characteristics of human cytomegalovirus specific t-cells ...362022/s4308500_mphil... · khanna...
TRANSCRIPT
1
Characteristics of Human Cytomegalovirus specific T-Cells in
Glioblastoma Multiforme
Hamish Stewart Alexander
BSc(Hons), MB.ChB
A thesis submitted for the degree of Master of Philosophy at
The University of Queensland in 2014
School of Medicine
QIMR Berghofer Medical Research Institute
2
Abstract
Glioblastoma Multiforme (GBM) is the most common primary brain malignancy and
invariably fatal. Evidence has emerged showing human cytomegalovirus (HCMV) infection
is present in GBM, providing a possible target for new immunotherapeutic agents. This thesis
aims to examine the functional characteristics of T-cell responses to GBM, including HCMV-
specific T-Cells. To examine the effects of exposure HCMV in patients with GBM an 80
patient cohort was assessed. No difference was seen between HCMV seropostive and
seronegative patients with median progression free survival of 344 days vs. 348 days
(p=0.453) and median overall survival s 719 days vs. 780 days (p=0.553). To further
investigate HCMV immunity in GBM, gene microarray analysis was performed on CD8+ T-
cells and HCMV specific T-cells from GBM patients. Analysis of CD8+ T-cells in GBM
patients showed down regulation of genes in two distinct subgroups related to effector
function and the exhausted phenotype of T-cells usually seen in persistent virus infection.
Included in this was significant up regulation of the co-inhibitory gene BTLA along with
down regulation of the gene for co-stimulatory molecule LIGHT. These are both ligands of
Herpes Virus Entry Mediator (HVEM) suggesting this may be an important pathway of
immune suppression in GBM. Flow Cytometry studies however failed to confirm differences
in surface expression of these molecules in T-cells of GBM patients. HVEM was however
seen to be expressed in tissue in a significant number of tumours examined by
immunohistochemistry. Further studies are required to define the role of these pathways and
their suitability for manipulation as a standalone therapy or adjunct to other
immunotherapeutic strategies.
3
Declaration by author
This thesis is composed of my original work, and contains no material previously published
or written by another person except where due reference has been made in the text. I have
clearly stated the contribution by others to jointly-authored works that I have included in my
thesis.
I have clearly stated the contribution of others to my thesis as a whole, including statistical
assistance, survey design, data analysis, significant technical procedures, professional
editorial advice, and any other original research work used or reported in my thesis. The
content of my thesis is the result of work I have carried out since the commencement of my
research higher degree candidature and does not include a substantial part of work that has
been submitted to qualify for the award of any other degree or diploma in any university or
other tertiary institution. I have clearly stated which parts of my thesis, if any, have been
submitted to qualify for another award.
I acknowledge that an electronic copy of my thesis must be lodged with the University
Library and, subject to the General Award Rules of The University of Queensland,
immediately made available for research and study in accordance with the Copyright Act
1968.
I acknowledge that copyright of all material contained in my thesis resides with the copyright
holder(s) of that material. Where appropriate I have obtained copyright permission from the
copyright holder to reproduce material in this thesis.
4
Publications during candidature
Andrea Schuessler, Corey Smith, Leone Beagley, Glen M Boyle, Sweera Rehan, Katherine
Matthews, Linda Jones, Tania Crough, Vijayendra Dasari, Kerenaftali Klein, Amy Smalley,
Hamish Alexander, David G Walker and Rajiv Khanna. Autologous T cell Therapy for
Cytomegalovirus as a Consolidative Treatment for Recurrent Glioblastoma. Cancer Res.
2014 Jul 1; 74(13):3466-76
Publications included in this thesis
No publications included
5
Contributions by others to the thesis
The conception, initial design and supervision of this thesis were provided by Professor Rajiv
Khanna from the Tumour Immunology Laboratory at the QIMR Berghofer Medical Research
Institute.
The experimental work for the gene expression study in Chapter 3 and flow cytometry study
in Chapter 4 was carried out with the help and guidance of Dr Andrea Schuessler and Dr
Corey Smith from the QIMR Berghofer Tumour Immunology Laboratory. Additional
healthy control data in chapter 3 was provided by Dr Schuessler.
HVEM Immunohistochemistry staining of GBM tissue in Chapter 4 was carried out by Clay
Winterford from the QIMR Berghofer Histotechnology Department. The grading of the
immunohistochemistry slides was performed with the assistance of Dr Mahindra Singh and
Dr Lakshmy Nandakumar from the Pathology Department of the Royal Brisbane Hospital.
Analysis of gene expression data using Gene Spring software and Pathway analysis using
IPA in Chapter 3 was performed by Prof Glen Boyle of the Cancer Drug Mechanism Group
at the QIMR Berghofer Medical Research Institute.
Statistical analysis of the clinical data in Chapter 2 was performed with the assistance of
Kerenaftali Klein of the Cancer and Population Studies Department at the QIMR Berghofer
Medical Research Institute.
Illustrations were prepared with the assistance of Madeline Flynn, Graphic Support Officer,
QIMR Berghofer Medical Research Institute.
Statement of parts of the thesis submitted to qualify for the award of another degree
None
6
Acknowledgements
Scholarships:
Sincere thanks to the NEWRO Foundation and BrizBrain and Spine along with the
Neurosurgical Society of Australasia and Synthes, for their financial support of this project.
• NEWRO Foundation and BrizBrain and Spine Research Fellowship for 2013
• The Neurosurgical Society of Australasia Research Scholarship - Sponsored by
Synthes 2013
Personal Acknowledgements:
Sincere thanks to Profs David Walker and Rajiv Khanna for agreeing to supervise this Thesis
and providing the environment to make it happen.
To the surgeons and staff at BrizBrain and Spine, thank you for all the encouragement and
support during the year.
Thanks to the QIMR staff and in particular the members of the Tumour Immunology Lab for
making me so welcome. Special thanks to Andrea Schuessler and Corey Smith for your time
and patience.
Thanks most of all to my family, Emily, Ted, Maggie and Olivia.
7
Keywords
Glioblastoma Multiforme, Human Cytomegalovirus, Immunotherapy, Cytotoxic T-Cells,
inhibitory receptor, immune response, gene expression analysis
Australian and New Zealand Standard Research Classifications (ANZSRC)
110709 Tumour Immunology 70%
111204 Cancer Therapy 10%
060405 Gene Expression 20%
Fields of Research (FoR)
1107 Immunology 80%
1112 Oncology and Carcinogenesis 20%
8
Contents
ABSTRACT .........................................................................................................................2
DECLARATION BY AUTHOR .........................................................................................3
PUBLICATIONS DURING CANDIDATURE ..................................................................4
CONTRIBUTIONS BY OTHERS TO THE THESIS........................................................5
KEYWORDS .......................................................................................................................7
AUSTRALIAN AND NEW ZEALAND STANDARD RESEARCH
CLASSIFICATIONS (ANZSRC) .......................................................................................7
FIELDS OF RESEARCH (FOR)........................................................................................7
LIST OF TABLES.............................................................................................................10
LIST OF FIGURES...........................................................................................................11
COMMON ABBREVIATIONS USED IN THIS THESIS...............................................12
CHAPTER 1: INTRODUCTION .....................................................................................13
1.1 GBM BIOLOGY .............................................................................................................15
1.2. GBM TREATMENT AND SURVIVAL ...............................................................................16
1.3. HCMV BIOLOGY .........................................................................................................17
1.4. TUMOUR INDUCED IMMUNE RESPONSE AND THE CNS .................................................20
1.5. TUMOUR MICROENVIRONMENT IMMUNOSUPPRESSION ...............................................25
1.6. HCMV IN GBM...........................................................................................................28
1.9. IMMUNOTHERAPY IN GBM ..........................................................................................35
1.9. HCMV TARGETED IMMUNOTHERAPY FOR GBM........................................................40
1.10. CURRENT STUDY ........................................................................................................41
CHAPTER 2: EFFECTS OF CMV SEROLOGY STATUS ON CLINICAL
OUTCOMES IN GBM ......................................................................................................42
9
2.1. INTRODUCTION............................................................................................................42
2.2. METHODS .....................................................................................................................42
2.3. RESULTS.......................................................................................................................44
2.4. DISCUSSION ..................................................................................................................49
CHAPTER 3: GENE EXPRESSION ANALYSIS OF CMV SPECIFIC AND TOTAL
CD8+ T-CELLS IN PATIENTS WITH GBM..................................................................51
3.1. INTRODUCTION.............................................................................................................51
3.2. METHODS .....................................................................................................................51
3.3. RESULTS.......................................................................................................................53
3.4. DISCUSSION ..................................................................................................................65
CHAPTER 4: THE HVEM, BTLA AND LIGHT NETWORK IN GBM PATIENTS...67
4.1. INTRODUCTION.............................................................................................................67
4.2. METHODS .....................................................................................................................69
4.3. RESULTS.......................................................................................................................71
4.4. DISCUSSION ..................................................................................................................77
CHAPTER 5: DISCUSSION.............................................................................................79
BIBLIOGRAPHY..............................................................................................................83
APPENDIX: GENES INCLUDED IN GENE EXPRESSION ANALYSIS ....................98
10
List of Tables
Chapter 1:
Table 1.1: HCMV biology and cancer hallmarks....................................................................32
Chapter 2:
Table 2.1: Descriptive statistics on the study cohorts.............................................................44
Table 2.2: Hazard Ratios for Mortality and Progression Free Survival..................................48
Chapter 3:
Table 3.1: Characteristics of Subjects for Gene Expression Analysis.....................................53
Table 3.2: Changes in Gene Expression of HCMV-Specific T-cells in GBM........................57
Table 3.3: Changes in Gene Expression of HCMV-Specific T-cells in GBM........................58
Table 3.4:Gene Expression; total CD8+ T-cells in GBM patients based on clinical outcomes
(time to progression and overall survival)............................................61
Chapter 4:
Table 4.1: HVEM Expression in GBM by Immunohistochemistry........................................73
11
List of figures
Chapter 1:
Fig 1.1: Adoptive T-cell transfer.............................................................................................14
Fig 1.2: Characteristics of anergic, exhausted, senescent and stem-like T-cells.....................23
Fig 1.3: Examples of Co-inhibitory and co-stimulatory molecules.........................................25
Chapter 2:
Fig 2.1: All cause mortality.....................................................................................................46
Fig 2.2: Progression free survival............................................................................................47
Chapter 3:
Fig 3.1: Changes in gene expression in CMV Specific T-Cells in GBM................................55
Fig 3.2: Changes in gene expression in the Total population of CD8+ T-Cells in GBM........56
Fig 3.3: Changes in gene expression in the Total population of CD8+ T-Cells in Long
surviving GBM patients.............................................................................................59
Fig 3.4: Changes in gene expression in CD8+ T-Cells of GBM patients based on progession
free survival................................................................................................................60
Fig 3.5: Key for Pathway Analysis Diagrams........................................................................62
Fig 3.6: Pathway Analysis of Gene Expression Data from the CMV Specific T-Cell
population in GBM patients.......................................................................................63
Fig 3.7: Pathway Analysis of Gene Expression Data from the total T-Cell population in GBM
patients.......................................................................................................................64
Chapter 4:
Fig 4.1: The HVEM/BTLA/LIGHT pathway..........................................................................68
Fig 4.2: The HVEM Network in Tumour Models...................................................................69
Fig 4.3: Analysis of BLTA, LIGHT and HVEM expression in CD8+ T-Cells.......................72
Fig 4.4: T-Cell Expression of BTLA, HVEM and LIGHT.....................................................73
Fig 4.5: CMV Specific T-Cell Expression of BTLA, HVEM and LIGHT.............................74
Fig 4.6: Immunohisotchemical staining for HVEM in GBM..................................................76
Common abbreviations used in this thesis
Adoptive cell transfer (ACT)
B and T Lymphocyte Attenuator (BTLA).
Central nervous system (CNS)
Chemokine receptor (CCR)
Cluster of differentiation (CD)
Dendritic cells (DC)
Epidermal growth factor receptor (EGFR)
Fluorescence Activated Cell Sorting (FACS)
Glioblastoma multiforme (GBM)
Glioma cancer stem cells (gCSCs)
Herpes simplex virus (HSV)
Herpes Virus Entry Mediator (HVEM)
Human cytomegalovirus (HCMV)
Immediate early (IE)
Immediate early antigen (IEA)
Immunoglobulin B and T lymphocyte attenuator (BTLA)
Immunoglobulin (Ig)
Interleukin (IL)
Lymphohotoxin-like, exhibits Inducible expression and competes with herpes simplex virus
Glycoprotein D doe HVEM, a receptor expressed by T-lymphocytes (LIGHT)
Major histocompatibility complex (MHC)
Peripheral blood mononuclear cells (PBMCs)
Polymerase chain reaction (PCR)
Tumour necrosis factor receptor (TNFR)
Tumour Necrosis Factor Super Family Member 14 (TNFSF14) also known as LIGHT
World Health Organisation (WHO)
13
Chapter 1: Introduction
Glioblastoma Multiforme (GBM) is the most common primary brain malignancy in adults
with an incidence of 3.15 per 100,000 in the United States [1]. It is characterised by an
aggressive clinical course and is invariably fatal. Current “gold standard” therapy consists of
maximal surgical debulking followed by radiotherapy and chemotherapy [2]. Therapy for
recurrent tumours is limited to palliative resection and second line chemotherapeutic agents
[3, 4]. Prognosis in GBM is remains uniformly poor with a median survival after “gold
standard” therapy of 14.6 months following diagnosis [5]. Given the limited success of
current therapeutic strategies there has been considerable effort into establishing a better
understanding of GBM biology and exploring potential therapeutic pathways such as
immunotherapy. In recent years significant evidence has emerged showing that human
cytomegalovirus (HCMV) is present in GBM and that it plays a oncomodulatory role [6]. The
presence of HCMV infection in GBM is not only a possible factor in tumour growth but also
provides potential antigen targets for novel therapeutic strategies such as immunotherapy..
Our laboratory is currently trialling HCMV specific adoptive T-cell transfer in GBM patients.
Specific adoptive cell transfer (ACT) refers to tumour specific therapy with cytotoxic T-Cells
harvested from a tumour bearing host. These tumour specific autologous cells then undergo
ex vivo expansion, activation and are transferred back to the patient [7]. This technique
provides a passive immunotherapy to give a tumour specific response. The benefits of this
approach are the ability to give large numbers of cells that have a high specificity for tumour
antigens and to bypass the immunosuppressive tumour microenvironment [8]. It also gives an
opportunity for selection of effector cells that show the best anti-tumour activity away from
the in vivo environment and effects of physiological or pathophysiological control [9].
14
Fig 1.1: Adoptive T-cell transfer: CD8+ T-Cells are isolated from peripheral blood samples
of patients with recurrent GBM after standard treatments with surgery, radiotherapy and
chemotherapy. These cells are then stimulated with CMV epitopes and IL2 ex-vivo then
transferred back to the patient. (Figure modified from original: www.attack-cancer.org)
Previous studies from this laboratory demonstrated that CMV seropositive GBM patients
have circulating HCMV-specific T-cells at levels that are similar to healthy individuals;
however, the majority of these cells are non-functional or display limited functionality [10].
The aim of this study is to better characterise the T-Cell response to GBM in order to identify
factors which are acting to inhibit the host anti-tumour response. This information could be
used to develop targets for manipulation that may potentiate immunotherapy, act as stand
alone therapies or facilitate the ex vivo engineering of T-Cells to be better suited to adoptive
T-Cell therapy.
15
1.1 GBM Biology
Gliomas are tumours arising from the glial cells of the central nervous system, most
commonly astrocytes, and are graded according to the World Health Organisation (WHO)
into four grades based on histological characteristics [11]. GBM represents WHO grade IV.
The peak incidence occurs between 45-75 years with a slight bias towards males. Typically
these tumours occur in the sub-cortical white matter of the cerebral hemisphere however they
may occur anywhere in the CNS. Histological examination shows characteristic pallisading
necrosis and microvascular proliferation. Risk factors for GBM are uncertain but include
ionising radiation and some genetic syndromes, however only 1-3% of GBM can be
attributed to identifiable genetic conditions. GBM occur in two settings; as de novo tumours
in usually older patients (primary) and in younger patients with a previous lower grade
astrocytoma (secondary). On a molecular level primary and secondary GBM have subtle
differences but tend to converge on the same pathways [11]. As an example secondary GBM
tends to have mutations of the tumour suppressor gene p53 whereas primary GBM are more
likely to show amplification of the MDM2 gene which codes for a suppressor of p53. An
important molecular marker of secondary vs. primary GBM is mutation of isocitrate
dehydrogenase 1 and 2 (IDH1 & IDH2) which is found in more than 80% of secondary
GBM but only few primary GBM [12]. This suggests a different genetic pathway for
primary and secondary tumours which may lead to different clinical courses. Common
molecular lesions seen in GBM are activation of RAS and PI-3 kinase and inactivation of p53
and Rb. These changes are seen in 80-90% of GBM.
Just as GBM are characterised by cellular heterogeneity they also display a wide array of
molecular changes and disruptions few of which have been fully defined [13] [14]. Studies of
GBM tissue using gene expression analysis have led to further classification of GBM based
on a molecular profile [13]. These sub-types are Proneural, Neural, Classical, and
Mesenchymal. The Classical, Mesenchymal, and Proneural subtypes are defined by changes
in gene expression of epidermal growth factor receptor (EGFR), neurofibromin 1 (NF1), and
alpha-type platelet-derived growth factor receptor (PDGFRA) respectively. These subtypes
appear to be of clinical relevance in that a more aggressive therapeutic strategy, defined as
concurrent chemotherapy and radiotherapy or greater than three cycles of chemotherapy, has
a greater benefit in the classical subtype but is of limited benefit in the proneural subtypes
[14].
16
The traditional model of tumourigenesis involves a mutation in a single cell that then
undergoes clonal expansion producing a tumour. The origin of this single cell mutation is
controversial and in GBM there are two theories that compete and may overlap [15, 16].
Firstly the “de-differentiation theory” proposes a series of changes conveying the hallmarks
of cancer such as resistance to apoptosis onto mature astrocytes transforming them into
malignant cells. The alternate theory involves quiescent stem cells that lie dormant until a
triggering event leads to development of a tumour. These sub-populations are then believed
to persist within the tumour that possessing tumourogenic capability driving the malignant
characteristics of GBM, with the rest of the cells acting as non-tumour initiating constituents
of the tumour. This hypothesis developed from the observation that subgroups of cells in
GBM have the characteristics of stem cells in that they can regenerate and differentiate into
different cell types (astrocytes and neurons). When these cells are injected into
immunocompromised animals they generate tumours [17]. These are referred to as glioma
Cancer Stem Cells (gCSC). A tremendous amount of research is now directed at tumour stem
cells to understand how they may act in resistance to standard therapy and be targets for new
agents. gCSC’s have been shown to have resistance to chemotherapy and radiotherapy when
compared to other GBM cells and this may underlie the relative resistance of GBM to
conventional therapies [18]. In addition these stem cells have immunosuppressive activity via
cytokines such as soluble colony-stimulating factor (sCSF-1), Transforming growth factor
beta 1 (TGF-β1), and macrophage inhibitory cytokine-1. They also induce an M2 phenotype
in microglia inducing immunosuppressive IL-10 and inhibiting proliferation of T-cells [19].
Interestingly recent studies have shown that HCMV shows tropism for gCSC’s and
macrophages. The production of HCMV IL-10 by these cells then feeds forward by
promoting monocytes to adopt a tumorigenic phenotype and express angiogenic vascular
endothelial growth factor (VEGF), immunosuppressive TGF-β1 and promote gCSC
migration [20].
1.2. GBM Treatment and survival
Surgical resection of GBM has proven to be non curative. However surgery remains a
mainstay of treatment, the rationale being three-fold; to obtain tissue for histological
diagnosis, relieve mass effect and reduce tumour load. Adjuvant external beam radiotherapy
17
to a dose of 60-65 Gray is given as standard treatment with concurrent chemotherapy[21].
Factors that have been shown to be predictive of longer survival are extent of surgery, age
<60yrs and Karnofsky grading at diagnosis [22, 23].
The biggest improvement in terms of survival in recent years has been the introduction of
Temozolomide chemotherapy [2]. Temozolomide causes cell death by apoptosis via
methylation of DNA at the N-7 or O-6 guanine residues, inducing cell cycle arrest at G2/M-
phase [24].The effectiveness of Temozolomide is variable with one factor being the silencing
of MGMT gene expression [25]. MGMT is a critical enzyme for repair of DNA and its loss
potentiates the chemotherapeutically induced DNA damage. Older chemotherapy strategies
such as Lomustine (CCNU) or the Procarbazine, Vincristine (PCV) regime have been
superseded in recent years and are now reserved for second line therapy [4]. Bevacizumab
(Avastin®, Genentech) is new chemotherapeutic agent approved by the US FDA for
recurrent GBM. It is a monoclonal antibody against VEGF and thus acts as an inhibitor of
angiogenesis [26]. The effectiveness of this drug on overall survival in GBM is unproven [27,
28]. Other treatments that have been recently evaluated for effectiveness include intra-tumour
placement of 1, 3-bis (2-chloroethyl)-1-nitrosourea (BCNU) wafers (Gliadel®, Arbor
Pharmaceuticals). This has shown promising results in terms of survival but concerns persist
regarding its toxicity profile preventing widespread adoption in clinical practice [29].
Despite improved understanding of GBM biology and new treatment strategies overall
prognosis remains dismal with only limited therapeutic options therefore further advances
both clinical and scientific are needed.
1.3. HCMV Biology
HCMV infection is common. It has prevalence of 40 to 100% depending on geographic and
socio-economic factors [30]. Viral transmission has been reported via saliva, sexual contact,
breast milk and placental transfer; as well as blood transfusions and solid organ or
hematopoietic stem cell transplantation [31]. Primary HCMV infection in an
immunocompetent host is normally asymptomatic, occasionally presenting with malaise,
headache and fever [32, 33]. Uncommon signs and symptoms include tonsillopharyngitis,
lymphadenopathy or splenomegaly. Rare serious complications include ulcerative colitis,
hepatitis, pneumonitis, meningitis, myocarditis and Guillian-Barre syndrome [34].
18
In contrast to the mild disease in adults HCMV is the leading infectious cause of
neurodevelopment defects and deafness in congenital infection. Primary infection in the 1st
Trimester of pregnancy has a 40% risk of transmission. 10-15% neonates affected will
display symptoms including pneumonia, hepatitis, chorioretinitis, growth retardation, central
nervous system (CNS) damage and microcephaly. This systemic infection is known as
‘cytomegalic inclusion disease’ and is fatal in 20-30% of symptomatic cases, often results in
neurological and visual impairment [35-37].
HCMV disease in adults has become increasingly important due to the rise in number of
immunocompromised hosts resulting from the acquired immunodeficiency disease
syndrome (AIDS) epidemic and immune compromising therapies such as organ and bone
marrow transplantation and cancer chemotherapy [37]. Reactivation or primary infection in
immunocompromised individuals can range from an asymptomatic viremia to fatal CMV
syndrome and tissue-invasive disease [33]. Severity of disease correlates to immune
dysfunction with worse disease seen in allogeneic bone marrow or stem cell transplant
recipients and AIDS patients with low CD4+ counts [38].
HCMV is also known as ubiquitous beta human herpes virus type-five (HHV-5) and is the
largest of the known human herpes viruses. It has a genome that spans 235 kilo bases
encoding 165 genes controlling various functions. The HCMV virion measures 200-300
nanometres. It consists of a DNA core surrounded by an icosahedral nuclear capisd and
proteinaceous matrix. This is in turn enclosed by a lipid bilayer envelope containing viral
glycoproteins [32, 39]. HCMV is known to infect fibroblasts, endothelial cells, epithelial
cells, monocytes, macrophages, smooth muscle cells, stromal cells, neutrophils, hepatocytes
and neuronal cells in vivo [40]. This broad range of susceptible cell types results in active
infection impacting many organs systems in an immunocompromised host. In the brain in
vitro studies revealed complete permissiveness to HCMV of neuronal cell lines and
undifferentiated glial cell lines from astrocytomas, but varying permissiveness to astrocytes,
and higher differentiated glial lines from glioblastomas [41].
Replication of HCMV after entry to the cell is slow with a cycle of 48-72hrs compared to
other herpesvirus [42]. Cell entry is achieved by envelope fusion or endocytosis to host cell
membrane and nucleocapsid release into cytoplasm. The nucleocapisd is then uncoated in
the nucleus and the viral DNA released to be transcribed by host RNA polymerase. Post
infection protein synthesis progresses in a sequential pattern typical of herpesviruses with
19
immediate early (IE); 0-2hrs, delayed early; 24hrs, and late; >24hrs [43]. The replicated
genomes form concatemers with genomic inversions, then are packaged in empty capsids
[44]. HCMV gene expression is further regulated, prior to complete virion production, at
posttranscriptional, translational and posttranslational levels [16]. As virions exit from the
cell, they acquire mature envelope from the Golgi membrane. This budding from the Golgi
membrane causes cytoplasmic inclusion bodies which are characteristic of HCMV infection.
Of the virions produced in this manner only 1% of are infectious, the rest consist largely of
protein or are enveloped particles lacking genomes [32].
Latency and Reactivation
One characteristic feature of all herpesvirus is persistence of the virus post primary infection
in a quiescent non-replicating state. This latent state of viral infection persists for the lifetime
of the host with periods of reactivation [45]. An important distinction must be made between
true latency in which the viral genome is maintained in the host cell with minimal or non-
detectable gene expression as opposed with persistent low level replication and shedding of
infectious viral particles resulting in minimal cytopathy and long term infection [46].
While lytic HCMV infection can affect large numbers of cells, those which support latent
infection are more select. Important sites of true latency of HCMV are monocytes in
peripheral blood, hematopoietic myeloid progenitor cells CD34+ and CD33+ granulocyte-
macrophage progenitors. CD34+ bone marrow cells, and immature dendritic cells [47]. Blood
vessel endothelial cells are also thought to be significant reservoir of HCMV replication,
likely to be a site of persistent infection with low grade shedding rather than true latency [48].
Latency is induced by disrupting replication via repression of the major immediate early
promoter (MIEP) and expression of a latency specific subset of viral genes [49]. These result
in signal modulation, restricted viral gene expression, and immune evasion. Reactivation of
HCMV is requires the reversal of those factors inducing latency in response to inflammatory
stress. TNF-α induced by inflammation in turn induces transcription factors NFκβ, API and
CRE which bind to the MIEP causing reactivation [50]. Chronic inflammation triggers
reactivation via stress induced catecholamines and pro-inflammatory prostaglandins [33].
20
1.4. Tumour induced Immune Response and the CNS
Tumours elicit an immune response primarily through CD8+ cytotoxic T-cells. These T-cells
are activated by antigen expressed by the tumour in the form of over expressed or mutated
self proteins, or tumour antigens generated by oncogenic viruses. While there is little
evidence that antibodies have a major anti-tumour role there is a known role for NK cells and
macrophages in the tumour associated immune response. The protective value of this immune
response in prevention of tumours is shown by the fact that immunodeficient hosts have
higher rates of cancer. Most cancers occur in immunocompetent hosts therefore tumours
must have immunosurveillance escape mechanisms. Several mechanisms have been
identified such as selective outgrowth of antigen-negative variants, loss of MHC expression,
loss of co-stimulation required for T-Cell activation, local immunosuppression, antigen
masking and apoptosis of cytotoxic T –cells.
Since the observation that GBM patients who develop a post operative infection survive
longer than those without there has been interest in the relationship between the immune
response and the brain and brain tumours [51]. The brain has traditionally been thought to
enjoy an immunologically privileged position due to a number of features unique to the CNS
namely the blood-brain barrier, absence of lymphatics, immunoregulatory factors and lack of
MHC expression in normal CNS cells.
This was believed to provide a mechanism by which the brain was protected from potentially
damaging inflammatory processes which occur hand in hand with an immune response. This
would mitigate the potential for cerebral oedema and its neurological sequelae. This
previously widespread concept of immune privilege for the CNS has been largely refuted by
the discovery that activated peripheral T-cells along with antibodies are able to cross the
blood-brain barrier and furthermore antigen specific T-cells are able to proliferate and
acquire effector function in the CNS however these are likely to have altered functional
ability [52-54]. Within the brain T-cells are rare, activation is required for T-cell entry into
the CNS but this need not be antigen specific [54].
In the CNS microglia act as macrophages by expressing MHC class II antigen along with co-
stimulatory molecules allowing T-cell activation, demonstrating that the CNS is in a dynamic
state with the peripheral nervous system [55]. Whilst the normal CNS has been shown to be
immunocompetent it has long been recognised that patients harbouring GBM have impaired
21
immune response [56, 57]. This immunosuppression in the GBM patient is due to a number
of mechanisms both systemic and local many related to T-cell function. The systemic
impaired immune response in GBM patients can been seen at a gross level with leucopoenia
on peripheral blood counts which appears to improve following cytoreductive surgery and
decline again at recurrence [58, 59]. In terms of T-cell function systemic responses are
inhibited through increased circulating Tregs [60]. Within the tumour micro environment
inhibitory factors include expression of Fas ligand by the tumour resulting in T-cell
apoptosis, high levels of infiltrating Tregs, and an overall inhibitory cytokine milieu.[61]
Simple physical factors such as hypoxia in necrotic tissue also play a role in T-cell
dysfunction in GBM [62].
The poor immune response to GBM can in part be attributed to the weak immunogenicity of
the tumour-associated anti-gens (TAA). TAA’s in GBM are classified in 4 ways: (i) those
antigens expressed due to translocations or mutations such as EGFRvIII, (ii) antigens from
genes derived from cancer-germ lines such as melanoma associated antigen, (iii) over
expressed genes such as human epidermal growth factor receptor 2 (HER2), (iv) antigens
from viruses such as the CMV antigens IE1 and pp65 [63]. These antigens are in their own
right weakly immunoreactive but are still able to act as targets for immunotherapy. One such
specific antigen targeted in GBM patients is the EGFR. These are regulators of cell migration
and survival and there gene has been shown to be amplified in GBM, specifically EGFRvIII
has only been seen in GBM [64, 65]. A study by Choi et al used Bio-specific antibodies
(bscAbs), specifically bio-specific T-cell engager (BiTE) subclass, to effectively redirect T
cells against GBM in mice [66]. They targeted a mutation of EGFRvIII, and found significant
survival benefit with a cure rate 75%. This demonstrates that with correct antigen targeting
the T-Cell response can have a very effective anti-tumour effect.
T-Cell Dysfunction in GBM
The T-Cell dysfunction seen in GBM patients is likely a result of overlapping mechanisms of
anergy, exhaustion, senescence and stemness (see fig 1.3).
T-cell Anergy refers to the state in which lymphocytes are functionally inactivated after
exposure to an antigen but persist in a hypo responsive state for a period of time [67]. T-Cell
activation requires both MHC and co-stimulatory molecule signals. Low stimulatory and/or
22
high inhibitory signal will result in anergy [68]. This is believed to be a mechanism for
avoiding autoimmune disease. In GBM patients profiling of T-cells has shown decreased
levels of mRNA transcripts for genes which are important for T-cell function and in
particular the process of TCR binding and activation along with the subsequent intracellular
signalling [69]. Such improper TCR signal transduction is thought to be important in the
development of T-Cell anergy [70].
T-Cell exhaustion refers to a state of poor function with reduced proliferation, cytotoxicity
and cytokine secretion in response to chronic stimulation from inflammation, chronic
infection or tumour antigen [71]. The mechanism of exhaustion is the focus of great interest
currently. A number of negative co-stimulatory factors have now been identified which may
induce a state of T-cell exhaustion [70]. These mechanisms likely evolved to limit tissue
damage post- infection but have been subverted by disease states. This was first observed in
chronic lymphocytic choriomeningitis (LCMV) in mice and has been shown in a number of
chronic viral infective states such as HIV, hepatitis B and C. The mechanism has been
localised to a number of regulatory molecules shown in fig 1.3. T-Cell exhaustion via these
pathways has also been seen in a number of cancers [72, 73]. There are now a multiple of
studies examining the role of the co-stimulatory factors such as PD-1 and CTLA-4 in the
setting of GBM, these studies have shown that these molecules are expressed in GBM and
may have direct effects on T cell function with potential clinical applications [74, 75].
T-cell senescence is the result of physiological processes as cells reach the end of their life
span. It is characterised by telomere shortening, cell cycle arrest and phenotypic changes. A
subpopulation of senescence T-cells have been identified in a number of cancers however
23
Figure 1.2: Characteristics of anergic, exhausted, senescent and stem-like T-cells. T-cells
can exisit in each of these states due to interactions with the receptors and subsequent
intracellular processes shown in this diagram Anergic T-cells are non-responsive with low
production of IL-2 as a mechanism of tolerance. Exhausted T-cells accumulate in chronic
disease in an unresponsive long lived state. Senescent T-cells accumulate due to aging or
chronic infection. Stem-like T-cells are quiescent, self renewing but long lived cells.
(Modified from: Crespo et al. T Cell anergy, exhaustion, senescence and stemness. Current
Opinion in Immunology 25:214-221. 2013)
24
there is no direct evidence of this mechanism being active against mature T-cells within
tumours.
The concept of “stem-like memory T-cells” refers to the ability of this cell population to self
renew and generate more differentiated memory T-cells. This concept has been supported by
evidence in mouse models that central memory T-cells are arrested pre-differentiation and
then after a second antigen exposure can generate effector T-cells [76]. This has also been
shown in tumour associated Th17 cells [77]. There is theoretical potential to manipulate
these stem like cells for therapeutic applications.
Immunomodulatory Molecules
T-cell function is modulated by interaction with co-signalling molecules. There are two
superfamilies of co-signalling molecules, tumour necrosis factor receptor (TNFR) family and
immunoglobulin (Ig) superfamily. Activation of T-cells requires co-stimulation via receptor
CD28, similarly there are a number of negative co-stimulatory factors now identified which
inhibit T-cell function. These mechanisms likely evolved to avoid autoimmunity and limit
tissue damage post infection but have been subverted by disease states. A number of
regulatory molecules such as PD-1, Tim-3, and LAG3 have been identified and are of
relevance in anti-tumour immunity [78]. These various inhibitory molecules can potentially
be blocked by antibodies and likewise stimulatory molecules could be activated by agonistic
antibodies as shown in fig 1.3. Manipulation of the immune system via these pathways has
been likened to control of a car using break (co-inhibitory molecules), clutch (blocking anti-
bodies) and accelerator (co-stimulatory molecules).
T-cell inhibition via these pathways has been seen in a number of cancers and manipulating
these pathways by targeting molecules such as CTLA-4 with the FDA approved drug
Ipilimumab (Yervoy; Bristol-Myers Squibb, NY, USA) have led to great interest in
evaluating the role of them in various diseases such as GBM.
25
Fig 1.3: Examples of Co-inhibitory and co-stimulatory molecules. These molecules can
potentially be stimulated or blocked by antibodies to manipulate T-Cell activation. (Modified
from Reardon et al. An update on vaccine therapy and other immunotherapeutic approaches
for Glioblastoma. Expert Rev. Vaccines 12(6), 597-615. 2013 )
1.5. Tumour Microenvironment Immunosuppression
GBM has the ability to evade immunosurveillance, create an immunosuppressive micro
environment and subvert the immune response to promote tumour growth [61]. Creating an
immunosuppressive microenvironment within the GBM appears to be an important factor in
its biology. Tumour infiltrating lymphocytes (TIL) have been shown to be present in GBM,
providing a survival benefit and demonstrating the presence of some anti-tumour immune
response [79]. Despite the presence of these cells an immune response in GBM fails, due to a
range of mechanisms active within the tumour microenvironment [80]. GBM patients have
impaired anti-tumour immune response through multiple mechanisms and blockade or
modification of these inhibitory pathways provides an attractive therapeutic strategy.
Examples of these mechanisms are immunosuppressive microglia; T-regulatory cells (Tregs);
cytokine secretion patterns and the signal transducer STAT3.
26
Microglia
Within the GBM microenvironment cytotoxic T-Cell function is suppressed by dysfunction
of microglia. Microglia have been shown to be the main MHC II presenting cells in the CNS
as astrocytes express only low levels of MHC II molecules [81]. Within GBM tissue
microglia has impaired APC function and they may suppress the effect of cytotoxic T-Cells
through expression tumour promoting chemokines.
In the setting of GBM the MHC class II expressing microglia may be hi-jacked into
secreting pro-glioma factors such as monocyte chemo attractant protein 1 (MCP-1) [82].
Studies have shown that this is a powerful tumour promoter with MCP-1 transfected rats
displaying a significant reduction in survival. In the rats there was also an increase in
microglia infiltration suggesting they further promote tumour growth [83].
Tregs
An important mechanism by which GBM subverts the immune response is via regulatory T-
Cells (Tregs). Tregs are believed to prevent auto-immunity by controlling the proliferation of
auto reactive T-cells. Tregs inhibit expression of MHC II molecules, CD40, CD80, and CD86
through inhibition of IL-2, IFN-γ and increasing production of the TH2 class of cytokines.
This results in suppression of antigen presentation in monocytes and macrophages. Tregs can
be identified by the expression of forkhead box P3 (FoxP3) which is up regulated in GBM
tissue compared to both autologous peripheral blood and healthy controls [84, 85]. In GBM
tissue FoxP3 expression has been seen to be correlated to WHO tumour grade[86].
Furthermore serum Tregs are seen at higher proportions in GBM patients compared to normal
controls thus contributing to the systemic immunosuppression in these patients [87].
Depletion of Tregs in a mouse model has been shown to provide an anti-tumour effect with
improved survival however human data has not been correlated to survival [60, 86].
Excitingly for potential therapeutic applications in vivo depletion of these Tregs cells has
potentiated the effects of adoptively transferred T-cells [88]. In addition the
chemotherapeutic drug Temozolomide which is usually given concurrently with
immunotherapy, is thought to inhibit Tregs trafficking possibly contributing to its clinical
effectiveness [89].
Cytokines
27
As a GBM progresses a series of immune suppressive changes take place in terms of cytokine
secretion. There is reduced expression of IL-12, IFN-γ and TNF-α along with increased
expression of IL-4, IL-5, Il-6, IL-10 and immunosuppressive protein TGF-β2 and
prostaglandin E2 (PG E2) [90]. These act on multiple pathways to affect immune function
with the effect more pronounced on CD4+ than CD8+ T-cells [91]. An example is the
proliferation of T-cells and down regulation of monocyte MHC II in response to increased
tumour secretion of IL-10 [92].
Blockade of individual cytokines has shown some possible therapeutic benefit. Anti-sense
oligonucleotide targeted at TGF-β2 as shown to provide a survival benefit compared to
standard chemotherapy regimes in phase I/II trials but phase II trials were stopped due to
slow subject recruitment [93, 94].
Conversely cytokines can be directed against the tumour as a form of immunotherapy. An
example of which is the strategy of “arming” oncolytic viruses such as HSV with cytokines
in particular IL12 to potentiate their anti-tumour effect [95].Given the multiple pathways
contributing to the immunosuppressive environment and the general cytokine milieu it
becomes difficult to identify the relative significance of each. It is therefore unlikely that
suppression or potentiation of a single cytokine pathway will have significant therapeutic
effect. Blockade of common pathways or multiple cytokines would be more likely to confer a
clinical benefit.
STAT-3
Signal transducer and activator of transcription type 3 (STAT-3) is an important molecular
switch for tumour cells to avoid immune detection as well as enabling proliferation,
promotion of angiogenesis and resistance to apoptosis. STAT-3 is controlled by growth factor
receptor tyrosine kinases, cytokines and G-protein receptors which are also among the most
commonly activated oncogenic proteins. An example in GBM is phosphorylation of STAT-3
by EGFRvIII [96].
STAT-3 activity has also been shown to be a key regulator of systemic immune function in
the tumour setting. STAT-3 activity is induced within the tumour microenvironment and
modulates the activity of tumour infiltrating immune cells, reducing cytotoxicity of NK cells
and neutrophils [97]. In DC’s STAT-3 activity down regulates expression of MHC class II,
CD80, CD86 and IL-12 blocking their ability to stimulate T-Cells. These effects have been
reversed with STAT-3 ablation in hematopoietic cells in mice resulting in marked
28
improvement in the host anti-tumour immune response [97]. In microglia isolated from
GBM patients blockade of STAT-3 induced up regulation of co-stimulatory molecules,
promoted secreting of proinflammatory cytokines and induced T-Cell proliferation and
activation [98].
1.6. HCMV in GBM
Viruses have been implicated in a number of cancers, for example the commonly known
associations of cervical carcinoma with human papilloma virus (HPV) and hepatocellular
carcinoma with Hepatitis B & C. There are many more examples with varying mechanism of
oncogenesis but the principle of virus leading to malignancy is widely accepted [99].
A number of viruses have been implicated in brain tumours and studies have suggested that
exposure to common viruses resulting in diseases such as chickenpox or shingles may have
an effect on development of glioma in later life [100]. Exposure to contaminated polio
vaccine had been suggested to increase likelihood of glioma and this was experimentally
supported in hamsters by induction of tumours with this virus (simian virus 40) [101]. This
theory has not been conclusively borne out by epidemiological studies.
In contrast Wrensch has reported an inverse relationship between the presence of Varicella-
zoster viral infection and rate of GBM [102]. This study also reported that GBM patients
were less likely to have anti-bodies against Epstein-Barr virus (EBV) [102, 103].
Interestingly an increased likelihood of antibodies against herpes simplex virus (HSV) and
HCMV in GBM patients compared with controls was seen.
These studies have lead on to work establishing that HCMV is present in GBM and has
triggered interest in its role in the disease and potential as a therapeutic target [6].
Evidence for HCMV in GBM
The potential role of HCMV in GBM has been controversial [6]. HCMV is a common herpes
virus that infects 40-100% of the human population, in comparison to the prevalence of GBM
at 0.0257% [104]. Foetal HCMV infection is the most common infective cause of hearing
loss, mental retardation and birth defects. Infection in immunocompetent patients is usually
subclinical and 40% of these occur within the first year of life. In immunocompromised
patients symptoms may occur with fever, malaise and pneumonia the most common. Once
29
infected the HCMV is never cleared by the host leading to lifelong viral persistence.
Reactivation in immunocompetent individuals elicits an immune response by innate and
adaptive pathways.
HCMV infection in the setting of GBM was first raised by Cobbs et al [105]. In his study
they found immunohistochemical evidence of expression of the HCMV specific IE1-72
(immediate early) protein in all 27 of 27 gliomas examined, compared to none in brain
samples taken from controls (meningiomas, Alzheimer’s patient, normal brain). This
expression was limited to malignant tissue and not seen in surrounding brain or areas of
necrosis within the specimens. Further in-situ hybridisation (ISH) techniques were then used
to show the presence of HCMV immediate-early gene mRNA in these tumours. Again
endothelial tumour cells were positive along with some vascular smooth muscle whereas
controls and surrounding normal brain were not. These findings were not initially reproduced
and studies were published showing both presence and absence of HCMV in glioma tissue
[106-108]. Scheurer et al published a study in 2008 stressing the technical considerations
when extracting from paraffin-embedded tissue and showing a 100% rate of IE-1 protein in
21 cases of GBM using immunohistochemical methods [109]. These findings were supported
by ISH and are consistent with Cobb’s findings. Ongoing studies have shown the presence of
HCMV proteins US28, pp65, gB, IL-10 and pp28 along with the most commonly studied IE-
1 [20, 105, 107, 109-112]
As described above there is now strong evidence that HCMV viral genes are expressed in
GBM. If HCMV infection is present in GBM it could exist in either a lytic or latent state. In
the lytic state viral replication is driven by the IE genes regulating transcription of host and
viral genes and inhibiting apoptosis resulting in a cytopathic effect. Latency is characterised
by lack of production of virions and absence of expression of the lytic genes such as IE-1.
Currently the evidence points to HCMV in GBM existing in neither of these classical states.
In GBM HCMV IE-1 expression is seen uniformly but there have been no published studies
showing production of HCMV virions in GBM. The human cytomegalovirus and glioma
symposium consensus was that HCMV acts in a similar model that proposed for
cardiovascular disease [6]. In this model, persistent HCMV infection of vascular endothelial
cells resulted in viral expression without a cytopathic effect. This results in production of
cytokines and renin and thus hypertension [113]. In a tumour this chronic infection is
proposed to lead to expression of HCMV genes that induce production of cytokines that may
contribute to oncogenesis or interact with the cell cycle to cause disruption to normal
30
regulatory mechanisms. This is in contrast to the example of HPV which leads to direct
malignant transformation of the infected cells.
Oncomodulatory Role of HCMV in GBM
For a virus to be considered oncogenic they should show certain features that are lacking in
HCMV infection, such as sustained expression of oncoproteins and integration into the host
genome. Viruses with known oncogenic properties such as human papilloma virus (HPV) and
Hepatitis B virus display these characteristics, HCMV does not. Cinatl et al showed that in
neuroblastoma cells lines HCMV infection while not responsible for the malignant
transformation produced phenotypic changes such as increased expression of oncoproteins
and decreased expression of tyrosine hydroxylase and dopamine beta hydroxylase. They
postulated that the HCMV thus acts as a co-factor in the development of a fully malignant
cell phenotype [114]. This oncomodulatory role for HCMV fits with the known molecular
biology of HCMV [115]. HCMV activity has been shown to affect pathways altering cell
physiology a manner consistent with the 10 hallmarks of cancer which are sustained
proliferative signalling, Evasion of growth suppression, Invasion and metastasis, Immortality,
Angiogenesis, Resistance to cell death, Deregulation of cellular energetics, Avoidance of
immune detection, Genome instability and Inflammation [53][116].
In Vivo animal data supports this oncomodulatory effect of CMV. In one such study
transgenic (Mut3) mice engineered to generate gliomas were perinatally infected with murine
CMV (mCMV) [117]. These mCMV mice showed a markedly shortened median survival
time and tumours with a more aggressive behaviour. Human data showed a relationship
between CMV antigen expression and survival in GBM patients. A total of 75 patients with
GBM, 74 of who were HCMV positive were studied. Using immunohistochemistry and in-
situ hybridization to assess and grade viral expression they showed high levels of expression
were associated with a shortening of survival time by 20mths compared with low levels of
expression [118]. Patients with less than 25% CMV infected tumour cells at diagnosis
showed a 2 year survival benefit compared with those with 25-90% CMV infected tumour
cells. Time to tumour progression was 22 vs. 9 months and at 24 months, 60% of the patients
with low-grade CMV infection were alive vs. 20% with high grade infection. However the
31
link between expression of HCMV in GBM tissue and poorer prognosis has not been seen
consistently and remains controversial [119].
Given the recent interest in CMV in GBM a meeting of experts in the field was held to
produce a consensus report. The outcome provides a summary of the “state of play” for the
role of CMV in GBM and concluded with four key points[6].
1. “there is sufficient evidence to conclude that HCMV sequences and viral gene
expression exist in most, if not all, malignant gliomas”
2. “there is sufficient evidence to support the hypothesis that HCMV could modulate the
malignant phenotype in Glioblastoma”
3. “HCMV could serve as a novel target for a variety of therapeutic strategies”
4. 2 areas of research were suggested
a. Epidemiological study to determine why only small minority of latent HCMV
infection evolve into GBM
b. “elaboration of how HCMV contributes to glioma malignancy could identify
novel therapeutic targets”
Mechanisms of Oncomodulation
As discussed the idea that HCMV is active in GBM in an oncomodulatory role can be
examined in relation to each of the hallmarks of cancer described previously and summarised
in Table 1.
32
Table 1.1: HCMV biology and cancer hallmarks (from Dziurzynski et al, Consensus on the
role of human cytomegalovirus in Glioblastoma. Neuro-oncology. 14(3): 246-255, 2012)
Hallmark HCMV Activity Protein Involved
Sustaining
Proliferative Signalling
Induces cell cycle progression to S phase
Induces expression of E2F genes
Phosphorylates Rb
IE2
pp71
UL97
Evading Growth
Suppressors
Activates EGDFR
Dysregulation of Cyclin E expression
Inhibits p53 degradation
Decreases levels of p21
Induces expression of p%3
Binds to p53
HCMV infection
IE1
mtrll
Activating Invasion
and metastasis
Activation of RhoA dependant motility of
U373 cells
Activates smooth muscle cells
US28
Enabling Relative
Immortality
Activation of telomerase IE1
Inducing angiogenesis Induction of VEGF expression
Induction of IL-8
US28
IE1
Resisting Cell Death Inhibits apoptosis
Activates PI3-K/Akt pathway
IE1
IE2 vMIA, vICA
Avoiding immune
destruction
Production of homologs to
immunosuppressive cytokines
Inhibits expression of MCH 1
Intracellular retention of NKG2D
Induces expression of TGF-α1
HCMV IL-10
US2
UL16
IE2
Genome Instability and
Mutation
Chromosome damage
Inhibits DNA damage repair
Increases mutation frequency
Induces chromosome aberrations in cell lines
Unidentified
protein
HCMV infection
pp65 and pp71
IE1, UL76
Tumour Promoting
Inflammation
Induces production of RANTES, fraktalkine,
MCP-1
NF-γβ activation & IL-6 production
HCMV infection
US28
33
Sustaining Proliferative Signalling, Evading Growth Suppressors and Resisting Cell Death
Disruption of cell cycle regulation by p53 and Rb is an important hallmark of all cancers and
of proven importance in GBM. HCMV related proteins, pp71 and UL97, are able both to
degrade and phosphorylate these tumour suppressors [94, 120]. This has not been specifically
reported in GBM cells as yet. Persistent expression of IE1 has been evaluated in a number of
cells lines, this was shown to affect cellular proliferation differently, some showing increase
and other unchanged or decreased potentially via the MAPK and AKT pathway [121].
Activating Invasion and metastasis
HCMV infection of GBM cells have been shown to phosphorylate focal adhesion kinase
(FAK) Tyr397 [122]. This is a critical kinase for integrin-mediated migration and invasion
and its expression correlated with invasiveness in many malignancies [123]. The activation of
FAK by HCMV leads to increased extracellular matrix-dependent migration of GBM cells,
which is not seen in normal astrocytes [122]. Furthermore HCMV has been shown to down
regulate neural adhesion molecule (NCAM) in neuroblastoma cells resulting in enhanced
penetration through the endothelium [124]. Through these mechanisms HCMV promotes
invasion and metastasis.
Enabling Relative Immortality
Avoidance of apoptosis and immortalization is critical to development of malignancy.
HCMV has shown to convey resistance to apoptosis in fibroblast and neuroblastoma cells
[125, 126]. Immortalization of cancer cells requires activation of telomerase by expression of
the telomerase reverse transcriptase (hTERT) gene. After HCMV infection it has been shown
that an important regular transcription factor specificity protein 1 (Sp1) binds to the hTERT
promoter in fibroblast cells. In the same study GBM cells were examined and a direct
correlation between HCMV immediate early antigen (IEA) and hTERT expression was seen
in all samples tested [112, 127].
Angiogenesis
Angiogenesis in GBM is thought to be promoted by HCMV infection via US28 pathway. US
28 is a HCMV encoded G-protein coupled receptor which is expressed in early infection and
has been isolated in GBM vascular endothelial cells. It binds certain chemokines and
34
persistent expression has shown to result in over expression of IL-6 and VEGF in certain cell
lines [111]. When these cells were implanted into a nude mouse model, tumours resulted
[128]. In addition US28 up regulates NF-kB which increases production of IL-6 which in
turn activates STAT3 [111]. STAT 3 has effects both paracrine and autocrine, it is a key
molecular component of tumourigenesis and suppression of immune response especially in
GBM [129, 130]. STAT 3 induction in mouse neural progenitor cells has induced increase
VEGF expression, angiogenesis and GBM formation [131].
Genome Instability and Mutation
HCMV induced damage to DNA repair mechanisms has been reported in the literature in a
number of settings however not specifically in GBM [132]. Genomic instability due to
HCMV infection as been shown by infection of human neural progenitor cells resulting in
abnormalities of attachment, migration, loss of multipotency and down regulation of certain
genes producing abnormal and premature differentiation [133]. Chronic infection of stem
cells by HCMV has shown to produce specific chromosomal damage and therefore
potentially oncogenic [134]. Virions from 3 strains of HCMV have been shown to cause
breaks in chromosome 1 at 1q42 and a loss of this band has been seen in a small number of
GBM samples [135]. In addition recent evidence has shown that HCMV infection can alter
DNA methylation machinery. Response to chemotherapeutic agents such as Temozolomide
has been linked to methylation status of key genes such MGMT, recent evidence point to the
methylation status across the whole genome being of importance [25, 136].
Avoiding immune destruction and Tumour Promoting Inflammation
Immune evasion is a feature of both malignancy and HCMV virulence. The virus utilises
several mechanisms to evade the host response and promote suppression of the immune
system in the tumour microenvironment. HCMV has been shown to latently infect bone
marrow CD34+ cells in most healthy subjects [137]. The viral genome is then passed through
the myeloid lineage until inflammation or immunosuppression leads to terminal
differentiation into dendritic cells or mature macrophages and viral re-activation [47]. Within
GBM it has been shown that macrophages and microglia, which make up a significant
component of the microenvironment, are infected with HCMV [20]. HCMV infection creates
a unique M1/M2 pattern in the macrophages and microglia [138]. The M1 phenotype is pro-
inflammatory via induction of cytokines such as IL-6 and TNF-γ. These cytokines have been
35
previously linked to oncogenesis in animal models of inflammatory induced cancer [139].
TNF-α may also promote reactivation of HCMV as seen in immunosuppressed patients [140].
The immunosuppressive M2 phenotype appears to be of particular importance in GBM.
Glioma cancer stem cells (gCSC’s) produce HCMV IL-10 which is a homolog of human IL-
10 which is an immunosuppressive cytokine promoting the M2 phenotype [19]. This suggests
a feed-forward mechanism whereby HCMV induces a M2 phenotype that stimulates
migration of gCSC’s via increased VGEF and TGF-γ.
The evidence outlined above demonstrates some of the potential pathways by which HCMV
acts as an oncomodulatory virus in GBM rather than a true oncogenic virus.
1.9. Immunotherapy in GBM
Despite many promising therapeutic strategies there has been little change in outcomes for
patients with GBM prior to the introduction of Temozolomide [141]. Immunotherapy is an
attempt to manipulate or augment the patients’ immune system to better identify tumour
antigens and then destroy tumour cells [142]. Immunotherapy for GBM has been studied
since the 1950’s and 60’s with Mahaley’s work using monoclonal anti-bodies in CNS
tumours [57]. Initial trials of therapeutic agents have until recently been limited to early
phase trials delivered to patients with advanced disease, and while proving safety and
tolerability, gave only limited clinical benefit, if any [143]. It is known that CD8+ T-cell
infiltration of GBM tissue is associated with prolonged outcome showing that an improved
immune response can have clinical benefit [144]. In other types of cancer immunotherapy is
more established. The United States Food and Drug Administration (FDA) has approved
interferon alpha, interleukin-2 (IL-2) and the CTLA4 anti-body Ipilimumab for melanoma
and Sipuleucel-T (Provenge) for metastatic prostate cancer [145, 146]. Immunotheraputic
approaches to GBM can be classified into four types: active, passive, adoptive and
immunmodulatory. Active strategies involve direct challenge to the immune system with
tumour specific antigens typically as a vaccine. Passive strategies involve targeting tumour
antigen using effector molecules such as antibodies. Adoptive therapy involves harvesting
cells from the host and manipulating them ex vivo prior to returning them to the patient.
Immunomodulatory strategies involve manipulation of co-inhibitory and co-stimulatory
molecules.
The strategies currently showing the most promise for GBM are vaccination and adoptive cell
strategies.
36
Vaccination
Dendritic cells pulsed with tumour lysates is a promising vaccination strategy for GBM. This
approach ensures immunogenicity against many tumour antigens and potentially allows for
therapy tailored to each individual patient. As with many immunotherapies there remain the
problems of delay whilst the vaccine is prepared and the risk of autoimmune reactions. In
addition this approach requires tumour samples thus surgery which may not be feasible for all
patients. A number of studies have demonstrated that this is treatment can be feasibly
delivered with minimal adverse events [61].
An alternate vaccination strategy is the use of synthetic tumour antigens for targeting the
therapy. An example of this approach is the study by Sampson et al which utilised a peptide,
PEP-3, which was derived from the mutated EGFRvIII found in 30-40% of GBM’s and
conjugated to keyhole limpet hemocynain (PEP-3-KLH). In an animal tumour model mice
vaccinated with PEP-3-KLH showed resistance to GBM [147]. Subsequent phase I trials
demonstrated safety and development of an antigen specific immune response [148]. Phase II
trials have shown a median OS of 26mths which was significantly improved over controls.
Patient who showed EGFRvIII-specific antibody reactions had Median OS of 47.7mths vs.
22.8 in other subjects [149]. A Phase III trial of PEP-3-KLH (rindopepimut; Celldex
Therapeutics, Ma, USA.) is underway (Clinical Trials.gov identifier NCT01480479) [150].
Adoptive Cell Transfer
Adoptive cell transfer (ACT) refers to cell therapies using NK cells, lymphokine-activated
killer (LAK), tumour specific cytotoxic T-Cells (CTL) or tumour infiltrating lymphocytes
(TIL) harvested from a tumour bearing host. These autologous tumour specific cells undergo
ex vivo expansion and activation then are transferred back to the patient [7]. Harvesting of
these cells can be from peripheral blood, tumour tissue or draining lymph nodes [8]. Up to
75% of CD8+ cells undergo successful ex vivo expansion with a greater number of
circulating T-cells at harvest achieving a better response [151, 152]. ACT has been applied to
GBM employing various techniques with some promising early results[153, 154].
This technique provides a passive immunotherapy to give a tumour specific response. The
benefits of this approach are the ability to give large numbers of cells that have a high
specificity for tumour antigens. It also gives an opportunity for selection of effector cells that
show the best anti-tumour activity away from the in vivo environment and effects of
physiological or pathophysiological control [9]. The main effector cells for ACT are T-cells
however other cell types may have a role such as NK cells. The cells suitable for ACT may
37
occur naturally such or if the desired characteristics are not available naturally they can be
engineered, for example incorporating cytokines or co-stimulatory molecules [155-157].
The disadvantages of ACT are difficulty in generation of sufficient numbers of cells for
successful transfusion, ensuring adequate tumour antigen specificity, long term engraftment
and maintaining effector function [158]. A significant problem with ACT is transferring cells
which retain the ability to self renewal and thus have persisting effector function.[159] To
overcome this problem immunosuppressive cells and factors can be depleted through pre
treatment conditioning with chemotherapy or whole body irradiation [160]. Myeloablative
treatment with stem cell rescue has also been used with some success to allow proliferation of
transferred T-Cells [161]. As always any myeloablative therapy has the potential to leave the
patient prone to opportunistic infections.
ACT can be further divided depending on the type of cell used for transfer into tumour.
Rather than T-cells harvested from peripheral blood, TILs, and engineered T cells (ETC) can
be used. TIL therapy has been studied extensively in melanoma immunotherapy [7]. In this
strategy the T-Cells are taken from fresh tumour. They then undergo selection and expansion
using IL-2 and cell culture. This is limited by T-Cell yield from the tumour with only 30-40%
of surgical specimens having sufficient cell numbers. Also 6 weeks is required for production
of cells for infusion which is significant for aggressive tumours such as GBM. Despite these
limitations it has proven to be the most effective of the ACT methods with objective response
rates of greater than 50% in a number of melanoma studies [9].
Some mouse models of ACT have shown a synergistic effect with vaccination [162, 163].
As noted previously limitations for ACT include developing specificity for the T-Cells and
prolonging the persistence of the transferred T-Cells. T-Cells engineered to express tumour
antigen specific receptors would potentially overcome the poor antigen city of some tumours.
Genetic modification of the T-Cells prior to transfer to enhance the life of these cells may
reduce the need for potentially harmful myeloablative therapies prior to transfer.
Immunotherapy as a synergist with Adjuvant Therapies
Whilst there has been some encouraging results for immunotherapy in various cancer settings
it has become more likely that immunotherapy will be of most clinical benefit in conjunction
38
with other treatments such as cytoreductive surgery, chemotherapy and radiotherapy [164,
165].
A corner stone of standard treatment for GBM is ionising radiation. The primary intention is
for a direct cytotoxic effect on tumour cells. In addition to the direct effects it appears that the
tumour micro-environment is also affected along with generation of inflammation [166]. T-
Cells are known to be sensitive to even low doses of radiation allowing for re-setting of the
immune environment of tumours [164]. this sensitivity has been demonstrated in tumours
resulting in clearance of the tumour field post radiotherapy allowing for repopulation with the
transferred T-cells [167, 168]. In addition radiotherapy improved antigen presentation [169].
In mouse models MHC class I expression was enhanced by radiotherapy and improved
response to immunotherapy [170].
Chemotherapy and immunotherapy have been traditionally regarded as antagonistic based on
the assumption that chemotherapeutic agents result in immunosuppressant effects thus
diluting any effect of therapy and that they generally act by inducing apoptosis in target cells
which is non-inflammatory [171]. This assumption has been challenged by recent evidence
that immunotherapy and chemotherapy can act in a synergistic fashion. This may be through
elimination Tregs cells, activation of ACP’s, improved antigen presentation lymphopenia
resulting in homeostatic T-Cell proliferation, disruption of the tumour microenvironment and
increased susceptibility of tumour cells [165]. Wheeler et al compared patients who had
received dendritic cell vaccination in conjunction with chemotherapy and those with
chemotherapy or vaccination alone [172]. Combined therapy gave a 42% 2yr survival
compared with 8% in single agent therapy; with significantly improve time to progression.
Standard chemotherapy in GBM is Temozolomide. This has been shown to induce
lymphopenia and decreased Tregs cells through blockade IL-2 receptors. Administration of
anti-IL-2 anti-body whilst the patient was lymphodepleted reduced Tregs while allowing
vaccine induced anti-tumour effector cell response [173].
Given the possible interactions with existing proven therapies much care must be taken in
when in the cycle of treatment immunotherapy should be introduced. Given the potential
immune reset effects of radiotherapy and potentially synergistic effects with standard
chemotherapy the most appealing time would be post radiation alternating with concurrent
metronomic TMZ therapy.
39
Barriers to Immunotherapy and Antigen-Loss Variants
The two major problems that hamper adoptive immunotherapy remain:
1. The relative cost and technical difficultly in the ex-vivo expansion of suitable cells.
2. The temporary tumour effect if the transferred cells fail to persist or engraft.
These problems have lead to significant interest in vaccine therapies to circumvent these
problems. In addition to these ex vivo problems the phenomena of antigen-loss variants may
limit the effectiveness ACT.
Tumour antigen specific immunotherapeutic approaches have the clear benefit of reducing
the risk of autoimmune complications. In tumours antigen heterogeneity is the norm, thus
individual tumour cells may express different phenotypes. Immunodominance refers to the
preferential immunodetection of the dominant epitope of the many that are expressed by a
target. The cells expressing the dominant epitope for which the therapy is targeted against
will be cleared and cells expressing the previously recessive epitopes will survive allowing
tumour regrowth [174].
Vaccine studies have demonstrated the phenomena of antigen loss variants. As discussed
earlier EGFRvIII has been targeted with a vaccine conjugated to keyhole limpet hemocyanin
(PEP-3-KLH). In a mouse GBM model treatment failure with this vaccine was associated
with loss of EGFRvIII expression in 80% of cases [147]. This was also seen in phase II
human trials of the PEP-3-KLH vaccine indicating escape variants may be a mechanism of
resistance [149, 175].
Strategies such as whole tumour lysates or multiple antigen cocktails have be used in vaccine
therapy to prevent the development of antigen loss variants. This carries an increased risk of
autoimmune reaction. A more elegant approach would be individual patient screening for
suitable target antigens; this would however be difficult for widespread use in terms of
resources and cost. Another alternative would be for identification of suitable target antigens
on a population basis such as current practice with Flu vaccination.
It is now clear that the heterogeneity inherent in GBM affects the response to
immunotherapy. A study by Prins et al showed that in a dendritic cell vaccine trial those
subjects with a mesenchymal subtype had increased CD3+ and CD8+ TIL’s and greater
response to the experimental treatment [176]. The mesenchymal subtype is characterized by
40
over expression of pro-inflammatory genes and may represent a sub-type of tumour more
suited to immunotherapy.
1.9. HCMV Targeted Immunotherapy for GBM
In a pilot study by a group at the University of California, a single GBM patient was
administered dendritic-cells that had been pulsed with autologous GBM lysates as an adjunct
therapy. A significant increase in the HCMV-specific T cell response was observed [110].
This study highlighted the relative ease of eliciting an immune response against viral antigens
which is in contrast to the difficulty of immunization against "self" tumour antigens.
Anti-viral drug Valganciclovir which is used to treat HCMV infection has also been trialled
in the setting of GBM. In a randomized, double-blind, placebo controlled trial with 42
patients Stragiotto et al studied the use of Valganciclovir as an adjunct to conventional
“Stupp protocol” treatment in a series of GBM patients [177]. They found no significant
difference between groups with median overall survival of 17.9 vs. 17.4 months. Post hoc
explorative analyses showed in these taking >6mths Valganciclovir survival was improved
24.1mth vs. 13.1mths, with survival at 4 yrs 27.3% vs. 5.9%. These results suggest long term
suppression of HCMV in GBM patients may reduce the aggressive characteristics of the
tumour, thus improving survival.
Our laboratory has been trialing the use of CMV specific adoptive T-cell therapy [178]. In a
phase I trial in recurrent GBM patients had peripheral blood mononuclear cells (PBMC’s)
were isolated from blood samples and stimulated in vitro with CMV peptides and IL-2 then
infused as an adoptive T-cell therapy. As one may expect recurrent GBM patients have poor
life expectancy and of the 19 patients who were enrolled only 13 successfully had CMV
specific T-cells expanded, 4 patients withdrew due to progressive disease prior to vensection
and 2 had insufficient T-cells expanded due to poor precursor frequency. Unfortunately 2
patients died prior to infusion and one after only 2 infusions, leaving 11 patients who
underwent the full course of T-Cell infusions. In this group median survival was 403 days,
ranging between 133 and 2,428 days. Importantly 4 of the subjects remained free of
progressive disease within the study period.
There is a number of developing therapeutic strategies targeting HCMV for GBM therapy
there is no firm evidence of the utility of this approach and the field remains in its infancy.
41
1.10. Current Study
CD8+ T-cells play a primary role in immune response to HCMV infection in
immunocompetent individuals [179]. There is now significant evidence that HCMV is
expressed in GBM. Studies on a small number of GBM patients have demonstrated that
although these patients had circulating HCMV-specific T-cells at detectable levels similar to
healthy controls the majority of these cells were non-functional or displayed limited
functionality [10].
The aim of this thesis is to examine the role of HCMV immune response in GBM patients.
This will be done through clinical study and functional characterisation of Total and HCMV
specific T-cell responses in GBM patients through gene expression analysis. This will guide
further pathway analysis to assess the mechanisms by which T-Cell function is inhibited in
response to GBM. The ultimate aim is to identify specific pathways impairing the host
response to GBM and adapt therapeutic strategies to circumvent these.
Key Points:
• Despite intensive efforts GBM remains an aggressive incurable disease
• The hosts immune response to GBM is ineffective due to a range of mechanisms
• Evidence now suggests HCMV is present in GBM and modulates some of the
malignant characteristics
• Immunotherapy is a promising treatment strategy possibly utilising HCMV as a target
42
Chapter 2: Effects of CMV Serology status on Clinical Outcomes in GBM
2.1. Introduction
First identified in 2002 by Cobbs as being present in GBM there is now considerable
evidence that HCMV sequences and viral gene expression exist in most if not all malignant
gliomas. The evidence from current literature points to the virus acting not as an oncogenic
virus but rather modulating the tumour and potentiating some of the malignant characteristics
of GBM [6]. Studies showing an increased likelihood of antibodies against HCMV in GBM
patients have been important in establishing the link between the virus and the tumour [103].
In addition studies have suggested that the strength of viral expression with in the tumour
tissue is correlated with clinical outcome, however these studies have been contradictory to
date [118, 119].
This study therefore aims to examine the relationship between HCMV serology status and
clinical end points of time to progression and overall survival.
2.2. Methods
Patients’ diagnosed with Glioblastoma Multiforme (GBM, WHO grade IV) were enrolled in
an observational trial (QIMR ref p158, ethics approval from QIMR Berghofer Human
Research Ethics Committee and Uniting Care Health Human Research Ethics Committee.)
These patients had demographic information recorded and peripheral blood samples taken
after informed consent.
Histology was based on the pathology reports available from specimens received at surgery
and slides were not reviewed. CMV serostatus was based on serum IgG >6IU and IgM
>0.8TV regarded as seropostive.
Clinical information regarding treatment was retrieved from retrospective review of patients’
notes. Performance score was graded by the Karnofsky scale based on review of initial
43
clinical consultation. Initial treatment was assessed as whether they underwent the “gold
standard” therapy of gross total resection (GTR), radiating therapy (Rx) and con-current
Temozolomide (TMZ). Insertion of GLIADEL®
Wafer (polifeprosan 20 with carmustine
implant; Eisai Inc, Woodcliff Lake, New Jersey, USA) was recorded in operation notes.
Metronomic TMZ was defined as cyclic dosing of monthly 5 day cycles of treatment in an
ongoing manner. At recurrence the treatment was defined as surgery, second line
chemotherapy i.e. not TMZ, and inclusion in the adoptive T-cell therapy phase 1 trial being
conducted in the same institution.
Date of death was recorded for patients from clinical notes. If the patient was alive the last
date of follow up was recorded. Progression was based on evidence of progression of disease
on imaging criteria in conjunction with clinical notes, in most cases magnetic resonance
imaging (MRI) was the standard modality. The date of progression was recorded as the date
of the scan showing progression. In cases where pseudo-progression was queried PET scans
were often performed for confirmation.
11 of the patients who were HCMV seropostive went on to be enrolled in T-Cell adoptive
therapy phase I trial after recurrence, however not all of these patients then received T-cell
therapy[178]. Therefore further analysis was performed excluding these patients to avoid any
bias that may have been introduced.
Statistical Methods
The summary statistics of patients were HCMV seropostive and HCMV seronegative were
presented respectively by number (percentage) or median (IQR) as appropriate.
All caused mortality and progression free survival for the both groups were compared using
Kaplan-Meier curves and the log-rank test, and the corresponding median survival time was
obtained. For the analyses, time to death or censoring was defined as the time between the
last follow-up date and the date of diagnosis. Time to progression or censoring was defined
as the time between the date of the first recurrence or the last follow-up date and the date of
diagnosis.
Furthermore, using multivariate Cox regression models, all caused mortality and progression
free survival for the both groups were compared after adjusting for age and KPS. The
44
aforementioned analyses were also repeated for the subset of patients who did not participate
in the T-cell Adoptive Therapy phase I trial.
2.3. Results
80 patients were enrolled in the study. There were 6 patients for whom no follow up data was
available so were excluded from analysis. A further 8 patients had follow up available until
recurrence but were lost to follow up prior to death; these patients were included in analysis
for progression free survival. The overall cohort characteristics are summarized in table 2.1.
Table 2.1: Descriptive statistics on the study cohorts
HCMV - HCMV+
N* 29 (39.2) 45 (60.8)
Male* 19 (65.5) 28 (62.2)
Ageγ 52 (42, 60) 54 (49, 62)
KPSγ 90 (80, 90) 85 (80, 90)
Histology
Primary GBM* 22 (75.9) 35 (79.5)
Secondary GBM* 3 (10.3) 3 (6.8)
AA (WHO III) * 3 (10.3) 6 (13.6)
Brainstem Glioma* 1 (3.4) 0 (0)
Initial Treatment
GTR/Rx/TMZ* 25 (86.2) 42 (95.5)
Gliadel* 13 (44.8) 22 (50)
Metronomic TMZ* 29 (100) 40 (93)
Treatment for Recurrence
Surgery* 18 (75) 28 (68.3)
2nd Line Chemotherapy* 14 (60.9) 31 (75.6)
T-cell Trial* 0 (0) 11 (26.8)
*Number (%); γMedian (IQR)
In the cohort of 29 patients with HCMV– serology status, 66 % were male, median (IQR) age
was 52 (42, 60) years, and median (IQR) KPS was 90 (80, 90). 76 % had primary GBM, 10
% had secondary GBM and AA respectively and 1 % had Brainstem Glioma. 86 % received
GTR/RX/TMZ as initial therapy, 45 % received Gliadel and 100 % received Metronomic
45
TMZ. When the condition recurred, 75 % received further surgery, 61 % received second line
chemotherapy and 0 % received T-cell treatment.
In the cohort of 45 patients with HCMV+ serology status, 62 % were male, median (IQR) age
was 54 (49, 62) years, and median (IQR) KPS was 85 (80, 90). 80 % had primary GBM, 7 %
and 14 % had secondary GBM and AA respectively. 96 % received GTR/RX/TMZ as initial
therapy, 50 % received Gliadel and 93 % received Metronomic TMZ. When the condition
recurred, 69 % received further surgery, 76 % received second line chemotherapy and 27 %
received T-cell treatment.
The median overall survival of patients with GBM who were HCMV -ve was 719 days
compared to 780 days for those who were HCMV+; this difference was not statistically
significant. When patients who were enrolled in the therapeutic trial were excluded, the
median overall survival was 613 days for HCMV+ patients and 673 days for HCMV–. Log
rank tests comparing these groups with and without the trial patients (fig 2.1) demonstrated
that there was no significant difference in overall survival (p=0.553 all subjects and p=0.462
with trial patients excluded).
Median progression free survival was 344 days for HCMV +ve patients and 348 days for
HCMV–patients. With trial patients excluded, median progression free survival was 327 days
for HCMV+ patients and 299 days for HCM- patients. Log rank tests comparing these groups
with and without the trial patients (fig 2.2) demonstrated that there was no significant
difference in overall survival (p=0.453 all subjects and p=0.989 with trial patients excluded).
46
a)
b)
Fig 2.1: All cause mortality. Kaplan-Meier curves for all caused mortality for patients with
HCMV+ and HCM- serology; a) all patients. b) Patients who did not participate in the T-cell
Adoptive Therapy phase I trial.
a)
47
b)
Fig 2.2: Progression free survival. Kaplan-Meier curves for progression free survival for
patients with HCMV+ and HCM- serology; a) all patients. b) Patients who did not
participate in the T-cell Adoptive Therapy phase I trial.
48
After adjusting for age and KPS, HCMV+ serology patients seemed to have higher mortality
(HR 1.34, 95% CI 0.59 – 3.07) and recurrence risks (HR 1.43, 95% CI 0.56 – 2.07) compared
to patients with HCMV- serology, though statistically not significant (Table 2.2). The results
for patients excluding those who participated in the T-cell Adoptive Therapy phase I trial
were also similar.
Table 2.2: Hazard Ratios for Mortality and Progression Free Survival. Association of
HCMV+ serology status with all caused mortality and progression free survival compared to
HCMV- serology status, adjusted for age and KPS: Cox regression analyses, for all patients
and for patients who did not participate in the T-cell Adoptive Therapy phase I trial.
HR (95 % CI) p-value
Mortality
CMV+ 1.34 (0.59, 3.07) 0.486
CMV+ (No Trial Pts) 1.65 (0.69, 3.93) 0.263
Progression Free Survival
CMV+ 1.43 (0.78, 2.63) 0.251
CMV+ (No Trial Pts) 1.08 (0.56, 2.07) 0.823
49
2.4. Discussion
This study shows no significant difference in rate of progression and overall survival in GBM
patients between those that have positive HCMV serology and those who are negative.
The existence of HCMV in GBM has been controversial but the evidence has now been
replicated in a number of studies with the weight of evidence now pointing to HCMV being
present in GBM tissue and acting as an oncomodulatory virus, potentiating the malignant
characteristics [6]. This has been supported by studies showing that the strength of viral
expression in the tumour is correlated with clinical outcome [118]. Despite the presence of
CMV within tumour tissue, epidemiological studies have failed to show a clear link between
CMV sero-status and glioma risk. The relationship between the existence of CMV in glioma
and clinical aspects of the disease remains unclear [180]. In terms of clinical outcome a
recently published study addressed the impact of anti-HCMV antibodies (IgG and IgM) on
GBM outcome. Using cohorts from the Harris County Case-Control Study they found an
increasing risk of glioma with low levels of anti-HCMV IgG however there was no
association with survival and antibody level [181]. The current data supports this study in
showing that CMV serostatus has no effect on clinical outcome.
Prior exposure to HCMV in the host could the theory result in improved anti-tumour
response against HCMV antigen expressing GBM tissue from a pre-primed immune system.
However, as discussed in the introductory chapter HCMV may also contribute to immune
escape of the tumour though the complex immunoescape mechanisms that allow it to persist
for life. In addition it is already known that HCMV specific T-cells in GBM patients are less
functional which would correlate with the lack of clinical benefit of pre-exposure to HCMV
[10]. Recent studies questioning the effects of expression of HCMV in clinical outcome
would suggest that not only does the presence of HCMV have limited effect on the tumour
characteristics but that it does not alter the immune-reactivity of the tumour [119]. This
would correlate with the lack of difference between the two groups in this study.
There are a number of criticisms with this study that mean interpretation must been cautious.
Firstly it is a retrospective study with the inherent bias data collection in this manner. An
example of such bias may be seen in the recurrence data. As these patients were not enrolled
in a therapeutic trial the follow up was performed on an ad hoc non-standardised basis
50
meaning the time to progression could be inaccurate. However survival data is a robust
endpoint which is not as affected by the retrospective nature of the data collection.
Unfortunately a relatively large number of patients had incomplete follow up data potentially
negating this benefit. The problem of incomplete data also exacerbates the relatively small
sample size of 80 patients for a study of this type. Also tumour related confounding factors
such as molecular sub-type and MGMT methylation status are not accounted for in this
analysis. Other factors that may impair the correlation to clinical outcomes include lack of,
and limited information on some confounding variables such as functional and
socioeconomic status.
Key Points:
• There is no difference in survival or time to tumour progression between patients with
GBM who have positive HCMV serology and those who do not.
• This suggests prior exposure to HCMV does not affect clinical outcomes in GBM.
51
Chapter 3: Gene Expression analysis of CMV Specific and Total CD8+ T-cells in
patients with GBM.
3.1. Introduction
This study aims to isolate CD8+ T-cells and CMV specific CD8+ T-cells from GBM patents
using Fluorescence Activated Cell Sorting (FACS) to then allow custom gene microarray
analysis of expression of 92 genes that have been described to be important for regulation
of CD8+ T-cells during HCMV infection [182]. These genes are relevant for several
functional categories: cell division and metabolism, survival and apoptosis, effector
molecules, migration and adhesion, transcription factors, chemokines and cytokines, cytokine
receptors, differentiation and exhaustion [182]. These studies will help to functionally
characterise HCMV-specific T-cell responses in GBM patients and help guide further
pathway analysis to assess the mechanisms by which T-Cell function is inhibited in response
to GBM.
3.2. Methods
Patient Selection
Patients’ diagnosed with high grade gliomas, either Anaplastic Astrocytoma (WHO III) or
Glioblastoma Multiforme (WHO grade IV), were enrolled in an observational trial (QIMR ref
p158, ethics approval from QIMR Berghofer Human Research Ethics Committee and Uniting
Care Health Human Research Ethics Committee). These patients had blood samples taken
after informed consent and PBMC’s isolated by Ficoll Paque (Amersham Pharmaica Biotech,
Uppsala, Sweden) density centrifugation. These PBMC’s were then cyropreserved. Subjects
were selected at random from samples of both CMV negative and positive groups. Age, Sex
and CMV serology matched healthy control donors were selected from the laboratory sample
bank.
Flow Cytometry
PBMCs were revived from cyropreserved stocks and left to recover for 1h at 37 °C in growth
medium (10ml RPMI 10% Foetal Calf Serum). Samples were stained with APC labelled
52
pMHC multimersif CMV seropositive, followed by staining for surface markers with
antibodies CD3-PE-Cy, CD14-pacific blue, CD19-AlexaFluor450, CD8-PerCpCy5.5 and
CD4-FITC to allow for separation of monocytes, B cells and T cells. After filtering the cell
suspension through a nylon mesh for removal of cell clumps, total CD8+ T cell and HCMV
specific CD8+ T cell populations were isolated using a BD FACSAria™ Cell Sorting
System (Becton, Dickinson and Company, Franklin Lakes, New Jersey, USA).
Gene expression analysis
The CD8+ and CMV specific CD8+ cells underwent RNA isolation using RNeasy Micro Kit
(Qiagen) according to manufacturer’s instructions. RNA equivalent to 3000 cells was then
transcribed to cDNA using AB high capacity RNA-to-cDNA kit. The cDNA then underwent
pre-amplification with Taqman PreAmp Master Mix kit prior to gene array analysis
performed with Taqman Gene expression assay and a TaqMan Array Micro Fluidic Cards
(see appendix 1 for list of genes, all reagents from Life Technologies) using ABI Viia 7 real
time PCR by the comparative CT method. Data was imported into GeneSpring v12.5, and
normalised to housekeeping genes 18s, actin and b2m. Data was filtered to remove any
target that was not expressed at a Ct value of less than 35. Groups were then compared using
an unpaired t-test with or without multiple testing correction (MTC) using the Benjamini-
Hochberg method. Results where p<0.05 were considered significant. Genes where there
were significant differences in expression between sample populations and controls were
entered into IPA (Ingenuity® Systems, www.ingenuity.com) software for pathway analysis.
Clinical Data
Clinical data was gathered from review of the patient files. Clinical outcomes were measured
by time to progression defined by progression on follow up imaging and overall survival
based on date of death. Sub-groups for analysis were time to progression < 1yr and > 1 yrs,
Overall survival >18mths and < 18mths.
53
3.3. Results
20 subjects were selected for analysis with 10 of these having CMV negative serology and 10
CMV positive. The sample population clinical data is shown in table 3.1.
Table 3.1: Characteristics of Subjects for Gene Expression Analysis
Subjects CMV+ Subjects CMV- Total
Number 10 10 20
Age (median) 49 yrs 58 yrs 55 yrs (range 21-74)
Gender
Male 3 5 8
Female 7 5 12
Histology
Primary GBM 7 8 14
Secondary GBM 2 1 4
Anaplastic
Astrocytoma
1 1 2
Time to recurrence
(median)
352.5 days
(range 84- 4651)
Overall Survival
(median)
649 days
(range 236- 4851)
In all subjects sufficient total CD8+ cells were isolated to allow for RNA extraction, however
the PCR reactions failed in 2 samples leaving 8 CMV +ve subjects and 10 CMV-ve subjects
for analysis.
Inadequate numbers of CMV specific CD8+ cells were isolated from the PBMC samples
following staining and FACS sorting to enable gene expression analysis in 4 CMV +ve
subjects leaving 6 samples of CMV specific CD8+ cells in which gene expression failed in 1
leaving 5 subjects for CMV specific analysis. After analysis statistically significant results
were reviewed.
54
Analysis of CMV specific T-cells in GBM patients compared with controls showed
statistically significant down regulation of a number of genes (table 3.2). These genes
represented two distinct subgroups related to effector function (GZMA, GZMB and SPON2)
and the exhausted phenotype of T-cells usually seen in persistent virus infection (KLRD1,
ZNF683, CD244, and PTGER2) [182]. There was also up regulation of the PDCD1 gene
which codes for PD-1 a co-inhibitory molecule. Analysis of the total CD8+ T-cell population
in GBM patients compared with controls showed a pattern of down regulation of effector
promoting genes and up regulation of inhibitory genes shown in table 3.2. The genes that
were up regulated included: BTLA a co-inhibitory molecule; TYMS and CDCA7 which code
for proteins critical in DNA replication; cytokine genes IL7R, CCL4L1 and CX3CR1;
GZMB which codes for a serine protease and NR3C2 a nuclear mineralocorticoid receptor
with unknown function in immune cells. There was down regulation of TNFS14 (LIGHT) a
co-stimulatory molecule and GZMK a granzyme protease.
The sub-group analysis based on clinical outcomes of time to progression and overall
survival is shown in table 3.3 and fig.3.3 and fig 3.4. In patients with longer progression free
survival there was down regulation of the pro-apoptotic BAD gene, chemokines receptor
CXCR5 and cell cycle enzyme TYMS. Improved overall survival was associated with down
regulation of BTLA and NR3C2.
55
Fig 3.1: Changes in gene expression in CMV Specific T-Cells of GBM. Heat map and graph showing the genes for whom statistically
significant Fold Change (shown with standard deviation) was observed in CMV specific T-Cells between GBM patients and age, sex and HLA
type matched healthy controls
GBM Control
56
Fig 3.2: Changes in gene expression in the total population of CD8+ T-Cells of GBM
patients. This table shows the gene for whom statistically significant Fold Change (shown
with standard deviation) was observed in the total population of CD8+ T-Cells between
GBM patients and age, sex and HLA type matched healthy controls.Table 3.2:Changes in
Gene Expression for CMV Specific T Cells in GBM patients.
57
Table 3.2:Changes in Gene Expression for CMV Specific T Cells in GBM patients.
Gene
Code Name
Regulatio
n
Fold
Chang
e
P=
Description Also known as
PDCD1 Programmed cell death
protein 1
up 4.223 0.002 Immune regulation CD279; PD-1; PD1; SLEB2; hPD-1;
hPD-l
CD244 n/a down -3.483 0.001 Natural Killer Cell Receptor 2B4; NAIL; NKR2B4; Nmrk;
SLAMF4
GZMA Granzyme A down -3.859 0.026 T-Cell and NK cell serine protease CTLA3; HFSP
GZMB Granzyme B down -3.95 0.036 T-Cell and NK cell serine protease CCPI; CGL1; CSP-B; CSPB; CTLA1;
CTSGL1; HLP; SECT
ZNF68
3
Zinc Finger Protein 683 down -4.836 0.017 May be involved in transcriptional regulation Hypothetical Protein MGC33414
PTGER
2
Prostaglandin E2 receptor down -4.972 0.003 Prostaglandin receptor for prostaglandin E2 EP2
KLRD
1
Killer cell lectin-like receptor
subfamily D 1
down -6.02 0.002 Killer cell lectin-like receptor CD94
SPON2 Spondin 2 down -12.936 0.007 Cell adhesion protein, a chemoattactant for
monocytes and macrophages
DIL-1, M-Spondin, Mindin
58
Table 3.3: Changes in Gene Expression for total CD8+ T-cells in GBM patients.
Gene
Code Name Regulation
Fold
Change P= Description Also known as
ZNF683 Zinc Finger Protein 683 up 3.84 0.02 May be involved in transcriptional regulation Hypothetical Protein
MGC33414
TYMS Thymidylate synthase up 3.71 0.017 A critical enzyme in DNA replication HST422; TMS; TS
CDCA7 Cell division cycle-associated protein
7
up 3.33 0.001 Induces growth and clonogenicity in
lymphoblastoid cells
JPO1
NR3C2 nuclear receptor subfamily 3, group C,
member 2
up 3.19 0.023 mineralocorticoid receptor MCR; MLR; MR; NR3C2VIT
CCL4L1 C-C motif chemokine 4-like up 3.17 0.019 Codes for CC motif chemokines 4- like AT744.2; CCL4L; LAG-1;
LAG1; SCYA4L
CX3CR1 CX3C chemokine receptor 1 up 3.00 0.029 Cytokine receptor CCRL1; CMKBRL1;
CMKDR1; GPR13; GPRV28;
V28
BTLA B- and T-lymphocyte attenuator up 2.82 0.013 T-Cell inhibition via interaction with tumor
necrosis family receptors
BTLA1; CD272
GZMB Granzyme B up 2.59 0.028 T-Cell and NK cell serine protease CCPI; CGL1; CSP-B; CSPB;
CTLA1; CTSGL1; HLP; SECT
CD244 n/a up 2.37 0.031 Natural Killer Cell Receptor 2B4; NAIL; NKR2B4; Nmrk;
SLAMF4
IL7R interleukin-7 receptor up 1.89 0.017 Cytokine receptor CD127
GZMK Granzyme K down -1.86 0.034 T-Cell and NK cell serine protease TRYP2
TNFSF14 Tumor necrosis factor ligand
superfamily member 14
down -2.18 0.015 A costimulatory ligand for HVEM CD258; HVEML; LIGHT;
LTg; TR2
59
Fig 3.3: Changes in gene expression in the total population of CD8+ T-Cells in Long
surviving GBM patients. Bar graph showing the fold change (with SD) for genes with
statistically significant differences in expression in the total population of CD8+ T-Cells
of GBM patients who survived > 18mths compared with those who survived <18ths.
60
Fig 3.4: Changes in gene expression in Total CD8+ T-Cells of GBM patients based on
progession free survival. Bar graph showing the fold change (with standard deviation) for
genes with statisicatlly significant differences in expression in the Total population of CD8+
T-Cells of GBM patients who had progession free survival > 12mths compared with those
who had progression free survival <12mths. Table 3.4:Changes in Gene Expression for
total CD8+ T-cells in GBM patients based on clinical outcomes (time to progression and
overall survival.
61
Table 3.4:Changes in Gene Expression for total CD8+ T-cells in GBM patients based on clinical outcomes (time to progression and overall
survival)
Gene
Code Name
Regulatio
n
Fold
Change P= Description Also known as
Total CD8+ T-cell population in GBM patients; Time to Progression >12mths vs. <12mths.
BAD Bcl-2-associated death promoter down -3.113 0.037 Pro-apoptotic member of the Bcl-2 gene family
involved in initiating apoptosis. BBC2; BCL2L8
CXCR5 Chemokine receptor CXCR3 down -4.502 0.037 G protein-coupled receptor in the CXC chemokine
receptor family BLR1; CD185; MDR15
TYMS Thymidylate synthase down -6.380 0.01 A critical enzyme in DNA replication HST422; TMS; TS
CD8+ T-cell population in GBM patients; overall survival >18mths vs. <18mths.
NR3C2 nuclear receptor subfamily 3, group
C, member 2 up 5.147 0.033 mineralocorticoid receptor
MCR; MLR; MR;
NR3C2VIT
BTLA B- and T-lymphocyte attenuator down -3.709 0.026 T-Cell inhibition via interaction with tumour necrosis
family receptors BTLA1; CD272
62
Pathway analysis of the changes in gene expression outlined above resulted in significant
association for the CMV specific group with the Cell Death and Survival, Cell-To-Cell
signalling and Interaction networks. The total T-Cell population analysis showed significant
association with the Cellular development, Cellular Growth and Proliferation,
Haematological System Development and Function network (Fig. 3.5 and 3.6).
Fig 3.5: Key for Pathway Analysis Diagrams. (IPA; Ingenuity® Systems, www.ingenuity.com)
63
Fig 3.6: Pathway Analysis of Gene Expression Data from the CMV Specific T-Cell
population in GBM patients. Coloured genes are those identified by the gene expression
analysis in the current study. The solid lines represent direct interactions between genes,
dotted lines represent and indirect interaction.
64
Fig 3.7: Pathway Analysis of Gene Expression Data from the total T-Cell population in
GBM patients. Coloured genes are those identified by the gene expression analysis in the
current study. The solid lines represent direct interactions between genes, dotted lines
represent and indirect interaction.
65
3.4. Discussion
This analysis of CMV specific CD8+T-Cells in GBM patients showed down regulation of
genes that are important for effector function and also a number of markers of the exhausted
phenotype of T-cells usually seen in persistent virus infection [183]. These genes were
GZMA, GZMB, SPON2, KLRD1, ZNF683, CD244, and PTGER2. Down regulation of
these genes could lead to the impaired functionality that has been demonstrated in this sub-
population of CD8+ cells in GBM patients.
KLRD1, CD244 and PTGER2 are markers of the exhausted T-Cell phenotype seen in
persistent virus infection. This is a subset of cells characterised by poor function with
reduced proliferation, cytotoxicity and cytokine secretion in response to stimulation [71].
This finding would contradict the suggestion that many of the CMV specific CD8+ cells are
of this exhausted phenotype. This may indicate inhibitory mechanisms other than the
exhaustion pathway may be of more importance in GBM mediated inhibition of T-cell
function. Supporting this idea is that ZNF683, a homologue of PRDM1 which codes for
Blimp-1 a transcription factor critical for establishment of the exhausted phenotype, was seen
to be down regulated [182]. There was up regulation of PDCD1 which codes for immune
modulating co-inhibitory molecule PD-1. Other key immune modulating molecules such as,
Tim-3 and LAG3 genes were unchanged.
In contrast to the pattern seen in the CMV specific cells the total CD8+ population showed
up regulation of many genes associated with T-cell exhaustion such as CD244, ZNF683
suggesting that this pathway may be of importance [71]. Other up regulated genes included
cytokine receptor IL7R and chemokine receptor CX3CR1 and CCL4L1; genes important for
the cell cycle and division CDCA7 and TYMS; the mineralocorticiod receptor NR3C2 which
has unknown function [182].
In addition to the genes already mentioned analysis of the total CD8+ T- Cell population in
GBM patients revealed up regulation of BTLA and down regulation of the gene coding for
LIGHT. These co-molecules when expressed in the surface membrane of the cell interact
with the same receptor herpes virus entry mediator (HVEM). BTLA forms an inhibitory
complex on interaction with HVEM and is usually down regulated as naive T-Cells
differentiate to effector cells. The LIGHT protein binds to HVEM to give a co-stimulatory
signal to T-Cells promoting proliferation and effector function [184]. The pattern of up
66
regulation of the inhibitory BTLA and the down regulation of LIGHT is therefore consistent
with overall inhibition of T-Cell function.
In patients with longer progression free survival there was down regulation of pro-apoptotic
BAD gene, chemokine receptor CXCR5 and cell cycle enzyme TYMS. Interestingly there
was an association with improved overall survival with down regulation of BTLA which
supports a possible role in tumour mediated immune suppression as discussed earlier. There
was also down regulation of NR3C2, the role of which is unclear in immune regulation.
All of the changes in gene expression identified by this data may be of importance and
require further pathway analysis to delineate their importance.
Network analysis of the gene expression data showed a common convergence in both the
CMV specific and total T-cell population on the IL12 complex. This cytokine has been of
interest as it is a promoter of the TH1 immune response [95]. IL12 has been used in oncolytic
virus therapy in GBM and other cancers. These viruses replicate within tumour but not
normal tissue and act to attack the tumour via their inherent cytolytic activity and by
generating an immune response [95]. Indirect interaction with IL12 shown in this pathways
analysis may suggest that this cytokine could be important in GBM T-cell interaction.
In summary this study identifies a number of candidate genes involved in T-Cell inhibition in
GBM. One potential pathway for further investigation from this study is the HVEM pathway
and its ligands BTLA and LIGHT.
Key Points:
• In GBM patents CMV Specific CD8+ T-cells show a pattern of down regulation of
genes promoting effector function
• In GBM patients the Total CD8+ T-cell population displays a pattern of up regulation
of inhibitory genes and down regulation of promoter genes
• The HVEM/BLTA/LIGHT pathway may be an important inhibitory pathway and
warrants further investigation
• Network analysis highlights the potential importance of IL12 in GBM immune
avoidance.
67
Chapter 4: The HVEM, BTLA and LIGHT network in GBM patients
4.1. Introduction
The results of gene expression analysis of CD8+ T-cells in the previous chapter revealed that
in GBM patients the total CD8+ population showed up regulation of the inhibitory co-
stimulatory molecule BTLA and down regulation of the co-stimulatory molecule LIGHT.
These molecules are both ligands for the HVEM receptor. This pathway is therefore a good
candidate for further investigation.
The aim of this set of experiments is to confirm the findings of the gene expression analysis
and confirm that the changes seen in BTLA/LIGHT/HVEM expression are seen in the
expression of the proteins they code for on the cell surface using FACS analysis.
This network consists of the Herpes Virus Entry Mediator (HVEM) receptor and its ligands
LIGHT ( Lymphohotoxin-like, exhibits Inducible expression and competes with herpes
simplex virus Glycoprotein D for HVEM, a receptor expressed by T-
lymphocytes/TNFSF14/CD258) and BTLA (B and T Lymphocyte Attenuator/CD272).
HVEM was first identified as a route of cell entry for HSV; it is now known to be an
important regulator of the immune response signalling through JNK1, NF-κB and AP-1 in
response to infection [185]. It is expressed in blood vessels, brain, heart, kidney, liver, lung,
prostate, spleen and thymus, along with resting T-cells and naive memory B-cells [94]. It has
the unique property of being able to interact with the members of the two superfamilies of co-
signalling molecules the TNFR family and Ig family in LIGHT and BLTA, respectively. The
extracellular compound of HVEM has three cysteine rich domains at the C-terminus, the
characteristic of the TNF superfamily. On the N terminus of HVEM has an area known as the
DARC (DARC for gD and BTLA binding site on the TNFR HVEM in CRD1) side which is
the site of BTLA binding [186]. HVEM is strongly expressed in T-cells in their resting state
and down regulated on activation [187].
BTLA is a 70KDa type I membrane glycoprotein of the CD28 family. BTLA binding with
HVEM results in SHP1/SHP2 phosphatase activity inhibiting T-cell activation. Normal
expression is seen on B cells, T-Cells and dendritic cells. During normal differentiation of
naive T-cells to effector cells BTLA expression is down regulated.
68
LIGHT is a type 2 transmembrane protein commonly expressed in spleen, brain, kidney and
activated CD8+ T-Cells. Normal expression of this receptor is on T-cells, B-Cells, NK cell,
monocytes and immature DC’s.
The binding of LIGHT to HVEM results in T-cell stimulation, B-cell co-stimulation, plasma
cell differentiation, Ig secretion and the maturation of DC’s. In addition LIGHT is believed to
down-regulate its own ligand, HVEM, as a self regulatory mechanism [188]. Interaction of
HVEM and LIGHT must be between adjacent cells (trans) due to the crystalline structure. On
the other hand BTLA can be co-expressed on the same cell resulting in interaction (cis).
Interaction of BTLA with HVEM in a cis or trans configuration forms a HVEM/BTLA
complex blocking interaction with LIGHT and overall an inhibitory signal is generated. Thus
HVEM has bidirectional signalling. This mechanism is believed to act in a cis- configuration
to regulate T-cell function by blocking immune “noise”. Within tumours HVEM expressed in
the tumour tissue is believed to act in a trans- configuration to modulate the immune
response [185].
Figure 4.1: The HVEM/BTLA/LIGHT pathway. HVEM and BTLA affect cell function
both through Trans interactions between cells and Cis configurations between molecules on
the same cell. ( modified from: Paulos et al. Putting the brakes on BTLA in T cell–mediated
cancer immunotherapy. J Clin Invest. 120(1): 76–80, 2010)
Interestingly a number of malignancies have shown elevated expression of HVEM [189].
This includes chronic lymphocytic leukaemia (CLL), lymphoma, a number of solid tumours
as well as metastases and melanoma [187]. In melanoma tumour specific T-cells have been
shown to have persistent BLTA expression and reduced production of IFN-γ [190]. Up-
regulation of BTLA and PD-1 has been shown to be important in inhibition of T-cell
69
expansion. Blocking of BTLA along with other inhibitory receptors has been shown to
potential reverse the hypo-responsiveness of T-cells in melanoma patients (Fig 1.7) [184]. In
tumour models forced expression of LIGHT has resulted in T-cell activation and anti-tumour
effect [191].
Figure 4.2: The HVEM Network in Tumour Models. HVEM/BTLA/LIGHT interaction in
a trans configuration with tumour specific T-cells in melanoma models resulting in an overall
inhibitory signal that is theoretically released by blockage of BTLA (modified from: Pasero
et al. The HVEM network: new directions in targeting novel co-stimulatory/co-inhibitory
molecules for cancer therapy. Current Opinion in Pharmacology. 12(4: 478-85, 2012.)
4.2. Methods
Peripheral blood samples were taken from patients enrolled in an observational trial (QIMR
ref p1558, ethics approval from Bancroft Centre human Research Ethics Committee and the
70
uniting Health Human Research Ethics Committee). PBMC’s were isolated and
cyropreserved as previously described. Twenty Subjects were selected with ten subjects
being HCMV seropostive and ten HCVM seropositive. Age, Sex and CMV serology matched
healthy control donors were selected from the laboratory sample bank. From the
observational trial cohort 15 subjects had tissue samples available for immunohistochemistry
analysis.
Flow Cytometry
PBMCs were revived from cyropreserved stocks and left to recover for 1h at 37 °C in growth
medium (RPMI 10% FCS). Samples were stained with APC labelled HCMV dextramers
followed by staining for surface markers with antibodies CD8-AF700 and CD4-FITC to
allow for separation of CD8+ T-cells. The cells where then stained with anti-BTLA-PerCP,
anti-HVEM, and anti-LIGHT-biotin. Secondary staining was performed with BV421-
Streptavidin and the cells were preserved in PBS-2% Paraformaldehyde. Analysis was then
performed using LSR 4 laser Fortessa Flow Cytometer (Becton, Dickinson and Company,
Franklin Lakes, New Jersey, USA with FACS Diva software (BD Biosciences). Post-
acquisition analysis was conducted using FlowJo software (TreeStar). In each group the % of
total CD8+ staining positive for each molecule was calculated along with mean florescence
intensity (MFI). Comparison was made between GBM patients and the control group along
with total CD8+ and CMV specific CD8+ groups using unpaired t-test on GraphPad Prism
version 6.00 for Windows, (GraphPad Software, La Jolla California USA,
www.graphpad.com.) with p < 0.05 considered significant.
Immunohistochemistry
Surgical samples were fixed in 4% neutral buffered formalin overnight then transferred to
70% ethanol before processing. Paraffin wax processing consisted of incubating the samples
in a series of ascending grade alcohols from 70% to 100%, clearing in xylene and infiltrating
with molten paraffin wax at 60oC over a period of several hours. 4µm paraffin sections were
cut from embedded samples and floated onto a 37oC waterbath and collected onto Menzel
Superfrost Plus slides and stored at 4◦C prior to immunohistochemical staining.
Immunohistochemical staining for HVEM was carried out using mouse monoclonal anti-
body to HVEM (MM0332-5S1, ab89479, Abcam, Cambridge, UK). Sections were dewaxed
and rehydrated through descending alcohol to water series using standard protocols.
Specimens were then subject to 20 minutes heat antigen retrieval at 121oC in a pressure
cooker. Endogenous alkaline phosphatase was blocked by incubating slides in Levamisole
71
and nonspecific antibody binding was inhibited by incubating the sections in Background
Sniper blocking reagent. The sections were then incubated overnight at 4oC with the primary
antibody diluted 1:70 in Da Vinci Green antibody diluents. After washing MACH2 Mouse -
alkaline phosphatase secondary antibody was applied for 45 minutes. Signals were developed
in Warp Red chromogen for 10-15 minutes (all reagents Biocare Medical). Finally sections
were lightly counterstained in Haematoxylin and dehydrated through ascending graded
alcohols.
Two independent Pathologists analysed each slide at high power. Expression of HVEM was
graded according to intracellular location (cytoplasm, membrane or intra nuclear) along with
intensity (grade 0 = no signal, 1 = weak, 2 = moderate, 3 = high).
4.3. Results
The % of total cells positive for each molecule along with the mean fluorescent intensity
(MFI) was calculated for each sample. A representative example of the gating strategy is
shown in Fig 4.1.
In the total CD8+ analysis mean BTLA expression in GBM vs. controls was 2.16% (SEM +/-
0.3187) vs. 3.804% (+/- 0.7839, p=0.0594) and 88.91 (+/- 5.978) MFI vs. 115.2 MFI (+/-
10.40, p= 0.0348). HVEM expression was 90.27% (+/- 1.335) vs. 94.60% (+/- 1.335,
p=0.1083) and1056 MFI (+/- 60.30) vs. 1405 (+/- 116.8, p= 0.0115). LIGHT expression was
6.967% (+/- 1.703)) vs. 14.53% (+/- 3.915, p=0.0846) and 124.9 MFI (+/- 9.757) vs. 160.6
(+/- 22.36, p= 0.1516). These results are summarized in Fig 4.2.
In the CMV specific CD8+ group analysis of mean BTLA expression in GBM vs. controls
was 5.150% (SEM +/- 1.756) vs. 2.329% (+/- 0.7465, p=0.1163) and 90.68 (+/- 17.56) MFI
vs. 91.97 MFI (+/- 9.603, p= 0.9447). HVEM expression was 92.80% (+/- 3.259) vs. 93.81%
(+/- 1.693, p=0.7648) and 964.5 MFI (+/- 82.47) vs. 1006 (+/- 87.37, p= 0.7605). LIGHT
expression was 70.60% (+/- 5.742) vs. 69.10% (+/- 7.251, p=0.8912) and 153.8 MFI (+/-
18.87) vs. 195.4 (+/- 48.21, p= 0.5474). These results are summarized in Fig 4.3.
72
Fig 4.3: Analysis of BTLA, LIGHT and HVEM expression in CD8+ T-Cells. A represenative
analysis of FACS gating for CMV Specific CD8+ T-Cells demonstrating expression of BLTA,
LIGHT and HVEM including FMO controls.
73
Hea
lthy
GBM
0
5
10
15
% B
TL
A+ o
f C
D8+ p = 0.05940
Hea
lthy
GB
M
0
50
100
150
200
250
MF
I fo
r B
TLA
in
CD
8+
p = 0.03475
Hea
lthy
GB
M
50
60
70
80
90
100
110
% H
VE
M+
of
CD
8+
p = 0.1083
Hea
lthy
GBM
0
500
1000
1500
2000
2500M
FI o
f H
VE
M i
n C
D8+ p = 0.01152
Hea
lthy
GB
M
0
20
40
60
80
% L
IGH
T+
ve o
f C
D8+
p = 0.08458
Hea
lthy
GB
M
0
200
400
600
MF
I o
f L
IGH
T in
CD
8+
p = 0.1516
Fig 4.4: T-Cell Expression of BTLA, HVEM and LIGHT. Plot of % and median flurecent
intencity (MFI) of total CD8+ cells expressing BTLA, HVEM and LIGHT .
74
Hea lt
hy
GBM
0
5
10
15
% B
TL
A +
of
CM
V S
pecif
ic C
D8+
p = 0.1163
Hea
lthy
GBM
0
50
100
150
MFI o
f B
TL
A in
CM
V S
pec
ifi c
CD
8+
p = 0.9447
Hea
lthy
GBM
75
80
85
90
95
100
105
% H
VE
M+
in
CM
V S
pecif
ic C
D8+
p = 0.7648
Heal
thy
GBM
0
500
1000
1500M
FI o
f H
VE
M in
CM
V S
pecif
ic C
D8+
p = 0.7605
Hea
lthy
GBM
0
50
100
150
% L
IGH
T +
of
CM
V S
pe
cif
ic C
D8
+
p = 0.8912
Hea
lthy
GBM
0
100
200
300
400
500
MF
I of
LIG
HT
in
CM
V S
pecif
ic C
D8
+
p = 0.5474
Fig 4.5: CMV Specific T-Cell Expression of BTLA, HVEM and LIGHT. Plot of % and
MFI of CMV specific CD8+ cells expressing BTLA, HVEM and LIGHT.
75
Immunohistochemistry
15 tumour samples were stained for HVEM. HVEM expression was seen in malignant cells
in 6 of the 15 samples as shown in Table 4.2. In all samples lymphocytes were identified and
assessed separately from the malignant tumour cells, however in only 1 specimen HVEM
staining was seen and this was not included in this analysis. The distribution of staining
analysis was all cytoplasmic in nature. Some intra-nuclear staining was seen but this was
thought to be artefact and not included as a positive result. Also membrane staining was seen
in only one sample and therefore not thought to be reliable and not included for analysis.
Table 4.1: HVEM Expression in GBM by Immunohistochemistry
HVEM Expression None 1 2 3
Number of Tumours (n=15) 9 2 3 1
Percentage of Tumours 60% 13% 20% 7%
76
Fig 4.6: Immunohisotchemical staining for HVEM in GBM. Slides showing garding scale of immunohisotchemical staining for HVEM in
GBM tissue. A) Normal Brain with no staining., B) Grade 1, C) Grade 2 and D) Grade 3.
77
4.4. Discussion
These results demonstrate that in total and CMV specific CD8+ T-Cells isolated from the
peripheral circulation there are no significant differences in surface expression of
BTLA/LIGHT and HVEM between healthy controls and patients with GBM.
There was significantly greater expression of BTLA and HVEM on total CD8+ T-cells in
healthy controls compared to GBM patients in MFI, which in the case of HVEM, was not
confirmed in terms of % of cells expressing that molecule. This was also at odds with the
results of the gene expression analysis which showed greater expression of BTLA in GBM
patients’ compared to healthy controls. This result therefore should be treated with some
scepticism.
The lack of difference between the GBM and control group in this study suggests that the
HVEM pathway may be unimportant in immuno-evasion by the tumour. Certainly it would
suggest that it may not be involved in the systemic immunosuppression seen in GBM
patients. There is however plausible reasons that could mask a potential active role in
immunosuppression in GBM patients.
In the early phase of acute infection BTLA is expressed at high levels in CMV specific T-
Cells with associated inhibition of function [192]. As the infection moves towards a chronic
state the long term effector cells become largely BTLA negative [190]. In the chronic disease
setting of persistent viral infection BTLA is down regulated. However recent studies have
shown BTLA expression persisting in the T-cells of patients with melanoma [190]. In
comparing Total CD8+ T-cells we are mainly comparing a mixed group of naive and effector
T-cells rather than a subgroup which has been activated by tumour or disease and therefore
we would expect more uniform high expression of BTLA and any changes in the gene
expression may require cell activation before becoming apparent in surface expression.
These results do not support the suggestion that BTLA/LIGHT/HVEM pathway is important
in immune suppression by GBM. However to fully explore these possibilities discussed
above these studies could be repeated using CD8+ cells isolated from tumour samples.
Results from samples taken from a single patient post adoptive T-cell transfer showed that
expression of a number of markers such as PD-1, TIM-3, CTLA, T-bet, EOMES and LEF-1
were different in tumour infiltrating cells compared to PBMC’s taken at similar time points
78
[193]. This suggests differences may exist between peripheral CD8+ T-cells in these patients
and the TIL’s within GBM tissue.
The immunohistochemical staining of GBM tissue showed expression in a third of tumours
studied. Immunohistochemistry as a technique has many limitations which may result in false
or artefact signal. In this case the presence of intra-nuclear staining was seen and though to be
artefact as the more expected staining pattern would be cytoplasmic and on the cell
membrane. Furthermore the small number of samples used in this study precludes any
meaningful analysis of the true extent of expression in GBM tissue and relation of this to
clinical effects such as impact of expression of HVEM on survival. While expression of
HVEM was only seen in a third of samples studied nonetheless the presence of HVEM
expression in some GBM suggests it may play a role in the biology of the tumour.
Key Points:
• In CMV Specific CD8+ T-Cells there is no difference between GBM patients and
Controls in expression of BTLA, LIGHT and HVEM.
• In the total CD8+ T-cells of GBM patients there is significantly less expression of
BTLA and HVEM in GBM patients compared to controls.
• HVEM is expressed in some GBM tissue; this supports the potential role of the
HVEM/LIGHT/BTLA pathway in creating immunosuppression in the tumour micro-
environment.
79
Chapter 5: Discussion
The aim of these studies was to further functionally characterise cytotoxic T-cell and HCMV-
specific T-cell responses in GBM patients. The hope is that this will aid in characterising the
relationship between the host immune response and the progression of the disease with the
ultimate goal being production of an effective adoptive T-cell therapy.
The presence of HCMV in GBM is controversial but is gaining acceptance. The impact of
the presence of HCMV as a prognostic factor remains unresolved. Two recent studies have
resulted in contradictory results [118, 119]. Both of these studies were based on analysis of
HCMV products in tumour tissue. The aim of the study described in chapter 2 was to assess
the effects of exposure to HCMV as measured by peripheral blood sample serology on
outcome in GBM. This test is readily available from simple serum samples and at small cost
compared to analysis of tumour tissues. The data from this study demonstrates no difference
in outcomes between those patients who have HCMV+ and HCMV– serology. These
findings are supported in the existing literature [181].
The impaired T-Cell response to malignancy is from multifactorial phenomena discussed in
the introduction. The gene array analysis in this thesis provides an opportunity to identify
potential mechanisms by which T-Cell function is impaired and define targets for further
investigation. Measurement of the gene expression does not however give any information on
the functional significance of these gene products. Also mRNA levels may not correlate to
protein expression levels once posttranslational modifications have taken place. Nevertheless
there were a large number of potential gene targets identified by this study all of which could
potentially be examined in relation to impaired T-Cell function in the setting of GBM. Given
constraints of time on this thesis the BTLA and LIGHT were chosen from the candidate
genes identified for further analysis. These genes were chosen as these two molecules act on
the same ligand and were shown to be altered in the total CD8+ population which in this
study had the largest number of subjects’ thus more robust data.
Analysis of BTLA/LIGHT/HVEM expression with FACS analysis and
immunohistochemistry did not clearly define an important role for this pathway in GBM
patients. In the CMV specific T-cell population there was no significant difference in
expression of these molecules. In the total CD8+ T-cell population expression of LIGHT was
unchanged between GBM and the control subjects. There was increased expression of BTLA
80
and HVEM in controls compared to GBM which was the opposite of the expected result from
the gene expression analysis. This difference was seen in both % counts and MFI for BTLA
but only in the MFI analysis for HVEM. It may be that activation of the T-cells may be
required to see the full changes in gene expression on the surface expression of the molecule
for which it codes due to post-translational modifications. If this were the case then
stimulation of the cell, via an activation assay, may be required to see the changes in gene
expression reflected in the surface expression in the T-Cell population.
It may simply be that BTLA/LIGHT is important in immunosuppression in GBM but is only
seen at the tumour microenvironment level. To evaluate if this is the case with
BTLA/LIGHT/HVEM the current experiment could be repeated examining expression of
these molecules in TILs rather than T-cells from PBMC’s. The difficulty with this approach
is that TIL’s have to be harvested from tumour samples which remains an onerous procedure
and requires immediate processing of tumour samples. Alternative strategies may involve
identification of a tumour specific marker and FACS cell sorting to isolate a tumour specific
population from the peripheral circulation.
While the results of the FACS analysis suggest a limited role for the BTLA/LIGHT/HVEM
pathway in GBM immune evasion the immunohistochemistry did show expression of HVEM
in tumour tissue which has not previously been reported. HVEM has been shown to be
present in melanoma and proposed to act via a “trans” interaction with BTLA to inhibit T-
cell function. The presence of HVEM in this study means that this interaction is at least
possible in GBM. Given limitations of this study discussed in the chapter these findings need
to be replicated with further immunohistochemistry, tumour cell cultures or flow cytometry
of studies of tumour lysates.
BTLA expression along with many other molecules such as PD-1 is believed to be a maker of
the “exhaustion” in T-cells [71]. BTLA is expressed on naive T-cells but is down regulated
by activation to effector cells. Interest in its role in malignancy has been sparked by the
discovery of persistent expression of BTLA in tumour specific T-cells in melanoma. Recent
studies have now provided conflicting evidence about the role of BTLA. In a mouse model of
adaptive cell transfer, blockade of the HVEM pathway resulted in tumour size reduction
[194]. This was presumed by blocking of the inhibitory signal from BTLA and allowing
HVEM to solely be stimulatory in response to LIGHT. However, in a human melanoma
adoptive cell therapy trial, BTLA expression was seen to be correlated with improved
81
response [195]. In fact there is evidence that BTLA may provide bidirectional signalling,
playing different roles at different stages of T-cell differentiation.
There are now suggestions that BTLA along with other markers such as PD-1 and Tims-3 are
less co-inhibitory molecules rather markers of a state of T-cell activation where the cell is
activated but not differentiated, an effector memory cell [196]. In this state the expression of
these inhibitory molecules could be a mechanism to regulate the immune response and
prevent harm. This argument has been supported in regard to PD-1 by gene expression
studies in healthy donors where CD8+ T-cells with high PD-1 expression were seen to be
phenotypically different from exhausted T-cells from HIV patients [197]. BTLA itself has not
been seen to be part of the exhaustion gene signature in gene expression studies on tumour
specific T-cells [198]. Instead these cells showed a phenotype similar to effector memory
cells. Thus in the tumour setting,CD8+ T-cells would be effector memory cells activated by
antigenic stimulation and the pro-inflammatory microenvironment but unable to differentiate
due to an inhibitory cytokine milieu or lack of correct co-stimulation.
In terms of impact for adoptive T-cell therapy it has been shown that ex vivo stimulation has
been able to overcome high levels of expression of BTLA and generate anti-tumour response
once transferred to patients [199]. The suggestion discussed above that the cell expressing these
exhaustion markers are not exhausted but rather incompletely differentiated means that they
may potentially prove to be a valuable sub-population for immunotherapy. If this sub-
population of cells can be targeted in a way to promote their continued differentiation to
effector cells they may provide significant anti-tumour immunity. Indeed in TIL’s isolated from
melanoma patients BTLA+ CD8+ cells had better proliferative response to stimulation with
IL2 and persistence of T-cell clones in vivo compared to BTLA- CD8+ TIL’s [200]. The same
group also co-cultured TIL’s with HVEM+ cells and observed improved survival through
activation of the PI3K-Akt pathway. It was therefore suggested that BTLA+ T-cells provide
both an inhibitory role to limit over stimulation and also promote cell survival via PI3K-Akt
pathway. Thus this pathway provides an immune fine tuning system.
Network analysis of the gene expression data shows a common convergence in both the
CMV specific and Total T-cell population on the IL12 complex. Changes in the cytokine
milieu within GBM has been well documented [90]. These cytokines act on multiple
pathways to affect immune function with an important effect being the tipping of the immune
environment toward the TH2 response [91]. IL 12 is of interest as it is a promoter of the more
82
immunostimulatory TH1 response. Interaction with IL12 via the pathways suggested by this
data supports that this cytokine could be important in potentiating tumour growth in GBM.
Therefore targeting IL12 becomes an attractive immunotherapy strategy to promote the TH1
environment and thus potentiate the effects of adoptive cell or vaccine therapies [201].
Immunotherapy for GBM appears a promising therapy as an adjunct to an overall treatment
regime involving surgery, radiotherapy and chemotherapy [178]. The identification of
HCMV in GBM tissue whilst of uncertain clinical significance does provide and attractive
target for immunotherapy strategies such as adoptive T-cell therapy. Significant barriers to
immunotherapy remain, some of which relate to the role of co-stimulatory and inhibitory
molecules in host response to malignancy and this is currently an area of intense interest. The
role of these various molecules in GBM is uncertain. However there is considerable potential
for therapeutic manipulation of these pathways. The outcome of the experiments detailed in
this thesis raise the possibility of the BTLA/LIGHT/HVEM pathway being of some
importance in the setting of GBM. Unfortunately the exact role remains unclear. Further
studies are required to define the role of these pathways and their suitability for manipulation
as a standalone therapy or adjunct to other immunotherapeutic strategies.
83
Bibliography
1. CBTRUS, Statistical Report: Primary Brain and Central Nervous System Tumors
Diagnosed in the United States in 2004–2006. 2010, Central Brain Tumor Registry of
the United States: IL, USA.
2. Stupp, R., et al., Radiotherapy plus concomitant and adjuvant temozolomide for
glioblastoma. N Engl J Med, 2005. 352(10): p. 987-96.
3. Anton, K., J.M. Baehring, and T. Mayer, Glioblastoma multiforme: overview of
current treatment and future perspectives. Hematol Oncol Clin North Am, 2012.
26(4): p. 825-53.
4. Hou, L.C., et al., Recurrent glioblastoma multiforme: a review of natural history and
management options. Neurosurg Focus, 2006. 20(4): p. E5.
5. Stupp, R., et al., Effects of radiotherapy with concomitant and adjuvant temozolomide
versus radiotherapy alone on survival in glioblastoma in a randomised phase III
study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol, 2009. 10(5): p. 459-
66.
6. Dziurzynski, K., et al., Consensus on the role of human cytomegalovirus in
glioblastoma. Neuro Oncol, 2012. 14(3): p. 246-55.
7. June, C.H., Adoptive T cell therapy for cancer in the clinic. Journal of Clinical
Investigation, 2007. 117(6): p. 1466-1476.
8. Yee, C., Adoptive T-Cell Therapy of Cancer. Hematology/Oncology Clinics of North
America, 2006. 20(3): p. 711-733.
9. Rosenberg, S.A. and M.E. Dudley, Adoptive cell therapy for the treatment of patients
with metastatic melanoma. Current Opinion in Immunology, 2009. 21(2): p. 233-240.
10. Crough, T., et al., Ex vivo functional analysis, expansion and adoptive transfer of
cytomegalovirus-specific T-cells in patients with glioblastoma multiforme. Immunol
Cell Biol, 2012. 90(9): p. 872-80.
11. WHO Classification of Tumours of the Central Nervous System. 4th ed. WHO
Classification of Tumours of the Central Nervous System, ed. F.T. Bosman, et al.
2007, Lyon: International Agency for Research on Cancer.
84
12. Zou, P., et al., IDH1/IDH2 mutations define the prognosis and molecular profiles of
patients with gliomas: a meta-analysis. PLoS One, 2013. 8(7): p. e68782.
13. Comprehensive genomic characterization defines human glioblastoma genes and core
pathways. Nature, 2008. 455(7216): p. 1061-8.
14. Verhaak, R.G., et al., Integrated genomic analysis identifies clinically relevant
subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR,
and NF1. Cancer Cell, 2010. 17(1): p. 98-110.
15. Stiles, C.D. and D.H. Rowitch, Glioma stem cells: a midterm exam. Neuron, 2008.
58(6): p. 832-46.
16. Goffart, N., J. Kroonen, and B. Rogister, Glioblastoma-initiating cells: relationship
with neural stem cells and the micro-environment. Cancers (Basel), 2013. 5(3): p.
1049-71.
17. Singh, S.K., et al., Identification of human brain tumour initiating cells. Nature, 2004.
432(7015): p. 396-401.
18. Murat, A., et al., Stem cell-related "self-renewal" signature and high epidermal
growth factor receptor expression associated with resistance to concomitant
chemoradiotherapy in glioblastoma. J Clin Oncol, 2008. 26(18): p. 3015-24.
19. Wu, A., et al., Glioma cancer stem cells induce immunosuppressive
macrophages/microglia. Neuro Oncol, 2010. 12(11): p. 1113-25.
20. Dziurzynski, K., et al., Glioma-associated cytomegalovirus mediates subversion of
the monocyte lineage to a tumor propagating phenotype. Clin Cancer Res, 2011.
17(14): p. 4642-9.
21. Stupp, R., et al., Radiotherapy plus Concomitant and Adjuvant Temozolomide for
Glioblastoma. The New England Journal of Medicine, 2005. 352(10): p. 987-996.
22. Laws, E.R., et al., Survival following surgery and prognostic factors for recently
diagnosed malignant glioma: data from the Glioma Outcomes Project. J Neurosurg,
2003. 99(3): p. 467-73.
23. Mineo, J.F., et al., Prognosis factors of survival time in patients with glioblastoma
multiforme: a multivariate analysis of 340 patients. Acta Neurochir (Wien), 2007.
149(3): p. 245-52; discussion 252-3.
24. Kanzawa, T., et al., Role of autophagy in temozolomide-induced cytotoxicity for
malignant glioma cells. Cell Death and Differentiation, 2004. 11(4): p. 448-57.
25. Hegi, M.E., et al., MGMT gene silencing and benefit from temozolomide in
glioblastoma. N Engl J Med, 2005. 352(10): p. 997-1003.
85
26. Cohen, M.H., et al., FDA drug approval summary: bevacizumab (Avastin) as
treatment of recurrent glioblastoma multiforme. Oncologist, 2009. 14(11): p. 1131-8.
27. Gilbert, M.R., et al., A randomized trial of bevacizumab for newly diagnosed
glioblastoma. N Engl J Med, 2014. 370(8): p. 699-708.
28. Chinot, O.L., et al., Bevacizumab plus radiotherapy-temozolomide for newly
diagnosed glioblastoma. N Engl J Med, 2014. 370(8): p. 709-22.
29. Bock, H.C., et al., First-line treatment of malignant glioma with carmustine implants
followed by concomitant radiochemotherapy: a multicenter experience. Neurosurg
Rev, 2010. 33(4): p. 441-9.
30. Michaelis, M., et al., Oncomodulation by human cytomegalovirus: novel clinical
findings open new roads. Medical Microbiology and Immunology, 2011. 200(1): p. 1-
5.
31. Gandhi, M.K. and R. Khanna, Human cytomegalovirus: clinical aspects, immune
regulation, and emerging treatments. The Lancet Infectious Diseases, 2004. 4(12): p.
725-738.
32. Mocarski ES, S.T., Pass R, Cytomegaloviruses, in Fields' virology, P.M.H. David
Mahan Knipe, Editor. 2007, Wolters Kluwer Health/Lippincott Williams & Wilkins.
33. Varani, S. and M.P. Landini, Cytomegalovirus-induced immunopathology and its
clinical consequences. Herpesviridae, 2011. 2(1): p. 6.
34. Sissons, J.G. and A.J. Carmichael, Clinical aspects and management of
cytomegalovirus infection. The Journal of Infection, 2002. 44(2): p. 78-83.
35. Ross, S.A. and S.B. Boppana, Congenital cytomegalovirus infection: outcome and
diagnosis. Seminars in pediatric infectious diseases, 2005. 16(1): p. 44-9.
36. Malm, G. and M.L. Engman, Congenital cytomegalovirus infections. Seminars in
fetal & neonatal medicine, 2007. 12(3): p. 154-9.
37. Boeckh, M. and A.P. Geballe, Cytomegalovirus: pathogen, paradigm, and puzzle.
The Journal of clinical investigation, 2011. 121(5): p. 1673-80.
38. Soderberg-Naucler, C., Does cytomegalovirus play a causative role in the
development of various inflammatory diseases and cancer? Journal of internal
medicine, 2006. 259(3): p. 219-46.
39. Chen, D.H., et al., Three-dimensional visualization of tegument/capsid interactions in
the intact human cytomegalovirus. Virology, 1999. 260(1): p. 10-6.
40. Reddehase, M.J., Cytomegaloviruses: Molecular Biology and Immunology, ed. M.J.
Reddehase. 2006, Mainz, Germany: Caister Academic Press.
86
41. Poland, S.D., et al., Cytomegalovirus in the brain: in vitro infection of human brain-
derived cells. The Journal of infectious diseases, 1990. 162(6): p. 1252-62.
42. Chatterjee, A., R. Livingston, and C.J. Harrison, Cytomegaloviruses, in eLS. 2011,
John Wiley & Sons, Ltd.
43. Stinski, M.F., Sequence of protein synthesis in cells infected by human
cytomegalovirus: early and late virus-induced polypeptides. Journal of virology,
1978. 26(3): p. 686-701.
44. McVoy, M.A. and S.P. Adler, Human cytomegalovirus DNA replicates after early
circularization by concatemer formation, and inversion occurs within the concatemer.
Journal of virology, 1994. 68(2): p. 1040-51.
45. Sinclair, J., Latency and reactivation of human cytomegalovirus. Journal of General
Virology, 2006. 87(7): p. 1763-1779.
46. Nelson, M.A.J.a.J.A., Molecular basis of persistence and latency, in Human
Herpesviruses: Biology, Therapy, and Immunoprophylaxis, C.-F.G. Arvin A,
Mocarski E, et al., Editor. 2007, Cambridge University Press: Cambridge.
47. Sinclair, J. and P. Sissons, Latency and reactivation of human cytomegalovirus. J Gen
Virol, 2006. 87(Pt 7): p. 1763-79.
48. Goodrum, F., K. Caviness, and P. Zagallo, Human Cytomegalovirus Persistence.
Cellular microbiology, 2012. 14(5): p.644-655.
49. Cheung, A.K.L., et al., Viral gene expression during the establishment of human
cytomegalovirus latent infection in myeloid progenitor cells. Blood, 2006. 108(12): p.
3691-3699.
50. Meier, J.L., and M. F. Stinski, Major immediate-early enhancer and its gene
products, in Cytomegaloviruses: molecular biology and immunology, M.J.
Reddehase, Editor. 2006, Caister Academic Press: Norfolk, United Kingdom. p. p.
151-166.
51. Bowles, A.P., Jr. and E. Perkins, Long-term remission of malignant brain tumors
after intracranial infection: a report of four cases. Neurosurgery, 1999. 44(3): p. 636-
42; discussion 642-3.
52. Hussain, S.F., et al., The role of human glioma-infiltrating microglia/macrophages in
mediating antitumor immune responses. Neuro Oncol, 2006. 8(3): p. 261-79.
53. Carson, M.J., et al., CNS immune privilege: hiding in plain sight. Immunol Rev, 2006.
213(1): p. 48-65.
87
54. Hickey, W.F., B.L. Hsu, and H. Kimura, T-lymphocyte entry into the central nervous
system. J Neurosci Res, 1991. 28(2): p. 254-60.
55. Hickey, W.F. and H. Kimura, Perivascular microglial cells of the CNS are bone
marrow-derived and present antigen in vivo. Science, 1988. 239(4837): p. 290-2.
56. Brooks, W.H., et al., Depressed cell-mediated immunity in patients with primary
intracranial tumors: characterisation of a humoral immunosuppressive factor. The
Journal of Experimental Medicine, 1972. 136(6): p. 1631-1647.
57. Mahaley, M.S., et al., Immunobiology of primary intracranial tumors. Part 1: studies
of the cellular and humoral general immune competence of brain-tumor patients.
Journal of Neurosurgery, 1977. 46(4): p. 467-476.
58. Koev, I., Immune Response in Malignant Glioma. Journal of IMAB - Annual
Proceeding (Scientific Papers), 2010. 16, book 3(2010): p. 17-19.
59. Dix, A.R., et al., Immune defects observed in patients with primary malignant brain
tumors. Journal of Neuroimmunology, 1999. 100(1-2): p. 216-232.
60. Ooi, Y.C., et al., The role of regulatory T-cells in glioma immunology. Clin Neurol
Neurosurg, 2014. 119: p. 125-32.
61. Reardon, D.A., et al., An update on vaccine therapy and other immunotherapeutic
approaches for glioblastoma. Expert Rev Vaccines, 2013. 12(6): p. 597-615.
62. Wei, J., et al., Hypoxia potentiates glioma-mediated immunosuppression. PLoS One,
2011. 6(1): p. e16195.
63. Zhang, J.G., et al., Antigenic profiling of glioma cells to generate allogeneic vaccines
or dendritic cell-based therapeutics. Clin Cancer Res, 2007. 13(2 Pt 1): p. 566-75.
64. Gan, H.K., A.H. Kaye, and R.B. Luwor, The EGFRvIII variant in glioblastoma
multiforme. Journal of Clinical Neuroscience, 2009. 16(6): p. 748-754.
65. Wikstrand, C., et al., The class III variant of the epidermal growth factor receptor
(EGFRvIII), characterization and utilization as animmunotherapeutic targe. Journal
of Neurovirology, 1998. 4(2): p. 148-158.
66. Choi, B.D., et al., Systemic administration of a bispecific antibody targeting
EGFRvIII successfully treats intracerebral glioma. Proc Natl Acad Sci U S A, 2013.
110(1): p. 270-5.
67. Schwartz, R., T cell anergy. Annual Review of Immunology, 2003. 21: p. 305-334.
68. Gimmi, C., et al., Human T-cell clonal anergy is induced by antigen presentation in
the absence of B7 costimulation. Proc Natl Acad Sci U S A, 1993. 90: p. 6586-6590.
88
69. Learn, C.A., et al., Profiling of CD4+, CD8+, and CD4+CD25+CD45RO+FoxP3+
T cells in patients with malignant glioma reveals differential expression of the
immunologic transcriptome compared with T cells from healthy volunteers. Clin
Cancer Res, 2006. 12(24): p. 7306-15.
70. Crespo, R., et al., Transcriptional and posttranscriptional inhibition of HMGCR and
PC biosynthesis by geraniol in 2 Hep-G2 cell proliferation linked pathways. Biochem
Cell Biol, 2013. 91(3): p. 131-9.
71. Wherry, E.J., T cell exhaustion. Nat Immunol, 2011. 12(6): p. 492-9.
72. Brahmer, J.R., et al., Safety and activity of anti-PD-L1 antibody in patients with
advanced cancer. N Engl J Med, 2012. 366(26): p. 2455-65.
73. Topalian, S.L., et al., Safety, activity, and immune correlates of anti-PD-1 antibody in
cancer. N Engl J Med, 2012. 366(26): p. 2443-54.
74. Avril, T., et al., Distinct effects of human glioblastoma immunoregulatory molecules
programmed cell death ligand-1 (PDL-1) and indoleamine 2,3-dioxygenase (IDO) on
tumour-specific T cell functions. J Neuroimmunol, 2010. 225(1-2): p. 22-33.
75. Wainwright DA, C.A., Dey M, Balyasnikova IV, Kim C, Tobias AL, Cheng Y, Kim
J, Zhang L, Qiao J, Han Y, Lesniak MS., Durable therapeutic efficacy utilizing
combinatorial blockade against IDO, CTLA-4 and PD-L1 in mice with brain tumors.
Clin Cancer Res, 2014. Epub ahead of print.
76. Muranski, P., et al., Th17 cells are long lived and retain a stem cell-like molecular
signature. Immunity, 2011. 35(6): p. 972-85.
77. Kryczek, I., et al., Human TH17 cells are long-lived effector memory cells. Sci Transl
Med, 2011. 3(104): p. 104ra100.
78. Driessens, G., J. Kline, and T.F. Gajewski, Costimulatory and coinhibitory receptors
in anti-tumor immunity. Immunol Rev, 2009. 229(1): p. 126-44.
79. Brooks, W.H., et al., Relationship of lymphocyte invasion and survival of brain tumor
patients. Annals of Neurology, 1978. 4(3): p. 219-224.
80. Black, K., et al., Inflammatory leukocytes associated with increased
immunosuppression by glioblastoma. Journal of Neurosurgery, 1992. 77(1): p. 120-
126.
81. da Fonseca, A.C. and B. Badie, Microglia and macrophages in malignant gliomas:
recent discoveries and implications for promising therapies. Clin Dev Immunol,
2013. 2013: ID. 264124.
89
82. Mantovani, A., et al., The origin and function of tumor-associated macrophages.
Immunology Today, 1992. 13(7): p. 265-270.
83. Platten, M., et al., Monocyte chemoattractant protein-1 increases microglia
infiltration and aggressiveness of gliomas. Annals of Neurology, 2003. 54(3): p. 388-
392.
84. El Andaloussi, A. and M.S. Lesniak, An increase in CD4+CD25+FOXP3+
regulatory T cells in tumor-infiltrating lymphocytes of human glioblastoma
multiforme. Neuro-Oncology, 2006. 8(3): p. 234-43.
85. Sonabend, A.M., C.E. Rolle, and M.S. Lesniak, The Role of Regulatory T Cells in
Malignant Glioma. Anticancer Research, 2008. 28(28): p. 1143-1150.
86. Heimberger, A.B., et al., Incidence and prognostic impact of FoxP3+ regulatory T
cells in human gliomas. Clin Cancer Res, 2008. 14(16): p. 5166-72.
87. Fecci, P., D. Mitchell, and J. Whitesides, Increased regulatory T-cell fraction amidst
a diminished CD4 compartment explains cellular immune defects in patients with
malignant glioma. Cancer Research, 2006. 66(6): p. 3294-3302.
88. Klebanoff, C., et al., Sinks, suppressors and antigen presenters: how lymphodepletion
enhances T cell-mediated tumor immunotherapy. Trends in Immunology, 2005. 26(2):
p. 111-117.
89. Jordan, J.T., et al., Preferential migration of regulatory T cells mediated by glioma-
secreted chemokines can be blocked with chemotherapy. Cancer Immunol
Immunother, 2008. 57(1): p. 123-31.
90. Skog, J., Glioma-specific antigens for immune tumor therapy Expert Review of
Vaccines, 2006. 5(6): p. 793-802.
91. Perrin, G., et al., Astrocytoma infiltrating lymphocytes include major T cell clonal
expansions confined to the CD8 subset. International Immunology, 1999. 11(8): p.
1337-1350.
92. Huettner, C., W. Paulus, and W. Roggendorf, Messenger RNA expression of the
immunosuppressive cytokine IL-10 in human gliomas. American Journal of
Pathology, 1995. 146(2): p. 317-322.
93. Schlingensiepen, K.H., et al., Targeted tumor therapy with the TGF-beta 2 antisense
compound AP 12009. Cytokine Growth Factor Rev, 2006. 17(1-2): p. 129-39.
94. Hume, A.J., et al., Phosphorylation of retinoblastoma protein by viral protein with
cyclin-dependent kinase function. Science, 2008. 320(5877): p. 797-9.
90
95. Cheema, T.A., et al., Multifaceted oncolytic virus therapy for glioblastoma in an
immunocompetent cancer stem cell model. Proc Natl Acad Sci U S A, 2013. 110(29):
p. 12006-11.
96. Mizoguchi, M., et al., Activation of STAT3, MAPK, and AKT in malignant astrocytic
gliomas: correlation with EGFR status, tumor grade, and survival. J Neuropathol Exp
Neurol, 2006. 65(12): p. 1181-8.
97. Kortylewski, M., et al., Inhibiting Stat3 signaling in the hematopoietic system elicits
multicomponent antitumor immunity. Nat Med, 2005. 11(12): p. 1314-21.
98. Hussain, S.F., et al., A novel small molecule inhibitor of signal transducers and
activators of transcription 3 reverses immune tolerance in malignant glioma patients.
Cancer Res, 2007. 67(20): p. 9630-6.
99. Pagano, J., et al., Infectious agents and cancer: criteria for a causal relation.
Seminars in Cancer Biology, 2004. 14(6): p. 453-471.
100. Wrensch, M., et al., Familial and personal medical history of cancer and nervous
system conditions among adults with glioma and controls. Am J Epidemiol, 1997.
145(7): p. 581-93.
101. Carbone, M., P. Rizzo, and H.I. Pass, Simian virus 40, poliovaccines and human
tumors: a review of recent developments. Oncogene, 1997. 15(16): p. 1877-88.
102. Wrensch, M., et al., Does prior infection with varicella-zoster virus influence risk of
adult glioma? Am J Epidemiol, 1997. 145(7): p. 594-7.
103. Wrensch, M., et al., Prevalence of antibodies to four herpesviruses among adults with
glioma and controls. Am J Epidemiol, 2001. 154(2): p. 161-5.
104. Colugnati, F.A., et al., Incidence of cytomegalovirus infection among the general
population and pregnant women in the United States. BMC Infect Dis, 2007. 7: p. 71.
105. Cobbs, C.S., et al., Human cytomegalovirus infection and expression in human
malignant glioma. Cancer Res, 2002. 62(12): p. 3347-50.
106. Lau, S.K., et al., Lack of association of cytomegalovirus with human brain tumors.
Mod Pathol, 2005. 18(6): p. 838-43.
107. Mitchell, D.A., et al., Sensitive detection of human cytomegalovirus in tumors and
peripheral blood of patients diagnosed with glioblastoma. Neuro Oncol, 2008. 10(1):
p. 10-8.
108. Poltermann, S., et al., Lack of association of herpesviruses with brain tumors. J
Neurovirol, 2006. 12(2): p. 90-9.
91
109. Scheurer, M.E., et al., Detection of human cytomegalovirus in different histological
types of gliomas. Acta Neuropathol, 2008. 116(1): p. 79-86.
110. Prins, R.M., T.F. Cloughesy, and L.M. Liau, Cytomegalovirus immunity after
vaccination with autologous glioblastoma lysate. N Engl J Med, 2008. 359(5): p. 539-
41.
111. Slinger, E., et al., HCMV-encoded chemokine receptor US28 mediates proliferative
signaling through the IL-6-STAT3 axis. Sci Signal, 2010. 3(133): p. ra58.
112. Straat, K., et al., Activation of telomerase by human cytomegalovirus. J Natl Cancer
Inst, 2009. 101(7): p. 488-97.
113. Cheng, J., et al., Cytomegalovirus infection causes an increase of arterial blood
pressure. PLoS Pathog, 2009. 5(5): p. e1000427.
114. Cinatl, J. et al., Modulatory effects of human cytomegalovirus infection on malignant
properties of cancer cells. Intervirology, 1996. 39(4): p. 259-69.
115. Michaelis, M., H.W. Doerr, and J. Cinatl, Jr., Oncomodulation by human
cytomegalovirus: evidence becomes stronger. Med Microbiol Immunol, 2009. 198(2):
p. 79-81.
116. Hanahan, D. and R.A. Weinberg, Hallmarks of cancer: the next generation. Cell,
2011. 144(5): p. 646-74.
117. Price, R.L., et al., Abstract 4815: Cytomegalovirus enhances glioblastoma via PDGF-
B/STAT3 pathway activation Cancer Research, 2012. 72(8 (Suppl 1)).
118. Rahbar, A., et al., Human cytomegalovirus infection levels in glioblastoma multiforme
are of prognostic value for survival. J Clin Virol, 2013. 57(1): p.36-42.
119. Ding, D., et al., Does the existence of HCMV components predict poor prognosis in
glioma? J Neurooncol, 2014. 116(3): p. 515-522.
120. Kalejta, R.F. and T. Shenk, Proteasome-dependent, ubiquitin-independent
degradation of the Rb family of tumor suppressors by the human cytomegalovirus
pp71 protein. Proc Natl Acad Sci U S A, 2003. 100(6): p. 3263-8.
121. Cobbs, C.S., et al., Modulation of oncogenic phenotype in human glioma cells by
cytomegalovirus IE1-mediated mitogenicity. Cancer Res, 2008. 68(3): p. 724-30.
122. Cobbs, C.S., et al., Human cytomegalovirus induces cellular tyrosine kinase signaling
and promotes glioma cell invasiveness. J Neurooncol, 2007. 85(3): p. 271-80.
123. Demuth, T. and M. Berens, Molecular mechanisms of glioma cell migration and
invasion. Journal of Neuro-Oncology, 2004. 70(2): p. 217-228.
92
124. Blaheta, R., et al., Human cytomegalovirus in- fection of tumor cells downregulates
NCAM (CD56): a novel mechanism for virus-induced tumor invasiveness. Neoplasia,
2004. 6(4): p. 323-331.
125. Zhu, H., Y. Shen, and T. Shenk, Human cytomegalovirus IE1 and IE2 proteins block
apoptosis. Journal of Virology, 1995. 69(12): p. 7960-7970.
126. Cinatl, J.J., et al., Persistent human cytomegalovirus infection induces drug resistance
and alteration of programmed cell death in human neuroblastoma cells. Cancer
Research, 1998. 58: p. 367-372.
127. Cinatl, J., Jr., et al., Activation of telomerase in glioma cells by human
cytomegalovirus: another piece of the puzzle. J Natl Cancer Inst, 2009. 101(7): p.
441-3.
128. Maussang, D., et al., Human cytomegalovirus-encoded chemokine receptor US28
promotes tumorigenesis. Proc Natl Acad Sci U S A, 2006. 103(35): p. 13068-73.
129. Brantley, E.C. and E.N. Benveniste, Signal transducer and activator of transcription-
3: a molecular hub for signaling pathways in gliomas. Mol Cancer Res, 2008. 6(5): p.
675-84.
130. Abou-Ghazal, M., et al., The incidence, correlation with tumor-infiltrating
inflammation, and prognosis of phosphorylated STAT3 expression in human gliomas.
Clin Cancer Res, 2008. 14(24): p. 8228-35.
131. Doucette, T.A., et al., STAT3 Enhances Tumour Formation and Grade in a mouse
model of Glioma. Neuro Oncol, 2010. 12(suppl 4): p. 130-131.
132. Gaspar, M. and T. Shenk, Human cytomegalovirus inhibits a DNA damage response
by mislocalizing checkpoint proteins. Proc Natl Acad Sci U S A, 2006. 103(8): p.
2821-6.
133. Luo, M.H., et al., Human cytomegalovirus infection causes premature and abnormal
differentiation of human neural progenitor cells. J Virol, 2010. 84(7): p. 3528-41.
134. Fortunato, E.A., M.L. Dell'Aquila, and D.H. Spector, Specific chromosome 1 breaks
induced by human cytomegalovirus. Proc Natl Acad Sci U S A, 2000. 97(2): p. 853-8.
135. Li, Y.S., et al., Cytogenetic evidence that a tumor suppressor gene in the long arm of
chromosome 1 contributes to glioma growth. Cancer Genet Cytogenet, 1995. 84(1): p.
46-50.
136. Liu, B.L., et al., Quantitative detection of multiple gene promoter hypermethylation in
tumor tissue, serum, and cerebrospinal fluid predicts prognosis of malignant gliomas.
Neuro Oncol, 2010. 12(6): p. 540-8.
93
137. Mendelson, M., et al., Detection of endogenous human cytomegalovirus in CD34+
bone marrow progenitors. J Gen Virol, 1996. 77 (12): p. 3099-102.
138. Chan, G., et al., NF-kappaB and phosphatidylinositol 3-kinase activity mediates the
HCMV-induced atypical M1/M2 polarization of monocytes. Virus Res, 2009. 144(1-
2): p. 329-33.
139. Grivennikov, S.I., F.R. Greten, and M. Karin, Immunity, inflammation, and cancer.
Cell, 2010. 140(6): p. 883-99.
140. Reinke, P., et al., Mechanisms of human cytomegalovirus (HCMV) (re)activation and
its impact on organ transplant patients. Transpl Infect Dis, 1999. 1(3): p. 157-64.
141. Tait, M.J., et al., Survival of patients with glioblastoma multiforme has not improved
between 1993 and 2004: analysis of 625 cases. Br J Neurosurg, 2007. 21(5): p. 496-
500.
142. Walker, D.G., Immunology of brain tumors and implications for immunptherapy, in
Brain Tumors, A.H. Kaye and E.R. Laws, Editors. 2012, Elsevier. p. 125-137.
143. Ge, L., et al., Immunotherapy of Brain Cancers: The Past, the Present, and Future
Directions. Clinical and Developmental Immunology, 2010. 2010: p. 1-19.
144. Yang, I., et al., CD8+ T-cell infiltrate in newly diagnosed glioblastoma is associated
with long-term survival. Journal of Clinical Neuroscience, 2010. 17(11): p. 1381-
1385.
145. Tsao, H., M. Atkins, and A. Sober, Management of Cutaneous Melanoma. The New
England Journal of Medicine, 2004. 351: p. 998-1012.
146. Kantoff, P., et al., Sipuleucel-T Immunotherapy for Castration-Resistant Prostate
Cancer. The New England Journal of Medicine, 2010. 363: p. 411-422.
147. Heimberger, A.B., et al., Epidermal growth factor receptor VIII peptide vaccination is
efficacious against established intracerebral tumors. Clin Cancer Res, 2003. 9(11): p.
4247-54.
148. Sampson, J.H., et al., An epidermal growth factor receptor variant III-targeted
vaccine is safe and immunogenic in patients with glioblastoma multiforme. Mol
Cancer Ther, 2009. 8(10): p. 2773-9.
149. Sampson, J.H., et al., Immunologic escape after prolonged progression-free survival
with epidermal growth factor receptor variant III peptide vaccination in patients with
newly diagnosed glioblastoma. J Clin Oncol, 2010. 28(31): p. 4722-9.
150. www.clinicaltrials.gov.
94
151. Zhou, J., et al., Persistence of multiple tumor-specific T cell clones is associated with
complete tumor regression in a melanoma patient receiving adoptive cell transfer
therapy. Journal of Immunotherapy, 2005. 28(1): p. 53-62.
152. Robbins, P., et al., Cutting edge: persistence of transferred lymphocyte clonotypes
correlates with cancer regression in patients receiving cell transfer therapy. Journal
of Immunology, 2004. 173(12): p. 7125-7130.
153. Katakura, R., et al., Therapeutic Efficacy of Adoptive Cell Transfer on Survival of
Patients with Glioblastoma Multiforme: Case Reports. Case Reports in Oncology,
2010. 3(2): p. 110-124.
154. Ishikawa, E., et al., Adoptive cell transfer therapy for malignant gliomas. Adv Exp
Med Biol, 2012. 746(9): p. 109-20.
155. Schumacher, T.N. and N.P. Restifo, Adoptive T cell therapy of cancer. Current
Opinion in Immunology, 2009. 21(2): p. 187-9.
156. Turtle, C.J., et al., Engineered T cells for anti-cancer therapy. Current Opinion in
Immunology, 2012. 24(5): p. 633-9.
157. Krebs, S., et al., Genetically modified T cells to target glioblastoma. Front Oncol,
2013. 3: p. 322.
158. Bhardwaj, N., Harnessing the immune system to treat cancer. Journal of Clinical
Investigation, 2007. 117(5): p. 1130-1136.
159. Sallusto, F., J. Geginet, and A. Lanzavecchia, Central memory and effector memory T
cell sub-sets: function, generation, and maintenance. Annual Review of Immunology,
2004. 22: p. 745-763.
160. Dudley, M.E., J.R. Wunderlich, and P.F. Robbins, Cancer regression and
autoimmunity in patients after clonal repopulation with antitumor lymphocytes.
Science, 2002. 298(5594): p. 850-854.
161. Wrzesinski, C., et al., Hematopoitetic stem cells promote the expansion and function
of adoptively transferred antitumor CD8+ T cells. The Journal of Clinical
Investigation, 2007. 117(2): p. 492-501.
162. Parviz, M., et al., Successful adoptive immunotherapy with vaccine-sensitized T cells,
despite no effect with vaccination alone in a weakly immunogenic tumor model.
Cancer Immunology, Immunotherapy, 2003. 52(12): p. 739-750.
163. Teshima, T., et al., Donor leukocyte infusion from immunized donors increases tumor
vaccine efficacy after allogeneic bone marrow transplantation. Cancer Research,
2002. 62: p. 796-800.
95
164. Gough, M.J. and M.R. Crittenden, Combination approaches to immunotherapy: the
radiotherapy example. Immunotherapy, 2009. 1(6): p. 1025-1037.
165. Ramakrishnan, R., S. Antonia, and D.I. Gabrilovich, Combined modality
immunotherapy and chemotherapy: a new perspective. Cancer Immunology,
Immunotherapy, 2008. 57(10): p. 1523-1529.
166. Demaria, S., N. Bhardwaj, and W.H. McBride, Combining radiotherapy and
immunotherapy: a revived partnership. International Journal of Radiation Oncology
Biology Physics, 2005. 63(3): p. 655-666.
167. Stewart, C. and C. Perez, Effect of irradiation on immune responses. Radiology, 1976.
118(1): p. 201-210.
168. Rosen, E., et al., The molecular and cellular basis of radiosensitivity: important
implications for understanding how normal tissues and tumors respond to therapeutic
radiation. Cancer Investigation, 1999. 17(1): p. 56-72.
169. Chakraborty, M., et al., External beam radiation of tumors alters phenotype of tumor
cells to render them susceptible to vaccine-mediated T-cell killing. Cancer Research,
2004. 64: p. 4328-4337.
170. Reits, E., et al., Radiation modulates the peptide repertoire, enhances MHC class I
expression, and induces successful antitumor immunotherapy. Journal of
Experimental Medicine, 2006. 203(5): p. 1259-1271.
171. Lake, R.A. and B.W.S. Robinson, Immunotherapy and chemotherapy — a practical
partnership. Nature Reviews Cancer, 2005. 5(5): p. 397-405.
172. Wheeler, C.J., Clinical Responsiveness of Glioblastoma Multiforme to Chemotherapy
after Vaccination. Clinical Cancer Research, 2004. 10(16): p. 5316-5326.
173. Mitchell, D., et al., Monoclonal antibody blockade of IL-2 receptor α during
lymphopenia selectively depletes regulatory T cells in mice and humans. Blood, 2011.
118(11): p. 3003-3012.
174. Schreiber, H., et al., Immunodominance and tumor escape. Semin Cancer Biol, 2002.
12(1): p. 25-31.
175. Sampson, J.H., et al., Greater chemotherapy-induced lymphopenia enhances tumor-
specific immune responses that eliminate EGFRvIII-expressing tumor cells in patients
with glioblastoma. Neuro Oncol, 2011. 13(3): p. 324-33.
96
176. Prins, R.M., et al., Gene expression profile correlates with T-cell infiltration and
relative survival in glioblastoma patients vaccinated with dendritic cell
immunotherapy. Clin Cancer Res, 2011. 17(6): p. 1603-15.
177. Stragliotto, G., et al., Effects of valganciclovir as an add-on therapy in patients with
cytomegalovirus-positive glioblastoma: A randomised, double-blind, hypothesis-
generating study. Int J Cancer, 2013. 133(5):1204-13.
178. Schuessler, A., et al., Autologous T cell Therapy for Cytomegalovirus as a
Consolidative Treatment for Recurrent Glioblastoma. Cancer Res, 2014.
74(13):3466-76.
179. Britt, W. and C. Alford, Cytomegalovirus, in Fields Virology. 1996. p. 2493-2523.
180. Dey, M., A.U. Ahmed, and M.S. Lesniak, Cytomegalovirus and glioma: putting the
cart before the horse. J Neurol Neurosurg Psychiatry, 2014. 86(2): 191-199.
181. Amirian, E.S., et al., Anti-human-cytomegalovirus immunoglobulin G levels in glioma
risk and prognosis. Cancer Med, 2013. 2(1): p. 57-62.
182. Hertoghs, K.M., et al., Molecular profiling of cytomegalovirus-induced human CD8+
T cell differentiation. J Clin Invest, 2010. 120(11): p. 4077-90.
183. Blackburn, S.D., et al., Coregulation of CD8+ T cell exhaustion by multiple inhibitory
receptors during chronic viral infection. Nat Immunol, 2009. 10(1): p. 29-37.
184. Paulos, C.M. and C.H. June, Putting the brakes on BTLA in T cell-mediated cancer
immunotherapy. J Clin Invest, 2010. 120(1): p. 76-80.
185. Steinberg, M.W., T.C. Cheung, and C.F. Ware, The signaling networks of the
herpesvirus entry mediator (TNFRSF14) in immune regulation. Immunol Rev, 2011.
244(1): p. 169-87.
186. Compaan, D.M., et al., Attenuating lymphocyte activity: the crystal structure of the
BTLA-HVEM complex. J Biol Chem, 2005. 280(47): p. 39553-61.
187. Pasero, C., et al., The HVEM network: new directions in targeting novel
costimulatory/co-inhibitory molecules for cancer therapy. Curr Opin Pharmacol,
2012. 12(4): p. 478-85.
188. Morel, Y., et al., Reciprocal expression of the TNF family receptor herpes virus entry
mediator and its ligand LIGHT on activated T cells: LIGHT down-regulates its own
receptor. J Immunol, 2000. 165(8): p. 4397-404.
189. www.proteinatlas.org.
190. Derre, L., et al., BTLA mediates inhibition of human tumor-specific CD8+ T cells that
can be partially reversed by vaccination. J Clin Invest, 2010. 120(1): p. 157-67.
97
191. Yu, P., et al., Priming of naive T cells inside tumors leads to eradication of
established tumors. Nat Immunol, 2004. 5(2): p. 141-9.
192. Serriari, N.E., et al., B and T lymphocyte attenuator is highly expressed on CMV-
specific T cells during infection and regulates their function. J Immunol, 2010.
185(6): p. 3140-8.
193. Schuessler, A., et al., Autologous T-cell therapy for cytomegalovirus as a
consolidative treatment for recurrent glioblastoma. Cancer Res, 2014. 74(13): p.
3466-76.
194. Lasaro, M.O., et al., Active immunotherapy combined with blockade of a coinhibitory
pathway achieves regression of large tumor masses in cancer-prone mice. Mol Ther,
2011. 19(9): p. 1727-36.
195. Radvanyi, L.G., et al., Specific lymphocyte subsets predict response to adoptive cell
therapy using expanded autologous tumor-infiltrating lymphocytes in metastatic
melanoma patients. Clin Cancer Res, 2012. 18(24): p. 6758-70.
196. Haymaker, C., et al., PD-1 and BTLA and CD8(+) T-cell "exhaustion" in cancer:
"Exercising" an alternative viewpoint. Oncoimmunology, 2012. 1(5): p. 735-738.
197. Duraiswamy, J., et al., Phenotype, function, and gene expression profiles of
programmed death-1(hi) CD8 T cells in healthy human adults. J Immunol, 2011.
186(7): p. 4200-12.
198. Baitsch, L., et al., Exhaustion of tumor-specific CD8(+) T cells in metastases from
melanoma patients. J Clin Invest, 2011. 121(6): p. 2350-60.
199. Wu, R., et al., Adoptive T-cell therapy using autologous tumor-infiltrating
lymphocytes for metastatic melanoma: current status and future outlook. Cancer J,
2012. 18(2): p. 160-75.
200. Haymaker, C., et al., Incovering a novel function of BTLa on tumour-infiltrating
CD8+ T cells. Journal for immunoTherapy of Cancer, 2013. 1(Suppl 1): p. 1.
201. Shimato, S., et al., Profound tumor-specific Th2 bias in patients with malignant
glioma. BMC Cancer, 2012. 12: p. 561.
98
Appendix: Genes included in Gene Expression Analysis
Functional group Gene name Assay ID
Cell Division and Metabolism BIRC5 Hs04194392_s1
CDCA7 Hs00230589_m1
CDK2 Hs01548894_m1
G0S2 Hs00274783_s1
MCM4 Hs00907398_m1
MKI67 Hs01032443_m1
SAP30 Hs01009154_g1
TYMS Hs00426586_m1
UHRF1 Hs01086727_m1
Survival and Apoptosis BAD Hs00188930_m1
BAK1 Hs00832876_g1
BAX Hs00180269_m1
BCL2L1 Hs00236329_m1
BCL2L11 Hs00708019_s1
BIK Hs00154189_m1
CASP1 Hs00354836_m1
FAS Hs00236330_m1
FASLG Hs00181225_m1
MCL1 Hs01050896_m1
TNFSF10 Hs00921974_m1
TNFSF14 Hs00542477_m1
XAF1 Hs00213882_m1
XIAP Hs00745222_s1
Effector Molecules GZMA Hs00989184_m1
GZMB Hs01554355_m1
GZMK Hs00157878_m1
PROK2 Hs01587689_m1
SPON2 Hs00202813_m1
Transcription Factors BCL6 Hs00153368_m1
EOMES Hs00172872_m1
FOXP3 Hs01085834_m1
GATA3 Hs00231122_m1
LEF1 Hs01547250_m1
NR3C2 Hs01031809_m1
PRDM1 Hs00153357_m1
RORC Hs01076122_m1
99
SCML1 Hs00232467_m1
SOCS3 Hs02330328_s1
TBX21 Hs00203436_m1
TSC22D3 Hs00608272_m1
ZNF395 Hs00608626_m1
ZNF683 Hs00543184_m1
Migration and Adhesion ADAM8 Hs00174246_m1
CCR1 Hs00928897_s1
CCR5 Hs99999149_s1
CCR7 Hs01013469_m1
CXCR1 Hs01921207_s1
CXCR3 Hs01847760_s1
CXCR5 Hs00540548_s1
CX3CR1 Hs01922583_s1
ITGA6 Hs01041011_m1
ITGAD Hs00236178_m1
ITGAL Hs00158218_m1
ITGB1 Hs00559595_m1
ITGB2 Hs00164957_m1
SELL Hs00174151_m1
Cytokine Receptors CXCR6 Hs01890898_s1
IL2RA Hs00907779_m1
IL2RB Hs01081697_m1
IL7R Hs00902334_m1
IL17RA Hs01064648_m1
IL18R1 Hs00977691_m1
IL28RA Hs00417120_m1
Chemokines and Cytokines CCL4 Hs00605740_g1
CCL5 Hs00174575_m1
CXCL13 Hs00757930_m1
IFNG Hs00989291_m1
IL2 Hs00174114_m1
IL4 Hs00174122_m1
IL8 Hs00174103_m1
IL10 Hs00961622_m1
IL17A Hs00174383_m1
IL21 Hs00222327_m1
IL32 Hs00992441_m1
TGFB1 Hs00998133_m1
Differentiation and Exhaustion BTLA Hs00699198_m1
CD27 Hs00386811_m1
CD28 Hs01007422_m1
100
CD160 Hs00199894_m1
CD244 Hs00175568_m1
CD40LG Hs00163934_m1
CTLA4 Hs03044418_m1
HAVCR2 Hs00262170_m1
ICOS Hs00359999_m1
KLRC1 Hs00970273_g1
KLRD1 Hs00233844_m1
KLRG1 Hs00929964_m1
LAG3 Hs00158563_m1
PDCD1 Hs01550088_m1
PTGER2 Hs00168754_m1
housekeeping genes 18S Hs99999901_s1
ACTB Hs01060665_g1
B2M Hs00984230_m1