 |
|
Laboratory Research
Howard A. Fine, M.D., Lab and Branch Chief
The overall goal of the Neuro-Oncology Laboratory is to develop novel therapeutic
strategies for the treatment of gliomas through an understanding, exploitation,
and eventual clinical translation of the principles underlying the molecular and
genetic pathogenesis of these tumors. Our approach is to leverage the unique
resources of the intramural NIH program, including its tremendous scientific and
clinical freedom to explore high risk yet high pay off projects, to build an
NIH-wide pre-clinical and clinical brain tumor experimental therapeutics center.
This Center works collaboratively and synergistically with both the NIH extramural
community as well as with the private sector to ensure the most efficient and rapid
development of novel approaches to the treatment of these devastating tumors.
Below is a brief summary of several of our current translational initiatives.
-
Glioma Molecular Diagnostic Initiative:
Although it is well recognized that human gliomas are
a heterogeneous group of tumors, there are to date no
pathologic classification schemas that reproducibly allow
the identification of biologically similar tumors or predict
for tumor-specific therapies. The lack of such a classification
schema significantly limits the ability of scientists
to unravel the molecular pathogenesis of different glioma
subtypes and precludes clinicians from offering therapeutic
options that are specific for a patient's particular tumor.
We have therefore initiated a large national effort in
collaboration with the Cancer
Genome Anatomy Project (CGAP) and NCI's Cancer
Therapy Evaluation Program (CTEP) to develop a comprehensive
and novel molecular classification schema for human gliomas
based on gene expression profiles using cDNA microarray
technology, comparative genomic hybridization (CGH), SNP
analyses, and high throughput sequencing. Based on cDNA
and SAGE libraries from more then 50 human gliomas, we
have constructed a 48,000 element cDNA/oligonucleotide
microarray chip enriched for genes that appear to be important
in glioma biology. In conjunction with the NCI's CTEP-sponsored
national Brain Tumor Consortia, we will receive hundreds
of tumor specimens and correlative, prospective corollary
clinical data. The data generated from this prospective
study will be assimilated into the molecular/clinical
database we are currently generating from the nearly 500
glioma specimens and historical clinical data we have
received from our extramural collaborators. It is our
expectation that within 5 years we will define a new molecular
classification schema for human gliomas that has both
prognostic and predictive therapeutic utility. Additionally,
this project will generate a public international molecular
and clinical database offering an unprecedented opportunity
for gene discovery, elucidation of signal transduction
pathways, and molecular target identification and validation
that will be available to investigators through out the
world. Finally, these data will be the first data to populate
an exciting new initiative known as the Cancer
Molecular Analysis Program (CMAP), a web-based bio-informatics
platform that will allow researchers and clinicians the
ability identify and evaluate molecular targets in cancer
through integration of basic and clinical cancer research
programs.
Back to Top
-
Bone Marrow-Derived Neural Stem Cells
(NSC):
NSC hold great promise for the treatment of a number of
serious neurological illnesses including brain tumors.
Nevertheless, the ability to obtain sufficient numbers
of NSC from adult tissue represents a significant technical
challenge. Although embryonic stem (ES) cells represent
a potential source of cells with NSC-like properties,
therapeutic use of ES cells are limited by questions regarding
tumorigenic potential, immunological tolerance, and ethical
concerns. We have now demonstrated that a subpopulation
of adult human bone marrow-derived cells that can be propagated
in large numbers in vitro, have morphology, gene expression
profiles, and migratory properties similar to human fetal
brain-derived NSC. These bone marrow-derived NSC (BMNSC)
can be induced to differentiate into all of the primary
lineages that make up the central nervous system including
astrocytes, oligodendrocytes, and different neuronal subtypes.
Additionally, our microarray gene expression studies have
allowed us to identify a set of 73 genes that appear to
play a pivotal role in the differentiation of both bone
marrow and brain-derived NSC toward the differentiated
neural lineages. Finally, we have demonstrated the ability
of BMNSC to migrate both to sites of tumor cell infiltration
within the brain and to sites of neural tissue damage.
Our data demonstrate that through their migratory capabilities,
these BMNSC transduced with therapeutic genes can mediate
brain tissue repair and brain tumor destruction. We have
a number of ongoing studies investigating issues related
to the therapeutic use of these cells for anti-tumor purposes,
for neural damage repair, and for better understanding
the molecular biology and genetics of these cells. In
particular, we are investigating how the set of genes
we have identified as being pivotal in NSC differentiation
are affected by specific mutations found within human
gliomas. On the clinical side, we will soon be initiating
the first clinical trial of these cells in which bone
marrow will be harvested from patients will malignant
brain tumors who are scheduled for surgery. BMNSC will
be generated from their marrows, genetically marked and
re-administered to the patient prior to surgery. Following
surgery, the efficiency of migration and transgene expression
of the genetically marked BMNSC will be evaluated by examining
the resected tumor specimen. This study will hopefully
establish the groundwork for a series of clinical studies
designed to evaluate the ability of these cells to mediate
brain tumor destruction and/or repair of damaged neural
tissue.
Back to Top
-
Glioma-Selective Gene Therapy:
Although therapeutic gene transfer offers a potentially
promising new therapeutic strategy for the treatment of
malignant gliomas, its clinical application has been limited
by issues related to the non-tumor selectivity and distribution
inefficiencies of current genetic delivery vector systems.
We have addressed the problem of tumor selectivity in
two different ways. First, we have developed a technology
that allows us to exploit the observation that nearly
100% of malignant gliomas have deregulated retinoblastoma
protein (rb) function secondary to genetic or epigenetic
mutations in the P16/cyclinD/CDK4/RB pathway. Thus, a
prediction, that until now, had not been demonstrated
experimentally in vivo, is that E2F responsive promoters
should be more active in tumor cells relative to normal
cells due to an excess of 'free' E2F and loss of pRB/E2F
repressor complexes. We have demonstrated that adenoviral
and adenoviral associated viral (AAV)-based vectors, containing
transgenes driven by the E2F-1 promoter, can mediate tumor
selective gene expression in vivo, allowing for eradication
of established gliomas with significantly less normal
tissue toxicity than seen with standard viral vectors.
In a second strategy to achieve glioma-selective vector
transduction we, in collaboration with Robert Kotin of
NHLBI, have recently demonstrated that a newly cloned
and never before utilized subtype of AAV (AAV-5), allows
for highly selective glioma cell transduction in vitro
and in vivo secondary to the expression of the AAV-5 co-receptor
(a specific 2,3-linked sialic acid glycopolysaccharide),
on glioma cells but not only normal cells within the central
nervous system. Finally, in collaboration with Ed Oldfield
and the Surgical Neurology Branch of the NIH, we have
addressed the problem of inefficient vector delivery by
demonstrating that a new delivery technology developed
here at the NIH, known as enhanced convection delivery,
can distribute particles as large as adenoviral virions
uniformly, at high concentrations, and safely throughout
a region of the brain as large as the total cerebral hemisphere.
All these studies have led us to the verge of initiating
a series of clinical studies that will be designed to
test the safety, vector distribution, glioma selective
transgene expression, and eventual efficacy of E2F-responsive
AAV-5 based vectors administered by direct intracerebral
enhanced convection in patients with malignant gliomas.
Back to Top
-
Endothelial Progenitor Cells:
We have identified a methodology for isolating and expanding
in vitro a sub-population of human hematopoietic cells
that are in fact endothelial progenitor cells (EPCs) or
angioblasts. We have demonstrated that we can genetically
engineer these cells ex-vivo to express marker genes or
the thymidine kinase (TK) gene using retrovirus-mediated
gene transfer. Genetically labeled EPCs were transplanted
into wild type and sub-lethally irradiated mice and found
to migrate and incorporate into the angiogenic vasculature
of growing tumors while maintaining transgene expression.
Ganciclovir (GCV) treatment resulted in tumor vascular
collapse with resultant tumor necrosis in animals previously
administered TK-expressing EPCs. These results demonstrate
the feasibility of utilizing genetically modified EPCs
as angiogenesis-selective gene-targeting vectors and demonstrate
the potential of this approach to mediate non-toxic and
systemic anti-tumor responses. Additionally, our EPC marker
experiments have demonstrated that as much as 15-30% of
tumor vasculature may in fact be bone marrow-derived.
If validated, these experiments suggests a paradigm shift
in that tumor-associated neovasculature may not be derived
just through angiogenic mechanisms as previously believed,
but also through the embryonic process of vasculogenesis.
We are exploring a number of in vitro and in vivo experiments
to better understand this process and to elucidate the
cellular and molecular biology of the EPCs, including
our recent demonstration that these cells appear to contribute
to the repair of the endothelial component of the blood-brain
barrier. Finally, we are preparing a pilot clinical study
where EPCs will be harvested from patients, propagated
and genetically marked ex vivo. These autologous cells
will then be re-administered back into the systemic circulation
from the donor patients with recurrent malignant gliomas,
scheduled for repeat tumor resection. We will evaluate
the resected tumor for incorporation of the genetically
marked EPCs into tumor vascular and for evidence of transgene
expression. This initial marker study will hopefully provide
the biological basis that will pave the way for similar
studies using EPCs carrying therapeutic genes.
Back to Top
-
Molecular Therapeutics:
The Neuro-Oncology Laboratories are dedicated to the development
of novel molecularly targeted anti-tumor agents. Toward
this end there are currently two major programs:
-
Selectively Targeted Ubiquination. We refined
a technology that allows us to select a cellular protein
(i.e. cell cycle regulator, oncogene, anti-apoptotic
gene) and selectively target it for ubiquination and
proteosome-mediated destruction. We have an active
program that utilizes viral vectors to transduce a
molecule that directs specific components of the cell
cycle into proteosome degradation initiating rapid
glioma cell apoptosis in vitro and tumor destruction
in vivo without causing any toxicity to normal cells.
We are currently working to bring this technology
to the clinic while developing similar strategies,
targeting other novel proteins, in the laboratory.
-
Phase Display: We have an active program to
identify novel molecular targets of 'signatures" on
both glioma cell membranes as well as on tumor-associated
endothelium utilizing random phage technology. We
have already identified a number of exciting candidates
targets and are currently confirming their tumor and/or
tumor endothelial specificity. We will be linking
these novel peptides to positron emitting nucleotides
for PET scan imaging as a novel diagnostic modality
and covalently binding these peptides to both cytotoxic
agents and high energy radionucleotides as a tumor
selective targeting technology.
Back to Top
-
Animal Brain Tumor Experimental Therapeutics
and Diagnostic Core:
One of the newest initiatives of the Neuro-Oncology Branch
is an animal brain tumor experimental therapeutics and
diagnostic core. The impetus for this initiative grows
out of realization that most academic centers and pharmaceutical/biotechnology
companies do not routinely evaluate new agents or diagnostic
modalities in brain tumor cell lines, or in vivo models.
The reasons for this are many, but include the technical
difficulties in establishing reliable orthotopic models,
and the relatively poor brain tumor models that were historically
available. This has resulted in inadequate pre-clinical
data for properly screening and selecting appropriate
agents to bring to clinical trials for brain tumor patients,
and for inadequate pharmacokinetic/toxicity data for designing
the optimal initial trials. For example, as locally infiltrative,
but non-metastasizing tumors, alternate methods of drug
delivery may be preferable to standard intravenous administration,
particular given issues related to the blood-brain barrier.
One such delivery technology known as enhanced convection
diffusion was developed here at the NIH and represents
a way of homogeneously delivering an agent throughout
the brain by direct infusion into the brain parenchyma.
There are few if any pharmaceutical companies, however,
that have gone to the effort to set up the methodologies
that could test whether convection delivery is a better
(more efficient, less toxic) method for delivering a novel
anti-tumor agents to a brain tumor. Furthermore, the lack
of thoughtful pre-clinical testing and modeling has made
the development of surrogate markers of the biologic activity
of any given drug virtual impossible; something particularly
pertinent for the clinical development of new classes
of drugs with cytostatic (rather then cytotoxic) properties
(i.e. differentiating and anti-angiogenic agents).
It is, therefore, our belief that this brain tumor therapeutics
core will be invaluable for the purpose of designing more
rationale methods of drug selection, for generating new
clinically useful surrogate markers of drug activity,
and for establishing better clinical study designs. This
core is available for testing new therapeautic agents
through collabarations with investigators both within
the intramural and extramural NCI program, investigators
at other academic medical institutions, and within the
private sector through special arrangements.
The experimental brain tumor therapeutics core consists
of five major components:
-
New Cytotoxic, Cytostatic, and Radiation Sensitizing
Agent Evaluation: This consists of the testing
of these new agents in vitro against a number of different
adult and pediatric glioma, primitive neuro-ectodermal,
neuronal tumor cell lines. The agents would be further
testing in orthotopic brain tumor models using stereotactic
technologies and evaluating a number of newer delivery
technologies (besides standard oral or intravenous
administration) including intracarotid administration,
delivery with or without selective or gross blood
brain barrier disruption, convection delivery, etc.
The most promising of these agents will then have
extensive pharmacology performed on them in animals
pretreated with anti-epileptic agents know to induce
the cytochrome P450 system in order to evaluate whether
the brain tumor population is likely to experience
altered pharmacokinetics of the agents compared to
patients with systemic tumors.
-
Central Nervous System Pharmacology (Blood-Brain
Barrier): In conjunction with exploring alternate
delivery modalities, the core has the capability of
performing extensive pharmacology with an emphasis
on assessing the ability of an agent to cross both
an intact and a compromised blood-brain barrier. Along
with cerebral spinal fluid pharmacology, the core
has the ability to study brain/spinal cord drug penetration
using techniques such as microdialysis and autoradiography.
Finally, in collaboration with Frank Balis in the
Pediatric Oncology Branch (NCI), the core has the
capability of performing central nervous system pharmacology,
including microdialysis, in non-human primates for
agents that demand such an evaluation.
-
Neurotoxicity: Evaluation of neurotoxicity
of anticancer agents alone or in combination (with
other agents and/or radiation) is rarely if ever undertaken
as part of the pre-clinical development of any new
agent. Neurotoxicity, particularly as it relates to
the developing central nervous system (i.e. pediatric
patients), however, is of vital importance for the
ultimate clinical utility of any new anticancer agent.
Thus, in conjunction with experts at NINDS and NIMH,
a series of behavior, radiographic, and pathologic
screens for therapy-induced neurotoxicity will be
developed and used as part of a standard screen for
agents that appear most promising for clinical development.
Information gained through these screens will hopefully
be presented to intramural and extramural investigators
who might be interested in exploring the mechanism
of injury in more depth. Having a routine animal screen
in place, will be a great incentive for the future
development of neuroprotective agents through both
academia and the pharmaceutical industry.
-
Evaluation of Novel Endpoints: As mentioned
above, many of the new classes of anti-tumor therapeutics
will have cytostatic rather then cytotoxic properties.
Evaluating which of these agents will have biologic
activity in humans in small, early clinical trials
is a challenge since the standard "response" criteria
are based on the determination of cytotoxic responses
(i.e. the tumor shrinks by 50%). The only truly valid
clinical parameter available for evaluating the activity
of a truly cytostatic agent is patient survival or
tumor progression-free survival. These, however, are
not useful parameters for screening drug activity
in small, early phase clinical trials. Thus, if surrogate
markers of biologic activity could be identified,
one could utilize these as early endpoints for screening
out agents with little or no activity in vivo. Toward
that end, the experimental brain tumor therapeutics
core will actively develop surrogate markers of drug
anti-tumor activity that can be utilized and validated
in clinical trials. There are three major areas that
well be explored:
-
Imaging: In collaboration with investigators
in NCI and NINDS, we will take advantage of the
enormous small animal imaging facilities that
are currently being constructed. Examples of the
kind of endpoints we are interested in developing
include techniques to quantitate microvasculature,
vascular permeability, cerebral edema, and tumor
metabolism. Currently we have the capabilities
of doing high resolution MRI, MR spectroscopy,
and PET scanning on small animals.
-
Gene Expression Profiling: A major effort
of the core is to generate gene expression profiles
using microarray technology, from given glioma
cell lines treated with a specific class of agents
(i.e. farnesyl transferase inhibitors). If characteristic
patterns could be identified that correspond with
anti-tumor activity, then clinical trials can/will
be devised to administer one of these agents to
patients with brain tumors immediately prior to
biopsy/surgery in order to attempt and identify
a similar genetic profile clinically. This effort
is complimented by the recently completed construction
of the brain tumor microarray chip built in the
laboratories of the Neuro-Oncology Branch on conjunction
with the Cancer Genome Anatomy project (CGAP).
Identification of characteristic expression profiles
may also lead to gene discovery and the identification
of new molecular targets for future therapeutic
intervention.
-
Proteinomics: Another major effort is
to exploit the technology of cell-wide protein
assessment through the developing initiatives
within the Center for Cancer Research (CCR) at
the NCI. In collaboration with investigators like
Lance Liotta in the Pathology Branch of the NCI,
we will begin to look at a wide array of proteins
(i.e. abundance, post-translational modification)
that are thought to be potentially relevant to
the mechanisms of action of the drug being evaluated
and/or proteins important to the biology of the
tumor cell in an effort to identify potentially
clinically useful surrogate markers of drug activity.
-
Serum Markers: Serum markers of anti-tumor
activity would be of enormous utility given their
ready accessibility. We and others have already
begun to evaluate the serum levels of angiogenic
peptides such as VEGF and bFGF as potential markers
of tumor "response", however, pre-clinical modeling
is imperative as is the continued identification
of additional markers for other classes of compounds.
-
Brain Tumor Repository of Mouse Models and Cell
Lines: Recently, a number of mouse brain tumor
models (medulloblastoma, oligodendroglioma, NF/schwanoma)
have been generated as have human brain tumor cell
lines. These reagents have been generally constructed
in individual laboratories, and there currently is
no central place to acquire such reagents. Furthermore,
these tumors and cell lines have not been well characterized.
Thus, it is our intention to collect these reagents
from our collaborators throughout the word and characterize
these tumors and cell lines, morphologically, cytogenetically,
immunohistochemically, and genetically through expression
profiling. We will utilize the most appropriate of
these models for our pre-clinical evaluations. The
animal brain tumor therapeutic core offers an outstanding
opportunity for the extramural NCI-sponsored brain
tumor consortia, the intramural NIH-wide Brain Tumor
Program, and the pharmaceutical industry to more thoroughly
evaluate and model novel anti-tumor agents in a pre-clinical
setting prior to moving into clinical trials. This
should result in superior clinical trial design with
greater certainty of obtaining the maximal amount
of useful data from every treated patient. Furthermore,
the core represents an outstanding resource to entice
potential partnerships between NCI and pharmaceutical/biotechnology
companies for the expressed purpose of using the core
to develop new anti-tumor and/or neural protective
agents.
Back to Top
-
Pediatric Laboratory Research:
Childhood brain tumors are diverse in their biology,
histology, and propensity for dissemination. A major research
objective in the Neuro-Oncology Branch is the development
of noninvasive methods of evaluating childhood brain tumors
using new imaging methods. These include Proton Nuclear
Magnetic Resonance Spectroscopic Imaging (1H-MRSI), 3-dimensional
imaging, and newer MR sequences. 1H-MRSI is
a noninvasive method of monitoring biochemical markers
in vivo within normal brain tissue and tumor. The
biochemical markers measured include N-acetyl aspartate
(NAA), a normal neural marker, creatine (Cr), a marker
of cell energy, choline (CHO), a constituent of the cell
membrane and therefore an indicator of cell number and
turnover, and lactate (LAC), a marker of anaerobic carbohydrate
metabolism. Data from our pilot 1H-MRSI study
indicate that the CHO:NAA ratio may be predictive of outcome
in children with recurrent primary tumors. We are continuing
these spectroscopy studies in an effort to define specific
spectroscopic patterns for tumor growth, tumor edema,
tumor necrosis and tumor response to chemotherapy in children
with brain tumors. Comparative studies using PET (Positron
Emission Tomography) scanning are planned. 1H-MRSI
is also being used to study neurotoxic effects of cancer
treatment in pediatric patients. Patients with or at risk
for neurotoxicity are currently being evaluated by MRI,
1H-MRSI and neuropsychologic testing. Correlation
of spectroscopic findings with the results of neuropsychological
testing will then be performed to determine if neurotoxicity
can be objectively defined.
Another important research objective is the development
of novel experimental therapeutic agents for children
with CNS tumors. We are particularly interested and committed
to development of innovative approaches, newer agents
that can overcome drug resistance, and agents that can
block or overcome the blood:brain barrier, thereby enhancing
drug delivery to the tumor. We have a strong collaboration
with the NCI Pediatric Oncology Branch and the Pharmacology
and Experimental Therapeutics Section in development of
new pharmacologic agents. Clinical pharmacokinetic and
pharmacodynamic studies are performed in the majority
of children enrolled on investigational studies.
Back to Top
|
|
