1.0E+04
0
20
40
60
so
Days post Ad-Cre Injection
10W
Figure 2.2 Longitudinal monitoring of tumor growth by BLL (A) Representative bioluminescent images
for a mouse imaged 7, 14, 2 1, and 28 days after Ad-Cre injection. (B) Individual measurements of
bioluminescent outputs from mice. The solid lines represent outputs from mice which had tumors, while
the dotted lines represent outputs from mice which did not form tumor.
45
Our results from BLI suggest that this transgenic mouse model of glioblastoma is highly
penetrant. 20 out of 23 mice (87%) injected with Ad-Cre showed continuous increase in
bioluminescent output with time, suggesting that tumors were growing in these mice (Figure
2.2). Histology was used to confirm the presence or absence of tumors. Bioluminescent output
generally increased exponentially and varied a lot among the mice. The time taken for the
exponential growth of tumor to occur varied from 20 to 60 days post tumor induction. Most
tumor-bearing mice suffered from poor body conditions and died or had to be euthanized after
the tumors grew exponentially. The survival of these mice ranged from approximately 3 to 10
weeks (Figure 2.3), which was similar to that published by Zhu et al. [128]. A major concern
with survival analysis was the development of lymphoma which led to decrease in survival.
Homozygous null mice for the Ink4a/ARF locus reportedly were reported to develop lymphoma
at a median age of 30 weeks [128]. Mice which were not manipulated at all had died from
lymphoma, which was characterized by splenomegaly (Figure 2.4). 19 out of 23 mice (83%)
used in this study were found to have enlarged spleens at the time of sacrifice. It might be
difficult to distinguish if the mice had succumbed to glioma or lymphoma.
The huge disparity in tumor growth and survival among tumor-bearing mice made
experiment planning, i.e. scheduling of MRI and treatment, very challenging. Not many mice
were evaluated by MRI as a result. T2 weighted images and multi-echo spin-echo images were
acquired (Figure 2.5). Hydrocephalus was common and tumor cores appeared to be necrotic in
some mice. Edematous peritumoral regions were not detected on the T2 weighted images. We
attempted to evaluate edema using T2 maps generated from and multi-echo spin-echo images,
since T2 signal correlates with water content. It was hard to demarcate tumor regions on T2
maps, although T2 signal of tumor was found to be slightly higher than that of normal cerebral
46
cortex. Low T2 signal from tumor necrotic core might affect the accuracy of using T2 signal to
evaluate edema. Some mice also did not survive the imaging sessions due to the long duration.
MRI might not be an appropriate tool to assess edema in this transgenic mouse model.
120
100
*80
60
-Tumor
-- No tumor
(
40
20
0
0
2
4
6
8
10
12
Weeks post Ad-Cre Injection
Figure 2.3 Kaplan-Meier survival curves of mice with tumor (n = 14) and mice with no tumor despite
Ad-Cre injection (n = 3).
Figure 2.4 Enlarged spleen, a sign of lymphoma, in a typical mouse.
47
A
B
T2-'
Figure 2.5 Tumor and edema assessment using MRL (A) Representative consecutive T2-weighted
coronal MR images of a mouse. Hydrocephalus and tumor with a necrotic core were noted. (B) A 72weighted image and its corresponding T2 map. Tumor region could be distinguished from normal brain
tissue on the T2-weighted image, but not with certainly on the T2 map. The green highlighted regions of
interest, normal cerebral cortex on the left and tumor on the right, were used to make T2 measurements on
the T2 map. T2 signal of tumor was higher than that of cerebral cortex.
0
C
0
U
82-,
*
--
81 -i
80
79
" Contralateral
cerebrum
" Ipsilateral
cerebrum
-4
YL
78
77
-I
76
-----
Control
T-
Tumor
Figure 2.6 Water content of the ipsilateral cerebrum and contralateral cerebrum. The water content of the
ipsilateral cerebrum was significantly higher than that of the contralateral cerebrum in tumor-bearing
mice (*P = 0.016), while the water content of the ipsilateral cerebrum and the contralateral cerebrum were
lower and not significantly different from each other in control animals. Data represent mean values
S.E.M. (n = 3 for control, n = 14 for tumor).
48
Brain water content was used to assess edema at the time of sacrifice (Figure 2.6). Water
content of both ipsilateral and contralateral cerebrums of tumor-bearing mice was higher than
that of control mice. The water content of the ipsilateral cerebrum was significantly higher than
that of the contralateral cerebrum in tumor-bearing mice (P = 0.016), while the water content of
the ipsilateral cerebrum and the contralateral cerebrum was not significantly different from each
other in control animals that were not manipulated. Presence of hydrocephalus and invasiveness
of tumor might explain the increase in water content of contralateral cerebrum.
2.3.2. CHARACTERIZATION OF 9L RAT MODEL
A
Day7
C
80.5
Z
E
80
Day9
120
-
B
Day5
7.60
-
79
-- Tumor
04
400
78.5
20S
0
78
Tumor
n=6
0
Notumor
n=3
10
20
Time (days)
Figure 2.7 Characteristics of 9L gliosarcoma rat model. (A) Representative T2-weighted coronal MR
images of a rat imaged 5, 7 and 9 days after intracranial tumor implantation. Tumor (outlined by a white
dotted line) and its associated edema (outlined by a black dotted line) can be seen on day 9. (B) Water
content of the ipsilateral cerebrum of tumor-bearing rats was significantly higher than that of the control
rats (****P < 0.0001). (C) Kaplan-Meier survival curve of rats with intracranial 9L gliosarcoma.
49
The intracranial 9L gliosarcoma rat model is more predictable and reproducible than the
transgenic mouse model. Longitudinal monitoring using MRI showed that edema usually started
to develop 9 days after tumor implantation (Figure 2.7A). Tumors were hard to distinguish on
MR images before day 9 and were shown to grow exponentially after day 9. Water content of the
ipsilateral cerebrum of tumor-bearing rats was significantly higher than that of control rats that
were not manipulated (Figure 2.7B). Rats with brain tumors survived for 13 to 16 days post
tumor implantation, with a median survival of 14 days (Figure 2.7C). Survival time after edema
development was about 4 to 7 days, thus this model is appropriate for testing effects of shortterm treatment but not long-term treatment of brain tumor associated edema.
2.4. CONCLUSIONS
Edema evaluation in the transgenic mouse model was challenging. We failed to measure
edema with MRI, and brain water content measurement might be obscured by hydrocephalus and
invasion of tumor across the midline. There was a huge variation in the time taken for the tumor
to grow exponentially, making it hard to plan treatment and MRI sessions. MRI techniques were
also limiting factors. Survival analysis of this mouse model could be complicated by the
development of lymphoma, in addition. We therefore concluded that this transgenic mouse
model was not appropriate for the preclinical study of treatment of brain tumor associated
edema, although it bears resemblance to human glioma in many aspects. 9L gliosarcoma rat
model is relatively more predictable and reproducible, thus subsequent in vivo therapeutic
efficacy studies were carried out using the 9L rat model.
50
CHAPTER 3
LOCAL DELIVERY OF DEXAMETHASONE
SODIUM PHOSPHATE
3.1. INTRODUCTION
Local drug delivery is a promising approach that can theoretically avoid the side effects
of systemic exposure, and allow increased local drug concentration and prolonged drug exposure
in the tumor by completely circumventing the blood brain barrier. High drug doses are generally
desirable at the tumor site to achieve sufficient anti-tumor activity but systemic toxicity has been
noted above a certain threshold. A dose escalation study was carried out in patients with
recurrent malignant glioma to determine the maximum tolerated dose (MTD) of carmustine
incorporated into the implantable polymers, in an attempt to improve the efficacy of the
commercially available Gliadel, which contains 3.85% of carmustine by weight [116]. Adverse
events, such as brain edema, seizures and infection, were more prominent in patients given
polymers with 20% carmustine by weight than in patients given lower doses; severe brain edema
and seizures were reported when the dose was escalated to 28% [116].
Although systemic toxicity was not identified in a preclinical study where local delivery
via polymer wafer incorporated with dexamethasone (DXM) was demonstrated to be effective in
reducing edema [106], intracranial implantation of similar wafer in normal rats resulted in
progressive weight loss over time, which was postulated to be a sign of systemic toxicity [130].
We hypothesize that optimizing DXM dosage in local delivery is important in achieving
treatment efficacy while minimizing systemic side effects. A reservoir-based, diffusion51
controlled drug delivery device that allowed for any drug to be used without further chemical
modification and for drug release rate to be easily altered, was developed based on the
microcapsule device design of Scott et al [107]. We loaded this device with different payloads of
dexamethasone sodium phosphate (DSP), a highly water soluble, inorganic ester of DXM, to
obtain varying drug release rates. The in vitro and in vivo release profiles of devices with
different payloads were characterized. An in vivo study in an intracranial 9L gliosarcoma rat
model was conducted to assess the effects of local delivery via these devices on reducing edema
and systemic toxicity. Brain water content, measured by the wet/dry weight method, was utilized
to evaluate edema while body weight and DXM level in plasma were used as indicators of
systemic toxicity.
3.2. MATERIALS AND METHODS
3.2.1. MATERIALS
Dexamethasone sodium phosphate injection solution was acquired from Baxter
(Deerfield, IL, USA). Dexamethasone, USP and dexamethasone sodium phosphate powder, USP
was obtained from Spectrum Chemicals & Laboratory Products (New Brunswick, NJ, USA).
High-performance liquid chromatography (HPLC) grade acetonitrile, water, sodium acetate,
methylprednisolone and dichloromethane were purchased from Sigma Aldrich (St. Louis, MO,
USA). HPLC grade glacial acetic acid was procured from Fisher Scientific (Fair Lawn, NJ,
USA). Phosphate buffered saline (PBS), pH 7.4, was bought from Corning (Manassas, VA,
USA).
52
3.2.2. DEVICE FABRICATION AND ASSEMBLY
Drug delivery devices, made of biocompatible Vectra MT1300 liquid crystal polymer,
were injection molded by microPEP (East Providence, RI, USA). Each device is comprised of a
reservoir and a cap. A single hole, 100 pm in diameter, was laser drilled in each cap at the
Massachusetts Institute of Technology (MIT)'s microsystems technology laboratories using a
laser ablation system. Profile of every laser-drilled hole was inspected under a VK laser scanning
microscope (Keyence Corporation, Elmwood Park, NJ, USA).
All the device parts were cleaned with 70% isopropanol before drug loading. DSP
powder was packed into a reservoir, which was then covered with a cap at the top. Biomedicalgrade epoxy was applied around the circumference of the cap to seal the cap to the reservoir. The
epoxy was allowed to dry for at least 24 h before drug release was initiated by removing residual
air trapped within the device using a vacuum system, while the device was submerged in sterile
PBS.
3.2.3. IN VITRO AND IN VIVO DRUG RELEASE
Activated devices (n = 3 for each payload) were incubated at 37C in PBS for in vitro
drug release. PBS was replaced every 24 h to maintain approximate sink conditions. The amount
of drug released in PBS was determined by HPLC.
Drug release rates in vivo were determined indirectly by subtracting the amount of drug
left in the explanted devices from the original payload. Explanted devices were retrieved from rat
brains at the end of the in vivo experiment on day 12 post tumor implantation, and their drug
contents were dissolved in sufficient PBS for measurement using HPLC.
53
3.2.4. HPLC ANALYSIS
HPLC analysis was performed using an Agilent 1200 series system (Agilent
Technologies, Wilmington, DE, USA) with the column temperature set at 30*C.
The HPLC method for determination of both DXM and DSP in in vitro samples was
adapted from Lamiable et al. [131]. The assay was conducted using an Agilent Zorbax Eclipse
Plus C18 column (75 x 4.6 mm, 3.5 im) with a mobile phase of 58% 2 mM sodium acetate
buffer (pH 4.8) and 42% acetonitrile ran at a flow rate at 1 mlImin, a detection wavelength at
246 nm, and an injection volume of 20 tiL. Good linearity was achieved for both DXM (R2
0.9997) and DSP (R2 =0.9997) in the concentration ranges of 1 pg/mL to 10 pg/mL and 1
pg/mL to 1,000 pg/mL respectively.
Blood samples were collected in heparinized tubes immediately after the animals were
euthanized by carbon dioxide asphyxiation on day 12 post tumor implantation. Plasma was
obtained by centrifuging heparinized blood at 4*C, 1,300 g for 10 min, and stored at -80*C until
analysis. The internal standard (1 jg of methylprednisolone) and 1 mL of dichloromethane were
added to 200 jL of the sample. The mixture was stirred for 15 min and centrifuged for 5 min at
2,600 g. The aqueous layer and creamy interface were discarded. The organic layer was
transferred to a reaction vial, evaporated to dryness at 450C under a stream of nitrogen and
reconstituted with 120
sL of mobile phase. The HPLC method used for the plasma samples was
similar to that used for the in vitro samples except that the mobile phase was delivered in a
gradient with the concentration of acetonitrile varying from 42% to 100% and then back to 42%
in a 15 min run. A linear calibration curve with R2 =0.9999 was obtained for DXM in the
concentration range of 0.1 Vg/mL to 10 [tg/mL.
54
3.2.5. IN VIVO EXPERIMENT
Male Fischer 344 (F344) rats, 200 - 300 g, were obtained from Charles River
Laboratories (Wilmington, MA, USA). All the animal experiments were conducted in
accordance with the MIT's Committee for Animal Care guidelines.
9L gliosarcoma was provided by Dr. Betty Tyler from Johns Hopkins University and
propagated in the flanks of carrier F344 rats. Tumor was harvested from a carrier rat and cut into
1 mm3 pieces for intracranial tumor implantation. Animals were anesthetized with intraperitoneal
(IP) injection of ketamine (75mg/kg) and xylazine (7mg/kg). The scalp was shaved and the
surgical site was prepped in a sterile fashion with betadine/ethanol scrub. A midline incision was
made followed by careful removal of the pericranium. A 3 mm burr hole was drilled over the left
hemisphere with its center 3 mm lateral to the sagittal suture and 5 mm posterior to the coronal
suture. The underlying cortex was gently aspirated to expose the sulcus, where the tumor piece
was placed. The skin was sutured and the animal was allowed to recover.
5 days after tumor implantation, rats were randomly assigned to the following groups (n
=5 per group): 1) no treatment (control), 2) intracranial blank devices (PBS Device), 3)
intracranial devices with an initial payload of 100 pg of DSP and a mean release rate of 7 pg/day
(DSP Device 7 ig/day), 4) intracranial devices with an initial payload of 800 pg of DSP and a
mean release rate of 36 Ig/day (DSP Device 36 jg/day), 5) intracranial devices with an initial
payload of 1,600 jg of DSP and a mean release rate of 50 pg/day (DSP Device 50 pg/day), 6)
intracranial devices with initial payload of 2,400 jg of DSP and a mean release rate of 88 pg/day
(DSP Device 88 jg/day), 7) daily IP injection of DSP at 1.644 mg/kg (DSP IP 411 pg/day*).
'411 pg/day is the dose for a 250g rat. Rats assigned to receive intracranial devices underwent a
second surgery for device implantation. The incision was reopened, and the connective tissue
55
overlying the burr hole was displaced. The device was implanted within the tumor, with its
orifices facing downwards and towards the tumor. Animals were euthanized 12 days after tumor
implantation. Rats in the DSP IP group were euthanized 2 h after the last IP dose. Their body
weights were recorded daily to monitor potential systemic toxicity.
3.2.6. BRAIN WATER CONTENT
Animals were euthanized by carbon dioxide asphyxiation on day 12 and their brains were
collected. Tumors were excised by blunt dissection before dividing the brains into ipsilateral and
contralateral cerebrums. Tumor, ipsilateral and contralateral cerebrums were weighed
immediately to obtain the wet weight, and then dried at 1000C for up to a week, until a final dry
weight was achieved. Tissue water content was calculated as water content = (wet weight - dry
weight)/ wet weight, and expressed as a percentage.
3.2.7. STATISTICAL ANALYSIS
All the data were expressed as mean
standard error of the mean (S.E.M.). Brain water
content, change in body weight and plasma DXM level were analyzed using ordinary one-way
analysis of variance (ANOVA) followed by Tukey's multiple comparisons tests for pair
comparisons among the different groups. P < 0.05 was considered statistically significant. All the
statistical analyses were performed using Graphpad Prism.
56
3.3. RESULTS AND DISCUSSION
3.3.1. DRUG RELEASE CHARACTERISTICS
A
2 mm
B
dx = 0.4 mm
rDrug isCap
releaExt
ednal
via orifice
2 mm
solution
Reservoir
Drug
solution
Figure 3.1 Drug delivery device. (A) Illustration of the device. (B) Schematics of how the device works.
Some drug inside the reservoir dissolved in PBS with the influx of PBS after vacuum. Drug then diffuses
across the orifice down a concentration gradient into the external solution. (C) Image of a laser-drilled
orifice, 100 pm in diameter, taken under a laser scanning microscope.
The drug delivery device, as shown in Figure 3.1, is a diffusion-controlled device
comprising of a reservoir of drug in its solid form and a cap with a 100 pm orifice through which
the drug is released. Influx of sterile PBS into the reservoir through the orifice, after removal of
residual gases within the device under vacuum, causes some drug to dissolve. Drug then diffuses
across the orifice in the cap down a concentration gradient into the external solution. Drug
release kinetics is determined by the drug payload, the solubility of the drug, the surface area of
diffusion (i.e. surface area of the orifice and number of orifices) and the diffusion distance (i.e.
distance between the solid drug and the exterior of the device).
DSP was chosen for its high solubility in water, estimated to be around 500 mg/mL. High
drug solubility was thought to be desirable since it would theoretically lead to a high drug release
rate, ensuring that sufficient dose was delivered locally for effective treatment. Our preliminary
study showed that implantation of a device that had a 100 pm orifice and a maximum payload of
57
2,400 pg of DSP led to weight loss in rats (data now shown). As the smallest orifice that can be
laser-drilled reliably is 100
sm in diameter,
the only way to decrease the release rate of DSP
from our device is to reduce the drug payload.
Figure 3.2A shows the release profiles of DSP from devices with different drug payloads
in PBS, plotted as cumulative amount of drug released against time, over the course of 7 days.
The amount of drug released decreased gradually with time for all the payloads. In vitro releases
were fairly reproducible as the coefficient of variation was < 10% for all the release rates. A
linear relationship was observed when plotting the cumulative release versus the square root of
time for every device (Figure 3.2B). The slope and correlation of the linear regression line of
best fit, in addition to the average release rate and coefficient of variation, for each payload are
shown in Table 3.1. The slopes of the lines of best fit varied linearly with the drug payloads
(Figure 3.2C). This linear relationship shows that the release kinetics of DSP from this drug
delivery device can be predicted based on the drug payload.
The Higuchi model [132] could be used to describe drug release from devices containing
payload above -880 jig . The basic equation of the Higuchi model is:
Mt
A
= VDC (2C - C)t,
C. > C,
(3.1)
where M, is the cumulative amount of drug released at time t, A is the surface area of diffusion, D
is the drug diffusivity in the polymer carrier, C, and C, are the initial drug concentration and the
solubility of the drug in the polymer respectively [132, 133]. Equation 3.1 can be expressed as:
--
= K %ft
(3.2)
58
where M, is the initial amount of drug loaded in the system, and K is a constant reflecting the
design variables of the system. An important condition is C, > C, for the justification of the
pseudo-steady state assumption. Our drug delivery system may behave according to the Higuchi
model if drug diffusion in the reservoir is not fast, resulting in a depletion layer adjacent to the
device cap (Figure 3.3) [134]. The device cap which has holes through which the drug is
released, acts as a rate-limiting membrane. C, can be assumed to be the saturated concentration
of DSP. A saturated solution is expected near the solid drug reservoir for drug payloads above
-880 pg in the early phase of drug release.
A proportionality between the cumulative amount of drug released and the square root of
time can also be derived from an exact solution of Fick's second law of diffusion for thin films
of thickness 8 under perfect sink conditions for C, < C. The amount of drug released is given by:
M
Dt 1/ 2
-- = 4-
1
~1/2 + 2
moo ,2
Y(-1)" ierfc
n=1
n6
2 DT
(3.3)
An approximation of equation 3.3 for short times and MtIM, <0.60 can be written as:
Mt
= 4
Dt 1/2
= k'fi
(3.4)
where k'is a constant. This equation might describe drug release from the device loaded with
100 pg of DSP.
59
A
+
+
+
800
600+
800 pg
1600 pg
2400 pg
100 pg (in vivo)
-4- 800 jig (in vivo)
400E 200
+
1600 pg (in vivo)
2400 pg (in vivo)
0
2
4
Time (days)
6
8
800
S600-
100 pg
S400800
jg
1600 pg
E
200
-
2400 pg
z-)
0
0
1
2
Square root of time (day
C
3
2
)
--100
ttg
400
y = 0.12x + 13.7
=?0.9676
300
200
2100CO
0-
0
1000
2000
Payload (pg)
3000
Figure 3.2 DSP release characteristics. (A) In vitro and in vivo release profiles of DSP from devices with
different drug payloads. In vitro release was performed in PBS while in vivo release at day 7 was obtained
indirectly by subtracting the amount of drug left in explanted device from the original drug payload. Data
represent mean values S.E.M. (n = 3 for each payload). (B) Same in vitro release data from (A) plotted
against the square root of time. Cumulative release was demonstrated to vary linearly with square root of
2
time for each payload. The lines represent linear regression lines of best fit. (C) A linear relationship (R
= 0.9) was observed between the slopes of the lines of best fit and the drug payloads.
60
Table 3.1 In vitro release data
Payload (jig)
Average release rate
ypg/day)
Coefficient of
variation (%)
Slope of line of best fit
(pg day-)
800
36
<2.76
130 (R =0.993)
2,400
88
5 6.87
319 (R2 =0.995)
A
h
CO
perfect
sink
.
CS
4
direction of depletion
direction of drug release
B
h
perfect
sink
-
CS
10
rate limiting
membrane
4-
direction of depletion
direction of drug release
Figure 3.3 (A) Schematic illustration of the Higuchi model. (B) Theoretical concentration profile within
the device reservoir where the drug transport is not fast. The device cap which has holes through which
the drug is released, acts as a rate-limiting membrane. The solid line in each diagram represents the
concentration gradient after time, t. The shift in concentration profile after interval of At is shown as a
dotted line.
In vivo drug release rates were determined indirectly by measuring the amount of drug
left in the explanted devices at the end of the in vivo experiment. The in vivo cumulative drug
release on day 7 was calculated by subtracting the amount of drug left in the explanted device
61
from the original payload. The in vivo cumulative drug release at the end of 7 days was found to
be close to the in vitro drug release for all the four payloads (Figure 3.2A), suggesting that in
vitro drug release could be a good approximate of actual in vivo release.
3.3.2. BRAIN WATER CONTENT
84
m
Control
PBS Device
m
M
82
0
DSP Device (7pg/day)
DSP Device (36pg/day)
DSP Device (50pg/day)
8OM
DSP Device (88pjg/day)
DSP IP (411pg/day*)
781
Tumor
ipsilateral
cerebrum
Contralateral
cerebrum
Figure 3.4 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals treated
with systemic or local administration of DSP had significantly lower water content in the tumor and
ipsilateral cerebrum than the untreated animals (P< 0.001). Animals that received PBS device are not
significantly different from the untreated animals. +411 pg/day is the DSP IP dose for a 250g rat. Data
represent the mean S.E.M. (n = 5 per group).
Water content of tumor, ipsilateral cerebrum and contralateral cerebrum were measured
by the wet/dry weight method and compared across all the groups relative to the control group
(Figure 3.4 and Table 3.2). Rats that received systemic and local DSP exhibited significantly
lower (P< 0.001) water content in the tumor and ipsilateral cerebrum than the rats in the control
and PBS device group at the end of the 7-day treatmenL Edema reduction appeared to be
independent of the route of administration and the dosage of DSP. There was no significant
62
difference among the groups treated with systemic and local DSP. Rats that received intracranial
blank devices were not significantly different from the untreated rats, suggesting that the device
alone did not worsen edema. The water content of the contralateral cerebrum was not
significantly different among all the groups. Based on these results, DSP seemed to exert its
effect on brain tumor associated edema only and not on normal brain tissues.
Table 3.2 Comparison of water content of tumor, ipsilateral cerebrum, contralateral cerebrum across all
the groups. *** P < 0.001, **** P < 0.0001, relative to the control group. Data represent the mean S.E.M.
(n= 5 per group). +411 pg/day is the dose for a 250g rat.
Water Content (%)
Treatment Group
Tumor
Ipsilateral Cerebrum
Contralateral Cerebrum
Control
83.4 + 0.2
80.0 i 0.1
78.9
0.0
DSP Device 7 pg/day
81.1
0.4
79.3
0.1
78.9
0.1
DSP Device 50 pg/day
80.8
0.4
79.0
0.1
78.8
0.1
DSP ]IP 411 pg/day'
80.7
0.4
78.9 i 0.1
78.9
0.1
3.3.3. SYSTEMIC DRUG EXPOSURE
DSP is converted to DXM by plasma enzymes. DXM levels in plasma samples, which
were analyzed by HPLC, are shown in Table 3.3. Dose-dependent systemic drug exposure was
observed with local delivery using our device. No DXM was detected in the plasma samples
collected from animals treated with DSP-loaded devices that had average release rates of 7
ig/day and 36 jig/day. The amount of DXM present could be too low to be detected by the
HPLC method used, which has a detection limit of 10 ng/mL. Only 2 out of 5 plasma samples
63
from animals that were given DSP-loaded devices with a mean release rate of 50 pg/day had
detectable amounts of DXM while all the plasma samples from animals that received DSP-
loaded devices with a mean release rate of 50 jg/day had measureable amounts of DXM. Our
results demonstrated that DSP released from our intracranial devices leaked into systemic
circulation at higher release rates. Average DSP release rate of not more than 36 pg/day appeared
to be ideal for minimizing systemic drug exposure, which can lead to a cascade of systemic side
effects. Peak DXM levels in plasma of rats in the DSP IP group, measured 2 h after the last IP
dose, were significantly higher than the DXM levels measured in all the other groups (**** P <
0.0001). Sustained, continuous, local drug delivery can prevent peak drug level in plasma which
is commonly seen in traditional routes of drug administration.
Table 3.3 Concentration of DXM in plasma samples. Data represent the mean
+ 411 pg/day is the dose for a 250g rat.
Treatment Group
DXM Concentration in Plasma (ng/mL)
DSP Device 36 pg/day
undetectable
DSP Device 88 pg/day
194.8
S.E.M. (n
5 per group).
29.5
3.3.4. DEXAMETHASONE-INDUCED WEIGHT LOSS
Systemically administered DXM has been shown to cause body weight loss in rodents
[135-138]. The percentage body weight change between day 5 and day 12 post tumor
implantation is plotted in Figure 3.5 to evaluate the body weight change induced by treatment.
64
10
_
0
-10
-20
-30
CO
9e
4O
4
PO
0
0
COO
Figure 3.5 Scatter plot of the percentage weight gain from day 5 to day 12 post tumor implantation, after
7 days of treatment. Dose-dependent weight loss was observed in animals treated with local delivery of
DSP. Systemic DSP dosing also led to significant weight loss. The lines represent the mean S.E.M. (n =
5 per group). ** P < 0.001, **** P < 0.0001, relative to the control group. '411 pg/day is the dose for a
250g rat.
The mean weight change of the rats that received blank devices (1.2
0.7%) was not
significantly different from that of the untreated rats, suggesting that the device did not induce
toxicity. Dose-dependent weight loss was observed in the animals that received intracranial DSPloaded devices. Increase in DSP release rate was associated with greater weight loss. Rats that
were given DSP devices with 1,600 ig of DSP, DSP devices with 2,400 pg of DSP and IP DSP
injection experienced weight changes of -9.2
and -17.1
1.9 % (P<0.001), -19.4
1.1 % (P< 0.0001),
0.9 % (P < 0.0001) respectively, which were significantly lower compared with that
of the untreated controls (-0.5
1.1 %). Rats that were given DSP devices loaded with 100 pg
65
and 800 gg of DSP, had weight changes of 0.2
1.1% and -1.5
0.9% respectively, which were
not significantly different from each other and from that of controls. It, therefore, appeared that
the mean DSP release rate should not exceed 36 pg/day in order to reduce edema without
systemic toxicity. This result is similar to that obtained from measuring DXM concentration in
plasma samples. Furthermore, the mean weight change of animals that were given the highest
local dose of 88 pg/day was not significantly different from that of animals dosed systemically at
411 pg/day, implying that continuous local drug delivery can be as toxic as once-a-day systemic
dosing even though a lower total dose was administered in local delivery.
3.4. CONCLUSIONS
We have developed a reservoir-based, diffusion-controlled intracranial drug delivery
device that was capable of delivering DSP in a reproducible and predictable manner for
treatment of brain tumor associated edema. Drug release was proportional to the initial payload
of DSP. Good in vitro-in vivo correlation was also observed. Local delivery of varying doses of
DSP was demonstrated to be as effective as systemic dosing in reducing edema. However, local
delivery of DSP can lead to dose-dependent systemic toxicity, as reflected by the concentration
of DXM in plasma and the change in body weight. Local administration of DSP via our device,
at an average rate of not exceeding 36 pg/day, was found to be effective in treating edema while
avoiding systemic side effects.
66
CHAPTER 4
LOCAL DELIVERY OF DEXAMETHASONE
4.1.
INTRODUCTION
Optimizing dosage in local drug delivery can potentially improve treatment outcomes by
achieving efficacy with minimal systemic toxicity. We have shown in Chapter 3 that only a very
little amount of dexamethasone sodium phosphate (DSP) needs to be locally delivered to treat
brain tumor associated edema without systemic complications. A low drug release rate of DSP
was previously achieved by using a low drug payload. Another approach to obtain low drug
release rate was to use drug of low solubility in water. Dexamethasone (DXM) is an ideal drug
candidate. The solubility of DXM in water (-50 gg/mL) is approximately 10,000 times less than
the solubility of DSP (-500 mg/mL).
The release rate of DXM will primarily depend on the surface area of diffusion, i.e. the
size and number of orifices in the device cap, instead of the initial payload. Varying drug release
rates were obtained by changing the number of orifices, each 300 pm in diameter. The drug
delivery device was micro-machined, instead of injection molded as in Chapter 3, to facilitate the
fabrication of the orifices. We investigated the in vitro stability and release kinetics of DXM
from this micro-machined intracranial drug delivery device. In vivo distribution of DXM in rat
brain was also examined. A more comprehensive in vivo study than the one in Chapter 3 was
carried out to evaluate the effects of local delivery, using the same intracranial 9L gliosarcoma
rat model. 2 local doses were examined and compared in this in vivo study. Therapeutic efficacy
was assessed by analyzing survival, measuring tumor and its associated edema using magnetic
67
resonance imaging (MRI), in addition to measuring brain water content by the wet/dry weight
method. Body weight, DXM and corticosterone levels in plasma were used to measure systemic
toxicity. Plasma DXM level was used as an indicator of systemic drug exposure while the
concentration of corticosterone, the equivalent of cortisol in human, in plasma was used to assess
hypothalamic-pituitary-adrenocortical (HPA) activity, which could be suppressed by DXM
administration.
4.2.
MATERIALS AND METHODS
4.2.1. MATERIALS
Dexamethasone sodium phosphate injection solution was acquired from Baxter
(Deerfield, IL, USA). Dexamethasone, USP and dexamethasone sodium phosphate powder, USP
was obtained from Spectrum Chemicals & Laboratory Products (New Brunswick, NJ, USA).
High-performance liquid chromatography (HPLC) grade acetonitrile, water, sodium acetate,
methylprednisolone and dichloromethane were purchased from Sigma Aldrich (St. Louis, MO,
USA). HPLC grade glacial acetic acid was procured from Fisher Scientific (Fair Lawn, NJ,
USA). Phosphate buffered saline (PBS), pH 7.4, was bought from Corning (Manassas, VA,
USA). Biocompatible polysulfone blocks and sheets were acquired from AmazonSupply
(Seattle, WA, USA).
4.2.2. DEVICE FABRICATION AND ASSEMBLY
Drug delivery devices, each comprising of a reservoir and a cap, were micro-machined
from biocompatible polysulfone blocks using the computer numerical control (CNC) micro
68
machining center (Cameron Micro Drill Presses, Sonora, CA, USA). The reservoirs and caps
were micro-machined in bulk separately (Figure 4.1).
The reservoirs were fabricated in a polysulfone block by first pocket milling to create the
insides of the reservoirs where the drug will be contained, and then profile milling to make the
walls of the reservoirs with attached bridges to the surrounding block. The bridges held the
machined reservoirs in place so that the block could be flipped and face milled to remove excess
material. The bridges were then manually cut away to obtain the reservoirs.
The caps were manufactured from a polysulfone sheet taped to a sacrificial block. The
polysulfone sheet was faced to the thickness of the cap before holes, 300 pm in diameter, were
drilled. Individual caps were produced by profile milling and then carefully removed from the
adhesive tape. Profile of every drilled hole was inspected under a VK laser scanning microscope
(Keyence Corporation, Elmwood Park, NJ, USA) for quality control.
All the device parts were cleaned with 70% isopropanol before drug loading. DXM
powder was packed into a reservoir, which was then covered with a cap at the top. Biomedicalgrade epoxy was applied around the circumference of the cap to seal the cap to the reservoir. The
epoxy was allowed to dry for at least 24 h before drug release was initiated.
4.2.3. IN VITRO DRUG RELEASE
Drug release was initiated by removing residual air trapped within the device using a
vacuum system, while the device was submerged in PBS. Activated devices (n= 5 for each type
of device) were incubated at 37"C in PBS for in vitro drug release. PBS was replaced every 24 h
to maintain approximate sink conditions. The amount of drug in PBS was determined by HPLC.
69
B. Device cap fabrication
A. Device reservoir fabrication
tape
-support
-
4,
4,
4,
bridges
H
C. Device assembly
A
Remove and clean
0 300 pm
holes
Cap
Remove
bridges
and
clean
I
Load drug
*
2 mm
Reservoir
Figure 4.1 Device fabrication and assembly. Fabrication of (A) device reservoir and (B) device cap were
performed separately. Schematics of the micro-machining processes were shown. (C) Device assembly
involved loading the drug into the reservoir, covering the reservoir with a cap, and then sealing the cap to
the reservoir with biomedical grade epoxy.
70
4.2.4. DRUG STABILITY
The stabilities of solid DXM and DXM in PBS at 37*C were evaluated over the course of
10 days. A 1 mg/mL stock solution of DXM in ethanol was divided into aliquots such that each
aliquot contained 10 pg of DXM. Ethanol was then evaporated to dryness at room temperature
overnight, and the resultant solid was incubated at 37C to assess the stability of solid DXM. A
10 pg/mL stock solution of DXM in PBS was divided into aliquots and incubated at 37*C to
examine the stability of DXM in PBS. The aliquots (n = 3) were collected at different time points
and stored at -20*C until analysis by HPLC. Aliquots containing DXM in PBS were thawed and
analyzed directly with HPLC while solid DXM was dissolved in PBS before HPLC analysis.
4.2.5. HPLC ANALYSIS
HPLC analysis was performed on an Agilent 1200 series system (Agilent Technologies,
Wilmington, DE, USA) with the column temperature set at 30C. The method for measuring
DXM in in vitro samples was adapted from Lamiable et aL [131]. The assay was conducted
using an Agilent Zorbax Eclipse Plus C18 column (75 x 4.6 mm, 3.5 pm) with a mobile phase of
58% 2 mM sodium acetate buffer (pH 4.8) and 42% acetonitrile ran at a flow rate at 1 mlImin, a
detection wavelength at 246 nm, and an injection volume of 20 pL. Good linearity was achieved
for DXM (R2 = 0.9997) in the concentration range of 1 pg/mL to 10 pg/mL.
4.2.6. ANIMAL CARE
Female Fischer 344 (F344) rats, 150 - 175 g, were bought from Charles River
Laboratories (Wilmington, MA, USA). All the animal experiments were done in accordance with
the Massachusetts Institute of Technology's Committee for Animal Care guidelines.
71
4.2.7. DRUG DISTRIBUTION IN RAT BRAIN
Rats were randomly assigned to one of the following groups (n = 3 per group): 1)
intracranial DXM-loaded device with 1 hole (DXM 1H Device), 2) intracranial DXM-loaded
device with 9 holes (DXM 9H Device), 3) daily intraperitoneal (IP) injection of DSP at 1.64
mg/kg (DXM IP). Rats assigned to receive intracranial devices underwent a surgical procedure
for device implantation. They were anesthetized with IP injection of ketamine (75mg/kg) and
xylazine (7mg/kg). The scalp was shaved and the surgical site was prepped in a sterile fashion
with betadine/ethanol scrub. A midline incision was made followed by careful removal of the
pericranium. A 3 mm burr hole was drilled over the left hemisphere with its center 3 mm lateral
to the sagittal suture and 5 mm posterior to the coronal suture. The dura was gently removed to
expose the underlying cortex, where the device was implanted with its orifice(s) facing
downwards and against the cortex. The skin was sutured and the animal was allowed to recover.
Animals were euthanized by carbon dioxide asphyxiation 14 days after start of treatment,
and their brains were harvested immediately. Rats in the DXM IP group were euthanized 2 h
after the last IP dose. Each brain was sectioned into 1 mm thick coronal slices using an Altom
stainless stain brain matrix (Roboz Surgical Instrument Co. Inc., Gaithersburg, MD, USA). Each
brain slice (-100 mg) was weighed, placed in -0.2 mL of ice-cold PBS with -100 mg of 0.5 mm
zirconium oxide beads (Next Advance, Averill Park, NY, USA) in a micro-centrifuge tube, and
homogenized for 3 min at the maximum speed setting using the Bullet Blender® (Next Advance).
The brain homogenate samples were centrifuged at 4 0C, 15,000 g for 15 min, and stored at -80 0C
until analysis. The concentration of DXM in plasma samples was determined by an enzymelinked immunosorbent assay (ELISA) kit (Neogen, Lexington, KY, USA) according to the
vendor's protocol. Samples were sufficiently diluted so that their DXM concentrations can be
72
read from the standard curve, which was plotted utilizing a 4-parameter logistic curve fit
algorithm for the concentration range of 0.5 ng/mL to 10 ng/mL.
4.2.8. EFFICACY STUDY
9L gliosarcoma was provided by Dr. Betty Tyler from Johns Hopkins University and
propagated in the flanks of carrier F344 rats. Tumor was harvested from a carrier rat and cut into
1 mm3 pieces for intracranial tumor implantation. Animals were anesthetized with intraperitoneal
(IP) injection of ketamine (75mg/kg) and xylazine (7mg/kg). The scalp was shaved and the
surgical site was prepped in a sterile fashion with betadine/ethanol scrub. A midline incision was
made followed by careful removal of the pericranium. A 3 mm burr hole was drilled over the left
hemisphere with its center 3 mm lateral to the sagittal suture and 5 mm posterior to the coronal
suture. The underlying cortex was gently aspirated to expose the sulcus, where the tumor piece
was placed. The skin was sutured and the animal was allowed to recover.
9 days after tumor implantation, rats were randomly assigned to one of the following
groups (n= 6 per group): 1) no treatment (control), 2) intracranial blank device (PBS Device), 3)
daily IP injection of DSP at 1.64 mg/kg (DXM IP), 4) intracranial DXM-loaded devics with 1
orifice (DXM 1H Device), 5) intracranial DXM-loaded device with 9 orifices (DXM 9H
Device). Rats assigned to receive intracranial devices underwent a second surgery for device
implantation. The incision was reopened, and the connective tissue overlying the burr hole was
displaced. The device was implanted within the tumor, with its orifices facing downwards and
towards the tumor. Animals were checked twice daily and euthanized when found in poor body
condition. Rats in the DXM IP group were euthanized 2 h after the last IP dose. The body
weights of all the animals were recorded daily from day 9 onwards to monitor their health.
73
4.2.9. MAGNETIC RESONANCE IMAGING
MRI was performed on 7T Varian MRI scanner (Agilent Technologies). Rats were
anesthetized with 2% isoflurane and placed prone in the cradle. T2 weighted images (TR =
4,000ms, TE = 40ms, 1mm slice thickness, 12 image slices, 128x128 matrix size, 32mm x 32mm
field of view) were acquired to assess tumor and its associated edema on day 9 and on day 12
post tumor implantation. Tumor and edema volumes were measured using OsiriX imaging
software.
BRAIN WATER CONTENT
4.2.10.
Animals were euthanized by carbon dioxide asphyxiation when they were found in poor
body condition. Their brains were collected immediately after sacrifice. Tumors were excised by
blunt dissection before dividing the brains into ipsilateral and contralateral cerebrums. Tumor,
ipsilateral cerebrum and contralateral cerebrum were weighed immediately to obtain the wet
weight, and then dried at 100C for up to a week until a final dry weight was achieved. The
tissue water content was calculated as water content = (wet weight - dry weight)/ wet weight,
and expressed as a percentage.
4.2.11.
PLASMA DEXAMETHASONE AND CORTICOSTERONE
Blood samples were collected in heparinized tubes immediately after the animals were
euthanized. All the animals were euthanized around the same time of the day, between 5pm and
7pm, to minimize the influence of daily fluctuations of corticosterone levels. Plasma was
obtained by centrifuging heparinized blood at 4C, 1,300 g for 10 min, and stored at -80C until
analysis. The concentration of DXM in plasma samples was determined by an enzyme-linked
74
immunosorbent assay (ELISA) kit (Neogen, Lexington, KY, USA) according to the vendor's
protocol. Samples were sufficiently diluted so that their DXM concentrations can be read from
the standard curve, which was plotted utilizing a 4-parameter logistic curve fit algorithm for the
concentration range of 0.5 ng/mL to 10 ng/mL. Plasma corticosterone level was also quantified
using an ELISA kit (Enzo Life Sciences, Farmingdale, NY, USA) following the manufacturer's
protocol. Samples were sufficiently diluted so that their corticosterone concentrations can be
determined from the standard curve, which was plotted utilizing a 4-parameter logistic curve fit
algorithm for the concentration range of 32 pg/mL to 20,000 pg/mL.
4.2.12.
STATISTICAL ANALYSIS
All the data were expressed as mean
standard error of the mean (S.E.M.). Tumor and
edema volumes were analyzed using repeated measures two-way analysis of variance (ANOVA)
in combination with Tukey's and Holm-Sidak's multiple comparisons tests for pair comparisons
among the groups and for comparison between day 9 and day 12 within each group respectively.
Log-rank (Mantel-Cox) test was used to compare the survival curves. One-way ANOVA
followed by Tukey's multiple comparisons test for pair comparisons was performed for all the
other experiments. P <0.05 was considered statistically significant. All the statistical analyses
were performed using Graphpad Prism.
75
4.3.
RESULTS
4.3.1. IN VITRO RELEASE
Our device is comprised of a reservoir of solid drug and a cap with hole(s) through which
drug is released. Influx of PBS into the device through the hole(s), after removal of residual
trapped air within the device via vacuum, caused some drug to dissolve. Drug then diffuses
across the hole(s) down a concentration gradient into the external solution (Figure 4.2).
Saturated
drug solution
External
solution
6 =0.4
mm
Figure 4.2 Illustration of the spatial concentration profile of the device. c2, the drug concentration inside
the device, is assumed to be the saturated drug concentration in the early phase of release while c1, the
drug concentration outside the device, is assumed to be zero in perfect sink conditions. 8, the distance
over which the concentration gradient forms, is assumed to be the thickness of the cap.
The release of DXM from our drug delivery system can be described by Fick's first law,
1= -D-
aC
ax
(4.1)
=dM
J = -D dx A dt
(4.2)
76
dMt
DA
dt
8
(4.3)
DA
Mt =(C2
- C1)t
(4.4)
where J is the diffusion flux, D is the drug diffusivity, dc/dx is the concentration gradient
(spatial derivative), M, is the cumulative amount of drug released up to time t, A is the surface
area for diffusion, c2 is the concentration of drug inside the device, c1 is the concentration of drug
outside the device and 8 is the distance over which the concentration gradient exists. Steadystate, unidirectional diffusion is assumed in equations 4.2, 4.3, and 4.4. c1 can be assumed to be
constantly zero in perfect sink conditions while c2 can be assumed to be the saturated
concentration of DXM in the device reservoir during the early phase of drug release. 8 can be
assumed to be the thickness of the cap. As the term DA (c2
-
c1 ) in equation 4.4 is a constant, M,
varies linearly with time as in zero-order release kinetics.
Releases of DXM from our drug delivery devices with different number of orifices in
PBS were observed to follow zero-order kinetics over the course of 10 days (Figure 4.3A). The
empirical release curves, plotted as cumulative amount of drug released against time, were fitted
using a linear regression model, and the mean release rates were determined from the slopes of
the linear regression lines of best fit. The mean release rates of DXM-loaded devices with 1 hole
(DXM 1H device), 5 holes (DXM 5H device), and 9 holes (DXM 9H device) were 0.58
pig/day (R2
=
0.998), 3.37
0.03 pg/day (R2
=
0.982), and 5.91
0.06 jIg/day (F
2
0.00
= 0.984)
respectively. The coefficients of variation were < 4.2 %<, 15.0 %, and S8.4 %for the release
77
rates of DXM 1H device, DXM 5H device, and DXM 9H device, suggesting that the releases
were fairly reproducible.
A
7o
A 70+
50
-+-
DXM 11H Device
DXM 5H Device
-
DXM 9H Device
y = 5.91x
R2 = 0.984
E4030
y =3.37x
R= 0.982
-20.
E10
10
y = 0.58x
0
6
8
4
Time (days)
2
R 2 =O0.998
12
10
B
y =0.66x
R= 0.998
=640
0
2
4
6
Number of orifices
8
10
Figure 4.3 In vitro releases of DXM from drug delivery devices. (A) Cumulative release curves of DXM
from devices with lhole (DXM 1H Device), 5 holes (DXM 5H Device) and 9 holes (DXM 9H Device).
Releases of DXM from these devices were shown to obey zero-order kinetics. The lines represent linear
regression lines of best fit. (B) Plot of the mean release rate against the number of orifices in the device. A
linear relationship (R2 = 0.998) was observed between the mean release rates, determined from the slopes
in (A), and the number of orifices. Data represent mean values S.E.M. (n = 5).
Equation 4.4 shows that drug release (M,) is directly proportional to the surface area for
diffusion (A), which is defined by the number of orifices in our drug delivery system. This is
consistent with our plot of the mean release rate versus the number of orifices (Figure 4.3B). A
linear relationship (R2 = 0.998) was revealed between the mean release rate and the number of
orifice, implying that the desirable DXM release rate can be obtained by manipulating the
number of holes in the device.
78
4.3.2. DRUG STABILITY
10
S
DXosli
0
-+DXM solid
-- DXM in PBS
80
0151
0
1
2
3
4
5
6
7
8
91
Time (days)
Figure 4.4 Stabilities of DXM in solid form and in PBS over the course of 10 days. Data were expressed
as the mean percentage of the initial amount of drug on day 0 S.E.M. (n = 3 for each time point).
DXM was stable in solid form and in PBS over the course of 10 days at 37"C (Figure
4.4). DXM was dissolved in PBS and its concentration was analyzed using HPLC for the
stability test of solid DXM, while the concentration of DXM in PBS was measured directly with
HPLC for the stability test of DXM in PBS. There was no evidence of drug degradation in both
tests. Stability of DXM in both solid and solution forms makes it a good drug candidate for
reservoir-based controlled drug delivery.
4.3.3. DRUG DISTRIBUTION IN RAT BRAIN
DXM penetration was limited in brains of rats that received intracranial devices (Figure
4.5). Systemic DXM IP dosing, on the contrary, resulted in a relatively uniform distribution of
DXM throughout the brain. DXM-loaded devices with 9 holes were found to give very high
DXM concentration, which was more than 3 times that found in systemically dosed rats, at the
79
site of device implantation. The concentration of DXM dropped exponentially away from the
device and no drug was measured more than 4 mm away from the device. When devices with a
single hole were implanted in the rat brains, the concentration of DXM at the site of implantation
was similar to that in systemically dosed rats, 2 h after the last dose. DXM levels dipped sharply
away from the device and were not detectable beyond 3 mm from the 1-hole device.
e
80
60-
--
DXM 1 H Device
DXM 9H Device
-r-
DXM IP
~40E
c
X
20
0-
-5 -4 -3
-2 -1
0
12
3
4
5
Distance from implantation site (mm)
Figure 4.5 Distribution of DXM in rat brain resulting from systemic DXM dosing (DXM IP), intracranial
implantation of DXM-loaded devices withl hole (DXM IH Device) and 9 holes (DXM 9H Device). Data
represent mean values S.E.M. (n = 3).
4.3.4. TUMOR AND EDEMA QUANTIFICATION FROM MRI
Representative T2-weighted coronal MR images of all the treatment groups before (day 9)
and 3 days after treatment (day 12) are shown in Figure 4.6A. Intracranial 9L gliosarcoma was
evident in the left hemisphere as an isointense or hyperintense region with a distinct boundary.
Peritumoral regions with hyperintense signal were identified as brain tumor associated edema.
When the tumor reached a minimum volume of 8 mm 3 on day 9, as measured on MR images,
tumor-bearing rats were randomized to one of the following treatment groups: 1) no treatment
80
(control), 2) intracranial implantation of blank device (PBS device), 3) daily IP injection of DSP
at 1.64 mg/kg (DXM IP), 4) intracranial implantation of DXM-loaded device with 1 hole (DXM
1H device), 5) intracranial implantation of DXM-loaded device with 9 holes (DXM 9H device).
Tumor volumes on day 12 were significantly higher (P < 0.05) than those on day 9 in all
the groups (Figure 4.6B). No significant difference was observed among all the groups on either
day 9 or day 12. The average relative increase in tumor size ranged from 4.6 to 6.1 fold, with no
statistical difference among all the groups (Figure 4.6C).
Treatment with DXM via either local or systemic administration was demonstrated to
impede the development of edema. The volume of edema increased significantly (P < 0.01) from
day 9 to day 12 in animals that were not treated with DXM, i.e. the control and PBS device
groups; but no significant increase was exhibited in the DXM IP, DXM 1H device, and DXM 9H
device groups - groups that were treated with DXM (Figure 4.6D). The volume of edema was
not significantly different among all the groups on day 9, before the start of treatment. After 3
days of treatment on day 12, rats that were treated with DXM had significantly less (P < 0.001)
edema than the untreated rats, while rats that received blank devices showed no significant
difference compared with the controls. The device alone did not appear to improve or worsen
edema. The effect of DXM on edema was not dose-dependent as the volume of edema was not
significantly different among the groups that received some form of DXM treatment. These
observations were consistent with the results obtained from plotting the relative change in edema
volume from day 9 to day 12 (Figure 4.6E). The relative change in edema volume was -0.48
0.08, -0.45
0.07, and -0.58
0.07 fold in the DXM IP, DXM 1H device, and DXM 9H device
groups respectively. These changes were significantly lower (P < 0.0001) than the relative
changes of 1.86
0.23 and 2.18
0.38 fold in the control and PBS device groups respectively.
81
PBS Device
Control
A
U...
U...
Day 9
Day 12
D
100
m
PBS Device
W DXM IP
M DXM 1H Device
80
E 60
m
DXM 9H Device
20mm
Day
.2
W DXM 9H Device
Izt
Day 12
ii
U
V
1111111111IMM
II
5-
I
I
A
S 0
A
A
0
E 3-
U
0
*
S
A
75
~
V
V
00
*
0
Eu
0
-U-
2-
E
*
42-
Day 12
Day 9
4.
*
I-
DXM IP
m DXMI H Device
VU
9
M
E
_____
*
-1
8-
PBS Device
E
E
10-
a 0
IM E
'U
Control
0
0
C
W
0
E
'"10-
40I
E
15
1
U
410
U.
-
E
DXMV 9H
Device
DXMV 1H
Device
DXM IP
-i LU
/
'/*
~
,. Vo
%I
OF
/
.?
4
r4
VYV
AAA
1
4
~
'i
Figure 4.6 Tumor and edema assessment using MRL (A) Representative T 2-weighted coronal MR images
before (day 9) and after (day 12) treatment. (B) Tumor volumes measured on MR images. There was no
significant difference among all the groups on both day 9 and day 12. (C) Relative change in tumor
volume calculated from volumes in (B) confirmed that increase in tumor was not significantly different.
(D) Edema volumes measured on MR images. Edema on day 12 was significantly reduced in groups
treated with systemic or local DXM compared with controls (***P < 0.001). (E) Relative change in edema
volume calculated from volumes in (D) confirmed that the decrease in edema was significant relative to
control (****P < 0.0001). Data are expressed as the mean values i S.E.M. (n= 6 per group).
82
4.3.5. BRAIN WATER CONTENT
86
8
Control
-
PBS Device
-
DXM 9H Device
DXM IP
M DXM IH Device
84-
80
8
Tumor
lpsilateral
Cerebrum
Contralateral
Cerebrum
+
Figure 4.7 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals treated
with systemic or local administration of DXM had significantly lower water content in the tumor and in
the ipsilateral cerebrum than the control animals (*P < 0.05, **P< 0.01). Data represent mean values
S.E.M. (n = 3 per group).
Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum were
measured by the wet/dry weight method and compared across all the groups relative to the
control group (Figure 4.7). Rats that were administered systemic or local DXM in the DXM IP,
DXM 1H device and DXM 9H device groups had significantly lower (P< 0.05) water content in
the tumor and in the ipsilateral cerebrum than the rats in the control and PBS device groups.
Edema reduction appeared to be independent of the route of administration and the dosage of
DXM since there was no significant difference among the 3 groups that were treated with DXM.
The device alone did not worsen edema as the brain water content of the rats that received
intracranial blank devices was not significantly different from that of the untreated rats. The
water content of the contralateral cerebrum was not significantly different among all the groups,
83
suggesting that DXM exerted its effect on brain tumor associated edema without affecting the
water content of normal brain tissues.
4.3.6. SURVIVAL
-0- Control
100
PBS Device
- -
IP
DXM IH Device
DXM
DXM 9H Device
50-0
0
9L
0
12
14
16
18
20
22
Time (days)
Figure 4.8 Kaplan-Meier survival curves. Systemic and local delivery of DXM led to significant survival
benefit (P< 0.05, n =6 per group).
The Kaplan-Meier survival curves of all the treatment groups are depicted in Figure 4.8.
Untreated control animals and animals that were given blank intracranial devices had median
survival of 14 and 14.5 days respectively, with all the animals succumbing to the tumor within
15 days of tumor implantation. Survival of the control group was not significantly different from
that of the PBS device group, implying that the blank device had no impact on survival. When
compared with the control group, survival was significantly prolonged (P< 0.05) in the DXM
IP, DXM 1H device, and DXM 9H device groups, which had median survival of 18.5, 17.5 and
17.5 days respectively. Rats in these 3 groups survived for at least 16 days and at most 21 days.
There was no significant difference in survival among these groups. The improvement in
survival in rats treated with DXM was independent of dose and route of DXM administration.
84
4.3.7. DEXAMETHASONE-INDUCED WEIGHT LOSS
A
11
+
Control
-U- PBS Device
DXMIP
1
DXM 1H Device
DXM 9H. Device
go.
9 10 11 12 13 14 15 16 17 18 19 20 21 22
Time (days)
B 110
m
.*
PBS Device
*
DXMIP
*
E 100
Control
I1M
DXM IH Device
1
DXM 9H Device
90
801
i
Day 16
Day 12
Figure 4.9 Relative body weight. (A) Plots of relative body weight versus time since the start of treatment
on day 9. All the groups had decreasing weight with time, with the DXM IP group experiencing the most
weight loss. (B) Comparison of the mean relative weights on day 12 and day 16. The DXM IP group had
significantly lower relative body weight than all other groups on both days (*P < 0.05). Data represent
mean values
S.E.M. (n = 6 per group).
Body weight loss was used as an indicator of systemic toxicity. The body weight of all
the groups, which was monitored daily from day 9 (the start of treatment) onwards, is plotted as
a percentage of the weight on day 9 in Figure 4.9A. All the groups displayed downward trends in
body weight with time, with the group that received systemic DXM losing the most weight.
Comparisons of the mean relative weights were made among all the groups on day 12
and day 16 to study the effect of short-term (3-days duration) and long-term (7-days duration)
85
treatment (Figure 4.9B). Both short-term and long-term systemic administration of DXM led to
significantly lower (P < 0.05) relative body weight than any other treatment. The DXM IP group
had a mean relative weight of 93.1
97.9
0.7 %, compared with 97.9
0.8 % for the PBS device group, 97.9
1.3 % for the control group,
1.3 % for the DXM 1H device group, and 97.9
0.4 % for the DXM 9H device group on day 12, after 3 days of treatment. There was no
significant difference among the control, PBS device, DXM 1H device, and DXM 9H device
groups, suggesting that local delivery of DXM could avoid DXM induced weight loss. 7 days
after treatment on day 16, rats in the DXM IP group, again, had significantly lower (P< 0.05)
relative body weight (88.1
DXM 9H device (92.1
1.2 %) than rats in the DXM 1H device (92.9
1.2 %) and the
0.3%) groups. No dose-dependent effect on body weight was observed
in the rats treated with locally administered DXM. The mean relative body weight of the DXM
1H device group was not significantly different from that of the DXM 9H device group on both
day 12 and day 16 although a higher local dose was delivered in the DXM 9H device group.
4.3.8. PLASMA DEXAMETHASONE AND CORTICOSTERONE
An important goal of local drug delivery is to minimize systemic drug exposure. The
concentration of DXM in plasma at the time of sacrifice was measured and compared among the
different groups to study the effect of treatment on systemic drug exposure (Figure 4.10A). Local
delivery of DXM in the brain failed to prevent DXM from leaking into systemic circulation, but
managed to reduce systemic exposure to DXM. DXM level in plasma of rats in the DXM IP
group, 2 h after the last IP dose, was found to be 3265
146 ng/mL, which was approximately 3
orders of magnitude higher than the DXM levels measured in plasma of rats treated with locally
delivered DXM . The concentration of DXM in plasma of rats that received local sustained
86
delivery of DXM via our 1-hole device and 9-holes device were 1.2
0.1 ng/mL and 8.1
0.8
ng/mL respectively, showing that the systemic level of DXM increased with rise in locally
administered DXM dose. Both route of administration and dosage, even in the case of local
delivery, were found to be important factors of systemic exposure to DXM.
A
B
1000
M000
1000
100
~1o0
E
10E
I
I
Figure 4.10 Plasma levels of (A) DXM and (B) corticosterone at the time of sacrifice. Plasma DXM level
in the DXM IP group was 3 orders of magnitude higher than the plasma DXM levels in the DXM 1H
device and DXM 9H device groups. Plasma corticosterone level in the DXM IP group was significantly
lower than that in any other group (P < 0.05). Data represent the mean values S.E.M. (n = 6 per group).
An adverse effect due to systemic exposure to DXM is corticosterone suppression.
Plasma corticosterone levels at the time of sacrifice, depicted in Figure 4.1OB, were measured to
assess systemic toxicity. The plasma concentration of corticosterone in rats that were treated with
systemic DXM, 6.4
0.9 ng/mL, was significantly lower (P< 0.05) than that in all other rats.
Plasma corticosterone levels of the control, PBS device, DXM 1H device and DXM 9H device
groups were 362.3
73.6 ng/mL, 399.0
92.3 ng/mL, 361.0
95.4 ng/mL and 349.0
93.4
ng/mL respectively, and there was no significant difference among these groups. Local delivery
of DXM in the brain, independent of the dose, was shown to avoid corticosterone suppression
which was associated with systemic dosing of DXM.
87
4.4. DISCUSSION
We have developed a reservoir-based, highly tunable drug delivery system for the local
delivery of DXM in the rat brain. A key feature of this system is the ability to modify drug
release rates by controlling the number of holes in the device. Sustained, zero-order releases of
DXM from devices with different number of orifices were achieved and accurately predicted by
Fick's first law. Drug degradation did not need to be accounted for during release studies since
DXM was found to be extremely stable at body temperature in both solid form and when
dissolved in PBS. The high stability of DXM also implies that DXM is suitable for our drug
delivery system. It is important that the solid DXM stored in the device reservoir will not
degrade when the device is implanted in the brain.
A similar type of drug delivery system was reported in an earlier study, but the parts were
manufactured by injection molding [107]. A novelty of this work is the use of micro-machining
for device fabrication. Micro-machining is an appealing method for making prototypes in pilot
studies as it enables fast turnaround time and flexibility in design alteration without
compromising on quality. The devices that we micro-machined have tight tolerances and can be
reliably reproduced, resulting in reproducible drug releases with these devices. The number of
orifices in our devices can be easily changed to obtain different release rates by simply changing
the number of times the devices are drilled during the micro-machining process. If the devices
were to be injection molded, different molds would be required for making devices with different
number of orifices. Micro-machining generally requires no tooling, therefore it is more cost
efficient than injection molding for production in small quantities. Although injection molding
can give a much lower cost per part than micro-machining in large-scale production, tooling cost,
88
which makes up the bulk of the upfront production cost, can make injection molding prohibitive
for small-scale prototype production.
Biocompatibility of the device is evident from the in vivo results. Blank devices did not
induce body weight loss, which is a reflection of systemic toxicity. No abnormalities were
observed. Blank devices also had no adverse or beneficial impact on tumor growth, edema
formation, and overall survival. Outcomes of rats that were given blank devices were essentially
indistinguishable from those of control rats.
In vivo evaluations were carried out with the 1-hole and 9-holes devices to investigate
whether different DXM release rates had an impact on drug distribution, therapeutic efficacy and
systemic side effects. The DXM release rate of the 9-holes device was an order of magnitude
higher than that of the 1-hole device, and less than twice that of the 5-holes device. Devices with
5 holes were not tested in in vivo experiments as it was thought that its performance would be
somewhere between that of the 1-hole and 9-holes devices.
Both 1-hole and 9-holes devices showed limited drug distribution profiles, which are
expected, since drug distribution occurs mainly by diffusion. The concentration of DXM was
high at the site of device implantation but dropped exponentially away from the device. The
concentration gradient and the distance over which DXM was distributed were greater for the 9holes device than that for the 1-hole device due to higher drug release. The concentration of
DXM at the site of device implantation obtained with 9-holes devices was about 4 times more
than that with 1-hole devices, although the drug release rate of the 9-holes device was about 10
times higher than that of the 1-hole device. Absence of perfect sink conditions at the site of
device implantation may explain the discrepancy. Furthermore, the concentration of DXM at the
site of implantation of the 1-hole device was found to coincide with the concentration of DXM in
89
brains of rats 2 h after the last systemic dose. Peak DXM levels in rat brain should be observed 2
h after systemic dosing, based on a previous pharmacokinetic study [1301. Our finding suggests
that the 1-hole device is capable of maintaining peak DXM level seen in systemic dosing, in the
tumor region throughout the duration of treatment The cumulative amount of DXM that the
tumor was exposed to could be similar or greater in rats that were treated with the 1-hole devices
intracranially than in rats that were dosed with IP injection of DXM. This cumulative amount of
DXM delivered in the tumor region is projected to be much higher in rats that were treated with
the 9-holes devices that in rats that received the 1-hole devices or daily systemic dosing.
Therapeutic efficacies of local DXM delivery via 1-hole and 9-holes devices and
systemic administration of DXM were not significantly different one another, despite the
differences in drug dosage and distribution in the brain. Local delivery of DXM using the
devices was shown to be as effective as systemic dosing of DXM in alleviating edema, as
measured by MRI and by brain water content. Edema reduction was associated with modest but
significant survival benefit, despite continuous tumor progression.
Systemically administered DXM is known to induce body weight loss and corticosterone
suppression in rodents [135-138]. Local administration of DXM using 1-hole and 9-holes
devices were shown to be equally effective in eliminating these systemic side effects, although
systemic level of DXM was higher in rats treated with 9-holes devices than in rats with 1-hole
devices. Leakage of locally delivered DXM into systemic circulation was minimal but inevitable,
despite the limited drug distribution profile of the devices. It was possible that systemic exposure
to DXM due to both 1-hole and 9-holes devices was below the threshold that would trigger
corticosterone suppression. Another possibility was that plasma corticosterone changes might be
detectable if the plasma samples were obtained at peak times in the circadian cycle. The samples
90
were drawn from euthanized rats within 2 hours before the end of the light cycle in our study so
corticosterone level would not be at its peak in a Fischer rat [139].
DXM has been shown to be a better drug candidate for our drug delivery system than
DSP, the highly water soluble, inorganic ester of DXM, when a low release rate is required. The
differences in release rates and release profiles between these 2 drugs arise from the huge
disparity in solubility. The release rate of DSP is higher and decreases gradually with time, while
the release rate of DXM is lower and characterized by zero-order kinetics. A device with
maximum load of DSP and a single 100 pm hole has a mean release rate of 88 pg/day; a device
with maximum load of DXM and a single 300 pm hole delivers DXM at a constant rate of 0.58
jig/day, which is 2 orders of magnitude lower than the DSP device, even though the DSP device
has a smaller hole. Delivery of DXM via our device therefore offers the advantage of long-term,
zero-order release, which is not possible with DSP. Another advantage of using DXM is that
maximum payload can be used to achieve the release rate(s) at which brain tumor associated
edema is reduced without systemic side effects. All the DXM-loaded devices used in this study
had maximum payloads. Drug delivery via our device, that had the smallest orifice possible and
the maximum payload of DSP, was shown to result in systemic toxicity in Chapter 3. The only
way to reduce release of DSP was to utilize lower initial payload. Lowering the payload helped
to eliminate systemic toxicity but the space within the device was not maximized. Maximizing
payload is essential for a device designed for implantation in the brain, since an intracranial
device should ideally be as small as possible to minimize additional mass effect, which can lead
to increased intracranial pressure within the rigid skull.
Local DXM delivered via the 1-hole and 9-holes devices were merely 2.36% and 0.23%
of the systemic dose (Table 4.1). The systemic dose used in our study was translated from typical
91
human dose (16 mg/day) using the body surface area normalization method [140]. Local drug
delivery with our device offers the benefit of using a small fraction of the dose used in systemic
delivery to achieve similar efficacy while minimizing the risk of toxicity associated with
systemic delivery.
Table 4.1 DXM dose comparison. 'Dose converted to DXM equivalent.
Comparison with
systemic dose (%)
Dose and Schedule
Total dose in 7
days/ pg
DXM IP (systemic)
250 gg once a day'
1750+
DXM IH Device (local)
sustained release, 5.9
41.4
2.36
DXM 9H Device (local)
sustained release,
4.1
0.23
sg/day
0.58 sg/day
-
Treatment
4.5. CONCLUSIONS
Reproducible, sustained release of DXM from our novel, micro-machined, drug delivery
device was achieved. Drug release was characterized by zero-order kinetics and could be easily
altered by changing the number of orifices in the device. Local delivery of DXM in rat brains via
our devices was demonstrated to be as effective as systemic delivery in alleviating edema.
Edema reduction was associated with survival benefit, despite continuous tumor progression.
Local DXM delivery also eliminated side effects, such as weight loss and corticosterone
suppression, caused by systemic dosing of DXM. Edema alleviation and systemic complications
were not dose-dependent with local delivery of DXM. Local drug administration using our drug
delivery system is superior to systemic administration as it allows the use of a much lower dose
to achieve similar efficacy with minimal risk of systemic toxicity. Its use can be extended to
other therapeutic agents targeting brain tumor to achieve similar outcome.
92
CHAPTER 5
LOCAL DELIVERY OF CEDIRANIB
5.1.
INTRODUCTION
Brain tumor associated edema is vasogenic in nature, characterized by the disruption of
blood brain barrier (BBB). The exact mechanism behind the breakdown of BBB is unclear. One
school of thought is that the tumor production of vascular endothelial growth factor (VEGF)
stimulates the formation of gaps in brain endothelium and leads to plasma leakage into the
extracellular space of brain parenchyma, resulting in edema [13]. VEGF has been shown to be
upregulated in both malignant gliomas [9] and metastatic tumors [10], and its level in high-grade
gliomas correlates with the occurrence of edema [12]. Therefore, agents that target the VEGF
signaling pathway are thought to be promising alternatives to corticosteroids, the standard of care
for brain tumor associated edema [421. Cediranib (AZD2171), an oral pan-VEGF receptor
tyrosine kinase inhibitor, induced vascular normalization within 24 hours and alleviated brain
edema in glioblastoma patients in a phase II trial [50]. Corticosteroid usage was reduced or
discontinued as a result [50, 51]. Bevacizumab, a monoclonal antibody against VEGF, also
demonstrated anti-edema activity and corticosteroid-sparing effect in a phase I study [47].
Treatment with systemically administered VEGF inhibitors, as with corticosteroids, is
associated with systemic side effects [55]. This is not surprising since VEGF is essential for
normal physiological functions. Another challenge with the systemic delivery of VEGF
inhibitors is the BBB. The brain distribution of cediranib was shown to be limited by Pglycoprotein, an active efflux transporter at the BBB [141].
93
Local drug delivery strategies can avoid the side effects of systemic exposure, and allow
increased drug concentration in the tumor by circumventing the BBB. One approach that has
much success is the implantation of diffusion-controlled, sustained-release polymer systems at
the tumor site. Treatment with Gliadel*, a carmustine impregnated biodegradable polymer wafer,
improved median survival by 2 months without systemic toxicity in patients with malignant
glioma [98, 99]. The matrix-based polymer system used in Gliadel@ has also been used to
deliver other chemotherapeutics, such as doxorubicin [102], taxol [103], and temozolomide
[104], in preclinical models. Another type of polymer system is the reservoir-based polymer
device. Local delivery of temozolomide using such a system has led to prolonged survival in a
9L rat model [107]. A reservoir-based system offers the advantages of high drug payload and
long-term, nearly zero-order drug release kinetics over a matrix-based system [96].
A reservoir-based drug delivery device, same as the one used in Chapter 4, was used to
test the hypothesis that local delivery of cediranib can reduce brain tumor associated edema
effectively with minimal systemic toxicity. Cediranib is practically insoluble in aqueous media.
Increasing its solubility, by employing the solid dispersion technique, was necessary to enhance
its delivery via diffusion. We investigated the in vitro stability, release kinetics and biological
activity of this solid dispersion of cediranib (AZD/PVP). Therapeutic efficacy was assessed by
analyzing survival, measuring tumor and its associated edema using magnetic resonance imaging
(MRI), in addition to measuring brain water content by the wet/dry weight method. Body weight
was used to measure systemic toxicity. The therapeutic efficacy and systemic toxicity associated
with local delivery of AZD/PVP were examined in an intracranial 9L gliosarcoma rat model, and
compared with other treatments, including local delivery of dexamethasone (DXM), systemic
delivery of DXM, and systemic delivery of cediranib.
94
5.2.
MATERIALS AND METHODS
5.2.1. MATERIALS
Cediranib was obtained from Selleck Chemicals (Houston, TX, USA). Dexamethasone
sodium phosphate (DSP) injection solution was acquired from Baxter (Deerfield, IL, USA).
Povidone K-30, polysorbate 80, and micronized dexamethasone, USP were bought from
Spectmm Chemicals & Laboratory Products (New Brunswick, NJ, USA). Dimethyl sulfoxide
(DMSO), high-performance liquid chromatography (HPLC) grade acetonitrile, water, methanol,
sodium acetate, ammonium formate and formic acid were purchased from Sigma Aldrich (St.
Louis, MO, USA). HPLC grade glacial acetic acid was procured from Fisher Scientific (Fair
Lawn, NJ, USA). Polysulfone blocks and sheets were obtained from AmazonSupply (Seattle,
WA, USA).
5.2.2. PREPARATION OF SOLID DISPERSION OF CEDIRANIB
Solid dispersion of cediranib in povidone K-30 (PVP) was prepared by the conventional
solvent evaporation method, adapted from Sethia and Squillante [142]. Briefly, cediranib and
PVP, in the weight ratio of 1 to 2, were dissolved in minimum volume of methanol. Methanol
was removed under vacuum in a rotavapor at room temperature for 15 minutes, followed by
evaporation under vacuum for at least 12 hours. The resultant solid dispersion was scraped out
with a spatula and refrigerated for 2 days at 4"C to harden. It was pulverized using a mortar and
pestle before use.
95
5.2.3. DEVICE FABRICATION AND ASSEMBLY
The drug delivery devices, each comprising of a reservoir and a cap, were micromachined as previously described in Chapter 4.2.2. All the caps used in this study had 9 holes,
each 300 pm in diameter. Profile of every drilled hole was inspected under a VK laser scanning
microscope (Keyence Corporation, Elmwood Park, NJ, USA) for quality control. All the device
parts were cleaned with 70% isopropanol before drug loading. Each reservoir was packed with
either AZD/PVP or DXM, and then covered with a cap at the top. Biomedical-grade epoxy was
applied around the circumference of the cap to seal the cap to the reservoir. The epoxy was
allowed to dry for at least 24 h before drug release was initiated.
5.2.4. IN VITRO DRUG RELEASE
Drug release was initiated by removing residual air trapped within the device using a
vacuum system while the device was submerged in sterile PBS. Activated devices (n=5 for each
drug) were incubated at 370C in PBS for in vitro drug release. PBS was replaced every 24 h to
maintain approximate sink conditions. The amount of drug in PBS was determined by HPLC.
5.2.5. DRUG STABILITY
The stabilities of solid AZD/PVP and AZD/PVP in phosphate buffered saline (PBS) at
37"C were evaluated over the course of 28 days. A 1 mg/mL stock solution of AZD/PVP in
methanol was divided into aliquots such that each aliquot contained 10 pg of AZD/PVP.
Methanol was then evaporated to dryness at room temperature overnight, and the resultant solid
AZD/PVP was incubated at 37"C to assess the stability of solid AZD/PVP. A 10 jpg/mL stock
solution of AZD/PVP in PBS was divided into aliquots and incubated at 37C to examine the
96
stability of AZD/PVP in PBS. The aliquots (n = 3) were collected at different time points and
stored at -20"C until analysis by HPLC. Aliquots containing AZD/PVP in PBS were thawed and
analyzed directly with HPLC while solid AZD/PVP was dissolved in PBS before analysis using
HPLC.
5.2.6. HPLC ANALYSIS
HPLC analysis was performed on an Agilent 1200 series system (Agilent Technologies,
Wilmington, DE, USA) with the column temperature set at 30'C. The HPLC method for
determination of DXM was adapted from Lamiable et al. [131]. 20 jLb of sample was quantified
using an Agilent Zorbax Eclipse Plus C18 column (75 x 4.6 mm, 3.5 pm) and a mobile phase,
consisted of 58% 2 mM sodium acetate buffer (pH 4.8) and 42% acetonitrile, delivered at 1
mi~min. The effluent was monitored at 246 nm. Concentration of cediranib was quantified using
a HPLC method modified from Wang et al. [1431. The assay was conducted using a Phenomenex
(Torrance, CA, USA) Kinetex C18 column (50 x 4.6 mm, 2.6 pm), with a mobile phase of 10
mM ammonium formate containing 0.1% formic acid: acetonitrile (62: 38) ran at a flow rate of 1
mlimin, a detection wavelength of 235 nm, and an injection volume of 20
sL.
5.2.7. IN VITRO HUVEC PROLIFERATION ASSAY
Cell cultures were maintained at 370 C with 5% carbon dioxide. Human umbilical vein
endothelial cells (HUVEC), purchased from Life Technologies (Carlsbad, CA, USA), were
grown in Medium 200 supplemented with low serum growth supplement (Life Technologies), at
a density of 1,000 cells/ well in 96 well plates. Stock solutions of cediranib and AZD/PVP were
prepared in DMSO at a concentration of 10 mM, and diluted to varying concentrations in an
97
assay medium composed of phenol red-free Medium 200, 2% v/v fetal bovine serum, 1 pg/mL
hydrocortisone (Life Technologies), 3 ig/mL heparin (Sigma Aldrich), and 3 ng/mL VEGF16 5
(R&D Systems, Minneapolis, MN, USA). After an attachment period of 24 h, the cells were
dosed with varying concentrations of cediranib or AZD/PVP. Cells dosed with vehicle (DMSO)
were used as control. The cell cultures were then incubated for 4 days and cellular proliferation
was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MIT) assay
(Sigma Aldrich). Briefly, M7TT was added and incubated for 4 h. The resultant formazan crystals
were dissolved in acidified isopropanol and the absorbance was measured by a microplate reader
at 570 nm with a reference wavelength of 690 nm. The experiments were performed in triplicate
and the relative cell viability was expressed as a percentage relative to the control.
5.2.8. IN VIVO EXPERIMENT
5.2.8.1. Animals
Female Fischer 344 (F344) rats, 150 - 175 g, were obtained from Charles River
Laboratories (Wilmington, MA, USA). All animal experiments were conducted in accordance
with the Massachusetts Institute of Technology's Committee for Animal Care guidelines.
5.2.8.2. Intracranial 9L gliosarcoma rat model
9L gliosarcoma was provided by Dr. Betty Tyler from Johns Hopkins University and
propagated in the flanks of carrier F344 rats. Tumor was harvested from a carrier rat and cut into
1 mm3 pieces for intracranial tumor implantation. Animals were anesthetized with intraperitoneal
(IP) injection of ketamine (75mg/kg) and xylazine (7mg/kg). The scalp was shaved and the
surgical site was prepped in a sterile fashion with betadine/ethanol scrub. Midline incision was
made, followed by careful removal of the pericranium. A 3 mm burr hole was drilled over the
98
left hemisphere with its center 3 mm lateral to the sagittal suture and 5 mm posterior to the
coronal suture. The underlying cortex was gently aspirated to expose the sulcus, where the tumor
piece was placed. The skin was sutured and the animal was allowed to recover.
5.2.8.3. Study design
9 days after tumor implantation, rats were randomly assigned to one of the following
groups (n =7 per group): 1) no treatment (control), 2) intracranial blank device (PBS device), 3)
daily IP injection of DSP at 1.64 mg/kg (DXM IP), 4) intracranial device delivering DXM
(DXM device), 5) daily oral gavage of cediranib suspended in 1% (w/v) aqueous polysorbate 80
at 4.63 mg/kg (AZD oral), 6) intracranial device delivering AZD/PVP (AZD device). Rats
receiving intracranial devices underwent a second surgery for device implantation. The incision
was reopened, and the connective tissue overlying the burr hole was displaced. The device was
implanted within the tumor, with its orifices facing downwards and towards the tumor. Animals
were checked twice daily and euthanized when found in poor body condition. Their body
weights were recorded daily from day 9 onwards to monitor their health.
5.2.8.4. Magnetic resonance imaging
MRI was performed on 7T Varian MRI scanner (Agilent Technologies). Rats were
anesthetized with 2% isoflurane and placed prone in the cradle. T2 weighted images (TR = 4,000
ins, TE = 40 ms, 1 mm slice thickness, 12 image slices, 128 x 128 matrix size, 32 mm x 32 mm
field of view) were acquired to assess tumor and its associated edema on day 9 and on day 12
post tumor implantation. Edema and tumor volumes were measured using Osirix imaging
software.
99
5.2.9. BRAIN WATER CONTENT
Animals were euthanized by carbon dioxide asphyxiation and their brains were collected.
Tumors were excised by blunt dissection before dividing the brains into ipsilateral and
contralateral cerebrums. Tumor, ipsilateral cerebrum and contralateral cerebrum were weighed
immediately to obtain the wet weight, and then dried at 100*C for up to a week until a final dry
-
weight was achieved. The tissue water content was calculated as water content = (wet weight
dry weight)/ wet weight, and expressed as a percentage.
5.2.10.
STATISTICAL ANALYSIS
All the data were expressed as mean
standard error of the mean (S.E.M.). Dose-
response curves were fitted with a nonlinear regression model to determine the drug
concentration where the response was reduced by half (IC50). Logarithms of IC50 values were
compared using extra sum-of-squares F test. Tumor and edema volumes were analyzed using
repeated measures two-way analysis of variance (ANOVA) in combination with Tukey's and
Holm-Sidak's multiple comparisons tests for pair comparisons among the groups and for
comparison between day 9 and day 12 within each group respectively. Log-rank (Mantel-Cox)
test was used to compare the survival curves. One-way ANOVA followed by Tukey's multiple
comparisons test for pair comparisons was performed for all other experiments. P < 0.05 was
considered statistically significant. All statistical analyses were performed using Graphpad
Prism.
100
5.3. RESULTS
5.3.1. IN VITRO DRUG RELEASE
The drug delivery device, as shown in Figure 5.1, is comprised of a reservoir of solid
drug and a cap with orifices through which drug is released. Influx of PBS into the reservoir,
after removal of residual gases within the device via vacuum, caused some drug to dissolve.
Drug then diffuses across the orifices down a concentration gradient into the external solution.
A
0 300 pm
holes
30p
cap
reservoir
2 mm
Figure 5.1 (A) Illustration and (B) photograph of an assembled drug delivery device with (C) 9 holes,
each 300 pm in diameter.
70
3
-e-DXM
- AZDIPVP
so0
504030'020-
S10
00
1
2
3 4 5 6 7
Time (day)
8
9 10 11
Figure 5.2 Cumulative release profiles of DXM and AZD/PVP. Data represent mean values
= 5 for each drug)
101
S.E.M. (n
Figure 5.2 shows the release profiles of DXM and AZD/PVP from the drug delivery
device in PBS, plotted as cumulative amount of drug released against time. The release of DXM
was observed to follow zero-order kinetics (R2 = 1.00) with a mean rate of 5.91
0.06 pg/day for
the duration of the 10-day release experiment. The release profile of AZD/PVP, on the other
hand, was characterized by an initial burst effect (13.22
0.43 pg) on the first day, followed by a
decreased, steady release at an average rate of 2.23 : 0.08 jg/day (R2= 0.95). In vitro releases of
both drugs were fairly reproducible. The coefficient of variation was 5 8.41% and
7.24% for
mean release rates of DXM and AZD/PVP respectively.
5.3.2. DRUG STABILITY
110
105
AZD/PVP solid
1a10-0-
-4- AZD/PVP in PBS
0
95-
090
C-
85
0
10
20
30
Time (days)
Figure 5.3 Stabilities of AZD/PVP in solid form and in PBS for 28 days. Data were expressed as the
mean percentage of the initial amount of drug on day 0 S.E.M. (n = 3 for each time point).
The stabilities of AZD/PVP in solid form and in PBS, over the course of 28 days at 37 0C,
are depicted in Figure 5.3. AZD/PVP was extremely stable in solid form with no evidence of
degradation throughout the course of 28 days. The concentration of AZD/PVP in PBS, in
contrast, decreased steadily with time, and approximately 10% of the drug was lost at the end of
102
28 days. AZD/PVP is expected to be protected from degradation when stored in solid state in a
device that is implanted intracranially. Stability of solid AZD/PVP at 37 0 C makes it a good drug
candidate for reservoir-based drug delivery.
5.3.3. BIOACTIVITY
VEGF-stimulated proliferation of HUVEC was inhibited by cediranib and AZD/PVP in a
similar dose-dependent manner, with IC50 values determined to be 21.92 nM and 48.33 nM
respectively (Figure 5.4). Logarithms of their IC50 values were not significantly different,
suggesting that AZD/PVP had similar biological activity as cediranib.
,120-
10--0-
AZD/PVP
100-
AZ
80
60
40
20
01
0.001 0.01
0.1
1
10 100 1000 10000
Drug concentration, nM
Figure 5.4 Dose-response curves of AZD/PVP and cediranib (AZD). Both drugs inhibited VEGFinduced HUVEC proliferation in similar dose-dependent fashion. Data represent mean values S.E.M.,
performed on triplicate experiments.
5.3.4. IN VIVO EFFICACY STUDY
The effects of different treatments on brain tumor, its associated edema, and survival
were evaluated in an intracranial 9L gliosarcoma rat model.
103
5.3.4.1. Tumor and edema quantification from MRI
Representative T2 -weighted coronal MR images before (day 9) and 3 days after treatment
(day 12) are shown in Figure 5.5A. Intracranial tumor was evident in the left hemisphere as an
isointense or hyperintense region with a distinct boundary. Peritumoral regions with hyperintense
signal were identified as brain tumor associated edema. Rats with tumors that were >8 mm3 on
day 9 were randomly assigned to one of the following groups: 1) no treatment (control), 2)
intracranial implantation of blank device (PBS device), 3) daily [P injection of DSP (DXM IP),
4) intracranial implantation of device delivering DXM (DXM device), 5) daily oral gavage of
cediranib (AZD oral), 6) intracranial implantation of device delivering AZD/PVP (AZD device).
Tumor volumes on day 12 were significantly higher (P< 0.0001) than those on day 9 in
all the groups (Figure 5.5B). No significant difference was observed among all the groups on
both day 9 and day 12. The average relative increase in tumor size ranged from 4.1 to 5.5 fold,
with no statistical difference among all the groups (Figure 5.5C).
Treatment with DXM or cediranib via either local or systemic administration was
demonstrated to impede the development of edema. The volume of edema increased significantly
(P< 0.0001) from day 9 to day 12 in the control and PBS device groups, but no significant
increase was detected in the DXM IP, DXM device, AZD oral and AZD device groups - groups
that were treated with systemic or local DXM or cediranib (Figure 5.5D). The volume of edema
was not significantly different among all the groups on day 9, before the start of treatment. After
3 days of treatment on day 12, rats in the DXM IP, DXM device, AZD oral and AZD device
groups had significantly less (P< 0.001) edema than the controls, while rats in the PBS device
group showed no significant difference compared with the controls. The device alone did not
appear to improve or worsen edema.
104
PBS Device
Control
A
AZD Oral
DXM Device
DXM IP
AZD Device
Day 9
Day 12
B
Control
PBS Device
150- M
- PBS Device
EDXIP
1
DX1Device
M, DXM r
1
DXM
Device
AZD Ora
-
= AZD Oral
= AZD Device
AZD Device
5
50
E-
Day9
C
Day 9
Day 12
E
15-
I
4-
Day 12
I
-
I
II
**.*
3.
-E
U
a
= 10-
C.>
S
4WZ0
S
A
so
AAA
0
0~
00
V
WV
V
V
U-
" I0
1-
a
I.
*
00
U
0-
0
0
V
-5.-
*
0
:~~4
A
~0
4
~
~
P
OP
cp~
~
,40
-~
-i~i.-.
0
0
oO
0
Figure 5.5 Tumor and edema assessment using MRIL (A) Representative Trweighted coronal MR images
before (day 9) and after (day 12) treatment. (B) Tumor volumes measured on MR images. (C) Relative
change in tumor volume calculated from the volumes in (B). Both (B) and (C) show that there was no
significant difference in tumor growth among all the groups. (D) Edema volumes measured on MR
images. Edema on day 12 was significantly reduced in groups treated with DXM or cediranib compared
with the control group (***P < 0.001). (E) Relative change in edema volume calculated from the volumes
in (D) confirmed that the decrease in edema was significant relative to control (****P <0.0001). Data
represent mean values S.E.M. (n = 7 per group).
105
These observations were consistent with the results obtained from plotting the relative
change in edema volume from day 9 to day 12 (Figure 5.5E). The relative change in edema
0.07, -0.45
volume was -0.53
0.04, -0.45
0.08 fold in the DXM IP, DXM
0.09 and -0.63
device, AZD oral and AZD device groups respectively. These changes were significantly lower
(P< 0.0001) than the relative change of 1.75
0.29 fold seen in the control and
0.30 and 1.90
PBS device groups respectively. No significant difference was observed among the groups
treated with DXM or cediranib.
5.3.4.2. Brain water content
86
M Control
PBS Device
m DXM IP
M-%*84
M
DXM Device
-
AZD Oral
M AZD Device
82
80-
78-
Tumor
Ipsilateral
Cerebrum
Contralateral
Cerebrum
Figure 5.6 Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum. Animals treated
with systemic or local administration of DXM and cediranib had significantly lower water content in the
tumor and ipsilateral cerebrum than the control animals (*P < 0.05). Data represent the mean S.E.M. (n
=4 per group).
Water content of the tumor, ipsilateral cerebrum and contralateral cerebrum were
measured by the wet/dry weight method and compared across all the groups relative to the
control group (Figure 5.6). DXM and cediranib were equally effective in reducing brain tumor
106
associated edema. Rats that were systemically or locally administered DXM or cediranib - in the
DXM IP, DXM device, AZD oral and AZD device groups - exhibited significantly lower (P <
0.05) water content in the tumor and ipsilateral cerebrum than the control group (Figure 5.6).
There was no significant difference among these 4 groups. The water content of the ipsilateral
cerebrum and the contralateral cerebrum were not significantly different in these 4 groups,
whereas the water content of the ipsilateral cerebrum was significantly higher (P < 0.01) than
that of the contralateral cerebrum in both the control and PBS device groups. The water content
of the contralateral cerebrum was not significantly different among all the groups, suggesting
that DXM and cediranib exerted their effects on brain tumor associated edema without affecting
the water content of normal brain tissues.
5.3.4.3. Survival
100
--
Control
-u- PBS Device
-+-
DXM Device
+
AZD Oral
--
AZD Device
-50
0+-4
0
12
14
16
DXM IP
18
20
22
Time (days)
Figure 5.7 Kaplan-Meier survival curves. Systemic and local delivery of both dexamethasone and
cediranib led to significant survival benefit (P< 0.05, n =7 per group).
107
The Kaplan-Meier survival curves of all the treatment groups are depicted in Figure 5.7.
Both the control and PBS device groups had the same median survival of 14 days, with all rats
succumbing to the tumor within 15 days of tumor implantation. No significant difference was
found between them. When compared with control, survival was significantly prolonged (P <
0.05) in the DXM IP, DXM device, AZD oral and AZD device groups, which shared the same
median survival of 16 days, minimum survival of 15 days and maximum survival of 19 days.
There was no significant difference in survival among these 4 groups. DXM and cediranib,
independent of the route of drug delivery, were equally effective in improving survival.
5.3.5. SYSTEMIC TOXICITY
Body weight was used as an indicator of systemic toxicity. The body weight of all the
groups, which was monitored daily from day 9 onwards, is plotted as a percentage of the weight
on day 9 in Figure 5.8A. All groups displayed decreasing body weight with time, with the DXM
IP group losing the most weight.
Comparisons of the mean relative weights were made among all the groups on day 12
and day 16 to study the effect of short-term (3-days duration) and long-term (7-days duration)
treatment (Figure 5.8B). Both short-term and long-term systemic administration of DXM led to
more severe weight loss than any other treatment. The DXM IP group had a significantly lower
relative weight of 92.3
97.0
0.6 %, compared with 97.0
0.7 % for the PBS device group (P< 0.05), 99.7
0.001), 100.7
1.4 % for the control group (P < 0.05),
1.0 %for the DXM device group (P<
0.7 % for the AZD oral group (P < 0.0001), and 101.0
1.6 % for the AZD
device group (P < 0.0001) on day 12. There was no significant difference among the control,
PBS device, DXM device, AZD oral and AZD device groups. Similar trend continued on day 16,
108
1.2 %) having significantly lower relative body weight than the
with the DXM IP group (83.2
DXM device (92.6
(93.1
1.5 %, P < 0.01), AZD oral (94.8
1.7 %, P < 0.001) and AZD device
2.1%, P < 0.01) groups. No significant difference was found among the DXM device,
AZD oral and AZD device groups. Local delivery of DXM in the brain prevented DXM induced
weight loss seen with systemic dosing of DXM. Cediranib, on the other hand, did not seem to
have an effect on body weight.
A
11o.
-+
control
-W-
PBS Device
,
DXM
DXM Device
iso-
.p_
AZDOral
~90.
AmD Device
9
10 11
12 13 14 15 16
17 18 19 20
Time (days)
B
110
**
m
Control
M
PBS Device
m DXMIP
mDXM
Device
SAZm Oial
mMD Device
Sol
Day 1
Day 12
Figure 5.8 Relative body weight as a measure of systemic toxicity. (A) Plots of body weight versus time
since start of treatment on day 9. All groups had decreasing weight with time, with the DXM IP group
experiencing the most weight loss. (B) Relative body weights of the groups on day 12 and day 16. Mean
relative weight of the DXM IP group is significantly lower than that of any other group on day 12 and day
16 (*P < 0.05, **P< 0.01, ***P < 0.001, ****P< 0.0001). Data represent mean S.E.M. (n = 7 per group).
109
5.4. DISCUSSION
The efficacy of corticosteroids in treating brain tumor associated edema is well
established [30]. Systemic corticosteroids have remained the standard of care as there is no better
option [5]. Corticosteroids are not only associated with numerous detrimental systemic
complications, but they may also antagonize the cytotoxic and anti-proliferative effects of
concurrently administered chemotherapy. The most commonly used corticosteroid, DXM, has
been shown to protect glioblastoma cell lines from apoptosis induced by temozolomide [144,
145] and cisplatinium [146].
Efforts in improving treatment outcome can be broadly classified into two categories:
drug development and development of drug delivery method. Cediranib and bevacizumab are
two VEGF inhibitors - a new class of anti-edema agents - that have demonstrated their potential
as alternatives to corticosteroids in clinical trials. Cediranib has been proven to be active in rats
[147]; while bevacizumab, a recombinant humanized antibody to human VEGF, does not bind to
rodent VEGF [148]. Cediranib was therefore chosen for testing in the rat intracranial 9L
gliosarcoma model, which is widely used for the study of brain tumor associated edema [21, 41,
106, 120, 121]. Furthermore, cediranib has been shown to synergistically sensitize glioma cells
to temozolomide [149, 150], unlike DXM, which has been shown to mitigate the efficacy of
temozolomide [144, 145]. Cediranib may therefore be a better drug candidate for the treatment of
brain tumor associated edema than DXM
Cediranib has always been given orally in studies of glioblastoma [50, 51, 126] but never
locally in the brain. Local delivery of cediranib in the brain was investigated in this study, with
the goal of alleviating brain edema while minimizing systemic toxicity. Common toxicities
110
associated with systemic cediranib include hypertension, fatigue and diarrhea [51]. Comparisons
between cediranib and DXM and between local and systemic delivery were made to determine if
local delivery of cediranib gave the best treatment outcomes.
Reproducible, sustained release of DXM and AZD/PVP from our device was achieved.
Release profile is dependent on drug solubility, surface area of diffusion (i.e. surface area of the
orifices) and diffusion distance (i.e. distance between the solid drug and the exterior of the
device). When a saturated or near saturated solution is maintained inside the device reservoir,
zero-order release kinetics is expected. This was demonstrated with the release of DXM.
Difference in the release profile of AZD/PVP from that of DXM might be explained by the
higher solubility of AZD/PVP which led to a greater initial release. Drug loss with this initial
burst decreased the chemical driving force. A concentration gradient formed within the device
reservoir and diffusion distance increased. Consequently, diffusion decreased and a near-steady
state was reached, giving rise to a diminished, steady release of AZD/PVP.
Our data validated that the bioactivity of AZD/PVP was comparable to that of cediranib.
It is important that AZD/PVP retains the biological activity of cediranib for it to be efficacious
when delivered locally in the brain. Cediranib is a highly potent inhibitor of the kinase activity of
VEGF receptor-2 (KDR), which plays a key role in mediating VEGF-induced responses [147]. It
has been reported to inhibit VEGF-stimulated HUVEC proliferation with an IC50 value of 0.4
nM [147], which is lower than what we determined. This discrepancy is probably due to different
culture conditions and the use of different assays to evaluate proliferation.
The device alone was established to have no adverse or beneficial effects on tumor
growth, edema formation, overall survival and body weight. Outcomes of the rats that were
given blank devices were essentially indistinguishable from those of the untreated control rats.
111
Local delivery using our device was demonstrated to be as effective as systemic delivery
in reducing edema for both DXM and cediranib. Both drugs were equally effective in alleviating
edema. Animals treated with either drug via either local or systemic administration experienced
significant decrease in edema compared with the control animals. There was no significant
difference among rats treated with DXM and cediranib. Importantly, edema reduction was
associated with survival benefit. Median survival showed a modest but significant improvement
from 14 to 16 days despite continuous tumor progression. The increase in tumor volume from
day 9 to day 12 observed in rats given DXM or cediranib was not significantly different from
that in controls, suggesting that DXM and cediranib had no effect on tumor growth. These
effects of DXM and cediranib on brain tumor and its associated edema mirror the observations in
other preclinical studies [106, 126].
Toxicity, together with efficacy, determines the success of a therapy. Treatment-related
adverse effects can worsen a patient's condition and lead to termination of therapy. Systemically
administered DXM has been shown to cause body weight loss in rodents [135-138]. It was found
to aggravate tumor-induced weight loss in our study. Local delivery appeared to eliminate the
toxicity that was observed with systemic delivery of DXM as weight loss in rats treated with
locally delivered DXM was significantly less than in rats given systemic DXM, but was not
statistically different from controls. The method of administration of cediranib, in contrast, had
no significant consequence on body weight in our study. This result is not surprising. No changes
in body weight or other signs of toxicity related to the systemic use of cediranib have been
reported in preclinical studies [126, 151, 152], although adverse effects of cediranib have been
observed in clinical trials [51, 153]. Systemic toxicity of cediranib might manifest in other ways.
Soft stools were present in all the rats that received cediranib by oral gavage but were absent in
112
all the other rats. It is uncertain if the soft stools were due to cediranib or polysorbate 80, the
vehicle used for oral gavage of cediranib.
Local DXM and cediranib doses delivered via our device were merely 2.36% and 0.21%
of their respective systemic doses (Table 5.1). The systemic doses used in our study were
translated from typical human doses (16 mg/day for DXM and 45 mg/day for cediranib) using
the body surface area normalization method [140J. Local drug delivery with our device offers the
benefit of using a small fraction of the dose used in systemic delivery to achieve similar efficacy
while minimizing the risk of toxicity associated with systemic delivery. Lower dose also makes
local delivery more cost-effective than traditional systemic routes of administration, especially
when it comes to delivering expensive therapeutic agents. Our device can be easily used with
other drugs without the need for intensive drug formulation. Local delivery of drugs targeting
brain tumor using our device is a promising approach to improve overall treatment outcome by
attaining efficacy without systemic toxicity.
Table 5.1 DXM and cediranib dose comparison. 'Dose converted to dexamethasone equivalent. 'Dose
converted to cediranib equivalent.
Total dose in 7
days/ pg
Dexamethasone
Systemic (DXM IP)
250 pg once a day'
1750+
Local (DXM Device)
sustained release, 5.9 pg/day
41.4
Cediranib
Systemic (AZD Oral)
690 pg once a day
4830
sustained release, 4.4 pIg on first
day, subsequent 0.74 pg/day'
10.2++
Local (AZD Device)
113
Comparison with
systemic dose (%)
-
Dose and Schedule
2.36
-
Treatment
0.21
5.5. CONCLUSIONS
A water-soluble solid dispersion of cediranib (AZD/PVP), which has similar bioactivity
as cediranib, was developed for diffusion-based drug delivery. Reproducible, sustained release of
both DXM and AZD/PVP from our novel, intracranial drug delivery device was achieved. Local
delivery of DXM and cediranib with our device was demonstrated to be as effective as systemic
delivery in alleviating edema. Edema reduction was associated with survival benefit, despite
continuous tumor progression. Local delivery of dexamethasone was shown to avoid
dexamethasone-related weight loss, a systemic adverse effect seen in animals treated with
systemic dexamethasone. Local drug administration using our device is superior to systemic
administration as it allows the use of a much lower dose to achieve similar efficacy with minimal
risk of systemic toxicity. Its use can be extended to other therapeutic agents targeting brain tumor
to achieve similar outcome.
114
References
[I] Papadopoulos MC, Saadoun S, Binder DK, Manley GT, Krishna S, Verkman AS. Molecular
mechanisms of brain tumor edema. Neuroscience. 2004;129:1011-20.
[2] Pope WB, Sayre J, Perlina A, Villablanca JP, Mischel PS, Cloughesy TF. MR imaging
correlates of survival in patients with high-grade gliomas. Am J Neuroradiol. 2005;26:2466-74.
[3] Wen PY, Kesari S. Malignant gliomas in adults. The New England journal of medicine.
2008;359:492-507.
[41 National Cancer Institute. PDQ@ Adult Brain Tumors Treatment. Bethesda, MD: National
Cancer Institute.
[5] Wen PY, Schiff D, Kesari S, Drappatz J, Gigas DC, Doherty L. Medical management of
patients with brain tumors. Journal of neuro-oncology. 2006;80:313-32.
[6] Baethmann A, Maierhauff K, Schurer L, Lange M, Guggenbichler C, Vogt W, et al.
RELEASE OF GLUTAMATE AND OF FREE FATTY-ACIDS IN VASOGENIC BRAIN
EDEMA. Journal of neurosurgery. 1989;70:578-91.
[7] Black KL, Hoff JT, McGillicuddy JE, Gebarski SS. INCREASED LEUKOTRIENE-C-4
AND VASOGENIC EDEMA SURROUNDING BRAIN-TUMORS IN HUMANS. Ann Neurol.
1986;19:592-5.
[8] Machein MR, Plate KH. VEGF in brain tumors. Journal of neuro-oncology. 2000;50:109-20.
[9] Carlson MRJ, Pope WB, Horvath S, Braunstein JG, Nghiemphu P, Tso CL, et al.
Relationship between survival and edema in malignant gliomas: Role of vascular endothelial
growth factor and neuronal pentraxin 2. Clinical Cancer Research. 2007;13:2592-8.
[10] Yano S, Shinohara H, Herbst RS, Kuniyasu H, Bucana CD, Ellis LM, et al. Expression of
Vascular Endothelial Growth Factor Is Necessary but not Sufficient for Production and Growth
of Brain Metastasis. Cancer research. 2000;60:4959-67.
[11] Schmidt NO, Westphal M, Hagel C, Ergun S, Stavrou D, Rosen EM, et al. Levels of
vascular endothelial growth factor, hepatocyte growth factor scatter factor and basic fibroblast
growth factor in human gliomas and their relation to angiogenesis. Int J Cancer. 1999;84:10-8.
[12] Strugar JG, Criscuolo GR, Rothbart D, Harrington WN. Vascular endothelial
growth/permeability factor expression in human glioma specimens: correlation with vasogenic
brain edema and tumor-associated cysts. Journal of neurosurgery. 1995;83:682-9.
115
[131 Dobrogowska DH, Lossinsky AS, Tarnawski M, Vorbrodt AW. Increased blood-brain
barrier permeability and endothelial abnormalities induced by vascular endothelial growth factor.
J Neurocytol. 1998;27:163-73.
[14] Koehler PJ. Use of Corticosteroids in Neurooncology. Anti-Cancer Drugs. 1995;6:19-33.
[15] Stummer W. Mechanisms of tumor-related brain edema. Neurosurg Focus. 2007;22:1-7.
[16] Kaal ECA, Vecht CJ. The management of brain edema in brain tumors. Current opinion in
oncology. 2004;16:593-600.
[17] Alberti E, Hartmann A, Schutz HJ, Schreckenberger F. EFFECT OF LARGE DOSES OF
DEXAMETHASONE ON CEREBROSPINAL-FLUID PRESSURE IN PATIENTS WITH
SUPRATENTORIAL TUMORS. Journal of neurology. 1978;217:173-81.
[18] Ostergaard L, Hochberg FH, Rabinov JD, Sorensen AG, Lev M, Kim L, et al. Early changes
measured by magnetic resonance imaging in cerebral blood flow, blood volume, and blood-brain
barrier permeability following dexamethasone treatment in patients with brain tumors. Journal of
neurosurgery. 1999;90:300-5.
[19] Jarden JO, Dhawan V, Moeller JR, Strother SC, Rottenberg DA. THE TIME COURSE OF
STEROID ACTION ON BLOOD-TO-BRAIN AND BLOOD-TO-TUMOR TRANSPORT OF
RB-82 - A POSITRON EMISSION TOMOGRAPHIC STUDY. Ann Neurol. 1989;25:239-45.
[20] McClelland S, 111, Long DM. Genesis of the use of corticosteroids in the treatment and
prevention of brain edema. Neurosurgery. 2008;62:965-7.
[21] Heiss JD, Papavassiliou E, Merrill MJ, Nieman L, Knightly JJ, Walbridge S, et al.
Mechanism of dexamethasone suppression of brain tumor-associated vascular permeability in
rats - Involvement of the glucocorticoid receptor and vascular permeability factor. Journal of
Clinical Investigation. 1996;98:1400-8.
[22] Barnes PJ. Molecular Mechanisms and Cellular Effects of Glucocorticosteroids. Immunol
Allerg Clin North Am. 2005;25:451-+.
[23] Kim H, Lee JM, Park JS, Jo SA, Kim YO, Kim CW, et al. Dexamethasone coordinately
regulates angiopoietin-1 and VEGF: A mechanism of glucocorticoid-induced stabilization of
blood-brain barrier. Biochem Biophys Res Commun. 2008;372:243-8.
[24] Vecht CJ, Hovestadt A, Verbiest HBC, Vanvliet JJ, Vanputten WLJ. Dose-Effect
Relationship of Dexamethasone on Karnofsky Performance in Metastatic Brain-Tumors - a
Randomized Study of Doses of 4, 8, and 16 mg Per Day. Neurology. 1994;44:675-80.
[25] Hempen C, Weiss E, Hess CF. Dexamethasone treatment in patients with brain metastases
and primary brain tumors: do the benefits outweigh the side-effects? Supportive Care in Cancer.
2002;10:322-8.
116
[26] Sturdza A, Millar B-A, Bana N, Laperriere N, Pond G, Wong RKS, et al. The use and
toxicity of steroids in the management of patients with brain metastases. Supportive Care in
Cancer. 2008;16:1041-8.
[27] Soffietti R, Ruda R, Mutani R. Management of brain metastases. Journal of neurology.
2002;249:1357-69.
[28] Roth P, Wick W, Weller M. Steroids in neurooncology: actions, indications, side-effects.
Current opinion in neurology. 2010;23:597-602.
[29] Bebawy JF. Perioperative Steroids for Peritumoral Intracranial Edema: A Review of
Mechanisms, Efficacy, and Side Effects. Journal of neurosurgical anesthesiology. 2012;24:1737.
[30] Dietrich J, Rao K, Pastorino S, Kesari S. Corticosteroids in brain cancer patients: benefits
and pitfalls. Expert review of clinical pharmacology. 2011;4:233-42.
[31] Ryken TC, McDermott M, Robinson PD, Ammirati M, Andrews DW, Asher AL, et al. The
role of steroids in the management of brain metastases: a systematic review and evidence-based
clinical practice guideline. Journal of neuro-oncology. 2010;96:103-14.
[32] Weissman DE, Janjan NA, Erickson B, Wilson FJ, Greenberg M, Ritch PS, et al. TwiceDaily Tapering Dexamethasone Treatment during Cranial Radiation for Newly Diagnosed Brain
Metastases. Journal of neuro-oncology. 1991;11:235-9.
[33] Marantidou A, Levy C, Duquesne A, Ursu R, Bailon 0, Coman I, et al. Steroid
requirements during radiotherapy for malignant gliomas. Journal of neuro-oncology.
2010;100:89-94.
[34] Jan Drappatz PYW. Management of vasogenic edema in patients with primary and
metastatic brain tumors. In: Posner JB, editor. UpToDate. Waltham, MA2013.
[35] Meixensberger J, Bendszus M, Licht K, Solymosi L, Roosen K. Peritumoural brain oedema
- Diagnosis and treatment approaches. Cns Drugs. 2000;13:233-51.
[36] Kesari S, Paleologos N, Vick N. Corticosteroids in Neuro-Oncology. In: Schiff D, Kesari S,
Wen P, editors. Cancer Neurology In Clinical Practice: Humana Press; 2008. p. 47-56.
[37] Tjuvajev J, Uehara H, Desai R, Beattie B, Matei C, Zhou Y, et al. Corticotropin-releasing
factor decreases vasogenic brain edema. Cancer research. 1996;56:1352-60.
[38] Villalona-Calero MA, Eckardt J, Burris H, Kraynak M, Fields-Jones S, Bazan C, et al. A
phase I trial of human corticotropin-releasing factor (hCRF) in patients with peritumoral brain
edema. Ann Oncol. 1998;9:71-7.
117
[39] Recht L, Mechtler LL, Wong ET, O'Connor PC, Rodda BE. Steroid-Sparing Effect of
Corticorelin Acetate in Peritumoral Cerebral Edema Is Associated With Improvement in Steroid-
Induced Myopathy. Journal of Clinical Oncology. 2013;31:1182-7.
[40] Badie B, Schartner JM, Hagar AR, Prabakaran S, Peebles TR, Bartley B, et al. Microglia
cyclooxygenase-2 activity in experimental gliomas: Possible role in cerebral edema formation.
Clinical Cancer Research. 2003;9:872-7.
[41] Portnow J, Suleman S, Grossman SA, Eller S, Carson K. A cyclooxygenase-2 (COX-2)
inhibitor compared with dexamethasone in a survival study of rats with intracerebral 9L
gliosarcomas. Neuro-oncology. 2002;4:22-5.
[42] Gerstner ER, Duda DG, di Tomaso E, Ryg PA, Loeffler JS, Sorensen AG, et al. VEGF
inhibitors in the treatment of cerebral edema in patients with brain cancer. Nature Reviews
Clinical Oncology. 2009;6:229-36.
[43] Norden AD, Drappatz J, Wen PY. Antiangiogenic therapies for high-grade glioma. Nat Rev
Neurol. 2009;5:610-20.
[44] Jo J, Schiff D, Purow B. Angiogenic inhibition in high-grade gliomas: past, present and
future. Expert Rev Neurother. 2012;12:733-47.
[45] Friedman HS, Prados MD, Wen PY, Mikkelsen T, Schiff D, Abrey LE, et al. Bevacizumab
Alone and in Combination With Irinotecan in Recurrent Glioblastoma. Journal of Clinical
Oncology. 2009;27:4733-40.
[46] Vredenburgh JJ, Cloughesy T, Samant M, Prados M, Wen PY, Mikkelsen T, et al.
Corticosteroid Use in Patients with Glioblastoma at First or Second Relapse Treated with
Bevacizumab in the BRAIN Study. Oncologist. 2010;15:1329-34.
[47] Kreisl TN, Kim L, Moore K, Duic P, Royce C, Stroud I, et al. Phase II Trial of Single-Agent
Bevacizumab Followed by Bevacizumab Plus Lrinotecan at Tumor Progression in Recurrent
Glioblastoma. Journal of Clinical Oncology. 2009;27:740-5.
[48] Pope WB, Lai A, Nghiemphu P, Mischel P, Cloughesy TF. MRI in patients with high-grade
gliomas treated with bevacizumab and chemotherapy. Neurology. 2006;66:1258-60.
[49] de Groot JF, Lamborn KR, Chang SM, Gilbert MR, Cloughesy TF, Aldape K, et al. Phase II
Study of Aflibercept in Recurrent Malignant Glioma: A North American Brain Tumor
Consortium Study. Journal of Clinical Oncology. 2011;29:2689-95.
[50] Batchelor TT, Sorensen AG, di Tomaso E, Zhang W-T, Duda DG, Cohen KS, et al.
AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and
alleviates edema in glioblastoma patients. Cancer Cell. 2007;11:83-95.
118
[511 Batchelor TT, Duda DG, di Tomaso E, Ancukiewicz M, Plotkin SR, Gerstner E, et al. Phase
H Study of Cediranib, an Oral Pan-Vascular Endothelial Growth Factor Receptor Tyrosine
Kinase Inhibitor, in Patients With Recurrent Glioblastoma. Journal of Clinical Oncology.
2010;28:2817-23.
[52] Conrad C, Friedman H, Reardon D, Provenzale J, Jackson E, Serajuddin H, et al. A phase
TAT trial of single-agent PTK 787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in
patients with recurrent glioblastoma multiforme (GBM). Journal of Clinical Oncology.
2004;22:11OS-S.
[53] Iwamoto FM, Lamborn KR, Robins HI, Mehta MP, Chang SM, Butowski NA, et al. Phase
H trial of pazopanib (GW786034), an oral multi-targeted angiogenesis inhibitor, for adults with
recurrent glioblastoma (North American Brain Tumor Consortium Study 06-02). Neurooncology. 2010;12:855-61.
[54] De Groot JF, Prados M, Urquhart T, Robertson S, Yaron Y, Sorensen AG, et al. A phase H
study of XL184 in patients (pts) with progressive glioblastoma multiforme (GBM) in first or
second relapse. Journal of Clinical Oncology. 2009;27.
[55] Armstrong TS, Wen PY, Gilbert MR, Schiff D. Management of treatment-associated
toxicites of anti-angiogenic therapy in patients with brain tumors. Neuro-oncology.
2012;14:1203-14.
[56] Pardridge WM. The blood-brain barrier bottleneck in brain drug development. NeuroRx:
the journal of the American Society for Experimental NeuroTherapeutics. 2005;2:3-14.
[571 Pardridge WM. CNS drug design based on principles of blood-brain barrier transport J
Neurochem. 1998;70:1781-92.
[58] Pardridge WM. Drug transport across the blood-brain barrier. J Cereb Blood Flow Metab.
2012;32:1959-72.
[591 Schinkel AH. P-glycoprotein, a gatekeeper in the blood-brain barrier. Advanced Drug
Delivery Reviews. 1999;36:179-94.
[601 Muldoon LL, Soussain C, Jahnke K, Johanson C, Siegal T, Smith QR, et al. Chemotherapy
delivery issues in central nervous system malignancy: A reality check. Journal of Clinical
Oncology. 2007;25:2295-305.
[611 Vogelbaum MA. Convection enhanced delivery for the treatment of malignant gliomas:
symposium review. Journal of neuro-oncology. 2005;73:57-69.
[62] Levin VA, Freemandove M, Landahl HD. PERMEABILITY CHARACTERISTICS OF
BRAIN ADJACENT TO TUMORS IN RATS. Archives of neurology. 1975;32:785-91.
119
[63] Blakeley JO, Olson J, Grossman SA, He X, Weingart J, Supko JG, et al. Effect of blood
brain barrier permeability in recurrent high grade gliomas on the intratumoral pharmacokinetics
of methotrexate: a microdialysis study. Journal of neuro-oncology. 2009;91:51-8.
[64] Donelli MG, Zucchetti M, D'Incalci M. Do anticancer agents reach the tumor target in the
human brain? Cancer chemotherapy and pharmacology. 1992;30:251-60.
[65] Groothuis DR. The blood-brain and blood-tumor barriers: A review of strategies for
increasing drug delivery. Neuro-oncology. 2000;2:45-59.
[66] Tomita T. Interstitial Chemotherapy for Brain-Tumors - Review. Journal of neuro-oncology.
1991;10:57-74.
[67] Walter KA, Tamargo RJ, Olivi A, Burger PC, Brem H. Intratumoral chemotherapy.
Neurosurgery. 1995;37:1128-45.
[68] Hochberg FH, Pruitt A. ASSUMPTIONS IN THE RADIOTHERAPY OF
GLIOBLASTOMA. Neurology. 1980;30:907-11.
[69] Reulen H-J. Bulk Flow and Diffusion Revisited, and Clinical Applications. In: Czernicki Z,
Baethmann A, Ito U, Katayama Y, Kuroiwa T, Mendelow D, editors. Brain Edema XIV:
Springer Vienna; 2010. p. 3-13.
[70] Voulgaris S, Partheni M, Karamouzis M, Dimopoulos P, Papadakis N, Kalofonos HP.
Intratumoral doxorubicin in patients with malignant brain gliomas. American Journal of Clinical
Oncology-Cancer Clinical Trials. 2002;25:60-4.
[71] Boiardi A, Eoli M, Salmaggi A, Zappacosta B, Fariselli L, Milanesi I, et al. Efficacy of
intratumoral delivery of mitoxantrone in recurrent malignant glial tumours. Journal of neurooncology. 2001;54:39-47.
[72] Boiardi A, Bartolomei M, Silvani A, Eoli M, Salmaggi A, Lamperti E, et al. Intratumoral
delivery of mitoxantrone in association with 90-Y radioimmunotherapy (RIT) in recurrent
glioblastoma. Journal of neuro-oncology. 2005;72:125-31.
[73] Nierenberg D, Harbaugh R, Maurer IH, Reeder T, Scott G, Fratkin J, et al. CONTINUOUS
INTRATUMORAL INFUSION OF METHOTREXATE FOR RECURRENT
GLIOBLASTOMA - A PILOT-STUDY. Neurosurgery. 1991;28:752-61.
[74] Patchell RA, Regine WF, Ashton P, Tibbs PA, Wilson D, Shappley D, et al. A phase I trial
of continuously infused intratumoral bleomycin for the treatment of recurrent glioblastoma
multiforme. Journal of neuro-oncology. 2002;60:37-42.
[75] Morantz RA, Kimler BF, Vats TS, Henderson SD. Bleomycin and brain tumors. A review.
Journal of neuro-oncology. 1983;1:249-55.
120
[76] Merchant RE, McVicar DW, Merchant LH, Young HF. TREATMENT OF RECURRENT
MALIGNANT GLIOMA BY REPEATED INTRACEREBRAL INJECTIONS OF HUMAN
RECOMBINANT INTERLEUKIN-2 ALONE OR IN COMBINATION WITH SYSTEMIC
INTERFERON-ALPHA - RESULTS OF A PHASE-I CLINICAL-TRIAL. Journal of neurooncology. 1992;12:75-83.
[77] Yoshida S, Tanaka R, Takai N, Ono K. LOCAL-ADMINISTRATION OF AUTOLOGOUS
LYMPHOKINE-ACTIVATED KILLER CELLS AND RECOMBINANT INTERLEUKIN-2
TO PATIENTS WITH MALIGNANT BRAIN-TUMORS. Cancer research. 1988;48:5011-6.
[78] Riva P, Franceschi G, Riva N, Casi M, Santimaria M, Adamo M. Role of nuclear medicine
in the treatment of malignant gliomas: the locoregional radioimmunotherapy approach. Eur J
Nucl Med. 2000;27:601-9.
[79] Bobo RH, Laske DW, Akbasak A, Morrison PF, Dedrick RL, Oldfield EH. Convectionenhanced delivery of macromolecules in the brain. Proceedings of the National Academy of
Sciences. 1994;91:2076-80.
[80] Raghavan R, Brady ML, Rodriguez-Ponce MI, Hartlep A, Pedain C, Sampson JH.
Convection-enhanced delivery of therapeutics for brain disease, and its optimization.
Neurosurgical FOCUS. 2006;20:E12.
[81] Lidar Z, Mardor Y, Jonas T, Pfeffer R, Faibel M, Nass D, et al. Convection-enhanced
delivery of paclitaxel for the treatment of recurrent malignant glioma: a Phase I/H clinical study.
Journal of neurosurgery. 2004;100:472-9.
[82] Tanner PG, Holtmannspotter M, Tonn JC, Goldbrunner R. Effects of drug efflux on
convection-enhanced paclitaxel delivery to malignant gliomas: Technical note. Neurosurgery.
2007;61:880-2.
[83] Laske DW, Youle RJ, Oldfield EH. Tumor regression with regional distribution of the
targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med. 1997;3:1362-8.
[84] Weaver M, Laske DW. Transferrin receptor ligand-targeted toxin conjugate (Tf-CRM107)
for therapy of malignant gliomas. Journal of neuro-oncology. 2003;65:3-13.
[85] Kunwar S, Prados MD, Chang SM, Berger MS, Lang FF, Piepmeier JM, et al. Direct
intracerebral delivery of cintredekin besudotox (IL13-PE38QQR) in recurrent malignant glioma:
A report by the Cintredekin Besudotox Intraparenchymal Study Group. Journal of Clinical
Oncology. 2007;25:837-44.
[86] Sampson JH, Archer G, Pedain C, Wembacher-Schroder E, Westphal M, Kunwar S, et al.
Poor drug distribution as a possible explanation for the results of the PRECISE trial Clinical
article. Journal of neurosurgery. 2010;113:301-9.
121
[87] Patel SJ, Shapiro WR, Laske DW, Jensen RL, Asher AL, Wessels BW, et al. Safety and
feasibility of convection-enhanced delivery of cotara for the treatment of malignant glioma:
Initial experience in 51 patients. Neurosurgery. 2005;56:1243-52.
[88] Weber F, Asher A, Bucholz R, Berger M, Prados M, Chang S, et al. Safety, tolerability, and
tumor response of IL4-Pseudomonas exotoxin (NBI-3001) in patients with recurrent malignant
glioma. Journal of neuro-oncology. 2003;64:125-37.
[891 Sampson JH, Akabani G, Archer GE, Berger MS, Coleman RE, Friedman AH, et al.
Intracerebral infusion of an EGFR-targeted toxin in recurrent malignant brain tumors. Neurooncology. 2008;10:320-9.
[90] Allard E, Passirani C, Benoit JP. Convection-enhanced delivery of nanocarriers for the
treatment of brain tumors. Biomaterials. 2009;30:2302-18.
[91] Bidros DS, Liu JK, Vogelbaum MA. Future of convection-enhanced delivery in the
treatment of brain tumors. Future Oncology. 2010;6:117-25.
[92] Schmidt NO, Ziu M, Carrabba G, Giussani C, Bello L, Sun Y, et al. Antiangiogenic
Therapy by Local Intracerebral Microinfusion Improves Treatment Efficiency and Survival in an
Orthotopic Human Glioblastoma Model. Clinical Cancer Research. 2004;10: 1255-62.
[93] Giussani C, Carrabba G, Pluderi M, Lucini V, Pannacci M, Caronzolo D, et al. Local
intracerebral delivery of endogenous inhibitors by osmotic minipumps effectively suppresses
glioma growth in vivo. Cancer research. 2003;63:2499-505.
[94] Zhu H, Woolfenden S, Bronson RT, Jaffer ZM, Barluenga S, Winssinger N, et al. The
Novel Hsp90 Inhibitor NXD30001 Induces Tumor Regression in a Genetically Engineered
Mouse Model of Glioblastoma Multiforme. Molecular Cancer Therapeutics. 2010;9:2618-26.
[95] Ikeda Y, Carson BS, Lauer JA, Long DM. Therapeutic Effects of Local-Delivery of
Dexamethasone on Experimental Brain-Tumors and Peritumoral Brain Edema. Journal of
neurosurgery. 1993;79:716-21.
[96] Saltzman WM. Drug Delivery - Engineering Principles for Drug Therapy. Oxford
University Press.
[97] Guerin C, Olivi A, Weingart JD, Lawson HC, Brem H. Recent advances in brain tumor
therapy: local intracerebral drug delivery by polymers. Investigational new drugs. 2004;22:2737.
[98] Brem H, Piantadosi S, Burger PC, Walker M, Selker R, Vick NA, et al. Placebo-Controlled
Trial of Safety and Efficacy of Intraoperative Controlled Delivery by Biodegradable Polymers of
Chemotherapy for Recurrent Gliomas. Lancet. 1995;345:1008-12.
122
[99] Westphal M, Hilt DC, Bortey E, Delavault P, Olivares R, Warnke PC, et al. A phase 3 trial
of local chemotherapy with biodegradable carmustine (BCNU) wafers (Gliadel wafers) in
patients with primary malignant glioma. Neuro-oncology. 2003;5:79-88.
[100] Leong KW, Brott BC, Langer R. Bioerodible Polyanhydrides as Drug-Carrier Matrices .1.
Characterization, Degradation, and Release Characteristics. Journal of Biomedical Materials
Research. 1985;19:941-55.
[101] Leong KW, Damore P, Marletta M, Langer R. Bioerodible Polyanhydrides as Drug-Carrier
Matrices .2. Biocompatibility and Chemical-Reactivity. Journal of Biomedical Materials
Research. 1986;20:51-64.
[102] LESNIAK MS, UPADHYAY U, GOODWIN R, TYLER B, BREM H. Local Delivery of
Doxorubicin for the Treatment of Malignant Brain Tumors in Rats. Anticancer research.
2005;25:3825-31.
[1031 Walter KA, Cahan MA, Gur A, Tyler B, Hilton J, Colvin OM, et al. Interstitial Taxol
Delivered from a Biodegradable Polymer Implant against Experimental Malignant Glioma.
Cancer research. 1994;54:2207-12.
[104] Brem S, Tyler B, Li K, Pradilla G, Legnani F, Caplan J, et al. Local delivery of
temozolomide by biodegradable polymers is superior to oral administration in a rodent glioma
model. Cancer chemotherapy and pharmacology. 2007;60:643-50.
[105] Tamargo RJ, Myseros JS, Epstein JI, Yang MB, Chasin M, Brem H. INTERSTITIAL
CHEMOTHERAPY OF THE 9L-GLIOSARCOMA - CONTROLLED RELEASE POLYMERS
FOR DRUG DELIVERY IN THE BRAIN. Cancer research. 1993;53:329-33.
[106] Tamargo RJ, Sills AK, Reinhard CS, Pinn ML, Long DM, Brem H. Interstitial Delivery of
Dexamethasone in the Brain for the Reduction of Peritumoral Edema. Journal of neurosurgery.
1991;74:956-61.
[107] Scott AW. Microcapsule drug delivery device for treatment of glioblastoma multiforme
2010.
[108] Wilkinson HA, Komblith P, Weems S. FOCAL CHEMOTHERAPY OF BRAIN
TUMORS USING SEMIPERMEABLE MEMBRANES. J Neurol Neurosurg Psychiatry.
1977;40:389-94.
[109] Rosenblu.MI, Bowie DL, Walker MD. DIFFUSION IN-VITRO AND IN-VIVO OF 1-(2CHLOROETHYL)-3-(TRANS-4-METHYLCYCLOHEXYL)-1-NITROSOUREA FROM
SILICONE-RUBBER CAPSULES, A POTENTIALLY NEW MODE OF CHEMOTHERAPY
ADMINISTRATION. Cancer research. 1973;33:906-14.
123
[110] Fung LK, Ewend MG, Sills A, Sipos EP, Thompson R, Watts M, et al. Pharmacokinetics
of interstitial delivery of carmustine, 4-hydroperoxycyclophosphamide, and paclitaxel from a
biodegradable polymer implant in the monkey brain. Cancer research. 1998;58:672-84.
[111] Fung LK, Shin M, Tyler B, Brem H, Saltzman WM. Chemotherapeutic drugs released
from polymers: Distribution of 1,3-bis(2-chloroethyl)-1-nitrosourea in the rat brain. Pharm Res.
1996;13:671-82.
[112] Pardridge WM. Drug delivery to the brain. J Cereb Blood Flow Metab. 1997;17:713-31.
[113] Buonerba C, Di Lorenzo G, Marinelli A, Federico P, Palmieri G, Imbimbo M, et al. A
comprehensive outlook on intracerebral therapy of malignant gliomas. Crit Rev Oncol/Hematol.
2011;80:54-68.
[114] Lesniak MS, Brem H. Targeted therapy for brain tumours. Nature Reviews Drug
Discovery. 2004;3:499-508.
[115] Sipos EP, Tyler B, Piantadosi S, Burger PC, Brem H. Optimizing interstitial delivery of
BCNU from controlled release polymers for the treatment of brain tumors. Cancer chemotherapy
and pharmacology. 1997;39:383-9.
[116] Olivi A, Grossman SA, Tatter S, Barker F, Judy K, Olsen J, et al. Dose escalation of
carmustine in surgically implanted polymers in patients with recurrent malignant glioma: A new
approaches to Brain Tumor Therapy CNS Consortium trial. Journal of Clinical Oncology.
2003;21:1845-9.
[1171 Barth RF, Kaur B. Rat brain tumor models in experimental neuro-oncology: the C6, 9L,
T9, RG2, F98, BT4C, RT-2 and CNS-1 gliomas. Journal of neuro-oncology. 2009;94:299-312.
[118] Assi H, Candolfi M, Lowenstein P, Castro M. Rodent Glioma Models: Intracranial
Stereotactic Allografts and Xenografts. In: Martfnez Murillo R, Martinez A, editors. Animal
Models of Brain Tumors: Humana Press; 2013. p. 229-43.
[119] Fomchenko EI, Holland EC. Mouse models of brain tumors and their applications in
preclinical trials. Clinical Cancer Research. 2006;12:5288-97.
-
[120] Guerin C, Wolff JEA, Laterra J, Drewes LR, Brem H, Goldstein GW. VASCULAR
DIFFERENTIATION AND GLUCOSE TRANSPORTER EXPRESSION IN RAT GLIOMAS
EFFECTS OF STEROIDS. Ann Neurol. 1992;31:481-7.
[121] Badruddoja MA, Krouwer HGJ, Rand SD, Rebro KJ, Pathak AP, Schmainda KM.
Antiangiogenic effects of dexamethasone in 9L gliosarcoma assessed by MRI cerebral blood
volume maps. Neuro-oncology. 2003;5:235-43.
124
[122] Reichman HR, Farrell CL, Delmaestro RF. Effects of Steroids and Nonsteroid
Antiinflammatory Agents on Vascular-Permeability in a Rat Glioma Model. Journal of
neurosurgery. 1986;65:233-7.
[123] Hossmann KA, Hurter T, Oschlies U. The Effect of Dexamethasone on Serum-Protein
Extravasation and Edema Development in Experimental Brain-Tumors of Cat. Acta
Neuropathologica. 1983;60:223-31.
[124] Jamshidi J, Yoshimine T, Ushio Y, Hayakawa T. Effects of Glucocorticoid and
Chemotherapy on the Peritumoral Edema and Astrocytic Reaction in Experimental Brain-Tumor.
Journal of neuro-oncology. 1992;12:197-204.
[125] Yamada K, Ushio Y, Hayakawa T, Arita N, Yamada N, Mogami H. Effects of
Methylprednisolone on Peritumoral Brain Edema - a Quantitative Autoradiographic Study.
Journal of neurosurgery. 1983;59:612-9.
[126] Kamoun WS, Ley CD, Farrar CT, Duyverman AM, Lahdenranta J, Lacorre DA, et al.
Edema Control by Cediranib, a Vascular Endothelial Growth Factor Receptor-Targeted Kinase
Inhibitor, Prolongs Survival Despite Persistent Brain Tumor Growth in Mice. Journal of Clinical
Oncology. 2009;27:2542-52.
[127] de Vries NA, Bruggeman SW, Hulsman D, de Vries HI, Zevenhoven J, Buckle T, et al.
Rapid and Robust Transgenic High-Grade Glioma Mouse Models for Therapy Intervention
Studies. Clinical Cancer Research. 2010;16:3431-41.
[128] Zhu H, Acquaviva J, Ramachandran P, Boskovitz A, Woolfenden S, Pfannl R, et al.
Oncogenic EGFR signaling cooperates with loss of tumor suppressor gene functions in
gliomagenesis. Proceedings of the National Academy of Sciences of the United States of
America. 2009; 106:2712-6.
[129] Woolfenden S, Zhu H, Charest A. A Cre/LoxP Conditional Luciferase Reporter
Transgenic Mouse for Bioluminescence Monitoring of Tumorigenesis. Genesis. 2009;47:659-66.
[130] Reinhard CS, Radomsky ML, Saltzman WM, Hilton J, Brem H. Polymeric Controlled
Release of Dexamethasone in Normal Rat-Brain. Journal of Controlled Release. 1991;16:331-40.
[131] Lamiable D, Vistelle R, Millart H, Sulmont V, Fay R, Caron J, et al. HIGHPERFORMANCE LIQUID-CHROMATOGRAPHIC DETERMINATION OF
DEXAMETHASONE IN HUMAN-PLASMA. Journal of Chromatography. 1986;378:486-91.
[132] Higuchi T. Rate of release of medicaments from ointment bases containing drugs in
suspension. Journal of Pharmaceutical Sciences. 1961;50:874-5.
[133] Siepmann J, Peppas NA. Modeling of drug release from delivery systems based on
hydroxypropyl methylcellulose (HPMC). Advanced Drug Delivery Reviews. 2001;48:139-57.
125
[134]
Paul DR. Elaborations on the Higuchi model for drug delivery. Int J Pharm. 2011;418:13-
7.
[135] Tonolo G, Fraser R, Connell JMC, Kenyon CJ. Chronic Low-Dose Infusions of
Dexamethasone in Rats - Effects on Blood-Pressure, Body-Weight and Plasma Atrial Natriuretic
Peptide. Journal of hypertension. 1988;6:25-31.
[136] Ma K, Mallidis C, Bhasin S, Mahabadi V, Artaza J, Gonzalez-Cadavid N, et al.
Glucocorticoid-induced skeletal muscle atrophy is associated with upregulation of myostatin
gene expression. American Journal of Physiology-Endocrinology and Metabolism.
2003;285:E363-E71.
[137] Rhee MS, Perianayagam A, Chen P, Youn JH, McDonough AA. Dexamethasone treatment
causes resistance to insulin-stimulated cellular potassium uptake in the rat. American Journal of
Physiology-Cell Physiology. 2004;287:C1229-C37.
[138] Gilson H, Schakman 0, Combaret L, Lause P, Grobet L, Attaix D, et al. Myostatin gene
deletion prevents glucocorticoid-induced muscle atrophy. Endocrinology. 2007;148:452-60.
-
[139] Griffin AC, Whitacre CC. SEX AND STRAIN DIFFERENCES IN THE CIRCADIANRHYTHM FLUCTUATION OF ENDOCRINE AND IMMUNE FUNCTION IN THE RAT
IMPLICATIONS FOR RODENT MODELS OF AUTOIMMUNE-DISEASE. J Neuroimmunol.
1991;35:53-64.
[140] Reagan-Shaw S, Nihal M, Ahmad N. Dose translation from animal to human studies
revisited. Faseb Journal. 2008;22:659-61.
[141] Wang T, Agarwal S, Elmquist WF. Brain Distribution of Cediranib Is Limited by Active
Efflux at the Blood-Brain Barrier. Journal of Pharmacology and Experimental Therapeutics.
2012;341:386-95.
[142] Sethia S, Squillante E. Physicochemical characterization of solid dispersions of
carbamazepine formulated by supercritical carbon dioxide and conventional solvent evaporation
method. Journal of Pharmaceutical Sciences. 2002;91:1948-57.
[143] Wang TL, Oberoi RK, Elmquist WF. Determination of cediranib in mouse plasma and
brain tissue using high-performance liquid chromatography-mass spectrometry. J Chromatogr B.
2011;879:3812-7.
[144] Das A, Banik NL, Patel SJ, Ray SK. Dexamethasone protected human glioblastoma
U87MG cells from temozolomide induced apoptosis by maintaining Bax:Bcl-2 ratio and
preventing proteolytic activities. Mol Cancer. 2004;3.
[145] Sur P, Sribnick EA, Patel SJ, Ray SK, Banik NL. Dexamethasone decreases
temozolomide-induced apoptosis in human gliobastoma T98G cells. Glia. 2005;50:160-7.
126
[1461 Wolff JEA, Denecke J, Jurgens H. Dexamethasone induces partial resistance to
cisplatinum in C6 glioma cells. Anticancer research. 1996;16:805-9.
[147] Wedge SR, Kendrew J, Hennequin LF, Valentine PJ, Barry ST, Brave SR, et al. AZD2171:
A highly potent, orally bioavailable, vascular endothelial growth factor receptor-2 tyrosine
kinase inhibitor for the treatment of cancer. Cancer research. 2005;65:4389-400.
[148] Lin YS, Nguyen C, Mendoza JL, Escandon E, Fei D, Meng YG, et al. Preclinical
pharmacokinetics, interspecies scaling, and tissue distribution of a humanized monoclonal
antibody against vascular endothelial growth factor. Journal of Pharmacology and Experimental
Therapeutics. 1999;288:371-8.
[149] Martinho 0, Silva-Oliveira R, Miranda-Goncalves V, Clara C, Almeida JR, Carvalho AL,
et al. In Vitro and In Vivo Analysis of RTK Inhibitor Efficacy and Identification of Its Novel
Targets in Glioblastomas. Transl Oncol. 2013;6:187-U264.
[150] Wachsberger PR, Lawrence RY, Liu Y, Xia X, Andersen B, Dicker AP. Cediranib
enhances control of wild type EGFR and EGFRvIII-expressing gliomas through potentiating
temozolomide, but not through radiosensitization: implications for the clinic. Journal of neurooncology. 2011;105:181-90.
[1511 Gomez-Rivera F, Santillan-Gomez AA, Younes MN, Kim S, Fooshee D, Zhao M, et al.
The Tyrosine Kinase Inhibitor, AZD2171, Inhibits Vascular Endothelial Growth Factor Receptor
Signaling and Growth of Anaplastic Thyroid Cancer in an Orthotopic Nude Mouse Model.
Clinical Cancer Research. 2007;13:4519-27.
[152] MEDINGER M, ESSER N, ZIRRGIEBEL U, RYAN A, JORGENSMEIER JM, DREVS
J. Antitumor and Antiangiogenic Activity of Cediranib in a Preclinical Model of Renal Cell
Carcinoma. Anticancer research. 2009;29:5065-76.
[153] Drevs J, Siegert P, Medinger M, Mross K, Strecker R, Zirrgiebel U, et al. Phase I Clinical
Study of AZD2171, an Oral Vascular Endothelial Growth Factor Signaling Inhibitor, in Patients
With Advanced Solid Tumors. Journal of Clinical Oncology. 2007;25:3045-54.
127