Note: Descriptions are shown in the official language in which they were submitted.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
1
CELL SEPARATION MATRIX
Background of the Invention
1. Field of the Invention
The present invention generally relates to a matrix for separating cells.
More particularly, the present invention relates to a cell-separation matrix
that may be
used to selectively isolate cells with metastatic potential. The cell-
separation matrix
may be used in the diagnosis of metastatic cancers and in the treatment of
cancer by
reducing circulating metastatic cells.
2. Description of the Related Art
Primary cancers frequently shed neoplastic cells into the circulation at an
early stage of metastases formation (Fidler I J, 1973, European Journal of
Cancer
9:223-227; Liotta LA et al., 1974, Cancer Research 34:9971004). Patients with
metastatic disease may release large numbers of cancer cells into the
circulation, in
many cases approaching release-rates of 10' to 109 cells per day (Glaves, D.,
RP Huben,
~ L. Weiss. 1988. Br. J. Cancer. 57:32-35). However, studies suggest that only
a minor
subpopulation of shed cancer cells, ranging from one of thousands to millions
of cells,
ZO are metastatic (Glaves, D., 1983, Br. J. Cancer, 48:665-673). The fact is
that the
majority of shed cancer cells do not survive in the circulation (Weiss and
Glaves, 1983;
Karczewski et al., 1994). Experimental data suggest that the initial release
of cancer
cells from a primary tumor is not the limiting factor in metastatic
development. When
tumor cells are introduced directly into the circulation of mice or rats, less
than 0.01
of such cells form tumor nodules. More commonly the efficiency is two or more
orders
of magnitude lower (Luzzi, K.J. et al. 1998. Am. J. Pathol. 153, 865-873).
It has been suggested that the adhesion of metastatic cells to the
extracellular matrix of basement membrane and connective tissue underlying
vessel
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
2
walls and subsequent tissue degradation are key events for metastases
formation in an
organ (Liotta et al., 1991, Cell 64:327-336). It is also believed that
angiogenesis, that
is the process of signaling new blood vessel growth into a growing tumor mass,
is
required for the survival, growth and metastasis of cancer cells (Folkman,
1995). It is
known that there a small number of endothelial cell progenitors or angioblasts
circulate
in human peripheral blood (Asahara et al., 1997). In addition, it is known
that a small
percentage of leukocytes in human peripheral blood that are activated to
associate with
circulating cancer cells. It is possible that during intravascular metastases
formation, a
small fraction of circulating cancer cells, as well as hematopoietic cells
comprising
endothelial cell progenitors and cancer cell-associating leukocytes,
preferentially attach
to sites where connective tissue structure has been modified due to local
wound or
inflammatory responses. The modified matrix may allow local invasion and
growth of
solitary cells (Clark et al., 1985)(AI-Mehdi, A. B. et al. 2000, Nature
Medicine, 6,
100-102).
The present inventor has hypothesized that it would be useful both for
diagnostic and therapeutic purposes to separate the small fraction of
circulating cancer
cells that are metastatic, as well as the rare endothelial cell progenitors
and cancer cell-
associating leukocytes, from the large number of other circulating cells in a
patient's
body. Two major problems have been identified with respect to such cancer cell
separation proposal: (1) the proposed method must isolate specifically viable
cancer
and related tissue cells but leave alone unrelated or damaged cells
(Karczewski et al.,
1994), and (2) that the proposed method must achieve the specificity in cell
separation
of one cell from over one million nucleate cells, or over one billion cells in
whole blood.
There are approximately 109 red cells and 10' white nucleate cells present in
one cubic
centimeter (c.c.) or gram of blood. It is estimated that among the order of 10
billion
total mononuclear cells harvested from a patient with metastatic cancer, there
are 25
thousand to 12 million contaminating cancer cells during traditional bone
marrow
harvest and leucopheresis procedures (Campana, D. et' al. 1995, Blood 85:1416=
34)(Brugger et al., 1999; Brugger et al., 1994; Brugger et al., 1995). These
contaminating cancer cells have been shown by genetic marking to contribute to
relapse (Rill, E R et al., 1994, Blood 84:380-383). Because of the danger
associated
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
3
with such cells, there exists a great need for efficient methods for removing
viable
cancer cells from a hematopoietic cell transplant (Gulati, S C et al. 1993,
Journal of
Hematotherapy, 2:467-71).
Several methods are known for detecting cancer cells from background
tissue cells. Traditional diagnosis utilizes the different morphology of tumor
cells, as
compared to normal cells of the blood and normal tissue cells, followed by
immunocytochemistry using developmental lineage tissue markers such as
antibodies
against hematopoietic and epithelial cells. For example, immuno-morphologic
analysis
may be performed by cytospin preparations or smears of marrow, peripheral
blood or
lymph node cell samples, followed by May Grunwald-Giemsa staining or stained
with
tissue specific antibodies, and examination by light microscopy (Molino et
al., 1991.
Cancer, 67:1033). Alternatively, rare circulating cancer cells have also been
detected
through the use of sensitive, reverse transcriptase polymerise chain reaction
(RT-PCR)
to amplify putative tumor markers or epithelial markers such as prostate
specific
antigen (PSA) mRNA or cytokeratin 19 mRNA (Peck et al., 1998; Wang et al.,
2000).
Microdissection methods are known for separating rare cancer cells from
major tissue cells one by one (Suarez-Quian et al., 1999, Biotechniques,
26:328-35;
Beltinger and Debatin, 1998, Mol. Pathol 51:233-6). These methods have several
disadvantages, particularly with respect to complicated sample processing, no
reference
for cell viability, and false-positive results. Alternative approaches to cell
separation are
based on physical characteristics of tumor cells such as shape, size, density
or electrical
charge (Vona et al., 2000). Circulating nucleated cells can be readily
separated from
large number of background red blood cells as a group called "bufFy coat" on
density
gradients by centrifugation (Dicke et al., 1970, Exp. Hematol. 20:126-130;
Olofsson et
al., 1980, Second J. Hematol. 24:254-262; Ellis et al., 1984, J. of
Immunological
Methods 66:9-16; Sabile et al., 1999, Am. J. Clin. Pathol. 112:171-8).
However, such
methods are dependent on the availability of the buoyant density and
morphology
unique to different nucleated cells, and various cancer cells seem to have
different
physical characteristics.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
4
Most recent approaches to cell separation are antibody-based.
Immuno-affinity methods involve affixing an antibody on a carrier or
fluorescent label,
in which antibody reacts to an antigenic epitope present on the surface of the
cells of
interest. The methods include affinity chromatography, immuno-precipitation,
and flow
cytometry or called fluorescence activated cell sorting (FACS). Flow cytometry
separates and detects individual cells one-by-one from a large number of
background
cells (Herzenberg et al., 1979, Proc. Natl. Aca. Sci. USA 76: 1453-5; Pituch-
Noworolska
et al., 1998, Int. J. Mol. Med. 1:573-8). It has been shown that breast
carcinoma cells
can be isolated and identified from a peripheral blood sample by flow
cytometry (Gross
et al., 1995. Proc. Natl. Aca. Sci. USA. 92:537). However, it could not
resolve cells that
existed in clusters, which may be the case in some cancers.
Other popular antibody-based, cell sorting approaches involve separating
cancer cells from a large number of background cells using antibody-coated
microbeads
in a centrifugation or filtration process (Dicke et al., 1968, Transplantation
6:562-570).
The antibody-coated microbeads may comprise a magnetic material to permit
separation of the cancer cell-bound antibody-coated microbeads from a
challenge
solution by way of a magnetic field (Shpall et al., 1991, Bone Marrow
Transplantation
7:145-151; Durrant et al., 1992, J. Immunol. Meth. 147:57-64; Denis et al.,
1997, Int.
J. Cancer 74:540-4; Racila et al., 1998, Proc. Natl. Acad Sci USA 95-4589-94).
There are numerous disadvantages associated with antibody-based cell
separation methods, including flow cytometry and magnetic cell separation. For
one,
cancer cells often variably express tumor- or tissue specific antigens (Sabile
et al.,
1999). There is also frequently significant non-specific antibody binding to
damaged
cells, with such techniques often including no reference for cell viability.
Overall such
antibody-based cell separation methods have a higher than desired false-
positive rate.
Furthermore, these cell separation methods are time consuming and cost
intensive.
In co-pending International PCT Application No. PCT/US01/26735, filed
August 28, 2001, claiming priority to U.S. Provisional Patent Application No.
60/231,517,
there is described a fibrous matrix scaffolding coated with blood-borne
adhesion
molecules, such as human plasma fibronectin, laminin and vitronectin, which
supports
the attachment of cancer cells and may be used to isolate metastatic cells
from other
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
cells. The fibrous matrix scaffolding of such application may be made of a
number of
materials including collagenous fibers, fibrin gels, purified cotton or
plastic fibers. The
matrix may be housed in a vessel. The cells captured by the matrix are assayed
ex vivo
as putative metastatic cells: (1) for their viability by apoptosis and
cytotoxicity assays,
5 (2) for their cell proliferation, and (3) for measurement of their
metastatic potential, i.e.,
assaying their ability to digest and internalize matrix fragments,
simultaneously. In
addition, conventional pathological methods for detecting cancer cells may be
used,
including cell size, nuclear shape, and immunocytochemical reactivity against
tissue
markers, such as PSA, cytokeratins, pan-epithelial antigen BerEP4 present on
normal
and neoplastic epithelial cells. The co-pending patent application is based on
the
observation that cancer cells present in the circulation of patients with
metastatic
diseases can attach to tissue fragments and form large cellular clusters. This
observation suggests that natural structural scaffolds promote attachment of
metastasized cancer cells, as well as hematopoietic cells associated with
metastasis.
Co-pending International PCT Application No. PCT/US01/26735 discloses that
type I/III
collagen, fibrin, purified cotton, and mechanically scratched surfaces of
tissue culture
plastic, absorb preferentially blood-borne adhesion components that promote
adhesion
of cancer cells.
Also described in co-pending International PCT Application No.
PCT/US01/26735 is a method for inhibiting the metastatic potential of cancer
cells by
administration of modulators of serine integral membrane proteases, in
particular those
inhibitors that interfere in the formation of a protease complex comprising
seprase and
dipeptidyl peptidase IV ("DPPIV").
Several cell separation systems are presently available for separation of
circulating cancer cells from blood of cancer patients. Table 1 summarizes
some of the
characteristics of the available methodologies, including a density gradient
centrifugation separating cells by cell density, a filtration based on cell or
clump size,
flow cytometry or microscopy of fluorescent antibody-targeted cells, magnetic
separation using cells bound by antibody-magnetic particles, and a functional
separation for viable cells based upon a matrix described in International PCT
Application PCT/US01/26735 (filed by the present inventor).
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
6
Table 1
Human circulating cancer cells resolved b~~ different methods
Methods Cells/mL* Emboli/mL**Cell References
viabili
(1) Antibody-antigen14 - 21,2093 - 1,462 Not knownGlaves etal.,
reaction
followed b centrifu 1988
ation
(2) Negative antibody2 - 5 2 - 5 Not knownTs'o et al.,
1977;
depletion followed Wang et al.,
by 2000
centrifu ation
(3) Autotransfusion 1032 - 56 - 8,370Mostly Karczewski
followed et al.,
b filtration b cell 101025 dead 1994
size
4 Filtration b cell 1- 3 1 - 12 Not knownVona et al.,
size 2000
(5) Antibody-fluorescence3,710 - Not known Not knownKraeft et al.,
2000
microsco is ima in 10 200
(6) Antibody-magnetic2 - 6 Not known Not knownRacila et al.,
1998
fluid/flow c omet
(7) Antibody-magnetic2 - 6 Not known Not knownBeitsch and
fluid/flow c omet Clifford, 2000
(8) Functional affinity182 - 18,0033 - 1,231 Viable Co-pending
to
matrix of International International
PCT PCT
Application PCT/US01/26735 Application
PCT/US01/26735
claiming priority
to U.S. Provisional
Patent Applic.
No.
60/231,517
*Range of putative cancer cells found in one milliliter of blood or in 106
equivalent nucleated blood cells by particular methods,
which have demonstrated the sensitivity of 1 cell per mL and background level
(no or few cells) in the blood from normal donor.
** Range of cell clusters or clumps containing 5 -100 putative cancer cells
found in one milliliter of blood or equivalent nucleated
blood cells by particular methods, which have demonstrated the sensitivity of
1 cell per mL and background level (no or few cells)
of blood from normal donor.
Summary of the Invention
The present invention provides a cell-separation matrix modified from
that described in co-pending PCT Patent Application PCT/US01/26735 (claiming
priority
to U.S. Provisional Patent Application No. 60/231,517) which provides an
improved
matrix for separating cells in a manner to isolate and detect metastatic
cells, and a
small fraction of hematopoietic cells associated with metastasis, from blood
and tissues
of patients inflicted with metastasic cancer. The modified-matrix provides a
"cancer cell
trap" that allows for the efficient removal of viable cancer cells from the
tissue fluids.
The modified-matrix is useful for separating over 99% of blood cells from such
metastatic, and associated-metastatic, cells. Metastatic cells may be
characterized by
in vitro assays including the local collagen or fibronectin degradation and
internalization,
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
7
cell proliferation, pathological and immunocytochemical identification, and
apoptotic
and cytolytic assays.
The modified-matrix of the present invention utilizes an intermediate
coating about a core material to effectuate improved absorption of blood-borne
adhesion components that promote the adhesion of cancer cells. The
intermediate
coating comprises materials, including, but not limited to, gelatin,
collagens, fibrin,
proteoglycans, hyaluronate, and dextran, that has the affinity, or efficiently
binds to
another material having the affinity, to bind blood-borne adhesion components
that
promote the adhesion of cancer cells, such as fibronectin, fibrin, heparin,
laminin,
tenascin or vitronectin, and synthetic compounds, such as synthetic
fibronectin and
laminin peptides and the like, and that has the ability to effectively coat
the core
material used in the matrix. For example, glutaraldehyde can be used to coat
bone
substrata and to bind blood-borne adhesion components that promote the
adhesion of
cancer cells. Gelatin has been found to be useful to coat core materials such
as whole,
denatured, polymers and fragments of bone, connective tissues (such as
collagens,
proteoglycans and hyaluronate), glass, inert polymeric materials (such as
magnetic
colloid, polystyrene, polyamide material like nylon, polyester materials
cellulose ethers
and esters like cellulose acetate, urethane foam material, DEAE-dextran), as
well as
other natural and synthetic materials, such as other foam particles, cotton,
wool,
dacron, rayon, acrylates and the like. The gelatin-coated core materials may
then be
crosslinked, for example, with glutaraldehyde, washed and the glutaraldehyde
cross-
linked, gelatin-coated, core material exposed to one or more blood-borne
adhesion
components that promote the adhesion of cancer cells. The blood-borne adhesion
components that promote adhesion of cancer cells, may comprise fibronectin,
fibrin,
laminin, heparin, and vitronectin, or biological mimics thereof, and may be
prepared by
purification from natural sources or synthesized by artificial means.
The modified-matrix system of the present invention more efficiently
captures and detects viable cancer and hematopoietic cells from tissue fluid
samples
derived from cancer subjects than that described in co-pending application PCT
Patent
Application PCT/US01/26735 (claiming priority to U.S. Provisional Patent
Application No.
60/231,517). The modified matrix of the invention has affinity for metastatic
cancer
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
8
cells and a small fraction of hematopoietic~cells, and it mimics the site at
the vessel wall
of arteriovenous anastomosis and loci of metastases, where extracellular
matrix (ECM)
components, including bone matrix, collagens, proteoglycans, fibronectin,
laminin,
fibrin, heparin, tenascin and vitronectin etc., have been modified during the
process of
intravasation. In essence, the modified matrix mimics a metastatic environment
capturing cancer cells. The cancer cells isolated by the methods of this
invention are
viable, grow ex vivo, and exhibit the invasive activity against the ECM, i.e.,
partially
degrading it followed by ingestion of ECM fragments by the cells.
In certain embodiments, bone fragments are used themselves as the core
material, which form shapes of planar substrata or beads. While the bone
substrata can
be used directly to bind the blood-borne adhesion components that promote the
adhesion of cancer cells, such substrata is more efficiently crosslinked with
glutaraldehyde, followed by blocking with blood-borne adhesion components that
promote the adhesion of cancer cells, for example, with 0.01 - 0.5 milligram
per
milliliter of human plasma fibronectin, fibrin, heparin, laminin, tenascin,
vitronectin, or
synthetic compounds, such as synthetic ~bronectin and laminin peptides and the
like.
The coated-cross-linked bone substrata or beads have been found to more
efficiently
capture viable cancer cells from a tissue fluid such as blood. Again, the bone
substrata
or beads are used as mimic of a natural matrix substrata that captures cancer
cells and
~0 a small fraction of hematopoietic cells from blood or other tissue fluids
related to
metastasis, and can be used to detect those cancer cells and small fractions
of
hematopoietic cells.
In other embodiments, surfaces of the core materials are activated
directly with bifunctional crosslinkers such as glutaraldehyde, washed and
blocked with
blood-borne adhesion components that promote the adhesion of cancer cells,
such as,
for example, 0.01 - 0.5 milligram per milliliter of human plasma fibronectin,
fibrin,
heparin, laminin, tenascin, vitronectin, or synthetic compounds, such as
synthetic
fibronectin and laminin peptides and the like in sterile and non-leaking
conditions. The
core materials including, but not limited to, bone, glass, inert polymeric
materials, such
as magnetic colloid, polystyrene, polyamide material like nylon, polyester
materials,
cellulose ethers and esters like cellulose acetate, urethane foam material,
DEAE-dextran,
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
9
as well as other natural and synthetic materials, such as other foam
particles, cotton,
wool, dacron, rayon, acrylates and the like. The blood-borne adhesion
components
coated core materials are used as mimic of a natural matrix substrata that
captures
cancer cells and a small fraction of normal cells that are related with
metastasis, and
may be used to detect such cells.
In yet other embodiments, forms of denatured collagens, called gelatin,
are used to coat core materials including, but not limited to, bone, glass,
inert
polymeric materials, such as magnetic colloid, polystyrene, polyamide material
like
nylon, polyester materials, cellulose ethers and esters like cellulose
acetate, urethane
foam material, DEAE-dextran, as well as other natural and synthetic materials,
such as
other foam particles, cotton, wool, dacron, rayon, acrylates and the like. The
gelatin-coated core materials are then crosslinked with glutaraldehyde, washed
and
blocked with blood-borne adhesion components that promote the adhesion of
cancer
cells, such as, for example, 0.01 - 0.5 milligram per milliliter of human
plasma
fibronectin, fibrin, heparin, laminin, tenascin, vitronectin, or synthetic
compounds, such
as synthetic fibronectin and laminin peptides and the like in sterile and non-
leaking
conditions. Once more, the gelatin-coated core materials crosslinked with
glutaraldehyde and blocked with blood-borne adhesion components are used as
mimic
of a natural matrix substrata that captures cancer cells and a small fraction
of
hematopoietic cells that are related with metastasis, and may be used to
detect such
cells.
Co-pending application PCT Patent Application PCT/US01/26735 (claiming
priority to U.S. Provisional Patent Application No. 60/231,517) discloses that
the cell-
adhesion matrices may comprise core materials comprising collagenous fibers,
fibrin
gels, purified cotton or plastic fibers. The present invention discloses that
many more
core materials may be used. Some of these core materials have been found to be
able
to be coated without an intervening intermediate layer with purified human
plasma
fibronectin or its fragments.
The modified matrix may be contacted directly with the fluid from which
the metastatic cancer cells are to be isolated, or may be applied as a thin
coating to a
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
cell separation vessel, such as a filter, tube, capillary, culture plate, cell
isolation
column, a flask etc., that are preferably sterilized. The thin coating is
preferably
immobilized to the cell separation vessel. The matrix-coated surfaces of the
cell
separation vessels are preferably designed maximize surface contact area.
Beads,
5 microbeads, or microcarriers may be used as a core material in order to
increase the
surface area available for contacting cells. The core material may also be in
the form
of micromeshes and/or packed beads. Matrix-coated beads and micromeshes form
filtration channels to maximize contact areas between matrix and cells
improving cell
separation efficiency.
10 The modified matrix may be used to remove metastatic cancer cells and
hematopoietic cells related to metastasis from a number of tissue fluids
including, but
not limited to, blood, bone marrow, ascites, lymph, urine, spinal and pleural
fluids,
sputum, airway and nipple aspirates. The cell separation method of this
invention may
also be used to isolate such cells from dissociated tumor tissue specimens and
cultured
tumor cells. Cancer cells that may be isolated using the modified matrix
include, but
are not limited to, carcinoma cells of prostate, breast, colon, brain, lung,
head & neck,
ovarian, bladder, renal & testis, melanoma, liver, pancreatic and other
gastrointestinal
cancer. Cancer cells that are particularly desired to be isolated include lung
carcinoma
cells, lung adenoma cells, colon adenocarcinoma cells, renal carcinoma cells,
rectum
adenocarcinoma cells, ileocecal adenocarcinoma ~ cells, gastric
adenocarcinoma,
pancreatic carcinoma, hepatoma cells, hepatocellular carcinoma cells, prostate
adenocarcinoma cells, bladder carcinoma cells, breast carcinoma, ovarian
carcinoma,
teratocarcinoma, amalanotic melanoma cells, malignant melanoma cells, squamous
cell
carcinoma of the cervix, esophagus, head & neck, air-way, larynx and of oral
origin;
glioblastoma cells, and endometrial adenocarcinoma cells. The present
invention
provides effective cell separation methods for diagnostic and therapeutic
applications in
patients with metastatic diseases, including, but not limited to, prostate,
breast, colon,
brain, lung, head & neck, ovarian, bladder, renal & testis, melanoma, liver,
pancreatic
and other gastrointestinal cancer.
Cell separation is performed by contacting the tissue fluid with the
modified matrix surface. Tissue fluids such as whole blood, huffy coat, bone
marrow,
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
11
ascites, and lymph are treated with anticoagulants to prevent coagulation
during the
cell separation procedure. For examples, blood and huffy coat may be pre-
diluted with
one tenth volume of medium containing 0.5 mM EDTA or with anticoagulant
citrate
dextrose (ACID; Baxter Healthcare Corporation, IL) containing 50 unit
heparin/ml.
The modified-matrix of the present invention can capture "viable" cancer
and the small fraction of hematopoietic cells circulating in the blood
involved in
metastasis, but has little affinity for over 99.99% of blood cells. The
invention is based
on the adhesive and invasive functions of cancer cells and the small fraction
of
hematopoietic cells involved in metastasis with respect to the modified
matrix. Cancer
cells that are isolated may be subjected to in vitro assays, demonstrating
that they are
viable, invasive and metastatic. As the matrices of the present invention are
non-toxic
they can also accommodate the growth of isolated cells. The matrix facilitates
cell
separation enabling one to count the number of isolated viable cells, analyze
genomic
changes, profile gene expression and proteomics, and treat the tissue fluid
where
targeted cells are present in very low concentrations. The sensitivities can
be on the
order of 1 cell to 1 gram of sample.
It may be desired that the separated cells remain viable. For example, it
may be desired to reuse certain of the separated cells therapeutically, or to
grow them
(e.g. the metastatic cancer cells) in an in vitro culture in order to amplify
a signal for
vaccine development. Conventional techniques such as the use of antibody-
affinity
microbeads typically subject the cells to a complicated and traumatizing
course which
not infrequently has an injurious effect on the cells. Considering the very
low
occurrence of the target cells, this phenomenon is particularly distressing.
This invention also provides an efficient method wherein viable cells
captured on the modified matrix can be released readily from the modified
matrix by
the use of digestive enzymes, including, but not limited to, trypsin/EDTA
solution
(purchased from GIBCO), collagenases and hyaluronases. Cell adhesion molecules
of
the modified matrix, including fibronectin, laminin, and vitronectin etc, are
sensitive to
digestion. These enzymes will cleave binding between the cells and the
modified matrix,
and release viable cells from the matrix into suspension.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
12
The cell separation method of the present invention may be used for
cancer diagnostic purposes, e.g. early detection, monitoring therapeutic and
surgical
responses, and prognostication of cancer progression. The enriched separated
cancer
cells can be used, for example, to determine the metastatic potential of the
patient's
cancer. The sensitivity and accuracy of measuring the metastatic potential of
a cancer
may be further enhanced using additional assays known to those of skill in the
art,
such. as determining the tissue origin of cancer cells, measuring the
angiogenic
capabilities of the cells, and determining the degree of reduction in
leukocyte count or
complement association.
Prognosis and therapeutic effectiveness may also be adjudged by assays
that count numbers of viable and rnetastatic cells in the blood or other
tissue fluids
during and post therapeutic intervention(s). For example, the modified matrix
may be
contacted with a blood sample from a cancer patient and the isolated cancer
and
hematopoietic cells associated with metastasis subsequently detected and
quantified
using a combination of antibody labeling and microscopic imaging or flow
cytometry.
Selection of chemotherapeutic regimen may be optimized by determining those
regimens that most effectively, without undue side effects, reduce the number
of.
cancer cells and hematopoietic cells associated with metastasis in the blood
sample as
detected by the matrix. Optimization of selection of chemotherapeutic regimen
may
also be performed by subjecting the isolated cancer and hematopoietic cells to
a
battery of chemotherapeutic regimes ex vivo. Effective doses or drug
combinations
could then be administered to that same patient.
The cell separation system of the present invention may also be used to
detect whether a new compound or agent has anti-cancer activity. For example,
the
number of viable cancer cells in whole blood can be determined before and
after the
administration of the compound or agent, with compounds or agents
significantly
reducing the number of viable cancer cells in .the._blood
after._administration being .
selected as potential anti-cancer candidates. Comparing the metastatic
potential of the
cancer cells throughout the treatment can follow the efficacy of the agent.
Agents
exhibiting efficacy are those, which are capable of decreasing number of
circulating
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
13
cancer cells, increasing number of viable associated leucocytes (host
immunity), and
suppressing cancer cell proliferation.
The modified matrix of the present invention may also be used as a
"cancer cell trap" that allows for the high yield and efficient removal of
viable cancer
cells from the tissue fluids. The cell separation method of the invention may
be
employed in respect of the autotransfusion of blood salvaged during cancer
surgery,
therapeutic bone marrow transplantation, peripheral blood stem cell
transplantation
and leucopheresis, in which autologous transfusions are done, from which
contaminating cancer cells have been removed.
The enriched cancer cells and their specific clusters of surface antigens
isolated using the modified matrix may be used in fusions with dendritic cells
for cancer
vaccine development. For example, the cancer cells of different carcinoma
cancers
may be subjected to ex vivo culture and expansion, and the cells used in
whole, or
purified for specific membrane structures or for specific antigens, to
interact with
dendritic cells to develop an effective tumor vaccine.
As would be understood by one of skill in the art, the cell fraction
enriched for cancer cells isolated using the disclosed matrices may also be
used as a
source of DNA, RNA and proteins in genomic, gene expression and proteomic
profiling
studies, for further discovery of genes, proteins and epitopes characteristic
of the
metastatic cell phenotype.
Further, the described matrices may be used to prevent full blown cancer
from occurring by removing cells capable of metastasis from the circulation.
Brief Description of the Drawings
Fig. IA depicts a front sectional view of an upright vacuum blood
collection-tube coated along its internal surface with a modified matrix film
capable of
segregating cells associated with metastasis that may be used in the diagnosis
of
metastatic cancer;
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
14
Fig. IBdepicts an enlarged front sectional view of a portion of the upright
vacuum blood collection tube of Fig. IA illustrating viable cancer and
hematopoietic
cells captured by the modified matrix film coated on the glass core material;
Fig. 2A depicts a front sectional view of an upright vacuum blood
collection tube containing cell separation beads coated with a modified matrix
film and
further comprising a separator for capturing and filtering the separation
beads when
the tube is inverted;
Fig. ZB depicts an inverted front sectional view of the vacuum blood
collection tube of Fig. 2A showing the cell separation beads trapped in the
filter
separator;
Fig. 3A depicts a front sectional view of an upright vacuum blood
collection tube containing cell separation microbeads or nanoparticles coated
with a
modified matrix film and having an intermediate magnetic coating;
Fig. 3B depicts a front sectional view of the upright vacuum blood
collection tube of Fig. 3A wherein a magnetic separator is applied to the tube
to
segregate the cell separation microbeads or nanoparticles from the
supernatant;
Fig. 4A depicts a three-dimensional view of a cell separation filter
containing within an inner confinement area cell separation beads coated with
modified
matrix which may be used in diagnostics, therapeutics or treatment according
to the
invention;
Fig. 4B is an expanded view of the portion of cell separation beads
designated in Fig. 4A, depicting the anastomosic channels formed by the cell
separation
beads within the inner confinement area;
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
Fig. 5 is a schematic representation of a method of the present invention
providing for the isolation of cancer cells from tissue samples using a
combination of
the function-affinity cell separation of the invention and immuno-affinity
purification.
5 Detaited Description of the Invention
Discussion:
The present invention provides an improved cell separation substratum
for separation of metastatic cancer and a small fraction of normal cells
associated with
metastasis from a tissue fluid sample. The improved cell separation substratum
10 comprises a supporting core material, comprising, but not limited to, bone
or tissue
fragments, magnetic colloid, plastic, glass and stainless steel, coated with
an
intermediate coating comprising material that has affinity, or efficiently
binds to
another material having the affinity, to bind blood-borne adhesion components
that
promote the adhesion of cancer cells, such as fibronectin, fibrin, heparin,
laminin,
15 tenascin, vitronectin, or their fragments, and that has the ability to
effectively coat the
core material used in the matrix. The improved cell separation substratum, or
matrix,
may be used to coat areas of objects which are intended to be in contact with
the
tissues from which the metastatic cancer cells and small fraction of normal
hemopoietic
cells that are to be isolated, such as a blood collection tube, plate, or
flask, the surface
of beads, or the inner lining of a capillary or filter, or may comprise the
material of the
object itself as the core material and another substance as the intermediate
coating
material. For example, a gelatin solution (2.5% gelatin w/v and 2.5% sucrose
w/v in
PBS) may be first coated on inner wall of a blood collection glass tube and
the gelatin
film fixed with 1 % glutaraldehyde, followed by PBS washing and masking by
human
plasma fibronectin, 0.1 mg/ml, in sterile condition. The modified matrix in
such case
would comprise glass coated with gelatin masked with human plasma fibronectin.
By permitting isolation of viable cancer cells in high efficiency (i.e.,
allowing one to isolate the relatively small number of cancer cells typically
seen in most
tissue samples), the present invention achieves a highly desirable objective,
namely
providing a method for the prognostic evaluation of subjects with cancer and
the
identification of subjects exhibiting a predisposition to developing
metastatic cancer.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
16
The invention encompasses a method for determining the number of
viable metastatic cells derived from a cancer subject comprising:
(a) adding a tissue fluid sample to a cell separation vessel, wherein the
wall contacting the fluid is coated with a modified matrix film under
conditions sufficient to specifically bind to cancer cells and a small
fraction
of hematopoietic cells associated with metastasis;
(b) washing the matrix films and removing unbound cells;
(c) treating the cell-bound matrix films with proteolytic enzymes; and
(d) eluting bound cells from the matrix films onto a solid support to
provide an enriched cell sample comprising cancer and the small fraction
of hematopoietic cells.
The enriched cell sample may be used for detecting and counting the
number of viable cancer and a small fraction of hematopoietic cells using
microscopic
imaging or flow cytometry, wherein a detection of increasing number of viable
cancer
cells is an indicator of cancer cells with metastatic potential, and
increasing number of
hematopoietic cells is an indicator for host immunity.
The enriched cell sample may also be used for identifying an agent that
inhibits metastasis of cancer cells by detecting and counting the number of
viable
cancer and hematopoietic cells treated with exogenous agents. A decrease in
the
number of cancer cells in the presence of the test agent, as compared to the
number of
cancer cells detected in the presence of a vehicle control, identifies a
compound that
inhibits metastases formation. On the other hand, an increase in the number of
hematopoietic cells in the presence of the test agent, as compared to the
number of
hematopoietic cells detected in the presence of a vehicle control, identifies
a compound
~5 that has immune activity-against metastases formation.
Metastatic cancer cells may be identified by particular functional assays
including:
(a) the intake of collagen or matrix fragments;
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
17
(b) the intake of acetylated low density lipoprotein (acLDL);
(c) the capacity of continued growth in culture in conditions containing
complement-inactivated human sera; and
(d) the recognition by antibodies against both epithelial and endothelial
markers but not by antibodies against leukocyte/monocyte common
antigens such as CD14, CD45, and CD68.
Enumeration of metastatic and hematopoietic cells in a given sample may be
performed
either by microscopic imaging or flow cytometry.
In accordance with one aspect of the invention, a crosslinked gelatin film
may be prepared using the three following steps (a) to (c):
(a) gelatin is prepared and isolated from connective tissues of human or
other animals;
(b) core material is covered with gelatin;
(c) the gelatin is crosslinked and the functional groups from the
crosslinking agent are blocked with fibronectin.
Gelatin may be crossed-linked as described in Chen and Singer, 1980; Chen et
al., 1994.
The gelatin-crosslinking method can be modified by persons of ordinary skill
in the art
to produce a gelatin-coating film having an affinity to viable cancer cells
and a specific
subset of hematopoietic cells associated with such viable cancer cells.
In one embodiment of the invention, there are provided cell separation
beads comprising the modified-matrix. The core material of the beads may
comprise,
without limitation, bone, glass, inert polymeric materials, such as magnetic
colloid,
polystyrene, polyamide material like nylon, polyester materials, cellulose
ethers and
esters like cellulose acetate, urethane foam material, DEAE-dextran, as well
as other
natural and synthetic materials, such as other foam particles, cotton, wool,
dacron,
rayon, acrylates and the like. The beads preferably have a diameter in the
range of
100 microns to 1,000 microns. The beads are coated on their surface to form a
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
18
modified matrix having tremendous surface areas for contacting cells in the
fluid. To
enhance handling of beads in fluid, the core can have an intermediate magnetic
coating,
allowing the beads to be subsequently separated from tissue sample, and/or
from each
other, in a magnetic field. The cell separation beads can be placed into a
blood
collection tube, plate, flask, capillary, etc. for providing a confined area
in which the
beads may contact the cells in the fluid.
The cell separation beads are preferably 100 microns to 1,000 microns in
diameter and may be coated with a crosslinked gelatin film. In one embodiment
the
crosslinked gelatin-coated beads are housed within a sterile vacuum blood
collection
tube with anticoagulant powder containing lithium heparin. In such embodiment
approximately, 0.1-mL of gelatin-coated beads are used for every 5-ml blood
that is to
be collected. The blood-bead mixture in the tube is placed on a shaker set at
slow
speed at 37°C for 30 minutes to 2 hours. The beads are then washed and
collected
using a mesh filter, preferably having mesh-opening widths of 75 +/- 12
microns.
The cell separation beads may be used to isolate cells associated with
metastasis using the following method:
(a) adding a tissue fluid sample to a vessel containing the cell separation
beads under conditions sufficiently allow the beads to bind to cancer cells;
(b) washing the beads and removing unbound cells through the use of a
filter, such filter preferably having mesh opening widths of 75 +/- 12
microns;
(c) treating the cells-bound beads with proteolytic enzymes; and
(d) eluting bound cells from the beads onto a solid support to provide an
enriched cell sample comprising cancer cells and typically a small fraction
of hematopoietic cells associated with metastasis.
Alternatively the core material may comprise fibers. Fibers selected must
be inert and compatible with the blood, and should be somewhat stiff to adhere
well to
the coating material, such as gelatin film. Preferably in a blood filter using
fibers as its
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
19
core material, the size of fibers should not typically exceed about 2 cm long,
and
should range from 10 microns to 1,000 microns in diameter. In blood filters,
if the
fibers are too big or too long, they can compact at high flow rates and less
channel
surface areas, and, therefore, be less efficient. In blood filters, the nature
of fibers
should be selected such that the fibers may adhere to the coating material and
create a
smooth anastomosic channel within the filter for blood flow. In forming a
preferred
filter, fibers may be packed tightly between two layers of meshes having mesh
opening
widths of 50 to 100 microns.
In magnetic cell separation applications involving the binding of
microbeads or nanoparticles to cell surfaces, the microbeads or nanoparticles
have a
diameter in the range of 20 nm to 20 microns. The cell separation microbeads
or
nanoparticles may be directly coated with cell adhesion molecules using an
attachment
agent such as glutaraldehyde to activate binding of cell adhesion molecules to
the
surface of magnetic colloid microbeads or nanoparticles. In one embodiment the
blood
borne-cell adhesion molecules-coated microbeads are housed within a sterile
vacuum
blood collection tube with anticoagulant powder containing lithium heparin. In
such
embodiment, approximately 50 millions of the modified matrix coated microbeads
or
nanoparticles are used for every 5-rnl blood that is to be collected. The
blood-microbead mixture in the tube is placed on a shaker set at slow speed at
37°C for
30 minutes to 2 hours. The microbeads are then washed and collected by passing
the
sample through a magnetic field to magnetically immobilize cells-microbeads
mixture.
The cell separation magnetic microbeads or nanoparticles may be used to
isolate cells
associated with metastasis using the following method:
(a) adding a tissue fluid sample to a vessel containing the cell separation
magnetic microbeads or nanoparticles under conditions sufficiently allow
the microbeads or nanoparticles to bind to cancer cells;
(b) washing the microbeads or nanoparticles and removing unbound cells
by passing the sample through a magnetic field to magnetically
immobilize microbeads or nanoparticles in the sample having cells bound
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
thereto to provide an enriched cell sample comprising cancer cells and
typically a small fraction of hematopoietic cells associated with metastasis.
A cell separation filter system comprising a pre-filter and the modified-
matrix of the present invention may also be used to separate metastatic cells
from a
5 tissue fluid sample preferably presented as a fluid suspension. Such a cell
separation
filter system may be fabricated using the following steps:
(a) building a pre-filter and a connecting tube;
(b) packing a filter container with a filtration unit containing core
materials comprising one or more of fibers, meshes and beads;
10 (c) bringing the pre-filter and filtration units into contact with a
coating
solution capable of coating the core material;
(d) removing the surplus amount of the coating solution;
(e) drying the coating solution on fibers, meshes and packed beads;
(f) crosslinking the coated film; and
15 (g) conjugating a cell adhesion molecule to the film surface.
Such cell separation filter system therefore comprises a pre-filter, called
the clump
screen, preferably having a mesh pore width of 150 to 500 microns, a
connecting blood
tube and a filter housing. In a preferred embodiment, the filter housing
comprises: an
inlet in the housing for the introduction of the blood to be filtered; an
outlet in the
20 housing for the removal of filtered blood or to return to a patient; and a
filter element
disposed within the housing, which element comprises core materials of beads,
preferably having a diameter in the range of 100 microns to 1,000 microns,
and/or
fibers ranging from 10 microns to 1,000 microns in diameters. Preferably the
core
beads or fibers are packed between two layers of meshes, having mesh-opening
widths
of 50 to 200 microns. The surfaces of both pre-filter screen and the filter
element are
coated with a material that has affinity, or efFciently binds to another
material having
the affinity, to bind blood-borne adhesion components that promote the
adhesion of
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
21
cancer cells, such as fibronectin, fibrin, heparin, laminin, tenascin,
vitronectin, or their
fragments, and that has the ability to effectively coat the core material, so
that the
pre-filter screen has affinity to cell clumps, called emboli, containing
viable cancer cells,
and the filter element contains anastomosic channels of tremendous surface
areas for
contacting cells in the fluid. The coated surfaces lining the anastomosic
channels
selectively remove viable cancer cells from the blood or other tissue fluids
to be filtered.
The cell separation filter system may be used in a manner to remove
cancer cells and emboli derived from blood or other tissue fluids of a cancer
system
using the following method:
(a) passing a tissue fluid sample through the cell filtration system, in
which pores of the pre-filter and anastomosic channels of the filter
comprise a modified matrix specific for adhesion and invasion by cancer
cells and emboli but not by majority of tissue cells, wherein pores of the
pre-filter and channels of the filter are under conditions sufficient to
specifically bind cancer cells;
(b) allowing a substantial part of cancer cells and emboli to be entrapped
in the cell filtration system; and
(c) removing adherent cancer cells and emboli; or returning the filtered
blood to the patient as needed.
A preferred cell separation filter system contains a pre-filter, preferably
having screen meshes with a pore width of 200 microns, positioned between a
blood
reservoir and the filter housing. The lining of pores in the pre-filter is
coated with a
crosslinked gelating film(s). The pre-filter removes large clumps form blood
containing
cancer cells (such large clumps can clog blood flow). The pre-filter unit may
be
disposable and can be modified form of the helically wound blood filter
described in
United States Patent No. 4,092,246 (comprising sheet material having a pore
width of
200 microns wound into a helical coil of desired tightness).
The cell separation ~Iter system containing the pre-filter may also be used
as a blood filter by subjects having metastatic cancer. The use of such a
filter system
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
22
involves the perfusion of the subject's blood through the modified-matrix
anastomosic
channels in the filter. In a preferred protocol, the subject's blood is
withdrawn and are
passed in contact with the modified matrix. During such passage, cancer cells
present
in the patient's blood preferentially adhere to the matrix and are removed
from the
circulation of a patient.
In a specific embodiment of the cell separation filter system useful for
filtering metastatic cancer cells from a patient's blood, the pre-filter and
the filter are
formed within a containment vessel. The containment vessel is connected to a
blood
input line which is operatively coupled to a conventional peristaltic pump or
to a
gravity-dependent blood flow system. A blood output line is also included.
Input and
output lines are connected to appropriate arterial or venous fistulas, which
are
implanted into, for example, the forearm of a subject. Citrate-phosphate-
dextrose
anticoagulant is automatically added into the blood flow in an appropriate
ratio.
Alternatively, apheresed peripheral blood can be applied in conjunction with
the cell
filtration system. Apheresis is initiated upon recovery of the white blood
cell count to
equal or more than 1 X 109/L. Apheresis or leucopheresis can be performed
using a
Cobe Spectra Cell Separator (Lalcewood, Colo.) at a rate of 80 ml/min for 200
min (total
volume of 16L).
A method of preventing metastases formation in a cancer subject using
such blood filter comprises:
(a) inoculating a cancer cell sample derived from a cancer subject onto
the cell filtration system;
(b) incubating the cancer cell sample for a time sufficient to allow
adhesion of cancer cells to the coated pores of the pre-filter and
anastomosic channels of the filter; and
(c) returning the filtered blood to the cancer patients.
Intraoperative autotransfusion of blood during major surgical procedures
for removal of primary tumors and bone marrow transplantation for
immunotherapy
can be applied. The salvaged blood samples such as blood harvested from
patients
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
23
undergoing abdominal surgery for resection of primary cancers are passed
through the
cell filtration system of the present invention in conjunction with a
commercial
gravity-dependent blood device such as OR Bloodbanker autotransfusion system
(International Technidyne, Edison, NJ) or the Cell Saver (Haemonetics, Natick,
MA).
Citrate-phosphate-dextrose anticoagulant is automatically added into the
salvaged
blood in an appropriate ratio. The use of the cell filtration system of the
invention
provides a novel method that can remove viable and still invasive cancer cells
from the
salvaged blood and bone marrow, which provides potentially significant
clinical benefit
of autotransfusion and bone marrow transplantation to cancer patients.
The invention encompasses a method for isolating metastatic and
angiogenic cells from a cancer subject comprising:
(a) passing a tissue fluid sample through the cell filtration system,
wherein pores of the pre-filter and channels of the filter are under
conditions sufficient to specifically bind cancer cells and emboli;
(b) washing the pre-filter and the filter and removing unbound cells;
(c) treating cell-bound anastomosic channels and pre-filter screen with
proteolytic enzymes; and
(d) eluting bound cells and emboli from the pre-filter and the filter onto a
solid support to provide an enriched cell sample comprising cancer cells
and emboli.
Given the ability of such modified-matrix filters to isolate viable cells
involved in metastasis and angiogenesis, the cells isolated by the present
invention
provide cellular sources for the discovery of cellular genes, RNAs, proteins
and antigens
important for prevention and intervention of metastases formation in a cancer
subject.
For example, DNA microarray technology has been used advantageously
in the identification of numerous genes differentially expressed in ovarian
tumor
samples (Welsh et al., 2001; Su et al., 2001; Giordano et al., 2001). From
these studies,
i~nany genes have emerged as promising biomarker candidates, including HE4, a
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
24
secreted protease inhibitor. Using a specialized array, many angiogenesis
genes were
found differentially regulated in ovarian cancer. In addition, serial analysis
of gene
expression (SAGE) was used to identify up-regulated genes in ovarian cancer,
including
I<op, SLPI, claudin-3 and claudin-4, making these products attractive
candidate
biomarkers. However, few have been linked to cancer progression and
metastasis. A
major problem encountered in linking the same as been the inability to obtain
highly
purified cancer cells to be used in the analysis. The fact is that tumors are
composed
of lots of different cell types. Many genes expressed at different levels are
actually
coming from non-tumor cells. A second major problem in linking up-regulated
genes in
ovarian cancer to cancer progression and metastasis is related to the
viability of the
cells. Apoptotic and necrotic tumor cells are common in larger tumor and
ascites. A
third major problem has been the lack of information concerning the invasive
phenotype of cells under investigation. In order to understanding gene
expression
patterns of cells during cancer progression and metastasis, it is, thus,
necessary to
separate the viable from the dying cancer cells, the aggressive from benign
cells, and
the cancer cells from the normal cells in tumor samples. The present invention
provides a method for separating and concentrating metastatic cancer cells.
It is known in ovarian cancer, that cancer cells can be found in primary
organs, in ascitic fluid blood or lymph, and in peritoneal micrometastases.
Cancer cells
shed in ascites and blood are numerous and they can be obtained by non-
invasive
means. It is postulated that only a small fraction of cancer cells in ascites
or blood may
exhibit ability to adhere to and invade connective tissue barriers, and have
potential for
metastasizing to a new site. These rare cancer cells in ascites or blood are
considered
as "metastatic" cells, which when grown in collagenous matrix may mimic
micrometastases and be considered as "metastasized" cells. By using the
present
invention to isolate metastatic cancer cells, a DNA microarray can be used to
select
robotically several sets of transcripts that were enriched in different
purified viable cell
types to address important questions of cancer progression and metastasis:
(i) higher in metastasized cells than in metastatic cells, indicating
potential genes driving the process of extravasation;
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
(ii) higher in metastatic cells than in primary tumor cells, indicating
potential genes driving the process of intravasation;
(iii) higher in both metastatic and tumor cells than in normal epithelial
cells, suggesting genes encoding early markers for cancer progression;
5 and
(iv) higher in both metastatic and tumor cells of ovarian epithelial
cancer than in cancer cells of other diseases, i.e., endometrioma or colon
adenocarcinoma, suggesting genes encoding possible cancer markers of
tissue origin.
10 The selected genes can be confirmed for their role in cancer progression
and
metastasis by a quantitative analysis using real time PCR on different cell
types derived
from normal, tumor and metastatic tissues. By a combination of DNA microarray
and
real time PCR, novel molecular markers and therapeutic targets for ovarian
cancer can
soon be discovered. Not only could the un-identified gene changes provide good
15 targets for chemotherapeutic drugs, but they may also provide molecular
markers to
help clinicians assess tumor aggressiveness.
As would be understood by one of ordinary skill in the art, the present
invention would likewise find use in other cancer types in characterizing the
roles of
genes, proteins, RNAs and antigens in cancer progression and metastasis.
20 EXAMPLES
EXAMPLE 1
Preparation of crosslinked gelatin films.
The following method may be followed to prepare crosslinked gelatin
films useful in respect of preparing a modified matrix embodiment of the
present
25 invention:
(a) gelatin is isolated from connective tissues of human or other animals
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
26
Type I collagen is purified from connective tissues of rat tails or
human placenta and heat-denatured by boiling for 5 minutes. The
gelatin solution is then allowed to dry at 100~C. in an oven under
vacuum. Gelatin powders include these produced by acid- or
heat-extraction and these from commercial sources including, but not
limited to, heat-denatured bovine type I collagen type A derived from
porcine skin, Sigma Chemical Co., St. Louis, M0, USA.
(b) Core materials are coated with gelatin
Gelatin powders are washed with chill distilled water three times by
stirring and centrifugation of the gelatin particles. The gelatin solution,
containing 2.5% gelatin w/v and 2.5% sucrose w/v, in PBS, pH 7.2, is
heated until boiling for five minutes to completely dissolve gelatin
particles. To coat a cell separation vessel, the gelatin solution is
maintained at 45~C, overlays the core materials, and immediately
removes excess gelatin fluid to leave a thin 1=Ilm covering the core
materials. The gelatin film is left at 45~C for 30 minutes until dried.
(c) The gelatin is crosslinked with a crosslinking agent and the functional
groups on the crosslinked gelatin due to the crosslinking agent are
blocked with fibronectin
The gelatin film-coated vessel walls are placed in a chill 1 % aqueous
glutaraldehyde solution. The mixture is kept at ambient temperature
for one to 24 hours and with weak agitation. The fixed films are
washed several times with distilled water to eliminate the excess
glutaraldehyde. The absence of reagent in the floating matter
resulting from washing is checked by measuring the optical density at
280 nm (adsorption wavelength of glutaraldehyde).
The free functions of the glutaraldehyde on the fixed gelatin film are
then blocked with fibronectin. The films are incubated in PBS
containing 0.1 mg human plasma fibronectin (Collaborative Research,
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
27
Inc., Bedford, NIA). The solution is maintained at 20 - 37°C for
20
minutes to 2 hours. To eliminate the free excess fibronectin present in
the floating matter, the gelatin films are then washed (several times)
with distilled water.
As would be understood by one of ordinary skill in the art given the
present disclosure, other embodiments using core material coating-agents other
than
gelatin, such as, but not limited to, fibril collagens, fibrin and
hyaluronates or synthetic
polymers such as dextran and crosslinked fibronectin fragments, may be used
without
exceeding the scope or departing from the spirit of the invention. In
addition, other cell
adhesion molecules having fibronectin-like activities, such as laminin,
fibrin, heparin
and vitronectin (Collaborative Research, Inc., Bedford, MA) or their
fragments, can be
used as blocking agents for the crosslinked gelatin films. Accordingly, it is
to be
understood that this example disclosure is proffered to facilitate
comprehension of the
invention, and should not be construed to limit the scope thereof.
In accordance with one aspect of the invention, cancer cells and
hemopoietic cells associated with metastasis may be separated and analyzed
using the
following steps:
(a) blood or buffy coat are prepared as sources of cells,
(b) viable cancer cells and a fraction of normal cells are separated on a
cell separation vessel comprising the modified matrix film, and
(c) cancer and related normal cells are detected and total cells for each
type counted.
EXAMPLE 2.
Blood cell separation using the modified matrix film.
(a) Blood or huffy coat are prepared as sources of cells
Five to ten ml of blood are drawn from control subjects or patients
with a diagnosis of the presence of primary tumor or metastatic
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
28
cancer into a blood collection tube (Vacutainer, Becton Dickinson,
green top, each tube holds 7-ml) containing lithium heparin as an
anticoagulant. Blood or cells collected from an in vivo source are
subjected to cell isolation within a relatively short time after their
collection because the cells may lose their viability. In order to
maintain the optimal isolation of cancer cells, it is preferred that blood
or tissue samples are stored at 4 °C and used within 24 hours after
their collection, most preferably, within four hours.
Buffy coat is processed from blood by conventional density gradient
centrifugation using Ficoll-Paque (Pharmacia) that removes the
majority of red cells leaving a thin layer of nucleate cells, called buffy
coat, which may contain cancer cells of interest.
(b) Viable cancer cells are isolated on a cell separation vessel comprising
the modified matrix .
The buffy coat is washed, and the nucleate cells are suspended in the
complete cell culture medium, consisting of a 1:1 mixture of
Dulbecco's modified Eagle's medium (DMEM) and RPM11640
supplemented with 10% calf serum, 15% Nu-serum (Collaborative
Research, Inc., Bedford, MA), 2 mM L-glutamine, 0.1 mM
non-essential amino acids, 1 mM sodium pyruvate, 1 unit/ml penicillin,
and 10 ug/ml streptomycin. The cells are seeded onto a 6-cm tissue
culture plate (NUNC) that were coated with the gelatin film. The cell
culture is then incubated in COz cell incubator for 30 minutes to 2
hours, and is washed gently with PBS to remove non-adherent cells.
The adherent cells on the matrix film are then suspended with
trypsin/EDTA solution (GIBCO) for 5 minutes, followed by washing
with complete medium. Cells in the washes are the enriched cell
sample comprising cancer and a small fraction of hematopoietic cells
that are frequently related to metastasis.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
29
(c) Detection of cancer cells and total cell count
The enriched cell sample is used for detecting and counting the
number of viable cancer and small fraction of hematopoietic cells
related to metastasis using microscopic imaging or flow cytometry.
Detection of increasing number of viable cancer cells is an indicator of
cancer cells with metastatic potential, and increasing number of
hematopoietic cells isolated by the matrix is an indicator for host
immunity. The metastatic cancer cells may be identified by functional
assays described below.
EXAMPLE 3.
Identification of Viable Cancer Cells
(a) Colon~r formation
A portion of enriched nucleate cells, i.e., equivalent to 0.1-ml blood
volume per well, are seeded onto a 16-well microtiter plate-glass slide (in 96-
well
microtiter plate format; Lab-Tek, Rochester, NY) comprising tissue culture
medium
containing 10% heat-inactivated human plasma (complement-inactivated human
sera)
or plasma. Cells are allowed to propagate for four days to two weeks thereby
allowing
the cancer cells to form colonies. It was estimated that, among approximately
100
putative metastatic cells isolated from the blood of patients with metastatic
diseases,
there was only one colony of carcinoma cells formed after one week of culture.
The
efficiency of colony growth in culture appears to be 10,000 folds higher than
what was
observed in vivo, suggesting that, free of host immunity, cultured cancer
cells increase
their capability to grow.
(b) ~optosis and c~~tolysis
Cells are cultured for one day and stained prior to fixation using Vybrant
Apoptosis Assay Kit #5 Hoechst/prodidium iodide (V13244, Molecular Probes, OR,
USA).
Within one day after isolation using the cell separation technology of the
invention and
in culture, approximately 1,000 putative metastatic cells and 100,000
leukocytes are
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
isolated from one milliliter whole blood (containing approximately 10,000,000
nucleate
white cells and 1,000,000,000 red cells) of patients with metastatic diseases.
Viable
cancer cells are resistant to Hoechst staining of nucleic acids within the
cells, and do
not uptake prodidium iodide while apoptotic or lysed cancer cells are stained
with
5 Hoechst staining. All leukocytes become apoptotic, as indicated by strong
nuclear
Hoechst staining, and some cells disintegrate, as indicated by red fluorescent
prodidium
iodide in the cells.
EXAMPLE 4.
Identification of Metastatic Cancer Cells
10 (a) Intake of collagen and acet~rlated low densit~i lipoprotein
The intake of collagen or matrix fragments and that of acetylated low
density lipoprotein (acLDL) by circulating cancer cells is indicative that the
cells are
invasive, angiogenic and metastatic.
Enriched cells are seeded on rhodamine-labeled collagen coated on a
15 16-well glass slide (Lab-Tek, Rochester, NY). The cells are grown on the
labeled
collagen for 12 to 24 hours. The cells are then incubated with fluorescein-
conjugated
acLDL for 1 hour. The cells are then stained by nuclear staining with Hoechst
dye for
10 minutes. For measurement of the invasive phenotype of these cells, cells
were
analyzed for the ability of the cell to adhere to, degrade and ingest
rhodamine-collagen
20 substratum. Metastatic cells exhibit extensive collagen-degrading/ingestion
activities.
Metastatic cells also exhibit the intake of fluoresceinacLDL, suggesting their
role in
angiogenesis, a process of metastasis. Neither leukocytes nor monocytes and
endothelial cells exhibit these properties. Furthermore, immuno- and
morphological
features of metastatic cells are characteristic of carcinoma cells (see
below).
25 (b) Immunoytochemistry
For the determination of possible developmental lineages of cancer cells,
the enriched cells from blood of cancer patients are analyzed for their
potential
epithelial origin by immunocytochemistry using antibodies against epithelial
specific
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
31
antigen (ESA), epithelial membrane antigen (EMA; Muc-1), and cytokeratins 4,
5, 6, 8,
10, 13, and 18 (PCK). Commercial sources of antibodies for epithelial markers
include
mouse mAb recognizing human epithelial specific antigen (ESA; clone VU-1 D9,
. NeoMarkers, CA, USA; SIGMA, MS, USA), Muc-1 epithelial membrane glycoprotein
(Muc-1; clone E29, NeoMarkers, CA, USA), cytokeratins 4, 5, 6, 8, 10, 13, and
18 (PCK;
clone C- 11, SIGMA, MS, USA). Furthermore, immunocytochemical staining using
antibodies against endothelial markers, including CD31/PECAM-1 endothelial
cell
marker (CD31; Clone JC/70A, NeoMarkers, CA, USA), Flk-1, a receptor for
vascular
endothelial growth factor (Flk-1, Clone sc-6251, Santa Cruz, USA), VE-cadherin
endothelial marker (VE-cad; Clone sc 9989, Santa Cruz, USA); CD34 peripheral
blood
stem cell marker (CD34; clone 581, Pharmingen, USA), may be used to confirm
the
above observation that metastatic cells may process endothelial function. A
preferred
antibody staining is to use fluorescein conjugated antibodies against Muc-1
epithelial
marker (EMA, DAKO, Denmark) or fluorescein conjugates of goat antibodies
against
factor VIII endothelial marker (F8; Atlantic), in the above functional
labeling of cancer
cells with rhodamine-collagen fragments to demonstrating the presence of both
fluorescein- epithelial or endothelial markers (green fluorescence) and
ingested
rhodamine-collagen fragments (red fluorescence) in same cancer cells.
It was estimated that less than 1% of leukocytes and peripheral blood
monocytes derived from cancer patients are co-isolated by the cell separation
method
of this invention. These hematopoietic cells are determined by antibodies
directed
against CD14, CD68 and CD45 leukocyte common antigen (CD45; clone T29/33,
DAKO,
Denmark).
In addition to the use of fluorescent labelings described above, alkaline-
phosphatase-anti-alkaline-phosphatase (APAAP) method may be used to generate
signals for antibody labeling. This allows one to visualize the cancer cells
with their
markers, and the cell morphology, by a high-resolution interference
differential contrast
(DIC) microscopy.
In one preferred embodiment of the present invention enriched cells are
seeded on 16-well chambered glass slides (Lab-Tek, Rochester, NY) coated with
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
32
rhodamine-collagen. The seeded cells are cultured on the same substratum for
12 - 24
hours in a C02 incubator at 37~C. Prior to fixation for immunocytochemistry,
the cells
are incubated with fluorescein-conjugated acLDL for 1 hour, followed by
nuclear
staining with Hoechst dye for 10 minutes. After fixation with~3%
paraformaldehyde in
PBS, pH 7.2, for 10 minutes and after blocking nonspecific binding sites with
2% BSA
for 30 minutes, mouse primary antibodies, or fluorescein-F8 or -Muc-1 (when
fluorescein-conjugated acLDL is not involved) are applied to the slides. The
samples are
incubated for 20 minutes at room temperature, washed twice in PBS for 5 min,
and
then exposed to secondary rabbit anti-mouse Ig (Z0259, Dako) for another 20
minutes.
After two more washes, the samples are incubated with
alkaline-phosphatase-anti-alkalinephosphatase (APAAP) mouse Ig complexes for
15 min.
Finally, the enzyme-substrate [NewFuchsin (Dako)] is added, resulting in the
development of red precipitates at the cells of interest.
Data from the APAAP test may be recorded by numerous methods known
to those of ordinary skill in the art, including by way of a Nikon Eclipse
E300 inverted
light microscope, or automatically scanning prepared slides using a Rare Event
Imaging
System (Georgia Instruments, Inc. (Roswell, GA)), in conjunction with a SONY
DC5000
Cat Eye Imaging system . Data may be stored on a computer server or other
device
for later analysis. The Rare Event Imaging System employs image processing
algorithms to detect rare fluorescent events and determine the total number of
cells
analyzed. It is comprised of an advanced computer-controlled microscope (Nikon
Microphot-FXA, Nikon, Japan) with autofocus, motorized X-, Y-, and Z-axis
control,
motorized filter selection, and electronic shuttering. Images are taken by an
integrating, cooled CCD detector and processed in a computer imaging
workstation.
Most metastatic cells in an enriched cell sample react positively with ESA,
Muc-1 or PCK and typically are of epithelial origin. Metastatic cells
generally do not
react with markers for leucocytes or monocytes, are usually larger than
hematopoietic
cells, and typically assume a carcinoma cell morphology on collagenous
matrices.
Circulating carcinoma cells are rare in blood of most normal donors, patients
with
benign disease, and cancer patients undergoing conventional chemotherapy. In
the
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
33
blood of cancer patients who are undergoing chemotherapy, both circulating
cancer
cells and leukocytes are generally not reactive to the modified matrix of this
invention.
(c) Anal sii s by flow cytometr)r
In order to better enumerate single metastatic cells in the blood, an
enriched cell sample of the present invention can be analyzed by flow
cytometry
following a manufacture's procedure. Alternately, cell samples containing
individual
cancer cells and emboli (clumps) can be automatically measured using a micro
capillary
fluorescent measurement system that can detect signals of both single cells
and clumps.
Enriched cell samples may be stained for fluorescence sorting using
procedures similar to these used in immunocytochemistry described above. The
cells
are determined for the metastatic propensity and apoptosis or cytolysis by
cellular
labeling prior to fixation using rhodamine-collagen substratum, fluorescein-
acLDL and
Vybrant Apoptosis Assay Kit #5 Hoechst/prodidium iodide (V-13244, Molecular
Probes,
OR, USA). In addition, the enriched cells are stained for cell type markers in
a solution
containing fluorescein-antibodies against Muc-1 (DAKO), fluorescein-antibodies
against
factor VIII endothelial marker (F8; Atlantic), phycoerythrin (PE)-conjugated
anti-CD31
endothelial marker (Becton-Dickinson) and peridinin chlorophyll protein
(PerCP)-labeled
anti-CD45 (Becton-Dickinson) for 15 minutes. In general, three fluorescent
labelings
are applied to a given sample: including the rhodamine-collagen, a fluorescein-
, PE-, or
PerCP- labeled cell type signal, and a Hoechst staining. Briefly, the staining
procedure
involves incubation with fluorescent dye-antibody conjugates and washing. The
labeled
cells are re-suspended in 0.5 ml of a buffer and the sample is analyzed on a
FACScan
or FACSVantage flow cytometer (Becton Dickinson).
EXAMPLE 5.
Determination of efficiency of recovery of cancer cells using the modified-
matrix
The human amelanotic melanoma cell line LOX was obtained from
Professor Oystein Fodstad, Institute for Cancer Research, the Norwegian Radium
Hospital, Oslo, Norway, and the human breast carcinoma cell lines MDA-MB-436
and
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
34
Hs578T were obtained from American Type Culture Collection (Rockville, MD).
Cells
were cultured in a 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM)
and
RPMI 1640 supplemented with 10% calf serum, 5% Nu-serum (Collaborative
Research,
Inc., Bedford, MA), 2 mM L-glutamine, 0.1 mM non-essential amino acids, 1 mM
sodium pyruvate, 1 unit/ml penicillin, and 10 ug/ml streptomycin.
LOX human malignant melanoma cells are tagged with a fluorescent dye,
PKH26 Red Fluorescent Cell Linker (Sigma), to determine the efficiency of
recovery of
cancer cells using the cell separation procedure of this invention.
Fluorescent-tagged
LOX cells were cultured on fibronectin-coated crosslinked gelatin films for
one day,
suspended and counted the fluorescently labeled cells using a hemocytometer.
They
were serially diluted and spiked into complete medium alone, and in parallel
into the
blood of a control normal donor. Graded doses of LOX cells were seeded into 1
mL
volumes of whole blood and complete medium, respectively, that were in 12-well
culture plates that were coated with crosslinked gelatin films, and incubated
for two
hours. After washing with complete medium and PBS, the adherent cells were
suspended by trypsin/EDTA (GIBCO). The enriched cell samples were further
seeded
onto a 16-well glass slide (Lab-Tek, Rochester, NY), cultured for over three
hours, and
counted by fluorescence microscopy for the number of fluorescent LOX cells in
each
well. Samples were analyzed for the number of cancer cells per well and
related to the
total cell count per milliliter of blood.
The efficiency of recovery of fluorescent-tagged LOX cells from whole
blood using the modified-matrix described above and cell separation procedure
of this
invention is shown in Table 1 below. Viable cancer cells were detected in
blood samples,
which initially contained as few as one cancer cell/mL (in three trials of the
one cell
experiment, two had detected one cell in the well). The result suggests that
the level
of sensitivity by the cell separation method is at 1 viable cancer cell per mL
of blood.
The_ recovery of viable cancer cells spiked into 1 mL of blood (10 -20 million
nucleate
white cells and one billion red cells) from a normal donor, as compared with
complete
medium was consistent over a frequency range, from 63.3% to 89.9% at all
cancer cell
doses, and has an average recovery of 75.9%. It appears that high cell density
in
whole blood does not significantly affect the efficiency of the procedure. The
average
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
recovery rate (75.9%) can be used to estimate the number of viable cancer
cells in the
circulation.
Table 1. EfFcieny of recover!/ of LOX cells from whole blood using the
cell separation procedure
LOX cells/ml blood LOX cells/ml medium % Cells recovered
8,545 9,834 86.9
2,193 2,440 89.9
1,054 1,213 86.9
547 612 89.4
248 299 82.9
61 83 73.5
28 44 63.6
13 20 65.0
7 11 63.6
4 6 66.7
2 3 66.7
0 0
Average = 75.9
5
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
36
EXAMPLE 6.
Use of isolated ovarian cancer cells for discovery of molecular markers and
therapeutic targets for ovarian cancer
(a) Purification of ovarian metastatic cancer cells
Ovarian cancer cells may be purified using a combination of
function-affinity cell separation of the invention and immuno-affinity
purification. Fig.
5shows the scheme of such cell separation from bodily tissues, such as tumors,
ascitic
fluid or blood. Tumor and adjacent normal tissues optimally should be obtained
immediately after surgical removal and digested with collagenase for 1 hour at
37°C to
yield a suspension of single cells and clumps (Step 51a). Ascites or salvaged
blood
samples, such as blood harvested from patients undergoing abdominal surgery
for
resection of primary cancers, preferably are removed of red blood cells by
density
gradient centrifugation procedure (Step 51b).
The first positive selection for purifying viable cancer cells involves
passing tumor and ascites cell suspensions through the function-affinity
matrix (Step
52). Briefly, cell suspension samples may be passed through a cell filtration
system
wherein pores of the pre-filter and channels of the filter contain materials
that under
specific conditions are sufficient to specifically bind cancer cells and
emboli. The
pre-filter and the filter are washed with PBS to remove unbound cells. The
cells and
emboli bound to the pre-filter and the filter may be released from the matrix
by
treating filter channels with proteolytic enzymes such as trypsin/EDTA, and
the
enriched cell sample collected. The enriched cells and other purified cells
may be
quantified, for example, using a hemocytometer.
The enriched cell samples may be further enriched by subjecting the
sample to a negative selection procedure. For example, a cocktail of anti-CD14
and
anti-CD45 immuno-magnetic beads (Dynal) may be used to remove hematopoietic
cells
as well as cancer cells binding non-specifically to the magnetic beads (Step
53).
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
37
The further enriched cell samples preferably are then subjected to a
second positive selection procedure involving antibody-affinity purification.
For example,
the epithelial cells remaining in the cell suspensions may be isolated by
binding to
anti-BerEP4 immunomagnetic beads (Dynal) (Step 54), the BerEP4 antibody
recognizing a pan-epithelial antigen present on normal and neoplastic
epithelium but
not present on hematopoietic or stromal cells (U. Latza, G. Niedobitek, R.,
Schwarting,
H. Nekarda, H. Stein, 1990. J. Clin. Pathol. 43, 213). Importantly, the BerEP4
bound
epithelial cells in ascites and blood express endothelial markers including
factor VIII,
CD31, and receptor for acetyl LDL, but BerEP4 bound primary tumor cells do
not. Thus,
the cancer cells in ascites and blood are also isolated by their binding to
anti-CD31
immuno-magnetic beads.
Isolated cells may be lyzed and RNA/DNA isolated for further analysis. A
portion of "metastatic" cancer cells isolated, as for example, from ascites
and blood
may also be cultured in a collagenous matrix (Step 55) for less than two days
to give
rise to a "metastasized" cell population mimicking cancer cells grown in
micrometastases. Other steps (Steps 56 and 57), as would be known to those of
ordinary skill in the art, could be performed to further improve the purity of
the
metastasized cell population.
In genetic studies, short-term cultures of ovarian surface epithelial cells
may be used as the control normal epithelial cell group. Ovarian cancer-
derived cell
lines, SK-OV-3 [American Type Culture Collection (ATCC) HTB-77], MDAH-2774
(ATCC
CRL-10303), and CAOV-3 (ATCC HTB-75), may be obtained from the ATCC and grown
in DMEM (Life Technologies, Rockville, MD), supplemented with 10% (vol/vol)
FCS and
penicillin/streptomycin. In addition, levels of gene expression of the above
three
cancer-cell types may be compared with those of stroma (fibroblastic) cells or
leukocytes and monocytes to rule out potential normal cell-contamination in
the
cancer-cell preparation. The results of such comparison may be used to help
discern
patterns of gene expression that are consistent with cancer progression and
development of the metastatic phenotype.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
38
(b) Microarray hybridization
Total RNA from the ovarian cancer cells isolated may be prepared with a
Qiagen RNeasy mini-kit according to the manufacturer's instructions (Step 58).
RNA
may be hybridized separately to large microarrays containing 16,000 human
genes
(Affymetrix; U95A). Arrays may be scanned using an Affymetrix confocal scanner
and
analyzed initially, for example, using GeneChip 3.1 (Affymetrix) as set forth
below.
(c) Microarray data analy
Microarray scanned image files may be visually inspected for artifacts and
analyzed with GeneChip 3.1 (Affymetrix) and GeneSpring 4.0 software (Silicon
Genetics). Each image may be scaled to an average hybridization intensity of
200,
which corresponds to approximately 3-5 transcripts per cell. The expression
level
(average difference) for each gene may be determined by calculating the
average of
differences in intensity (perfect match-mismatch) between its probe pairs.
Genes with
average hybridization intensities <0 across all samples may be excluded from
further
analysis. GeneSpring 4.0 software (Silicon Genetics) is used to select, group,
and
visualize genes whose expression varied across the samples with SD>_250.
Hierarchical
clustering of the samples and gene expression levels within the samples may be
used
to lead to the unambiguous separation of normal, primary tumor and malignant
cells,
as well as the identification of three subsets of ovarian cancer cell samples,
i.e.,
primary, metastatic and metastasized.
To identify potential tumor markers, the hybridization intensity of each
gene in normal and malignant cell samples may be compared, and three different
estimates for population differences (difference of means, fold change, and
unpaired t
test) may be applied in parallel. The genes are ranked according to each
metric, and
the sum of the metrics was used to derive a semipuantitative estimate of the
differential abundance of each transcript. Four categories of potential genes
encoding
molecular markers are:
(i) For a gene to be selected as enhanced during extravasation, it has to
be expressed in all "metastasized" cell samples (BerEP4+ or CD31 +
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
39
cancer cells from ascites or blood samples with culture) at least 5
times higher than in all "metastatic" cell samples (BerEP4+ or CD31 +
cancer cells from ascites or blood samples without culture), with
experiments done in duplicate.
(ii) For a gene to be selected as enhanced during intravasation, it has to
be expressed in all "metastatic" and "metastasized" cell samples
(BerEP4+ or CD31 + cancer cells from ascites or blood samples with
and without culture, respectively) at least 5 times higher than in the
tumor cell samples, with experiments done in duplicate.
(iii)For a gene to be selected as enhanced during ovarian cancer
progression, it has to be expressed in all tumor, "metastatic" and
"metastasized" cell samples at least 5 times higher than in the normal
epithelial and stoma culture samples, with experiments done in
duplicate.
(iv)For a gene to be selected as enhanced and as a specific marker for
ovarian cancer, it has to be among genes fit in category (iii) above,
and expressed at least 5 times higher than in all ascites "metastatic"
and "metastasized" cell samples of endometrioma or other cancer,
with experiments done in duplicate.
(d) Validation of cell-specific gene expression
Microarray results of differential expression of genes may be validated in
at least three distinct ways. First, fragments of genes of interest may be
amplified by
RT-PCR from the RNAs of distinct cell types in triplicates to determine the
overexpression of specific genes in specific cell types. Second, the National
Center for
Biotechnology Information (NCBI) "gene-to-tag" databases, available through
UniGene
(htt~//www.ncbi.nim.nih.gov/UniGene/), for gene expression patterns of these
same
three genes in tumor cells and tissues may be queried. LU and HE4 are
typically
highly expressed in primary ovarian tumors, as well as in other tumors and
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
micrometastases. Third, quantitative large-scale analysis of gene expression
in different
cancer cell types may be performed using real-time RT-PCR as described below.
(e) Real-time RT-PCR
To validate and extend previous findings of genes differentially expressed
5 in ovarian tumor tissues, real-time RT-PCR, a highly sensitive and
reproducible
technique, may be chosen, preferably using robotic means, in validation of a
potential
set of markers for diagnostic and prognostic applications for treating
patients with
ovarian cancer (Hough et al., 2001). This method allows highly quantitative
analysis of
gene expression on a large number of specimens. In addition, it requires a
relatively
10 low amount of RNA, typically less than 1 pg. Real-time RT-PCR does not
require large
amounts of starting RNA in each purified cell type, and it can measure levels
of gene
expression of 32 RNA samples at one time. This approach would allow an
accurate
determination of the frequency and extent of overexpression of many genes
relevant to
ovarian cancer. The approach may also take advantage of genes selected from
the vast
15 screen assay of DNA microarray. Such genes may be tested stringently to
determine
those genes that are consistently and highly upregulated in a set of over 100
well-defined cancer cells from ovarian cancer in order to determine the
"ovarian cancer
gene cassettes" that are useful in diagnostic and prognostic applications for
treating
patients with ovarian cancer. Furthermore, real-time RT-PCR is feasible to be
used in
20 measuring the genuine up-regulated ovarian cancer genes in 5mL blood of any
patients
who are in high risk of developing cancer.
In a typical procedure, one picogram of total RNA from each sample is
used to generate cDNA using the Taqman reverse transcription reagents (PE
Applied
Biosystems, Foster City, CA). Mock template preparations are prepared in
parallel
25 without the addition of reverse transcriptase. Quantitative PCR is
performed with an
iCycler (Bio-Rad, Hercules, CA) using Pico Green dye (Molecular Probes,
Eugene, OR),
and threshold cycle numbers are obtained using iC~cler software v2.3.
Representative
conditions for amplification are: one cycle of 95°C, 2 min followed by
35 cycles of 95°C,
15 sec, 58°C, 15 sec, and 72°C, 15 sec. Quantitative PCR
reactions are typically
30 performed in triplicate and threshold cycle numbers averaged. RT-PCR
products should
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
41
meet two criteria to be included in a study: (1) the signal from the reverse
transcriptase (RT)-derived cDNA should be at least 100 fold greater than that
of control
reactions performed without reverse transcriptase, and (2) the PCR products
from the
reactions with RT should be the expected size upon gel electrophoresis. Gene
expression may be normalized to that of beta-actin, a gene that is uniformly
expressed
in all ovarian cells as assessed by DNA microarray.
(f) Methods for Preparation and Analysis of DNA
Genomic DNA may be prepared from purified epithelial cells using the
Qiagen DNA Easy Purification Kit (Qiagen). Preferably at least six independent
replicates on each DNA sample are performed in order to assess gene copy
number.
Real-time PCR may be carried out as described above for the expression
analysis,
except that the control reactions should be carried out without any genomic
DNA
template. Appropriate primers may be used to design primers for genomic PCR,
such
as Primer 3 (http://www.genome.wi.mit.edu/cqi-bin/primer/primer3www.cqi).
Representative conditions for amplification are: one cycle of 95°C, 2
min followed by
35 cycles of 95°C, 15 sec, 58°C, 15 sec, 72°C, 15 sec.
EXAMPLE 7.
Functional proteomics studies to aid in diagnosis of metastatic phenotype
and in monitoring chemotherapy effect
Genomic methodologies described in EXAMPLE 6 provide significant
information about gene structure and expression as well as other events such
as
splicing. However, the vast array of posttranslational modifications and
surface
localization commonly observed in proteins cannot be studied or be predicted
accurately. Proteomic techniques are a solution to definitively study
posttranslational
modifications of abundant proteins but they alone have restricted value in
understanding surface localization and interaction of minor functional
proteins such as
enzymes (Mann et al., 2001). To enrich minor proteins of specific function,
advanced
separation methodologies far isolating specific cell types as described above,
membrane structures or protein complexes must be used in combination with
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
42
sophisticated proteomic technology (Bell et al., 2001; Mann et al., 2001;
Pawson and
Scott, 1997).
2-D DIGE (Differential In Gel Electrophoresis)-mass spectrometry system
(Amersham Pharmacia Biotech) may be used in conjunction with the function-
based,
cell separation method of the invention to facilitate the studies on molecular
structures
underlying the metastatic phenotype. To identify structure of membranes, an
invadopodia membrane separation method (Mueller et al., 1999) may be used to
isolate invadopodia proteins. To further enrich protein complexes of interest,
affinity-based purification can be performed using immobilized antibody
against the
epitope, followed by competitive elution with peptide encoding the epitope as
described,
for example, in Mann et al., 2001. By proteomic analysis on a defined cell
product
exhibiting the metastatic phenotype, the targeted proteins and their
endogenous
inhibitors could be identified.
(a) Determination of whether natural substrates and inhibitors associated
with seprase exist as enzyme-substrate complexes at invadopodia
The 2-D DICE-mass spectrometry system (Amersham Pharmacia Biotech)
may be used for the identification of: (a) natural substrates (inhibitors) of
a cell surface
enzyme involved in cell invasion (a member of invadopodia proteases), called
seprase,
and (b) novel proteins associated with the enzyme in physiological complexes.
Recent
data from membrane purification and immunoprecipitation experiments suggest
the
existence of invadopodia complexes that contain seprase and form the
structural basis
for expression of the metastatic phenotype. However, there are many proteins
involved, including proteases, their substrates in degradative process,
integrins, kinases,
cytoskeletal and signal molecules in their isoforms.
The 2D profiles of seprase-associated proteins in the presence of seprase
inhibitors (experimental) and the absence of protease inhibitors (control) may
be
compared and analysed. For example, approximately 109 LOX human malignant
melanoma cells that express seprase and other invadopodia antigens are lysed
in 0.15
M NaCI, 4% CHAPS, 30 mM Tris, pH 8.5. Proteins associated with seprase are
immunoprecipitated using monoclonal antibodies directly conjugated on Agarose
beads.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
43
A pair of immunoprecipitates are incubated at 37 °C in the presence
(experimental) of
and the absence (control) of seprase enzymatic inhibitors (5.0 mM ABESF, and
300 pM
H-Ile-Pro-NHO-pNB). Proteins are then eluted from the column with 6 M urea, 4%
CHAPS, 30 mM Tris, pH 8.5 (4 °C) that contain the epitope peptide, and
concentrated
by ultra filtration with MW 5,000 cut-off to a total protein concentration of
approximately 0.5 mg/rnl. These two samples are the control and experimental
groups,
and are labeled with 2 different CyTM dyes developed for DIGE that do not have
apparent alteration on electrical mobility of proteins, and resolved by 2D gel
electrophoresis followed by mass spectrometric identification. Among
approximately
400 analytic spots, approximately 50 spots are showing more than a 2-fold
increase,
and approximately 50 spots showing more than a 2-fold decrease are picked and
their
peptide sequences analyzed using MALDI MS to indicate the identity of putative
natural
substrates for seprase. Similarly, form those spots (among approximately the
400
analytic spots) that show less than a 2-fold change and that show sharp spot
match,
approximately 50 best-fitted spots may be picked and their peptide sequences
analyzed
using MALDI MS to indicate the identity of putative associated proteins for
seprase that
are involved in cell invasion.
The 2-D DICE system is high throughput and ideal for complex analysis.
In a preliminary study described above, immuno-affinity purified proteins
derived from
malignant human melanoma cells (LOX cell line) detect 428 analytical spots in
a 2D gel,
suggesting the feasibility of using the 2D DICE system in such a complex
analysis.
Interestingly, tenascin-X was identified as a potential natural substrate for
seprase as it
has 5-fold higher amount in the seprase-complex treated with seprase
inhibitors.
(b) The monitoring of ovarian cancer therapy using functional proteomic
studies
In order to determine the efficacy of ovarian cancer therapy, the 2D
profiles of cell-matrix contact membranes (including invadopodia that invade
in the
collagenous film) derived from metastatic cells described above, in the
presence of
therapeutic agents ex vivo (experimental) and in the absence of therapeutic
agents ex
vivo (control), may be compared and analyzed. For example, the experimental
group
may comprise cells cultured ex ~ivo in the presence of conventional
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
44
Taxol/carboplatinum chemotherapy (Taxol 175 mg/m2 over 3 hours, carboplatinum
AUC=7.5) or experimental therapeutics such as angiogenesis-MMPI (AG3340
Agouron/Warner-Lambert; Bay 12-9566 Bayer; or Marimastat British Biotech),
while
the control group may comprise cells cultured in the absence of therapeutic
agent.
A 2-D DICE and mass spectrometric analysis may be performed on
proteins from both experimental and control groups. Briefly, cell membranes
(invadopodial membranes) in contact with a collagenous matrix are isolated and
membrane proteins are partitioned into Triton X-114 according, for example, to
the
method described in Mueller et al., 1999. This procedure can be used to
produce
invadopodial membranes having 51% purity as determined by morphometry and
immuno-labeling, and 122-fold enrichment over the membranes for the
invasiveness
marker seprase. The control and experimental membrane proteins are cyedyed
with 2
different dyes and run upon a single 2D gel. Approximately 50 spots from those
that
show the highest increase, and approximately 50 spots showing highest decrease
in
the comparison, are picked and their peptide sequences analyzed using MALDI MS
to
indicate the identity of proteins associated with expression of the malignant
phenotype.
Resulting major membrane proteins are used to assess the overall proteomic
profiling
and to correspond to known invadopodia residents, proteases (seprase and MTi-
MMP),
and integrins, a3~1 and a5~il, for cell surface proteolytic cascades and
integrin
signaling pathway, respectively. The goal is to resolve functional proteomics
of cancer
malignancy by correlating the identification and analysis of invadopodia
proteins to the
function of genes or proteins. This approach may be used to provide
information that
may help develop targeted therapeutic agents.
Description of the Fi ures
Now turning to the figures, referring to Figs IA and IB there is shown a
general schematic of a contained cell separation system (10) comprising a
vacuum
blood collection tube (11) in which whole blood (12) may be stored. Vacuum
blood
collection tube (11) is coated along its inner surface of the tube's walls
(15) with a
modified-matrix (14) comprising core material, such as inert glass and
polymeric
materials, such as magnetic colloid, polystyrene, polyamide material like
nylon,
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
polyester materials, cellulose ethers and esters like cellulose acetate, an
intermediate
coating about the core material comprising material that has the affinity, or
efficiently
binds to another material having the affinity, to bind blood-borne adhesion
components
that promote the adhesion of cancer cells (such as fibronectin, fibrin,
heparin, laminin,
5 tenascin, vitronectin, or their fragments), such as gelatin, collagens,
fibrin, dextran and
hyaluronate, and blood-borne adhesion components of natural or synthetic
origin.
Referring to Fig. 2A there is shown a general schematic of a cell
separation system (20) incorporating cell separation beads (22) in a vacuum
blood
collection tube (21). As seen in Fig. 2A, a conventional mesh filter (23) is
incorporated
10 in a blood collection tube to facilitate the washing and the collection of
cell-bound
beads. Alternatively, the mesh filter (23) can be an independent unit outside
the tube.
In this case, after incubation of the blood-bead mixture, the mixture can pull
into the
mesh filter (23) outside the tube to facilitate the washing and the isolation
of viable
cancer cells from whole blood of patients with metastatic diseases. As seen in
Fig, 2B,
15 when incorporated into a vacuum blood collection tube (21) the mesh filter
(23)
collects the beads.
Referring to Figs .3A and 3B there is shown a general schematic of a cell
separation system (30) comprising a vacuum blood collection tube (31)
incorporating
microbeads or nanoparticles (32), preferably having a diameter in the range of
20 nm
20 to 20 microns, having an intermediate magnetic coating (34) which is
attracted to a
magnetic source (33). The beads (32) are typically suspended in blood (35)
containing an anticoagulant such as lithium heparin. After incubation of the
blood-bead
mixture in the tube on shaker with slow speed at 37°C for 30 minutes to
2 hours, the
sample tube is passed through a magnetic field using a magnetic separator (33)
to
25 magnetically immobilize microbeads or nanoparticles in the sample having
cells bound
thereto. This provides a gentle means of washing and collection of cell-bound
microbeads or nanoparticles. The microbeads or nanoparticles can capture
viable
cancer cells and related tissue cells.
Referring to Fig. 4A there is shown a general schematic of a cell filtration
30 system (40) comprising a filter housing (48) having an inlet (41) in the
housing for the
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
46
introduction of the blood (46) to be filtered; an outlet (42) in the housing
for the
removal of filtered blood and to return to a patient; and a filter element
(47) disposed
within the housing, which element comprises a plurality of beads (44),
preferably
having a diameter in the range of 200 microns to 1,000 microns, which beads
are
packed tightly between two layers of meshes (43), preferably having mesh
opening
widths of 50 to 200 microns. The core beads are held back by the meshes. Thus,
the
filter system allows it to be back-washed with wash liquid.
Fig. 4B is an expanded view of the portion of cell separation beads
designated in Fig. 4A, depicting the anastomosic channels formed by the cell
separation
beads within the inner confinement area and the size and nature of the core
bead (44)
to be employed. Core beads (44) are coated with an intermediate coating (45)
(such
as gelatin, collagens, fibrin, hyaluronates and dextran) comprising material
that has the
affinity, or efficiently binds to another material having the afi'Inity, to
bind blood-borne
adhesion components that promote the adhesion of cancer cells (such as
fibronectin,
fibrin, heparin, laminin, tenascin, vitronectin, or their fragments) as well
as to bind to
the core bead substrate.
In a blood filter, the core material selected must be inert and compatible
with the blood, and should be somewhat stiff to adhere well to the
intermediate
coating (45). Typically materials which may be employed in a blood filter
would
include, but not be limited to: inert polymeric materials, such as
polystyrene, polyamide
material like nylon, polyester materials, cellulose ethers and esters like
cellulose acetate,
urethane foam material, DEAE-dextran, as well as other natural and synthetic
materials,
such as other foam particles, cotton, wool, dacron, rayon, acrylates and the
like. The
core material is preferably a polyester, such as a 40 mil 3 denier natural
polyester, or
non-porous polystyrene plastic. Preferably, in a blood filter, the size of the
core
materials of beads should not typically exceed about 1 mm, or preferably 200
microns
to 1,000 microns in diameter. The core bead material is preferably sorted by
their sizes.
In blood filters, if the beads are too big, they can compact at high flow
rates and less
channel surface areas, and, therefore, be less efficient. The nature of the
core material
to be selected is such that the beads may adhere to the intermediate coating,
such as
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
47
gelatin, collagens, fibrin, hyaluronates and dextran, and create a smooth
anastomosic
channel within the filter for blood flow.
As seen in Fig. 4B, a modified matrix (45) is coated on the surfaces of
packed beads (44) and mesh openings to create anastomosic channels of
tremendous
surface areas for contacting cells in the fluid. The matrix-coated lining of
channels
selectively remove viable and aggressive cancer cells and related tissue cells
from the
blood to be filtered. The filter of this invention is sterile, non-toxic and
non-leaking of
proteins or particles into blood flow.
Fig, 5 in a schematic representation of a method of the present invention
providing for the isolation of viable cancer cells and related tissue cells
from tissue
samples using a combination of the function-affinity cell separation of the
invention and
immuno-affinity purification. Fig. Sis described in detail in EXAMPLE 6.
References
AI-Mehdi, A. B. et al. 2000. Intravascular origin of metastasis from the
proliferation of
endothelium-attached tumor cells: a new model for metastasis. Nature Medicine.
6,
100-102.
Asahara,T., Murohara,T., Sullivan,A., Silver,M., Van der Zee,R., Li,T.,
Witzenbichler,B.,
Schatteman,G., and Isner,J.M. (1997). Isolation of putative progenitor
endothelial cells
for angiogenesis. Science 275, 964-967.
Beitsch,P.D. and Clifford,E. (2000), Detection of carcinoma cells in the blood
of breast
cancer patients. American Journal of Surgery 180, 446-448.
BeII,A.W., Ward,M.A., Blackstock,W.P., Freeman,H.N.M., Choudhary,J.S.,
Lewis,A.P.,
Chotai,D., FazeI,A., Gushue,J.N., Paiement,J., PaIcy,S., Chevet,E.,
LafreniereRoula,M.,
Solari,R., Thomas,D.Y., RowIey,A., and Bergeron,J.J.M. (2001). Proteomics
Characterization of Abundant Golgi Membrane Proteins. Journal of Biological
Chemistry
276, 5152-5165.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
48
Beltinger,C.P. and Debatin,K.M. (1998). A simple combined microdissection and
aspiration device for the rapid procurement of single cells from clinical
peripheral blood
smears. Molecular Pathology 51, 233-236.
Brugger,W., Bross,K.J., Glatt,M., Weber,F., Mertelsmann,R., and Kanz,L.
(1994).
Mobilization of tumor cells and hematopoietic progenitor cells into peripheral
blood of
patients with solid tumors [see comments]. Blood 83, 636-640.
Brugger,W., Buhring,H.J., Grunebach,F., VogeI,W., KauI,S., Muller,R.,
Brummendorf,T.H., ZiegIer,B.L., Rappold,l., Brossart,P., Scheding,S., and
Kanz,L.
(1999). Expression of MUC-1 epitopes on normal bone marrow: implications for
the
detection of micrometastatic tumor cells. J. Clin. Oncol. 17, 1535-1544.
Brugger,W., Heimfeld,S., Berenson,R.J., Mertelsmann,R., Kanz, and L. (1995).
Reconstitution of hematopoiesis after high-dose chemotherapy by autologous
progenitor cells generated ex vivo [see comments]. New England Journal of
Medicine
333, 283-287.
Campana, D. et al. (1995). Detection of minimal residual disease in acute
leukemia:
methodologic advances and clinical significance, Blood 85:1416-34.
Chen, W.-T., and S.J. Singer. 1980. Fibronectin is not present in the focal
adhesions
formed between normal cultured fibroblasts and their substrata. Proc. Nat.
Acad. Sci.
U.S.A.77:7318-7322.
Chen, W.-T., J.-M. Chen, S.J. Parsons, and J.T. Parsons. 1985. Local
degradation of
fibronectin at sites of expression of the transforming gene product pp60src.
Nature 316:
156-158.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
49
Chen, W.-T., K. Olden, B. A. Bernard, F. F. Chu. 1984. Expression of
transformation-
associated protease(s) that degrade fibronectin at cell contact sites. J. Cell
Biol.
98:1546-1555.
Chen,W.-T. and Singer,S.J. (1980). Fibronectin is not present in the focal
adhesions
formed between normal cultured fibroblasts and their substrata. Proc. Natl.
Acad. Sci. U.
S. A. 77, 7318-7322.
Chen,W.-T., Yeh,Y., and Nakahara,H. (1994). An in vitro cell invasion assay:
determination of cell surface proteolytic activity that degrades extracellular
matrix. J.
Tiss. Cult. Meth. 16, 177-181.
Clark,R.A., Folkvord,J.M., and Wertz,R.L. (1985). Fibronectin, as well as
other
extracellular matrix proteins, mediate human keratinocyte adherence. Journal
of
Investigative Dermatology 84, 378-383.
Cote et al., 1991. Prediction of Early Relapse in Patients With Operable
Breast Cancer
by Detection of Occult Bone Marrow Micrometastases, Journal of Clinical
Oncology, vol.
9, No. 10, pp. 1749-1756.
Dicke et al., 1968. The selective elimination of immunologically competent
cells from
bone marrow and lymphatic cell mixtures," Transplantation 6:562-570.
Dicke et al., 1970, °Avoidance of acute secondary disease by
purification of
hemopoietic stem cells with density gradient centrifugation," Exp. Hematol.
20:126-130.
Durrant et al., 1992, A rapid method for separating tumor infiltrating cells
and tumor
cells from colorectal tumors" J. Immunol. Meth. 147:57-64.
Ellis et al., 1984, "The use of discontinuous Percoll gradients to separate
populations of
cells from human bone marrow and peripheral blood," J. of Immunological
Methods
66:9-16.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
Fidler, LJ. (1973). The relationship of embolic homogeneity, number, size and
viability
to the incidence of experimental metastasis. European Journal of Cancer
(Oxford) 9,
223227.
5
Folkman,J. (1995). Seminars in Medicine of the Beth Israel Hospital, Boston.
Clinical
applications of research on angiogenesis. N. England J. Med. 333, 1757-1763.
Giordano,T.J., Shedden,K.A., Schwartz,D.R., Kuick,R., Taylor,J.M.G., Lee,N.,
Misek,D.E.,
10 Greenson,J.K., Kardia,S.L.R., Beer,D.G., Rennert,G., Cho,K.R., Gruber,S.B.,
Fearon,E.R.,
and Hanash,S. (2001). Organ-Specific Molecular Classification of Primary Lung,
Colon,
and Ovarian Adenocarcinomas Using Gene Expression Profiles. Am J Pathol 159,
1231-1238.
15 Glaves, D., 1983. Correlation between circulating cancer cells and
incidence of
metastasis. Br. J. Caneer, 48:665-673.
Glaves,D., Huben,R.P., and Weiss,L. (1988). Haematogenous dissemination of
cells
from human renal adenocarcinomas. British Journal of Cancer 57, 32-35.
Griwatz et al., 1995. An immunological enrichment method for epithelial cells
from
peripheral blood, Journal of Immunological Methods, vol. 183, pp. 251-265.
Gross et al, 1995. Moody study detecting breast cancer cells in peripheral
blood
mononuclear cells at frequencies as low as 10(-7). Proc. Natl. Aca. Sci. USA.
92:537.
Gulati, S C et al. (1993) Rationale for purging in autologous stem cell
transplantation.
Journal of Hematotherapy, 2:467-71.
Herzenberg et al., 1979, Fetal cells in the blood of pregnant women: detection
and
enrichment by fluorescence-activated cell sorting. Proc. Natl. Aca. Sci. USA
76: 1453-5.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
51
Hough,C.D., Cho,K.R., Zonderman,A.B., Schwartz,D.R., and Morin,P.J. (2001).
Coordinately Up-Regulated Genes in Ovarian Cancer. Cancer Research 61, 3869-
3876.
Karczewski,D.M., Lema,M.J., and Glaves,D. (1994). The efficiency of an
autotransfusion
system for tumor cell removal from blood salvaged during cancer surgery.
Anesthesia &
Analgesia 78, 1131-1135.
Kraeft,S.K., Sutherland,R., Gravelin,L., Hu,G.H., Ferland,L.H., Richardson,P.,
Elias,A.,
and Chen,L.B. (2000). Detection and analysis of cancer cells in blood and bone
marrow
using a rare event imaging system. Clinical Cancer Research 6, 434-442.
Liotta,L.A., Kleinerman,J., and SaideI,G.M. (1974). Quantitative relationships
of
intravascular tumor cells, tumor vessels, and pulmonary metastases following
tumor
implantation. Cancer Research 34, 997-1004.
Mann,M., Hendrickson, R.C., and Pandey,A. (2001). ANALYSIS OF PROTEINS AND
PROTEOMES BY MASS SPECTROMETRY. Annual Review of Biochemistry 70, 437473.
Mueller,S.C., Ghersi,G., Akiyama,S.K., Sang,Q.X., Howard,L., Pineiro-
Sanchez,M.,
Nakahara,H., Yeh,Y., and Chen,W.-T. (1999). A novel protease-docking function
of
integrin at invadopodia. J. Biol. Chem. 274, 24947-24952.
Pawson,T. and Scott,J.D. (1997). Signaling Through Scaffold, Anchoring, and
Adaptor
Proteins. Science 278, 2075-2080.
Peck,K., Sher,Y.P., Shih,J.Y., ROffIer,S.R., Wu,C.W., and Yang,P.C. (1998).
Detection
and quantitation of circulating cancer cells in the peripheral blood of lung
cancer
patients. Cancer Res. 58, 2761-2765.
Racila,E., Euhus,D., Weiss,A.J., Rao,C., McConneII,J., Terstappen,L.W., and
Uhr,J.W.
(1998). Detection and characterization of carcinoma cells in the blood.
Proceedings of
the National Academy of Sciences of the United States of America 95, 4589-
4594.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
52
Sabile,A., Louha,M., Bonte,E., Poussin,fC., Vona,G., Mejean,A., Chretien,Y.,
Bougas,L.,
Lacour,B., Capron,F., Roseto,A., Brechot,C., and Paterlini-Brechot,P. (1999).
Efficiency
of Ber-EP4 antibody for isolating circulating epithelial tumor cells before RT-
PCR
detection. American Journal of Clinical Pathology 112, 171-178.
SU,A.I., WeISh,J.B., Sapinoso,L.M., ICern,S.G., Dimitrov,P., Lapp,H.,
SchuItZ,P.G.,
PoweII,S.M., Moskaluk,C.A., Frierson,H.F., Jr., and Hampton,G.M. (2001).
Molecular
Classification of Human Carcinomas by Use of Gene Expression Signatures.
Cancer
Research 61, 7388-7393.
Suarez-Quian,C.A., Goldstein,S.R., Pohida,T., Smith,P.D., Peterson,J.l.,
Wellner, E,
Ghany,M., and Bonner,R.F. (1999). Laser capture microdissection of single
cells from
complex tissues. BioTechniques 26, 328-335.
Ts'o,P.O., Pannek,J., Wang, Z.P., Lesko,S.A., Bova,G.S., Partin, and AW.
(1997).
Detection of intact prostate cancer cells in the blood of men with prostate
cancer.
Urology 49, 881-885.
Vona,G., Sabile,A., Louha,M., Sitruk,V., Romana,S., Schutze,K., Capron,F.,
Franco,D.,
Pazzagli,M., Vekemans,M., Lacour,B., Brechot,C., and Paterlini-Brechot,P.
(2000).
Isolation by Size of Epithelial Tumor Cells : A New Method for the
Immunomorphological and Molecular Characterization of Circulating Tumor Cells.
Am J
Pathol 156, 57-63.
Wang,Z.P., Eisenberger,M.A., Carducci,M.A., Partin,A.W., Scher, HI, and
Ts'o,P.O.
(2000). Identification and characterization of circulating prostate carcinoma
cells.
Cancer 88, 2787-2795.
Weiss, L. and Glaves,D. (1983). Cancer cell damage at the vascular
endothelium.
[Review] [62 refs]. Annals of the New York Academy of Sciences 416, 681-692.
CA 02463467 2004-04-08
WO 03/035888 PCT/US02/11749
53
WeIsh,J.B., Zarrinkar,P.P., Sapinoso,L.M., Kern,S.G., Behling,C.A., Monk,B.J.,
Lockhart,D.J., Burger,R.A., and Hampton,G.M. (2001). Analysis of gene
expression
profiles in normal and neoplastic ovarian tissue samples identifies candidate
molecular
markers of epithelial ovarian cancer. PNAS 98, 1176-1181.
