Improving Oncolytic Viral Therapy for Primary and Metastatic Tumors in the Brain

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Walter Hans Meisen

Graduate Program in Integrated Biomedical Science Program

The Ohio State University

2015

Dissertation Committee:

Balveen Kaur, PhD, Advisor

Timothy Cripe, MD, PhD

Jonathan Godbout, PhD

Susheela Tridandapani, PhD

Walter Hans Meisen

2015

Abstract

Oncolytic viruses [OVs] are an exciting cancer immunotherapy which has received considerable attention in recent years. OVs are viruses designed to specifically destroy cancer cells and promote anti-tumor immune responses. Promising phase 3 clinical trial data in melanoma patients suggests FDA approval for OV therapy may soon be a reality.

Despite successes in cancers like melanoma, the efficacy of these viruses in central nervous system [CNS] cancers has been limited. While these viruses have proved safe in early phase clinical trials, durable responses have not been achieved. The unique brain tumor microenvironment and restrictive blood brain barrier make these cancers a therapeutic challenge for researchers and clinicians. The goal of these studies was to develop novel strategies to improve OV therapy for CNS cancers.

In part one of this dissertation we examined the therapeutic efficacy of a targeted OV for breast cancer [BC] brain metastases. The 2 year survival rate of patients with this disease is less than 2%. Treatment options for BC brain metastases are limited and there is an unmet need to identify novel therapies for this disease. Brain angiogenesis inhibitor 1

[BAI1] is a G-protein coupled receptor involved in tumor angiogenesis, invasion, phagocytosis, and synaptogenesis. Here, we found BAI1 expression was significantly reduced in BC compared to normal breast tissue and higher expression was associated

ii with better patient survival. Nestin is an intermediate filament whose expression is up-

regulated in several cancers. We found higher Nestin expression significantly correlated

with BC lung and brain metastases. These findings suggested both BAI1 and Nestin could be therapeutic targets for this disease. We then demonstrated the ability of an OV,

34.5ENVE, to target and kill high nestin expressing cells and deliver therapeutic

Vstat120 [extracellular fragment of BAI1]. Finally, we demonstrated 34.5ENVE could

extend the survival of mice in two models of BC brain metastases.

In the context of OV therapy, the immune response is a double-edged sword. While it

has the potential to generate long-term anti-tumor immune responses, early innate

immune responses to viral infection reduce OV replication, tumor destruction, and

efficacy. In part two of this dissertation, we characterized the antiviral effects of

macrophages and microglia on OV therapy for glioblastoma. Glioblastoma is one of the

most common and deadly types of primary brain tumors, and patients diagnosed with

these tumors have a median survival of only 15 months. We identified

microglia/macrophage secreted tumor necrosis factor α [TNFα] as a major factor which

reduces OV replication through the induction of apoptosis in infected tumor cells. We

demonstrated that the transient inhibition of TNFα could significantly enhance OV

replication and anti-tumor efficacy in vivo. The results of these studies suggest clinical

trials of FDA approved TNFα inhibitors are warranted in combination with OV in

patients with glioblastoma.

iii

One of the challenges to creating novel therapeutics, such as OVs, for CNS cancers is developing animal models and non-invasive imaging modalities in which to evaluate them. Live animal imaging is particularly challenging in brain tumor models because the skull significantly limits options for monitoring tumor growth and treatment responses.

Bioluminescent imaging [BLI] and magnetic resonance imaging [MRI] are two non- invasive imaging modalities which can significantly enhance intracranial tumor studies.

These imaging modalities allow real time measurements of tumor volume, cancer cell viability, and therapeutic responses, but they vary in the information they can provide. In part three of this dissertation we evaluated BLI and MRI in three murine glioblastoma models. We found BLI and MRI output was significantly affected by tumor necrosis, hemorrhaging, tumor depth, extra-cranial growth, and animal positioning. In synthesizing the data from this study, we created a multi-modality imaging paradigm for analyzing changes in tumor growth and biology while reducing cost-prohibitive and time consuming MRI for preclinical brain tumor studies.

Dedication

This work is dedicated to my amazing wife Isabel, my parents Maureen and Walter, my brother Paul, my sister Tori, and all of my family. I am incredibly grateful for your love, support, and encouragement. This document work also dedicated to my daughter Isabel

Veronica. Even though you will not arrive until June, you already hold a special place in my heart.

Acknowledgments

I would like to thank my advisor, Dr. Balveen Kaur, for her incredible support over the last 5 years. In addition to being an outstanding scientist and administrator, she is an exceptional mentor who dedicates herself completely to her students. She is a hardworking and passionate researcher, and she inspires us to be the same. I am

incredibly grateful for her guidance and friendship.

I would also like to recognize my thesis committee members- Drs. Cripe, Godbout,

Kwon, and Tridandapani for their valuable suggestions and support. Thank you for

collaborating with me on projects, providing me with feedback on experiments, and for

inviting me into your labs to teach me new techniques. Your help and guidance went well

beyond committee meetings and I am very grateful.

I was very fortunate to collaborate with some incredible scientists at OSU. I would like to

acknowledge Katie Thies, Dr. Haritha Mathsyaraja, Dr. Michael Ostrowski, Dr. Steven

Sizemore, Dr. Arnab Chakravarti, Dr. Norm Lehman, Dr. Bradley Elder, Peter Boyer, Dr.

Anna Bratasz, Dr. Kimerly Powell, Dr. Eric Wohleb, Dr. Jonathan Godbout, Dr. Matthew

Old, Dr. Jianfeng Han, Dr. Jianhua Yu, and Dr. Michael Caligiuri, Robin Nakkula, Dr.

Rolf Barth, Dr. Amy Lovett-Racke, Dr. Michael Racke, Dr. Mireia Guerau-de-Arellano, vi

Dr. Ira Racoma, Dr. Altaf Wani , Dr. Kuzuo Okemoto, Dr. Jaime Imatola, Dr. Feng

Geng, and Dr. Deliang Guo. I am very grateful for your help on my projects, and I was

honored to participate in your research. I would also like to acknowledge Dr. Joanna

Groden, Dr. Jeffrey Parvin, and Ms. Amy Lahmers for their great work with the

Biomedical Sciences Graduate Program.

To my lab mates: Dr. Yeshavanth Banasavadi-Siddegowda, Dr. Chelsea Bolyard, Sam

Dubin, Dr. Kamaldeen Muili, Dr. Cristina Jaime-Ramirez, Theresa Relation, Luke Russell,

Dr. Ji Young Yoo, Dr. Jun-Ge Yu, Dr. Ji Eun Son, Teja Nallanagulagari, George Kutras,

Alessandra Welker, Lindsay Boyd, Dr. Jayson Hardcastle, Dr. Amy Haseley-Thorne, Dr. Jeff

Wojton, Dr. Nina Dmitrieva, Nick Denton, The Cripe Lab, Louvenia Broadnax, John Smith,

and Lisa Denning. Thank you for all of your support and advice- you have all helped shape

me as a scientist in one way or another and I am forever thankful.

To my amazing friends and family, who include my lab mates, I would not have been able to

complete this journey without you. I will always cherish the fun times we have shared

together. Lastly, I would like to acknowledge my incredible wife, Isabel. Thank you for all of

your love and support, and for standing by me no matter how many times I almost burn down

the kitchen.

vii

Vita

2003...... Colonial Forge High School

2007...... B.S. Biology and Chemistry, The University

of Virginia

2009...... M.S. Microbiology, Georgetown University

2010 to present ...... Graduate Research Associate, Department

of Neurological Surgery, The Ohio State

University

Publications

1. Qin, Y., Meisen, W.H., Hao, Y. & Macara, I.G. Tuba, a Cdc42 GEF, is required

for polarized spindle orientation during epithelial cyst formation. The Journal of

cell biology 189, 661-669 (2010). PMID: 20479467

2. Meisen WH and Kaur B. How can we trick the immune system into overcoming

the detrimental effects of oncolytic viral therapy to treat glioblastoma? Expert Rev

Neurother 13(4), 341-3 (2013). PMID: 23545048

viii

3. Wojton J, Chu Z, Mathsyaraja H, Meisen WH, Denton N, Kwon CH, Chow LM,

Ostrowski M, Palascak M, Franco R, Bourdeau T, Thornton S, Kaur B, and Qi X.

Systemic delivery of SapC-DOPS has antiangiogenic and antitumor effects

against glioblastoma. Molecular Therapy 21(8): 1517-25 (2013). PMID:

23732993

4. Racoma IO, Meisen WH, Wang QE, Kaur B, and Wani AA. Thymoquinone

inhibits autophagy and induces cathepsin-mediated, caspase-independent cell

death in glioblastoma cells. PLOS One 8(9) (2013). PMID: 24039814

5. Okemoto K, Wagner B, Meisen WH, Haseley A, Kaur B, and Chiocca E.A.

STAT3 activation promotes oncolytic HSV1 replication in glioma cells. PLOS

One 8(8) (2013). PMID: 23936533

6. Okemoto K, Kasai K, Wagner B, Haseley A, Meisen WH, Bolyard C, Mo X,

Wehr A, Lehman A, Fernandez S, Kaur B, and Chiocca EA. DNA demethylating

agents synergize with oncolytic HSV1 against malignant gliomas. Clinical

Cancer Research, 19(21): 5952-9 (2013). PMID: 240056786.

7. Thorne AH, Meisen W.H., Russell LO, Yoo JY, Bolyard CM, Lathia JD, Rich J,

Puduvalli VK, Mao H, Yu J, Caligiuri MA, Tridandapani S, Kaur B. Role of

CCN1 in macrophage-mediated oncolytic virus clearance. Molecular Therapy.

2014. PMID: 24895995.

8. Van Brocklyn JR, Wojton J, Meisen WH, Kellough DA, Ecsedy JA, Kaur B, and

Lehman NL. Aurora-A inhibition offers a novel therapy effective against

intracranial glioblastoma. Cancer Research. 2014. PMID: 25106428.

9. Wojton J, Meisen WH, Jacob NK, Thorne AH, Hardcastle J, Denton N, Chu Z,

Dmitrieva N, Marsh R, Van Meir EG, Kwon CH, Chakravarti A, Qi X, Kaur B.

SapC-DOPS-induced lysosomal cell death synergizes with TMZ in glioblastoma.

Oncotarget. 2014. PMID: 25210852

10. Meisen WH, Dubin S, Sizemore ST, Mathsyaraja H, Thies K, Lehman N, Jaime-

Ramirez AC, Elder JB, Ostrowski M, Kaur B. Oncolytic Viral Therapy Enhances

the Survival of Mice in a Novel Model of Breast Cancer Brain Metastasis.

Molecular Cancer Therapeutics. 2014. PMID: 25376607[E-published ahead of

print].

11. Meisen WH, Wohleb E, Jaime Ramirez AC, Bolyard C, Yoo JY, Russell L,

Hardcastle J, Dubin S, Godbout J, Kaur B. The impact of macrophage and

microglia secreted TNFα on oncolytic HSV-1 therapy in the glioblastoma tumor

microenvironment. Accepted Clinical Cancer Research. (2015). PMID: 25829396

[E-published ahead of print].

Fields of Study

Major Field: Integrated Biomedical Science Program

Table of Contents

Abstract ...... ii

Dedication ...... v

Acknowledgments...... vi

Vita ...... viii

Publications ...... viii

Table of Contents ...... xii

List of Tables ...... xviii

List of Figures ...... xix

Chapter 1: Introduction ...... 1

Section 1: Breast Cancer Brain Metastases ...... 1

Epidemiology and Etiology ...... 1

Pathological Features and Classification ...... 2

Diagnosis ...... 3

Treatment ...... 4

Emerging Therapies ...... 4

Section 2: Glioblastoma ...... 9 xii

Epidemiology and Etiology ...... 9

Pathological Features and Classification ...... 10

Diagnosis ...... 11

Treatment ...... 12

Emerging Therapies ...... 13

Therapeutic Resistance ...... 15

Section 3: Oncolytic Viruses and Targeting the Brain Tumor Microenvironment ...... 17

Clinical Status of Oncolytic Viruses for GB and Breast Cancer Brain Metastases .. 17

First Generation oHSVs for GB and Breast Cancer Brain Metastases ...... 18

Second and third generation oHSVs for GB and Breast Cancer Brain Metastases ... 21

Vasculostatin expressing oHSVs for GB and Breast Cancer Brain Metastases ...... 24

Section 4: Immune Responses to Oncolytic Viral Therapy ...... 26

Chapter 2: Changes in BAI1 and Nestin Expression Are Prognostic Indicators for

Survival and Metastases in Breast Cancer and Provide Opportunities for Dual Targeted

Therapies ...... 32

Abstract ...... 32

Introduction ...... 33

Results ...... 35

BAI1 expression in reduced in breast cancer and is associated with patient survival35

xiii

Nestin expression is up-regulated in breast cancer and is associated with metastases

...... 35

Oncolytic virus is cytotoxic in multiple human BC subtypes ...... 36

Syngeneic model of breast cancer brain metastases in an HSV-1 sensitive strain .... 37

34.5ENVE is cytotoxic to murine breast cancer cells in vitro ...... 38

34.5ENVE treatment extends survival in vivo ...... 39

Discussion ...... 41

Future Directions: PARP Inhibitors in Combination with oHSV therapy for Breast

Cancer Brain Metastases ...... 42

Materials and Methods ...... 44

Cell Lines and Viruses ...... 44

Cell Viability Assays ...... 45

Virus replication assay ...... 45

Animal surgery ...... 45

Immunohistochemistry/immunofluorescence ...... 46

Image Acquisition...... 46

MRI Analysis ...... 46

Statistical Analysis ...... 47

Figures and Tables ...... 48

xiv

Chapter 3: The impact of macrophage and microglia secreted TNFα on oncolytic HSV-1 therapy in the glioblastoma tumor microenvironment ...... 62

Abstract ...... 62

Introduction ...... 64

Results ...... 67

oHSV Therapy Increases Macrophage Infiltration into the Brain Tumor

Microenvironment ...... 67

oHSV Therapy Increases Macrophage Activation in the Brain Tumor

Microenvironment ...... 68

Co-culture of oHSV-Infected Tumor Cells with Microglia or Macrophages Reduces

Viral Replication in Vitro ...... 69

Macrophage and Microglia Secreted TNFα inhibits virus replication ...... 70

Secreted TNFα induces apoptosis in oHSV infected cells ...... 72

Inhibition of macrophage or microglia secreted TNFα increases oHSV replication in

vitro ...... 72

Inhibition of TNFα increases virus replication in vivo ...... 73

Discussion ...... 75

Future Directions: Vstat120 expressing oHSVs ...... 82

Materials and Methods ...... 86

Cell Lines ...... 86 xv

Viruses and virus replication assays ...... 87

Co-culture Assays ...... 87

Western Blot ...... 87

Microglia and macrophage antibody staining ...... 88

Image Acquisition...... 88

Animal surgery ...... 89

Cell Viability Assay...... 89

Isolation of microglia and macrophages ...... 89

Murine Bone Marrow Macrophage Generation ...... 90

Statistical Analysis ...... 91

Figures and Tables ...... 92

Chapter 4: Analysis of bioluminescent and magnetic resonance imaging modalities identifies important strategies for monitoring Glioblastoma tumor growth in vivo ...... 110

Abstract ...... 110

Introduction ...... 111

Results ...... 114

Patient derived GB xenografts recapitulate the biology of human tumors ...... 114

GB Tumor Volume Increases Exponentially by MRI ...... 115

Luciferase Signal Increases with GB Tumor Growth ...... 116

xvi

The relationships between BLI and MRI vary for each tumor model ...... 117

Sources of BLI signal artifacts ...... 118

Discussion ...... 121

Methods ...... 126

Cell Lines ...... 126

Animal surgery ...... 127

Luciferase Imaging ...... 127

IHC Analyses and Image Acquisition ...... 128

MRI Studies ...... 128

Figure and Tables ...... 129

Conclusions and Future Directions ...... 137

References ...... 140

xvii

List of Tables

Table 1: Description of BC molecular subtypes and Nestin expression in a panel of human BC cells...... 50

Table 2: 34.5ENVE treatment reduces tumor volumes in Met-1 BCBM tumors by MRI.

...... 58

xviii

List of Figures

Figure 1: Reduced BAI1 is associated with breast cancer patient survival ...... 48

Figure 2: Increased Nestin expression is associated breast cancer metastases ...... 49

Figure 3: Structure of 34.5ENVE Virus ...... 51

Figure 4: Oncolytic HSV derived therapeutics target and kill multiple BC subtypes in

vitro ...... 52

Figure 5: Characterization of three murine models of BCBM for preclinical evaluation of

oncolytic HSV-1 derived therapeutics ...... 53

Figure 6: 34.5ENVE replicates in murine BC cells ...... 54

Figure 7: 34.5ENVE replicates in and kills murine BC cells in vitro ...... 55

Figure 8: 34.5ENVE infected murine BC cells secrete Vstat120 and show enhanced

cytotoxicity when targeted with a Nestin-driven OV ...... 56

Figure 9: Anti-tumor efficacy of 34.5ENVE in mice bearing established BCBM ...... 57

Figure 10: The combination of 34.5ENVE with olaparib kills MDA-MB-468 BC cells synergistically...... 59

Figure 11: The combination of 34.5ENVE with olaparib kills SKBR3 BC cells

synergistically...... 60

Figure 12: Olaparib does not affect 34.5ENVE virus replication...... 61

xix

Figure 13: OV treatment increases microglia activation and induces macrophage infiltration into the tumor microenvironment...... 92

Figure 14: Percoll gradient isolated CD11b+CD45hi cells are predominantly infiltrating macrophages...... 93

Figure 15: OV treatment increases macrophage activation...... 94

Figure 16: Microglia and macrophages reduce virus replication in tumor cells in vitro. . 95

Figure 17: Images of uninfected and infected co-cultures 12 hours post infection...... 96

Figure 18: Species specificity of murine TNFα ELISA ...... 97

Figure 19: Microglia and macrophage secreted TNFα inhibits virus [rHSVQ1] replication in vitro...... 98

Figure 20:oHSV treatment increases secreted TNFα in vivo ...... 99

Figure 21: Murine TNFα reduces viral replication in infected human glioma cells...... 100

Figure 22: TNFα increases cell proliferation and is not cytotoxic to uninfected GB cells.

...... 101

Figure 23: TNFα induces apoptosis in OV infected glioma cells...... 102

Figure 24: Inhibition of microglia/macrophage secreted TNFα increases virus [rHSVQ1] replication in vitro...... 103

Figure 25: TNFα inhibition increases virus replication and efficacy in vivo...... 104

Figure 26: RAMBO reduces monocyte infiltration in OV intracranial tumors...... 105

Figure 27: RAMBO Reduces Macrophage and Microglia Activation in treated intracranial tumors...... 106

Figure 28: RAMBO Replicates Better than a control virus [rHSVQ1] In Vivo and in cultures with microglia...... 107

Figure 29: RAMBO reduces the expression and secretion of TNFα by microglia co- cultured with infected glioma cells...... 108

Figure 30: Inhibition of TNFα Rescues Differences in Virus Replication between

RAMBO and rHSVQ1...... 109

Figure 31: GB intracranial xenograft models recapitulate the human disease...... 129

Figure 32: GB tumor volume increases with time by MRI...... 130

Figure 33: BLI signal increases with time in GB intracranial xenograft models...... 131

Figure 34: Changes in the tumor microenvironment increases BLI and MRI signal variability...... 132

Figure 35: The relationship between BLI and MRI varies between tumor models...... 133

Figure 36: BLI signal fluctuates with animal positioning...... 134

Figure 37: Tumor implantation depth, extracranial growth, and necrosis can create BLI signal artifacts...... 135

Figure 38: Multi-modality imaging strategy for intracranial GB tumor models...... 136

xxi

Chapter 1: Introduction

Section 1: Breast Cancer Brain Metastases

Epidemiology and Etiology

Breast cancer [BC] is the leading cause of cancer deaths in women worldwide. More than

1.3 million women are diagnosed and over 500,000 women die from the disease each

year [1]. The National Cancer Institute’s Surveillance, Epidemiology, and End Results

Program estimates a woman’s risk of developing breast cancer during her lifetime is 1 in

8 [2]. The median age at the time of diagnosis is 61, and a majority of new cases are

diagnosed between 50-69 years of age [2]. The incidence of BC is highest in Caucasian

women, followed by African Americans, Hispanics, and Asians [2]. A variety of lifestyle

factors increase the risk of BC, including lack of exercise, obesity, alcohol consumption,

and post-menopausal hormone replacement therapy [3]. Genetic mutations and family

history are also strongly associated with BC. Approximately, 20% of hereditary breast

cancers are caused by mutations in the DNA repair genes BRCA1/2 [4].

Approximately 15% of BC patients will develop symptomatic brain metastases [5]. Based on post-mortem analyses, however, the actual incidence of BC brain metastases is near

30% [6]. The median latency between BC diagnosis and the development of brain metastases is 2-3 years [7]. The incidence of CNS metastases is higher in certain subtypes

of BC. HER2 positive BC patients have a 35% risk of developing brain metastases, and

triple negative patients [lack Estrogen Receptor (ER), Progesterone Receptor (PR), and

human epidermal growth factor receptor 2 (HER2) expression] have a 25-46% chance of

developing CNS metastases [5]. The 2 year survival rate of patients with BC brain

metastases is less than 2% [8].

Pathological Features and Classification

A vast majority of BC brain metastases occur within the parenchyma and only 8% of

patients present with leptomeningeal metastases [8]. The number of CNS lesions can vary

significantly, and a majority of patients have more than one site of metastases. In one

study, about 26% of patients presented with a single lesion at the time of diagnosis [9].

Macroscopically, a majority of brain metastases appear as spherical masses with well

delineated borders [10]. However, a histological examination of patient biopsy samples

reveals some of these metastatic lesions do invade into the surrounding parenchyma [11].

Edema and hemorrhaging is not uncommon in CNS metastases and larger tumors are

often necrotic [10].

Breast cancers are classified into six intrinsic subtypes: luminal A, luminal B, HER2- enriched, basal-like, claudin-low and normal breast [12]. Approximately 50-60% of BCs are luminal A, and this subtype is associated with a favorable prognosis. These cancers express high levels of the ER, PR, cytokeratin 8/18, and Bcl-2. This subtype lacks HER2 expression and proliferates at a fairly slow rate [low Ki67 staining]. The Luminal B subtype comprises 10-20% of BCs, and they are more aggressive than Luminal A breast

2 cancers. Luminal B cancers are ER+ and/or PR+ with HER2 expression [or HER2- with high Ki67]. Approximately 15-25% of BCs express HER2. These cancers have a high proliferative rate and a large frequency of p53 mutations. The Basal-like subtype accounts for 15-20% of breast cancers. This subtype expresses high levels of CK5 and

CK17, P-cadherin, caveolin 1 and 2, Nestin, CD44, and EGFR. Basal-like BCs do not express the receptors ER, PR, and HER2 and are frequently referred to as triple negative

BCs. Triple negative BCs also include the rare Normal Breast and Claudin-low subtypes.

This dissertation focuses on HER2+ and triple negative BCs because 25-55% of these patients will develop brain metastases [5].

Diagnosis

The clinical manifestations of brain metastases are similar to those of primary brain tumors. Motor issues, cognitive deficits, and seizures are common in over 25% of patients [13]. Headaches are the most frequent symptom and occur in nearly 48% of patients [14]. Similar to primary tumors, BC brain metastases are typically diagnosed by

Gadolinium enhanced MRI. CT scans are utilized as a secondary option because MRI is a more sensitive imaging modality for detecting brain metastases [15]. While PET imaging has been utilized as a diagnostic tool for GB, it is not recommended for CNS metastases.

In one study, PET was successful in detecting a lesion less than 1 cm diameter in only

40% of cases [16]. Additional information is needed to determine the usefulness in detecting leptomeningeal metastases by PET. Brain biopsies are utilized when physicians are unsure of the diagnosis.

Treatment

Corticosteroids and anti-epileptics are utilized to reduce seizures, control edema, and

provide symptomatic relief [17]. Radiotherapy is the standard treatment for patients with

CNS metastases, and it has been shown in several studies to improve patient survival

compared to corticosteroids alone [18, 19]. Surgical removal of parenchymal metastases

is associated with improved patient outcomes. In a randomized study of 48 patients with

a single CNS metastasis, the combination of surgery with whole brain radiation [WBRT]

significantly improved the survival of patients compared to WBRT alone [median

survival 40 weeks vs 15 weeks] [20]. Surgery is typically performed when patients

present with one lesion usually >3-4 cm [no more than 3 lesions]. Smaller lesions are

usually treated with WBRT or stereotactic radiosurgery [19]. While the value of

removing of multiple lesions remains unclear, one study of 148 patients found that

surgical resection in combination with WBRT improved survival even in patients with

multiple metastases [21]. A variety of systemic chemotherapies such as cisplatin, TMZ,

etoposide, capecitabine, and epothilone B analogues are also commonly administered to

BC patients. Following these first-line chemotherapies, CNS objective response rates of

up to 40% were observed in some patients [22]. The effectiveness of these drugs against

brain metastases in heavily pre-treated BC patients is unclear.

Emerging Therapies

While traditional treatment regimens have shown some significant benefits, new therapies are needed in order to improve patient outcomes. CNS metastases represent a

therapeutic challenge due to the unique tumor microenvironment and the blood brain

barrier which reduces the penetration of drugs into the CNS.

Approximately 30% of breast cancers are HER2+, and uncontrolled brain metastases are

the cause of death in 36% of these patients [23, 24]. A humanized monoclonal antibody against HER2, trastuzumab, has shown great efficacy in controlling HER2+ BC. In a clinical trial with 469 patients, trastuzumab in combination with standard chemotherapy significantly improved patient survival compared to chemotherapy alone [median survival 25.1 vs 20.3 months] [25]. As a result of these studies, trastuzumab was approved in 1998 for the treatment of metastatic HER2+ BCs. Interestingly, HER2+ patients receiving trastuzumab also have a higher incidence of brain metastases. In a retrospective study, Clayton et al found 25% of patients developed symptomatic brain metastases while on trastuzumab. In this study the CNS was the first site of symptomatic disease progression in 82% of patients who developed cerebral metastases while on trastuzumab [26]. Similar studies observed 34-43% of patients developed brain metastases while on trastuzumab [26, 27]. These figures are well above the ranges observed in other BC subtypes, and the reason for the increased rate of CNS metastases is unclear. One theory suggests poor antibody penetration into the brain causes it to function as a “sanctuary site” where small micrometastases can grow uncontrolled [28]. To increase antibody penetration, several groups have tested trastuzumab in combination with radiotherapy in order to disrupt the blood brain barrier. In a small study with 6 patients, radiation was found to increase antibody penetration by 45%, but these levels

were still very low compared to serum concentrations of trastuzumab [226 ng/mL and

20,185 ng/mL, respectively] [29].

Lapatinib is a HER2 and EGFR tyrosine kinase inhibitor. The combination of lapatinib

with capecitabine was approved in 2007 as a second line treatment for HER2 metastatic

BC. In 2010, the combination of lapatinib with letrozole was approved as a first line

therapy for women with HER2+ metastatic BC [30]. Lapatinib has shown promising results in preclinical studies for BC brain metastases. In a preclinical brain metastases prevention model, Gril et al found lapatinib reduced the formation of large brain metastases by 54% [31]. The uptake of lapatinib into the brain is considerably better than trastuzumab, but still relatively poor. In a preclinical study with radiolabeled lapatinib, the concentration of drug in brain lesions was found to be 80-90% less than doses

observed in peripheral metastases [32]. In clinical trials for patients with breast cancer brain metastases, lapatinib has produced mixed results. A recent phase 2 clinical trial evaluated the efficacy of lapatinib in combination with capecitabine as primary systemic treatment for BC brain metastases. In this study, 65% of patients had at least a 50% reduction in tumor volumes [33]. Unfortunately, nearly a third of patients in the study experienced at least one severe adverse event due to toxicity. Additional studies are needed to determine if this drug combination is better than standard WBRT as a first line therapy [33-35]. Lapatinib is also being investigated as a radiosensitizer. In preclinical studies the combination of lapatinib with radiation significantly reduced tumor size in

HER2+ murine xenografts [36]. Currently, a phase II clinical trial combining lapatinib

with radiation for patients with HER2+ BC brain metastases is underway

[NCT01622868].

Approximately 25-46% of triple negative BC patients develop brain metastases. There are relatively few treatments for patients with triple negative BC, and the median survival of these patients with CNS metastases is only 6 months [5, 37]. PARP inhibitors disrupt

DNA damage repair, and these drugs are particularly effective in cancers with defects in

DNA repair pathways such as those with BRCA1/2 mutations. [38, 39]. In December

2014, the PARP inhibitor olaparib was approved for the treatment of advanced ovarian cancer in patients with BRCA1/2 mutations. An estimated, 30% of triple negative BCs possess BRCA1/2 mutations and are sensitive to PARP inhibitors in preclinical studies. A phase 1 clinical trial testing the combination of the PARP inhibitor veliparib [ABT-888] with WBRT was recently completed for patients with brain metastases and the results are pending [NCT00649207].

Anti-angiogenic agents are also being evaluated for CNS metastases. Recently, Kim et al found that BC cells that preferentially metastasize to the brain express higher levels of

VEGF and IL-8 [40]. A phase II clinical trial testing the combination of bevacizumab

with carboplatin is ongoing for patients with progressive BC brain metastases

[NCT01004172]. Preliminary results presented at the 2013 ASCO meeting found a 45%

objective response rate by RECIST criteria in these patients.

Mutations in the PI3K pathways are commonly observed in HER2+ and triple negative

BCs, and high PI3K pathway activation is frequently observed in BC brain metastases

[41]. Additionally, this pathway is commonly up-regulated in HER2+ BCs resistant to trastuzumab [42]. As a result, the PI3K and mTOR inhibitors everolimus and BKM120

are currently being explored in clinical trials for patients with BC brain metastases

[NCT01783756, NCT01305941, NCT01132664] [43].

Section 2: Glioblastoma

Epidemiology and Etiology

Glioblastoma Multiforme [GB] is one of the most common and deadly forms of malignant brain tumors. Approximately 13,000 people are diagnosed each year in the

U.S. with GB, and the median survival for these patients is only 15 months [44]. GB is

typically discovered later in life, and the median age at diagnosis is 64. The incidence of

GB is slightly higher in men than women [3.86 per 100,000 people compared with 2.39

per 100,000 people, respectively], and Caucasians are diagnosed with GB at higher rates

than people of African or Asian descent [44, 45]. No specific occupations or

environmental carcinogens have been conclusively linked with brain tumors. While

several small epidemiological studies have suggested a higher incidence of gliomas in

physicians, firefighters, farmworkers, butchers, electricians, pathologists, embalmers, cell

phone users, and in persons with diets high in N-nitroso compounds, larger studies have

not supported these hypotheses [45-47]. Interestingly, while cigarette smoking has been

associated with a variety of cancers, a recent prospective study of nearly 500,000 people

found no causal link between heavy smoking and glioma development [48]. Radiation is

the only environmental factor known to promote brain tumor development in adults and

children [49, 50]. Several familial syndromes have also been associated with a higher

incidence of CNS tumors including: Li Fraumeni syndrome, Neurofibromatosis Type 1,

Neurofibromatosis Type 2, Tuberous sclerosis, Turcot Syndrome B, and Cowden disease

[51-55]. While these genetic disorders are associated with an increased risk of tumor

development, the majority of gliomas are thought to arise from spontaneous mutations.

Pathological Features and Classification

GB is a form of malignant glioma. Based on their histological characteristics, gliomas are assigned a prognostic grade by the World Health Organization. Grade I tumors, such as pilocytic astrocytomas, are slow growing and possess relatively well circumscribed borders. Diffusively infiltrative astrocytic tumors with cellular atypia are defined as

Grade II gliomas [diffuse astrocytoma]. Grade III and IV tumors are invasive with high

mitotic activity and significant anaplasia. Grade IV tumors [Glioblastoma] are

distinguished from Grade III tumors [anaplastic astrocytoma] by the presence of

microvascular proliferation and necrosis [56]. While the histological grade is only one

component utilized to predict therapy responses and outcomes, patients diagnosed with

higher grade gliomas typically have a poorer prognosis.

Genomic analyses of GB tumors have identified 4 clinically relevant GB subtypes:

Classical, Mesenchymal, Neural, and Proneural. These tumors are classified by the

presence of genetic mutations, amplifications, and deletions [57]. Classical GB tumors

possess a high number of EGFR and EGFRvIII amplications/mutations, an absence of

p53 mutations, and deletions in CDKN2A. The Mesenchymal GB subtype contains

deletions in NF1 and PTEN. Genes in the TNF and NF-κB families are also highly

expressed in this subtype. The Neural GB subtype is characterized predominantly by the

expression of neuronal markers and is not defined by mutations in any one specific tumor

suppressor or oncogene. Proneural tumors have alterations in PDFRA and IDH1. This

group also expresses high levels of the oligodendrocyte genes NKX2-2 and OLIG2.

Recent epigenetic analysis has identified a subset of Proneural tumors with a CpG island

methylator phenotype [G-CIMP] [58]. G-CIMP positive tumors are most prevalent in low

grade gliomas and secondary GB [low grade gliomas that progress into GB]. Patients

with G-CIMP positive tumors are diagnosed at a younger age and have move favorable outcomes. Among the 4 GB subtypes, proneural patients tend to live the longest, but their tumors are also the least responsive to therapy. Patients diagnosed with Classical GB

have significantly better responses to intensive therapy than the other subtypes [57].

Diagnosis

GB patients present with a variety of symptoms including gait imbalance, cognitive

issues, depression, visual field disturbances, and incontinence [59]. Approximately 50%

of patients have severe headaches, and 20-40% of patients have seizures [14, 60]. Clinical presentations can vary based on the size and location of the tumor, and as a result GB is sometimes mistaken for other conditions. Malignant glioma is initially diagnosed by

magnetic resonance imaging [MRI]. Contrast enhancement is utilized to identify areas of

hemorrhage and necrosis. When MRI is not applicable [e.g. patient with a pacemaker],

tumors are imaged by computer tomography. Positron emission tomography [PET] is also

being evaluated for its ability to diagnose GB and monitor therapeutic responses [61].

PET imaging provides information on tumor viability and metabolic activity. In clinical

trials, the combination of PET with MRI significantly improved the identification of GB

tumor areas within the brain [62]. Following tumor imaging, GB diagnosis is confirmed

by a biopsy and histological examination by a pathologist.

Treatment

Symptomatic treatment of GB patients includes corticosteroids to reduce edema and anti-

epileptic drugs to reduce seizures. Surgical resection of the tumor is performed when

possible, but 20-30% of tumors are inoperable [63]. The degree of tumor debulking has

also been found to impact patient survival. In a phase 3 clinical trial with 242 patients,

“complete” resection of the tumor increased median survival by 40% compared to

patients with residual tumor post-surgery [as determined by contrast enhanced MRI] [64,

65]. Carmustine wafers [BCNU] are often placed in the resected tumor bed following

surgery. While a phase 3 trial demonstrated a survival benefit for patients with glioma, a

significant increase in survival was not observed for GB [66-68].

Following tumor resection, radiotherapy and Temozolomide [TMZ] are administered to patients. Radiotherapy with concomitant and adjuvant TMZ has been shown to significantly enhance the survival of GB patients [69, 70]. The cytotoxic properties of

TMZ are dependent upon its ability to alkylate DNA [71]. The enzyme O6- methylguanine–DNA Methyltransferase [MGMT] removes methylated O6 guanine and confers resistance to TMZ. In clinical trials, GB patients with a methylated MGMT promoter [low MGMT enzyme expression/activity] had significantly better survival than

patients with an unmethylated MGMT promoter [70]. In this study, MGMT promoter methylation status was the strongest predictor of patient outcomes from TMZ therapy

[72].

Despite significant efforts, GB tumors always recur. The average progression free

survival with standard therapy is 6.9 months, and tumors typically recur within 3-4 cm of

the original tumor bed [73, 74]. Tumor recurrence near the site of surgical resection highlights the infiltrative nature of GB cells. Bevacizumab, a humanized monoclonal antibody against Vascular Endothelial Growth Factor [VEGF], is the only FDA approved treatment for recurrent GB. The 2009 approval was based on two clinical trials where

Bevacizumab significantly reduced tumor size in recurrent GB [as assessed by MRI] [75,

76]. Clinical trials with Bevacizumab in combination with TMZ and radiation for newly diagnosed GB demonstrated an improvement in progression free survival but not in overall survival [77, 78]. Patients in the Bevacizumab treated group had more adverse

events, reduced neurocognitive function, and a poorer quality of life. The clinical benefits

of Bevacizumab require further investigation.

Emerging Therapies

To date, no other drugs are approved for newly diagnosed or recurrent GB, but there are a

number of ongoing Phase 3 clinical trials. A phase 2/3 trial for newly diagnosed GB

patients with TMZ and the PARP inhibitor Veliparib is ongoing [NCT02152982]. PARP inhibitors, such as Veliparib, have demonstrated strong anti-tumor efficacy against GB in

13 preclinical studies via their ability to inhibit DNA damage repair [79-82]. The NovoTTF-

100A System is another treatment being investigated in clinical trials. This system utilizes alternating electric fields to induce catastrophic, anti-mitotic effects on dividing cells [83]. Recent phase 3 clinical trial data presented at the 2014 Society for

NeuroOncology Meeting showed newly diagnosed GB patients treated with the

NovoTTF-100A system and TMZ lived significantly longer than patients treated with

TMZ alone [median overall survival of 19.6 months compared to 16.6 months, respectively]. A phase 4 clinical trial (sentry study) with the NovoTTF-100A system is currently recruiting patients with recurrent GB [NCT01756729]. The dendritic cell vaccine, DCVax®-L, is currently in a phase 3 trial for newly diagnosed GB

[NCT00045968]. Autologous dendritic cells are pulsed with autologous tumor cells obtained via surgical resection in order to generate an anti-tumor immune response.

Vaccination begins following surgery, radiotherapy, and chemotherapy. In unpublished clinical trial data, DCVax®-L treated patients had a median survival of 36.4 months [84,

85]. While promising, the sample size for this trial was small [19 patients] and Northwest

Biotherapeutics has not published any of their study data. Immunotherapies for recurrent

GB are also in late-stage clinical trials. The combination of the anti-PD1 antibody,

Nivolumab, and the anti-CTLA-4 antibody, Ipilimumab, is currently in phase 3 clinical trials for recurrent GB [NCT02017717]. A variety of other targeted chemotherapeutic drugs and immunotherapies, such as OVs, are in phase I and II trials for brain tumors [86,

87] . Some of these trials and therapies will be discussed in later sections.

Therapeutic Resistance

Despite initial responses, GB tumors possess a strong ability to resist conventional

therapies. Extensive tumor heterogeneity at the cellular and molecular levels contributes to therapeutic resistance and recurrence. A variety of genetic alterations in signaling pathways controlling proliferation, invasion, angiogenesis, and cell death all contribute to

GB resistance. Common pathways altered in GB include: the Rb pathway, the p53 pathway, mitogenic pathways, PI3K pathway, and receptor tyrosine kinase [RTK] signaling pathways [87]. While many of these signaling cascades, such as the Rb pathway, are almost universally affected in GB, the mutations that alter them can vary significantly between individuals [88]. These alterations make generating widely applicable therapies difficult. Mutations can also vary significantly within the tumor itself. A truncated form of the EGF receptor, EGFRvIII, is expressed in 25% of GB patient tumors. The truncation of this receptor confers ligand independent signaling and enhanced tumorigenicity [89, 90]. Despite the growth advantage mediated by EGFRvIII, only a small proportion of the cells in these tumors express this receptor variant [91, 92].

The heterologous nature of GB tumors means targeted therapies likely only destroy a subset of cells and are unlikely to prevent tumor recurrence. In addition to the activation

of pathways which control cell growth, GB tumors also possess a variety of mechanisms

to escape radiation and chemotherapy induced cell death. High levels of the anti-

apoptotic proteins Bcl-2, Bcl-xL, and Mcl-1 as well as decreased levels of apoptotic

proteins such as BAX are commonly observed in recurrent GB and demonstrate the

ability of these tumors to resist caspase mediated cell death [93].

The presence of GB cancer stem-like cells [GSC] within tumors also contributes to therapeutic resistance. GSCs are a small, subpopulation of cells capable of initiating tumor formation. GSCs are difficult to destroy with conventional therapies because they replicate slowly, express high levels of drug export proteins, and are not dependent on the oncoproteins targeted by newer drugs for survival [94]. These self-renewing cells persist within the tumor, and they give rise to more differentiated malignant cells needed to regenerate tumors following therapy [95, 96]. GSCs typically reside in the nutrient rich perivascular niche [97]. Here, endothelial cells and immune cells provide growth factors which maintain the GSC population. The GSC population is also thought to be plastic, and recent work has suggested non-stem cancer cells can acquire and stem-cell like phenotype [98]. A variety of GSC markers have been identified including CD133,

Integrin alpha 6, and SSEA-1, and research to target these cells is ongoing [96, 99, 100].

Section 3: Oncolytic Viruses and Targeting the Brain Tumor Microenvironment

Oncolytic viruses [OV] represent a promising therapeutic modality for the treatment of cancers. OVs are genetically engineered to replicate in and destroy tumor cells while preserving normal tissues. These attenuated viruses have proved safe in early phase clinical trials, and recent phase 3 clinical trials have demonstrated the therapeutic potential of this immunotherapy [101]. These trials have resulted in second and third generation OVs. These viruses combine the cytolytic properties of these viruses with gene therapy to target the tumor microenvironment.

Clinical Status of Oncolytic Viruses for GB and Breast Cancer Brain Metastases

China approved the use of the H101 adenovirus for the treatment of various solid tumors in 2005, but similar viruses have not yet obtained clinical approval in the US [102].

Currently, there are several promising phase III trials testing the therapeutic efficacy of

OVs in patients diagnosed with melanoma, bladder, and advanced head and neck cancers

[NCT01438112, NCT00769704, NCT01166542]. Despite this promising work, there are

currently no trials for patients with breast cancer brain metastases and only ongoing

phase I/II trials for patients with GB [NCT02197169, NCT00157703, NCT01174537,

NCT00528684, NCT01301430, NCT01582516, NCT00314925, NCT01956734, and

NCT02062827]. Several different OVs have been evaluated in clinical trials for glioma including: adenovirus [ONXY-015, DNX-2401], Herpes Simplex Virus [G207, 1716,

G47∆, M032], Measles Virus [MV-CEA], New Castle Disease virus [NDV-HUJ],

Parvovirus [H-1PV], Poliovirus [PVS-RIPO], Reovirus [Reolysin], Seneca Valley Virus

[SVV-001], and Retrovirus [non-lytic Toca-511] [103]. Oncolytic HSV-1 viruses [oHSV] are the focus of this dissertation, and the remainder of the document will focus on these viruses for the treatment of GB and breast cancer brain metastases.

First Generation oHSVs for GB and Breast Cancer Brain Metastases

HSV-1 is a double-stranded DNA virus. The virus’s large genome encodes about 84 proteins with unique modulatory functions [104]. Extensive analysis of these proteins allows neuropathogenic, immuno-regulatory, and viral replication genes to be removed or altered in order to generate OVs with strong safety and tumor specificity profiles.

Additionally, HSV-1 viral DNA does not integrate with the host’s DNA during

replication and its spread is limited via anti-herpetic drugs such as acyclovir.

Collectively, these features make HSV-1 a promising viral vector for GB therapy.

First generation oHSVs were generated via the deletion of one or both of two key viral

genes: UL-39 and RL-1. UL-39 encodes ICP6, part of an enzyme important for viral

DNA synthesis. Viruses lacking ICP6 are able to utilize mammalian ribonucleotide

reductase [RR] to generate deoxyribonucleotides required for viral DNA synthesis and

replication [105]. A deletion in the ICP6 gene enhances oHSV tumor specificity because

the virus can only replicate in rapidly dividing tumor cells expressing high levels of the

RR enzyme and not in normal, quiescent cells. Given these studies, the oHSV hrR3 was

created. This virus contains a gene-disabling lacZ insertion in ICP6. In glioma

xenografts, hrR3 significantly prolonged the survival of mice with subcutaneous tumors

[106]. While a promising first step, this singly attenuated virus displayed some in vivo toxicity [107].

The HSV-1 RL-1 gene encodes the neurovirulence protein ICP34.5. HSV-1 infection rapidly shuts off host cell translation through the activation of PKR and the phosphorylation of eif2α. ICP34.5 recruits the phosphatase PP1α that dephosphorylates eif2α and results in viral protein synthesis. ICP34.5 is also involved in a variety of other processes in infected cells such as autophagy, interferon β expression, and virus

replication via its interactions with proliferating cell nuclear antigen [108, 109]. Deletion

of the RL-1 gene reduces the replication of HSV-1 in normal cells [110]. However, in

replicating tumor cells with activated Ras and Mitogen active protein kinase pathways,

PKR is inactivated and the RL-1 deletion does not inhibit viral protein synthesis [110].

The oHSV 1716 is deleted for both copies of the RL-1 gene [111]. 1716 has been found

to be safe and have significant anti-tumor effects in pre-clinical and clinical studies [103].

In a phase I GB trial with 1716, 3/9 patients had significant responses and 5/9 had stable

disease [112]. In a second phase I study utilizing 1716, 3/12 patients were disease free for

15 months [113]. Most importantly in all three clinical trials, no patients treated with

1716 experienced significant toxicities following virus administration [114, 115].

Radiotherapy is part of the standard of care given to GB patients, and 1716 has also been

evaluated for its ability to synergize with radiation. In two preclinical studies with a virus

similar to 1716 [R3616- both copies of RL-1 deleted], radiation was found to increase

19 virus replication and anti-tumor efficacy in GB xenografts compared to virus alone [116,

117].

The G207 oHSV contains deletions in both the UL39 and RL-1 genes [118]. The G207 virus was one of the first oHSVs tested in clinical trials. In a phase I GB study with

G207, Markert et al. observed tumor shrinkage in 8/21 patients by MRI [119, 120]. G207 has also been evaluated for its ability to synergize with radiotherapy. In vitro and in GB xenografts, radiation was found to increase virus replication and tumor destruction by upregulating RR [121]. Based on these results, a phase I clinical trial for recurrent GB was conducted that combined G207 with radiotherapy. No toxicities were reported and

3/9 patients had significant radiographic responses [122]. In all 3 clinical trials conducted with G207 the virus was found to be safe in patients with no reported toxicities [119,

123]. G207 has also been found to synergize with TMZ and improve the survival of mice with intracranial GB xenografts [124]. Similar to studies with radiotherapy, the synergy of G207 with TMZ was found to be dependent upon the activation of DNA damage repair genes and the up-regulation of RR. G207 viral therapy has been investigated for breast cancer brain metastases. In mice bearing MDA-MB-435 intracranial tumors the intratumoral administration of G207 significantly increased the survival of mice [125].

Second and third generation oHSVs for GB and Breast Cancer Brain Metastases

While encouraging, initial clinical trial data with oHSVs suggested targeting tumor cells alone was not sufficient to generate long term GB remissions. Tumor cells and cancer stem-like cells make up only a small proportion of the brain tumor microenvironment.

Microglia, macrophages, T-cells, astrocytes, neurons, fibroblasts, endothelial cells, and extracellular matrix all contribute to tumor growth and therapeutic responses.

Additionally, the restrictive blood-brain-barrier and high interstitial pressure within the tumor limit the diffusion and effectiveness of traditional chemotherapies [126]. In order to generate potent, long lasting anti-tumor responses new therapies must target multiple components of the tumor microenvironment. As a result, second and third generation viruses have been created that target both tumor cells and the tumor microenvironment.

Several viruses have been designed to enhance oHSV targeting and killing of GSCs. The rQestin34.5 virus contains a GFP-disabling insertion in UL-39 and it possesses one copy of RL-1 driven by a Nestin promoter. Nestin is an intermediate filament whose expression is upregulated in several cancers and it is highly expressed in cancer stem cells. Cancer stem cells promote resistance and progression [96, 127-129]. In intracranial

GB xenografts, rQNestin34.5 significantly enhanced the survival of mice compared to a control virus lacking Nestin driven γ34.5 [130]. A clinical trial with rQNestin34.5 is being planned. The G47∆ virus is a derivative of the G207 oHSV, and it possesses deletions in UL-39, RL-1, and α47. The deletion of the α47gene enhances MHC I presentation in infected cells and increases immune cell recognition [131]. The removal

of α47 also places the US11 gene under the control of an immediate/early viral promoter.

The early expression of US11 compensates for the RL-1 deletion, and inhibits PKR

mediated protein synthesis shutoff [132, 133]. In vitro, the more virulent G47∆ replicated

better and was more cytotoxic to GSCs than G207 and R3616 oHSVs [134]. The

combination of G47∆ with TMZ was also found to kill GSCs synergistically and prolong

the survival of mice with GSC enriched intracranial tumors [135].

In order to promote long term anti-tumor immune responses new oHSVs have been generated that express potent, immunostimulatory molecules in infected cells. Several oHSVs have been created to express cytokines such as IL-18, IL-10, IL-4, and IL-12

[136, 137]. Of these viruses, the M032 oHSV is the most promising for GB. Similar to its parent G207, the M032 virus is deleted for both copies of RL-1 but the UL39 gene has been left intact. The M032 virus also expresses human IL-12. IL-12 is produced by dendritic cells, B-lymphocytes, and monocytes and it enhances cytotoxic T-cell and NK cell activity. IL-12 is also anti-angiogenic. Treatment of neuroblastoma tumors with the murine version of the virus increased T-cell infiltration into the tumor and increased animal survival [138]. Based on these results, a phase I clinical trial with M032 is currently being conducted for recurrent GB [NCT02062827]. Building on the success of

M032, a version of the G47∆ virus expressing IL-12 has been generated. Compared to

G47∆, the G47∆ -mIL12 virus inhibited angiogenesis, reduced the numbers of T- regulatory cells in the tumor microenvironment, and extended the survival of mice in an immune competent GSC model [139].

Second and third generation oHSVs have also been evaluated for their efficacy against breast cancer brain metastases. The intracarotid administration of G47∆ after blood brain barrier disruption with mannitol was shown to significantly improve the survival of mice with intracranial MDA-MB-435 breast cancer brain tumors [140]. G207-like viruses expressing immune-modulatory cytokines such as IL-4, CD40L, and secondary lymphoid tissue chemokine have also been evaluated in breast cancer brain metastasis mouse models. In mice with 4T1 breast cancer brain tumors, viruses expressing these cytokines had significantly stronger anti-tumor effects than a control OV [137].

Several oHSVs have also been developed that target brain tumor extracellular matrix.

Tumor extracellular matrix promotes immune suppression, tumor growth, and angiogenesis. Tumor extracellular matrix proteins, such as collagen, have been shown to act as a physical barrier to oHSV spread and efficacy [141-143]. Proteases such as collagenase, matrix metalloproteinases [MMPs], and chondroitenase ABC [Chase] have been shown to significantly enhance viral dispersion and tumor cell killing [144-147].

Recently, an oHSV expressing Chase was created to target Chondroitin-sulfate- proteoglycans secreted by GB tumors. The Chase oHSV virus expresses Chase under an immediate-early [IE] 4/5 viral promoter, contains deletions in both copies of 34.5, and possesses a gene disabling GFP insertion in ICP6. Compared to a control OV, Chase oHSV significantly improved virus replication, viral spread, and survival in GB xenografts [148].

Tumor vasculature is critical for tumor maintenance and growth. Tumor blood vessels

provide nutrients and oxygen, remove waste, produce growth factors and cytokines, as

well as recruit tumor promoting immune cells. These features make tumor endothelium

and pro-angiogenic factors attractive therapeutic targets. Several oHSVs have been

generated that deliver angiostatic and anti-angiogenic gene therapy. A G47∆ derived

oHSV expressing the anti-angiogenic protein Platelet factor 4 [PF4] has been created.

The oHSV directed expression of PF4 reduced endothelial cell migration in vitro and

induced strong anti-tumor effects in subcutaneous GB xenografts [149]. The same group

took a complimentary approach and generated an oHSV expressing a dominant negative

fibroblast growth factor [FGF] receptor. FGF is overexpressed in several cancers, such as

GB, and it strongly promotes tumor cell growth and angiogenesis [150]. In subcutaneous

GB xenograft tumors, the G47∆ virus expressing a dominant negative FGF receptor reduced tumor vasculature and increased survival significantly more than G47∆ [151].

Vasculostatin expressing oHSVs for GB and Breast Cancer Brain Metastases

Vasculostatin [Vstat120] is a secreted, anti-angiogenic protein produced by the cleavage of Brain Angiogenesis Inhibitor 1 [BAI1]. The loss of BAI1 expression is observed in several cancers including glioblastoma, colorectal cancer, gastric cancer, and renal cell carcinoma [152]. The re-expression of BAI1 or Vstat120 exerts potent anti-angiogenic

and anti-tumor effects in animal models of glioblastoma and renal cell carcinoma [153,

154]. Two oHSVs that express Vstat120 have been created- RAMBO and 34.5ENVE.

The RAMBO virus expresses Vstat120 under an IE4/5 viral promoter, contains deletions

24 in both copies of RL-1, and possesses a gene disabling GFP insertion in UL-39 [155].

The 34.5ENVE virus is derived from the rQestin34.5 backbone and it expresses Vstat120 under an IE4/5 viral promoter [156]. Both the RAMBO and 34.5ENVE virus demonstrated strong anti-angiogenic and anti-tumor effects in murine models of glioblastoma [155-157]. The expression of Vstat120 by RAMBO and 34.5ENVE reduced endothelial cell migration and tube formation. In GB survival studies, the Nestin driven

34.5ENVE virus performed significantly better than the RAMBO virus in 4 different xenograft models. Interestingly, mice treated with either Vstat120 expressing virus had better survival than mice treated with rQNestin34.5 [156]. These results highlight the strong anti-angiogenic effects of Vstat120 in controlling tumor growth. Recently,

34.5ENVE was also shown to significantly enhance the survival of mice in two different models of breast cancer brain metastases [Discussed in detail in Chapter 2] [11].

Section 4: Immune Responses to Oncolytic Viral Therapy

In the context of OV therapy, the immune response is a double-edged sword. On one hand, the innate immune responses result in rapid viral clearance and decreased OV efficacy, while on the other hand, immune responses elicited after viral infection also have the potential to activate an adaptive anti-tumor immune response to promote tumor eradication. Here we summarize some of the challenges and recent progress made by investigators in manipulating the immune response with respect to OV therapies for GB.

The rapid innate immune response induced by OVs is thought to promote viral clearance, inhibit viral replication, and reduce tumor cell killing. In the brain, the influx of monocytes, neutrophils, and natural killer (NK) cells following OV treatment has been correlated with reduced viral propagation limiting efficacy through the up-regulation of chemokines and cytokines. Several studies testing the efficacy of immunosuppressive agents given in conjunction with virotherapy to modulate these early defenses have shown promise. The most well studied of these agents is cyclophosphamide [CPA], and it has been shown to improve viral load and efficacy in numerous animal studies when given in conjunction with OV [158]. At high doses, the primary mechanism of action of

CPA is thought to be through its cytotoxic effects on immune cells, but the drug has also been shown to reduce the levels of circulating IgM and anti-viral antibodies.

Interestingly, treatment of tumor bearing animals with a low dose CPA has also been shown to inhibit T-regulatory cells and enhance NK cell anti-tumor activity activated by viral treatment. In mouse melanoma studies, low dose CPA in conjunction with Reovirus

and IL-2 were found to significantly enhance viral efficacy through its

immunostiumulatory effects.[159] Irrespective of the underlying mechanism, all of these

studies collectively observed increased anti-tumor responses when OV is administered in conjunction with CPA. The ability of CPA to enhance OV therapy is currently being investigated with oncolytic measles virus (MV-NIS) in patients diagnosed with myeloma

(NCT00450814).

The role of phagocytic macrophage cells in limiting the efficacy of oncolytic viral

therapy has also been studied using agents such as clodronate liposomes (CLs). CLs can

destroy monocytic/macrophage cells in vivo. While treatment with CLs increased tumor

viral load it did not enhance the survival of rats bearing intracranial GBs.[160] This result

is thought to be due primarily to the inability of CLs to cross the blood brain barrier and

neutralize the resident microglia. While the nervous system and brain are considered to

be “immune privileged”, immune cells, such as NK cells, infiltrate the CNS upon OV

infection. A recent study investigating the negative impact of NK cells on efficacy of OV

in treating intracranial GB in mice, found that the deletion of NK cell cytotoxicity

receptors improved oHSV therapy, and suggests the targeting of these receptors may help

improve OV efficacy in patients.[161] This study highlights the importance of targeting

multiple cell types involved in the initial immune responses to OV infection in the brain.

Tumor cell invasion into the normal brain is one of the hallmarks of glioblastoma, and so

treatment with a systemic agent that can reach distant invading cells is considered

optimal. The systemic delivery of most OVs has remained a challenge due to their rapid

serum neutralization. The use of Cobra Venom Factor (CVF) to inactivate the C3

component of complement has been shown to improve virus stability in serum.[162]

Copper present in serum has also been shown to inhibit the ability of oncolytic HSV to

destroy and reach intracranial tumors. Interestingly, copper is also vital for tumor

angiogenesis, and the anti-neoplastic effects of copper chelation are currently being

evaluated in several clinical trials (NCT00383851, NCT00405574, NCT00176800).

Treatment of animals bearing GB tumors with the copper chelating agent ATN-224

improved tumor virus loads, increased tumor cell killing, and enhanced animal

survival.[163] Copper also plays a key role in immune cell regulation, and it is important

for neutrophil, NK cell, and macrophage function.[164] While the impact of copper

chelation on reducing innate immune cell function was not directly examined in the

ATN-224 study, these experiments provide a foundation for future work examining the

combination of copper chelators with OV therapies. It is important to note, however, that

copper is also important for T-cell maintenance and function and thus the effects of

copper chelation on the development of anti-tumor immune response following OV

treatment remain to be elucidated.

The type I interferon (IFN) response is one of the major antiviral responses that limits

OV replication. Agents that can transiently suppress this response have been shown to

promote virus replication and enhance OV efficacy in various preclinical studies. Histone

Deacetylases (HDACs) are important gene regulators and the inhibition of HDAC

28 activity has been shown to inhibit cell growth and induce apoptosis.[165] Importantly,

HDAC inhibitors have also been shown to inhibit the IFN-mediated antiviral response.[166] HDAC inhibitors such as Trichostatin A and Valproic Acid have been shown to improve the oncolytic effects of OV against brain tumors in preclinical animal models.[167] The ability of HDAC inhibitors to target cancer cells and modulate immune responses makes combination studies highly significant. We have previously demonstrated that OV treatment induces the expression of the extracellular matrix protein

Cysteine Rich 61 (CCN1).[168] CCN1 is known to promote tumor angiogenesis, and we recently reported that this protein also plays a significant role in the induction of a Type-1

IFN antiviral response resulting in the activation of the Jak/Stat Signaling pathway.[169]

Future studies will unveil the significance of disrupting CCN1 signaling in oncolytic viral efficacy.

Interestingly, strategies to limit tumor angiogenesis in conjunction with viral therapy have been investigated for their immunomodulatory effects.[170] Anti-angiogenic agents not only restrict the flow of oxygen and nutrients into the tumor, but they also destroy the

“highways” used by immune cells to infiltrate into the tumor microenvironment. In one study, the treatment of animals with the anti-angiogenic agent Cilengitide decreased the influx of CD45 positive immune cells into the tumor microenvironment following OV therapy and also improved the survival of rats bearing intracranial GBs.[171] Similarly, the combination of Bevacizumab with an oncolytic virus expressing the anti-angiogenic protein angiostatin (G47Δ-mAngio) also significantly reduced tumor vasculature and macrophage accumulation in the tumor microenvironment resulting in increased virus

29 distribution, tumor cell killing, and animal survival.[172] A randomized Phase II study testing safety and efficacy of bevacizumab with Reolysin in patients with metastatic colorectal cancer is ongoing will uncover the clinical efficacy of this strategy

(NCT01622543).

While the transient suppression of the innate immune response increases virus replication and tumor cell killing, it is important to note that the generation of a strong anti-tumor immune response is considered to be just as important in creating successful OV therapies. For example, the long term administration of the immune suppressive corticosteroid, dexamethasone, with OV was unable inhibit tumor growth in mice with subcutaneous neuroblastoma tumors. This observation was thought to be the result of dexamethasone mediated suppression of tumor-specific cytotoxic T lymphocytes, and it highlights the importance of generating an anti-tumor immune response.[173] The creation of “armed” viruses to activate and amplify the anti-tumor immune response is currently an intense area of study. oHSVs expressing IL-12 and IL-4 in order to help generate more potent T-cell responses and anti-tumor immunity against treated brain tumors have shown improved anti-tumor efficacy.[136, 138] While the generation of an antitumor immune response leading to the eradication of tumors is currently being tested, the activation of an unbridled immune response against central nervous system tumors has to be approached with caution.

Currently, there is an array of pharmacological inhibitors as well as an emerging number of second generation viruses designed to affect different aspects of the immune response to OV therapy. Future work focusing on the combination of drugs targeting antiviral immune responses with rational, “armed” OVs to generate the optimal anti-tumor immune response will uncover the potential of this very promising therapy.

Reproduced with the permission of Informa Healthcare:

Meisen WH and Kaur B. How can we trick the immune system into overcoming the detrimental effects of oncolytic viral therapy to treat glioblastoma? Expert Rev Neurother

13(4), 341-3 (2013). PMID: 23545048

Chapter 2: Changes in BAI1 and Nestin Expression Are Prognostic Indicators for

Survival and Metastases in Breast Cancer and Provide Opportunities for Dual Targeted

Therapies

Abstract

The 2 year survival rate of patients with breast cancer [BC] brain metastases is less than

2%. Treatment options for BC brain metastases are limited and there is an unmet need to identify novel therapies for this disease. Brain angiogenesis inhibitor 1 [BAI1] is a GPCR involved in tumor angiogenesis, invasion, phagocytosis, and synaptogenesis. For the first time, we identify BAI1 expression is significantly reduced in BC and higher expression is associated with better patient survival. Nestin is an intermediate filament whose expression is up-regulated in several cancers. We found higher Nestin expression significantly correlated with BC lung and brain metastases, suggesting both BAI1 and

Nestin can be therapeutic targets for this disease. Here, we demonstrate the ability of an oncolytic virus, 34.5ENVE, to target and kill high nestin expressing cells and deliver

Vstat120 [extracellular fragment of BAI1]. Finally, we created two orthotopic immune competent murine models of BC brain metastases and demonstrated 34.5ENVE extended the survival of immune competent mice bearing intracranial breast-cancer tumors.

Introduction

Breast cancer is one of the leading causes of brain metastases. The 2 year survival rate of

patients with breast cancer [BC] brain metastases is less than 2% [8]. Treatment options for patients refractory to standard surgery and radiation are limited. These tumors are frequently resistant to conventional chemotherapeutic drugs and antibody-based therapies that poorly penetrate the blood brain barrier [174, 175]. There is an unmet need to identify novel, targeted strategies to treat this disease.

Vstat120 is a cleaved and secreted fragment of Brain Angiogenesis Inhibitor 1 [BAI1].

The loss of BAI1 expression is observed in several cancers including glioblastoma, colorectal cancer, gastric cancer, and renal cell carcinoma.[152] The re-expression of

BAI1 or Vstat120 exerts potent anti-angiogenic and anti-tumor effects in animal models of glioblastoma and renal cell carcinoma [153, 154]. Surprisingly, its expression and function has not been examined in BC. Nestin is an intermediate filament whose expression is upregulated in several cancers [127-129]. Nestin is also expressed in cancer stem cells. Cancer stem cells are known to promote cancer resistance and progression

[96]. Here we determined the roles of BAI1 and Nestin gene expression in breast cancer metastases and patient survival. We found lower BAI1 expression correlates with poorer patient survival, and high Nestin expression is associated with an increased probability of metastases.

34.5ENVE is an oncolytic Herpes Simplex Virus that expresses Vstat120 and its

replication is driven by a cancer stem cell specific Nestin promoter [156]. 34.5ENVE has

demonstrated unparalleled anti-tumor efficacy in murine models of glioblastoma [156].

Given the roles of Nestin and BAI1 in BC brain metastases, we tested the ability of

34.5ENVE to target BC cells in vitro. In these studies, 34.5ENVE killed BC cells of

varying molecular subtypes, including those known to frequently metastasize to the brain.

To test the therapeutic efficacy of a Vstat120 expressing oncolytic virus [OV] in vivo, we

created three new, orthotropic models of BC brain metastasis in immune-competent

FVB/NJ mice. Two of these models recapitulated the biology of human BC brain metastases [BCBM]. In both of these models, we found 34.5ENVE treatment significantly improved the survival of mice with established BCBM.

Results

BAI1 expression in reduced in breast cancer and is associated with patient survival

To determine the relevance of BAI1/Vstat120 in BC we analyzed patient derived gene

expression data from The Cancer Genome Atlas [TCGA]. We observed a 52% reduction

in BAI1 expression in invasive ductal breast carcinomas [n=389] compared to normal

breast tissue [n=61; P<0.0001] [Figure 1A].[176] Further analysis revealed low BAI1

expression was also associated with decreased disease free survival [n=324; P<0.03]

[Figure 1B]. An examination of BAI1 expression in 50 breast cancer cell lines from the

Neve et al dataset showed BAI1 mRNA levels were reduced in 38% of breast cancer cell

lines compared to the MCF-10A breast epithelial cell line [19 of 50 cell lines] [Figure

1C] [177]. These results suggest the loss of BAI1 promotes BC tumorigenesis and the

restoration of BAI1/Vstat120 may have therapeutic effects in BC.

Nestin expression is up-regulated in breast cancer and is associated with metastases

Nestin is up-regulated in several metastatic cancers, and its high expression correlates

with reduced BC patient survival [127-129]. Nestin is also expressed in cancer stem cells known to promote cancer resistance and progression [96]. Analysis of a cohort of 166 patients stratified by median Nestin expression revealed a significant association between

Nestin expression and incidence of brain and lung metastases [n=164; P<0.02] [Figure

2A] [178]. Of the BC subtypes, we found high Nestin expression was most strongly

associated with triple negative breast cancers [TNBC]. This subtype is highly prone to

brain metastases [Figure 2B] [128]. Additional analysis of the Neve et al microarray

dataset showed Nestin was up-regulated in 100% of the BC cell lines examined [50 of

50] [Figure 2C] [177]. These results suggest that Nestin may be a strong therapeutic

target for aggressive and metastatic BCs.

Oncolytic virus is cytotoxic in multiple human BC subtypes

34.5ENVE expresses Vstat120 and its replication is driven by a Nestin promoter, thus we

hypothesized 34.5ENVE might be therapeutically relevant for BC brain metastases. To

test this hypothesis, we tested infection, replication, and cytotoxicity of 34.5ENVE in a

variety of human BC cell subtypes [Figure 3 and Table 1]. Over the course of 3 days,

34.5ENVE infected and replicated in human BC cells as determined by increasing virus

encoded GFP expression [Figure 4A]. In vitro cytotoxicity of human BC cells to

34.5ENVE infection was dose dependent and increased with time [Figure 4B-C]. Four

days after infection at an MOI of 0.05, we observed 83.4%, 75.7%, 80.6%, and 90.1%

cell death in MDA-MB-231, SKBR3, MCF7, and MDA-MB-468 cells, respectively.

Most BC treatments are targeted to particular subtypes, but 34.5ENVE killed BC cells

across multiple subtypes. Importantly, we observed significant killing in the Her2+ and

TNBC subtypes. TNBCs are notoriously resistant to conventional therapies, and Her2+

targeted antibody treatments poorly penetrate the blood-brain-barrier. As a result, 25-55% of these BC patients will develop brain metastases.[5] These results highlight the therapeutic potential of 34.5ENVE to treat BC brain metastases.

HSV-1 replication and cytotoxicity is enhanced by the viral neurovirulence gene ICP34.5

[156]. To improve the safety and targeting of the 34.5ENVE virus, the expression of

ICP34.5 is driven by a Nestin promoter. Nestin expression was increased in all 50 of the

BC cell lines examined suggesting it is a relevant therapeutic target for BC [Figure 4C].

In order to determine the efficacy of Nestin driven ICP34.5 on tumor cell killing, we

compared the cytotoxicity of 34.5ENVE with a similar virus lacking Nestin driven

ICP34.5 expression [RAMBO] [155, 156]. We observed 54.14% increased killing in the

TNBC MDA-MB-468 cells in the Nestin-driven 34.5ENVE virus as compared to a virus

without ICP34.5 [P<0.001] [Figure 4D]. These results further support the use of

34.5ENVE for the treatment of this disease. This is the first study to specifically use

Nestin expression to target BC.

Syngeneic model of breast cancer brain metastases in an HSV-1 sensitive strain

While there are several excellent models to study the biology of brain metastasis

development in immune compromised mice, there are currently few immune competent

models to test the safety and efficacy of its potential therapies.[8] Cody et al previously described an immune competent model of BCBM to test OVs, but the virus had limited anti-tumor efficacy in vitro and in vivo suggesting it was not an optimal model to evaluate oncolytic HSV derived therapeutics.[179] For these studies we characterized three novel, murine BC [DB-7, Met-1, and Mvt1] models. DB-7 and Met-1 cells are derived from transgenic FVB/N mice expressing polyoma virus middle T oncogene

[PyVmT] under the control of a mammary epithelium promoter [180]. PyVmT serves as

a surrogate for activated receptor tyrosine kinase signaling pathways, such as Her2,

commonly activated in BC [5]. Mvt1 cells are derived from tumors of MMTV-c-

myc/VEGF bitransgenic mice [181]. In these studies, DB-7, Met-1, and Mvt1 BC cells

were implanted intracranially into the brains of FVB/NJ mice. DB-7 and Met-1 tumor borders were generally demarcated from the normal brain parenchyma with localized invasion in the Met-1 tumors. Similar to patient specimens, these tumors contained

significant tumoral vascularization as well as tumor associated microglia/macrophages

[Figure 5]. Mvt1 tumors were highly invasive and infiltrated into distant brain structures

including the ventricles and meninges. This phenotype was characteristic of rare

leptomeningeal metastases and did not resemble the parenchymal brain metastases

commonly observed in patients [182]. Histological comparisons of patient BC brain

metastases with all three murine tumors indicated DB-7 and Met-1 derived tumors most

closely recapitulated the human CNS metastases and so these BCBM models were

selected for further analysis [Figure 5] [10].

34.5ENVE is cytotoxic to murine breast cancer cells in vitro

Human tropic viruses often replicate poorly in murine cells, so there are very few

immune competent models of cancer to study OVs derived from HSV-1. To determine if we could evaluate the therapeutic effects of 34.5ENVE in this murine BCBM model, we

examined the ability of the virus to infect and replicate in the DB-7 and Met-1 tumor

derived cell lines. For these assays human glioma cells were used as a positive control.

We observed an increase in virus encoded GFP expression over 48 hours following

infection at a low MOI consistent with virus replication and spread in these cells [Figure

6]. Quantification of virus replication revealed that DB-7 and Met-1 murine cells

supported replication at a similar rate compared to human glioma cells [Figure 7A].

34.5ENVE killed tumor cells in a dose and time dependent manner at levels comparable

to human glioma cells [Figure 7B-C]. Three days following infection at an MOI of 0.01,

we observed 91.5%, 82.5%, and 88.2% cell death in DB-7, Met-1, and human glioma

cells, respectively. We also verified these cells secreted virally expressed Vstat120 and

demonstrated improved cytotoxicity of the ICP34.5 expressing 34.5ENVE virus [Figure

8A-B]. The characterization of this HSV-1 sensitive murine model will aid in the future evaluation of preclinical toxicity and efficacy of novel, HSV-1 derived therapeutics.

34.5ENVE treatment extends survival in vivo

We utilized MRI to non-invasively evaluate the antitumor response of 34.5ENVE in mice

with established Met-1 brain tumors. Mice were treated intratumorally with a single dose

of HBSS or 34.5ENVE [n= 6/group] 14 days post-tumor cell implantation [average initial

tumor volume 4.94 mm3]. Figure 9A shows representative coronal T1-weighted MRI

images from mice treated with PBS or 34.5ENVE 1 day pretreatment and on days 6 and

10 post-treatment. The tumor volumes in mice treated with HBBS grew rapidly, and

obtained an average tumor volume of 59.01 mm3 within 10 days of treatment [Figure 9B,

Table 2]. Significantly, 34.5ENVE treated tumors showed substantial decreases in tumor

volume [3.43 mm3 average tumor volume 10 days post viral therapy; P<0.02].

Interestingly, we observed initial pseudoprogression of tumors [by volume] in 34.5ENVE

39 treated mice prior to tumor regression possibly due to tumor destruction and immune cell infiltration. Following these mice over time, we observed 34.5ENVE treatment significantly enhanced the survival of mice bearing Met-1 BC brain tumors. Control treated mice had a median survival of only 36 days, whereas mice receiving 34.5ENVE therapy survived significantly longer [median survival 52 days; P<0.038] [Figure 9C].

We next tested the antitumor effects of 34.5ENVE in the DB-7 BCBM model. Mice treated with 34.5ENVE showed a 100% increase in median survival compared to control mice with DB-7 tumors. HBSS [n= 5] and 34.5ENVE [n=7] treated mice showed median survival times of 17 and 34 days, respectively [Figure 9D] [P<0.0004].

Discussion

BC brain metastases continue to present a significant therapeutic challenge. A recent BC

BM clinical trial with lapatinib and capecitabine noted that nearly a third of patients

experienced at least one severe adverse event due to toxicity [33]. Conversely, OV

therapies which are currently in clinical trials for a variety of solid tumor malignancies

including BC and brain tumors [NCT01656538, NCT02031965, NCT01174537,

NCT00794131] have proven to be safe and well tolerated. In this study, we identified

BAI1/Vstat120 and Nestin as novel therapeutic targets for BCBMs. We demonstrated an

OV, 34.5ENVE, expressing anti-angiogenic Vstat120 and ICP34.5 under a Nestin promoter had significant cytotoxic effects in BC cells of varying molecular subtypes, including Her2+ and TNBC. Significantly, we also described two novel, immune competent murine models of BCBMs that closely recapitulates the human disease.

Finally, we demonstrated that a single, intratumoral dose of 34.5ENVE virus significantly

enhanced the survival of mice with established metastatic BC brain tumors. The results of

these studies warrant further investigation of BAI1 and Nestin dual targeted therapies to

treat established BC brain metastases.

Reproduced with the permission of AACR:

Meisen WH, Dubin S, Sizemore ST, Mathsyaraja H, Thies K, Lehman N, Jaime-Ramirez

AC, Elder JB, Ostrowski M, Kaur B. Oncolytic Viral Therapy Enhances the Survival of

Mice in a Novel Model of Breast Cancer Brain Metastasis. Molecular Cancer

Therapeutics. 2014. PMID: 25376607[E-published ahead of print]. 41

Future Directions: PARP Inhibitors in Combination with oHSV therapy for Breast

Cancer Brain Metastases

Approximately, 25-55% of HER2+ and Triple Negative BC patients will develop brain metastases [5]. There are relatively few treatments for patients with BC brain metastases, and the 2 year survival rate is less than 2% [8]. In the previous study we found

34.5ENVE oHSV therapy may be a viable treatment option for patients with this disease.

In order to improve upon this potential therapy, we are currently testing the combination of 34.5ENVE with other agents being evaluated for BC brain metastases [11]. PARP inhibitors disrupt DNA damage repair, and these drugs are particularly effective in cancers with defects in DNA repair pathways such as those with BRCA1/2 and PTEN mutations [38, 39, 183]. In December 2014, the PARP inhibitor olaparib was approved for the treatment of advanced ovarian cancer in patients with BRCA1/2 mutations.

Additionally, the PARP inhibitor veliparib [ABT-888] was recently tested in a phase I clinical trial for BC brain metastasis [NCT00649207- results pending]. Given these recent clinical trials, we are currently examining the combination of oHSV therapy with PARP inhibitors for BC brain metastases. For these studies we selected the PARP inhibitor olaparib because it is FDA approved.

Results

We have conducted preliminary in vitro studies to test the combination of 34.5ENVE and olaparib. For these studies we utilized a triple negative [MDA-MB-468] and a HER2+

[SKBR3] cell line because these BC subtypes are most likely to metastasize to the brain.

The combination of 34.5ENVE with olaparib induced significantly more cell killing than either agent alone [Figures 10A and Figure 11A]. Chou-Talalay analysis of this data indicated the combination of 34.5ENVE with olaparib induced synergistic cell death in both cell lines [Combination Index values <1] [Figures 10B and 11B]. We also performed virus replication assays and determined the combination of both agents did not reduce virus replication [Figure 12A-B].

Conclusions

This preliminary data suggests the combination of oHSV therapy with FDA approved

PARP inhibitors warrants further investigation for BC brain metastases. We are currently testing the combination of 34.5ENVE with olaparib in vivo.

Materials and Methods

Cell Lines and Viruses

Vero, DB-7, Met-1, U251-T3, MCF7, MDA-MB-231, and MDA-MB-468 cells were maintained in DMEM supplemented with 10% fetal bovine serum [FBS]. SKBR3 cells were maintained in McCoy's 5A Medium supplemented with 10% FBS. All cells were incubated at 37oC in an atmosphere with 5% carbon dioxide and maintained with 100

units of penicillin/mL, and 0.1 mg of streptomycin/mL. U251 cells were obtained from

Dr. Erwin G. Van Meir [Emory University, Atlanta, Georgia] and authenticated by us

through the University of Arizona Genetics Core in July 2013. U251-T3 cells were

created in our lab [May 2009] as a tumorigenic clone of U251 cells by serially passaging

these cells three times in mice [these cells have not been separately authenticated]. DB-7,

Met-1 [murine BC], MCF7, MDA-MB-231, SKBR3, and MDA-MB-468 [human BC]

cells were obtained in December 2012 from Dr. Michael C. Ostrowski [Ohio State

University, Columbus, OH] and have not been authenticated since receipt [180, 184].

Monkey kidney epithelial derived Vero cells were obtained in April 2005 from Dr. E

Antonio Chiocca [Ohio State University, Columbus, Ohio]. These cells have not been

authenticated since receipt. All cells are routinely monitored for changes in morphology

and growth rate. All cells were negative for mycoplasma. RAMBO and 34.5ENVE

viruses were prepared and titered as previously described [156].

Cell Viability Assays

Cells were plated in 96 well plates and infected simultaneously with 2% FBS in DMEM containing 34.5ENVE at the indicated Multiplicity of Infection [MOI]. Viability was assessed as described previously using a standard MTT [3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide] assay [185].

Virus replication assay

500,000 cells plated in 6 well plates were infected with 34.5 ENVE at an MOI of 0.005. 3 days later cells and supernatants were harvested and the viral titers were determined via a standard plaque forming unit assay.

Animal surgery

All animal experiments were performed in accordance with the Subcommittee on

Research Animal Care of The Ohio State University guidelines and were approved by the institutional review board. 6-8-week-old, female FVB/N mice [The Jackson Laboratory,

Bar Harbor, Maine], were used for in vivo tumor studies. Intracranial surgeries were performed as previously described with stereotactic implantation of 100,000 DB-7, Met-

1, or Mvt1 cells [156]. Tumors were treated with HBSS or 34.5ENVE virus at the location of tumor implantation. Animals were euthanized when they showed signs of morbidity.

Immunohistochemistry/immunofluorescence

Mouse BCBM tumors were fixed in zinc formalin [Anatech Ltd] and paraffin embedded.

Tumors were sectioned at 5 μm and stained using the following antibodies: anti-

MECA32 [TROMA-1], anti-F4/80 [Invitrogen; MF48000], Alexa Fluor 594 [Invitrogen].

Human tumor immunohistochemistry was performed using antigen retrieval at pH 6.0.

Tumors were stained with anti-CD163 [Leica Microsystems Novocastra], anti-CD31

[Dako; M0823], and a HRP-linked secondary antibody. Specimens were visualized with

DAB.

Image Acquisition

Immunofluorescent images were acquired using a Nikon Eclipse E800 epifluorescence microscope equipped with a Photometrics Coolsnap camera and Nikon Plan Fluor objectives. MetaVue software [Molecular Devices] was used for image acquisition.

Immunohistochemical staining was imaged using a Nikon Eclipse 50i microscope equipped with an Axiocam HRC camera [Zeiss] and Nikon Eclipse Ci microscope equipped with a DS-Fi2 camera system.

MRI Analysis

Mice-bearing intracranial tumors were treated with HBBS or virus. Anatomic imaging

was performed on the days indicated using a Gadolinium enhanced T1-weighted imaging

sequence. For data analysis, a region-of-interest [ROI] that included the tumor was

46 manually outlined. Tumor volumes were calculated from ROIs as previously described

[156].

Statistical Analysis

Student's t-test was used to analyze changes in cell killing, viral plaque forming assays, and tumor volume measurements. A P < 0.05 was considered statistically significant. In survival assays, Kaplan–Meier curves were plotted and the log rank test was utilized to determine statistical significance. All statistical analyses were performed with the use of

Graph Pad Prism software [version 5.01].

Figures and Tables

Figure 1: Reduced BAI1 is associated with breast cancer patient survival

A. Mean BAI1 expression in 61 normal mammary gland and 389 invasive ductal breast carcinoma samples in the TCGA data set [P = 0.000137896]. B. Kaplan Meier curves representing DFS in patients of the TCGA cohort with high [n = 162] or low [n = 162] BAI1 expression [P=0.03]. C. Analysis of BAI1 expression in the Neve et al. breast cancer cell line microarray database. Expression levels are normalized to the non- tumorigenic, epithelial cell line MCF-10A.

Figure 2: Increased Nestin expression is associated breast cancer metastases

A. The Bos et al cohort was stratified by median NES expression [83 NES high and 83 NES low patients] to examine the probability of brain and lung metastases in BC patients [P = 0.02]. B. NES expression correlates with TNBC status in the Curtis et al dataset. 1975 samples in the Curtis et al breast cancer patient cohort were stratified by TNBC status. There are 250 TN and 1725 patients with other biomarker status in this cohort. Mean NES expression is significantly higher [P = 2.79E-37] in TNBCs. C. Analysis of NES expression in the Neve et al. breast cancer cell line microarray database. Expression levels are normalized to the non-tumorigenic, epithelial cell line MCF-10A.

Cell Line ER PR HER2 Nestin MDA-MB-231 + MDA-MB-468 + MCF7 + + + SKBR3 + + Table 1: Description of BC molecular subtypes and Nestin expression in a panel of human BC cells.

ER, Estrogen Receptor; PR, Progesterone Receptor; HER2

Figure 3: Structure of 34.5ENVE Virus

The 34.5ENVE Virus is deleted for both copies of γ34.5 and contains a gene disrupting GFP insertion in ICP6. 34.5ENVE expresses and secretes the anti-angiogenic gene Vstat120 under a viral IE4/5 promoter. A single copy of γ34.5 driven by a Nestin promoter is reinserted in the viral backbone.

Figure 4: Oncolytic HSV derived therapeutics target and kill multiple BC subtypes in vitro

A. The ability of 34.5ENVE to infect and replicate in a panel of human BC cells was determined using 34.5ENVE directed GFP expression. The human glioma line, U251-T3, was used a positive control. Dose dependent [B.] and temporal [C.] viability of human BC cells treated with 34.5ENVE. Data shown are mean cell viability ± SD. D. Relative cytotoxicity of Nestin driven 34.5ENVE or control RAMBO virus at an MOI of 0.01 in MDA-MB-468 cells. Data shown are mean cell viability ± SD [P<0.001].

Figure 5: Characterization of three murine models of BCBM for preclinical evaluation of oncolytic HSV-1 derived therapeutics

Representative panel of DB-7, Met-1, Mvt1, and human breast cancer brain metastases tumors. Human biopsy sample was stained for H&E, macrophages [CD163], endothelial cells [CD31] [100x magnification]. Murine specimens were stained for H&E, macrophages [F4/80], and endothelial cells [MECA-32] [20X Magnification].

Figure 6: 34.5ENVE replicates in murine BC cells

The ability of 34.5ENVE to infect and replicate in two murine breast cancer cells, DB-7 and Met-1, was determined using GFP imaging. The human glioma line, U251-T3, was used a positive control. 500,000 cells were plated in 2% FBS media then infected with 34.5ENVE at 0.001 MOI.

Figure 7: 34.5ENVE replicates in and kills murine BC cells in vitro A. 72 hour viral titers of DB-7, Met-1, and U251-T3 cells were infected with 34.5ENVE [0.005 MOI]. Data shown are mean viral titers ± SD [U251-T3 to DB7 P<0.01; U251-T3 to Met-1 P<0.001]. B. 48 hour cell viability of BC cells treated with 34.5ENVE at the indicated MOIs. Data shown are mean cell viability ± SD. C. Temporal response of murine BC cells treated with 34.5ENVE at 0.01 MOI for 3 days. Data shown are mean cell viability ± SD.

Figure 8: 34.5ENVE infected murine BC cells secrete Vstat120 and show enhanced

cytotoxicity when targeted with a Nestin-driven OV

A. Verification of Vstat120 production and secretion in vitro. TOP- To determine production of Vstat120 by 34.5ENVE we probed for Vsat120 via immunoblot of DB-7 and Met-1 cells. Cell lysates were harvested 24 hours after infection with 34.5ENVE at a MOI of 0.5 as described in Materials and Methods. BOTTOM- Immunoblot of DB-7 and Met-1 cell supernatants, infected and harvested as described above and concentrated by centrifugation in a 50kDa filter. B. 72 hour cell viability of murine BC cells infected with 34.5ENVE or same virus lacking Nestin promoter driven ICP34.5 [RAMBO] at an MOI of 0.01 [P<0.001].

Figure 9: Anti-tumor efficacy of 34.5ENVE in mice bearing established BCBM

A. Representative T1-weighted MRI images of coronal sections of mice with Met-1 BM tumors treated with HBSS or 34.5ENVE. Scans were performed on the days indicated after treatment [TX] [Day 14 post tumor cell implantation]. White arrows indicate tumor location. B. Mean tumor volumes [mm3] ± SD of Met-1 BM tumors treated with HBSS or 34.5ENVE treated mice [n=6 mice/group]. C. Kaplan-Meir survival curve for A and B, mice treated on day 14 with HBSS or 2.0 x 105 pfu 34.5ENVE [n=6 mice/group, P=0.038]. D. Kaplan-Meir survival curve of DB-7 BM tumors treated on day 7 with HBSS or 2.5 x 105 pfu 34.5ENVE [n=5 HBSS; n=7 34.5ENVE, P=0.0004].

Days Post Implantation HBSS (mm3) 34.5ENVE (mm3) p-Value 13 6.16 3.72 0.51 20 25.60 8.66 0.14 24 59.01 3.43 0.02 Table 2: 34.5ENVE treatment reduces tumor volumes in Met-1 BCBM tumors by MRI.

Data shown are mean tumor volumes [mm3] of Met-1 BM tumors treated with HBSS or 34.5ENVE from Figures 5A-C [n=6 mice/group].

Figure 10: The combination of 34.5ENVE with olaparib kills MDA-MB-468 BC cells synergistically.

A. MDA-MB-468 cells were treated with varying concentrations of olaparib and 34.5ENVE. After 3 days, cell viability was assessed via a standard MTT [3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay. B. Chou-talalay analysis was utilized to determine if the combination of olaparib with 34.5ENVE resulted in synergistic cell death. Combination Index values less than 1 are considered synergistic.

Figure 11: The combination of 34.5ENVE with olaparib kills SKBR3 BC cells synergistically.

A. SKBR3 cells were treated with varying concentrations of olaparib and/or 34.5ENVE. After 3 days, cell viability was assessed via a standard MTT [3-(4,5-dimethylthiazol-2- yl)-2,5-diphenyltetrazolium bromide] assay. B. Chou-talalay analysis was utilized to determine if the combination of olaparib with 34.5ENVE resulted in synergistic cell death. Combination Index values less than 1 are considered synergistic.

Figure 12: Olaparib does not affect 34.5ENVE virus replication.

500,000 MDA-MB-468 cells [A.] or 300,000 SKBR3 cells [B.] were plated in 6 well plates. Cells were treated with the calculated LD50 doses of olaparib and/or 34.5ENVE. Two days later cells and supernatants were harvested and the viral titers were determined via a standard plaque forming unit assay [n=3-4/group].

Chapter 3: The impact of macrophage and microglia secreted TNFα on oncolytic HSV-1

therapy in the glioblastoma tumor microenvironment

Abstract

Oncolytic herpes simplex viruses [oHSV] represent a promising therapy for glioblastoma

[GB], but their clinical success has been limited. Early innate immune responses to viral infection reduce oHSV replication, tumor destruction, and efficacy. Here, we characterized the antiviral effects of macrophages and microglia on viral therapy for GB.

Quantitative flow cytometry of mice with intracranial gliomas [± oHSV] was utilized to

examine macrophage/ microglia infiltration and activation. In vitro co-culture assays of

infected glioma cells with microglia/macrophages were utilized to test their impact on

oHSV replication. Macrophages from TNFα knockout mice and blocking antibodies were

used to evaluate the biological effects of TNFα on virus replication. TNFα blocking

antibodies were utilized to evaluate the impact of TNFα on oHSV therapy in vivo. Flow

cytometry analysis revealed a 7.9 fold increase in macrophage infiltration after virus

treatment. Tumor infiltrating macrophages/microglia were polarized towards a M1, pro- inflammatory phenotype and they expressed high levels of CD86, MHCII, and Ly6C.

Macrophages/microglia produced significant amounts of TNFα in response to infected glioma cells in vitro and in vivo. Utilizing TNFα blocking antibodies and macrophages derived from TNFα knockout mice we discovered TNFα induced apoptosis in infected

tumor cells and inhibited virus replication. Finally, we demonstrated the transient

blockade of TNFα from the tumor microenvironment with TNFα blocking antibodies

significantly enhanced virus replication and survival in GB intracranial tumors.

Translational Relevance

Glioblastoma is one of the most common and deadly types of primary brain tumors, and

patients diagnosed with these tumors have a median survival of only 15 months.

Oncolytic herpes simplex viruses [oHSV] represent a promising therapy for glioblastoma,

and these viruses are currently being tested in patients for safety and efficacy. Innate

immune responses to viral infection are thought to reduce oHSV replication, tumor

destruction, and efficacy. In this study, we investigated the anti-viral functions of

microglia and macrophages in oHSV therapy for glioblastoma. We identified

microglia/macrophage secreted tumor necrosis factor α [TNFα] as a major factor that reduces viral replication through the induction of apoptosis in infected cells. We demonstrated the inhibition of TNFα could significantly enhance virus replication and efficacy in vivo. The results of these studies suggest FDA approved TNFα inhibitors may significantly enhance patient outcomes in oHSV clinical trials.

Introduction

Glioblastoma [GB] is one of the most common and deadly types of primary brain tumors.

These tumors are characterized by widespread invasion, extensive angiogenesis, and resistance to cell death [186]. These features along with a restrictive blood brain barrier severely limit treatment options and result in a median patient survival of 15 months [73].

Oncolytic Herpes Simplex Viruses [oHSVs] are viruses genetically modified to specifically infect, replicate in, and target cancer cells for destruction. oHSVs represent a promising treatment modality for patients with GB, and in clinical trials these viruses are safe and well tolerated [187]. Early phase clinical trials have produced promising results and there is currently a phase III clinical trial for patients with advanced melanoma

[NCT00769704] [122, 188, 189].

The success of oHSV derived therapeutics is thought to depend on the oncolytic destruction of tumor cells and the activation of anti-tumor immune responses. These immune responses can potentially lead to long term cancer remission. However, the pro- inflammatory immune responses generated by viral infection can also antagonize oHSV replication and spread. Innate immune responses destroy replicating virus and reduce tumor cell killing, and several studies have demonstrated the negative effects of innate immune responses to oHSV treatment [161, 190, 191].

Microglia and infiltrating macrophages are thought to be significant mediators of the

innate immune response to viral infection in the CNS [192-196]. Depletion of these cells with clodronate liposomes or cyclophosphamide [CPA] reduces antiviral responses and improves oHSV efficacy [160, 197-201]. As a result of these preclinical studies the combination of oncolytic measles virus with CPA is currently being evaluated in a phase

I clinical trial for multiple myeloma [ClinicalTrials.gov Identifier: NCT00450814].

While these studies highlight the importance of modulating early immune responses to oHSV infection, the depletion of all phagocytic cells with clodronate liposomes or total immune suppression with high doses of CPA does not specifically address the mechanism by which macrophages and microglia limit oHSV replication, spread, and efficacy.

In this study, we investigated the impact of microglia and macrophages in oHSV therapy for GB. Quantitative flow cytometry analysis of mice with intracranial gliomas treated with oHSV revealed significant changes in the activation and infiltration of macrophages and microglia in oHSV treated animals relative to untreated mice. To evaluate the impact of these cells on oHSV propagation, we developed an in vitro co-culture system with infected glioma cells and microglia/macrophages. In these studies, macrophages and microglia significantly reduced virus replication. Furthermore, we identified microglia/macrophage secreted tumor necrosis factor α [TNFα] as a major factor that reduces viral replication through the induction of apoptosis in infected cells. In co-culture assays, we were able to rescue changes in virus replication with TNFα knockout

65 macrophages or TNFα function blocking antibodies. Finally, we demonstrated the specific inhibition of TNFα produced by the tumor microenvironment could significantly enhance virus replication and efficacy in vivo. The results of these studies suggest FDA approved TNFα inhibitors may significantly enhance patient responses in oHSV clinical trials.

Results

oHSV Therapy activates microglia in vivo

In GB animal models, microglia comprise 13-34% of all viable cells in the tumor [202].

Similar ranges are seen in human tumors, and these observations underscore the

importance of this cell type in the context of oHSV therapy [203, 204]. The ability of microglia in the tumor microenvironment to switch from a glioma supportive role to an anti-viral state following oHSV treatment has not been well studied. To examine changes in microglia activation following oHSV infection in vivo, we treated mice with established U87ΔEGFR intracranial tumors with oHSV or PBS [injection control]. These mice were euthanized 3 days following treatment, and we analyzed the tumor- and non- tumor bearing hemispheres for microglia [CD11b+CD45lo] MHCII expression [Figure

13A]. We observed an 8.75 fold increase in microglia MHCII expression in oHSV treated

mice compared to PBS treated animals [P<0.001] [Figure 13B]. Interestingly, oHSV

therapy up-regulated mean microglia MHCII expression in both the tumor and non tumor

bearing hemispheres of the brain, but this increase was higher in the tumor bearing

hemisphere [16.33% MHCII+] compared to the non-tumor bearing hemisphere [8.23%

MHCII+] [Figure 13C].

oHSV Therapy Increases Macrophage Infiltration into the Brain Tumor

Microenvironment

Microglia activation induces the expression of various cytokines and chemokines that can

stimulate the migration of immune cells into the CNS [196]. Macrophages are important

mediators of this innate immune response to viral infection, but the extent of macrophage

infiltration into the CNS following oHSV therapy is unknown. To quantify the impact of

oHSV induced macrophage migration, we treated mice with established intracranial

U87ΔEGFR tumors with oHSV or PBS as described earlier. We observed a 7.96 and 5.70

fold increase in macrophage [CD11b+CD45hi] infiltration into the tumor and non-tumor

bearing hemispheres following oHSV infection, respectively [n=5/group; P<0.001 and

P<0.05] [Figure 13D-E]. While, oHSV therapy strongly induced macrophage infiltration

into both hemispheres, this increase was significantly higher in the tumor bearing

hemisphere [P<0.001] [Figure 13D-E]. While macrophages comprised the bulk of the

innate immune cell infiltrate, other innate immune cells populations are known are

known to migrate into the CNS following viral infection [161, 205]. We examined the

percoll isolated cell populations for Ly6G+ neutrophils and CD160+ natural killer [NK] cells, and we found few NK cells or neutrophils in the CD11b+CD45+ populations at this

time point [Figure 14A].

oHSV Therapy Increases Macrophage Activation in the Brain Tumor Microenvironment oHSV therapy induced significant macrophage infiltration into the brain tumor microenvironment, but the phenotype and activation of these cells remained unknown.

Depending on their polarization, macrophages can promote an immune-suppressive or pro-inflammatory tumor microenvironment. The activation status of these infiltrating cells is crucial to understanding how these cells contribute to oHSV therapy for GB. To determine the polarization status of infiltrating macrophages, we evaluated the expression

of the classic activation markers CD86, Ly6C, and MHCII. We observed significant

increases in the percentages and cell numbers of macrophages [CD11b+CD45hi] positive

for CD86+ and LY6C+ following oHSV treatment compared to PBS treatment [P<0.001;

P<0.001, respectively] [Figure 15A-B]. While the percentages of MHCII positive macrophages [CD11b+CD45hi] in the tumor environment between oHSV and control

treatments did not change, we observed a 9 fold increase in the total numbers of MHCII

positive macrophages infiltrating the tumor bearing hemisphere following oHSV therapy

compared to control treated mice [P<0.001] [Figure 15C]. The surface expression of

these three activation markers increased on macrophages[CD11b+CD45hi] in both the

treated and untreated hemispheres, but the treated hemispheres contained higher

percentages of macrophages that expressed CD86 and LY6C [Figure 15A-C]. Together,

these data suggested the infiltrating macrophages were polarized toward a pro-

inflammatory state.

Co-culture of oHSV-Infected Tumor Cells with Microglia or Macrophages Reduces Viral

Replication in Vitro

oHSV treatment significantly activated microglia/macrophages in vivo, but the effects of

these polarized immune cells on virus replication remained unknown. To determine the

functional consequences of this microglia/macrophage activation we developed an in

vitro co-culture system. Human glioma cells were infected with oHSV at a MOI of 2,

washed to remove unbound virus, and then overlaid with murine macrophages

[RAW264.7] or microglia [BV2] [Figure 16A]. To specifically examine the

microglia/macrophage response toward infected cells and not towards free virus, the

infected cells were cultured for less than 12 hours to prevent the lytic burst of tumor cells

and the infection of microglia/macrophages. Compared to infected glioma cells alone,

culturing infected cells with microglia or macrophages reduced viral titers by 37.28% and

69.99%, respectively [P<0.01; P<0.001] [Figure 16B-C]. This decrease in virus

replication was also accompanied by significant phenotypic changes 12 hours post

infection. Uninfected glioma cells were adherent with extensive filopodia. Following

infection, these cells became rounded but remained adherent [Figure 16D]. Similarly,

uninfected glioma cultured with microglia/macrophages revealed no significant changes

in morphology. Interestingly, when infected glioma cells were cultured with macrophages

or microglia, the microglia/macrophages surrounded the infected tumor cells and formed

tight rosette-like clusters that became non-adherent [Figure 16D; Figure 17A-B].

Macrophage and Microglia Secreted TNFα inhibits virus replication

Culturing infected glioma cells with microglia or macrophages significantly decreased virus replication, but how these cells reduced virus propagation remained to be

elucidated. TNFα is a pleiotropic cytokine whose expression is significantly up-regulated

in response to viral CNS infections [196, 206]. To test if TNFα produced by activated macrophages/microglia could limit viral replication in glioma cells, we determined the levels of TNFα secreted by microglia and macrophages in our co-culture system utilizing a species specific ELISA [Figure 18]. Using a murine specific TNFα ELISA, we

observed microglia and macrophages produced significant amounts of TNFα in response

to infected tumor cells. Compared to uninfected co-cultures, oHSV infection increased

macrophage and microglia secreted TNFα by 35.42 and 9.00 fold, respectively [P<0.001;

P<0.001] [Figure 19A-B]. Interestingly, we observed a 57.11% and 33.66% decrease in

macrophage and microglia secreted TNFα when these cells were cultured with uninfected

tumor cells compared to being cultured alone, respectively [Figure 19A-B]. In support of this in vitro data, we also observed a significant increase in murine secreted TNFα in the brain and serum following oHSV treatment in intracranial xenografts [P<0.001; P<0.01, respectively] [Figure 20A-B]. This result is concordant with previously published work demonstrating macrophages and microglia produce large amounts of TNFα in response to

HSV infection in the CNS [206]. Next, we determined if the levels of TNFα produced in these co-cultures were sufficient to reduce virus replication in glioma cells. Treatment of infected cells with 1000 pg/mL or 2000 pg/mL of recombinant human TNFα resulted in

34.29% and 40.73% reductions in virus replication, respectively [P<0.05; P<0.01]

[Figure 19C]. Similar results were obtained when infected glioma cells were treated with recombinant murine TNFα [Figure 21A]. Visual inspection of infected cells treated with soluble TNFα also revealed surprising morphological changes. oHSV infected glioma cells treated with TNFα became rounded and non-adherent. The cells resembled infected cultures with microglia/macrophages [Figure 19D, Figure 16D and Figure 21B]. We did not observe any morphological changes or reductions in cell viability in multiple uninfected glioma cell lines treated with TNFα [Figure 22A-B]. Additional experiments with uninfected glioma cells treated with varying doses of TNFα for 60 hours also did not reduce cell proliferation [P<0.001 for all doses] [Figure 22C].

Secreted TNFα induces apoptosis in oHSV infected cells

Macrophage and microglia secreted TNFα significantly reduced virus replication in

oHSV infected cells, but the mechanism of TNFα directed virus inhibition remained to be

determined. High magnification images of oHSV infected cells treated with TNFα

revealed significant changes in cell morphology. Unlike oHSV infected cells alone, the

addition of TNFα resulted in significant membrane blebbing [white arrows], cell shrinkage, and a loss of adherence, all features characteristic of cells undergoing apoptosis [Figure 19D]. Based on these observations, we hypothesized TNFα induced apoptosis in infected cells resulting in reduced virus titers. To investigate if TNFα was inducing apoptosis in infected cells, we conducted immunoblot assays for caspase 8, cleaved caspase 3, and cleaved PARP. We observed significant caspase 8, caspase 3, and

PARP activation in cells treated with oHSV and TNFα. We did not observe significant

activation of these proteins in glioma cells treated with TNFα or oHSV alone [Figure

19E; Figure 23].

Inhibition of macrophage or microglia secreted TNFα increases oHSV replication in

vitro

Since macrophage and microglia secreted TNFα reduced virus replication by inducing

apoptosis in infected cells, we hypothesized the inhibition of macrophage/microglia

produced TNFα would significantly improve virus replication. In order to determine if

inhibiting macrophage TNFα was sufficient to rescue virus replication, we conducted co- culture assays with freshly isolated wild type or TNFα knock out [TNFα-/-] bone marrow

72 derived macrophages [BMDM]. Phase contrast microscopy of these cultures revealed significant morphological differences between the two groups. Consistent with our previous results, a majority of the infected glioma cells cultured with wild type BMDMs formed non-adherent clusters and exhibited significant membrane blebbing indicative of apoptosis. In contrast to these observations, infected glioma cells cultured with TNFα-/-

BMDMs were adherent and showed substantially less membrane blebbing [Figure 24A].

These observations correlated with changes in virus titers; culturing infected glioma cells with TNFα-/- BMDMs significantly rescued virus replication compared wild type

BMDMs [P<0.05] [Figure 24B]. In a similar experiment, we found the addition of murine specific TNFα blocking antibodies rescued the reduction in virus replication in infected glioma cells when cultured with BV2 microglia [P<0.05] [Figure 24C-D].

Inhibition of TNFα increases virus replication in vivo

Blockade or knockout of macrophage/microglia secreted TNFα significantly enhanced oHSV replication in vitro. In order to assess the translational relevance of these results for oHSV therapy, we tested if TNFα blockade could enhance virus replication in vivo. In these experiments athymic nude mice were implanted subcutaneously with U87ΔEGFR human GB tumors. When the tumors reached an average volume of 143 mm3 the mice were treated with a single dose of oncolytic virus. Mice were also administered a murine specific TNFα blocking antibody or a control antibody. Mice were given antibody 1 day prior to virus injection, the day of virus administration, and on days 1, 3, and 5 post virus treatment. In these studies, a luciferase expressing oHSV was utilized to visualize virus

replication. We observed a significant enhancement in virus propagation in vivo [as

measured by luciferase encoded by virus] in mice treated with a TNFα blocking antibody

as compared to a control IgG antibody on days 1, 2, and 3 following oHSV

administration [n=5/group] [P<0.02; P<0.01; P<0.02, respectively] [Figure 25A-B].

These results suggested the inhibition of TNFα produced by macrophages and the tumor

microenvironment was sufficient to increase virus replication in vivo.

Finally, we conducted intracranial GB studies to determine if TNFα blockade could

enhance the survival of mice treated with oHSV. In these studies, mice were implanted

intracranially with U87ΔEGFR human GB cells and treated with oHSV 8 days later

[2x105 pfu rHSVQ1-luc]. The antibody dosing regimen from the subcutaneous tumor

experiments was utilized in this study. Mice treated with oHSV and a murine specific

TNFα blocking antibody lived significantly than those treated with oHSV and an isotype

control antibody [P=0.026], TNFα blocking antibody alone [P=0.0003], or with an

isotype control antibody alone [P=0.0003] [Figure 25C]. These results suggested the

combination of TNFα blocking antibodies may enhance oHSV therapeutic efficacy for

GB.

Discussion oHSV therapy is a promising treatment modality for GB. The success of oHSV derived therapeutics depends on both the oncolytic destruction of tumor cells and the activation of long-term, anti-tumor immune responses. While the innate immune response is important for activating adaptive responses, the innate responses to oHSV therapy can also inhibit virus replication and oncolytic tumor cell killing. Depletion of macrophages and microglia with clodronate liposomes and CPA has previously been shown to reduce antiviral responses and improve oncolytic virus efficacy for GB [160, 171, 191, 197-201,

207, 208]. The combination of oncolytic measles virus with CPA is currently being evaluated in a phase I clinical trial for multiple myeloma [ClinicalTrials.gov Identifier:

NCT00450814].

Recently, NK cells were shown to help coordinate the innate immune response to oHSV therapy and the depletion of these cells was found to enhance oncolytic virus [OV] efficacy for GB [161]. Neutrophils have also been shown to limit OV dissemination in part through the release of neutrophil extracellular traps [209]. Collectively, these studies suggest modulating early innate immune responses to achieve the optimal balance between viral replication and inflammation is critical to the clinical success of oHSV therapies.

While microglia and infiltrating macrophages are thought to be the primary mediators of the innate immune response to oHSV infection for GB, the mechanism by which these

cells limit virus replication and therapeutic efficacy has not been well studied [160].

Here, we quantified the extent of microglia/macrophage activation and infiltration

following oHSV treatment. While microglia are the resident immune cells of the CNS, in

this study we observed infiltrating macrophages outnumbered microglia more than 2:1 in

the tumor microenvironment following oHSV infection. These results suggested

monocyte derived macrophages may be the dominant cell type that controls oHSV

infection.

While infiltrating macrophages primarily increased in the tumor bearing hemisphere,

there was also significant activation and infiltration of immune cells in the contra-lateral hemisphere. These results suggested oHSV infection induced a global inflammatory response in the CNS rather than a localized immune response confined to the tumor.

Activation signals such as TNFα are propagated throughout the CNS in response to inflammatory stimuli. While this study focuses on the anti-viral effects of TNFα, in response to virus infection many signals such as IL-1β, IL-6, interferons, and nitric oxide are released in order to control oHSV infection [193]. These pro-inflammatory mediators signal in an autocrine and paracrine manner to activate immune cells such as macrophages and enhance their ability to respond to viral infection. oHSV associated inflammation in the non-tumor bearing hemisphere and surrounding healthy brain parenchyma has not been well studied. These observations may have implications in the treatment of brain tumor patients with oHSVs where unchecked inflammation can be detrimental.

In these studies, we observed a significant inflammatory response to viral infection until

at least 3 days post treatment. In vivo flow cytometry experiments indicated microglia

and infiltrating macrophages were polarized towards an M1, pro-inflammatory state. We

demonstrated the anti-viral consequence of microglia and macrophage activation in co-

culture studies and found both macrophages and microglia reduced virus replication in

glioma cells. Together these results confirmed the anti-viral capabilities of these cell

types in modulating oHSV replication in vivo. This data also supports previous studies

that identify the anti-viral activity of macrophages and microglia against wild-type

Herpes Simplex Virus 1 [HSV-1] infections [210, 211].

In this study, we identified TNFα as a major macrophage/microglia secreted factor which

reduces oHSV replication. TNFα is a pleotropic cytokine important for the recruitment

and activation of immune cells. Macrophages and microglia are also major producers of

TNFα. TNFα is known to limit wild-type HSV replication in the CNS, and it has previously been shown to mediate anti-viral effects in studies with wild-type vesicular stomatitis virus, adenovirus-2, encephalomyocarditis virus, HSV-1, HSV-2, respiratory syncytial virus, and influenza through a variety of mechanisms [206, 212-219].

While TNFα is detrimental to virus replication, TNFα signaling in cancer cells can result in increased tumor cell growth, angiogenesis, invasion, and progression [220-222].

Higher levels of the anti-apoptotic proteins Bcl-2, Bcl-xL, and Mcl-1 as well as decreased

levels of apoptotic proteins such as BAX are commonly observed in recurrent GB and

demonstrate the ability of these tumors to resist caspase mediated cell death [93].

Consistent with these published studies, we observed TNFα was not toxic to uninfected

GB cells in vitro. In infected glioma cells, however, we observed TNFα activated the

extrinsic apoptotic pathway resulting in premature cell death, reduced virus replication,

and decreased anti-tumor efficacy. The precise mechanism of how the combination TNFα

with oHSV induces apoptosis is unclear. While HSV-1 has been shown to inhibit apoptosis, previous work has demonstrated the inability of HSV-1 to prevent apoptosis in infected cells exposed to environmental stimuli such as TNFα [223]. TNFα induced cell death was found to be cell-type dependent, and in the case of glioma this process may depend on the expression of pro- and anti-apoptotic proteins within the cells.

Oncolytic HSVs expressing TNFα have previously been tested for their ability to enhance oHSV anti-tumor efficacy [224]. In these studies, TNFα expressing viruses did not enhance anti-tumor efficacy in an immune-competent lymphoma model compared to a control oHSV that did not express TNFα. Additionally, in human squamous carcinoma xenografts, the antitumor efficacy of an oHSV expressing high levels of TNFα was significantly less than an oHSV expressing low levels of TNFα. In support of our findings, these results suggest TNFα elicits strong antiviral responses that may be detrimental to oncolytic HSV therapy.

While TNFα blockade lead to increased virus propagation, its effect on toxicity in the

context of HSV-1 infections is not clear. Both virus-mediated and immune-mediated

mechanisms contribute towards the pathology of HSV-1 infections. In studies with mice infected with wild-type HSV-1, TNFα knockout mice had higher virus titers and were more susceptible to fatal HSV encephalitis than wild type mice. These results highlight the protective, anti-viral functions of TNFα [206, 212]. While TNFα is important for controlling virus replication, high levels of TNFα have also been shown to induce blood brain barrier disruption leading to increased inflammation [225]. Interestingly, HSV-1

infected mice treated with TNFα blocking antibody showed reduced signs of viral

encephalitis and lived longer than those treated with virus alone [226]. Thus, a transient

blockade of TNFα during virotherapy could increase virus replication and reduce

neurotoxicity due to acute inflammation while still allowing for an immune response to

eventually clear the infection. Importantly, in our studies we observed no toxicity

associated with TNFα antibody administration in combination with our attenuated,

oncolytic virus.

Radiation and chemotherapy also induce the production of cytokines such as TNFα [116,

135, 227]. Oncolytic virotherapy for GB is often administered following tumor resection

and concurrently with radiation and chemotherapy. As a result, patients may benefit from

the transient use of TNFα inhibitors prior to oHSV administration in order to enhance

oncolytic tumor cell killing and reduce CNS inflammation. The TNFα inhibitors

etanercept, adalimumab, certolizumab, and golimumab are currently FDA approved for a

79 variety of diseases and could be readily utilized in oHSV clinical trials. These inhibitors may be more effective than general immune suppressants such as high dose myeloablative CPA that can have significant toxicities in patients. The combination of oHSV with TNFα inhibitors could enhance virus replication, reduce TNFα driven tumor proliferation, angiogenesis, and invasion, as well counter the negative effects of chemotherapy/radiation induced inflammation. This transient inhibition of TNFα could then be removed to allow for the activation of long-term, anti-tumor immune responses that may be more potent due to increased virus mediated cell killing and antigen release.

In subcutaneous and intracranial tumor studies, we found the inhibition of TNFα secreted by the tumor microenvironment significantly enhanced virus replication and therapeutic efficacy. In these experiments we utilized a TNFα blocking antibody because the current

FDA approved TNFα inhibitors are antibody based. While the integrity of the blood- brain-barrier is disrupted in glioblastoma, the ability of therapeutic antibodies to cross the blood-tumor-barrier [BTB] is thought to be limited [228, 229]. While we observed up to

9 fold increases in viral luciferase expression in subcutaneous tumors, the therapeutic effect in the intracranial tumor studies was more modest. In addition to antibody penetration into the brain tumor microenvironment following oHSV therapy, we hypothesize the increase in animal survival may have also been through the ability of the antibody to bind TNFα in the serum following oHSV therapy. The future development of specific, soluble TNFα inhibitors that better penetrate the BTB may further increase the anti-tumor efficacy we observed. These experiments support the future use of TNFα inhibitors in combination with oHSV for GB.

Reproduced with the permission of AACR:

Meisen WH, Wohleb E, Jaime Ramirez AC, Bolyard C, Yoo JY, Russell L, Hardcastle J,

Dubin S, Godbout J, Kaur B. The impact of macrophage and microglia secreted TNFα on oncolytic HSV-1 therapy in the glioblastoma tumor microenvironment. Accepted Clinical

Cancer Research. (2015). PMID: 25829396 [E-published ahead of print].

Future Directions: Vstat120 expressing oHSVs

The results from this project identified microglia and macrophage secreted TNFα as a

major factor that reduces oHSV efficacy for GB [230]. These findings suggested agents

which transiently reduce macrophage/microglia recruitment and/or TNFα secretion may

improve oHSV efficacy for brain tumors. Our lab previously developed a novel OV,

RAMBO, which expresses the anti-angiogenic fragment Vasculostatin (Vstat120). We

demonstrated mice with intracranial GB tumors treated with the RAMBO virus live

significantly longer than mice treated with a parent virus lacking Vstat120 expression

[rHSVQ1] [155]. Anti-angiogenic agents have previously been shown to reduce the

recruitment of macrophages and microglia to oHSV treated GB tumors [171, 231]. We

speculated anti-angiogenic Vstat120, secreted by the RAMBO virus, may also be

impacting immune responses to virally treated tumors. The primary objective of this

preliminary study is to investigate the effects RAMBO directed Vstat120 expression on

microglia/macrophage recruitment/activation and OV efficacy.

Results

Macrophage recruitment and activation is reduced in RAMBO treated tumors

We first examined how each OV affected macrophage recruitment to treated GB tumors.

In these experiments, mice implanted with U87ΔEGFR intracranial tumors were treated

with PBS, rHSVQ1, or RABMO. Three days post OV treatment, the mice were

euthanized and the tumor and non-tumor bearing hemispheres were separated via gross dissection [Figure 26A]. We examined the infiltration of macrophages into the tumor

microenvironment following OV therapy. Similar to our previously observed results, the

treatment of intracranial tumors with rHSVQ1 resulted in significant monocyte

infiltration into the tumor (p<0.05) [Figure 26B-C]. Surprisingly, this infiltration was reduced 4.5 fold in tumors treated with the RAMBO virus as compared to rHSVQ1

(p<0.05) [Figure 26B-C]. Infiltrating monocytes from RAMBO treated tumors also had

significantly less MHCII, Ly6C, and CD86 than rHSVQ1 treated tumors (p<0.05)

[Figure 27C]. The microglia of RAMBO treated tumors had significantly less MHCII and

CD206 expression than those treated with rHSVQ1 (p< 0.05) [Figure 27A-B]. These

results suggested Vstat120 expressed by the RAMBO virus may be mediating anti-

angiogenic independent effects on microglia and macrophages which temper their

responses to viral infection.

The RAMBO virus replicates better than a control virus in vivo and in co-cultures with

microglia

To further examine the interaction of Vstat120 with microglia/macrophages we

developed a co-culture system with murine microglia and infected human glioma cells

[Figure 28A]. In plaque forming unit [PFU] assays with U251-T2 glioma cells and

microglia, we observed a 2.6 fold increase in infectious virus in RAMBO treated co-

cultures as compared to rHSVQ1 co-cultures (p<0.01) [Figure 28B]. There was no

difference in replication between RAMBO and rHSVQ1 in glioma cells or microglia

alone [Figure 28C]. We also observed these differences in virus replication in vivo. Mice

with U87∆EGFR intracranial tumors were treated with 1 x 105 pfu RAMBO or rHSVQ1.

The treated animals were sacrificed at various time points and virus replication was

examined by quantitative PCR for viral ICP4 gene expression. Higher levels of viral

ICP4 were observed in RAMBO treated tumor on days 3 and 7 [p<0.05] [Figure 28D].

No virus was detected in tumors by day 13.

Microglia in RAMBO treated co-cultures produce less TNFα

To explore the mechanism by which RAMBO enhanced virus replication, we examined microglia and macrophage cytokine production in the co-cultures. X12v2 glioma cells were treated with PBS, rHSVQ1, or RAMBO and co-cultured with microglia [Figure

29A]. We observed a 75-fold induction of TNFα expression in microglia cultured with rHSVQ1 infected glioma cells compared to untreated microglia [p<0.05]. Interestingly, we observed a 3.4-fold reduction in microglia TNF-α expression in RAMBO treated co- cultures compared to rHSVQ1 co-cultures [p<0.01] [Figure 29B]. These results corresponded with ELISA data where we observed a 6.9 fold decrease in microglia TNFα secretion in RAMBO treated co-cultures [p<0.01] [Figure 29C]. Similar results were in co-cultures with infected U87∆EGFR glioma cells [Figure 29A-C]. Treating these co- cultures with a microglia (murine) specific TNFα blocking antibody rescued the differences in viral replication between rHSVQ1 and RAMBO co-cultures [Figure 30].

These results suggest the RAMBO virus modulates microglia/macrophage anti-viral activity by reducing TNFα.

Conclusions

These results suggest Vstat120 enhances OV replication by reducing

microglia/macrophage secreted TNFα. Since the ability of TNFα blocking antibodies to cross the blood-tumor-barrier [BTB] is limited, the RAMBO virus expressing Vstat120 may be an appealing alternative [228, 229]. In addition to Vstat120 anti-angiogenic functions, this preliminary data suggests Vstat120 a novel role in modulating macrophage and microglia immune responses. Future work will identify the mechanism by which

Vstat120 mediates these effects on microglia and macrophages.

Materials and Methods

Cell Lines

Vero, LN229, U87ΔEGFR, U251-T2, and U251-T3-mCherry cells were maintained in

DMEM supplemented with 10% fetal bovine serum [FBS]. U251-T2 and U251-T3-

mCherry cells were created in our lab [May 2009] as tumorigenic clones of U251 cells by

serially passaging these cells two and three times in mice, respectively. Monkey kidney

epithelial derived Vero cells and U87ΔEGFR cells were obtained in April 2005 from Dr.

E Antonio Chiocca [Ohio State University, Columbus, Ohio]. LN229 cells were obtained

in January 2005 from Erwin Van Meir [Emory University, Atlanta, Georgia]. GB30

neurospheres were originally received in 2012 from Dr. EA Chiocca [Ohio State

University, Columbus, OH]. GB30 neurospheres were maintained as tumor spheres in

Neurobasal Medium supplemented with 2% B27, human EGF [50 ng/ml], and bFGF [50

ng/ml] in low-attachment cell culture flasks as previously described [166]. Vero cells

have not been authenticated since receipt. U87ΔEGFR [January 2015], LN229 [July

2013], GB30 [January 2015], and U251 [January 2015] cells were authenticated by the

University of Arizona Genetics Core via STR profiling. Murine BV2 microglia were

maintained in DMEM supplemented with 2% FBS. BV2 cells were obtained in January

2009 from J. Godbout [Ohio State University, Columbus, Ohio]. Murine RAW264.7

macrophages were obtained in RPMI supplemented with 10% FBS. RAW264.7

macrophages were received in June 2010 from S. Tridandapani [Ohio State University,

Columbus, Ohio]. Murine BV2 and RAW264.7 cells have not been authenticated since

receipt. All cells were incubated at 37oC in an atmosphere with 5% carbon dioxide and

86 maintained with 100 units of penicillin/mL, and 0.1 mg of streptomycin/mL [Penn/Strep].

All cells are routinely monitored for changes in morphology and growth rate. All cells are negative for mycoplasma.

Viruses and virus replication assays

rHSVQ1, rHSVQ1-Luciferase, 1716, hrR3 and rQNestin34.5 were prepared and titered on Vero cells via a standard plaque forming unit assay as previously described.[156]

Co-culture Assays

550,000 glioma cells were plated in 6 well Falcon tissue culture plates and infected with virus at a multiplicity of infection [MOI] of 1 or 2 in DMEM supplemented with 0.05%

FBS. Cells were washed 3 times over the course of an hour to remove unbound virus.

Infected cells were then overlaid with 1,000,000 microglia or macrophages [2:1 ratio of macrophages/microglia to glioma cells] for 12 hours [pre-virus burst]. For TNFα blocking antibody assays, 1800 ng/ml of mouse TNFα neutralizing antibody [D2H4; Cell

Signaling] or an isotype control was utilized. Concentrations of antibody were determined experimentally based on manufacturer’s specifications.

Western Blot

Cells were cultured with virus, TNFα, and/or microglia/macrophages as described above.

BCA analysis [Pierce Biotechnology, Rockford, IL] was used to determine protein concentration. Equal amounts of protein were separated on a 4-20% Tris-HCL gel and

87 transferred to a PVDF membrane. Caspase 8, cleaved Caspase 3, Cleaved PARP, and

GAPDH [Cell signaling] were used at 1:1000 except GAPDH [1:5000] which was used as a loading control.

Microglia and macrophage antibody staining

Staining of surface antigens were performed as previously described.[232, 233] Briefly,

Fc receptors were blocked with anti-CD16/CD32 antibody [eBioscience, San Diego,

CA]. Cells were then incubated with the appropriate antibodies: CD45, CD11b, MHCII,

CD86, LY6C, LY6G, and CD160 [eBioscience, San Diego, CA] for 45 minutes. Cells were re-suspended in FACS buffer [2% FBS in HBSS with 1 mg/ml sodium azide] for analysis. Non-specific binding was assessed via isotype-matched antibodies. Antigen expression was determined using a Becton-Dickinson FACS Caliber four color cytometer. Ten thousand events were recorded for each sample and isotype matched- conjugate. Data was analyzed using FlowJo software [Tree Star, CA].

Image Acquisition

Fluorescent and bright field images were acquired using an Olympus IX81 epifluorescence microscope equipped with a QImaging Retiga 2000R FAST camera and

Olympus objectives. Image-Pro software [Version 6.2] was used for image acquisition.

For luciferase imaging, mice received an intraperitoneal [IP] injection of luciferin

[Caliper Life Sciences] and the luciferase signal was visualized/quantified utilizing a

IVIS Lumina II imaging system.

Animal surgery

All animal experiments were performed in accordance with the Subcommittee on

Research Animal Care of The Ohio State University guidelines and were approved by the

institutional review board. 6-8-week-old, female athymic nude mice [NCI], were used for in vivo tumor studies. Intracranial surgeries were performed as previously described with stereotactic implantation of 100,000 U87ΔEGFR [156]. Tumors were treated with HBSS, rQNestin34.5, or rHSVQ1-Luciferase virus at the location of tumor implantation. For antibody studies, mice were treated via IP injection at the days indicated with 400ug of anti-murine TNFα antibody [XT3.11] or isotype control antibody [BE0094; BE0088] from BioXCell. Animals were euthanized when they showed signs of morbidity.

Cell Viability Assay

Cells were plated in 96 well plates with 2% FBS in DMEM with varying concentrations

of TNFα [Human (Gibco); Mouse (affymetrix eBioscience)]. Viability was assessed as

described previously using a standard MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide] assay.[185]

Isolation of microglia and macrophages

Microglia and macrophage populations were isolated for flow cytometry analysis from

murine brain homogenates as previously described.[234, 235] Briefly, brains were split

into the tumor and non-tumor bearing hemispheres and then passed through a 70 µm

nylon cell strainer. Resulting homogenates were centrifuged, the supernatants removed,

and the cell pellets resuspended in 70% isotonic Percoll [GE-healthcare, Uppsala,

Sweden]. A discontinuous Percoll density gradient was layered as follows: 70%, 50%,

35%, and 0%. The gradient was centrifuged and the microglia/macrophages populations

were collected from the interphase between the 70% and 50% Percoll layers.[236, 237]

Murine Bone Marrow Macrophage Generation

Bone marrow derived macrophages were isolated as previously described.[238] Briefly, the tibia and femurs of euthanized mice were flushed with PBS several times to remove bone marrow cells. Cells were centrifuged and plated in RPMI supplemented with 10%

FBS and 1% Penn/Strep. 20 ng/mL murine macrophage colony stimulating factor [R&D] and 10 ug/mL of polymyxin B [Calbiochem] were added to the cultures and the cells were allowed to mature for 8 days. TNFα-/- and strain control mice were obtained from the Jackson Laboratory [Ben Harbor, Maine].

Statistical Analysis

Student's t-test or one-way ANOVA with Bonferroni multiple comparision post hoc tests were used to analyze changes in cell killing, viral plaque forming assays, luciferase imaging experiments, and flow cytometry assays. In survival assays, Kaplan–Meier curves were plotted and the log rank test was utilized to determine statistical significance.

All statistical analyses were performed with the use of Graph Pad Prism software

[version 5.01]. A P<0.05 was considered statistically significant. Derived P values are identified as *P<0.05; **P<0.01; ***P<0.001.

Figures and Tables

Figure 13: OV treatment increases microglia activation and induces macrophage

infiltration into the tumor microenvironment.

A. Diagram of mice with intracranial U87ΔEGFR tumors [red dot] treated with 1 x 105 pfu of OV [rQNestin34.5] or PBS 7 days post tumor cell implantation. 3 Days post OV treatment, the mice were euthanized and the tumor and non-tumor bearing hemispheres were separated via gross dissection [midline drawn between two hemispheres]. B. Quantification of MHCII expression of the tumor bearing hemispheres of PBS and OV treated mice. Data shown is mean percent positive ± SD [n= 5/group]. C. Representative scatter plot of MHCII expression on microglia [CD11b+CD45lo] in tumor and non-tumor bearing mice treated with PBS or OV. D. Quantification of macrophage infiltration into the tumor bearing hemisphere following OV therapy or PBS injection. Data shown are mean percent positive ± SD [n=5/group; P<0.001]. E. Representative scatter plot of macrophage [CD11b+CD45hi] infiltration following OV therapy.

Figure 14: Percoll gradient isolated CD11b+CD45hi cells are predominantly infiltrating macrophages.

Percoll gradient isolated CD11b+CD45hi cells are predominantly infiltrating macrophages. Cell numbers of neutrophils [CD11b+CD45hiLY6G+], NK cells [CD11b+CD45hiCD106], or infiltrating macrophages [CD11bCD45hi] in the tumor bearing hemispheres treated with OV or PBS. Data shown is mean cell number ± SD.

Figure 15: OV treatment increases macrophage activation.

A-C. Left- Representative scatter plots of CD86+ [A], LY6C+ [B], and MHCII+ [C] macrophages in the tumor and non-tumor bearing hemispheres following OV [rQNestin34.5] or PBS treatment. Right- Quantification of the percentage and cell numbers of macrophages expressing the indicated marker following OV or PBS treatment in the tumor and non-tumor bearing hemisphere [ n=5/group]. Quantified data is mean values ± SD.

Figure 16: Microglia and macrophages reduce virus replication in tumor cells in vitro.

A. Schematic of microglia/macrophage and tumor cell co-cultures. Tumor cells [yellow cells] were infected with OV [rHSVQ1] [red dots] at a MOI of 2. Unbound virus was washed away and microglia or macrophages [blue cells] were overlaid on the infected glioma cells. The cells were cultured for 12 hours and the viral titers were determined by a standard plaque formation assay. B-C Viral titers of glioma cells infected alone or cultured with BV2 microglia [B.] or RAW264.7 macrophages [C.]. Data shown is mean virus titer ± SD. D. Images of glioma cells cultured with microglia and macrophages with and without OV infection 12 hours post infection.

Figure 17: Images of uninfected and infected co-cultures 12 hours post infection.

U251-T3-mCherry glioma cells were infected with oHSV [hrR3] at a MOI of 2. Unbound virus was washed away and microglia [green labeled cells] were overlaid on the infected glioma cells [red labeled cells]. Representative GFP, RFP, brightfield, and merged images of uninfected [A.] and infected [B.] co-cultures are shown 12 hours post infection.

Figure 18: Species specificity of murine TNFα ELISA .

Species specificity of murine TNFα ELISA kit for murine (filled squares) and not human (circles) TNFα. Data shown is mean concentration.

Figure 19: Microglia and macrophage secreted TNFα inhibits virus [rHSVQ1] replication in vitro.

A-B. Quantification of TNFα secreted by BV2 microglia [A.] or RAW264.7 macrophages [B.] alone, and when cultured with uninfected or infected U251-T2 glioma cells for 12 hours. Data shown is mean concentration TNFα ± SD. C. Viral titers of glioma cells infected at an MOI of 2 alone, with recombinant human TNFα [1000 or 2000 pg/mL], or with BV2 microglia for 12 hours. Data shown is mean virus titer ± SD D. Representative images of U251-T2 glioma cells infected with OV [rHSVQ1] at an MOI of 2 with vehicle [Left] or with TNFα [2000 pg/mL] [Right] for 12 hours. White arrows indicated membrane blebbing in infected cells treated with TNFα. E. Western blot of U251-T2 glioma cells alone, treated with TNFα [5000 pg/mL], infected with OV at an MOI of 2, or treated with TNFα and OV for 12 hours. Infected U251-T2 glioma cells cultured with BV2 microglia is also shown. Caspase 8, cleaved Caspase 3, Cleaved PARP, and GAPDH are shown. Caspase 8 blot shows full length protein [1], cleaved intermediate protein [2], and active protein [3].

Figure 20:oHSV treatment increases secreted TNFα in vivo

A. Nude mice bearing U87ΔEGFR intracranial tumors were treated intratumorally with PBS or 2 x 105 pfu of oHSV [rHSVQ1]. 3 days following virus administration the mice were euthanized and their brains were removed. The brains were homogenized in 1.5 mLs DMEM and the supernatants were analyzed a murine specific TNFα ELISA [PBS=3; rHSVQ1=5]. B. Nude mice bearing GB30 intracranial tumors were treated intratumorally with PBS or 1 x 104 pfu of oHSV [1716]. 3 days following virus administration the mice were euthanized and blood collected by cardiac puncture. The serum was separated and analyzed on a murine specific TNFα ELISA [PBS=4; 1716=4].

Figure 21: Murine TNFα reduces viral replication in infected human glioma cells.

A. Viral titers of U251-T2 glioma cells infected at an MOI of 1 with varying doses of murine TNFα for 12 hours. Infected U251-T2 glioma cells cultured with BV2 microglia was used as a positive control. Data shown is mean viral titer ± SD. B. Representative images of OV infected cells cultured with TNFα or BV2 microglia.

100

Figure 22: TNFα increases cell proliferation and is not cytotoxic to uninfected GB cells.

A. Representative images of U251-T2 glioma cells alone or with TNFα [1000 or 5000 pg/mL] for 12 hours. B. Cell viability of U87ΔEGFR, LN229, or U251-T2 GB cells treated with varying concentrations of TNFα for 12 hours. C. U251-T2 gloma cells were treated with 3 different concentrations of TNFα for 60 hours. Cell viability was examined at 12, 36, and 60 hours. Data shown is mean cell viability ± SD normalized to untreated cells at each time point.

101

Figure 23: TNFα induces apoptosis in OV infected glioma cells.

Western blot of U251-T2 glioma cells alone, treated with TNFα at 2000 pg/mL, infected with OV at an MOI of 2, or treated with TNFα [1000 and 2000 pg/mL] and OV for 12 hours. Infected U251-T2 glioma cells cultured with BV2 microglia was used as a positive control. Caspase 8, Cleaved PARP, and GAPDH are shown. Caspase 8 blot shows full length protein [1], cleaved intermediate protein [2], and active protein [3].

102

Figure 24: Inhibition of microglia/macrophage secreted TNFα increases virus [rHSVQ1]

replication in vitro.

A. Representative images of U251-T2 glioma cells infected at an MOI of 2 cultured with bone marrow derived macrophages derived from wild-type or TNFα knockout mice for 12 hours. Large representative images are taken at a 4x magnification with the insets taken at a 20x magnification (white arrows indicate blebbing). B. 12 hour viral titers of cultures described in A. Data shown is mean virus titer ± SD C. Schematic of experimental setup utilizing murine specific TNFα antibodies to block microglia [blue cells] secreted TNFα in co-cultures with infected glioma cells [yellow cells]. D. Quantification of virus titer obtained from infected glioma cells cultured with BV2 microglia with IgG or anti-murine TNFα blocking antibody [1800 ng/ml]. Data shown is mean virus titer ± SD.

103

Figure 25: TNFα inhibition increases virus replication and efficacy in vivo.

Nude mice with U87ΔEGFR subcutaneous tumors were treated with 1 x 106 pfu of an OV expressing luciferase [rHSVQ1-luc]. Murine specific TNFα or isotype control antibodies were administered on days -1, 0, 1, 3, and 5 post OV therapy. A. Data shown are quantification of luciferase gene activity in U87ΔEGFR subcutaneous tumors treated with control or TNFα blocking antibodies on the days indicated after virus treatment. Data shown is total flux in each mouse [n=5/group]. B. Representative luciferase images of OV treated mice with TNFα blocking or isotype control antibodies at the days indicated [n=5/group]. C. Kaplan-Meier survival curve of mice bearing U87ΔEGFR intracranial tumors treated with PBS or 2x105 pfu rHSVQ1 with IgG or TNFα blocking antibody [IgG + Saline n=10; anti-TNFα + Saline n=11; IgG +rHSVQ1 n= 14; anti- TNFα + rHSVQ1 n=15]. 104

Figure 26: RAMBO reduces monocyte infiltration in OV intracranial tumors.

A. Diagram of mice with intracranial U87ΔEGFR tumors [red dot] treated with 1 x 105 pfu of OV [rHSVQ1] or PBS 7 days post tumor cell implantation. 3 Days post OV treatment, the mice were euthanized and the tumor and non-tumor bearing hemispheres were separated via gross dissection [midline drawn between two hemispheres]. B. Monocytes (CD11b+hi CD45+hi) were isolated from the brains of OV treated mice and examined for the infiltration of monocytes. Representative scatter plots shown. C. Quantitative analyses of flow cytometry data examining monocyte infiltration in vivo following OV treatment (n=4/group).

105

Figure 27: RAMBO Reduces Macrophage and Microglia Activation in treated intracranial tumors.

A. Microglia [CD11b+hi CD45+int/lo] were isolated from the brains of OV treated mice and examined for the expression of activation markers. Representative scatter plot shown. B. Quantitative analyses of flow cytometry data examining microglia activation in vivo following OV treatment [n=4/group]. C. Quantitative analyses of flow cytometry data examining monocyte activation in vivo following OV treatment [n=4/group].

106

Figure 28: RAMBO Replicates Better than a control virus [rHSVQ1] In Vivo and in cultures with microglia.

A. Schematic of microglia and tumor cell co-cultures. Tumor cells [yellow cells] were infected with OV [rHSVQ1] [red dots]. Unbound virus was washed away and microglia or macrophages [blue cells] were overlaid on the infected glioma cells. The cells were cultured for 12 hours and the viral titers were determined by a standard plaque formation assay. B. Viral titers of infected U251-T2 glioma cells cultured with BV2 microglia C. Viral titers of glioma cells or microglia infected alone. D. Mice with U87∆EGFR intracranial tumors were treated with 1 x 105 pfu RAMBO or a control virus (rHSVQ1). The treated animals were sacrificed on the days indicated post-OV therapy. RNA was isolated from the whole tumors [n=4/group] and QPCR was performed for the viral gene ICP4. Higher levels of viral ICP4 were observed in RAMBO treated tumor on days 3 and 7 [p<0.05]. No virus was detected in tumors by day 13.

107

Figure 29: RAMBO reduces the expression and secretion of TNFα by microglia co-

cultured with infected glioma cells.

A. Schematic of microglia and tumor cell co-cultures. Tumor cells [yellow cells] were infected with OV [rHSVQ1] [red dots]. Unbound virus was washed away and microglia or macrophages [blue cells] were overlaid on the infected glioma cells. B. Infected U87ΔEGFR and X12v2 cells human glioma cells were co-cultured with BV2 murine microglia. 12 hours following the infection, the cells were harvested, and the RNA isolated for qPCR analyses. Murine primers were utilized to interrogate the expression of murine microglia cytokines. C. Infected U87ΔEGFR and X12v2 cells human glioma cells were co-cultured with BV2 murine microglia. 12 hours following the infection, the supernatants were collected to analyze TNF-α production by BV2 murine microglia.

108

Figure 30: Inhibition of TNFα Rescues Differences in Virus Replication between

RAMBO and rHSVQ1.

Glioma cells infected with rHSVQ1 or RAMBO were co-cultured with microglia in the presence of IgG or a mouse specific TNFα blocking antibody. After 12 Hours, PFU assays were conducted to examine virus replication. As expected, the RAMBO virus replicated significantly better than rHSVQ1 in the co-cultures with IgG antibody [p<0.05]. In the presence of the microglia specific TNFα blocking antibody, the differences in virus replication between rHSVQ1 and RAMBO were rescued. Viral titers are quantified in the chart on the right.

109

Chapter 4: Analysis of bioluminescent and magnetic resonance imaging modalities

identifies important strategies for monitoring Glioblastoma tumor growth in vivo

Abstract

Bioluminescent imaging [BLI] and magnetic resonance imaging [MRI] are two non- invasive imaging modalities which can significantly enhance intracranial tumor studies.

In this study we created a multi-modality imaging paradigm for analyzing changes in tumor growth and biology in order to maximize data acquisition while reducing costs and time. We evaluated BLI and MRI modalities in three different glioblastoma [GB] xenograft models. We selected the classical GB tumor model, U87ΔEGFR, as well as two GB neurosphere lines derived from primary patient tumor samples. In correlative studies, we found the relationships between BLI and MRI varied in the three GB models.

BLI and MRI output was significantly affected by tumor necrosis, hemorrhaging, tumor depth, extracranial growth, and animal positioning. Finally, we developed an imaging strategy to overcome the challenges of in vivo GB tumor studies.

110

Introduction

Glioblastoma [GB] is one of the most common and deadly types of primary brain tumors.

These aggressive tumors are characterized by widespread invasion, unregulated cellular proliferation, extensive angiogenesis, and resistance to cell death [186]. These features severely limit treatment options and result in a median patient survival of only 15 months

[239].

Despite exciting advances in preclinical research, translating basic science discoveries into clinically effective treatments remains a challenge. Among these challenges, animal models are a significant barrier to successful therapeutic research. Currently, investigators can select from a variety of spontaneous, orthotopic, immune-competent, and xenograft tumor models [240]. Testing novel treatments in multiple models is a necessity as results can vary significantly between models. Investigators must also decide on when to initiate therapy, dosing schedules, and treatment duration in animal studies.

These decisions are especially difficult in GB studies, where the skull significantly limits options for monitoring tumor growth, treatment responses, and anti-tumor efficacy.

Bioluminescent imaging [BLI] and magnetic resonance imaging [MRI] are two non- invasive imaging modalities which can significantly enhance intracranial tumor studies.

These imaging modalities allow real time measurements of tumor volume, tumor growth, and therapeutic responses. BLI studies utilize cells that stably express luciferase, which emits light when interacting with a substrate such as D-Luciferin. This methodology has

111

been utilized to examine a variety of CNS and peripheral cancers [241-244]. While BLI is a rapid, inexpensive, and sensitive imaging technique, it is also susceptible to a variety of artifacts due to animal positioning, substrate availability, as well as tissue mediated light scattering, absorption, and reflection [245]. MRI, is used clinically as the standard for determining mass diameters and volumes [246]. Like BLI, however, the information provided by MRI is limited and investigators struggle to assess metrics such as tumor cell viability, invasion, necrosis, and angiogenesis. MRI animal studies are also costly and time consuming. Imaging strategies for GB tumor studies which employ both BLI and

MRI techniques could provide more efficient data acquisition to investigators. The choice of how to best utilize the strengths of each modality at different points of tumor development is not clear.

In this study, we evaluated BLI and MRI modalities in three different GB xenograft models. The overall objective of this study was to develop a multi-modality imaging strategy for monitoring GB tumor growth rate and therapeutic responses. We selected the classical GB tumor model, U87ΔEGFR, as well as two GB neurosphere lines derived from primary patient tumor samples. Compared to traditional serum cultured cell lines,

GB neurospheres are thought to retain more of the phenotypic characteristics of human

GB tumors [247]. Few comprehensive imaging studies with neurosphere derived GB tumors have been conducted. Therefore, we selected these models in order to understand how their unique biology affects in vivo imaging. In correlative studies, we found the relationships between BLI and MRI varied in all three GB models. Additionally, we also

112 identified several strengths and weakness of each imaging for evaluating changes in tumor growth in vivo. We found the output from these imaging modalities was significantly impacted by tumor depth, necrosis, hemorrhage, animal positioning, and extracranial tumor growth. In synthesizing the data from this study, we created a multi- modality imaging paradigm for analyzing changes in tumor growth and biology while reducing cost-prohibitive and time consuming MRI.

113

Results

Patient derived GB xenografts recapitulate the biology of human tumors

Glioblastoma is characterized by high cellularity, excessive mitotic activity, nuclear

atypia, microvascular proliferation, necrosis, and invasion [56]. GB xenograft models are

commonly used to evaluate novel therapeutics, but the biology of these tumors often varies significantly from human patients. Here, we examined three murine xenograft GB models in order to determine how well these tumors recapitulated the human disease and to allow us to identify histopathological features of each model which may affect MRI and BLI. Mice were implanted intracranially with GB30 neurospheres, GB169 neurospheres, or serum-cultured U87ΔEGFR cells. The animals were sacrificed when the mice displayed signs of morbidity and tumor burden. We observed significant histological differences in the tumor biology of serum-cultured and neurosphere derived

GB tumors. U87ΔEGFR tumors grew rapidly in a large, expansive mass of tightly packed cells [Figure 31A]. The tumor borders were well demarcated from the normal brain, and we observed no tumor cell invasion [Figure 31A- black inset]. The tumors contained significant tumor vasculature, and we did not view any large areas of necrosis. GB169 tumors grew significantly more slowly than U87ΔEGFR tumors. These tumors possessed well delineated borders, and we did not observe any microscopic or interhemispheric invasion [Figure 31B and Figure 31B black inset]. While these tumors were highly vascularized, they also contained significant necrosis with pseudopalisades indicative of hypoxia [Figure 31B red inset]. GB30 neurosphere derived tumors grew rapidly at a rate similar to U87ΔEGFR tumors. Like the GB169 tumors, GB30 tumors also contained

114

pseudopalisades surrounding areas of necroses [Figure 31C and Figure 31C red inset].

These tumors were highly vascularized and we observed glomeruloid bodies present

within the tumor [Figure 31C green inset]. GB30 tumor cells also invaded into the normal

brain [Figure 31C black inset], but this invasion was limited to a few millimeters into the

surrounding parenchyma and we did not observe any tumor cells in the contralateral

hemisphere. While all three GB models possessed features characteristic of human GB

tumors, we found the neurosphere derived tumors best recapitulated the patient disease.

GB Tumor Volume Increases Exponentially by MRI

To determine how tumor volume changed with time, we examined tumor growth by MRI

in mice implanted with GB30 neurospheres, GB169 neurospheres, or serum-cultured

U87ΔEGFR cells. U87ΔEGFR cells engrafted quickly in the parenchyma and we observed tumors as early as 6 days following cell implantation [Figure 32A]. These tumors grew rapidly, and large masses in the right hemisphere were seen 14 days post tumor cell implantation [Figure 32B; Figure 34A] [n=12]. In accordance with the tumor histology, these sphere-like tumors appeared as solid mass of cells with little visible necrosis, hemorrhaging, or structural heterogeneity by MRI. The GB169 neurosphere derived tumors had a notably different appearance by MRI. The tumor shape was irregular and contained large amounts of necroses characteristic of patient GB tumors

[n=10]. GB169 tumor engraftment was not noticeable by MRI until 30-45 days after implantation. However, once identified, the small tumor foci grew rapidly and occupied a majority of the implanted hemisphere within several weeks [Figure 32C; Figure 34B].

115

This slower growing GB169 model possessed significantly more variation in tumor

volume growth compared to the U87ΔEGFR and GB30 models. The GB30 neurosphere

derived tumors grew rapidly at a rate similar to the U87ΔEGFR tumors. We observed

little variation in tumor volumes in this fast growing model [n=8] [Figure 32D; Figure

34C]. Unlike the U87ΔEGFR tumors, GB30 tumors also had significant necrosis and

structural heterogeneity which resembled the tumor histology described earlier. Also, in

accordance with fixed tumor sections, the large tumors were irregular in shape and possessed poorly defined borders reflective of invasion. While all of the tumors had different growth kinetics, especially at earlier time points, after visible tumor foci formed the tumor grew at a rapid, exponential rate.

Luciferase Signal Increases with GB Tumor Growth

In vivo luciferase signal is reflective of cell viability [245]. In these studies we measured

the in vivo photon flux in mice implanted with U87ΔEGFR, GB30, or GB169 tumor

cells. U87ΔEGFR derived tumors grew the fastest of the three models. Luciferase signal

was detected starting at 8 days post tumor cell implantation and rapidly increased over 2

weeks [Figure 33A]. We observed the photon flux emitted from the tumors was

consistent between mice and the signal grew exponentially by day 14 [n=12 mice]

[Figure 33B; Figure 34A]. Similar to these observations, luciferase signal was detected in

GB30 implanted mice starting at 8 days [n= 8 mice] [Figure 33A]. Interestingly, we

observed that BLI flux did not grow in a uniformly exponential pattern [Figure 33C;

Figure 34B]. At some later points BLI signal decreased despite an increase in tumor

116

growth by MRI. These observations suggested substantial tumor necrosis or

hemorrhaging were reducing cell viability and resulting in decreased BLI signal at later

time points. GB169 tumors grew much slower than U87ΔEGFR or GB30 tumors, and

this reduced growth rate was reflective in the BLI signal [n=10] [Figure 33A]. Initial BLI signaling was observable starting at 9 days in some of the mice. This finding was much earlier than observed by MRI [30-45 days]. While GB169 tumors also had necrosis, the luciferase flux increased fairly consistently over time and did not appear to have the same variability as the GB30 tumors [Figure 33D; Figure 34C]. Interestingly, however, we did observe at later time points with larger tumor volumes there was a plateau effect or decrease in photon flux suggesting tumor necrosis or other changes in the tumor microenvironment were affecting cell viability and luciferase signal output.

The relationships between BLI and MRI vary for each tumor model

We next assessed the relationship between BLI and MRI in each tumor model. In these studies we plotted photon flux versus tumor volume irrespective of time. We observed a strong linear relationship between BLI and MRI in the U87ΔEGFR tumor model [R2=

0.7525] [Figure 35A]. This result was consistent with our previous histological and imaging data. These homogenous tumors grew exponentially with relatively little deviation in BLI or MRI between mice at each time point. The relationship between photon flux and tumor volume was significantly weaker in the GB169 tumor model [R2=

0.5796] [Figure 35B]. These tumors contained substantial necrosis and hemorrhaging and we observed significant fluctuations in the BLI signal in these mice over time. Similarly,

117 we expected the relationship between BLI and MRI in the GB30 tumor model would also be poor. Like the GB169 model, these tumors were necrotic, hemorrhagic, and invasive.

We also observed substantial changes in BLI at later time points. Unexpectedly, we observed a high coefficient of determination [R2= 0.7478] [Figure 35C]. However, we believe this relationship is an artifact of too few MRI time points. Due to the rapid growth of this model we were only able to image this cohort of mice 3 times.

Sources of BLI signal artifacts

Signaling artifacts can result in the misinterpretation and misreporting of BLI imaging data [248]. We observed BLI signal varied significantly dependent on animal body positioning. Animals implanted with GB169 tumors were measured by BLI when they had substantial tumor growth. Here animals were imaged in the prone position and then again in the supine position. The signaling intensity dropped 5.8 fold in the prone versus supine position [Figure 36A]. This example highlights how alterations in head/body positioning can affect BLI signal and reinforces the importance of consistent animal placement. Interestingly, we observed with proper positioning BLI could be utilized to qualitatively identify sites of tumor growth. BLI was measured in mice implanted with

U87ΔEGFR tumor cells at 8, 10, 12, and 14 days post tumor cell implantation along with axial and coronal T2 MRI. Qualitative analysis demonstrated BLI provided a reliable estimate of tumor location [Figure 36B-C]. Furthermore, luciferase intensity appeared to increase proportionately throughout the longitudinal study at the site of tumor implantation.

118

Another potential source of BLI signal variation is tumor depth. In these studies, 2 mice

implanted were with GB169 neurospheres at different depths. One mouse was implanted

with cells at a depth of 1.85 mm and the second at a depth of 4 mm. The mice were

followed longitudinally, and the tumor volume and BLI signal was measured at 43, 50,

and 57 days post implantation. Both tumors grew at similar rates, and possessed

comparable tumor volumes at all three time points [Figure 37A]. Interestingly, however,

their BLI signals varied by 34.8, 13.4, and 5.8 fold at days 43, 50, and 57, respectively

[Figure 37B]. These results demonstrate the necessity of consistent surgical techniques as

well as the utilization of a significantly large cohort of animals to account for signal

variation due to differences in tumor depth.

Extracranial tumor growth also created BLI signal distortion. In these experiments we

implanted mice with GB169 tumors and we selected for a mouse with a tumor which

possessed substantial extracranial growth. We observed the BLI signal was 50.61%

higher in a mouse with extracranial tumor growth compared to a tumor within the

parenchyma even though this tumor was 24.32% smaller [20.71mm3 versus 27.61mm3]

[Figure 37C].

We next determined how tumor necrosis, hemorrhaging, and other changes in the tumor microenvironment altered BLI signaling. We implanted 3 mice with GB30 tumor cells and measured the BLI signal in these mice when they displayed substantial

119 necrosis/hemorrhaging by MRI [17 days post tumor cell implantation]. In mice with similar tumor volumes we observed tumor necrosis/hemorrhaging substantially impacted

BLI signal. These tumors exhibited 134.86 and 3.05 fold less luciferase signal than a similar sized tumor without observable necrosis/hemorrhage [white arrows], respectively

[Figure 37D]. This study highlights the importance of monitoring viable tumor cells via

BLI when examining therapeutic responses where MRI derived tumor volumes may show no difference.

120

Discussion

GB xenograft models are useful tools for evaluating experimental therapeutics and studying tumor biology. Both MRI and BLI are valuable imaging modalities for examining changes in tumor growth and biology. In this study we evaluated a serum cultured GB cell line and two neurosphere GB lines. While we demonstrated all three models possessed characteristics of human GB tumors, the GB169 and GB30 neurosphere derived tumors best recapitulated the disease. We found BLI generally correlated with changes in tumor volume in all three models evaluated, but that these relationships changed in the presence of tumor necrosis and hemorrhaging. Previously published work with U87MG derived tumors demonstrated a strong correlation of BLI with MRI tumor volumes, similar to the results we observed with the U87ΔEGFR model

[249]. In the GB30 and GB169 models, however, we found the BLI signal fluctuated substantially at different time points due to changes in the tumor microenvironment

[Figure 33]. We observed in GB30 tumors of a similar size that tumors with high levels of necrosis or hemorrhaging had a lower BLI signal. Similar studies have also noted that changes in the tumor microenvironment can result in alteration or plateau of BLI signals

[250]. In vivo, BLI signal intensity is reflective not only changes in the total number of cells, but also of viable cells capable of expressing the luciferase transgene. We observed that in models which strongly recapitulate the human disease BLI can offer unique insights to the changes occurring within the tumor beyond alterations in tumor volume. In a preclinical GB efficacy study with the anti-HGF drug for Ficlatuzumab, researchers found BLI was significantly more sensitive than MRI in detecting small drug responses

121

[251]. The ability of BLI to detect changes in cell viability allowed them to analyze drug

effects despite the absence of anatomical changes by MRI. Indeed, other studies have

noticed reduced correlations between BLI and MRI following treatments which effect

cell viability [252].

BLI was also advantageous in detecting early tumor engraftment. In the fast growing

U87ΔEGFR and GB30 models, we observed that detecting tumor formation by BLI and

MRI occurred roughly at similar timepoints. In the slower growing GB169 neurosphere

model, however, we found BLI was capable of detecting tumor engraftment prior to

appearing on MRI. Generally, we found when the BLI flux reached 1x106 p/s we could detect a visible tumor by MRI. While this flux value should be determined experimentally for each cell line, 1x106 p/s may serve as a general metric for

investigators wishing to know when to image intracranial tumors by MRI.

In this study we gained valuable insight into the sources of error which can occur from in

vivo tumor luciferase imaging. While BLI signal was useful for qualitatively identifying

tumor location, we observed this localizing capability was quickly lost through changes

in body positioning [Figure 36]. These observations highlight the importance of reliable

and consistent imaging practices. We also observed significant alterations in signal

intensity dependent on tumor depth and extracranial cell growth. Changes in tumor depth

of only 2.21 mm altered BLI signal intensity by as much as 34.8 fold, a finding which

highlights the importance of consistent surgical techniques [Figure 37]. For studies,

122 where treatments must be administered intratumorally, examining tumor depth by MRI prior to therapy may be particularly valuable.

Tumor cell growth through the needle track resulting in extracranial tumor formation is also common to some GB models. In our study, we observed the GB169 model was more prone to extracranial tumor formation than GB30 or U87ΔEGFR cells. Extracranial growth can also be the result of improper tumor cell implantation. Injecting large volumes at high flow rates can result in backflow out of the needle track and the growth of tumor cells on the surface of the skull. Yamada et al previously tested the effects of changes in flow rate and volume on indigo carmine dye dispersion in the brain, and found large volumes and rapid injection rates caused dye to spread in to the subarchanoid space and ventricles [253]. They found slow injection times with minimal volumes resulted in more consistent tumor formation. In our studies, we injected cells in a volume of 2 ul using an automated injector to inject cells at a rate of 0.4 ul/min. Following the injection of tumor cells the needle was left in the animal for 3 minutes and then slowly retracted over 5 minutes to prevent pulling cells up through the needle track [depth 3mm]. We found this strategy significantly reduced tumor cell backflow and extracranial growth in

GB models. It is possible that larger volumes can be injected at a slower injection rates, though we have not evaluated this possibility experimentally. While BLI signal intensity changed drastically with extracranial growth, this artifact is easy to screen for by palpating the surface of the animal’s head. We recommend using a higher number of mice so animals with these extracranial growths can be excluded from efficacy studies.

123

In Figure 38 we have proposed a general model for monitoring changes in tumor growth by BLI and MRI. After the creation of a luciferase expressing tumor cell line we recommend testing the tumor cell line in vivo to ensure it has maintained its tumorgenicity [Figure 38A]. In beginning the study we advise preliminary luciferase imaging to confirm tumor engraftment and growth prior to MRI [Figure 38B]. Once a viable signal is acquired, preliminary MRI should be obtained to setup a baseline for future tumor growth [Figure 38C]. At this stage, investigators should verify tumor positioning and depth. This step is particularly important if intratumoral injections will be performed. Following initial MRI volume analyses the study then enters the tracking stage [Figure 38D]. In stages D.1 and D.2 animals are monitored by both BLI and MRI.

This stage will allow the investigator to develop a growth model for subsequent studies.

After generating a tumor growth model, investigators can reduce cost prohibitive MRI and utilize regular BLI to determine treatment schedules and therapeutic responses. For example, BLI can be used with a calculated growth curve to determine when tumors have reached an adequate treatment size. The ability to demonstrate the presence of established tumors by BLI or MRI will significantly enhance the impact of therapeutic studies. We believe utilizing this model will improve efficacy, enhance repeatability, and reduce variability. Furthermore, the utilization of this imaging paradigm will allow direct comparisons between uniquely derived cell lines, treatments, and their tumor growth curves.

124

We demonstrated BLI is an extremely useful, cost-effective, and sensitive tool for

monitoring intracranial tumor growth. Research continues to improve BLI sensitivity, and

the recent development of a synthetic luciferin, CycLuc1, resulted in better

biodistribution and stronger BLI signaling in intracranial imaging studies than D-luciferin

[254]. Positron Emission Tomography [PET] is another non-invasive imaging modality which has been evaluated in preclinical GB studies. In comparative studies, BLI has been shown to be more sensitive than PET imaging in monitoring therapeutic responses in vivo [251, 255]. Other constructs using fluorescent or sodium iodide symporter reporters are emerging as new technologies for in vivo non-invasive imaging [256]. In

summary, the use of BLI in conjunction with MRI is a promising strategy for monitoring

GB tumor growth and therapeutic responses in vivo. Each modality offers unique data

which allows researchers to analyze changes in the tumor microenvironment. We have

identified the strengths and limitations of BLI and MRI and proposed an imaging

paradigm which takes advantage of each modality while reducing time and costs.

125

Methods

Cell Lines

U87ΔEGFR-luciferase GB cells were maintained in DMEM supplemented with 10% fetal bovine serum. U87ΔEGFR cells were obtained in April 2005 from Dr. E Antonio

Chiocca [Ohio State University, Columbus, Ohio]. GB30-luciferase and GB169- luciferase were obtained under an approved Institutional Review Board [IRB] protocol from the office of Responsible Research Practices at The Ohio State University Medical

Center where written consent for the collection of surgical specimens was obtained from patients, as described [166]. Patient-derived GB30-luciferase and GB169-luciferase neurospheres were isolated and maintained as tumor spheres in Neurobasal Medium supplemented with 2% B27, human EGF [50 ng/ml], and bFGF [50 ng/ml] in low- attachment cell culture flasks as previously described [166] All cells were incubated at

37oC in an atmosphere with 5% carbon dioxide and maintained with 100 units of penicillin/mL, and 0.1 mg of streptomycin/mL [Penn/Strep]. GB30 and GB169 neurospheres were originally received in 2012 from Dr. EA Chiocca [Ohio State

University, Columbus, OH]. U87ΔEGFR [January 2015], and GB30 [January 2015] cells were genotyped by the University of Arizona Genetics Core via STR profiling. GB169 cells have not been genotyped since receipt. All cells are routinely monitored for changes in morphology and growth rate. All cells were negative for mycoplasma.

126

Animal surgery

All animal experiments were performed in accordance with the Subcommittee on

Research Animal Care of The Ohio State University guidelines and were approved by the institutional review board. 6-8-week-old, female athymic nude mice [NCI], were used for in vivo tumor studies. Intracranial surgeries were performed as previously described with stereotactic implantation of 50,000 U87ΔEGFR, 100,000 GB30, or 100,000 GB169 cells.[156] Animals were euthanized when they showed signs of morbidity.

Luciferase Imaging

The luciferase signal was visualized/quantified utilizing an IVIS Lumina II imaging system. Briefly, mice were anesthetized with 2.5% isoflurane mixed with 1 liter per minute carbogen and maintained with 1-1.5% isoflurane during imaging. Mice received an intraperitoneal injection of D-luciferin [150 mg/kg] [Caliper Life Sciences] approximately 15 minutes before imaging. Throughout the longitudinal study, animals were imaged with consistent camera parameters chosen to maximize signal without over saturating the image. These parameters were set to a FOV of 24 x 24 cm [FOV24 lens], binning set as medium, an F/stop [aperture] of 1, and sequenced exposure times of 30 seconds. This exposure time was selected after experimenting with a range of exposure times [5, 10, 30, 60 and 120 seconds] to determine the most effective and accurate setting during the longitudinal study. This also ensured consistency in comparisons at each point as the tumor signal intensity and region would continue to grow. Each group of mice throughout the study was imaged with this same protocol.

127

IHC Analyses and Image Acquisition

Brain tissues were fixed with 4% PFA and dehydrated for paraffin embedding. 5 µm

paraffin sections were cut and processed for H&E staining. Images of H&E staining were

obtained by scanning whole slides with a ScanScopeXT microscope from Aperio

Techonologies. Images were prepared for presentation using ImageScope v12.1.0.5029

MRI Studies

Anatomic imaging of tumor bearing mice was performed on the days indicated using a

T2-weighted RARE sequence [TR = 2500ms, TE=12ms, Rare Factor = 8, FOV=2.0 x 2.0 cm, Image Matrix = 256 x 256, Slice 19 Thickness = 0.5 or 1mm, 32 or 16 Slices]. Each mouse was placed in the magnet head-first in the prone position. The head was secured

with ear prongs and a bite bar. Images were acquired contiguously in the axial plane. For

data analysis, a region-of-interest [ROI] that included the tumor was manually outlined.

Tumor volumes were calculated from ROIs as previously described [156].

128

Figure and Tables

Figure 31: GB intracranial xenograft models recapitulate the human disease.

Representative images of immunohistochemistry for hematoxylin and eosin (H&E) are shown for U87ΔEGFR [A.], GB169 [B.], and GB30 [C.]. A. U87ΔEGFR tumors were tightly packed with no tumor cell invasion. B. GB169 tumors showed no tumor cell invasion (black inset) and psuedopallisades indicative of hypoxia (red inset). C. GB30 tumors were invasive (black inset), and contained psuedopalisades around area of necroses (red inset) in addition to glomeruloid bodies (green inset). Whole brain image scale bars are 200mm and inset scale bars are 200um.

129

Figure 32: GB tumor volume increases with time by MRI.

A. Representative coronal sections of tumor bearing mice obtained by T2-weighted MRI. Images were taken at the days indicated post tumor cell implantation. Quantitative tumor volume analysis of mice implanted intracranially with [B.] U87ΔEGFR [n=12], [C.] GB169 [n=10], or [D.] GB30 [n=8] tumors. Data shown is total tumor volume [mm3] for each animal at a given time point.

130

Figure 33: BLI signal increases with time in GB intracranial xenograft models.

A. Representative images of mice bearing intracranial tumors taken at the days indicated post tumor cell implantation. Quantitative BLI analysis of mice implanted intracranially with [B.] U87ΔEGFR [n=12], [C.] GB30 [n=8], or [D.] GB169 [n=10] tumors. Data shown is total photon flux [p/s] for each animal at a given time point.

131

Figure 34: Changes in the tumor microenvironment increases BLI and MRI signal

variability.

The photon flux [p/s] and tumor volume [mm3] was plotted for the [A.] U87ΔEGFR [n=12], [B.] GB169 [n=10], or [C.] GB30 [n=8] tumor models. Each mouse is represented as a dot. The average photon flux or tumor volume at each time point is displayed as a dashed line in order to show deviation from the mean.

132

Figure 35: The relationship between BLI and MRI varies between tumor models.

Tumor volume versus photon flux was plotted for the [A.] U87ΔEGFR [n=12], [B.] GB169 [n=10], or [C.] GB30 [n=8] tumor models. Linear regressions were performed to evaluate the relationship between each modality irrespective of time. Equations with the coefficient of determination [R2] are shown for each plot.

133

Figure 36: BLI signal fluctuates with animal positioning.

A. Mice implanted intracranially with GB169 tumors were imaged in the prone position [Left] then quickly flipped and imaged in the supine [Right] position. Average BLI flux of mice [n=3] is shown. B. U87ΔEGFR tumor growth was monitored in vivo via MRI and BLI. Axial [Top row], coronal [Middle Row], and BLI [Bottom row] images were taken of the same mouse at the time points indicated post tumor cell implantation.

134

Figure 37: Tumor implantation depth, extracranial growth, and necrosis can create BLI

signal artifacts.

A. GB169 neurospheres were implanted intracranially in mice at a depth of 1.85 mm [Top Row] or 4.06 mm [Bottom Row]. The tumor depth was determined by measuring the center mass depth 43 days post tumor cell implantation. Tumor volume was followed longitudinally by T2 MRI. Data shown are representative coronal images at each time point. B. Total photon flux of mice described in [A.] over time. Red arrows indicate the time points shown in [A.]. C. Representative coronal T2 MRI images of mice bearing GB169 intracranial tumors with [Left] and without [Right] extracranial growth. Data shown below are tumor volume and total photon flux 57 days post tumor cell implantation for each mouse. D. Representative coronal T2 MRI images of mice implanted with GB30 tumors. Data shown below are total tumor volume and total photon flux for each mouse 17 days post tumor cell implantation. White arrows indicate areas of putative necrosis or hemorrhaging.

135

Figure 38: Multi-modality imaging strategy for intracranial GB tumor models.

After the creation and validation of a luciferase expressing cell line [Step A]. Animals are monitored by BLI until tumor engraftment is detected [Step B]. After obtaining a BLI signal, baseline MRI is conducted to verify tumor location and size [Step C]. Step D.1 and D.2 utilize BLI supplemented with MRI to monitor and track changes in tumor growth. This strategy allows researchers to create a tumor growth model for each cell line. Time and cost intensive MRI can then be reduced in subsequent studies and BLI can be used to determine treatment times and changes in tumor size due to treatment effects.

136

Conclusions and Future Directions

In part one of this dissertation we demonstrated the therapeutic potential of a targeted OV

to treat breast cancer [BC] brain metastases. In two models of BC brain metastases, we

were able to substantially increase the survival of mice with a single, intratumoral dose of the 34.5ENVE virus. This work is especially significant given that to date, clinical trials

for metastases patients have proved ineffective and been marred by toxicity. A recent BC

brain metastases clinical trial with lapatinib and capecitabine noted that nearly a third of

patients experienced at least one severe adverse event due to toxicity [33]. In our studies

we observed no toxicity from OV administration. We believe the administration of OVs

may be beneficial for patients with cancers unresponsive to traditional chemotherapies, with multiple lesions, or with inoperable tumors. In our preliminary data we

demonstrated the combination of 34.5ENVE with the PARP inhibitor olaparib killed BC

cells synergistically. This finding is particularly important because olaparib is already

FDA approved for ovarian cancer, and the drug is currently in clinical trials for BC brain

metastases. Collectively, these studies support future clinical trials testing the efficacy of

OVs for brain metastases.

In part two of this dissertation, we identified TNFα as a major macrophage/microglia

secreted factor which reduces OV replication. We observed the combination of OV with

TNFα blocking antibodies significantly enhanced the survival of mice with intracranial

137 brain tumors. The TNFα inhibitors etanercept, adalimumab, certolizumab, and golimumab are currently FDA approved for a variety of diseases and could be readily utilized in OV clinical trials. Additionally, oncolytic virotherapy for glioblastoma is often administered following tumor resection and concurrently with radiation and chemotherapy. Radiation and chemotherapy are known to induce the production of cytokines such as TNFα, and thus patients may benefit from the transient use of TNFα inhibitors prior to OV administration in order to enhance oncolytic tumor cell killing and reduce CNS inflammation [116, 135, 227]. While antibody mediated inhibition of TNFα was sufficient to enhance animal survival, the development of specific, soluble TNFα inhibitors which better penetrate the blood-brain-barrier or the creation OVs which reduce TNFα may further increase OV efficacy. In preliminary data we demonstrated the expression of Vstat120 by the RAMBO virus significantly decreased microglia and macrophage secreted TNFα and improved OV anti-tumor efficacy. Future studies comparing RAMBO to rHSVQ1 with TNFα blocking antibodies are ongoing.

Cumulatively, these experiments provide strong evidence which support TNFα modulation in combination with OV therapy for brain tumors.

In part three of this dissertation, we evaluated BLI and MRI in 3 different murine models of glioblastoma. We found BLI and MRI output was significantly affected by tumor necrosis, hemorrhaging, tumor depth, extra-cranial growth, and animal positioning. In synthesizing the data from this study, we created a multi-modality imaging model for analyzing changes in tumor growth and biology. We believe this model will improve

138 efficacy, enhance repeatability, and reduce variability. Furthermore, we believe the utilization of this imaging paradigm will allow direct comparisons between uniquely derived cell lines, treatments, and their tumor growth curves. The combination of BLI and MRI modalities will significantly aid in the pre-clinical evaluation OV based therapeutics for brain tumors.

While oncolytic viruses represent a promising treatment for brain tumors, significant improvements are needed before this exciting immunotherapy is approved for clinical use. The unique brain tumor microenvironment and restrictive blood brain barrier make these cancers a therapeutic challenge. Research to overcome these obstacles is ongoing, and OV therapies will play a large part in the future of cancer treatment.

139

References

1. Benson JR, Jatoi I. The global breast cancer burden. Future Oncol. 2012;8(6):697-702.

2. Hayat MJ, Howlader N, Reichman ME, Edwards BK. Cancer statistics, trends, and multiple primary cancer analyses from the Surveillance, Epidemiology, and End Results (SEER) Program. The oncologist. 2007;12(1):20-37.

3. Li C. Breast Cancer Epidemiology: Springer New York; 2010.

4. Li ML, Greenberg RA. Links between genome integrity and BRCA1 tumor suppression. Trends in biochemical sciences. 2012;37(10):418-24.

5. Lin NU, Amiri-Kordestani L, Palmieri D, Liewehr DJ, Steeg PS. CNS metastases in breast cancer: old challenge, new frontiers. Clinical cancer research : an official journal of the American Association for Cancer Research. 2013;19(23):6404-18. Epub 2013/12/04.

6. Lee YT. Breast carcinoma: pattern of metastasis at autopsy. Journal of surgical oncology. 1983;23(3):175-80.

7. Cheng X, Hung MC. Breast cancer brain metastases. Cancer metastasis reviews. 2007;26(3-4):635-43.

8. Weil RJ, Palmieri DC, Bronder JL, Stark AM, Steeg PS. Breast cancer metastasis to the central nervous system. Am J Pathol. 2005;167(4):913-20. Epub 2005/09/30.

9. Altundag K, Bondy ML, Mirza NQ, Kau SW, Broglio K, Hortobagyi GN, et al. Clinicopathologic characteristics and prognostic factors in 420 metastatic breast cancer patients with central nervous system metastasis. Cancer. 2007;110(12):2640-7.

10. Pekmezci M, Perry A. Neuropathology of brain metastases. Surgical neurology international. 2013;4(Suppl 4):S245-55. Epub 2013/05/30.

11. Meisen WH, Dubin S, Sizemore ST, Mathsyaraja H, Thies K, Lehman NL, et al. Changes in BAI1 and nestin expression are prognostic indicators for survival and metastases in breast cancer and provide opportunities for dual targeted therapies. Molecular cancer therapeutics. 2015;14(1):307-14. Epub 2014/11/08.

140

12. Eroles P, Bosch A, Perez-Fidalgo JA, Lluch A. Molecular biology in breast cancer: intrinsic subtypes and signaling pathways. Cancer treatment reviews. 2012;38(6):698- 707.

13. Zimm S, Wampler GL, Stablein D, Hazra T, Young HF. Intracerebral metastases in solid- tumor patients: natural history and results of treatment. Cancer. 1981;48(2):384-94.

14. Forsyth PA, Posner JB. Headaches in patients with brain tumors: a study of 111 patients. Neurology. 1993;43(9):1678-83.

15. Schellinger PD, Meinck HM, Thron A. Diagnostic accuracy of MRI compared to CCT in patients with brain metastases. Journal of neuro-oncology. 1999;44(3):275-81.

16. Rohren EM, Provenzale JM, Barboriak DP, Coleman RE. Screening for cerebral metastases with FDG PET in patients undergoing whole-body staging of non-central nervous system malignancy. Radiology. 2003;226(1):181-7.

17. Hempen C, Weiss E, Hess CF. Dexamethasone treatment in patients with brain metastases and primary brain tumors: do the benefits outweigh the side-effects? Supportive care in cancer : official journal of the Multinational Association of Supportive Care in Cancer. 2002;10(4):322-8.

18. Lin NU, Bellon JR, Winer EP. CNS metastases in breast cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2004;22(17):3608-17.

19. May N. Tsao DR, Andrew Wirth. Radiotherapeutic and surgical management for newly diagnosed brain metastasis(es): An American Society for Radiation Oncology evidence- based guideline Practical Radiation Oncology. 2012;2(3):210-25.

20. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, Kryscio RJ, et al. A randomized trial of surgery in the treatment of single metastases to the brain. The New England journal of medicine. 1990;322(8):494-500.

21. Iwadate Y, Namba H, Yamaura A. Significance of surgical resection for the treatment of multiple brain metastases. Anticancer research. 2000;20(1B):573-7.

22. Lim E, Lin NU. New insights and emerging therapies for breast cancer brain metastases. Oncology. 2012;26(7):652-9, 63.

23. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science. 1987;235(4785):177-82.

141

24. Xu ZY, Marko NF, Chao ST, Angelov L, Vogelbaum MA, Suh JH, et al. Relationship Between HER2 Status and Prognosis in Women With Brain Metastases From Breast Cancer. Int J Radiat Oncol. 2012;82(5):E739-E47.

25. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Bajamonde A, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. The New England journal of medicine. 2001;344(11):783-92.

26. Clayton AJ, Danson S, Jolly S, Ryder WD, Burt PA, Stewart AL, et al. Incidence of cerebral metastases in patients treated with trastuzumab for metastatic breast cancer. British journal of cancer. 2004;91(4):639-43.

27. Bendell JC, Domchek SM, Burstein HJ, Harris L, Younger J, Kuter I, et al. Central nervous system metastases in women who receive trastuzumab-based therapy for metastatic breast carcinoma. Cancer. 2003;97(12):2972-7.

28. Pestalozzi BC, Brignoli S. Trastuzumab in CSF. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2000;18(11):2349-51.

29. Stemmler HJ, Schmitt M, Willems A, Bernhard H, Harbeck N, Heinemann V. Ratio of trastuzumab levels in serum and cerebrospinal fluid is altered in HER2-positive breast cancer patients with brain metastases and impairment of blood-brain barrier. Anti-cancer drugs. 2007;18(1):23-8.

30. Schwartzberg LS, Franco SX, Florance A, O'Rourke L, Maltzman J, Johnston S. Lapatinib plus letrozole as first-line therapy for HER-2+ hormone receptor-positive metastatic breast cancer. The oncologist. 2010;15(2):122-9.

31. Gril B, Palmieri D, Bronder JL, Herring JM, Vega-Valle E, Feigenbaum L, et al. Effect of lapatinib on the outgrowth of metastatic breast cancer cells to the brain. Journal of the National Cancer Institute. 2008;100(15):1092-103.

32. Taskar KS, Rudraraju V, Mittapalli RK, Samala R, Thorsheim HR, Lockman J, et al. Lapatinib distribution in HER2 overexpressing experimental brain metastases of breast cancer. Pharmaceutical research. 2012;29(3):770-81.

33. Bachelot T, Romieu G, Campone M, Dieras V, Cropet C, Dalenc F, et al. Lapatinib plus capecitabine in patients with previously untreated brain metastases from HER2-positive metastatic breast cancer (LANDSCAPE): a single-group phase 2 study. The lancet oncology. 2013;14(1):64-71. Epub 2012/11/06.

34. Nieder C. Brain metastases from HER2-positive breast cancer. Lancet Oncology. 2013;14(1):E2-E3.

142

35. Bartsch R, Preusser M. Primary systemic treatment of breast-cancer brain metastases. Lancet Oncology. 2013;14(1):8-9.

36. Sambade MJ, Kimple RJ, Camp JT, Peters E, Livasy CA, Sartor CI, et al. Lapatinib in combination with radiation diminishes tumor regrowth in HER2+ and basal-like/EGFR+ breast tumor xenografts. International journal of radiation oncology, biology, physics. 2010;77(2):575-81.

37. Anders CK, Carey LA. Biology, metastatic patterns, and treatment of patients with triple- negative breast cancer. Clinical breast cancer. 2009;9 Suppl 2:S73-81.

38. Greenup R, Buchanan A, Lorizio W, Rhoads K, Chan S, Leedom T, et al. Prevalence of BRCA mutations among women with triple-negative breast cancer (TNBC) in a genetic counseling cohort. Annals of surgical oncology. 2013;20(10):3254-8.

39. Rios J, Puhalla S. PARP inhibitors in breast cancer: BRCA and beyond. Oncology. 2011;25(11):1014-25.

40. Kim LS, Huang S, Lu W, Lev DC, Price JE. Vascular endothelial growth factor expression promotes the growth of breast cancer brain metastases in nude mice. Clinical & experimental metastasis. 2004;21(2):107-18.

41. Adamo B, Deal AM, Burrows E, Geradts J, Hamilton E, Blackwell KL, et al. Phosphatidylinositol 3-kinase pathway activation in breast cancer brain metastases. Breast cancer research : BCR. 2011;13(6):R125.

42. Berns K, Horlings HM, Hennessy BT, Madiredjo M, Hijmans EM, Beelen K, et al. A functional genetic approach identifies the PI3K pathway as a major determinant of trastuzumab resistance in breast cancer. Cancer cell. 2007;12(4):395-402.

43. Lim E, Lin NU. Updates on the management of breast cancer brain metastases. Oncology. 2014;28(7):572-8.

44. Schwartzbaum JA, Fisher JL, Aldape KD, Wrensch M. Epidemiology and molecular pathology of glioma. Nature clinical practice Neurology. 2006;2(9):494-503; quiz 1 p following 16. Epub 2006/08/26.

45. Ohgaki H, Kleihues P. Epidemiology and etiology of gliomas. Acta neuropathologica. 2005;109(1):93-108. Epub 2005/02/03.

46. De Roos AJ, Stewart PA, Linet MS, Heineman EF, Dosemeci M, Wilcosky T, et al. Occupation and the risk of adult glioma in the United States. Cancer causes & control : CCC. 2003;14(2):139-50. Epub 2003/05/17.

143

47. Chen H, Ward MH, Tucker KL, Graubard BI, McComb RD, Potischman NA, et al. Diet and risk of adult glioma in eastern Nebraska, United States. Cancer causes & control : CCC. 2002;13(7):647-55.

48. Braganza MZ, Rajaraman P, Park Y, Inskip PD, Freedman ND, Hollenbeck AR, et al. Cigarette smoking, alcohol intake, and risk of glioma in the NIH-AARP Diet and Health Study. British journal of cancer. 2014;110(1):242-8. Epub 2013/12/18.

49. Preston DL, Ron E, Yonehara S, Kobuke T, Fujii H, Kishikawa M, et al. Tumors of the nervous system and pituitary gland associated with atomic bomb radiation exposure. Journal of the National Cancer Institute. 2002;94(20):1555-63.

50. Neglia JP, Meadows AT, Robison LL, Kim TH, Newton WA, Ruymann FB, et al. Second neoplasms after acute lymphoblastic leukemia in childhood. The New England journal of medicine. 1991;325(19):1330-6.

51. Li FP, Fraumeni JF, Jr., Mulvihill JJ, Blattner WA, Dreyfus MG, Tucker MA, et al. A cancer family syndrome in twenty-four kindreds. Cancer research. 1988;48(18):5358-62. Epub 1988/09/15.

52. Huson SM, Harper PS, Compston DA. Von Recklinghausen neurofibromatosis. A clinical and population study in south-east Wales. Brain : a journal of neurology. 1988;111 ( Pt 6):1355-81. Epub 1988/12/01.

53. Al-Saleem T, Wessner LL, Scheithauer BW, Patterson K, Roach ES, Dreyer SJ, et al. Malignant tumors of the kidney, brain, and soft tissues in children and young adults with the tuberous sclerosis complex. Cancer. 1998;83(10):2208-16. Epub 1998/11/25.

54. Hamilton SR, Liu B, Parsons RE, Papadopoulos N, Jen J, Powell SM, et al. The molecular basis of Turcot's syndrome. The New England journal of medicine. 1995;332(13):839-47. Epub 1995/03/30.

55. Liaw D, Marsh DJ, Li J, Dahia PL, Wang SI, Zheng Z, et al. Germline mutations of the PTEN gene in Cowden disease, an inherited breast and thyroid cancer syndrome. Nature genetics. 1997;16(1):64-7. Epub 1997/05/01.

56. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A, et al. The 2007 WHO classification of tumours of the central nervous system. Acta neuropathologica. 2007;114(2):97-109.

57. Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, 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):98-110.

144

58. Noushmehr H, Weisenberger DJ, Diefes K, Phillips HS, Pujara K, Berman BP, et al. Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma. Cancer cell. 2010;17(5):510-22.

59. Omuro A, DeAngelis LM. Glioblastoma and other malignant gliomas: a clinical review. Jama. 2013;310(17):1842-50.

60. Glantz MJ, Cole BF, Forsyth PA, Recht LD, Wen PY, Chamberlain MC, et al. Practice parameter: anticonvulsant prophylaxis in patients with newly diagnosed brain tumors. Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2000;54(10):1886-93.

61. Chen W. Clinical applications of PET in brain tumors. Journal of nuclear medicine : official publication, Society of Nuclear Medicine. 2007;48(9):1468-81.

62. Pauleit D, Floeth F, Hamacher K, Riemenschneider MJ, Reifenberger G, Muller HW, et al. O-(2-[18F]fluoroethyl)-L-tyrosine PET combined with MRI improves the diagnostic assessment of cerebral gliomas. Brain : a journal of neurology. 2005;128(Pt 3):678-87.

63. Preusser M, de Ribaupierre S, Wohrer A, Erridge SC, Hegi M, Weller M, et al. Current concepts and management of glioblastoma. Annals of neurology. 2011;70(1):9-21.

64. Stummer W, Reulen HJ, Meinel T, Pichlmeier U, Schumacher W, Tonn JC, et al. Extent of resection and survival in glioblastoma multiforme: identification of and adjustment for bias. Neurosurgery. 2008;62(3):564-76; discussion -76.

65. Pichlmeier U, Bink A, Schackert G, Stummer W, Group ALAGS. Resection and survival in glioblastoma multiforme: an RTOG recursive partitioning analysis of ALA study patients. Neuro Oncol. 2008;10(6):1025-34.

66. Perry J, Chambers A, Spithoff K, Laperriere N. Gliadel wafers in the treatment of malignant glioma: a systematic review. Current oncology. 2007;14(5):189-94.

67. Westphal M, Hilt DC, Bortey E, Delavault P, Olivares R, Warnke PC, et al. A phase 3 trial of local chemotherapy with biodegradable carmustine (BCNU) wafers (Gliadel wafers) in patients with primary malignant glioma. Neuro Oncol. 2003;5(2):79-88.

68. Westphal M, Ram Z, Riddle V, Hilt D, Bortey E, Executive Committee of the Gliadel Study G. Gliadel wafer in initial surgery for malignant glioma: long-term follow-up of a multicenter controlled trial. Acta neurochirurgica. 2006;148(3):269-75; discussion 75.

69. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. The New England journal of medicine. 2005;352(10):987-96.

145

70. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ, Janzer RC, 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. The lancet oncology. 2009;10(5):459-66.

71. Newlands ES, Stevens MF, Wedge SR, Wheelhouse RT, Brock C. Temozolomide: a review of its discovery, chemical properties, pre-clinical development and clinical trials. Cancer treatment reviews. 1997;23(1):35-61.

72. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. The New England journal of medicine. 2005;352(10):997-1003.

73. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. The New England journal of medicine. 2005;352(10):987-96. Epub 2005/03/11.

74. Gaspar LE, Fisher BJ, Macdonald DR, LeBer DV, Halperin EC, Schold SC, Jr., et al. Supratentorial malignant glioma: patterns of recurrence and implications for external beam local treatment. International journal of radiation oncology, biology, physics. 1992;24(1):55-7.

75. Friedman HS, Prados MD, Wen PY, Mikkelsen T, Schiff D, Abrey LE, et al. Bevacizumab alone and in combination with irinotecan in recurrent glioblastoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2009;27(28):4733-40.

76. Kreisl TN, Kim L, Moore K, Duic P, Royce C, Stroud I, et al. Phase II trial of single- agent bevacizumab followed by bevacizumab plus irinotecan at tumor progression in recurrent glioblastoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2009;27(5):740-5.

77. Gilbert MR, Dignam JJ, Armstrong TS, Wefel JS, Blumenthal DT, Vogelbaum MA, et al. A randomized trial of bevacizumab for newly diagnosed glioblastoma. The New England journal of medicine. 2014;370(8):699-708.

78. Chinot OL, Wick W, Mason W, Henriksson R, Saran F, Nishikawa R, et al. Bevacizumab plus radiotherapy-temozolomide for newly diagnosed glioblastoma. The New England journal of medicine. 2014;370(8):709-22.

79. Barazzuol L, Jena R, Burnet NG, Meira LB, Jeynes JC, Kirkby KJ, et al. Evaluation of poly (ADP-ribose) polymerase inhibitor ABT-888 combined with radiotherapy and temozolomide in glioblastoma. Radiation oncology. 2013;8:65.

146

80. Lin F, de Gooijer MC, Roig EM, Buil LC, Christner SM, Beumer JH, et al. ABCB1, ABCG2, and PTEN determine the response of glioblastoma to temozolomide and ABT- 888 therapy. Clinical cancer research : an official journal of the American Association for Cancer Research. 2014;20(10):2703-13.

81. Tentori L, Leonetti C, Scarsella M, D'Amati G, Vergati M, Portarena I, et al. Systemic administration of GPI 15427, a novel poly(ADP-ribose) polymerase-1 inhibitor, increases the antitumor activity of temozolomide against intracranial melanoma, glioma, lymphoma. Clinical cancer research : an official journal of the American Association for Cancer Research. 2003;9(14):5370-9.

82. Donawho CK, Luo Y, Luo Y, Penning TD, Bauch JL, Bouska JJ, et al. ABT-888, an orally active poly(ADP-ribose) polymerase inhibitor that potentiates DNA-damaging agents in preclinical tumor models. Clinical cancer research : an official journal of the American Association for Cancer Research. 2007;13(9):2728-37.

83. Kirson ED, Schneiderman RS, Dbaly V, Tovarys F, Vymazal J, Itzhaki A, et al. Chemotherapeutic treatment efficacy and sensitivity are increased by adjuvant alternating electric fields (TTFields). BMC medical physics. 2009;9:1.

84. Wheeler CJ, Black KL. DCVax-Brain and DC vaccines in the treatment of GBM. Expert opinion on investigational drugs. 2009;18(4):509-19.

85. Rolle CE, Sengupta S, Lesniak MS. Challenges in clinical design of immunotherapy trials for malignant glioma. Neurosurgery clinics of North America. 2010;21(1):201-14.

86. Jackson C, Ruzevick J, Phallen J, Belcaid Z, Lim M. Challenges in immunotherapy presented by the glioblastoma multiforme microenvironment. Clinical & developmental immunology. 2011;2011:732413.

87. Furnari FB, Fenton T, Bachoo RM, Mukasa A, Stommel JM, Stegh A, et al. Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes & development. 2007;21(21):2683-710.

88. Ueki K, Ono Y, Henson JW, Efird JT, von Deimling A, Louis DN. CDKN2/p16 or RB alterations occur in the majority of glioblastomas and are inversely correlated. Cancer research. 1996;56(1):150-3.

89. Gan HK, Kaye AH, Luwor RB. The EGFRvIII variant in glioblastoma multiforme. Journal of clinical neuroscience : official journal of the Neurosurgical Society of Australasia. 2009;16(6):748-54.

90. Huang HS, Nagane M, Klingbeil CK, Lin H, Nishikawa R, Ji XD, et al. The enhanced tumorigenic activity of a mutant epidermal growth factor receptor common in human

147

cancers is mediated by threshold levels of constitutive tyrosine phosphorylation and unattenuated signaling. The Journal of biological chemistry. 1997;272(5):2927-35.

91. Nishikawa R, Sugiyama T, Narita Y, Furnari F, Cavenee WK, Matsutani M. Immunohistochemical analysis of the mutant epidermal growth factor, deltaEGFR, in glioblastoma. Brain tumor pathology. 2004;21(2):53-6.

92. Inda MM, Bonavia R, Mukasa A, Narita Y, Sah DW, Vandenberg S, et al. Tumor heterogeneity is an active process maintained by a mutant EGFR-induced cytokine circuit in glioblastoma. Genes & development. 2010;24(16):1731-45.

93. Strik H, Deininger M, Streffer J, Grote E, Wickboldt J, Dichgans J, et al. BCL-2 family protein expression in initial and recurrent glioblastomas: modulation by radiochemotherapy. Journal of neurology, neurosurgery, and psychiatry. 1999;67(6):763- 8. Epub 1999/11/24.

94. Stiles CD, Rowitch DH. Glioma stem cells: a midterm exam. Neuron. 2008;58(6):832-46.

95. Chen J, Li Y, Yu TS, McKay RM, Burns DK, Kernie SG, et al. A restricted cell population propagates glioblastoma growth after chemotherapy. Nature. 2012;488(7412):522-6.

96. Bao S, Wu Q, McLendon RE, Hao Y, Shi Q, Hjelmeland AB, et al. Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature. 2006;444(7120):756-60. Epub 2006/10/20.

97. Venere M, Fine HA, Dirks PB, Rich JN. Cancer stem cells in gliomas: identifying and understanding the apex cell in cancer's hierarchy. Glia. 2011;59(8):1148-54. Epub 2011/05/07.

98. Heddleston JM, Li Z, McLendon RE, Hjelmeland AB, Rich JN. The hypoxic microenvironment maintains glioblastoma stem cells and promotes reprogramming towards a cancer stem cell phenotype. Cell Cycle. 2009;8(20):3274-84. Epub 2009/09/23.

99. Lathia JD, Gallagher J, Heddleston JM, Wang J, Eyler CE, Macswords J, et al. Integrin alpha 6 regulates glioblastoma stem cells. Cell stem cell. 2010;6(5):421-32.

100. Son MJ, Woolard K, Nam DH, Lee J, Fine HA. SSEA-1 is an enrichment marker for tumor-initiating cells in human glioblastoma. Cell stem cell. 2009;4(5):440-52.

101. Kaufman HL, Ruby CE, Hughes T, Slingluff CL, Jr. Current status of granulocyte- macrophage colony-stimulating factor in the immunotherapy of melanoma. Journal for immunotherapy of cancer. 2014;2:11.

102. Garber K. China approves world's first oncolytic virus therapy for cancer treatment. Journal of the National Cancer Institute. 2006;98(5):298-300. Epub 2006/03/02. 148

103. Russell SJ, Peng KW, Bell JC. Oncolytic virotherapy. Nature biotechnology. 2012;30(7):658-70.

104. Taylor TJ, Brockman MA, McNamee EE, Knipe DM. Herpes simplex virus. Frontiers in bioscience : a journal and virtual library. 2002;7:d752-64.

105. Goldstein DJ, Weller SK. Herpes simplex virus type 1-induced ribonucleotide reductase activity is dispensable for virus growth and DNA synthesis: isolation and characterization of an ICP6 lacZ insertion mutant. Journal of virology. 1988;62(1):196-205.

106. Mineta T, Rabkin SD, Martuza RL. Treatment of malignant gliomas using ganciclovir- hypersensitive, ribonucleotide reductase-deficient herpes simplex viral mutant. Cancer research. 1994;54(15):3963-6.

107. Nakamura H, Kasuya H, Mullen JT, Yoon SS, Pawlik TM, Chandrasekhar S, et al. Regulation of herpes simplex virus gamma(1)34.5 expression and oncolysis of diffuse liver metastases by Myb34.5. The Journal of clinical investigation. 2002;109(7):871-82.

108. Kanai R, Zaupa C, Sgubin D, Antoszczyk SJ, Martuza RL, Wakimoto H, et al. Effect of gamma34.5 deletions on oncolytic herpes simplex virus activity in brain tumors. Journal of virology. 2012;86(8):4420-31.

109. Harland J, Dunn P, Cameron E, Conner J, Brown SM. The herpes simplex virus (HSV) protein ICP34.5 is a virion component that forms a DNA-binding complex with proliferating cell nuclear antigen and HSV replication proteins. Journal of neurovirology. 2003;9(4):477-88. Epub 2003/08/09.

110. Smith KD, Mezhir JJ, Bickenbach K, Veerapong J, Charron J, Posner MC, et al. Activated MEK suppresses activation of PKR and enables efficient replication and in vivo oncolysis by Deltagamma(1)34.5 mutants of herpes simplex virus 1. Journal of virology. 2006;80(3):1110-20.

111. MacLean AR, ul-Fareed M, Robertson L, Harland J, Brown SM. Herpes simplex virus type 1 deletion variants 1714 and 1716 pinpoint neurovirulence-related sequences in Glasgow strain 17+ between immediate early gene 1 and the 'a' sequence. The Journal of general virology. 1991;72 ( Pt 3):631-9.

112. Rampling R, Cruickshank G, Papanastassiou V, Nicoll J, Hadley D, Brennan D, et al. Toxicity evaluation of replication-competent herpes simplex virus (ICP 34.5 null mutant 1716) in patients with recurrent malignant glioma. Gene therapy. 2000;7(10):859-66.

113. Harrow S, Papanastassiou V, Harland J, Mabbs R, Petty R, Fraser M, et al. HSV1716 injection into the brain adjacent to tumour following surgical resection of high-grade glioma: safety data and long-term survival. Gene therapy. 2004;11(22):1648-58.

149

114. Papanastassiou V, Rampling R, Fraser M, Petty R, Hadley D, Nicoll J, et al. The potential for efficacy of the modified (ICP 34.5(-)) herpes simplex virus HSV1716 following intratumoural injection into human malignant glioma: a proof of principle study. Gene therapy. 2002;9(6):398-406.

115. Markert JM, Liechty PG, Wang W, Gaston S, Braz E, Karrasch M, et al. Phase Ib trial of mutant herpes simplex virus G207 inoculated pre-and post-tumor resection for recurrent GBM. Molecular therapy : the journal of the American Society of Gene Therapy. 2009;17(1):199-207.

116. Bradley JD, Kataoka Y, Advani S, Chung SM, Arani RB, Gillespie GY, et al. Ionizing radiation improves survival in mice bearing intracranial high-grade gliomas injected with genetically modified herpes simplex virus. Clinical cancer research : an official journal of the American Association for Cancer Research. 1999;5(6):1517-22. Epub 1999/07/02.

117. Advani SJ, Sibley GS, Song PY, Hallahan DE, Kataoka Y, Roizman B, et al. Enhancement of replication of genetically engineered herpes simplex viruses by ionizing radiation: a new paradigm for destruction of therapeutically intractable tumors. Gene therapy. 1998;5(2):160-5.

118. Mineta T, Rabkin SD, Yazaki T, Hunter WD, Martuza RL. Attenuated multi-mutated herpes simplex virus-1 for the treatment of malignant gliomas. Nature medicine. 1995;1(9):938-43.

119. Ning J, Wakimoto H. Oncolytic herpes simplex virus-based strategies: toward a breakthrough in glioblastoma therapy. Frontiers in microbiology. 2014;5:303.

120. Markert JM, Medlock MD, Rabkin SD, Gillespie GY, Todo T, Hunter WD, et al. Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene therapy. 2000;7(10):867-74.

121. Stanziale SF, Petrowsky H, Joe JK, Roberts GD, Zager JS, Gusani NJ, et al. Ionizing radiation potentiates the antitumor efficacy of oncolytic herpes simplex virus G207 by upregulating ribonucleotide reductase. Surgery. 2002;132(2):353-9.

122. Markert JM, Razdan SN, Kuo HC, Cantor A, Knoll A, Karrasch M, et al. A phase 1 trial of oncolytic HSV-1, G207, given in combination with radiation for recurrent GBM demonstrates safety and radiographic responses. Molecular therapy : the journal of the American Society of Gene Therapy. 2014;22(5):1048-55. Epub 2014/02/28.

123. Haseley A, Alvarez-Breckenridge C, Chaudhury AR, Kaur B. Advances in oncolytic virus therapy for glioma. Recent patents on CNS drug discovery. 2009;4(1):1-13.

150

124. Aghi M, Rabkin S, Martuza RL. Effect of chemotherapy-induced DNA repair on oncolytic herpes simplex viral replication. Journal of the National Cancer Institute. 2006;98(1):38-50.

125. Toda M, Rabkin SD, Martuza RL. Treatment of human breast cancer in a brain metastatic model by G207, a replication-competent multimutated herpes simplex virus 1. Human gene therapy. 1998;9(15):2177-85.

126. Charles NA, Holland EC, Gilbertson R, Glass R, Kettenmann H. The brain tumor microenvironment. Glia. 2012;60(3):502-14.

127. Kleeberger W, Bova GS, Nielsen ME, Herawi M, Chuang AY, Epstein JI, et al. Roles for the stem cell associated intermediate filament Nestin in prostate cancer migration and metastasis. Cancer research. 2007;67(19):9199-206. Epub 2007/10/03.

128. Liu C, Chen B, Zhu J, Zhang R, Yao F, Jin F, et al. Clinical implications for nestin protein expression in breast cancer. Cancer science. 2010;101(3):815-9. Epub 2009/12/24.

129. Apostolou P, Toloudi M, Chatziioannou M, Ioannou E, Papasotiriou I. Cancer stem cells stemness transcription factors expression correlates with breast cancer disease stage. Current stem cell research & therapy. 2012;7(6):415-9. Epub 2012/10/16.

130. Kambara H, Okano H, Chiocca EA, Saeki Y. An oncolytic HSV-1 mutant expressing ICP34.5 under control of a nestin promoter increases survival of animals even when symptomatic from a brain tumor. Cancer research. 2005;65(7):2832-9.

131. Todo T, Martuza RL, Rabkin SD, Johnson PA. Oncolytic herpes simplex virus vector with enhanced MHC class I presentation and tumor cell killing. Proceedings of the National Academy of Sciences of the United States of America. 2001;98(11):6396-401.

132. Mulvey M, Poppers J, Ladd A, Mohr I. A herpesvirus ribosome-associated, RNA-binding protein confers a growth advantage upon mutants deficient in a GADD34-related function. Journal of virology. 1999;73(4):3375-85.

133. Cassady KA, Gross M. The herpes simplex virus type 1 U(S)11 protein interacts with protein kinase R in infected cells and requires a 30-amino-acid sequence adjacent to a kinase substrate domain. Journal of virology. 2002;76(5):2029-35.

134. Wakimoto H, Kesari S, Farrell CJ, Curry WT, Jr., Zaupa C, Aghi M, et al. Human glioblastoma-derived cancer stem cells: establishment of invasive glioma models and treatment with oncolytic herpes simplex virus vectors. Cancer research. 2009;69(8):3472- 81.

151

135. Kanai R, Rabkin SD, Yip S, Sgubin D, Zaupa CM, Hirose Y, et al. Oncolytic virus- mediated manipulation of DNA damage responses: synergy with chemotherapy in killing glioblastoma stem cells. Journal of the National Cancer Institute. 2012;104(1):42-55. Epub 2011/12/17.

136. Andreansky S, He B, van Cott J, McGhee J, Markert JM, Gillespie GY, et al. Treatment of intracranial gliomas in immunocompetent mice using herpes simplex viruses that express murine interleukins. Gene Ther. 1998;5(1):121-30. Epub 1998/04/16.

137. Ino Y, Saeki Y, Fukuhara H, Todo T. Triple combination of oncolytic herpes simplex virus-1 vectors armed with interleukin-12, interleukin-18, or soluble B7-1 results in enhanced antitumor efficacy. Clinical cancer research : an official journal of the American Association for Cancer Research. 2006;12(2):643-52.

138. Parker JN, Gillespie GY, Love CE, Randall S, Whitley RJ, Markert JM. Engineered herpes simplex virus expressing IL-12 in the treatment of experimental murine brain tumors. Proceedings of the National Academy of Sciences of the United States of America. 2000;97(5):2208-13. Epub 2000/02/19.

139. Cheema TA, Wakimoto H, Fecci PE, Ning J, Kuroda T, Jeyaretna DS, et al. Multifaceted oncolytic virus therapy for glioblastoma in an immunocompetent cancer stem cell model. Proceedings of the National Academy of Sciences of the United States of America. 2013;110(29):12006-11.

140. Liu R, Martuza RL, Rabkin SD. Intracarotid delivery of oncolytic HSV vector G47Delta to metastatic breast cancer in the brain. Gene therapy. 2005;12(8):647-54.

141. Mok W, Stylianopoulos T, Boucher Y, Jain RK. Mathematical modeling of herpes simplex virus distribution in solid tumors: implications for cancer gene therapy. Clinical cancer research : an official journal of the American Association for Cancer Research. 2009;15(7):2352-60.

142. Currier MA, Adams LC, Mahller YY, Cripe TP. Widespread intratumoral virus distribution with fractionated injection enables local control of large human rhabdomyosarcoma xenografts by oncolytic herpes simplex viruses. Cancer gene therapy. 2005;12(4):407-16.

143. Kaur B, Cripe TP, Chiocca EA. "Buy one get one free": armed viruses for the treatment of cancer cells and their microenvironment. Current gene therapy. 2009;9(5):341-55.

144. McKee TD, Grandi P, Mok W, Alexandrakis G, Insin N, Zimmer JP, et al. Degradation of fibrillar collagen in a human melanoma xenograft improves the efficacy of an oncolytic herpes simplex virus vector. Cancer research. 2006;66(5):2509-13.

152

145. Mok W, Boucher Y, Jain RK. Matrix metalloproteinases-1 and -8 improve the distribution and efficacy of an oncolytic virus. Cancer research. 2007;67(22):10664-8.

146. Kim Y, Lee HG, Dmitrieva N, Kim J, Kaur B, Friedman A. Choindroitinase ABC I- mediated enhancement of oncolytic virus spread and anti tumor efficacy: a mathematical model. PloS one. 2014;9(7):e102499.

147. Mahller YY, Vaikunth SS, Ripberger MC, Baird WH, Saeki Y, Cancelas JA, et al. Tissue inhibitor of metalloproteinase-3 via oncolytic herpesvirus inhibits tumor growth and vascular progenitors. Cancer research. 2008;68(4):1170-9.

148. Dmitrieva N, Yu L, Viapiano M, Cripe TP, Chiocca EA, Glorioso JC, et al. Chondroitinase ABC I-mediated enhancement of oncolytic virus spread and antitumor efficacy. Clinical cancer research : an official journal of the American Association for Cancer Research. 2011;17(6):1362-72.

149. Liu TC, Zhang T, Fukuhara H, Kuroda T, Todo T, Martuza RL, et al. Oncolytic HSV armed with platelet factor 4, an antiangiogenic agent, shows enhanced efficacy. Molecular therapy : the journal of the American Society of Gene Therapy. 2006;14(6):789-97.

150. Auguste P, Javerzat S, Bikfalvi A. Regulation of vascular development by fibroblast growth factors. Cell and tissue research. 2003;314(1):157-66.

151. Liu TC, Zhang T, Fukuhara H, Kuroda T, Todo T, Canron X, et al. Dominant-negative fibroblast growth factor receptor expression enhances antitumoral potency of oncolytic herpes simplex virus in neural tumors. Clinical cancer research : an official journal of the American Association for Cancer Research. 2006;12(22):6791-9.

152. Cork SM, Van Meir EG. Emerging roles for the BAI1 protein family in the regulation of phagocytosis, synaptogenesis, neurovasculature, and tumor development. J Mol Med (Berl). 2011;89(8):743-52. Epub 2011/04/22.

153. Kaur B, Cork SM, Sandberg EM, Devi NS, Zhang Z, Klenotic PA, et al. Vasculostatin inhibits intracranial glioma growth and negatively regulates in vivo angiogenesis through a CD36-dependent mechanism. Cancer research. 2009;69(3):1212-20. Epub 2009/01/30.

154. Kudo S, Konda R, Obara W, Kudo D, Tani K, Nakamura Y, et al. Inhibition of tumor growth through suppression of angiogenesis by brain-specific angiogenesis inhibitor 1 gene transfer in murine renal cell carcinoma. Oncology reports. 2007;18(4):785-91. Epub 2007/09/06.

155. Hardcastle J, Kurozumi K, Dmitrieva N, Sayers MP, Ahmad S, Waterman P, et al. Enhanced antitumor efficacy of vasculostatin (Vstat120) expressing oncolytic HSV-1.

153

Molecular therapy : the journal of the American Society of Gene Therapy. 2010;18(2):285-94. Epub 2009/10/22.

156. Yoo JY, Haseley A, Bratasz A, Chiocca EA, Zhang J, Powell K, et al. Antitumor efficacy of 34.5ENVE: a transcriptionally retargeted and "Vstat120"-expressing oncolytic virus. Molecular therapy : the journal of the American Society of Gene Therapy. 2012;20(2):287-97. Epub 2011/10/28.

157. . !!! INVALID CITATION !!!

158. Peng KW, Myers R, Greenslade A, Mader E, Greiner S, Federspiel MJ, et al. Using clinically approved cyclophosphamide regimens to control the humoral immune response to oncolytic viruses. Gene Ther. 2012. Epub 2012/04/06.

159. Kottke T, Thompson J, Diaz RM, Pulido J, Willmon C, Coffey M, et al. Improved systemic delivery of oncolytic reovirus to established tumors using preconditioning with cyclophosphamide-mediated Treg modulation and interleukin-2. Clinical cancer research : an official journal of the American Association for Cancer Research. 2009;15(2):561-9. Epub 2009/01/17.

160. Fulci G, Dmitrieva N, Gianni D, Fontana EJ, Pan X, Lu Y, et al. Depletion of peripheral macrophages and brain microglia increases brain tumor titers of oncolytic viruses. Cancer research. 2007;67(19):9398-406. Epub 2007/10/03.

161. Alvarez-Breckenridge CA, Yu J, Price R, Wojton J, Pradarelli J, Mao H, et al. NK cells impede glioblastoma virotherapy through NKp30 and NKp46 natural cytotoxicity receptors. Nature medicine. 2012;18(12):1827-34. Epub 2012/11/28.

162. Ikeda K, Wakimoto H, Ichikawa T, Jhung S, Hochberg FH, Louis DN, et al. Complement depletion facilitates the infection of multiple brain tumors by an intravascular, replication-conditional herpes simplex virus mutant. Journal of virology. 2000;74(10):4765-75. Epub 2000/04/25.

163. Yoo JY, Pradarelli J, Haseley A, Wojton J, Kaka A, Bratasz A, et al. Copper chelation enhances antitumor efficacy and systemic delivery of oncolytic HSV. Clinical cancer research : an official journal of the American Association for Cancer Research. 2012;18(18):4931-41. Epub 2012/07/04.

164. Bonham M, O'Connor JM, Hannigan BM, Strain JJ. The immune system as a physiological indicator of marginal copper status? The British journal of nutrition. 2002;87(5):393-403. Epub 2002/05/16.

165. Fouladi M. Histone deacetylase inhibitors in cancer therapy. Cancer investigation. 2006;24(5):521-7. Epub 2006/08/31.

154

166. Otsuki A, Patel A, Kasai K, Suzuki M, Kurozumi K, Chiocca EA, et al. Histone deacetylase inhibitors augment antitumor efficacy of herpes-based oncolytic viruses. Molecular therapy : the journal of the American Society of Gene Therapy. 2008;16(9):1546-55. Epub 2008/07/24.

167. Liu TC, Castelo-Branco P, Rabkin SD, Martuza RL. Trichostatin A and oncolytic HSV combination therapy shows enhanced antitumoral and antiangiogenic effects. Molecular therapy : the journal of the American Society of Gene Therapy. 2008;16(6):1041-7. Epub 2008/04/05.

168. Kurozumi K, Hardcastle J, Thakur R, Shroll J, Nowicki M, Otsuki A, et al. Oncolytic HSV-1 infection of tumors induces angiogenesis and upregulates CYR61. Molecular therapy : the journal of the American Society of Gene Therapy. 2008;16(8):1382-91. Epub 2008/06/12.

169. Haseley A, Boone S, Wojton J, Yu L, Yoo JY, Yu J, et al. Extracellular matrix protein CCN1 limits oncolytic efficacy in glioma. Cancer research. 2012;72(6):1353-62. Epub 2012/01/28.

170. Wojton J, Kaur B. Impact of tumor microenvironment on oncolytic viral therapy. Cytokine & growth factor reviews. 2010;21(2-3):127-34. Epub 2010/04/20.

171. Kurozumi K, Hardcastle J, Thakur R, Yang M, Christoforidis G, Fulci G, et al. Effect of tumor microenvironment modulation on the efficacy of oncolytic virus therapy. Journal of the National Cancer Institute. 2007;99(23):1768-81. Epub 2007/11/29.

172. Zhang W, Fulci G, Buhrman JS, Stemmer-Rachamimov AO, Chen JW, Wojtkiewicz GR, et al. Bevacizumab with angiostatin-armed oHSV increases antiangiogenesis and decreases bevacizumab-induced invasion in U87 glioma. Molecular therapy : the journal of the American Society of Gene Therapy. 2012;20(1):37-45. Epub 2011/09/15.

173. Todo T, Rabkin SD, Chahlavi A, Martuza RL. Corticosteroid administration does not affect viral oncolytic activity, but inhibits antitumor immunity in replication-competent herpes simplex virus tumor therapy. Human gene therapy. 1999;10(17):2869-78. Epub 1999/12/10.

174. Fidler IJ, Balasubramanian K, Lin Q, Kim SW, Kim SJ. The brain microenvironment and cancer metastasis. Mol Cells. 2010;30(2):93-8. Epub 2010/08/28.

175. Gonzalez-Angulo AM, Hortobagyi GN. Brain metastases and breast cancer subtypes. Onkologie. 2010;33(4):143-4. Epub 2010/04/15.

176. Network TCGA. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. Epub 2012/09/25.

155

177. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, et al. A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer cell. 2006;10(6):515-27. Epub 2006/12/13.

178. Bos PD, Zhang XH, Nadal C, Shu W, Gomis RR, Nguyen DX, et al. Genes that mediate breast cancer metastasis to the brain. Nature. 2009;459(7249):1005-9. Epub 2009/05/08.

179. Cody JJ, Scaturro P, Cantor AB, Yancey Gillespie G, Parker JN, Markert JM. Preclinical evaluation of oncolytic deltagamma(1)34.5 herpes simplex virus expressing interleukin- 12 for therapy of breast cancer brain metastases. International journal of breast cancer. 2012;2012:628697. Epub 2013/01/25.

180. Borowsky AD, Namba R, Young LJ, Hunter KW, Hodgson JG, Tepper CG, et al. Syngeneic mouse mammary carcinoma cell lines: two closely related cell lines with divergent metastatic behavior. Clinical & experimental metastasis. 2005;22(1):47-59. Epub 2005/09/01.

181. Pei XF, Noble MS, Davoli MA, Rosfjord E, Tilli MT, Furth PA, et al. Explant-cell culture of primary mammary tumors from MMTV-c-Myc transgenic mice. In vitro cellular & developmental biology Animal. 2004;40(1-2):14-21. Epub 2004/06/08.

182. Scott BJ, Kesari S. Leptomeningeal metastases in breast cancer. American journal of cancer research. 2013;3(2):117-26. Epub 2013/04/18.

183. Mendes-Pereira AM, Martin SA, Brough R, McCarthy A, Taylor JR, Kim JS, et al. Synthetic lethal targeting of PTEN mutant cells with PARP inhibitors. EMBO molecular medicine. 2009;1(6-7):315-22.

184. Zabuawala T, Taffany DA, Sharma SM, Merchant A, Adair B, Srinivasan R, et al. An ets2-driven transcriptional program in tumor-associated macrophages promotes tumor metastasis. Cancer research. 2010;70(4):1323-33. Epub 2010/02/11.

185. Wojton J, Chu Z, Mathsyaraja H, Meisen WH, Denton N, Kwon CH, et al. Systemic delivery of SapC-DOPS has antiangiogenic and antitumor effects against glioblastoma. Molecular therapy : the journal of the American Society of Gene Therapy. 2013;21(8):1517-25. Epub 2013/06/05.

186. Wang Y, Jiang T. Understanding high grade glioma: molecular mechanism, therapy and comprehensive management. Cancer letters. 2013;331(2):139-46. Epub 2013/01/24.

187. Mohyeldin A, Chiocca EA. Gene and viral therapy for glioblastoma: a review of clinical trials and future directions. Cancer J. 2012;18(1):82-8. Epub 2012/02/01.

156

188. Russell SJ, Federspiel MJ, Peng KW, Tong C, Dingli D, Morice WG, et al. Remission of disseminated cancer after systemic oncolytic virotherapy. Mayo Clinic proceedings. 2014;89(7):926-33. Epub 2014/05/20.

189. Kaufman HL, Bines SD. OPTIM trial: a Phase III trial of an oncolytic herpes virus encoding GM-CSF for unresectable stage III or IV melanoma. Future Oncol. 2010;6(6):941-9. Epub 2010/06/10.

190. Haralambieva I, Iankov I, Hasegawa K, Harvey M, Russell SJ, Peng KW. Engineering oncolytic measles virus to circumvent the intracellular innate immune response. Molecular therapy : the journal of the American Society of Gene Therapy. 2007;15(3):588-97. Epub 2007/01/25.

191. Ikeda K, Ichikawa T, Wakimoto H, Silver JS, Deisboeck TS, Finkelstein D, et al. Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses. Nature medicine. 1999;5(8):881-7. Epub 1999/07/30.

192. Aravalli RN, Hu S, Rowen TN, Palmquist JM, Lokensgard JR. Cutting edge: TLR2- mediated proinflammatory cytokine and chemokine production by microglial cells in response to herpes simplex virus. J Immunol. 2005;175(7):4189-93. Epub 2005/09/24.

193. Lokensgard JR, Cheeran MC, Hu S, Gekker G, Peterson PK. Glial cell responses to herpesvirus infections: role in defense and immunopathogenesis. The Journal of infectious diseases. 2002;186 Suppl 2:S171-9. Epub 2002/11/09.

194. Hashimoto Y, Moki T, Takizawa T, Shiratsuchi A, Nakanishi Y. Evidence for phagocytosis of influenza virus-infected, apoptotic cells by neutrophils and macrophages in mice. J Immunol. 2007;178(4):2448-57. Epub 2007/02/06.

195. Rock RB, Gekker G, Hu S, Sheng WS, Cheeran M, Lokensgard JR, et al. Role of microglia in central nervous system infections. Clinical microbiology reviews. 2004;17(4):942-64, table of contents. Epub 2004/10/19.

196. Marques CP, Cheeran MC, Palmquist JM, Hu S, Urban SL, Lokensgard JR. Prolonged microglial cell activation and lymphocyte infiltration following experimental herpes encephalitis. J Immunol. 2008;181(9):6417-26. Epub 2008/10/23.

197. Fulci G, Breymann L, Gianni D, Kurozomi K, Rhee SS, Yu J, et al. Cyclophosphamide enhances glioma virotherapy by inhibiting innate immune responses. Proceedings of the National Academy of Sciences of the United States of America. 2006;103(34):12873-8. Epub 2006/08/16.

198. Peng KW, Myers R, Greenslade A, Mader E, Greiner S, Federspiel MJ, et al. Using clinically approved cyclophosphamide regimens to control the humoral immune response to oncolytic viruses. Gene Ther. 2013;20(3):255-61. Epub 2012/04/06.

157

199. Currier MA, Gillespie RA, Sawtell NM, Mahller YY, Stroup G, Collins MH, et al. Efficacy and safety of the oncolytic herpes simplex virus rRp450 alone and combined with cyclophosphamide. Molecular therapy : the journal of the American Society of Gene Therapy. 2008;16(5):879-85. Epub 2008/04/05.

200. Qiao J, Wang H, Kottke T, White C, Twigger K, Diaz RM, et al. Cyclophosphamide facilitates antitumor efficacy against subcutaneous tumors following intravenous delivery of reovirus. Clinical cancer research : an official journal of the American Association for Cancer Research. 2008;14(1):259-69. Epub 2008/01/04.

201. Lun XQ, Jang JH, Tang N, Deng H, Head R, Bell JC, et al. Efficacy of systemically administered oncolytic vaccinia virotherapy for malignant gliomas is enhanced by combination therapy with rapamycin or cyclophosphamide. Clinical cancer research : an official journal of the American Association for Cancer Research. 2009;15(8):2777-88. Epub 2009/04/09.

202. Badie B, Schartner JM. Flow cytometric characterization of tumor-associated macrophages in experimental gliomas. Neurosurgery. 2000;46(4):957-61; discussion 61- 2. Epub 2000/04/14.

203. Morantz RA, Wood GW, Foster M, Clark M, Gollahon K. Macrophages in experimental and human brain tumors. Part 2: studies of the macrophage content of human brain tumors. Journal of neurosurgery. 1979;50(3):305-11. Epub 1979/03/01.

204. Morantz RA, Wood GW, Foster M, Clark M, Gollahon K. Macrophages in experimental and human brain tumors. Part 1: Studies of the macrophage content of experimental rat brain tumors of varying immunogenicity. Journal of neurosurgery. 1979;50(3):298-304. Epub 1979/03/01.

205. Currier MA, Eshun FK, Sholl A, Chernoguz A, Crawford K, Divanovic S, et al. VEGF blockade enables oncolytic cancer virotherapy in part by modulating intratumoral myeloid cells. Molecular therapy : the journal of the American Society of Gene Therapy. 2013;21(5):1014-23. Epub 2013/03/14.

206. Sergerie Y, Rivest S, Boivin G. Tumor necrosis factor-alpha and interleukin-1 beta play a critical role in the resistance against lethal herpes simplex virus encephalitis. The Journal of infectious diseases. 2007;196(6):853-60. Epub 2007/08/19.

207. Kambara H, Saeki Y, Chiocca EA. Cyclophosphamide allows for in vivo dose reduction of a potent oncolytic virus. Cancer research. 2005;65(24):11255-8. Epub 2005/12/17.

208. Lamfers ML, Fulci G, Gianni D, Tang Y, Kurozumi K, Kaur B, et al. Cyclophosphamide increases transgene expression mediated by an oncolytic adenovirus in glioma-bearing mice monitored by bioluminescence imaging. Molecular therapy : the journal of the American Society of Gene Therapy. 2006;14(6):779-88. Epub 2006/09/26. 158

209. Jenne CN, Wong CH, Zemp FJ, McDonald B, Rahman MM, Forsyth PA, et al. Neutrophils recruited to sites of infection protect from virus challenge by releasing neutrophil extracellular traps. Cell host & microbe. 2013;13(2):169-80. Epub 2013/02/19.

210. Morahan PS, Morse SS, McGeorge MG. Macrophage extrinsic antiviral activity during herpes simplex virus infection. The Journal of general virology. 1980;46(2):291-300. Epub 1980/02/01.

211. Kodukula P, Liu T, Rooijen NV, Jager MJ, Hendricks RL. Macrophage control of herpes simplex virus type 1 replication in the peripheral nervous system. J Immunol. 1999;162(5):2895-905. Epub 1999/03/11.

212. Lundberg P, Welander PV, Edwards CK, 3rd, van Rooijen N, Cantin E. Tumor necrosis factor (TNF) protects resistant C57BL/6 mice against herpes simplex virus-induced encephalitis independently of signaling via TNF receptor 1 or 2. Journal of virology. 2007;81(3):1451-60. Epub 2006/11/17.

213. Minami M, Kita M, Yan XQ, Yamamoto T, Iida T, Sekikawa K, et al. Role of IFN- gamma and tumor necrosis factor-alpha in herpes simplex virus type 1 infection. Journal of interferon & cytokine research : the official journal of the International Society for Interferon and Cytokine Research. 2002;22(6):671-6. Epub 2002/08/07.

214. Wong GH, Goeddel DV. Tumour necrosis factors alpha and beta inhibit virus replication and synergize with interferons. Nature. 1986;323(6091):819-22. Epub 1986/10/05.

215. Kulu Y, Kawasaki H, Donahue JM, Kasuya H, Cusack JC, Choi EW, et al. Concurrent chemotherapy inhibits herpes simplex virus-1 replication and oncolysis. Cancer gene therapy. 2013;20(2):133-40. Epub 2013/01/26.

216. Mestan J, Digel W, Mittnacht S, Hillen H, Blohm D, Moller A, et al. Antiviral effects of recombinant tumour necrosis factor in vitro. Nature. 1986;323(6091):816-9. Epub 1986/10/05.

217. Cirino NM, Panuska JR, Villani A, Taraf H, Rebert NA, Merolla R, et al. Restricted replication of respiratory syncytial virus in human alveolar macrophages. The Journal of general virology. 1993;74 ( Pt 8):1527-37. Epub 1993/08/01.

218. Seo SH, Webster RG. Tumor necrosis factor alpha exerts powerful anti-influenza virus effects in lung epithelial cells. Journal of virology. 2002;76(3):1071-6. Epub 2002/01/05.

219. Lokensgard JR, Hu S, Sheng W, vanOijen M, Cox D, Cheeran MC, et al. Robust expression of TNF-alpha, IL-1beta, RANTES, and IP-10 by human microglial cells during nonproductive infection with herpes simplex virus. Journal of neurovirology. 2001;7(3):208-19. Epub 2001/08/23.

159

220. Fajardo LF, Kwan HH, Kowalski J, Prionas SD, Allison AC. Dual role of tumor necrosis factor-alpha in angiogenesis. Am J Pathol. 1992;140(3):539-44. Epub 1992/03/01.

221. Balkwill F. TNF-alpha in promotion and progression of cancer. Cancer metastasis reviews. 2006;25(3):409-16. Epub 2006/09/05.

222. Ryu J, Ku BM, Lee YK, Jeong JY, Kang S, Choi J, et al. Resveratrol reduces TNF-alpha- induced U373MG human glioma cell invasion through regulating NF-kappaB activation and uPA/uPAR expression. Anticancer research. 2011;31(12):4223-30. Epub 2011/12/27.

223. Galvan V, Roizman B. Herpes simplex virus 1 induces and blocks apoptosis at multiple steps during infection and protects cells from exogenous inducers in a cell-type- dependent manner. Proceedings of the National Academy of Sciences of the United States of America. 1998;95(7):3931-6. Epub 1998/05/09.

224. Han ZQ, Assenberg M, Liu BL, Wang YB, Simpson G, Thomas S, et al. Development of a second-generation oncolytic Herpes simplex virus expressing TNFalpha for cancer therapy. The journal of gene medicine. 2007;9(2):99-106. Epub 2007/01/30.

225. Candelario-Jalil E, Taheri S, Yang Y, Sood R, Grossetete M, Estrada EY, et al. Cyclooxygenase inhibition limits blood-brain barrier disruption following intracerebral injection of tumor necrosis factor-alpha in the rat. The Journal of pharmacology and experimental therapeutics. 2007;323(2):488-98. Epub 2007/08/21.

226. Boivin N, Menasria R, Piret J, Rivest S, Boivin G. The combination of valacyclovir with an anti-TNF alpha antibody increases survival rate compared to antiviral therapy alone in a murine model of herpes simplex virus encephalitis. Antiviral research. 2013;100(3):649-53. Epub 2014/01/15.

227. Multhoff G, Radons J. Radiation, inflammation, and immune responses in cancer. Frontiers in oncology. 2012;2:58. Epub 2012/06/08.

228. Liebner S, Fischmann A, Rascher G, Duffner F, Grote EH, Kalbacher H, et al. Claudin-1 and claudin-5 expression and tight junction morphology are altered in blood vessels of human glioblastoma multiforme. Acta neuropathologica. 2000;100(3):323-31. Epub 2000/08/31.

229. Wolburg H, Noell S, Fallier-Becker P, Mack AF, Wolburg-Buchholz K. The disturbed blood-brain barrier in human glioblastoma. Molecular aspects of medicine. 2012;33(5- 6):579-89. Epub 2012/03/06.

230. Meisen WH, Wohleb ES, Jaime-Ramirez AC, Bolyard CM, Yoo JY, Russell L, et al. The impact of macrophage and microglia secreted TNFalpha on oncolytic HSV-1 therapy in the glioblastoma tumor microenvironment. Clinical cancer research : an official journal of the American Association for Cancer Research. 2015. Epub 2015/04/02.

160

231. Meisen WH, Kaur B. How can we trick the immune system into overcoming the detrimental effects of oncolytic viral therapy to treat glioblastoma? Expert review of neurotherapeutics. 2013;13(4):341-3. Epub 2013/04/03.

232. Henry CJ, Huang Y, Wynne A, Hanke M, Himler J, Bailey MT, et al. Minocycline attenuates lipopolysaccharide (LPS)-induced neuroinflammation, sickness behavior, and anhedonia. J Neuroinflammation. 2008;5:15. Epub 2008/05/15.

233. Huang Y, Henry CJ, Dantzer R, Johnson RW, Godbout JP. Exaggerated sickness behavior and brain proinflammatory cytokine expression in aged mice in response to intracerebroventricular lipopolysaccharide. Neurobiol Aging. 2008;29(11):1744-53. Epub 2007/06/05.

234. Henry CJ, Huang Y, Wynne AM, Godbout JP. Peripheral lipopolysaccharide (LPS) challenge promotes microglial hyperactivity in aged mice that is associated with exaggerated induction of both pro-inflammatory IL-1beta and anti-inflammatory IL-10 cytokines. Brain Behav Immun. 2009;23(3):309-17. Epub 2008/09/26.

235. Wynne AM, Henry CJ, Huang Y, Cleland A, Godbout JP. Protracted downregulation of CX3CR1 on microglia of aged mice after lipopolysaccharide challenge. Brain Behav Immun.24(7):1190-201. Epub 2010/06/24.

236. Frank MG, Wieseler-Frank JL, Watkins LR, Maier SF. Rapid isolation of highly enriched and quiescent microglia from adult rat hippocampus: immunophenotypic and functional characteristics. J Neurosci Methods. 2006;151(2):121-30. Epub 2005/08/30.

237. Wohleb ES, Hanke ML, Corona AW, Powell ND, Stiner LM, Bailey MT, et al. beta- Adrenergic receptor antagonism prevents anxiety-like behavior and microglial reactivity induced by repeated social defeat. J Neurosci.31(17):6277-88. Epub 2011/04/29.

238. Fang H, Pengal RA, Cao X, Ganesan LP, Wewers MD, Marsh CB, et al. Lipopolysaccharide-induced macrophage inflammatory response is regulated by SHIP. J Immunol. 2004;173(1):360-6. Epub 2004/06/24.

239. Sone ED, Zubarev ER, Stupp SI. Supramolecular templating of single and double nanohelices of cadmium sulfide. Small. 2005;1(7):694-7. Epub 2006/12/29.

240. Fomchenko EI, Holland EC. Mouse models of brain tumors and their applications in preclinical trials. Clinical cancer research : an official journal of the American Association for Cancer Research. 2006;12(18):5288-97. Epub 2006/09/27.

241. Edinger M, Sweeney TJ, Tucker AA, Olomu AB, Negrin RS, Contag CH. Noninvasive assessment of tumor cell proliferation in animal models. Neoplasia. 1999;1(4):303-10. Epub 2000/08/10.

161

242. Dinca EB, Sarkaria JN, Schroeder MA, Carlson BL, Voicu R, Gupta N, et al. Bioluminescence monitoring of intracranial glioblastoma xenograft: response to primary and salvage temozolomide therapy. Journal of neurosurgery. 2007;107(3):610-6. Epub 2007/09/25.

243. Contag CH, Jenkins D, Contag PR, Negrin RS. Use of reporter genes for optical measurements of neoplastic disease in vivo. Neoplasia. 2000;2(1-2):41-52. Epub 2000/08/10.

244. Sweeney TJ, Mailander V, Tucker AA, Olomu AB, Zhang W, Cao Y, et al. Visualizing the kinetics of tumor-cell clearance in living animals. Proceedings of the National Academy of Sciences of the United States of America. 1999;96(21):12044-9. Epub 1999/10/16.

245. O'Neill K, Lyons SK, Gallagher WM, Curran KM, Byrne AT. Bioluminescent imaging: a critical tool in pre-clinical oncology research. The Journal of pathology. 2010;220(3):317-27. Epub 2009/12/08.

246. Henson JW, Ulmer S, Harris GJ. Brain tumor imaging in clinical trials. AJNR American journal of neuroradiology. 2008;29(3):419-24. Epub 2008/02/15.

247. Lee J, Kotliarova S, Kotliarov Y, Li A, Su Q, Donin NM, et al. Tumor stem cells derived from glioblastomas cultured in bFGF and EGF more closely mirror the phenotype and genotype of primary tumors than do serum-cultured cell lines. Cancer cell. 2006;9(5):391-403. Epub 2006/05/16.

248. De Vocht N, Reekmans K, Bergwerf I, Praet J, Hoornaert C, Le Blon D, et al. Multimodal imaging of stem cell implantation in the central nervous system of mice. Journal of visualized experiments : JoVE. 2012(64):e3906.

249. Szentirmai O, Baker CH, Lin N, Szucs S, Takahashi M, Kiryu S, et al. Noninvasive bioluminescence imaging of luciferase expressing intracranial U87 xenografts: correlation with magnetic resonance imaging determined tumor volume and longitudinal use in assessing tumor growth and antiangiogenic treatment effect. Neurosurgery. 2006;58(2):365-72; discussion -72. Epub 2006/02/08.

250. Jost SC, Collins L, Travers S, Piwnica-Worms D, Garbow JR. Measuring brain tumor growth: combined bioluminescence imaging-magnetic resonance imaging strategy. Molecular imaging. 2009;8(5):245-53. Epub 2009/10/03.

251. Mittra ES, Fan-Minogue H, Lin FI, Karamchandani J, Sriram V, Han M, et al. Preclinical efficacy of the anti-hepatocyte growth factor antibody ficlatuzumab in a mouse brain orthotopic glioma model evaluated by bioluminescence, PET, and MRI. Clinical cancer research : an official journal of the American Association for Cancer Research. 2013;19(20):5711-21. Epub 2013/08/29. 162

252. Rehemtulla A, Stegman LD, Cardozo SJ, Gupta S, Hall DE, Contag CH, et al. Rapid and quantitative assessment of cancer treatment response using in vivo bioluminescence imaging. Neoplasia. 2000;2(6):491-5. Epub 2001/03/03.

253. Yamada S, Khankaldyyan V, Bu X, Suzuki A, Gonzalez-Gomez I, Takahashi K, et al. A method to accurately inject tumor cells into the caudate/putamen nuclei of the mouse brain. The Tokai journal of experimental and clinical medicine. 2004;29(4):167-73. Epub 2005/02/19.

254. Evans MS, Chaurette JP, Adams ST, Jr., Reddy GR, Paley MA, Aronin N, et al. A synthetic luciferin improves bioluminescence imaging in live mice. Nature methods. 2014;11(4):393-5. Epub 2014/02/11.

255. Ray P, Wu AM, Gambhir SS. Optical bioluminescence and positron emission tomography imaging of a novel fusion reporter gene in tumor xenografts of living mice. Cancer research. 2003;63(6):1160-5. Epub 2003/03/22.

256. Amber Miller LS, Shruthi Naik, Rebecca Nace, Mark Federspiel, Kah Whye Peng, Stephen J Russell. Reporter gene imaging identifies intratumoral infection voids as a critical barrier to systemic oncolytic virus efficacy. Molecular Therapy — Oncolytics 2014;1(1).

163