Note: Descriptions are shown in the official language in which they were submitted.
~ WO95/07105 21 712 J~ I PCT~S94110217
HAPLOTYPE-MATCHED CYTOKINE-SECRETING CELLS AND METHODS OF USING TO
STIMULATE AN IMMUNE RESPONSE
BACKGROUND OF THE lNV~.lION
FIELD OF THE lNvhN~l~loN
This invention relates generally to the fields of
gene therapy and immunotherapy of cancer and, more
specifically, to a composition comprising haplotype-matched
tumor cells which have been genetically modified to express
a cytokine gene and to a composition comprising autologous
fibroblasts which have been genetically modified to express
a cytokine gene and unmodified tumor cells. The invention
also relates to methods for using the haplotype-matched
cytokine-secreting cells to stimulate an immune response
against a tumor located in the central nervous system of a
cancer patient. The invention further relates to a method
of inhibiting or preventing the growth of tumor cells in
the central nervous system of a patient by ;mmlln;zing the
patient with the haplotype-matched cytokine-secreting
cells.
BACKGROUND INFORMATION
Cytokines are immune system modulators that
mediate many of the immune responses involved in anti-tumor
;mmlln;ty. Several cytokines have been produced using
recombinant DNA methodology and evaluated for their
effectiveness in treatlng cancer patients. The
~m;nistration of lymphokines, which are cytokines produced
by lymphocytes, and related immunomodulators has produced
positive responses in patients with various types of
neoplasms. However, ~m; n; stration of cytokines frequently
is associated with toxicity, which limits the therapeutic
value of these agents.
WO95/07105 ~ 2~ } PCT~Sg~/10217
Interleukin-2 (IL-2) is a lymphokine having a
central role in the generation of anti-tumor ; mmlln; ty
(Rosenberg et al., J. Natl. Canc. Inst., 10:73-77 (1990)).
In response to tumor antigens, helper T cells secrete small
quantities of IL-2, which acts locally at the site of a
tumor antigen to activate cytotoxic T cells and natural
killer cells. The latter cells effect systemic tumor cell
destruction.
Intravenous, lntralymphatic and intralesional
10 ~m; ni stration of high doses of IL-2 have produced
clinically significant responses in some cancer patients.
However, severe toxicity, including hypotension and edema,
limit the dose and, therefore, the usefulness of
intravenous and intralymphatic IL-2 ~m; n; stration (see
Lotze et al, J. Amer. Med. Assn., 256:3117-3124 (1986);
Pizza et al., Lymphokine Res., 7:45-48 (1988~; Gandolfi et
al., Hepato-Gastroenterology, 36:352-356 (1989); Sarna et
al., J. Biol. ResP. Mod., 9:81-86 tl990~). The toxicity of
systemically ~m; n; stered lymphokines is not surprising as
these agents mediate local cellular interactions and
normally are secreted in very small quantities.
Other cytokines, such as interleukin-4,
interleukin-7, colony st mulating factors, alpha-interferon
and gamma-interferon also have been used to stimulate
immune responses to tumor cells (see, for example, Hock et
al., J. Exp. Med., 174:1291-1298 (1991); Thomassen et al.,
Canc. Res., 51:857-862 (1991), each of which is
incorporated herein by reference). Like IL-2, however,
current modes of administration of these cytokines cause
adverse side effects to the patient.
To circumvent the toxicity of systemic cytokine
in;stration, the effectiveness of intralesional
injection of IL-2 has been investigated. While this
approach el; m; n~tes the toxicity associated with systemic
~ WO95/07105 PCT~S94/10217
,, 2171211
IL-2 administration, multiple intralesional injections are
required to optimize therapeutic efficacy (Bubenik et al.,
Immunol. Lett., 19:279-282 (1988); Bubenik et al., Immunol.
Lett., 23:287-292 (1989/1990)). Thus, these injections are
impractical for many patients, particularly when tumor
sites are inaccessible for injection or create a risk of
patient morbidity.
An alternative approach to cytokine therapy
involves inserting cytokine genes into tumor cells. Using
this method, the expression of cytokine gene products
following cytokine gene transfer into the tumor cells
abrogates the tumorigenicity o~ the cytokine-secreting
tumor cells when implanted into syngeneic hosts. The
transfer of genes encoding ~L-2 (Fearon et al., Cell, 387-
403 (1990); Gansbacher et al., J. Exp. Med., 172:1217-1224
(1990)), gamma-interferon (Watanabe et al., Proc. Natl.
Acad. Sci., USA, 86:9456-9460 (1989)), interleukin-4
(Tepper et al., Cell, 57:503-512 (1989)) or granulocyte-
macrophage colony stimulating factor (Dranoff et al., Proc.
Natl. Acad. Sci. USA, 90:3539-3543 (1993)) significantly
reduces or eliminates the growth of several different
histological types of murine tumors.
In the studies employing tumor cells genetically
modified to express IL-2, treated ~n;m~l S rejected the
cytokine-secreting tumor cells and developed systemic
immllnity against the tumor cells and were protected against
subsequent tumor cell cha'lenge with unmodiied tumor cells
(see Fearon et al., l99C, and Gansbacher et al., 1990).
Similar inhibition of tumor growth and protective immllnity
30 also was demonstrated when ;mmlln;zations were performed
with a mixture of unmod-fied parental tumor cells and
genetically modified tumor cells, which expressed IL-2. No
toxicity associated with localized lymphokine transgene
expression was reported in these ~nim~l tumor studies (see
Fearon et al., 1990; Gansbacher et al., 1990; and Tepper et
WO95/07105 ~ PCT~S9~/10217
~ 4
al., 1989; see, also, Kriegler et al., Gene Transfer and
Expression: A laboratory manual (Stoc~ton Press 1990),
which is incorporated herein by reference).
It was estimated that 15,600 new cases of primary
brain and CNS tumors would occur in 1990 (Ransohoff et al.,
Cancer of the Central Nervous System and the Pituitary, in
American Cancer Society Textbook of Clinical OncoloqY,
Holleb et al., ed., Chap. 24 (American Cancer Society, Inc.
1991)). In addition, many non-CNS primary tumors
metastasize to the CNS. As the lifespan of humans
increases, the number of cancer patients having CNS
involvement is expected to increase.
The prognosis for brain cancer patients is poor.
For example, of the estimated 12,000 new cases of primary
CNS tumors in 1987, 1,100 deaths were pro~ected in 1990
(Ransohoff et al. 1991). Furthermore, all patients with
primary CNS tumors of the glioma series will eventually die
of their disease. Although there is no clear line of
demarcation, tumors of the glioma series are generally
classified as benign astrocytomas, anaplastic astrocytomas
or glioblastoma (referred to herein generally as "gliomas")
(Ransohoff et al. 1991).
For example, gliomas currently are treated by
surgery, radiotherapy or chemotherapy, either alone or in
various combinations (Levin et al., Int. J. Rad. Oncol.
Biol. PhYs., 18:321-324 (1990)). In addition,
investigational treatments have been utilized, including
local administration of I~-2 in combination with autologous
lymphokine activated killer cells (Ransohoff et al. 1991).
However, the effectiveness of this method has not been
established and, in any case, suffers from the morbidity
problems associated with localized injection of such a
formulation at the site O r the tumor as discussed above.
WO95/07105 ~ ~ PCT~S94/10217
5 ?17l2ji
In view of the problems associated with localized
treatment in the CNS using, for example, cytokine-secreting
tumor cells or lymphocyte activated killer cells, it would
be advantageous if an active immune response could be
induced in a site other than the CNS, yet still provide an
effect at the site of the tumor. However, it is unclear
whether the induction of an active immune response outside
the CNS will provide a therapeutic advantage against a
tumor within the CNS. For example, the CNS is considered
to be a partially immunologically privileged site (Oldfield
et al., Hum. Gene Ther., 4:39-69 (1993)), suggesting that
immune effector cells are not active across the blood-brain
barrier. This "immunologic privilege" makes it merely
speculative that immlln;zation with genetically modified
tumor cells will induce an immune response against the
tumor cells within the CNS. In fact, a recent report by
Ram et al. (Abstract, presented at the Meeting of the Amer.
Assn. Neurosurg., Boston, MA, 1993) indicates that an
immune response against tumor cells that were genetically
modified to express the IL-2 gene does not occur in the
CNS. Using the rat brain tumor model, 9L, Ram et al.
genetically modified the 9L tumor cells by introducing a
gene encoding IL-2. When the IL-2-secreting tumor cells
were injected peripherally, the majority of ~n;m~ls did not
form tumors. However, when the IL-2-secreting 9L cells
were injected in the CNS, tumors were observed to grow in
the rat CNS. These results in the CNS are in contrast to
the results reported in other tumor systems using tumor
cells genetically modified to express IL-2 (see, for
example, Fearon et al., l990, and Gansbacher et al., l990).
These results indicate that the CNS is an immunologically
privileged site and that results using cytokine gene
therapy in other tumor systems are not predictive of the
therape~tic value of such therapy in treating CNS tumors.
Mahaley et al., J. Neurosurq., 59:201-207 (1983),
;mm~ln;zed glioma patients with subcutaneous injections of
W095/07105 ~ PCT~S94/10217
~ ~ 6
allogeneic human glioma tissue culture cell lines. The
patients also were treated with levamisole and with
radiotherapy and BCNU chemotherapy. Patients inoculated
with the U-251MG cell line had significantly longer
survival compared to non-imml~n;zed historical control
patients treated with levamisole, radiotherapy and
chemotherapy. However, while the results of Mahaley et al.
appeared promising, more recent studies using similar
methods have produced equivocal results (Bullard et al.,
Sem. Oncol., 13:94-109 (1986); Frank and Tribolet,
Neurosurg. Rev., 9:31-37 (1986)).
In an ~ n i m~ 1 study of gliomas, immune rat
cytotoxic T cells primed ex vivo against a rat glioma were
A~m; n; stered intravenously to rats with intracerebral
glioma and induced regression of these brain tumors
(Holladay et al., J. Neurosurg., 77:757-762 (1992);
Holladay et al., Neurosurq., 30:499-504 (1992); Holladay et
al., Neurosurq., 31:528-533 (1992), each of which is
incorporated herein by reference). Thus, passive
;mmlln;zation of glioma-bearing rats provided a therapeutic
advantage in this study. However, due to the partially
immunologically privileged location of a CNS tumor, methods
for using active immunotherapy to inhibit or prevent the
growth of primary tumors of the CNS and metastatic lesions
within the CNS have not yet been developed.
Thus, there exists a need for an effective means
to stimulate the immune response of a patient having a
primary CNS tumor or metastatic lesions located within the
CNS, such that the patient's immune system can prevent or
inhibit the growth of the tumor cells. The present
invention satisfies this need and provides, in addition,
related advantages.
WO95/0710~ 1 71 21 PCT~S94/10217
SUMMARY OF THE lNv~NlION
The invention provides a composition comprising
haplotype-matched tumor cells which have been genetically
modified to express a cytokine gene and to a composition
comprising autologous fibroblasts which have been
genetically modified to express a cytokine gene and
unmodified tumor cells. The invention also provides
methods for using the haplotype-matched cytokine-secreting
cells to stimulate an im~une response against a tumor
located in the central nervous system of a cancer patient.
The invention further relates to a method of inhibiting or
preventing the growth of tumor cells in the central nervous
system of a patient by ;mmnn;zing the patient with the
haplotype-matched cytokine-secreting cells.
A cytokine expressed by a haplotype-matched
genetically modified cell is secreted at the ;mmlln;zation
site in an effective amount, which is defined as a level of
cytokine that is sufficient to induce or augment a systemic
anti-tumor immune response. The haplotype-matched
genetically modified cell can be a tumor cell, which
contains the appropriate tumor antigen required to induce
an immune response. In addition, the haplotype matched-
genetically modified cell can be an autologous fibroblast,
in which case tumor antigen is provided by including
unmodified tumor cells in the composition used to ;mmlln;ze
the patient. Tmmlln;zation can be either at the site of the
tumor in the CNS or at a site other than the CNS. An
effective amount of cytokine secretion does not result in
unacceptable patient toxicity because the level of secreted
cytokine does not signiflcantly affect systemic cytokine
concentrations.
Since the amount of cytokine secreted by the
haplotype-matched genetically modified cells is sufficient
to induce anti-tumor imml~nity but is too low to produce
W095/07105 ' ~i~ PCT~S9V10~17
unacceptable patient toxicity, the present approach
provides the benefit of localized cytokine ~m; n; stration
without producing undesirable side effects. Furthermore,
the continuous localized expression of a cytokine at the
site of ;mmllnization more effectively augments an immune
response against the patient's tumor cells as compared to
intermittent cytokine injections. The disclosed invention
also provides the advantage of localized imm11n;zation with
the haplotype-matched genetically modified cells and,
therefore, avoids the necessity of cumbersome intravenous
infusions required for immunotherapy with cells activated
ex vivo such as expanded population of tumor infiltrating
lymphocytes (see, for example, Rosenberg et al., New Enql.
J. Med., 323:570-578 (l990).
BRIEF DESCRIPTION OF THE DRAWING
Figure ~ shows schematic diagrams of retroviral
vectors DC/ADA/R/IL-2 and GlNaCvi2.23.
DETAILED DESCRIPTION OF THE lwv~NLION
The invention provides a composition comprising
haplotype-matched tumor cells which have been genetically
modified to express a cytokine gene and provides a
composition comprising autologous fibroblasts which have
been genetically modified to express a cytokine gene and
unmodified tumor cells. The invention also provides
methods for using the haplotype-matched cytokine-secreting
cells to stimulate an immune response against a tumor
located in the central nervous system of a cancer patient.
The invention further provides a method of inhibiting or
preventing the growth of tumor cells in the central nervous
system of a patient by imm11n;zing the patient with the
haplotype-matched cytokine-secreting cells.
WO95/~71~5 1211 PCT~59~/10217
The invention contemplates, in part, the
stimulation of a patient's im.mune response against a
primary CNS tumor or metastatic lesions located within the
CNS by (a) obt~; n; ng tumor cells having a haplotype which
is matched to the patient's haplotype; (b) introducing into
the haplotype-matched tumor cells a gene encoding a
cytokine such as IL-2 and, if desired, a suicide gene such
as the herpes simplex virus thymidine kinase gene (HSV-TK),
wherein the cytokine gene product is expressed and secreted
in an effective amount by the haplotype-matched tumor cells
and wherein expression of the suicide gene can be induced
if desired; (c) if desired, irradiating the tumor cells so
as to prevent the cells from proliferating in vivo; and (d)
;mm1ln;zing the patient with the haplotype-matched cytokine-
secreting tumor cells, such that expression and secretionof the cytokine gene product stimulates the patient's
immune response but does not produce unacceptable patient
toxicity.
The invention further contemplates the
stimulation of a patient's im~une response against a
primary CNS tumor or metastatic lesions located within the
CNS by (a) obt~; n; ng autologous fibroblasts, which
inherently have a haplotype that is matched to the
patient's haplotype; (b) introducing into the autologous
fibroblasts a gene encoding a cytokine such as IL-2 and, if
desired, a suicide gene such as the herpes simplex virus
thymidine kinase gene (HSV-TK), wherein the cytokine gene
product is expressed and secreted in an effective amount by
the cytokine-secreting fibroblasts and wherein expression
of the suicide gene can be induced if desired; (c)
obtaining tumor cells, which provide a source of tumor
antigen, (d) irradiating said tumor cells so as to prevent
the cells from proliferating in vivo; and (e) ;mmlln;zing
the patient with the cytokine-secreting fibroblasts and the
irradiated unmodlfied tumor cells, such that expression and
secretion of the cytokine gene product stimulates the
WO95/0710~ 2 ~ . PCT~S94/10217
1~,' 10
patient~s immune response but does not produce unacceptable
patient toxicity.
As used herein, "gene" means a nucleotide
sequence encoding a desired gene product such as a cytokine
or an active fragment of a protein or peptide having
cytokine activity. A "gene product," therefore, is a
protein or a peptide, wherein the protein or peptide may be
an active fragment of the protein or peptide as it is
normally expressed in a cell. As used herein, an "active
fragment" means that the peptide or protein has cytokine
activity. Such activity can be readily determined using
assays well known in the art and described herein.
The invention provides haplotype-matched cells
which have been genetically modified to express a cytokine
gene. As used herein, "haplotype-matched" means that a
genetically modified cell such as a tumor cell and the
patient being treated share one or more major
histocompatibility locus hap'otypes. For example, if it is
determined that a patient with a glioma expresses the major
histocompatibility locus HLA-A2 haplotype, the patient will
be ;mmtln;zed with HLA-A2 glioma cells that have been
genetically modified to express and secrete a cytokine gene
product. The haplotype of the patient can be readily
determined using methods wel' known in the art.
Haplotype-matched tumor cells can be autologous
or allogeneic. For example, the cytokine-secreting cells
can be autologous fibroblasts or tumor cells obtained from
the patient. The autologous cells, which are grown in
tissue culture and genetica'ly modified, inherently are
haplotype-matched to the patient.
In addition, since various HLA-A haplotypes are
known to be present in the human population, a panel of
genetically modified tumor cells can be created. A panel
~ WO9~0710~ 71211 ~CT~S94/l~Zl7
of such allogeneic tumor cells can express, for example,
the various different HLA-A haplotypes present in a
population. In addition, various panels can represent
tumors of different histologic origin such as glioma,
neuroblastoma and other primary CNS tumor cells as well as
other non-CNS tumors such as lung carcinoma, breast
carcinoma, melanoma and other tumors that metastasize to
the CNS. Thus, the lnvention provides haplotype-matched
cytokine-secreting cells useful for im~lln;zing cancer
patients expressing various haplotypes and having various
types of tumors in the CNS.
As used herein, the term "genetically modified"
means that the haplotype-matched cells have been subjected
to recombinant DNA techniques such that the cells can
express and secrete a cytokine gene that has been
introduced into the tumor cells. Methods for introducing
a cytokine gene into a cell are well known in the art and
described below.
For ;mmlln;zation~ the tumor cells are of the same
histologic origin as the patient's tumor. Tumor cells
having a desired haplotype can be obtained from established
allogeneic cells lines or can be autologous cells obtained
from the patient to be treated. Where the tumor cell is
obtained from a patient, the tumor cells will be grown in
culture using methods well known by one skilled in the art
of tissue culture. For example, methods for culturing
primary human glioblastoma cells have been described by
Bigner et al., J. Neuropathol. Exp. Neurol., 40:201-229
(1981), which is incorporated herein by reference. If
- 30 desired, the cells can then be genetically modified using
methods described herein or well known in the art.
Alternatively, the tumor cells can remain unmodified and
can be injected with cytokine-secreting fibroblasts to
stimulate an immune response in a patient.
W095/07105 ~ 2l~ PCT~594/10~17
Numerous cytokine genes have been cloned and are
available for use in th-s protocol. For example, the genes
encoding various interieukins, gamma-interferon and
granulocyte-macrophage colony stimulating factor are
available from the American Type Culture Collection (see
ATCC/NIH Repository Catalogue of Human and Mouse DNA Probes
and Libraries, 6th ed., '9g2). In addition, genes encoding
cytokines, including interleukin-6, granulocyte colony
stimulating factor and human stem cell factor, are
available commercially (Amgen, Thousand Oaks, CA; see, for
example, Patchen et al., Exptl. Hematol., 21:338-344 (1993)
and Broudy et al., Blood, 82:436-444 (1993), each of which
is incorporated herein by reference~. Similarly, gene
encoding various isoforms of TGF-B, including TGF-B1, TGF-
~2, TGF-~3, TGF-B4 and TGF-B5, also are available to those
in the art.
In addition, selectable marker genes such as the
neomycin resistance tNeoR) gene are available commercially
and the use of such selectable marker genes is described,
for example, in Sambrook et al., 1989. Incorporation of a
selectable marker gene allows for the selection of tumor
cells that have successfully received and express a desired
gene.
A suicide gene can be incorporated into a
haplotype-matched genetically modified tumor cell to allow
for selective inducible k-llina of the tumor cell after
stimulation of the immune response. As used herein, a
"suicide gene" means a gene, the expression of which can
result in the death of the cel' expressing the suicide gene
when the cell is exposed to certain drugs. An example of
a suicide gene useful in the invention is the HSV-TK gene.
A tumor cell induced to express a transferred HSV-TK gene
is selectively killed when exposed to a drug such as
acyclovir or gancyclovir. A suicide gene also can be a
gene encoding a non-secreted cytotoxic polypeptide. A
WO95/07105 7rl 21~ PCT~S94/10217
13
suicide gene can be attached to an inducible promoter and,
when destruction of a haplotype-matched cytokine-secreting
tumor cell is desired, an agent that induces the promoter
can be A~mi n; stered such that expression of the cytotoxic
polypeptide kills the haplotype-matched cytokine-secreting
tumor cell. However, destruction of a haplotype-matched
cytokine-secreting tumor cell is not m~n~Atory and may not
be desired.
Numerous methods are available for introducing a
nucleic acid sequence into a cell in vitro ( see Kriegler et
al., l9gO, and Sambrook et al., 1989). For example, an
appropriate nucleic acid sequence can be inserted into an
expression vector such as a plasmid or a viral vector,
which is introduced into a cell using methods well known in
the art such as transfection, transduction, electroporation
and lipofection. Examples of useful viral vectors include
adenovirus and adeno-associated vectors (see, for example,
Flotte, J. Bioenerq. Biomemb., 25:37-42 (1993) and
Kirshenbaum et al., J. Clin. Invest, 92:381-387 (1993),
each of which is incorporated herein by reference).
Vectors are particularly useful when the vector contains a
promoter sequence, which can provide constitutive or
inducible expression of a cloned nucleic acid sequence.
Such vectors are well known in the art (see, for example,
Methods in EnzymologY, Vol. 185, D.V. Goeddel, ed.
(Academic Press, Inc., 1990)) and available from commercial
sources (eg., Promega, Madison, WI).
An effective method for transferring a gene or
other nucleic acid sequence into a cell is by using
retroviral gene transduction. When retroviruses are used
for gene transfer, replication competent retroviruses
theoretically can develop by recombination between the
retroviral vector and viral gene sequences in the packaging
cell line utilized to produce the retroviral vector.
However, packaging cell lines in which the production of
WO95/07105 ~ 2 11 PCT~S9~/10217
14
replication competent v rus by recombination has been
reduced or el;minAted can be used. In any case, all
retroviral vector supernatants used to infect patient cells
can be screened for replication competent virus by stAn~Ard
assays such as PCR and reverse transcriptase assays (see,
for example, Rosenberg et al., New Enal. J. Med., 323:570-
578 (1990), which is incorporated herein by reference).
Retroviral vectors useful for expressing a
cytokine can be constructed using methods well known in the
art. For example, a retroviral vector expressing an IL-2
gene product, DC/A~/R/IL-2, was described by Gansbacher et
al., Canc. Res., 50:7820-7825 (1990); Gansbacher et al.,
Blood, 80:2817-2825 (1992); Gastl et al. Canc. Res.,
52:6229-6236 (1992), each of which is incorporated herein
by reference (see Figure 1). In addition, a cytokine-
expressing retroviral vector, designated GlNaCvi2.23, was
obtained from Genetic Therapy, Inc. (Gaithersburg, MD; see
Figure 1).
Prior to ;mmllnization~ the tumor cells can be
irradiated so as to preven' the tumor cells from
proliferating in vivo. Approximately 106 to 107 genetically
modified cytokine-secreting cells are required for each
;mmlln;zation. The number of cells, however, can be
adjusted so as to provide a sufficient number of cells to
secrete an effective amount of the cytokine. As used
herein, an "effective amount" of a cytokine is an amount
that induces the patient's immune response without
producing unacceptable toxicity in the patient. For
example, in the first patient treated using the disclosed
method, transient erythema at ;mmlln;zation sites and tumor
necrosis were not observed until the IL-2 dose exceeded 100
unit/24 hours. Thus, therapy can be initiated with
transduced tumor cells that secrete this dose of IL-2.
Since transduced cells typically secrete approximately 20-
40 units of IL-2/105 cells/24 hours, initial ;mmlln;zation
WO95/07105 1 712¦1 PCT~S94/10217
requires injection of approximately 5 x 106 genetically
modified cytokine-secreting ce~ls. The appropriate number
of cytokine-secreting cells along with unmodified tumor
cells, if required, can be injected subcutaneously,
intramuscularly or in any manner acceptable for
~ i mmlln; zation.
A nucleic acid sequence of interest also may be
introduced into a haplotype-matched cell using methods
which do not require the initial introduction of the
nucleic acid sequence into a vector. For example, a
nucleic acid comprising a cytokine gene and a selectable
marker can be introduced into a cell using a cationic
liposome preparation (Morishita et al., J. Clin. Invest.,
91:2580-2585 (1993), which is incorporated herein by
reference). In addition, a nucleic acid can be introduced
into a haplotype-matched cell using, for example,
adenovirus-polylysine DNA complexes (see, for example,
Michael et al., J. Biol. Chem., 268:6866-6869 (1993), which
is incorporated herein by reference). Other methods of
introducing a nucleic acid sequence into a tumor such that
a gene contained within the nucleic acid can be expressed
are well known and described, for example, in Methods in
EnzYmoloqy, Vol. 185, l9g0).
The following examples are intended to illustrate
but not limit the scope of the invention.
EXAMPLE I
PREPARATION OF HAPLOTYPE-MATCHED
GENETICALLY MODIFIED CELLS
This example illustrates the methods used to
culture glioblastoma cells and genetically modify the cells
such that the tumor cells express and secrete a cytokine
gene product.
WO95/0710S r ~ ; PCT~S94/10217
~ 6
Establishment of Primar~ Human Glioblastoma Cell Lines
Methods have been developed that permit the
establishment of over 50% of primary glioblastoma tumors in
continuous cultures sultable for retroviral gene transfer
(Bigner et al., 1981). Briefly, the patient's tumor was
obtained from a clinically indicated surgical resection,
minced and placed in Richter's zinc option me~; ~ . An
aliquot of the cells was centrifuged, washed in Richter's
zinc option media, then cryopreserved as a "back-up"
culture in a solution cont~in;ng 10% dimethylsulfoxide and
50% fetal calf serum. A portion of the established tumor
cell line was expanded for transduction with the IL-2
retroviral vector and for application in immune response
monitoring assays.
The glioblastoma cell culture was prepared by
transferring the tumor tissue to a 60 mm tissue culture
plate and resecting from the normal brain and necrotic
tissue a sample of "pure" tumor using sterile forceps,
scissors and scalpel. The selected tumor pieces were diced
with sterile scissors into the smallest pieces possible.
Four ml of 0.4% collagenase n serum-free medium cont~;n;ng
gentamicin 50 ~g/ml was added to the tumor tissue in the
tissue culture plate, which was then incubated for 1-4
hours at 37 C in a C02 incubator. (For larger tumor
samples, the tumor pieces are placed in a lO0 mm tissue
culture plate and 8 ml of medium is added, as described
above). The plates were checked hourly and the sample was
worked up and down in a pipette to encourage dissociation
and to assess the optimal time for further processing.
When the tumor cells were freely dissociated, the
entire sample was trans~erred to a 50 ml tube for
centrifugation. The plates were rinsed with serum-free
medium to collect all cells. Centrifugation was performed
in an IEC PR6 centrifuge at lO00 rpm for five minutes. The
WO95/07105 2171~ PCT~S94/10217
supernatant was aspirated and the pellet resuspended in
Richter's zinc option culture medium with gentamicin 50
~g/ml in an amount appropriate to distribute the cells into
the number of dishes adequate to accept 7 x 106 cells/100 mm.
dish with 10 ml of medium containing 20% fetal calf serum
as described (Bigner et al., 1981). The cells then were
incubated at 37 C in a 5~ C02 incubator. Unless the
culture medium became extremely acidic, the original medium
was not changed before 48 hours to allow the cells to
attach. Gentamicin-free medium was used for subsequent
medium changes. As the tumor cells reached confluency,
they were detached from the plates using trypsin and
passaged at low split ratios of 1:1 or 1:2 until the
cultures grew ade~uately. The HLA-A2 glioblastoma cell
line obtained was designated GT9.
Cytogenetic and other cell line characterization
studies are performed to identify, for example, p53, PDGF,
EGFR and TGF-~ genotypes and phenotypes. These studies are
performed within the first 72-96 hours to determine the
presence of malignant cells and are repeated at intervals
of 20-30 passage levels and at the 70th passage level as
the tumors are established. To avoid chromosome breaks
from ordinary fluorescent light bulbs, Westinghouse F40G0
(Gold) bulbs are used in the l~m; n~r flow hoods and cell
culture rooms in which the cultures are being established.
Panels of genetically modified tumor cell
vaccines can be prepared using HLA-typed primary glioma
cell cultures as described by Bigner et al. (1981). The
cell panels can represent several different histologic
- types of tumor cells and can express HLA-A2 or HLA-A1 loci,
which are expressed by approximately 40% and 25% of the
North American population respectively. The availability
of this panel of tumor cells having various haplotypes
affords the opportunity to develop genetically modified
whole cell vaccines matched at these loci for a significant
WO 95/07105 2~ PCT/US91/10217
18
proportion of the North American population. Studies in
melanoma patients have indicated that the ~LA-A2 locus is
a do~;nAnt haplotype for tumor antigen presentation which
can meA; ~te MHC-restricted tumor destruction by cytotoxic
T cells (Crowley et al., Canc. Res., 50:492 (1990); Crowley
et al, J. Immunol., 146:1692-1699); Pandolfini et al.,
Canc. Res., 51:3164-3170 (1991)). Thus, autologous HLA-A2
tumor cells such as GT9 that have been genetically modified
to express IL-2, for example, can be used to stimulate the
immune response of a significant fraction of glioma
patients.
Preparation of Primary Cultures of Autoloqous Fibroblasts:
Primary cultures of autologous cultures can be
obtained using methods well known in the art. Fibroblasts
can be obtained from a skin punch biopsy.
Transduction of the PrimarY ~uman Glioblastoma Tumor Cells
With an IL-2-cont~;n;nq Retroviral Vector:
Standard retroviral gene transfer methods were
used to transduce the glioblastoma cultures with the IL-2
retroviral vectors. Cultured tumor cells (5 x 104 cells/10
cm plate) were incubated with supernatant from the
appropriate packaging cell line in the presence of
polybrene (8 mg/ml) as described by Xu et al., Virology,
171:311-341 (1989) and by Miller and Rosman, BioTechniques,
7:980 (1989), each of which is incorporated herein by
reference. After 24 hr, the cells were washed, then
cultured in medium cont~in;ng 100-150 yg/ml of the neomycin
analogue, G418, to select for transduced cells. The cells
then were cultured for 48 hr in DMEM supplemented with 10%
fetal calf serum (FCS). Transfected cells were selected
10-14 days after selection with G418 was begun. The G418
resistant cells were tested for IL-2 gene expression by
measuring IL-2 in the culture supernatant using the ELISA
~ WO95107105 21 7I21 1 PCT/US94/10217
19
assay described below. Aliquots of the G418 resistant
cells were stored at -70 C until required for
; mmll n;zations.
Similar methods were used to tranduce autologous
fibroblasts obtained from the patient.
Measurement of IL-2 Expression:
Transduced cell culture supernatants were
analyzed ~or IL-2 secretion levels employing commercially
available enzyme linked immunosorbent assay (ELISA) kits
cont~in;ng antibodies specific for human IL-2 (Genzyme or
T Cell Sciences). Briefly, 96-well plastic microtiter
plates coated with the primary antibody were incubated with
the test sample, washed, then incubated with the
appropriate secondary antiserum con~ugated to peroxidase or
alkaline phosphatase. The enzymatic reaction was developed
using a chromogen substrate and the optical density read on
a micro-ELISA plate reader. These kits contain
substitution control antibodies and standard IL-2 solutions
of known concentration to permit quantitation of IL-2
levels.
EXAMPLE II
IMMUNIZATION OF A PATIENT WITH AUTOLOGOUS
IL-2-SECRETING CELLS
This example illustrates the effectiveness of
treating a human patient with autologous glioma cells which
have been genetically modified to express and secrete IL-2
and with a combination of autologous fibroblasts which have
been genetically modified to express and secrete IL-2 and
autologous irradiated, unmodified tumor cells.
WO95/07105 ` PCT~S9~/10217 ~
2~ 2~ 20
Patient History:
A glioblastoma multiforme (GBM) patient was
treated with IL-2 gene therapy. The patient is a 52 year
old female with GBM of the right temporal lobe diagnosed
in December 1992. She was initially treated with surgical
resection, conventional radiotherapy and PCV chemotherapy
(procarbazine, CCNU and vincristine). Nine months later,
a second resection was performed for tumor recurrence.
Tumor pathology revealed a GBM at re-resection. The
patient's tumor progressed after experimental treatment
with accutane and with Iodine-131-labeled anti-tenacin
monoclonal antibody. Subsequently, the patient was treated
with experimental stereotactic radiation therapy designed
to encompass the site of tumor involvement.
IL-2 Gene Therapy:
IL-2 gene therapy was initiated in January 1993,
approximately one year after the first tumor resection.
The patient received nine subcutaneous ;mmlln;zations at 2
to 4 week intervals with either autologous, irradiated IL-2
transduced tumor cells (GT9 cells, as described in Example
I) or a mixture of irradiated unmodified tumor cells and
IL-2-transduced fibroblasts. The treatment protocol i5
shown in Table 1.
Two IL-2 retroviral vectors were employed in this
study. The retroviral vector, DC/AD/R/IL-2, utilized an
adenosine d~Am;n~se promoter to drive IL-2 expression
(Figure 1; see, also, Gansbacher et al. 1990, 1992; Gastl
et al., 1992). The retroviral vector GlNaCvi2.23 employed
a cytomegalovirus promoter (Figure 1; Genetic Therapy,
Inc., Gaithersburg, MD).
Table 1 lists the transduced cell types and IL-2
doses ~m;nistered for each ;mmlln;zation. IL-2 secretion
PCT/US94/10217
~WO95/0710~ 7t2I1 21
I ible 1 ~2 GENE T~ERAPY OF GLIOBLASTO~IA
TREAT~IENT SU~IMARY
D~te o~ UYI~; # of Tûtsl
~,.. ,.. "",,.. ~ n VcctûrTrD~ nr~d Tr:ln~r~uced Il.-2 Secretion Dose o~2
Rs ~ Cdl T,vpeCells10~j ce~/24 hr
114/93 Rs t~ 1 DC/AD~Rm,2 tumor1.25 s 10a 43 units j um~s
Lt27t93 R~ 1~ 2 DC/AD/Rm~2 tumor2.60 s 10a 10 unus 3 uni~
2ngtg3 R~ # 3 DCIADIRIII,-2 fibrobl;lstS.60 s 10a 40 units 1~ u nitt
315193 Rs i~4 DC/AD/Rm,l fibrobl:~stS20 s 10a 51 units 21 u~
3126193 R~ t~5 DClADtRlIL 2 fibroblast2.25 :c 10a 3S units 8 nnits
S/16193 Rs~6DC/ADJRm,2 -' lobl~120 s 10a91 units 11 uni~
SJ12t93 Rs ~7 GlN~CYi2.`23 fibrobl&st1 2_a0 s loa 427 units 107 uni~
~127t93 Rs~8GINaCYi2.23 mrobiast' 320 ~ loa328 untts12auniu
tumor2.00 s 10~1 10 unlts
6111193 Rs~9~ G~NaCYi2.23 fibroblas~2.00 r loa1 550 units 440 unlts
tllmor1.00 s 1071 33 units
SUBSI ITUTE SHEET ~RULE 2~
WO95/07105 ~ 22 PCT~S94/10217
by the transduced cells in vitro was measured by ELISA.
Tumor cells transduced with DC/AD/R/IL-2 or GlNaCvi2.23
expressed similar amounts of IL-2 in vitro (10-43 units
IL-2/106 cells/24 hrs). However, fibroblasts transduced
with the GlNaCvi2.23 vector secreted approximately 5-10
fold higher levels of IL-2 compared to those transduced
with the DC/AD/R/I~-2 vector (Table 1). The total
~m; n; stered IL-2 dose ranged from 3 to 440 units/24 hrs.
The total tumor cell dose for each ;mmlln;zation was 10'
cells, the dose being adjusted using unmodified tumor
cells.
Clinical Course:
No significant adverse reaction was observed at
the ;mmlln;zation sites and no treatment related
abnormalities were observed during monitoring of the
patient's complete blood count, serum chemistry and urine
specimens. Transient, mild erythema lasting less than 24
hr was observed at the injection site following
;mmlln;zation at TL-2 doses greater than 100 units/24 hrs.
Tamoxifen (2 x 80 mg/day) was ~m; n; stered beginning
approximately 3 months after the first ;mmlln;zation.
Magnetic resonance imaging (MRI) scans were
performed at approximately 4 week intervals during the
first five months of treatment. The scans revealed modest
changes in overall tumor size with waxing and waning of
peritumoral edema associated with alterations in decadron
doses (not shown). The MRI scan performed six months after
the initiation of treatment (4 weeks after the final and
highest dose of IL-2 was administered) revealed marked
tumor necrosis with significant peritumoral edema (not
shown).
Clinically, the MRI findings were associated with
an exacerbation of the patient's baseline left-sided
WO95/0710S 21 7 PCT~S9~/10217
weakness. However, this weakness gradually has improved
following ~m; ni stration of increased doses of decadron,
which then were gradually tapered. Stereotactic
intraventriculostomy was performed to relieve increased
pressure in the left third ventricle. Cytological
evaluation of cerebrospina' fluid revealed the presence of
inflammatory cells without detectable tumor cells.
In summary, IL-2 gene therapy resulted in no
significant toxicity at the sites of immlln;zation and was
associated with the generation of a cellular anti-glioma
immune response (see below). Marked tumor necrosis was
observed following the final IL-2 imm~ln;zation dose. Thus,
the results establish the potential therapeutic value of
the disclosed method for inhibiting or preventing the
growth of tumor cells ln the CNS by stimulating the
patient's immune response by ; mmlln; zation with haplotype-
matched cytokine-secreting tumor cells.
Immune Responsiveness:
Peripheral blood mononuclear cells and serum
samples from the patient were analyzed to assess the
cellular and humoral anti-glioma cell immune response
against autologous tumor cells.
To determine cell mediated imm~ln;ty~ standard
chromium release assays were used. Briefly, peripheral
blood mononuclear (PBM) cells were isolated by Ficoll-
Hypaque density centrifugation of heparinized blood and
were stimulated in vitro by incubating the cells with
irradiated autologous tumor cells at various ratios of
PBM:tumor cells in 96-wel' round-bottomed plates in the
presence of IL-2 for 7 days. The cells then were washed
and restimulated for six additional days. Target tumor
cells were labelled overnight with lO0 ~Ci of chromium-51
at 37 C. The labelled cells were extensively washed and
-
WO95/07105 PCT~S94/10217
2~ t,- 24
mixed with various numbers of ef~ector cells in 96-well V-
bottom plates. After a 4 hr incubation at 37 C, the
plates were centrifuged at 400 x g for 5 min and the
radioactivity determined in a l00 ~l aliquot of the culture
supernatant. The percent specific lysis was calculated
using the formula:
{ (cpm e~cp ~ Cpmbkgd ) / ( Cpm total CPmbkgd ) } X
100 .
The results of the chromium release assay
revealed that at a ratio of 30:l peripheral blood
mononuclear cells:tumor cells, the level of tumor cell
cytolytic activity increased 3-4 fold above the baseline
level following the third and subsequent imml]nizations (not
shown). These findings are consistent with the generation
of a cellular anti-glioma cell immune response.
The humoral immune response was measured using
indirect immunofluorescence to identify antitumor
antibodies present in the patient's serum. No humoral
response against the autologous tumor cells was observed.
EXAMPLE III
TREATMENT OF PATIENTS HAVING A TUMOR IN THE CNS
USING HAPLOTYPE-MATCHED CYTOKINE-SECRETING CELLS
This example illustrates the general application
of the claimed invention to patients having primary CNS
tumors or metastatic lesions in the CNS.
Patient Selection:
Patients will have a histologically confirmed
diagnosis indicating the presence of a primary CNS tumor or
metastatic lesions present within the CNS. Patients with
tumors that must be resected for therapeutic purposes or
Sll~S~lTUTE SHEET ~RULE 2~)
WO95107105 ~1 712 PCT~S94/10217
disclosed herein. Autologous fibroblasts and tumor cells
can be cultured using methods as described above or
otherwise known to one in the art.
However, the above-described patients as well as
patients in which tumor cell samples are unavailable can be
immuniæed with allogeneic haplotype-matched genetically
modified tumor cells, so long as such tumor cells are of
the same histologic origin as the patient's tumor. For
example, where a patient with a HLA-A2 haplotype has a
tumor of the glioma series, ;mmlln;zation can utilize
genetically modified GT9 cells, as described in Example I.
Other appropriate allogeneic haplotype-matched genetically
modified tumor cells can be obtained from a panel of such
tumor cells that have been established as continuously
cultured cells.
Pretreatment Evaluation:
Standard pretreatment evaluations are performed
as follows:
l) History and physical eX~mi n~tion including
a description and quantitation of disease activity and
tissue-typing of the patient.
2) Performance Status Assessment
0 = Normal, no symptoms
l = Restricted, but ambulatory
2 = ~p greater than 50% of waking
hours, capable of self-care
3 = Greater than 50% of waking hours
confined to bed or chair, limited
self-care
4 = Bedridden
WO95/07105 ~ PCT~S91110217
26
3) Pretreatment laboratory analysis, including
complete blood count, including differential count,
platelet count, PT, PTT, glucose, BUN, creatinine,
electrolytes, SGOT, SGPT, LDH, alkaline phosphatase,
bilirubin, uric acid, calcium and total protein albumin.
Other analyses are performed as deemed
appropriate, including ur nalysis, serum complement levels
and immunophenotyping of ~eripheral blood B cell and T cell
subsets. In addition, pretreatment evaluations can include
chest X-ray and other diagnostic studies including
computerized tomography (CT), magnetic resonance imaging
(MRI) or radionuclide scans to document and ~uantify the
extent of disease activity. Follow-up evaluations of these
assessments are performed at regular intervals during the
course of therapy (approxlmately every l to 3 months) to
monitor the patient's response to therapy and to identify
potential signs of toxicity, thus permitting adjustments in
the number and distribution of ;mmllnizations.
Restrictions on Concurrent Therapy:
For optimal effects of this treatment, patients
should receive no concurrent therapy which is known to
suppress the immune system.
Treatment Protocol:
Each patient will receive subcutaneous
;mmlln;zations with autologous or allogeneic haplotype-
matched cytokine-secreting tumor cells, which can be
genetically modified to express and secrete, for example,
IL-2, and with genetically modified cytokine-secreting
autologous fibroblasts and unmodified irradiated tumor
cells. Prior to ;mmlln;zation~ tumor cells will be
irradiated with approximately 7000 rads of radiation, so as
to render the tumor cells lncapable of proliferation in
WO95/07105 1 71%1 l PCT~S94/10217
27 ~
vivo. Treatment will proceed essentially as described in
Example II.
In general, a tumor biopsy is taken approximately
two months prior to the initiation of ;mmllnization. The
tumor cells are adapted to tissue culture and, if desired,
genetically modified to express a cytokine gene. Cytokine-
secreting tumor cells can be isolated and used for
;mmllnization. However, if autologous tumor cells are
unavailable or cannot be adapted to grow in tissue culture,
allogeneic haplotype-matched cytokine-secreting tumor cells
can be used for ;mmlln;zing the patient.
The patient is ;mmnn;zed subcutaneously with
haplotype-matched cytokine-secreting tumor cells or with
cytokine-secreting autologous fibroblasts and irradiated
unmodified tumor cells at l-~ week intervals, with
adjustments to the ;mmnn;zation schedule made as required.
Where ;mmlln;zation involves, for example, the use of IL-2-
secreting cells, the level of IL-2 secreted at the site of
;mmnn;zation will be escalated from 1OO units/24 hr early
in the ;mm~n; zation schedule to 400 units/24 hr later in
the schedule. The number of injected IL-2-secreting cells
will remain relatively constant at approximately l x 106 to
l x 107 tumor cells/;mmlln;zation site by adding an
appropriate number of irradiated unmodified tumor cells to
the IL-2-secreting tumor cells required to secrete the
appropriate level of IL-2 as determined by one skilled in
the art of tumor immunotherapy. Multiple ;mmlln;zation
sites can be used if it is deemed desirable to increase the
IL-2 dose to the patient. The patient will be physically
e~m;ned on each of the three consecutive days following
;mmlln;zation and physical and laboratory evaluations will
be made at weekly intervals.
Alternatively, a patient may be treated at the
site of the tumor in the CNS. For example, during a
r' ~
WO 95/07105 PCT/US94/10217
217121~ 28
surgical procedure to remove a CNS tumor, haplotype-matched
cytokine-secreting cells can be placed in the region from
which the tumor was removed surgically. In most cases,
;mmlln;zation with cytokine-secreting tumor cells at the
time of surgery will utilize allogeneic haplotype-matched
genetically modified tumor ce'ls selected from a panel of
genetically modified tumor cells. However, if autologous
tumor cells had been available prior to surgery, such
autologous tumor cells can be genetically modified and used
to ;mmllnize the patient at the site of the tumor in the CNS
or unmodified tumor cells can be ~ministered in
combination with cytokine-secreting autologous fibroblasts.
In addition to using stereotactic surgical procedures to
place genetically modified cytokine-secreting cells at the
site of a tumor within the CNS, ultrasound- or computerized
tomography-directed fine needle insertion can be employed
to introduce cytokine-secreting cells into the site of the
tumor.
Dose Ad~ustments:
TmmllnizationS using cytokine-secreting cells are
~mi nistered at intervals o_ 1-4 weeks. The patient is
physically e~m;ned on each of the three consecutive days
following ;mmnnization and physical and laboratory
evaluations will be made at weekly intervals. In addition,
the immunoresponsiveness of the patient is determined using
the assays described above, including, for example, assays
to determine changes in the activity of the cellular immune
response in the patient.
So long as no toxicity is observed, subsequent
;mmlln;zations are ~mi nistered at intervals of 1-4 weeks.
The results of the cellular and humoral
immunoresponsiveness assays and tumor monitoring studies
can be used to optimize the treatment protocol as
determined by one skilled in the art of tumor
~ WO95/07105 21 71211 - ~ - PCT~S94/10217
29
immunotherapy. Although toxic side effects are not
expected to result from these ;mmlln;zations~ potential side
effects are treated as described above.
Treatment of Potential Toxicity:
Unacceptable toxic side effects at the site of
immunization were not observed in the patient treated as
described in Example II and, therefore, are not expected to
result from these ;mm~ln;zations. However, potential side
effects of these ;m~lln;zations can be treated as required.
For example, if massive tumor cell lysis results, any
resulting uric acid nephropathy, adult respiratory distress
syndrome, disseminated intravascular coagulation or
hyperkalemia will be treated using st~n~Ard methods well
known in the art. Local toxicity at the sites of
;mmun;zation will be treated with either topical steroids
and, if necessary, surgical excision of the injection site.
Generalized hypersensitivity reactions such as "the
chills," fever or rash will be treated symptomatically with
antipyretics and antihistamines. Patients should not be
treated prophylactically. Edema, arthralgia,
lymphadenopathy or renal dysfunction can be treated using
corticosteroids and/or antihistamlnes. Anaphylaxis will be
treated by standard means such as A~m; n; stration of
epinephrine, fluids and steroids.
Other AssaYs:
Provided that sufficient material is available
for evaluation the following assays also are performed.
St~n~Ard immunofluorescence flow cytometry procedures are
useful to evaluate changes in the percentage of T-cells,
natural killer cells and B-cells associated with cytokine
gene therapy. Monoclonal antibodies specific for T cells
(CD2, CD3, CD4, CD8), na~ural killer cells (CDl6, CD57,
WO95/07105 PCT~S94/10217 ~
2~ ~2 30
CD58) and B cells (CDl9, CD20) can be used for these
studies.
Briefly, Fico'l-Hypaque purified mononuclear
cells are incubated with the primary antibody for l hr at
room temperature, washed, then incubated with fluorochrome
conjugated secondary antibody. The cells are washed, fixed
and the percentage of positive cells are determined using
a Coulter Epics 4 flow cytometer. Incubation of the cells
with isotype-matched control antibody instead of the
primary antibody is useful as a negative substitution
control.
St~n~Ard immunohistological methods employing
monoclonal antibodies specific for the hematopoietic cell
subsets described above can be used to characterize the
immune effector cell infiltrates observed in delayed-type
hypersensitivity skin test biopsy sites. Methods for
immunohistological evaluations of fresh frozen cryostat
tissue sections are well known in the art.
EXAMPLE IV
ANIMAL STUDIES
The rat glioma tumor model described by Holladay
et al., l990, 1992, demonstrates the usefulness of the
disclosed method of stimulating an immune response in a
subject against the subject's tumor. Gliomas are produced
in the rats, as described, and various treatment modalities
are employed.
Briefly, glioma-bearing rats are treated with
haplotype-matched glioma cells, which are genetically
modified to express a cytokine gene, or with unmodified
tumor cells and genetically modified cytokine-secreting
autologous fibroblasts. Tmmllnization is at a site other
than the CNS or at a site within the CNS. The stimulation
WO95/07105 PCT~S94/10217
_, 21 712 .` . ~
of a cellular and humoral immune response is determined as
described above. In additlon, the effect of treatment on
the tumor is monitored by sacrificlng rats at various times
after initiating treatment and e~m; n; ng the gross and
histological anatomy of the tumor. The ability of
;mmllni zed An;m~l s to reject a subsequent tumor challenge
also is determined.
Although the invention has been described
with reference to the dlsclosed examples, it should be
understood that various modifications can be made without
departing from the spiri' of the invention. Accordingly,
the invention is limited only by the following claims.
