Preclinical Evaluation of Oral Metronomic Topotecan

Home , Canertinib, Tegafur

PRECLINICAL EVALUATION OF ORAL METRONOMIC TOPOTECAN

AND PAZOPANIB FOR THE TREATMENT OF AGGRESSIVE

EXTRACRANIAL PEDIATRIC SOLID TUMORS

Sushil Kumar

A thesis submitted in conformity with the requirements for the degree of

Doctor of Philosophy

Institute of Medical Sciences, Faculty of Medicine

University of Toronto

Toronto, Ontario, Canada

PRECLINICAL EVALUATION OF ORAL METRONOMIC TOPOTECAN AND PAZOPANIB FOR THE TREATMENT OF AGGRESSIVE EXTRACRANIAL PEDIATRIC SOLID TUMORS

Sushil Kumar

Doctor of Philosophy, 2013

Institute of Medical Sciences, University of Toronto

Abstract

Low Dose Metronomic (LDM) chemotherapy, combined with VEGF pathway inhibitors, is a highly effective strategy to coordinately inhibit angiogenesis and tumor growth. We have tested the efficacies of daily oral LDM topotecan alone and in combination with pazopanib, in three pediatric extracranial solid tumors mouse models. We also investigated the effect of prolonged combination therapy with the combination on tumor behavior in a neuroblastoma mouse xenograft model.

In-vitro dose-response study of topotecan and pazopanib was conducted on several cell lines. In-vivo antitumor efficacies of drugs, as single agents and combination, were tested in immunodeficient mice models. For studying the mechanisms of resistance to our therapy, a time-response study (28, 56 and 80 days) was conducted in SK-N-BE(2) xenografts model, treated in same way as earlier.

In vitro, topotecan caused a dose-dependent decrease in viabilities of all cell lines, while pazopanib did not. In vivo, the combination of topotecan and pazopanib demonstrated significant anti-tumor activity compared to the respective single agents

in all models. Reductions in the levels of viable Circulating Endothelial Progenitors and/or Circulating Endothelial Cells and tumor microvessel density were correlated with tumor response and therefore confirmed the antiangiogenic activity of the regimens. However, the combination also caused significantly higher myelotoxicity than single agents. Pharmacokinetic study did not reveal any interaction between the two co-administered drugs.

In the time-response study, we found that only combination treated animals survived till 80 days. However, tumors in these animals started growing gradually after 50 days. Unlike single agents, all three durations of combination treatment significantly lowered tumor microvessel densities, compared to the control. However, tumors treated with the combination for 56 and 80 days had higher pericyte coverage. The combination increased the hypoxia, angiogenic expression and proliferative index and caused metabolic reprogramming of tumor cells.

We conclude that the combination of LDM topotecan and pazopanib has superior efficacy than either single agents, which is attributed to superior antiangiogenic activity. However, prolonged treatment with the combination can have additive myelotoxicity and may encounter adaptive resistance associated with metabolic reprogramming and increased proliferation of the tumor cells.

iii

Acknowledgments

I express my profound gratitude to my PhD supervisor Dr. Sylvain Baruchel for accepting me as his student and for entrusting me with a research project of great translational relevance. His continuous guidance and encouragement had not only helped me with this project, but also gave me confidence which will prove beneficial in all my future endeavours.

I am extremely thankful to Dr. Herman Yeger who helped me enhance my knowledge and experimental skills in the area of molecular and cellular biology and gave me the opportunity to access the facilities available in his lab. Dr. Yeger‟s advices were very crucial in the successful completion of my research project.

I am extremely thankful to my Program Advisory Committee members, Dr. Yigal Dror and Dr Shinya Ito for their expert advice and comments, especially with regards to

Clinical Pharmacology & Toxicology and Pathobiology of cancer.

I thank my dearest friend and fellow PhD student Mr. Reza Bayat Mokhtari who had taught me several experimental techniques relevant to my research and also supported me during many difficult phases of my project. I thank my lab members, Dr. Libo Zhang and Ms. Bing Wu for helping me and giving positive feedbacks regarding my work. I am thankful to Drs. Kai Du, Karen Aitken and Syed Islam for their scientific discussions and valuable suggestions. I also thank our ex-lab member, Dr. Diana Stempak for helping me during the initial days of my PhD program.

I hereby acknowledge the generous financial supports from James Fund for

Neuroblastoma Research, Dr. Dina Gordon Malkin/Queen Elizabeth-II Scholarship in

Science and Technology (University of Toronto) and Glaxosmithkline, Collegeville, PA, for my PhD program. The departments, Analytical Facility for Bioactive Molecules,

Pathology Research Laboratory and the Lab Animal Services Facility at the Hospital for

Sick Children have substantially contributed to this research project.

Finally, I am grateful to my parents Mr. K.K. Muralidharan and Mrs. Savitri Kurup and for their unconditional love and encouragement and for nurturing ambition in me. I am thankful to my father-in-law Mr. T.R.S. Nair and mother-in-law Mrs. Sushila Nair for all their best wishes. I would never have completed my PhD without the career advices and support from my wife Sunita during every situation. I am thankful to my brother Sujit who always looked up to me and made me feel special. I am pleased to have my little daughter Shreya in my life who is always there to make me smile during the difficult times.

Above all, I thank God for showing me the right direction and giving me all the wonderful people who have helped me achieve my career goal.

Table of Contents

Abstract ...... ii

Acknowledgments ...... iv

Table of Contents ...... vi

List of Tables ...... xi

List of Figures ...... xii

List of Abbreviations ...... xiv

1 Introduction ...... 1

1.1 Pediatric cancers ...... 1

1.2 Treatment options for extracranial pediatric solid tumors ...... 2

1.3 Limitations of conventional chemotherapy ...... 7

1.3.1 Acute and chronic adverse effects:...... 7

1.3.2 Drug resistance ...... 11

1.4 Recurrence and relapse in pediatric cancers ...... 13

1.4.1 The role of Minimal residual disease in cancer relapses ...... 14

1.4.2 Maintenance therapy ...... 14

1.5 Angiogenesis: a newer target for maintenance chemotherapy ...... 15

1.5.1 Modes of angiogenesis ...... 16

1.5.2 Molecular mechanisms of sprouting angiogenesis and vasculogenesis ... 17

1.6 Angiogenesis and angiogenic factors in pediatric solid tumors ...... 23

1.7 Antiangiogenic strategies ...... 26

1.8 Targeted antiangiogenic agents ...... 27

1.8.1 Classification ...... 27

1.8.2 Anti-VEGF antiangiogenic agents...... 29

1.8.3 Anti-VEGF therapy in pediatric cancers ...... 31

1.9 Low Dose Metronomic Chemotherapy ...... 32

1.9.1 Advantages of LDM chemotherapy over conventional MTD (Maximum Tolerated Dose) chemotherapy ...... 33

1.9.2 Mechanisms of LDM chemotherapy: ...... 34

1.9.3 Examples of efficacies of single agent LDM chemotherapy ...... 37

1.10 Combination of LDM chemotherapy and antiangiogenic agents ...... 43

1.10.1 Rationale for combining LDM chemotherapy with antiangiogenic agents ...... 43

1.10.2 Pioneering studies on the combination of LDM chemotherapy and antiangiogenic agents ...... 44

1.10.3 Previous works on LDM chemotherapy with antiangiogenic drugs ...... 45

1.10.4 Induction of tumor dormancy by maintenance therapy involving metronomic scheduling ...... 51

1.11 Adverse effects of antiangiogenic therapy ...... 52

1.12 Topotecan ...... 54

1.12.1 Chemistry ...... 54

1.12.2 Mechanism of action ...... 54

1.12.3 Pharmacokinetics and metabolism ...... 55

1.12.4 Indications and adverse effects ...... 55

1.12.5 Resistance to topotecan ...... 56

1.12.6 Topotecan in pediatric cancers ...... 56

1.12.7 Antiangiogenic property of topotecan ...... 61

1.12.8 Metronomic topotecan ...... 61

1.13 Pazopanib ...... 62

1.13.1 Chemistry ...... 62

1.13.2 Mechanism of action ...... 62

1.13.3 Pharmacokinetics and metabolism ...... 63

1.13.4 Preclinical antitumor activity ...... 63 vii

1.13.5 Clinical activity of pazopanib ...... 63

1.13.6 Comparison of pazopanib with other VEGF receptor inhibitors ...... 70

1.14 Biomarkers of antiangiogenic therapy ...... 70

1.14.1 Need for markers of antiangiogenic therapy ...... 71

1.14.2 Common biomarkers of antiangiogenic therapy ...... 72

1.15 Resistance mechanisms to antiangiogenic therapy ...... 79

1.15.1 Up-regulation of angiogenic factors ...... 79

1.15.2 Involvement of bone marrow derived cells ...... 80

1.15.3 Pericyte coverage ...... 81

1.15.4 Increased invasive potential of tumor cells ...... 82

1.15.5 Metabolic switch of tumor cells ...... 82

1.15.6 Involvement of cancer stem cells...... 83

2 Thesis overview ...... 84

2.1 Rationale ...... 84

2.2 Hypothesis ...... 86

2.3 Objectives ...... 86

2.4 Thesis outline ...... 87

2.4.1 To test the efficacy and safety of LDM topotecan and pazopanib in immunodeficient mice models of extracranial pediatric solid tumors: neuroblastoma, rhabdomyosarcoma and osteosarcoma...... 87

2.4.2 To conduct a time–response study to investigate the changes in tumor xenograft behavior in response to prolonged therapy with LDM topotecan and pazopanib in a neuroblastoma mice model...... 88

3 Materials ...... 89

3.1 Drugs and reagents ...... 89

3.2 Cell lines ...... 89

4 Effectiveness of LDM topotecan and pazopanib in mouse models immunodeficient mice models of extracranial pediatric solid tumors: neuroblastoma, rhabdomyosarcoma and osteosarcoma...... 90 viii

4.1 Methods ...... 90

4.1.1 In-vitro cytotoxicity ...... 90

4.1.2 In-vivo evaluation of topotecan and pazopanib ...... 91

4.1.3 Immunohistochemistry and histopathology ...... 92

4.1.4 Analysis of CEPs and CECs by flow cytometry ...... 93

4.1.5 Bone marrow progenitor assay ...... 94

4.1.6 PK of topotecan and pazopanib ...... 94

4.1.7 Statistical analysis ...... 96

4.2 Results ...... 97

4.2.1 Drug-induced in vitro cytotoxities ...... 97

4.2.2 LDM topotecan and pazopanib in neuroblastoma mouse models ...... 101

4.2.3 Effect of LDM topotecan and pazopanib on the tumor growth in sarcoma models ...... 107

4.2.4 Effect of treatment on tumor microvessel densities ...... 110

4.2.5 Effect of the treatments on CECs and CEPs ...... 113

4.2.6 Safety of TP and PZ in mice ...... 116

4.2.7 PK did not reveal drug interaction between topotecan and pazopanib in TP+PZ group ...... 118

4.3 Discussion ...... 121

5 To conduct a time–response study to investigate the changes in tumor xenograft behavior in response to prolonged therapy with LDM topotecan and pazopanib in a neuroblastoma mice model...... 131

5.1 Methods ...... 133

5.1.1 In-vivo tumor treatment ...... 133

5.1.2 Immunohistochemistry ...... 133

5.1.3 Western blot ...... 134

5.1.4 Real Time PCR...... 135

5.1.5 Statistics ...... 136 ix

5.2 Results ...... 137

5.2.1 Treatment with TP and PZ ...... 137

5.2.2 The effect of treatments on the apoptosis ...... 139

5.2.3 Effect of treatments on hypoxia and angiogenic gene expression ...... 141

5.2.4 The effect of treatments on tumor vessel density and pericyte coverage144

5.2.5 The effect of treatments on proliferative index and mitotic index of tumor cells ...... 146

5.2.6 Effect of treatments on indicators of elevated glycolysis in the tumor cells ...... 149

5.3 Discussion ...... 152

6 Summary and Conclusions ...... 168

6.1 Thesis Summary ...... 168

6.2 Thesis conclusions ...... 169

7 Future Directions ...... 171

8 Limitations of this study ...... 173

9 References ...... 175

List of Tables

Table 1: Standard therapy for pediatric solid tumors ...... 3

Table 2: Long-term adverse effects of pediatric cancer therapy ...... 9

Table 3: Preclinical examples of single agents LDM chemotherapy ...... 37

Table 4:Examples of clinical trials for LDM chemotherapy used alone ...... 40

Table 5: Recent preclinical examples of combination of LDM chemotherapy with antiangiogenic drugs ...... 46

Table 6: Recent clinical trials involving combination of LDM chemotherapy with antiangiogenic drugs ...... 48

Table 7: Clinical trials of conventional topotecan in pediatric cancers ...... 58

Table 8: List of clinical trials involving single agent pazopanib ...... 65

Table 9:The cell lines used for in-vitro experiment and their characteristics ...... 90

List of Figures

Figure 1: Regulation of HIF-1alpha and VEGF in tumor cells...... 19

Figure 2: The various molecules involved in tumor angiogenesis...... 22

Figure 3 The classification of antiangiogenic agents based upon the mechanism of action ...... 28

Figure 5: In-vitro dose- response of topotecan and pazopanib on HUVEC ...... 98

Figure 6: In-vitro dose-response of pazopanib on tumor cell lines ...... 98

Figure 7: In-vitro dose-response study of topotecan alone and in combination with 5000 ng/ml pazopanib on neuroblastoma cell lines ...... 99

Figure 8: In-vitro dose-response study of topotecan alone and in combination with 5000 ng/ml pazopanib on sarcoma cell lines ...... 100

Figure 9: In-vivo efficacy of metronomic topotecan and pazopanib in SK-N-BE(2) subcutaneous xenograft model ...... 103

Figure 10: In-vivo efficacy TP and PZ in SH-SY5Y subcutaneous xenograft model...... 104

Figure 11: In-vivo efficacy of TP and PZ in NUB-7 IV metastatic neuroblastoma model...... 105

Figure 12: In-vivo efficacy of TP and PZ in BE(2)-c IV metastatic neuroblastoma model...... 106

Figure 13: Efficacy of TP and PZ in osteosarcoma KHOS subcutaneous xenograft model...... 108

Figure 14: Efficacy of TP and PZ in rhabdomyosarcoma RH30 subcutaneous xenograft model...... 109

xii

Figure 15: Effect of treatment with TP and PZ on tumor microvessel density in neuroblastoma mice model...... 111

Figure 16: Effect of treatment with TP and PZ on tumor microvessel density of tumors in sarcoma models...... 112

Figure 17: The effect of the treatment regimens on CEP levels in blood ...... 114

Figure 18: The effect of the treatment regimens on CEC levels in blood...... 115

Figure 19: Toxic effect of the treatment regimens bone marrow ...... 117

Figure 20: Plasma concentration-time profiles of topotecan ...... 119

Figure 21: Plasma concentration-time profiles of pazopanib ...... 120

Figure 22: Effect of treatment on tumor growth and survival...... 138

Figure 23: Apoptosis in the tumor tissues after treatment ...... 140

Figure 24: Hypoxia and angiogenic gene expression in tumors after treatment ...... 142

Figure 25: The effect of treatments on tumor vasculature ...... 145

Figure 26: The effect of treatments on proliferation and mitotic index of tumors. ....147

Figure 27: The effect of treatments on markers of glycolysis...... 150

Figure 28: The downstream targets of HIF-1 alpha...... 157

Figure 29: Diagrammatic representation of glucose metabolism under normoxia and hypoxia ...... 163

xiii

List of Abbreviations

ABC ATP Binding Cassette

ALK Analplastic Lymphoma Kinase

ALL Acute Lymphoblastic Leukemia

ALT Alanine Amino Transferase

AMEM Alpha Minimum Essential Medium

Ang Angiopoeitin

AST Aspartate Amino Transferase

BCRP Breast Cancer Resistant Protein bFGF Basic Fibroblast Growth Factor

CAIX Carbonic Anhydrase-IX

CDK Cyclin Dependent Kinase

CDKI Cyclin Dependent Kinase Inhibitor

CEC Circulating Endothelial Cells

CEP Circulating Endothelial Progenitors

CFU-GM Colony Forming Units-Granulocytes/Macrophages

xiv

CNS Central Nervous System

COG Children‟s Oncology Group

COMBAT Combined Oral Maintenance Biodifferentiating and Antiangiogenic

Therapy

CR Complete Response

DMEM Dulbecco Minimum Essential Medium

EFS Event Free Survival

EGFR Epidermal Growth Factor Receptor

ERK Extracellular Signal Regulated Kinase

FACS Fluorescent Activated Cell Sorting

FBS Fetal Bovine Serum

FGF Fibroblast Growth Factor

GI Gastrointestinal

GIST Gastrointestinal Stromal Tumor

HCC Hepatocellular Carcinoma

HIF Hypoxia Inducible Factor

HRP Horse Radish Peroxidases

HUVEC Human Umbilical Vein Endothelial Cells

IAP Inhibitor of Apoptosis

LDM Low Dose Metronomic

LRP Lung Resistance Protein

MAPK Mitogen Activated Protein Kinase

MDR Multidrug Resistant

MMP Matrix Metalloproteases

MTD Maximum Tolerated Dose mTOR Mammalian Target of Rapamycin

NOD/SCID Non-Obese Diabetic/ Severely Combined Immunodeficient

NSCLC Non-Small Cell Lung Cancer

OBD Optimal Biological Dose

OS Overall Survival

PBS Phosphate Buffered Saline

PBST PBS containing 0.1% tween 20

PDGF Platelet Derived Growth Factor

PDGFR Platelet Derived Growth Factor Receptor

xvi

PFS Progression Free Survival

PgP P-Glycoprotein

PH3 Phospho Histone H3

PHD Prolyl Hydroxylases

PI3K Phosphatidylinositol-3-Kinase

PK Pharmacokinetic

PlGF Placental Growth Factor

PPTP Pediatric Preclinical Testing Program

PR Partial Response

PZ Pazopanib regimen, 150 mgéKg, daily, orally

RCC Renal Cell Carcinoma

RTKI Receptor Tyrosine Kinase Inhibitor

RT-PCR Real Time Polymerase Chain Reaction

SD Stable Disease

SDF-1α Stromal Derived Factor-1α siCAM Soluble intercellular Adhesion Molecule

SMA α-Smooth Muscle Actin

xvii

TP LDM Topotecan, 1 mg/Kg, daily, orally

TP+PZ Combination of LDM topotecan and pazopanib

TSP-1 Thrombospondin-1 uPA Urokinase Type Plaminogen Activator vCAM Vascular Cell Adhesion Molecule

VEGF Vascular Endothelial Growth Factor

VEGFR Vascular Endothelial Growth Factor Receptor

VHL Von Hippel Lindau vWF von Willebrand Factor

WBC White Blood Cells

xviii 1

1 Introduction

1.1 Pediatric cancers

Cancer is the second leading cause of deaths in Canada and the leading cause of death in US among children under the age of 14 [1, 2]. Even though cancer-related incidence and mortality is negligible compared to adult cancers, it has significant social impact. The most common childhood cancers in children and adolescents are leukemia (31%), brain/CNS cancers (16.6%), neuroblastoma (7.8%), bone cancers

(6%) and soft-tissue sarcoma (7.4%) [2]. The survival in pediatric cancer patients has increased substantially since 1970s, despite increase in the incidence of invasive cancers in children (environmental exposure being a suspected cause). In

US, the 5-year survival rate has increased from 58.1% in 1975-77 to 79.6% in 1996-

2003 [2]. In Canada, 5-year survival rate has increased from 71% in 1980 to 82% in early 2000s [1]. This increase in survival is attributed to advancements in treatments which led to cure or long-term remission in childhood cancers. Among childhood cancers, as per 1995 statistics, leukemia accounted for highest mortality (34%) followed by CNS cancers (25%). The death rate in leukemia and CNS cancers has fallen by 50% and 23%, respectively, in 1975-1995 [2]. Whereas, the survival in pediatric cancers such as Hodgkin disease, retinoblastoma, Wilm‟s tumor and germ cell tumors are > 90%, the survival in some cancers are still much lower than the average survival rate for all pediatric cancers. These are neuroblastoma (70%), bone cancers (63%), rhabdomyosarcoma (64%). Though low risk diseases among these cancers can be treated successfully, the advanced stage and recurrent diseases

2 attribute to the low survival rates, despite the latest multimodal therapy. Therefore newer strategies need to be explored to treat the advanced or high-risk diseases.

1.2 Treatment options for extracranial pediatric solid tumors

The advent of multimodal therapy consisting of radiation, surgery, chemotherapy and stem cell transplant has tremendously improved the long-term survival of pediatric patients. The standard treatment options for some pediatric solid tumors are described in Table 1 [2].

Table 1: Standard therapy for pediatric solid tumors (OS: Overall Survival; EFS: Event Free Survival)

Ewing Sarcoma Local (5 yr EFS Surgery, radiation, chemotherapy

and OS: 70%) (Vincristine+doxorubicin+cyclophos-

phamide; ifosfamide+etoposide)

Metastatic (6 yr Chemotherapy, local control with

EFS : 28% and surgery and/or radiation therapy

OS: 30% (vincristine+doxorubicin+cyclophosph

amide; ifosfamide+etoposide)

Recurrent (5 yr Chemotherapy (cyclophosphamide+

survival: 10-15%) topotecan; irinotecan+temozolomide)

New molecular targeted therapy

Osteosarcoma Localized (OS: Surgery, preoperative chemotherapy

65-70%) (Combination of high-dose

methotrexate, doxorubicin, cisplatin,

ifosfamide, etoposide, carboplatin)

Metastatic/ Surgical resection; pre- and post-

recurrent (OS: operative chemotherapy(high-dose

<25%) methotrexate, doxorubicin, cisplatin,

high-dose ifosfamide, etoposide,

carboplatin, cyclophosphamide)

Rhabdomyo- Previously Radiation, surgery (preferred only if sarcoma untreated (OS: structural/functional integrity is

>90%) preserved), chemotherapy

(vincristine+dactinomycin)

Intermediate/ Radiation; chemotherapy

high risk (OS < (vincristine+ dactinomycin+

50%) cyclophosphamide)

Recurrent Chemotherapy

(OS<50%) (carboplatin+etoposide;

ifosfamide+carboplatin+etoposide;

cyclophosphamide+topotecan;

irinitecan+ vincristine;

gemcitabine+docetaxel; topotecan+

vincristine+ doxorubicin)

New molecular targeted therapy

Neuroblastoma Low-risk Observation only in low risk and

(includes INSS stage 4S patients

stages 1, 2 and In other low risk patients Surgery 4s, non MYCN alone; Surgery with chemotherapy amplified; OS > (moderate dose of carboplatin, 90%) cyclophosphamide, doxorubicin or

etoposide

Intermediate risk Surgery followed by chemotherapy

(Stages 2B, 3 (carboplatin, cyclophosphamide,

and 4s; non- doxorubicin or etoposide)

MYCN amplified;

OS:85%)

High risk (Stages 1)Neoadjuvant chemotherapy (high-

2, 3 and 4 MYCN dose cyclophosphamide, doxorubicin,

amplified; OS ≈ etoposide, carboplatin, cisplatin,

30%) vincristine, topotecan, ifosfamide)

2) Post-surgical myeloabalative

chemotherapy with stem cell

transplant.

3) Radiation

4) 3-cis retinoic acid and

immunotherapy

Retinoblastoma Intraocular Enucleation, radiation, local therapy,

systemic chemotherapy

(vincristine+carboplatin+etoposide);

subconjuctival chemotherapy

(carboplatin); ophthalmic artery

infusion (melphalan, topotecan,

carboplatin)

Extraocular Chemotherapy (Vincristine,

cyclophosphamide, doxorubicin,

platinum, epipodophylotoxin;

methotrxate, cytarabine,

hydrocortisone)

Recurrent Local radiation; aggressive

chemotherapy

Hepatoblastoma Stage I/II Surgery, chemotherapy (doxorubicin;

(OS>90%) cisplatin+vincristine+fluorouracil;

cisplatin+ doxorubicin; cisplatin)

Stage III Surgery, pre-operative chemotherapy

(OS: 60%-65%) (cisplatin+doxorubicin; cisplatin+

vincristine+fluorouracil; ifosfamide+

cisplatin+doxorubicin)

Stage IV (OS: Surgery, pre-operative chemotherapy

20%-60% (cisplatin+vincristine+fluorouracil;

cisplatin+ doxorubicin)

Liver transplantation in selected

indications

1.3 Limitations of conventional chemotherapy

Despite the overall improvement of survival in pediatric cancers, the high-risk or metastatic disease still presents worse prognosis in some pediatric cancers. The drawbacks of conventional chemotherapy are acute and chronic adverse effects, and the development of drug resistance.

1.3.1 Acute and chronic adverse effects:

Dose-limiting acute toxicities and long-term adverse effects are the major drawbacks of conventional chemotherapy. Anticancer therapy has various short-term toxicity and long-term sequelae.

Acute toxicity: Majority of acute toxicities of cancer chemotherapeutics is due to the fact that drugs non-specifically target all types of proliferating cells, including the normal non-cancerous cells [3, 4].

 Immunosuppression and myelosuppression: Cytotoxic drugs target the

bone marrow cells leading to haematological side effects such as

thrombocytopenia, neutropenia and anemia. Immune suppression renders

the patient susceptible to severe infections.

 Gastrointestinal (GI) side effects: Cytotoxic agents targeting the

proliferating cells of GI mucosa cause nausea, vomiting and mucositis.

Chronic adverse effects: Though the advances in treatments have resulted in ≈ 80% pediatric cancer patients surviving into adulthood, the survivors are vulnerable to long-term morbidity compared to healthy individuals of the same age group [5]. Many of these chronic effects are not apparent until several years after the completion of therapy. Nearly 75% of pediatric cancer survivors have reported at least one adverse

8 event and ≈ 40% have reported life-threatening or debilitating ailments. Some common long-term adverse effects of conventional therapy for pediatric cancers are mentioned in Table 2 [2].

Table 2 : Long-term adverse effects of pediatric cancer therapy [2]

Types Examples

Cardiovascular Anthracyclin related (cardiomyopathy, arrhythmia, subclinical

system left ventricular dysfunction)

Vinca alkaloids (vasospastic attacks, autonomic dysfunction)

CNS Platinum agents (Peripheral sensory neuropathy)

Vinca alkaloids (peripheral sensory or motor neuropathy)

Methotrexate (clinical leukoencephalopathy)

Methotrexate, cytarabine (neurocognitive defects).

Digestive Hepatic dysfunction; sinusoidal obstructive syndrome

(methotrexate, mercaptopurine and actinomycin D)

Damage to testicular germinal epithelium by the combination Reproductive of alkylating agent and procarbazine; impaired

spermatogenesis.

Impaired ovarian function by combination of alkylating agents

and procarbazine; amenorrhoea by high-dose

cyclophosphamide, premature menopause.

Higher incidence of infertility in both male and female

survivors due to retrograde ejaculation or abnormal uterine

structure.

Higher incidence of miscarriages in female survivors and

partners of male survivors.

Respiratory Bleomycin, busulfan and nitrosoureas induce lung damage on

their own or potentiate damaging effects of radiation. The

manifestations include subclinical pulmonary dysfunction;

interstitial pneumonitis; pulmonary fibrosis; restrictive lung

disease; obstructive lung disease

Sensory Ototoxicity, hearing loss, vertigo and tinnitus due to platinum

agents

Cataract and blindness due to busulfan and corticosteroids

Urinary Cyclophosphamide induced bladder toxicity

Platinum agents and ifosfamide induced renal toxicity

Methotrexate induced renal toxicity

Subsequent Alkylating agent related and topoisomerase-II related neoplasms leukemia.

Solid neoplasms of skin, breast, thyroid, brain, bone, lung and

GI system and sarcomas

1.3.2 Drug resistance

Resistance to anticancer drugs has always been a major concern with

conventional cytotoxic drugs and targeted therapies alike. Acquired resistance

occurs by following mechanisms:

 Upregulation of efflux transporters

 Downregulation of influx transporters

 Activation of detoxification mechanism

 Blockade of apoptotic pathway

 Alteration of cell cycle checkpoints

 Alterations in the target protein

 Upregulation of efflux transporters: Cancer cells acquire multidrug resistance by

upregulation of efflux transporters, which shunt the drug molecules back into

extracellular matrix. These ATP binding cassette (ABC) transporters are the

products of Multidrug Resistance superfamily of genes (MDR). P-glycoprotein

(PgP), Multidrug resistance protein-1 (MRP1) and Breast Cancer Resistance

Protein (BCRP) are common efflux transporters [6]. PgP is the product of MDR-1

gene. It is upregulated in pediatric cancers such as neuroblastoma,

rhabdomyosarcoma, osteosarcoma, and is associated with metastatic disease [7,

8]. PgP binds to neutral and positive charged drug molecules and its main targets

are doxorubicin, vinblastine and paclitaxel. MRP-1 transports negatively charged

molecules and the phase-II metabolites of drugs [6]. MRP-1 expression has been

reported in neuroblastoma and rhabdomyosarcoma cell lines [9, 10]. Lung

Resistance Protein (LRP), associated with nuclear drug transport has been found

to be upregulated after therapy consisting of combinations vincristine +

actinomycin D + cyclophosphamide; etoposide / vincristine / actinomycin /

ifosfamide / adriamycin; epirubicin / vindesine / ifosfamide in rhabdomyosarcoma

[11]. BRCP or ABCG2 has been found to be involved in efflux of mitoxantrone in

acute lymphoblastic leukemia (ALL) and its increased expression has been

reported in relapsed and refractory acute myeloid leukemia [12, 13].

 Downregulation of influx transporters: The reduced intake of drugs may also

result due to mutation of transporters involved in the drug uptake. Mutation of

folate transporters, which transports folate analogues such as methotrexate is an

example [6]. This mechanism of resistance to methotrexate and trimetrexate has

been observed in leukemia patients [14].

 Activation of detoxification pathway: High expression of drug detoxifying enzymes

has been attributed to drug resistance in cancers. Some of the examples are

polymorphisms in dihydrofolate reductase (methotrexate) and elevated aldehyde

dehydrogenase (doxorubicin in Ewing‟s sarcoma) [15, 16].

 Resistance to apoptosis: Response of tumor cells to cytotoxic agents depends

upon the balance of proapoptotic and antiapoptotic factors within the cells.

Proapoptotic proteins include proapoptotic BCl-2 proteins (BCl-xs, Bak, Bad) and

P53, whereas antiapoptotic factors comprise of BCl-2 class of antiapoptotic

proteins (BCl-2, BCLxl, MCl-1) [17]. Tumor cells acquire resistance to cytotoxic

drugs either by upregulation of antiapoptotic genes or by downregulation or

mutation of proapoptotic genes. Also over-activation of proliferation pathways

such as MAPK/ERK, PI3K/Akt also induces apoptotic resistance to cancer cells

[18].

 Alterations in cell cycle checkpoints: Cell cycle is divided into G1, S, G2 and M

phase. Cell cycle is regulated by balance between Cyclin Dependent Kinases

(CDK)/cyclins and CDK Inhibitors (CDKIs) [19]. CDKs bind and activate cyclins,

which in turn facilitate the transition of cells from one phase to next phase. CDKIs

(P21, P27) inhibit the activity of CDKs thereby causing cell cycle arrest and

apoptosis. Cancer cells acquire resistance to cytotoxic agents by upregulating

CDKs or downregulating CDKIs.

 Alterations in the target protein: This type of resistance is commonly observed

with targeted therapy [20]. Secondary mutations in target proteins such as

receptor tyrosine kinases leading to alterations in drug binding site may hinder

the binding of targeted drug. Some examples are imatinib resistant chronic

myelogenous leukemia, gefitinib and erlotinib resistant EGFR mutant Non Small

Cell Lung Cancer (NSCLC) and imatinib resistant Gastrointestinal Stromal

Tumor (GIST) resulting due to KIT mutations [21].

1.4 Recurrence and relapse in pediatric cancers

One of the consequences of resistance to chemotherapy is recurrence and relapse despite initial response to therapy which is a major cause of low survival rate in high- risk pediatric cancers [22, 23]. Relapsed or recurrent tumors are more resistant compared to untreated cancers to the chemotherapy. A small fraction of patients

14 experience delayed recurrence or relapse after 5 years of completion of therapy [24].

The reported cumulative incidence of late recurrence or relapse is 4.4%, 5.6% and

6.2% after 10, 15 and 20 years, respectively, after initial diagnosis [24].

1.4.1 The role of Minimal residual disease in cancer relapses

Minimal residual disease refers to a small number of tumor cells that are unaffected by the therapy, even during remission, and is a major cause of cancer relapse in hematological malignancies, neuroblastoma and rhabdomyosarcoma [25-27]. These are the cells which escape therapy in sanctuary sites characterized by poor oxygenation and drug penetration, acquire resistance after exposure to the chemotherapy or due to the genomic instability of oligometastases [28, 29].

Therefore, relapsed cancers are not responsive to initial therapy. The presence of minimal residual disease in bone marrow after treatment has been associated with poor EFS in advanced neuroblastoma patients [30]. The presence of micrometastasis indicates worse prognosis in patients with localized Ewing‟s sarcoma [31]. In both alveolar and embryonal rhabdomyosarcoma, micrometastasis in bone marrow and peripheral blood is associated with worse prognosis [32]. In

ALL, the presence of minimal residual disease in blood and bone marrow is an important predictor of subsequent relapse after the therapy [33,34].

1.4.2 Maintenance therapy:

As a preventive measure to avoid relapse long-term maintenance therapy, aimed at targeting minimal residual disease, is employed after multimodal therapy. The requirements of maintenance therapy are [28]: 1) it should be minimally toxic because the patients who have undergone high-dose chemotherapy are unable to

15 tolerate toxic chemotherapy. 2) it should not have cross-resistance with previously administered drugs. In ALL, maintenance regimen consists of daily oral mercaptopurine, weekly methotrexate, vincristine and corticosteroids administered for a period of 2-3 years [2]. In neuroblastoma, most frequent site for minimal residual disease is bone marrow, followed by bone. Presently in neuroblastoma, myeloabalative chemotherapy followed by maintenance therapy with 13-cis retinoic acid is used to target minimal residual disease. 13-cis retinoic acid in combination with anti-GD2 antibody and Granulocyte/Macrophage Colony Stimulating Factor has resulted in superior overall survival (OS) in neuroblastoma [35]. In a stage-IV pediatric soft-tissue sarcoma trial, oral maintenance regimen (consisting of trofosphamide+etoposide and trofosfamide+idarubucin) demonstrated a survival rate of 57.8% after 57.4 months compared to high-dose therapy (thiotepa+ cyclophosphamide and melphalan+etoposide) which had a survival rate of 24.4%

[36]. Vinorelbine and cyclophosphamide represent another modality of maintenance therapy in high risk sarcoma [37]

1.5 Angiogenesis: a newer target for maintenance chemotherapy

Angiogenesis is the mechanism by which tumor forms its own vasculature, which is required for its growth and metastasis. Even after malignant transformation, tumor cells have to acquire capability to form blood vessels if they have to grow beyond few micrometers [38].

The significance of angiogenesis in the tumor progression was first reported by Dr.

Judah Folkman (1971) [39]. He discovered that the growth of micrometastatic tumors in immunodeficient mice was restricted to ≈0.2 mm when angiogenesis was

16 inhibited, due to dynamic balance i.e. the number of proliferating cells were balanced by approximately same number of apoptotic cells. When the angiogenesis inhibition was removed, tumors resumed growth. Due to the significant impact of angiogenesis on tumor progression, it is considered as one of the hallmarks of cancer [40].

Therefore, growth inhibition of microscopic tumors or minimal residual disease by inhibition of angiogenesis is an effective strategy for maintenance chemotherapy.

Investigations into the mechanisms of angiogenesis have paved way for discovery of several antiangiogenic strategies [41].

1.5.1 Modes of angiogenesis

Though sprouting angiogenesis was considered to be the predominant mechanism of tumor angiogenesis, other modes of tumor vasculature formation have emerged

[38].

Sprouting angiogenesis refers to the process of formation of tumor vasculature from pre-existing blood vessels. Here endothelial cells from pre-existing vasculature proliferates and migrate into tumor tissue. Vasculogenesis is the process where endothelial progenitor cells from bone marrow are recruited into tumor tissue, where they differentiate into endothelial cells. Several cell surface markers have been used to isolate these cells. Asahara et al isolated CD34+ and Flk-1+ peripheral blood mononuclear cells, which differentiated into mature endothelial cells in-vitro [42].

Songs et al isolated PDGFRβ+ progenitor cells which were capable of differentiating into pericytes [43]. Several cytokines play role in the mobilization of bone marrow progenitors, including Stromal Derived Factor-1 α (SDF1α) and Vascular Endothelial

Growth Factor (VEGF).

Intesusseption is the formation of blood vessels by division of pre-existing capillary plexus, without actual mitosis of endothelial cells [44]. In vessel co-option, tumors co- opt pre-existing blood vessels. Vascular mimicry is a phenomenon where tumors form conduits lined by tumor cells which function to transport nutrients and oxygen to the tumor. Transdifferentiation of tumor cells into endothelium is characterized by genetically unstable endothelial cells derived from cancer stem cells.

1.5.2 Molecular mechanisms of sprouting angiogenesis and vasculogenesis

The driving factor behind tumor angiogenesis is hypoxia. The factors responsible for tumor hypoxia are: A) higher tumor cell turnover compared to endothelial cells, due to which tumor cells farther from capillaries do not get adequate nutrients and oxygen; B) abnormal architecture of tumor vasculature characterized by large endothelial cell junction and loose pericyte coverage leading to leakiness and reduced efficiency of blood vessels [38]. This hypoxic condition leads to stabilization of the alpha subunit of Hypoxia Inducible Factors (HIF) (Figure 1).

HIF-1 is a transcription factor for number of genes involved in angiogenesis, cellular adaptation to hypoxia and apoptosis. It consists of two subunits namely HIF-1α and

HIF-1β. Under normoxic condition, HIF-1 alpha is unstable (half life ≈ 5 min) [45]. In the presence of oxygen, HIF-1 α is hydroxylated at two proline residues by prolyl hydroxylases (PHDs) in the presence of Fe++ and O2. Hydroxylated HIF-1 α binds to E3 ubiquitin-Von-Hippel Lindau (VHL) complex, a process known as poly- ubiquitination, which marks the HIF-1 α for degradation by proteosome [45-47]. In hypoxia, activity of PHDs is inhibited due to the absence of its co-substrate O2. Thus

HIF-1 α becomes stabilized and translocates from cytoplasm to nucleus, where it

18 forms a trancriptionally active complex with HIF-1 β. Active HIF-1 binds to the

Hypoxia Responsive Elements (HRE) of regulatory regions of target genes and thus induce gene expression. Target genes of HIF-1 include genes involved in angiogenesis, glucose metabolism, proliferation /survival and mitosis.

HIF-1 alpha is also upregulated by dysregulated PI3K/AKT pathway, which occurs in cancers due to receptor mutations or downregulation of tumor suppressor genes

[48].

Figure 1: Regulation of HIF-1alpha and VEGF in tumor cells (reproduced, with permission, from Liao, D. and R.S. Johnson, Hypoxia: a key regulator of angiogenesis in cancer. Cancer Metastasis Rev, 2007. 26(2): 281-90) [48].

The most potent proangiogenic gene induced by HIF-1 alpha is VEGF [38, 44, 49].

Other angiogenic genes such as Fibroblast Growth Factor (FGF) and Platelet

Derived Growth Factor (PDGF) are also regulated by HIF-1 alpha. However, VEGF is the most critical and most specific angiogenic factor which promotes endothelial cell division, growth and survival [50].

VEGF was first discovered by the name Vascular Permeability Factor in 1983

(Senger and Colleagues) as it induced vascular leakage [51]. Later, Ferrara and colleagues in 1989 confirmed the pro-angiogenic function of this soluble mitogen and named it Vascular Endothelial Growth Factor (VEGF) [52]. The various isoforms of VEGF are VEGF-A, VEGF-B, VEGF-C, VEGF-D and Placental Growth Factor

(PlGF) [50]. VEGF-A is further sub-divided into isoforms such as VEGF121, VEGF165,

VEGF189, VEGF206, containing 121, 165, 189 and 206 amino acids, respectively.

VEGF-A induce endothelial cell division and migration. VEGF-B induce embryonic angiogenesis whereas VEGF-C and VEGF-D induce lymphangiogenesis. Placental

Growth Factor (PlGF) is involved in embryonic vasculogenesis. VEGF binds to

VEGF receptors (VEGFRs) on the surface of endothelial cells leading to heterodimerization and autophosphorylation of VEGFRs [53]. VEGF-A binds to

VEGFR1 (Flt1) and VEGFR-2 (Flk-1). VEGFR-2 mediates most of the functions of

VEGF. VEGFR3 mediates lymphangiogenesis by binding VEGF-C and VEGF-D.

Though VEGF is the most important pro-angiogenic molecule, other factors such as bFGF, angiopoietins, interleukin-8 (IL-8), and PlGF also play role in tumor angiogenesis [41, 54].

The heterodimerization and phosphorylation of VEGFRs leads to signal transduction via pathways such as MAPK/ERK and PI3K/AKT/mTOR which results in endothelial cell division [55, 56]. The newly formed endothelial cells migrate along the VEGF gradient towards the tumor. Upon reaching the vicinity of tumor tissue, these cells degrade the vascular membrane and extracellular matrix by secreting matrix metalloproteinases (MMPs) and form the endothelial lining in hypoxic regions of tumor [57, 58] (Figure 2).

HIF-1 alpha, VEGF and stromal derived factor- alpha (SDF1α) also play the key role in vasculogenesis, as they activate, mobilize and recruit endothelial progenitor cells in bone marrow towards tumor tissues [59, 60] (Figure 2). It is noteworthy that

SDF1α, secreted by hypoxic stromal cells of tissue, also acts as a chemotactic agent for highly tumorigenic side population of neuroblastoma and rhabdomyosarcoma cells [61]. HIF-1 alpha, VEGF and SDF1α up-regulate MMP-9 within bone marrow progenitor cells, which activates soluble kit ligand. Soluble kit activates quiescent hematopoietic and endothelial progenitor cells to proliferating state.

Figure 2: The various molecules involved in tumor angiogenesis (reproduced, with permission, from Ferrara, N. and R.S. Kerbel, Angiogenesis as a therapeutic target. Nature, 2005. 438(7070): p. 967-74 [41]) In this figure BMC denotes bone marrow progenitor cells. The figure also explains the cross-talk between tumor cells and stromal cells which plays an important role in tumor progression by influencing processes such as angiogenesis, tumor cell survival, proliferation and invasion.

1.6 Angiogenesis and angiogenic factors in pediatric solid tumors

Angiogenesis is a very critical step in tumor progression because 1) the tumor vasculature provides nourishment and oxygen for the growing tumor. 2) blood is the main medium of metastatic spread of tumor cells. 3) angiogenesis is required for the establishment and growth of cancer at metastatic sites [62]. Tissue examination of adults who died of reasons other than cancer have revealed that microscopic tumors are present in a significant fraction of humans (< 39%) whereas only a tiny fraction

(1-1.5%) develop clinically detectable cancers [63]. Thus, microscopic tumors can remain dormant for years, one of the reasons being the absence of angiogenic switch. The degree of angiogenesis has correlated with worse prognosis in many adult cancers including breast cancer, malignant melanoma, bladder cancer, renal cancer, prostate cancer and lung cancer [64]. Majority of pediatric cancers such as leukemia, CNS cancers, neuroblastoma and pediatric sarcoma are angiogenesis dependent [65, 66].

Neuroblastoma is an angiogenesis dependent cancer. High vascular index has found to be strongly correlated with disseminated disease, worse prognosis and N-MYC amplification [67]. MMP2 and MMP9 have been found to be higher in stage-IV neuroblastoma compared to lower stages [68, 69]. High levels of VEGF-A, VEGF-B,

VEGF-C, bFGF, Ang-1, Ang-2, TGF-alpha and PDGF-A was detected in a panel of

37 primary tumors and 22 cell lines [70]. Here the tumor samples representing advanced stage disease had significantly higher expression of these factors. This redundancy in angiogenic expression suggests possible synergism and that the inhibition of VEGF pathway alone may not be effective antiangiogenic strategy for neuroblastoma. Another study reports that all the cell lines and primary tumors had

high expression of VEGF and VEGFR2 (Flk1) [71]. Upregulation of VEGF165 mRNA in tumor tissues and VEGF protein in serum samples in stage-III neuroblastoma patients, compared to lower stages, has been reported [72]. Hypoxia enhances the

VEGF secretion in neuroblastoma cell lines [73]. In preclinical models, the serum levels of VEGF165 has been correlated with tumor size of SH-SY5Y xenografts [74].

Apart from its role in angiogenesis, VEGF is also reported to act as a survival factor for neuroblastoma by upregulation of BCl-2 expression [75]. Under hypoxia,

VEGF/Flt-1 autocrine loop provides survival advantage to SK-N-BE(2) neuroblastoma cells via MAPK/ERK-1/2 pathway [76]. Serum levels of Hepatocyte

Growth Factor (HGF) and VEGF have correlated with the genetic markers of poor outcome in neuroblastoma patients [77].

In soft tissue sarcoma, high VEGF concentration was associated with poor survival and a greater probability of local recurrence and metastasis [78]. Here VEGF concentration, but not tumor microvessel density, was found to be an independent prognostic factor for disease outcome. In another study, tumor microvessel density correlated positively with EFS in childhood embryonal rhabdomyosarcoma patients

[79]. However, it needs to be noted that apart from its role in angiogenesis, VEGF is also an autocrine growth factor for rhabdomyosarcoma cells [80]. Higher serum and urine levels of angiogenic factors such as VEGF, bFGF and Interleukin-8 (IL-8) has been reported in embryonal rhabdomyosarcoma patient [81].

In bone cancers, high VEGF serum concentration has been detected in patients with osteosarcoma, chondrosarcoma and ewing‟s sarcoma [82]. In ewing‟s sarcoma, vasculogenic mimicry and vasculogenesis have also been reported to contribute to

tumor vasculature [83, 84]. VEGF165, but not VEGF189 is found to stimulate vasculogenesis in ewing‟s sarcoma [84, 85]. VEGF upregulated by oncoproteins

EWS-ETS is an independent negative predictor of survival in ewing‟s sarcoma [86].

VEGFR2 inhibition by DC101 has been reported to suppress tumor growth in ewing‟s sarcoma mice model [87]. Other than VEGF, SDF-1 alpha also contributes to blood vessel formation in ewing‟s sarcoma [86]. In one study, high microvessel density was associated with progression-free and overall survival in osteosarcoma [88], whereas in another study, high microvessel density was associated with metastasis and poor prognosis [89]. In a third study, no correlation was observed between angiogenesis and disease outcome in osteosarcoma patients [90].

Glioblastoma Multiforme is a highly vascularized and one of the most studied cancer with regards to angiogenesis in pediatric cancers [91]. Neovascularization in GBM occurs by sprouting angiogenesis, vasculogenesis, intesussception, vascular mimicry and transdifferentiation of tumor stem cells to endothelial cells. VEGF plays a major role in glioblastoma angiogenesis and vasculogenesis. SDF-1/CXCR4 also contributes to glioblastoma angiogenesis and metastasis. Endothelial hyperplasia is the hallmark and criteria for grading glioblastoma [92, 93]. Microvessel density is the best predictor of progression-free survival in optic pathway/hypothalamic glioma [94].

Glioblastoma is characterized by abnormal vasculature which leads to leakiness and hypoxia resulting in upregulation of VEGF [93]. Stem-like cells in GBM have been found to form more vascularized tumors, driven by VEGF, than non-stem like cells

[95].

1.7 Antiangiogenic strategies

The concept of antiangiogenic therapy was first proposed by Judah Folkman in

1970s, who postulated that since angiogenesis is an essential mechanism for the growth and metastasis of tumor, cancer can be starved to remission by cutting off the blood supply [96]. In 1990s, antiangiogenic potentials of several drugs were explored. These included endogenous antiangiogenic molecules (angiostatin, endostatin), endothelial growth inhibitors (TNP-470, thalidomide, IL-12), neutralizers of angiogenic factors and receptors, MMP inhibitors and inhibitors of vascular adhesion molecules [97]. In 2000, the unconventional scheduling of conventional cytotoxic agents, also referred to as Low Dose Metronomic Chemotherapy (LDM) demonstrated antitumor efficacy by antiangiogenic mechanism [98, 99]. In 2004, bevacizumab, VEGF-neutralizing monoclonal antibody was approved by FDA for colorectal cancer [100]. Thereafter, small molecule Receptor Tyrosine Kinase

Inhibitors (RTKIs) such as sunitinib, sorafenib, pazopanib and axitinib, which inhibit

VEGFR autophosphorylation were approved [100]. Lately, gold nanoparticles and antivascular ultrasound stimulated microbubbles have demonstrated antiangiogenic efficacy [101, 102].

Since VEGF signaling pathway is the major pathway in tumor angiogenesis, majority of antiangiogenic strategies have focused on inhibition of VEGF pathway, though other molecular targets such as angiopoietin/Tie2, MMPs, endoglins and Analplastic

Lymphoma Kinase (ALK1) are also being explored [103]. At present, LDM chemotherapy and targeted antiangiogenic drugs are the most widely employed antiangiogenic strategies in cancers.

Antiangiogenic strategies have previously demonstrated the ability to target minimal residual disease. Treatment of osteosarcoma patients with endostatin following primary tumor removal decreased the angiogenesis inducing ability in patients with post-operative recurrence, whereas preclinical studies have demonstrated its ability to prevent lung metastatsis [104, 105]. ZD4190 has prevented outgrowth of minimal residual squamous carcinoma in deep tissues in mice model [106]. Bevacizumab has demonstrated the ability to delay tumor progression in an experimental metastatic neuroblastoma mice model [107]. Thrombospondin-1 (TSP-1) has prevented the development of pulmonary metastasis in melanoma mice model [108].

1.8 Targeted antiangiogenic agents:

1.8.1 Classification:

Antiangiogenic agents can be broadly classified as direct-acting or indirect-acting

[109]. Direct acting antiangiogenic agents inhibit the angiogenesis by directly affecting the endothelial cells, whereas indirect-acting antiangiogenic agents inhibit angiogenesis by targeting the growth factors or receptors involved in endothelial stimulation. Figure 3 describes the various classes of antiangiogenic agents [54,

110-114].

Figure 3 The classification of antiangiogenic agents based upon the mechanism of action [54, 110-114]. (RTKI = Receptor Tyrosine Kinase Inhibitors)

1.8.2 Anti-VEGF antiangiogenic agents:

Since this study involves a VEGF receptor inhibiting antiangiogenic agent, this section covers the antiangiogenic drugs that are designed to target VEGF pathway.

The description of currently approved and experimental VEGF pathway blocking antiangiogenic drugs are given below.

First generation

Bevacizumab: It is a monoclonal antibody that forms a protein complex with VEGF, thus making it incapable to bind to its receptors [100]. It is currently approved for the treatment of metastatic colorectal cancer, first line treatment in NSCLC, metastatic

Renal Cell Carcinoma (RCC), and recurrent glioblastoma multiforme. Major side effects of bevacizumab are hypertension, risk of bleeding, perforations in intestine and nasal septum.

Second generation

Major drawback of first generation VEGF pathway inhibitor bevacizumab is that it is effective only in combination with other chemotherapeutics. This is because it causes the normalization of residual vasculature leading to enhanced delivery of chemotherapeutics [115]. Since it only inhibits VEGF ligand, leaving other proangiogenic tyrosine kinase pathways still active, it is not effective as a single agent. On the other hand, RTKIs can suppress multiple tyrosine kinases, hence are more effective as single agents [116]. Following are some of the VEGFR blocking small molecule RTKIs, which are classified as second generation antiangiogenic agents:

Sunitinib: It inhibits VEGFR1, VEGFR2, VEGFR3, Platelet Derived Growth Factor

Receptors (PDGFRs) and c-kit receptors [117]. It has been approved for GIST, especially in imatinib resistance due to kit mutations, meningioma, pancreatic neuroendocrine tumors in adults, and RCC. It has shown significant efficacy in Phase-II trials in metastatic breast cancer and refractory NSCLC [118, 119]. Adverse effects are manageable, most common of which are hypertension, diarrhea, nausea and hypothyroidism [120]. Hypertension is a biomarker for its efficacy in RCC. Sunitinib revolutionized antiangiogenic therapy as it was first agent to show efficacy in imatinib resistant GIST and first agent to show OS > 2 years [117].

Sorafenib: It is a small molecule tyrosine kinase inhibitor which inhibits VEGFR,

PDGFR and raf kinases [121]. It is approved for the treatment of renal cancers and hepatocellular carcinoma (HCC) [122, 123].

Pazopanib: Pazopanib inhibits VEGFR (1, 2, and 3), PDGFR (α and β) and c-Kit

[124]. It has been approved by USFDA for RCC and soft tissue sarcoma. In its phase-II trial as a single agent, it has demonstrated efficacy in RCC, soft-tissue sarcoma, ovarian cancer and NSCLC [125]. It showed objective response rate in

Phase-II study with RCC patients who had either progressed or were intolerant to sunitinib or bevacizumab. In another study, pazopanib was effective in RCC patients who were refractory to targeted therapies like sunitinib, sorafenib, temsirolimus, everolimus and bevacizumab [126].

Axitinib: It inhibits VEGFR (1,2, and 3), PDGFR (α and β) and c-Kit and is approved for advanced RCC [127].

Vandetanib: It is an inhibitor of VEGFRs and EGFR and is approved for the treatment of thyroid cancer [128]. Other experimental VEGF RTKIs, which block

VEGFRs, are cediranib, semaxinib, telatinib [129, 130].

VEGF trap (aflibercept): It is a novel recombinant protein which blocks VEGF A,

VEGF-B and PlGF. In Phase-III trials, it significantly enhanced the efficacy of 5-fluoro uracil in metastatic colorectal cancer patients who were refractory to oxaliplatin and who had received prior bevacizumab therapy [100].

1.8.3 Anti-VEGF therapy in pediatric cancers:

A COG phase-I trial of bevacizumab in refractory solid tumors and another trial in pediatric CNS tumors have found that it was well tolerated in children [131]. Though adverse effects such as lymphopenia, rash, mucositis, proteinuria, and rise in blood pressure were observed, these were not dose-limiting. In a Phase-II trial, its efficacy in combination with irinotecan in pediatric gliomas was found to be inferior to that in high grade recurrent adult glioma, though the toxicity profiles were similar [132].

Here, 10 out of 12 patients tolerated the regimen, PR and stable disease was achieved in two and four patients, respectively. Median PFS and OS were 2.25 months and 6.25 months respectively. Prolonged therapy has been found stabilize low-grade pediatric gliomas [133, 134].

A Children‟s Oncology Group (COG) clinical trial for refractory solid tumors reports that common adverse effects with 15 mg/m2/ day (Maximum Tolerated Dose; MTD) sunitinib for 28 days followed by 2 weeks break were leucopenia, neutropenia and hypertension [135, 136]. Here, two patients who were pre-treated with anthracyclins developed severe cardiotoxicity due to which treatment had to be discontinued. A

COG phase-I study recommends dose of sorafenib to be 200 and 150 mg/m2 (every

12 h) for pediatric refractory solid tumors and leukemia, respectively [137]. In a COG phase-I trial in refractory pediatric solid tumors, aflibercept (VEGFtrap) achieved stable disease for >13 weeks in patients with HCC, hepatoblastoma and clear cell sarcoma [138]. The Pediatric Preclinical Testing Program (PPTP) study showed that pazopanib delayed the tumor growth and enhanced the survival of animals in mice models of rhabdomyosarcoma and ewing‟s sarcoma [139].

1.9 Low Dose Metronomic Chemotherapy:

According to the definition, Low Metronomic Dose (LDM) chemotherapy is chronic administration of low dose of cytototoxic agent without any drug free breaks for a prolonged duration [140]. Metronomic scheduling lays great emphasis on the ability of a regimen to prevent the endothelial cell recovery during the course of a therapy.

Since cytotoxic agents target all types of proliferating cells, activated endothelial cells are also targeted when used in metronomic fashion. Conventional therapy also targets endothelial cells, but unlike LDM therapy, prolonged drug-free breaks causes endothelial recovery. The concept of LDM chemotherapy was first postulated by

Kerbel et al in 1991 and was proved by two independent studies by Klement et al and Browder et al. Klement et al (2000) confirmed the efficacy of LDM vinblastine in neuroblastoma mouse model [98]. Browder et al confirmed the efficacy of LDM cyclophosphamide in the xenograft models of breast cancer and lewis lung carcinoma, developed from cell lines which were resistant to the same drugs [99].

Thereafter, LDM chemotherapy has demonstrated safety and efficacy various preclinical and clinical studies (Tables 3 and 4).

Though the term “Low Dose Metronomic” was coined in the year 2000, prolonged low-dose chemotherapy was earlier used as maintenance regimen for cancers. In

ALL, the maintenance therapy involves low doses of weekly, oral methotrexate and daily 6-mercaptopurine [141], Low-dose chemotherapy with carboplatin and etoposide achieved OS of 100% in infants with localised, unresectable non-MYCN amplified neuroblastoma.[142]. However, the contribution of antiangiogenic activity to these efficacies cannot be established. Protracted low-dose topoisomerase-I inhibitors demonstrated preclinical activity against melphalan and vincristine resistant pediatric solid tumor xenografts [143]. The efficacy observed in this preclinical study was translated into clinic in a pharmacokinetically-guided dosing schedule of topotecan in pediatric solid tumor patients [144-146].

1.9.1 Advantages of LDM chemotherapy over conventional MTD (Maximum Tolerated Dose) chemotherapy [140]

 LDM chemotherapy has lower acute toxicity due to lower dose of the cytotoxic

agents. In the pioneering study conducted by Klement et al, in-vitro, Human

umbilical vein endothelial cells (HUVEC) was much more sensitive than

neuroblastoma cell lines to anti-proliferative effect of vinblastine, where the

thymidine incorporation in HUVEC was only 6.2% of that of tumor cell lines.

This observation supports the concept that lower doses of cytotoxic agents

can target endothelial proliferation but not tumor cells. Such a low dose, which

can inhibit angiogenesis without any direct effect on tumor cells, can have

antitumor effect by starving the tumor cells without any acute toxicity

commonly observed with conventional regimen.

 LDM chemotherapy has been proved to be active on tumors resistant to MTD

schedule of same cytotoxic agents in the pioneering study by Browder et al,

where the breast cancer and Lewis lung carcinoma cell lines resistant to

cyclophosphamide were sensitive to the LDM regimen of same drug [99].

 LDM chemotherapy has been reported to enhance chemosensitivity of

endothelial cells contrary to MTD (where cross resistance between vinblastine

and paclitaxel was observed) [147].

1.9.2 Mechanisms of LDM chemotherapy:

Although initially it was considered that the predominant mechanism of LDM therapy is antiangiogenesis, several other mechanisms have been reported. So far the proposed mechanisms of LDM therapy are antiangiogenesis, anti-vasculogenesis, immune stimulation, induction of endogeneous antiangiogenic molecules.

Immune stimulation: CD4+/CD25+ regulatory T cells (Treg) suppress T cell mediated anti-tumor immune response [148]. Its presence is related to tumor progression and inversely correlated with efficacy of treatment [149]. Low dose cyclophosphamide has been reported to deplete CD4+/CD25+ cells and therefore increase T lymphocyte proliferation and T memory cells in mice models as well as in end-stage cancer patients [148, 150], whereas low dose temozolomide has depleted T lymphocyte proliferation in rat glioma model [151].

Some cytotoxic agents have been found to cause maturation of dendritic cells of immune system. These included topoisomerase inhibitors, antimicrotubule agents

(vinblastine and paclitaxel), alkylating agents (mechlorethamine and diaziquone) and

35 an antimetabolite (cladribine) [152]. In a separate experiment low doses of vinblastine induced dendritic cell maturation in immunocompetent C57BL/6 melanoma mice model, but not in immunodeficient mice [153].

Inhibition of Circulating Endothelial Progenitor (CEP) mobilization: LDM chemotherapy induces the mobilization of CEP from bone marrow, whereas high dose chemotherapy does the opposite. In adult cancer patients, CEP and VEGF level increased following high dose chemotherapy , whereas LDM trofosfamide with or without celecoxib significantly reduced CEPs while VEGF concentration was unchanged [154]. In lymphoma bearing mice, after few days of completion of MTD cyclophosphamide, robust CEP mobilization was observed and tumor became drug resistant, whereas LDM cyclophosphamide maintained a consistent decrease in

CEPs and more durable tumor growth inhibition [155]. Metronomic irinotecan, with or without bevacizumab, decreased CEP levels significantly in colon cancer bearing mice, whereas MTD irinotecan did not show any CEP reduction [156] .

Induction of endogenous antiangiogenic molecules: LDM cyclophosphamide caused increase in TSP-1 in mice models of adult cancers, an effect which was not observed in TSP-1-null C57BL/6 mice [157]. Metronomic ceramide analogues were antiangiogenic and significantly increased TSP-1 levels in mice models of pancreatic cancer [158]. TSP-1 induces endothelial cell apoptosis by increased expression of proapoptotic Bax, downregulating Bcl-2 and promoting caspase-3 activation [159].

TSP1 is reported to have inhibited VEGF-mediated Bcl-2 expression in endothelial cells in vitro and as a result demonstrated antiangiogenic effect in vivo.

However, there are exceptions regarding the statement that LDM chemotherapy is superior to MTD in terms of efficacy against resistant cancers. PC3 prostate cancer xenografts, which acquired resistance to LDM cyclophosphamide, were found to be sensitive to MTD regimen of the same drug [160].

1.9.3 Examples of efficacies of single agent LDM chemotherapy

Table 3: Preclinical examples of single agents LDM chemotherapy († indicate pediatric cancer studies)

Drug Disease model

Irinotecan Superior activity to MTD irinotecan (every 2 wks) with

significant CEP reduction in colon cancer [156]

Topotecan Prostate cancer (superior to conventional dosing in mice)

[161]

Cyclophosphamide HCC (superior to conventional MTD dosing; suppressed

(20mg/Kg, twice a growth and metastsis) [162].

wk)

Temozolomide Superior to conventional regimen in in-vitro HUVEC cell

migration and tube formation assays [163].

Docetaxel Gastric cancer (in-vitro upregulation of TSP1 in HUVEC)

[164]

Cyclophosphamide Rat breast cancer (eradicated tumor and had lower

cardiotoxicty) [165]

Cyclophosphamide Rat breast cancer (doxorubicin slowed tumor growth,

and doxorubicin whereas cyclophosphamide eradicated tumor growth)

[165]

Cyclophosphamide Lymphoma (Compared to untreated, treated groups

showed tumor regression in rats and maintained a low IL-

10 level; delayed tumor growth in nude mice) [166]

Etoposide Lewis lung carcinoma and glioblastoma (Inhibited VEGF

and FGF induced corneal neovascularization, increased

levels of endostatin, inhibited primary tumor growth and

prevented spontaneous lung metastasis in lewis lung

carcinoma) [167]

TW-37 (Bcl2 Human squamous cell carcinoma (inhibited endothelial inhibitor) sprout formation, induced differentiation and reduced the

invasive phenotype in mice xenografts) [168]

Gemcitabine Pancreatic ductal adenocarcinoma (Both MTD and

metronomic regimens delayed tumor growth, however,

metronomic regimen was associated with the lowering of

pro-angiogenic factors) [169]

Topotecan Ovarian cancer (1 mg/Kg daily, oral, topotecan significantly

delayed tumor growth) [170]

Topotecan Ovarian cancer (0.5 mg/Kg daily, oral topotecan reduced

tumor growth by 40-59% compared to the untreated mice)

[171]

Cisplatin HCC (superior antiangiogenic and higher safety compared

to MTD regimen) [172].

Topotecan † Wilm‟s tumor (significant tumor size reduction even at 0.36

mg/Kg with no observable adverse effects) [173]

Topotecan † Hepatoblastoma (0.36 mg/Kg IV topotecan 5 days/week

for 6 weeks ssignificantly reduced tumor sizes and

vascularity) [174]

Vinblastine † Neuroblastoma (in-vitro, vinblastine was more toxic to

HUVEC than to neuroblastoma cells; in-vivo it caused a

significant tumor growth delay and reduction in tumor

perfusion) [98]

Methotrexate † Osteosarcoma (in rats, 1.2 mg/Kg twice a week for 8

weeks significantly reduced the tumor volume) [175]

Table 4: Examples of clinical trials for LDM chemotherapy used alone († represents pediatric cancer trial)

Cytotoxic drug Disease

Capecitabine HCC (case study where 2 months therapy caused

disappearance of nodules in 64 yr old man with

advanced HCC) [176].

Vinorelbine Phase-I trial indicated minimal toxicity in patients with

advanced cancers [177]

Trofosfamide, Embryonal rhadbomyosarcoma (Higher event free

idarubicin, and survival and lower relapse when high dose

etoposide † chemotherapy was followed by maintenance

regimen) [178]

Docetaxel Angiosarcoma (retarded progression and consequent

lung metastasis in 65 yr old patient) [179]

Capecitabine Well tolerated and maintained the response for 18

months in HCC [180].

Cyclophosphamide Literature review suggests its effectiveness in

castration resistant prostate cancer [181].

Capecitabine Elderly gastric cancer patients (effective and well

tolerated as palliative therapy) [182]

Alternate cycles of NB90 trial in neuroblastoma [183]. metronomic melphalan/etoposide and vincristine/cyclophosphamide †

Vincristine/cyclophosph Pediatric refractory cancers (well tolerated and

-amide/methotrexate † associated with disease stabilization)[184].

Cyclophosphamide Disease stabilization in androgen ablation/refractory

prostate cancer patient[185]

Topotecan (0.8 Recurrent pediatric brain tumor. Regimen was safe in mg/m2/day) for 21 all patients (regimen was safe and achieved days, repeated every remission in 2 out of 26 patients who are alive 7 and

28 days † 9.5 years after therapy) [186].

Zoledronic acid Breast cancer with bone metastasis (also reduced

serum VEGF unlike conventional regimen)[187]

Vinorelbine Metastatic breast cancer [188]

Trofosfamide (case Docetaxel refractory prostate cancer (achieved long- study) term remission) [189]

Cyclophosphamide Superior to megestrol acetate in advanced cancer

patients having exhausted of all therapies [190].

Etoposide Soft tissue sarcoma [191]

Temozolomide Effective and tolerated in glioblastoma multiforme

patients refractory to conventional temozolomide

[192].

Paclitaxel and cisplatin Parotid gland carcinoma [193]

(case study)

Temozolomide † Pediatric brain-stem glioma (median duration was

three cycles of 6 weeks therapy (85 mg/m2 daily); the

first cycle was given with induction radiotherapy;

Median OS : 9.8 m, prolonged hematological toxicity

was observed) [194].

Temozolomide † Recurrent pediatric brain tumors (of 28 patients, 2

Complete Response (CR) and 2 PR; metronomic

scheduling was associated with higher commulative

drug exposure and lower grade 3/4 toxicity compared

with conventional schedule) [195].

4-drug regimen † 56 days (8 weeks) with weekly vinblastine, daily

cyclophosphamide, twice weekly methotrexate and

daily celecoxib.

1.10 Combination of LDM chemotherapy and antiangiogenic agents

1.10.1 Rationale for combining LDM chemotherapy with antiangiogenic agents:

Though metronomic dosing of chemotherapeutics demonstrates cytotoxic activity on endothelial cells, this effect can be lost due to the activation of proangiogenic pathways. In response to LDM chemotherapy induced antiangiogenesis, tumor cells upregulate VEGF, which antagonizes the antiangiogenic effect of LDM chemotherapy [196]. The antiangiogenic effect of docetaxel in in-vivo matrigel plugs was hindered by the protective function of VEGF and bFGF [197]. The monoclonal antibody against VEGF and antiangiogenic agent methoxyestradiol was able to reverse this resistance. Apart from being a growth factor for endothelial cells, VEGF also acts as a survival/antiapoptotic factor to endothelial cells by following mechanisms:

 In-vitro, VEGF has been reported to upregulate anti-apoptotic protein survivin

(in the presence of cytotoxic agent paclitaxel) Bcl-2 and A1 in endothelial cells

[198, 199]. In another study, VEGF was reported to induce Inhibitor of

Apoptosis (IAP) family of proteins, survivin and XIAP, in endothelial cells.

 Activation of Flk-1/KDR (VEGFR2) by VEGF or its mutant was found to

regulate endothelial cell survival via PI3K/Akt pathway, whereas activation of

Flt-1-specific VEGF mutant was unable to show this effect [200].

 VEGF has been found to increase the expression of urokinase type

plasminogen activator (uPA). uPA binds to its receptors on endothelial cells,

which causes activation of plasminogen to plasmin. Plasmin is involved in the

matrix degradation, which helps in the mobilization of endothelial cells into

tumor tissues [201].

As for antiangiogenic agents, substantial number of patients are refractory or do not respond at all (pancreatic cancer). Even if effective, antiangiogenic therapy only extends the Progression Free Survival (PFS), but not OS of patients. Combining

LDM chemotherapy with the antiangiogenic agents can overcome the limitations of both LDM chemotherapy and antiangiogenic agents.

1.10.2 Pioneering studies on the combination of LDM chemotherapy and antiangiogenic agents

The first study demonstrating the superiority of such combination was the study involving LDM vinblastine in combination with anti-VEGFR2 antibody DC101 [98].

Here continuous administration of both the agents caused sustained and full regression of neuroblastoma xenografts in mice. The treatment lasted for 6 months without causing toxicity or drug resistance. Simultaneously, in another study

(Browder et al), metronomic cyclophosphamide demonstrated three fold higher efficacy than the conventional schedule, while its combination with TNP-470 eradicated the tumor in mice models of drug resistant breast cancer and lewis lung carcinoma [99]. Here, double staining of von Willebrand Factor and TUNEL assay revealed that the endothelial apoptosis preceded tumor cell apoptosis. Prior to these studies, there were other reports where antiangiogenic agents have enhanced the efficacies of conventional regimen of cytotoxic agents. In one preclinical study, antiangiogenic drugs such as TNP-740, minocycline, suramin and genistein increased the antitumor activity of minocyclin and cyclophosphamide in Lewis Lung

Carcinoma mice model [202]. In another study, antiangiogenic regimen consisting of

Tetrahydrocortisol , beta-cyclodextrin tetradecasulfate in a 1:1 molar ratio and minocycline enhanced the anti-tumor and antimetastatic activity of standard regimen

45 of cytotoxic drugs such as cisplatin, melphalan, cyclophosphamide,adriamycin, bleomycin, and radiation therapy in mice lewis lung carcinoma model [203].

1.10.3 Previous works on LDM chemotherapy with antiangiogenic drugs

LDM chemotherapy combined with antiangiogenic agents has been found to be advantageous recently in several preclinical and clinical trials. Some examples of such combinations are mentioned in Tables 5 and 6.

Table 5: Recent preclinical examples of combination of LDM chemotherapy with antiangiogenic drugs († indicates study involving pediatric cancer)

Cytotoxic drug Combination / Disease

Topotecan A4.6.1 (Wilm‟s tumor, neuroblastoma) [204, 205]

Doxorubicin DC101 (soft tissue sarcoma) [206]; bevacizumab

(hepatocellular carcinoma) [207]

Carboplatin TSP1/endostatin (testicular germ cell tumor) [208]

CHS 828 TNP-470 or SU5416 neuroblastoma) [209]

Paclitaxel VEGF peptide mimics (breast cancer) [210]

Cyclophosphamide IL-12 (colorectal carcinoma, by decreasing the number of

Treg cells) [211] ; TSP1 & Pigment Epithelium Derived

Factor (colon cancer) [212]; 2-methoxyestradiol analogue

ENMD-1198 (breast cancer) [213]

Capecitabine (-) epigallocatechin-3-gallate (gastric cancer) [214]

Cisplatin Viral targeted endostatin (head&neck squamous cell

carcinoma) [215]

Doxifluridine TNP-470 (uterine carcinosarcoma) [216]

S1 Vandetanib (hepatocellular carcinoma) [217]

Uracil + tegafur Sorafenib (HCC; delayed the resistance which was

encountered with single agent sorafenib) [218]

Etoposide Celecoxib and rosiglitazone (Lewis Lung Carcinoma and

glioblastoma) [167].

Gemcitabine Sunitinib (pancreatic cancer) [219]

Topotecan Pazopanib (ovarian cancer) [170, 171]

Cyclophosphamide Celecoxib (Mammary adenocarcinoma)[220]

Cyclophosphamide OXi4503 (breast cancer, melanoma) [221]

Cyclophosphamide Axitinib (rat gliosarcoma) [222]

Topotecan † A4.6.1 (Wilm‟s tumor; lowest rebound tumor and no lung

metastasis compared to untreated and single agents)

[204]

Vinblastine † DC101 (Neuroblastoma; caused significant tumor growth

delay and reduced tumor perfusion compared to control

and both single agents; stable tumor size was observed

for 210 days in response to continued treatment with the

combination which was tolerated by mice) [98]

Table 6 : Recent clinical trials involving combination of LDM chemotherapy with antiangiogenic drugs († indicates pediatric cancer trial)

Drug Combination/disease Major observation

Cyclophosphamide Cisplatin & bevacizumab

(refractory anaplastic

ependymoma) [223]

Cyclophosphamide Methotrexate & vandetanib Changes in platelet

(breast cancer) [224] proteomics served

as marker

Cyclophosphamide Bevacizumab& erlotinib (HER2

& capecitabine negative breast cancer) [225]

Cyclophosphamide Bevacizumab (metastatic breast Increase in Mean

& capecitabine cancer) [226] corpuscular volume

was antiangiogenic

marker

Cyclophosphamide Veliparib (adult refractory solid

tumors & lymphomas) [227]

Temozolomide Bevacizumab, octreotide

(malignant Neuroendocrine

Tumors) [228]

Cyclophosphamide Pazopanib (platinum resistant

recurrent ovarian cancer) [229]

Cyclophosphamide Bevacuzumab (heavily

pretreated advanced ovarian

cancer) [230]

Vinorelbine Bevacizumab (metastatic breast

cancer) [231]

Tegafur Gefitinib (adrenocarcinoma) Superior to gefitinib

[232] in patients with

EGFR mutations

Cyclophosphamide Bevacizumab (recurrent Long term remission

endometrial cancer) [233] in a patient

Cyclophosphamide Bevacizumab (recurrent ovarian Safe and efficacious

cancer) [234] (39% patients had

PFS at least 6

months)

Tegafur/uracil Sorafenib (HCC) [235] Enhanced efficacy

of tegafur

Cyclophosphamide Celecoxib (non-hodgkin‟s 37% response and

lymphoma) [236] 22% SD

Vinblastine or Celecoxib (pediatric recurrent 13 of 33 patients cyclophosphamide solid tumors) [237] had stable disease.

†

Cyclophosphamide Zolendronic acid Of 21 patients, 1 PR

† (recurrent/refractory and 9 stable disease

neuroblastoma) [238]

4-drug regimen † Weekly vinblastine, daily 56 days (8 weeks)

cyclophosphamide, twice weekly treatment was well

methotrexate and daily celecoxib tolerated and

(refractory and relapsed achieved disease

pediatric tumors) [239] stabilization.

1.10.4 Induction of tumor dormancy by maintenance therapy involving metronomic scheduling

LDM single agent therapy is known to induce tumor dormancy. Oral LDM chemotherapy, consisting of alternate cycles of thalidomide+celecoxib and cyclophosphamide+etoposide for 6 months, was well tolerated by heavily pretreated pediatric population and also induced tumor dormancy [240]. In another follow-up study, 80% pediatric CNS cancer survivors who were treated with LDM therapy following high dose intensive chemotherapy had stable disease [241]. COMBAT

(Combined Oral Maintenance Biodifferentiating and Antiangiogenic Therapy) consisting of celecoxib, 13-cisretinoic acid, metronomic temozolomide and low-dose etoposide, administered to extensively pretreated patients, induced prolonged disease stabilization in 9/14 children with relapsed and/or high-risk solid tumors

[242]. In another study, COMBAT regimen (low-dose daily temozolomide, etoposide, celecoxib, vitamin D, fenofibrate and retinoic acid) achieved 2-year OS in 43.1% patients [243]. In pediatric recurrent solid tumors, prolonged treatment with LDM vinblastine or cyclophosphamide with celecoxib resulted in stable disease in 13 / 33 patients for a duration range 28-76 weeks [237]. LDM cyclophosphamide and zoledronic acid (daily, orally) for 28 days, repeated every 28 days, showed 1 PR and

9 stable disease out of 21 neuroblastoma patients (median range of courses: 4.8)

[238]. Maintenance chemotherapy with 12 alternating cycles of oral melphalan/etoposide and cyclophosphamide/vincristine for one year achieved median nine year EFS of 31±5% in high-risk neuroblastoma patients [183]. In a case study report, a child with disseminated medulloblastoma, which appeared six years after primary tumor removal, achieved remission with LDM cyclophosphamide, etoposide and zoledronic acid (3 week cycle with one week break for 18 months)

[244]. Antiangiogenic regimen consisting of bevacizumab, thalidomide, celecoxib, fenofibrate, etoposide, and cyclophophamide administered to children with recurrent brain tumors achieved OS of 33 months (alive at the time of reporting) in 10/16 patients [245]. In a phase-II study in 26 children with recurrent brain tumors treated with oral daily metronomic topotecan, 2 achieved complete remission 7 and 9.5 years after study and 4 achieved stable disease (median 4.6 months) [186]. In a 4- drug regimen consisting of weekly vinblastine, daily cyclophosphamide, twice weekly methotrexate and daily celecoxib, the treatment was well tolerated and achieved disease stabilization in refractory and relapsed pediatric tumors [239]. Here, out of

16 patients, 1 had an objective response and 4 showed disease stabilization lasting for at least 24 weeks with continued treatment. At the time of this report, 43% patients were alive. 4 grade IV toxicities and 24 grade-III toxicities were observed in only 62.5% patients.

Treatment with metronomic etoposide/cyclophosphamide/celecoxib regimen in children with refractory cancer achieved stable disease lasting for 20 weeks in 41% patients with only 35% patients experiencing grade III/IV toxicities [246]. In patients with recurrent medulloblastoma, all the five patients treated with metronomic temozolomide combined with radiation achieved complete response after median follow-up of 28 months, without any neurological toxicity [247].

1.11 Adverse effects of antiangiogenic therapy: Contrary to the earlier belief, LDM chemotherapy and antiangiogenic agents are not devoid of adverse effects. Though the incidence of acute toxicities is lower than conventional cytotoxic chemotherapy, serious delayed toxicities are observed with

53 antiangiogenic therapy. VEGF pathway inhibition causes hypertension and endothelial dysfunction in normal tissues [248, 249]. Chronic side effects are fatigue, diarrhea, nausea, GI and vaginal fistulations, skin/hair depigmentation, elevation of aspartate aminotransferase (AST) & alanine aminotransferase (ALT), and hematological side effects (neutropenia, lymphopenia and thrombocytopenia) [250, 251].

Myelosuppression is observed with VEGF pathway inhibiting antiangiogenic agents because VEGF is a cytokine that is also involved in hematopoesis. Leukopenia, lymphopenia and thrombocytopenia have been observed in patients undergoing pazopanib therapy [252].

On the other hand, LDM chemotherapy, even though, administered at a lower dose compared to conventional chemotherapy, leads to high total cumulated doses of anticancer agents after chronic administration [140]. This may cause delayed toxicities.

For example, metronomic temozolomide for 42 days caused prolonged thrombocytopenia and lymphopenia [140]. In prostate cancer patients treated with metronomic chemotherapy, thrombocytopenia was observed due to which one patient required repeated blood transfusion over a period of two months [253]. In mice, LDM cyclophosphamide did not cause intestinal toxicity, but did cause sustained lymphopenia [140].

Though the side effects of antiangiogenic therapy are manageable when administered as single agents, in combination with other chemotherapeutics these can cause dose- limiting toxicities [250, 254-256].

1.12 Topotecan

1.12.1 Chemistry:

Topotecan is a water soluble derivative of camptothecin [257]. Its chemical formula is

(S)-10-[(dimethylamino)methyl]-4-ethyl-4,9-dihydroxy-1H-pyrano[3',4':6,7] indolizino

[1,2-b] quinoline-3,14(4H,12H)-dione monohydrochloride. It undergoes pH dependent reversible hydrolysis from lactone form to carboxylate form. Lactone form predominates in acidic pH whereas carboxylate form predominates in basic pH. At any pH, equilibrium exists between the lactone form and carboxylate form. Lactone form is the active form and the concentration of lactone form determines its cytotoxicity.

1.12.2 Mechanism of action

Topotecan is a topoisomerase-I inhibitor [257]. Topoisomerase-I is the enzyme that catalyzes breaking of a single DNA strand during replication or transcription. This break is necessary to prevent the torsional strain which develops ahead of the replication fork which may prevent further transcription or replication. Topotecan forms a stable

DNA/topoisomerase-I complex which prevents the DNA relaxation and stops transcription or replication, thus leading to cell cycle arrest and apoptosis. Topotecan is most active in S phase of the cell cycle.

1.12.3 Pharmacokinetics and metabolism

Topotecan is a water soluble molecule which exhibits topoisomerase-I inhibition in its active lactone form. Topotecan is administered by oral and IV routes. Topotecan is rapidly absorbed from gastrointestinal tract, however efflux proteins such as PgP and

BCRP limit topotecan absorption [258]. Approximately 30% of administered topotecan

(IV) has been found to exist in lactone form at physiological pH. Its plasma protein binding ranges from 7-35%. Topotecan follows a linear pharmacokinetics. In vitro studies report that topotecan is metabolized to N-demethylated derivative [257].

Phenytoin, an inducer of CYP3A4, enhances the metabolism of topotecan, suggesting the involvement of this enzyme in its biotransformation [259]. Topotecan is also reported to decrease the bioavailability of docetaxel, which is also metabolized by

CYP3A4 [260].

The oral bioavailability of topotecan is approximately 40% and Tmax ranges from 1-2 hrs and can be affected by food [258, 261]. Plasma half-life is approximately 3 hrs. The volume of distribution ranging from 128 litres to 176 litres has been reported, indicating distribution in tissues [257, 262]. The plasma clearance is 30 litres/h/m2 [257]. Renal impairment reduces topotecan clearance. Topotecan penetrates the blood brain barrier.

1.12.4 Indications and adverse effects

Topotecan was approved for the treatment of ovarian cancer, cervical cancer (in combination with cisplatin) and NSCLC [257]. 5 days/ week administration of topotecan

(1.5-4.5 mg/m2) is the most common route of administration, followed by continuous IV infusion (0.6-2.1 mg/m2/day for 5days, 3-4 weeks) [257]. Oral dosing schedule consists

56 of 5 days/ week (2.3 mg/m2/day). The most common adverse effects of topotecan are myelosuppression (neutropenia), mucositis and diarrhoea [257] .

1.12.5 Resistance to topotecan

Efflux transporters are commonly reported cause of resistance to topotecan. BCRP has been reported to confer resistance to topotecan in breast cancer and ovarian cancer

[263, 264]. PgP and MRP-2 have been reported to confer resistance to breast cancer and leukemia cells [265]. Higher expression of PgP has been observed in patients with non-localized neuroblastoma [266]. Presence of side population cells over-expressing efflux transporters, ABCG2 and ABCA3 have been reported in neuroblastoma, glioblastoma, breast cancer and lung cancer cell lines [267].

1.12.6 Topotecan in pediatric cancers

Pediatric Preclinical Testing Program (PPTP) has demonstrated the efficacy of topotecan in ALL, Wilm‟s tumor, rhabdomyosarcoma, Ewing‟s sarcoma and neuroblastoma [268]. The efficacy of topotecan (5 days/ week for 21 days) by IV route has been correlated with systemic exposure and tumor extracellular fluid concentration in neuroblastoma xenograft models [269, 270]. The efficacy of low dose, protracted schedules (5days/ week for 12 weeks) of topotecan was proved in preclinical chemoresistant rhabdomyosarcoma models [143]. This was followed by the clinical trial of protracted low dose pharmacokinetically guided dosing (5 days/week for 2 weeks) of topotecan in heavily pretreated children with solid tumors [144]. This schedule was tolerated by patients and partial responses were observed in 5 out of 29 patients. Later clinical trial using same strategy i.e. pharmacokinetically guided dosing to achieve AUC of 80-120 ng/ml.h, was well tolerated and demonstrated 60% response rate in

57 neuroblastoma patients [146]. Topotecan achieved 40% response rate in patients with

Wilm‟s tumor [1.8 mg/m2 (range, 0.7 to 3.2 mg/m2), to achieve AUC of 80 ng/ml.h] when administered as 30 min infusion for 5 days/ week for 2 weeks [271].

Table 7 : Clinical trials of conventional topotecan in pediatric cancers

Indication Median Schedule Target AUC Result

Dose

Solid tumor [272] 3 mg/m2 30 min 80-120 Tolerated and 5

and 4 infusion, 5 ng/ml.h and out of 29 patients

mg/m2 days/week for 120-180 achieved PR.

2 weeks ng/ml.h

Neuroblastoma 2.7 mg/m2 30 min 80-120 60% response

[146] infusion, 5 ng/ml.h rate

days/week for

2 weeks

Wilm‟s tumor [271] 1.8 mg/m2 30 min 80 ng/ml.h 40% response

infusion, 5 rate

days/week for

2 weeks

Ewing sarcoma 0.3 Continuous Well tolerated,

osteosarcoma, soft mg/m2/day 21 days but limited activity

tissue sarcomas, infusion

medulloblastoma,

primitive

neuroectodermal

59 tumor, astrocytoma, neuroblastoma recurrent or refractory to conventional therapy [272]

High-risk 30 min 120-160 11.1% CR, 16.6% medulloblastoma infusion, 5 ng/ml.h PR, 47.2% stable and primitive days/week for disease neuroectodermal 2 weeks tumor [145]

High-grade glioma 0.4 mg/m2 Oral, daily Median MTD was

[273] 0.9 mg/m2 after 4

weeks. Objective

response in 2 of

13 patients

Recurrent/refractory 0.8, 1.1, Oral, 5 MTD was 1.8 solid tumors [274] 1.4, 1.8, days/week for mg/m2/d; stable

and 2.3 2 weeks, disease in 9 out

mg/m2/d every 28 days of 19 patients

for a

maximum of

six courses

Acute Dose 30 min MTD: 2.4 mg/m2/

Nonlymphoblastic escalation infusion for 5 day for 9 days; 1

Leukemia (ANLL) from 2.0 days CR and 4 PR in and ALL [275] mg/m(2)/d ANLL; 1 CR and

1 PR in ALL

1.12.7 Antiangiogenic property of topotecan

Topotecan has been reported to be antiangiogenic in preclinical trials. Its i.p. administration in mice orthotopic models of wilm‟s tumor (0.36-3.0 mg/Kg, 5 days/week,) and hepatoblastoma (0.6 mg/Kg, 5days/week) inhibited angiogenesis [174,

204]. It demonstrated cytotoxicity on HUVEC and inhibited in vivo angiogenesis in surgically implanted sponge discs (1 mg/Kg, every other day for 2 weeks, i.m.) [276]. In mouse corneal model, it inhibited angiogenesis when administered at 3.5 mg/Kg, daily for 6 days by i.m. route [277]. Apart from direct cytotoxicity to endothelial cells, topotecan exerts its antiangiogenic activity by inhibition of HIF-1 alpha and consequent down-regulation of VEGF expression. In vitro, topotecan has reduced the expression of

HIF-1 alpha and VEGF in ovarian cancer cells and neuroblastoma cells [278, 279, 280].

Daily oral administration [1 mg/kg, 10 doses] is reported to have significantly reduced

HIF-1 alpha in glioblastoma xenografts compared to intermittent dosing (10 mg/kg, every 4 days×3) [281]. Treatment of glioblastoma patients with oral topotecan (1.2 mg/m2, 5 days/week for 2 weeks) resulted in significant reduction of HIF-1 alpha in tumor biopsy specimens [282].

1.12.8 Metronomic topotecan

Oral metronomic topotecan (0.5, 1.0 and 1.5 mg/Kg, daily, 28 days) was found to be antiangiogenic compared to oral MTD topotecan (7.5 and 15 mg/Kg, weekly, 28 days) in mice models of ovarian cancer [278]. Oral metronomic topotecan (1 mg/Kg daily) significantly reduced the CEPs after 7 and 28 days treatments in another ovarian cancer mice model [170]. In colon cancer orthotopic model, oral metronomic topotecan

(1mg/Kg, daily) prolonged mice survival and significantly delayed liver metastasis [283].

Metronomic topotecan (20µg/Kg, daily intratumoral injection) demonstrated higher

62 antitumor efficacy compared to weekly topotecan (160 µg/ml, intratumoral injection) in prostate cancer xenografts [161]. In a phase-II study conducted in recurrent or refractory pediatric brain tumor patients, metronomic topotecan (0.8 mg/m2/day for 21 days, every 28 days) was well tolerated (not more than 15% patients with grade 4 toxicity) and induced remission in 2 of 25 patients (7 and 9.5 years off study) and stable disease in 4 other patients (median 4.6 months) [186].

1.13 Pazopanib

1.13.1 Chemistry

Pazopanib is a water insoluble RTKI, discovered and marketed by Glaxosmithkine. Its

IUPAC name is 5-[[4-[(2,3-Dimethyl-2H-indazol-6-yl) methylamino]-2- pyrimidinyl]amino]-2-methylbenzolsulfonamide. Its molecular weight is 437.5 g/mole

(base) and 474 g/mole (hydrochloride salt).

1.13.2 Mechanism of action

Pazopanib is a targeted antiangiogenic drug, which acts by inhibiting VEGFR-1

(IC50=10nM), VEGFR-2 (IC50=30nM), VEGFR-3 (IC50=47 nM), PDGFR-α, PDGFR-β and c-kit receptors [124]. It inhibits the VEGF induced proliferation of HUVEC

(IC50=21nM) more selectively than the bFGF induced proliferation (IC50=721nM).

1.13.3 Pharmacokinetics and metabolism

Pazopanib is metabolized by CYP3A4. Its half-life is 35 hrs. It is highly protein bound in plasma (> 99%). Plasma concentration of at least 18 µg/ml (40nm) has been found to inhibit VEGFR2 phosphorylation and subsequent angiogenesis [124]. Its metabolism can be inhibited by CYP3A4 inhibitors. Pazopanib has been reported to be a weak inhibitor of CYP3A4 and CYP2D6, whereas it had no effect on CYP1A2, CYP2C9, and

CYP2C19 in advanced solid tumor patients. Low or high fat diet has been found to cause two-fold increase in bioavailability of pazopanib, compared to fasting state, as observed in a Phase-I study in patients with advanced solid tumors [284].

1.13.4 Preclinical antitumor activity

Pazopanib has shown antitumor activity in preclinical mice models of breast cancer, colon cancer, multiple myeloma, renal cancer, lung cancer, head & neck cancer and prostate cancer [124]. In multiple myeloma, it demonstrated direct cytotoxic effect on tumor cells. Pazopanib demonstrated efficacy and tolerance in pediatric cancer models

(rhabdomyosarcoma and Ewing‟s sarcoma) in a PPTP study [139].

1.13.5 Clinical activity of pazopanib

Pazopanib has been approved for treating advanced stage RCC and adult soft-tissue sarcoma [285, 286]. In RCC clinical trial, oral daily administration of pazopanib (800 mg/day) achieved a PFS of 9.2 months whereas that in placebo group was 4.2 months.

800 mg daily administration achieves a steady state trough concentration of 40 µm

(18µg/ml) [287].

The clinical trial of pazopanib in combination with cytotoxic agents such as gemcitabine

(advanced solid tumors) [288], paclitaxel and carboplatin(advanced solid tumors and

64 gynecologic cancer) [256, 289], cyclophosphamide (platinum resistant ovarian cancer)

[229] have been reported.

Table 8: List of clinical trials involving single agent pazopanib

Disease Dose Result Side effect (> grade

3) (mg)

Advanced urothelial 800 Objective Response in Hypertension, fatigue,

cancer (Phase-II) 17.1% patients GI & vaginal fistulation

[290]

Advanced thyroid 800 No RECIST responses Hypertension (13%)

cancer; (Phase-II) Pharygeo-larygeal Median time to [291] pain (13%) progression: 62d;

median survival time:

111d.

Non-adipocytic soft 800 Median PFS: 4.6m Fatigue (65% for

tissue sarcoma; (pazopanib); 1.6m pazopanib; 49% for

(Phase-III) [292] (placebo) placebo)

OS: 12.5m (pazopanib); Diarrhea (58% for

10.7m (placebo) pazopanib vs 16% for

placebo)

Nausea (54% for

pazopanib vs 28% for

placebo)

Hypertension: (41%

for pazopanib vs 8%

for placebo)

Weight loss (48% for

pazopanib vs 20% for

placebo)

Castration sensitive 800 4/18 patient had Diarrhea (2 events); prostate cancer; progressive disease; Hypertension (3

Phase-II [293] 13/18 disenrolled due to events)

grade 1/2 toxicity

Recurrent/metastatic 800 PR: 6.1%; Stable Fatigue (15.2%); nasopharyngeal Disease (SD): 48.5%; Hand-foot syndrome carcinoma; Phase-II progressive disease: (15.2%); Anorexia

[294] 33.3%. (9.1%), Diarrhea

(6.1%); vomiting 22% patients receiving (6.1%) >4months treatment had

PR/stable disease of at

least 6 months

HCC; Phase-I [295] Escalation 73% PR or SD 2/5 patients

(200-800) administered 800 mg

had Dose Limiting

Toxicity. The common

adverse effects were

Diarrhea, skin hypo

pigmentation and

AST.

Metastatic thyroid 800 PR: 49% Fatigue (29/39); cancers; Phase-II skin/hair pigmentation

[296] (28/39); diarrhea

(27/39);

nausea(27/39)

Recurrent/metastatic 800 PR:5%; SD:55% (20% ≥ Transaminitis, breast cancer; 6m); Progressive hypertension,

Phase-II [297] disease: 35% (median neutropenia, intestinal

time to hemorrhage

progression:5.3m)

Recurrent ovarian 800 Overall response: 18% Elevation of ALT (8%) cancer; Phase-II and AST (8%)

[298]

NSCLC, Phase-II 800 Tumor reduction in 86% None (Grade 2

[299] patients hypertension, diarrhea

and fatigue)

Recurrent 800 PR: 2 patients Leukopenia, glioblastoma, Phase- (radiographic); 9 lymphopenia,

II [252] patients had reduced thrombocytopenia,

contrast enhancement, increase in AST and

but <50% tumor ALT, CNS

reduction. hemorrhage. 8/35

patients required dose Median PFS: 12 weeks; reduction. 1 patient had PFS >6m

Metastatic RCC; Median PFS 9.2m for Diarrhea,

Phase-III [126] pazopanib vs. 4.2m for hypertension, hair

placebo; objective depigmentation,

response: 30% for nausea, anorexia,

pazopanib vs 3% for vomiting

placebo.

Relapsed/ refractory 800 PFS and OS prolonged Hypertension,

Soft-tissue sarcoma compared to historical hyperbilirubinemia,

[286] controls treated with fatigue

second line

chemotherapy.

Multiple myeloma, 800 No patient achieved

Phase-II [300] remission; 10/16

69 patients had progressive disease before 6wk.

Median time for progression was 52d.

1.13.6 Comparison of pazopanib with other VEGF receptor inhibitors

Comparison of sunitinib and pazopanib in renal cell cancer showed no significant

difference in overall toxicity (grade 1-4) [301]. However, grade 2-4 toxicity was

significantly higher in sunitinib treated patients; mucositis (42% in sunitinib vs 15% in

pazopanib; P=0.01) and nausea (9% for sunitinib vs none for pazopanib). Diarrhea and

fatigue were significantly higher in pazopanib treated patients. Hematological toxicity

was 4% for sunitinib and none for pazopanib. Pazopanib was found to be more

effective than lapatinib in terms of PFS and OS in patients with advanced and

recurrent cervical cancer [302]. Here, grade 3 adverse effects (diarrhea) for pazopanib

and lapatinib were 11% and 13% respectively. Grade 4 adverse effects for pazopanib

and lapatinib were 12% and 9% respectively.

1.14 Biomarkers of antiangiogenic therapy

Biomarker is defined as a distinct biological indicator of a process, event or a condition [303]. In the context of cancer therapeutics, there are following types of biomarkers [303]:

 Prognostic biomarkers provide information about overall cancer outcome,

regardless of therapy.

 Predictive biomarker is used to estimate the response of a specific patient(s)

to specific treatment(s) compared with another treatment.

 Pharmacodynamic biomarkers are those whose changes in response to

treatment are associated with target modulation.

 Surrogate markers are intended to serve as substitute for a clinical end point.

1.14.1 Need for markers of antiangiogenic therapy Since antiangiogenic therapies have indirect action on the tumor, the techniques used for response monitoring of conventional therapies are not useful for antiangiogenic therapy. Newer biomarkers for antiangiogenic therapy are needed due to following reasons:

 Predicting response: Due to heterogeneity of the tumor biology, inter-patient

variability in response to antiangiogenic therapy is observed. A predictive

biomarker can be useful in selecting patients who are likely to respond to the

therapy [304] .

 To optimize the dose and schedule of anti-angiogenic therapy: Unlike

cytotoxic drugs, MTD is not associated with maximum efficacy of

antiangiogenic therapy. Pharmacodynamic biomarkers can enable the design

of Optimal Biologic Dose [303, 305].

 Monitoring the therapeutic response: Response Evaluation Criteria In Solid

Tumors (RECIST), which relies on changes in tumor burden in response to

therapy, may not accurately predict the response to antiangiogenic therapy.

First, unlike cytotoxic chemotherapy, antiangiogenic therapy is cytostatic,

hence, tumor shrinkage may not be a realistic measure of efficacy [306]. For

instance, increase in OS with combination of bevacizumab and chemotherapy

was not associated with tumor shrinkage. Second, the purpose of

antiangiogenic therapy in the present context is to target microscopic disease

rather than large tumors [306]. The changes in tumor vasculature caused by

antiangiogenic therapy may not necessarily lead to reduction in tumor volume.

Thus, specific biomarkers of antiangiogenic therapy are valuable for

monitoring its response and resistance.

 To identify the time window for effectiveness because the benefit of

antiangiogenic therapy is often transient. Extending the therapy beyond this

window is not only ineffective but can also cause severe toxicities and may

also enhance the invasiveness of tumors [303, 307].

 Contrary to earlier expectations, evasive and adaptive resistance to

antiangiogenic therapy is responsible for its loss of efficacy [305]. Hence,

biomarkers of resistance will be useful in identifying new targets.

 Antiangiogenic therapies have unique adverse effects, hence predictive

biomarkers of toxicity can identify the patients who are at higher risk of

serious toxicity [303].

 Reduce the attrition in clinical trials: Validated biomarkers can improve the

efficacy and safety of a candidate therapy by enabling individualization of

optimal biological dose in clinical trials, therefore, reduce the clinical trial

failure [303].

1.14.2 Common biomarkers of antiangiogenic therapy

 Hypertension: It is the most predictive biomarker for anti-VEGF therapy. Rise

in blood pressure has been observed after treatment with bevacizumab and

RTKIs. Diastolic blood pressure ≥ 90 after single agent treatment correlated

with increased OS in five Phase-II trials in refractory or advanced melanoma,

RCC, NSCLC and thyroid cancer [308].

 Circulating protein markers: Several proangiogenic and antiangiogenic

molecules have been investigated for their utility as the markers of

angiogenesis and antiangiogenic efficacy. VEGF is the most widely

investigated marker of angiogenesis and antiangiogenic efficacy [309].

Increased levels of VEGF have been observed in rhabdomyosarcoma

patients. Increase in circulating VEGF levels have been observed with

bevacizumab and RTKIs. Elevated VEGF levels correlated with the clinical

benefit of sunitinib in patients with RCC, whereas PDGF and VEGFR2 had no

predictive value [310].

As the markers of patient selection, low VEGFR2 and high VEGF correlated

with low OS and bevacizumab treatment enhanced the survival in breast

cancer patients with low VEGF [311]. In another study, no association

between pre-treatment VEGF and clinical response to bevacizumab was

observed in RCC patients [312]. Baseline levels of VEGF-C were a predictor

of clinical outcome in sunitinib treated HCC patients [313]. VEGF, also as a

marker for therapeutic efficacy, has given mixed results. In Phase-I trial in

rectal cancer patients, bevacizumab increased VEGF, PlGF and Circulating

Endothelial Cells (CEC) [314], whereas, in phase-III trial in metastatic renal

cell carcinoma did not correlate with the response [312]. Sunitinib efficacy has

correlated with increase in VEGF and/or PlGF in several studies [315, 316].

Reduction in VEGFR2 levels has also correlated with sunitinib efficacy in

clinical trials [316]. However, in another study with vandetanib, increase in

VEGF did not correlate with the efficacy [317]. In renal cell carcinoma patients

treated with sorafenib, increase in VEGF inversely correlated with patients

survival [318]. Clinical trial with metronomic vinblastine and celecoxib, VEGF,

bFGF, vascular cell adhesion molecules (vCAM-1), soluble intercellular

adhesion molecule (sICAM), TSP-1 and endostatin did not correlate with the

efficacy [131, 237]. In a trial involving metronomic, S1 upregulated TSP1,

whereas vandetanib up-regulated VEGF. The combination, S1+vandetanib,

up-regulated both the markers [217]. TSP1 has been used as marker of

efficacy of metronomic chemotherapy [319]. Higher levels of SDF-1 alpha, IL-

6 and soluble kit correlated with worse prognosis in sunitinib treated HCC

patients [320]. In NSCLC, treatment with pazopanib caused significant

decrease in VEGFR2 and increase in PlGF, which was associated with tumor

shrinkage.

 Circulating Endothelial Cells (CECs) and Circulating Endothelial Progenitors

(CEPs): CECs are the activated endothelial cells which are shed off from the

vasculature due to elevated angiogenesis [321]. CEPs are mobilized from

bone marrow and are recruited into tumor tissue where they differentiate into

mature endothelial cells. In normal blood CEC and CEP constitute 0.01%-

0.0001% of mononuclear cells, however in cancers their levels are elevated.

Fluorescence Activated Cell Sorting (FACS) is used for the quantitation of

CECs and CEPs in peripheral blood. CEPs are identified by the presence of

VEGFR2, progenitor markers (CD133 and CD34), and the absence of

hematopoietic progenitor marker CD45 [322]. CECs are identified as the cells

lacking CD133, CD34 and expressing markers of mature endothelial cells

(CD31, von Willebrand Factor and VE cadherin). Elevated CECs have been

75 reported in lymphoma and breast cancer patients compared to healthy subjects, which correlated with increase in VEGF levels [321] . Significantly higher CEC levels has been reported in progressive disease compared to stable disease in adult cancer patients [321]. In another breast cancer study it correlated with microvessel density [321]. In breast cancer patients, CECs have been reported be a predictor of response to metronomic chemotherapy, with higher CEC levels associated to poor prognosis [323]. CECs have been reported as a predictive marker for bevacizumab therapy in colorectal cancer patients [324].

CEPs are a highly predictive marker for angiogenesis and antiangiogenic therapy. High levels of CEPs have been observed in patients with pediatric malignancies, compared to the healthy subjects [325]. In this study, significantly higher CEP levels was observed in children with metastatic disease compared to those with localized disease, whereas VEGF, VEGFR2 and CECs did not correlate with the disease stages. In another study, endothelial colony forming units characterized as CD31+, CD34+, CD45-,

AC133-, CD14-, CD41a-, CD235a-, 7AAD- blood cells were elevated in pediatric solid tumor patients compared to healthy controls [326]. Elevated levels of AC133+, CD34+ and VEGFR2+ CEPs have been reported in hemangioma [327]. Elevated levels of CD34+/Flk-1+ CEPs have been reported in Kaposi‟s sarcoma patients [328].

In adult cancers, high VEGFR2+ CEPs have been correlated with the disease stage. High CEP level correlated with poor prognosis in advanced HCC

76 patients treated with metronomic chemotherapy and sorafenib [329].

Metronomic chemotherapy prevents mobilization of CEPs from bone marrow.

Therefore, CEPs were successfully employed as pharmacodynamic marker to establish optimum biological dose of metronomic cyclophosphamide, vinblastine, vinorelbine and cisplatin in mice models of melanoma, breast cancer and erythroleukemia [330]. It has correlated with the antiangiogenic activity of axitinib in neuroblastoma xenograft model [331]. High levels of

CEPs after treatment with sunitinib correlated with poor prognosis in HCC patients. CEPs are also a reliable predictive marker of antiangiogenic efficacy

[320]. In metastatic RCC patients treated with sunitinib or sorafenib, baseline

CEP levels correlated with PFS and OS [332]. In our previous study involving

LDM cyclophosphamide and celecoxib in non-hodgkin‟s lymphoma patients, after 8.4 months (median) follow-up CECs, CEPs and VEGF remained low in responders, whereas VEGF levels increased in non-responders [236].

CEPs and CECs, as surrogate markers, are subject to bias due to inter- patient variability and the differences between the features of various cancers.

In a previous study conducted by Dr. Baruchel‟s laboratory in patients from three participating hospitals of North America, CEC and CEP levels did not correlate with the disease stage and clinical features in osteosarcoma patients

[333]. According to authors, there are several reasons for the lack of predictability in this study. First possibility could be methodological flaws, which the authors contend, would be unlikely because the sample collection and flow cytometry methods were conducted by experienced technicians using pre-validated method. Second possibility is the inter-patient variability.

Third possibility is the differences between angiogenic behaviour of different

types of cancers, e.g., the factors that regulate angiogenesis and the

difference in the characteristics of tumor vasculature between carcinoma and

sarcoma.

 Microvessel density: Microvessel density is the most definitive marker of

angiogenesis and the best predictor of disease outcome among all the

markers of angiogenesis. Microvessel density, as determined by

immunohistochemistry or immunofluorescence, employs endothelial cell

markers such as von Willebrand factor, CD31 and CD34. It has been an

indicator of worse prognosis in many cancers, including pediatric cancers

such as multiple myeloma [334], Wilm‟s tumor [335], osteosarcoma [67, 89],

retinoblastoma [336], neuroblastoma [67] and glioma [94].

Microvessel density is also an appropriate marker for patient stratification for

antiangiogenic therapy, with higher microvessel density indicating better

response [337]. Microvessel density has been employed as a marker of

antiangiogenic efficacy in preclinical studies involving single agents such as

sunitinib [338, 339], pazopanib [171], sorafenib [340], axitinib [222],

metronomic topotecan [278] and metronomic cyclophosphamide [222]. It has

correlated with the superior efficacy of the combination of metronomic

chemotherapy with targeted antiangiogenic drugs, compared to the single

agents in several studies such as S1+vandetanib (HCC) [217], topotecan +

pazopanib (ovarian cancer) [171], taxanes+AEE788 (ovarian cancer) [341],

cyclophosphamide+OXi4503 [221], cyclophosphamide +axitinib [222].

 Imaging markers: Imaging markers are the most convenient and least

invasive biomarkers of antiangiogenic therapy. Changes in blood volume,

blood flow and vascular permeability are the parameters employed for the

assessment of efficacy of antiangiogenic therapy, using DCE-MRI or CT scan.

Decrease in Ktrans or Ki (measure of vascular permeability), decrease in blood

flow and blood volume have correlated with the efficacy of various

antiangiogenic therapy in clinical trials [303]. Changes in Dynamic Contrast

Enhanced MRI marker K (trans) correlated with the tumor response and

survival in HCC patients treated with metronomic tegafur/uracil+sorafenib

[342]. High FET uptake (PET) is used as response to metronomic

temozolomide+celecoxib in recurrent glioblastoma [343]. VEGF labeled with

technicium 99 binds with VEGFR and gets internalized, and therefore has

been used as imaging marker for functionally active VEGF receptors after

treatment with pazopanib in colon cancer mice model [344].

 Other biomarkers: Tumor pO2, determined by oxymetry, has been used as a

marker to identify therapeutic window [345]. Alpha-fetoprotein, as an

antiangiogenic marker, correlated with the response of metronomic 5-

fluoropyrimidine and sorafenib, bevacizumab or thalidomide in HCC patients

[346]. Elevated corpuscular volume has been used as an antiangiogenic

marker for metronomic capecitabine and cyclophosphamide in combination

with bevacizumab [226].

1.15 Resistance mechanisms to antiangiogenic therapy

The success of antiangiogenic therapy observed in preclinical studies has not translated into the clinic. The advantage of VEGF pathway inhibiting antiangiogenic agents has been modest. The antiangiogenic agents, even in combination with chemotherapeutics, have not produced any significant enhancement in overall survival [347-351]. Cancers either remain refractory to the maintenance regimen of antiangiogenic therapy or show recurrence and relapse.

Inhibition of angiogenesis as a possible strategy to avoid resistance was put forth by

Kerbel et al (1991), the rationale being that endothelial cells are genetically stable unlike the tumor cells [352]. However, afterwards resistance to antiangiogenic therapies were observed after an initial response phase [196]. These resistance mechanisms are different from those encountered with agents directly targeting the tumor cells. In case of antiangiogenic therapy, these mechanisms are indirectly activated in response to the changes in tumor microenvironment caused by the treatment. In prostate cancer PC3 xenografts 25 differentially expressed genes were identified indicative of chemoresistance to metronomic cyclophosphamide, 3 of which were thioredoxin containing protein 5, cathepsin B, and annexin A3; [353].

Following are some of the mechanisms of antiangiogenic resistance:

1.15.1 Up-regulation of angiogenic factors Up-regulation of pro-angiogenic factors and their redundancy have been proved as a mechanism of antiangiogenic resistance in several studies. In one study, mice bearing pancreatic islet tumors treated with DC101 initially responded to the therapy as indicated by reduction in tumor size and vascularity [354]. However, this was followed by tumor re-growth promoted by alternate pro-angiogenic factors such as

FGF-1, FGF-2, ephrin A1, ephrin A2 and Angiopoetin-1. When these mice were treated with FGF-trap at the peak of responses, the tumor recurrence was delayed, indicating the role of alternate FGF in therapeutic resistance to DC101. In mice implanted with HIF-1 alpha knockout colon cancer cells, angiogenesis occurred by the up-regulation of IL-8 by HIF-1 alpha independent mechanism [355]. Up regulation of VEGF, PDGF-A and FGF-2 has been reported in response to treatment with angiogenic inhibitors (TSP1, tumstatin and endostatin) [356]. Glioblastoma patients treated with VEGFR2 inhibitor cediranib initially showed response which was followed by a relapse [255]. Here, the FGF-2 level in blood was higher in relapsing phase than in response phase. Sunitinib increases the plasma VEGF level; hence it is a marker for antiangiogenic efficacy of sunitinib. However, when non- tumor bearing mice were treated with sunitinib, the plasma levels of proangiogenic factors (G-SCF, SDF-1α, SCF, osteopoetin) increased which represents systemic response to the treatment [357]. Therefore, the use of pro-angiogenic factors as the markers of efficacy has been questioned, as VEGF can be a toxicity marker during the therapy and also a marker of drug resistance after the withdrawal of therapy.

1.15.2 Involvement of bone marrow derived cells Hypoxia induced HIF-1alpha up-regulates several factors which are involved in mobilization of CEP to hypoxic areas [358]. In glioma, cytokines such as VEGF, bFGF, SDF-1 and Tie-2 are known to cause antiangiogenic resistance by promoting

CEP recruitment into the tumor tissues [359]. In glioblastoma multiforme, HIF-

1alpha+ cells, but not HIF-1 alpha knock-out cells, caused the bone marrow recruitment of CD45+ cells [360].

1.15.3 Pericyte coverage Pericyte recruitment is a critical step in vascular maturation. Tumor vasculature is characterized by immature blood vessels without pericyte coverage or loosely associated pericytes around endothelium [361]. Absence of pericyte coverage renders the endothelial cells susceptible to antiangiogenic therapy. Pericyte recruitment occurs by: (a) migration of cells positive for α-smooth muscle actin from mature blood vessels [362]; (b) recruitment of PDGFRβ+ perivascular progenitor cells from bone marrow and its subsequent differentiation in the tumor tissue [42].

The molecules that play key role in pericyte recruitment are PDGF-B and its receptor

PDGFRβ. Abramsson et al (2003) and Lindblom et al. (2003) demonstrated that extracellular gradient of PDGF-B, secreted by endothelial cells, is required for the mobilization and tight contact of PDGFRβ+ perivascular cells to the tumor endothelium [363]. In another study, PDGF-C is reported to be involved in glioblastoma resistance to DC101 by inducing pericyte recruitment to the tumor endothelium [361]. MMP-9 expressed by vascular cell is also reported to contribute to pericyte recruitment in SK-N-BE(2) neuroblastoma xenografts [364]. Cross-talk between pericytes and endothelial cells resulting from surface to surface contact causes up-regulation of VEGF-A within the endothelial cells, which further upregulates antiapoptotic protein BCl-w [365].

Role of pericyte coverage in antiangiogenic resistance have been reported in several studies. Increased presence of pericytes (NG2+, PDGFRβ+, desmin+) has been observed in melanoma tumors resistant to antiangiogenic drug PTK787/ZK222584 and bevacizumab resistant tumors obtained from melanoma patients [366]. In a

82 retinoblastoma mice model, though antiangiogenic therapy reduced the total microvessel density, it did not reduce the number of mature vessels [367].

1.15.4 Increased invasive potential of tumor cells Even though it retards the tumor growth, several studies have reported enhanced invasive potential of tumor cells in response to antiangiogenic therapy. In mice models of glioblastoma, treatment with anti-VEGF antibody or anti-VEGFR2 antibodies prolonged the mice survival, but the treated tumors showed satellite tumor formation and higher metastatic potential [368]. In other models of glioblastoma and pancreatic cancers, VEGF pathway inhibiting antiangiogenetic therapy or genetic ablation of VEGF-A resulted in increased malignant potential of tumor cells [360,

369]. Sunitinib treatment in mice orthotopic and IV metastatic models of melanoma and breast cancer have revealed increased metastatic potential of tumor cells [370].

Metronomic temozolomide + celecoxib treated glioblastoma multiforme patients had distant metastasis along neural axis [371]. This study reports that even though VEGF signaling supports the tumor growth by promoting angiogenesis, it also prevents the invasiveness in cancer cells. This conclusion, however, conflicts with the studies which correlate VEGF with metastatic stage in neuroblastoma, sarcoma and glioblastoma [372-374]. Taken together, these observations may suggest that though pro-angiogenic activity of VEGF facilitates the metastasis of invasive tumor cells from the primary site and subsequent tumor growth at metastatic sites, it may not be involved in enhancing the inherent invasive potential of tumor cells.

1.15.5 Metabolic switch of tumor cells Metabolic switch to glycolysis is a mechanism of antiangiogenic resistance wherein, tumor cells acquire ability to survive and proliferate even in hypoxia induced by

83 antiangiogenic agents. Anti-VEGF treatment with bevacizumab is reported to have encountered resistance in mice models of glioblastoma, ovarian cancer and squamous cell carcinoma by this mechanism [375, 376]. The metabolic switch in hypoxic cancer cells is induced by HIF-1 alpha, where it upregulates the enzymes involved in glucose transport and glycolysis. In this type of resistance, hypoxic cells compensate for this low output of ATP by increasing the availability of glucose molecules and up regulation of glycolytic enzymes, Glut-1 and hexokinase-II, commonly employed as markers of elevated glycolysis [377].

1.15.6 Involvement of cancer stem cells Cancer stem cells have been implicated in the resistance to antiangiogenic therapy.

The residual stem cell population has been reported to be responsible for relapse after metronomic cyclophosphamide treatment in HCC xenografts [378]. Hypoxia induced by treatment with sunitinib has been reported to increase the stem cell population in breast cancer xenografts [379]. Hypoxia induced by antiangiogenic therapy possibly enhances cancer stem cell population, either by selection of hypoxia-resistant cancer stem cells which evade therapy or due to acquisition of stem-like phenotype by cancer cells [380]. HIF-1 alpha is known to induce the expression of Oct-4 and activate notch pathway which are required for self-renewal and multipotency of cancer stem cells [381]. Previously, the role of hypoxia in enhancing the stem-like phenotype in neuroblastoma cells have been proved in our laboratory [61]. Cancer stem cells isolated from glioma patients have been found to express elevated VEGF than non-stem like cancer cells, therefore can be comparatively resistant to antiangiogenic therapy [95].

2 Thesis overview

2.1 Rationale Neuroblastoma and sarcoma are the most common extracranial childhood solid tumors.

Though the 5-year overall survival rates in childhood cancers have reached 80%, the survival rates of a subtype of these aggressive cancers are still below 50% and 70%, respectively [2]. New strategies targeting alternate mechanism of resistance and minimal residual disease are needed to improve the outcome of these aggressive cancers. Discovery of new maintenance treatment regimens targeting residual disease represent an exciting opportunity to improve survival rate. Most of these recurrent patients would have received dose intensifying chemotherapy as an induction or consolidation therapy, therefore minimally toxic regimen targeting alternate targets should be used in this context.

Targeting angiogenesis by the use of LDM chemotherapy and targeted antiangiogenic therapy have demonstrated efficacy in several types of cancers including pediatric cancers [109]. Here, we have used LDM scheduling of topoisomerase-I inhibitor topotecan. Drawbacks to using conventional topotecan include resistance of tumor cells, cross-resistance with topoisomerase-II inhibitors in neuroblastoma cell lines and acute toxicity such as neutropenia [257, 263, 382]. LDM regimen of topotecan can possibly overcome some of these limitations.

Despite its advantages, response to single agent LDM chemotherapy is short-lived.

VEGF, a survival factor for endothelial cells, may be responsible, at least in part, for the loss of antitumor efficacy of LDM chemotherapy [383]. As a result, combining metronomic therapy with agents that target the VEGF-signalling pathways is associated

85 with an overall increase in anti-tumor activity [98, 225, 341, 384]. Pazopanib, a second generation antiangiogenic RTKI, has lesser cardiotoxicity than sunitinib. Since previously, patients pre-treated with anthracyclins were intolerant to sunitinib, we believe that a lesser cardiotoxic drug like pazopanib will be a better substitute for sunitinib in such patients [248]. Previously, two independent studies have reported the marked superiority of the combination of daily, oral, LDM topotecan and the small molecule RTKI pazopanib compared to either single agent therapy in models of advanced ovarian cancer [170, 171].

Neuroblastoma was one of the preclinical tumor models to validate the concept of combining metronomic chemotherapy with anti-angiogenic therapy [98]. However the mechanism of increased efficacy and safety of metronomically administered drug combinations and their pharmacokinetics (PK) has never been studied widely in pediatric cancers. Despite reports regarding anti-tumor activity of such combinations, their effectiveness in a particular pediatric cancer model cannot be predicted on the basis of its effects on other cancer models. In our experience combination of metronomic cyclophosphamide and sunitinib did not have any advantage over sunitinib monotherapy when tested in a neuroblastoma preclinical xenograft model

[338]. Also, in a previous study, the combination of axitinib with metronomic cyclophosphamide was less effective than metronomic cyclophosphamide alone in gliosarcoma model [222]. Therefore, the benefit of combining metronomic chemotherapy with a particular RTKI should be confirmed preclinically and the proper dose and preclinical PK need to be established before moving to phase-I clinical trials.

Other than efficacy and safety, therapeutic resistance is another concern with antiangiogenic therapy [350]. Despite the initial enthusiasm, metronomic chemotherapy and targeted antiangiogenic agents, even in combination with chemotherapeutics, have not produced any significant enhancement in overall survival [348, 349, 351].

Therefore, there is a need to understand the factors contributing to resistance of cancers to prolonged antiangiogenic therapy.

Considering the need of effective maintenance therapies in pediatric solid tumors and to gain understanding of tumor response to prolonged antiangiogenic therapy, we have attempted to test the efficacy, safety in mice models of pediatric solid tumors, and to understand the changes in tumor behaviour in response to prolonged therapy with LDM topotecan and pazopanib in a neuroblastoma mouse model.

2.2 Hypothesis

• The combination of LDM topotecan and pazopanib will have superior efficacy than either single agents.

• A time-response study will inform us about the changes in tumor behaviour in response to prolonged therapy with LDM topotecan and pazopanib.

2.3 Objectives  To test the efficacy and safety of LDM topotecan and pazopanib in

immunodeficient mice models of extracranial pediatric solid tumors:

neuroblastoma, rhabdomyosarcoma and osteosarcoma.

 To conduct a time–response study to investigate the changes in tumor

xenograft behavior in response to prolonged therapy with LDM topotecan and

pazopanib in a neuroblastoma mice model.

2.4 Thesis outline

2.4.1 To test the efficacy and safety of LDM topotecan and pazopanib in immunodeficient mice models of extracranial pediatric solid tumors: neuroblastoma, rhabdomyosarcoma and osteosarcoma.

 In-vitro effect of topotecan and pazopanib on endothelial cell lines

(HUVEC) and tumor cell lines.

 In-vivo efficacy of LDM topotecan and pazopanib in Non-Obese

Diabetic / Severely Combined Immunodeficient (NOD/SCID) mice

models of neuroblastoma (2 subcutaneous xenograft models and 2 IV

metastatic models), osteosarcoma (KHOS subcutaneous xenograft

model) and rhabdomyosarcoma (RH30 subcutaneous xenograft

model).

 Biomarkers of antiangiogenic efficacy: CEP and CEC were measured

by Fluorescence Assisted Cell Sorting (FACS). Microvessel densities

of the tumor xenografts were done by immunohistochemistry /

immunofluorescence staining of markers of endothelial cells.

 Safety of LDM topotecan and pazopanib: The markers of toxicity were

White Blood Cells (WBC) levels in blood in neuroblastoma and both

sarcoma models and bone marrow progenitor assay of mice bone

marrows collected from rhabdomyosarcoma mice model.

 Pharmacokinetic (PK) studies: The 24 h bioavailability of topotecan and

pazopanib in mice plasma after single drug administration were

determined in non-tumor bearing mice.

2.4.2 To conduct a time–response study to investigate the changes in tumor xenograft behavior in response to prolonged therapy with LDM topotecan and pazopanib in a neuroblastoma mice model.

 Treatment of mice bearing SK-N-BE(2) xenografts with LDM topotecan

and pazopanib for different durations or till the end point, whichever is

earlier.

 Comparison of tumor microvessel densities and pericyte coverage

between different treatment groups by double staining for the marker of

endothelial cells (CD31) and pericytes (α-smooth muscle actin) and

detection by immunofluorescence.

 Comparison of hypoxia and proangiogenic expression by

immunohistochemistry, western blot and Real-time Polymerase Chain

Reaction (Real Time PCR) between treatment groups.

 Comparison of apoptotic indices between different treatment groups in

treatment groups by immunofluorescence of cleaved caspase-3.

 Comparison of proliferative and mitotic indices between different

treatment groups by immunohistochemistry of tumor xenografts.

 Comparison of markers of glycolysis between different treatment

groups by immunohistochemistry.

3 Materials

3.1 Drugs and reagents Topotecan (S)-10-[(dimethylamino)methyl]-4-ethyl-4,9-dihydroxy-1H-yrano [3',4':6,7] indolizino[1,2-b] quinoline-3,14 (4H,12H)-dione monohydrochloride and Pazopanib (5-

[[4-[(2,3-Dimethyl-2H-indazol-6-yl)methylamino]-2-pyrimidinyl]amino]-2-methyl benzolsulfonamide were provided by GlaxoSmithKline, Collegeville, PA. Topotecan-d6 was purchased from Toronto Research Chemicals (catalogue # T542502).

3.2 Cell lines Neuroblastoma cell lines (SK-N-BE(2) and SH-SY5Y,) osteosarcoma cell line (KHOS) and rhabdomyosarcoma cell lines (RH30 and RD) and HUVEC were obtained from

American Type Culture Collection (ATCC) (Manassas, VA); BE(2)-c was obtained from

Dr. Michelle Haber (Children's Cancer Institute for Medical Research, Lowry Cancer

Research Centre, Randwick, Australia); NUB-7, an I-type neuroblastoma cell line was obtained from Dr. Herman Yeger (The Hospital for Sick Children, Toronto, Ontario).

Neuroblastoma cell lines were grown in alpha Minimum Essential Medium (AMEM; #

310-010-CL, Wisent Bioproducts, St. Bruno, Quebec, Canada), while sarcoma cell lines were grown in Dulbecco Minimum Essential Medium (DMEM; # 319-010-CL, Wisent

Bioproducts,St. Bruno, Quebec, Canada), both containing 10% fetal bovine serum

(FBS) and 1% antibiotic mixture in humidified atmosphere at 37C with 5% CO2.

4 Effectiveness of LDM topotecan and pazopanib in mouse models immunodeficient mice models of extracranial pediatric solid tumors: neuroblastoma, rhabdomyosarcoma and osteosarcoma.

This chapter represents the work which has been published: “Kumar S, Mokhtari RB,

Sheikh R, Wu B, Zhang L, Xu P, Man S, Oliveira ID, Yeger H, Kerbel RS, Baruchel S

(2011). Metronomic oral topotecan with pazopanib is an active antiangiogenic regimen in mouse models of aggressive pediatric solid tumor. Clin Cancer Res;17(17):5656-67”.

4.1 Methods

4.1.1 In-vitro cytotoxicity

Table 9 : The cell lines used for in-vitro experiment and their characteristics

Cell line Cancer type Characteristics

SK-N-BE(2) [385, Neuroblastoma MYCN amplified, P53

386] mutated, N-type

BE(2)-c [386] Neuroblastoma MYCN amplified, I-type

SH-SY5Y [387] Neuroblastoma Non-MYCN amplified

KHOS [388] Osteosarcoma Ras mutated

RH30 [389] Alveolar rhabdomyosarcoma PAX3/FKHRfusion gene

RD [389] Embryonal rhabdomyosarcoma

50,000 cells were seeded in 48 well plates and incubated for 48 h, after which they were treated with topotecan and/or pazopanib (in triplicate) for 72 h. Cell viability was determined by Alamar blue assay. Alamar blue (10% of total volume) was added to each well three hours prior to fluorometric detection. Fluorometric detection was performed using the SPECTRAmax gemini Spectrophotometer at excitation wavelength of 540nm and emission wavelength of 590nm. The mean fluorescence measurements of blank (wells without cells) was subtracted from the individual measurement for each well. This corrected value was used for calculating the IC50 by graph pad prism (refer 4.1.7).

4.1.2 In-vivo evaluation of topotecan and pazopanib

Xenograft models: For subcutaneous xenograft studies, we used SK-N-BE(2), SH-

SY5Y, KHOS and RH30. 1 x 106 cells were implanted subcutaneously into the inguinal fat pad of each of NOD/SCID mice. When tumors reached 0.5 cm in diameter, the animals were randomized into four groups and treated daily by oral gavage. The animals were grouped as: Control group, LDM topotecan group or „TP‟

(1.0 mg/Kg topotecan), pazopanib group or „PZ‟ (150 mg/Kg pazopanib) and combination group or „TP+PZ‟ (1.0 mg/Kg topotecan+150 mg/Kg pazopanib). In order to compare pulse topotecan to TP in KHOS osteosarcoma model, PZ was replaced by weekly oral dose of pulse topotecan or „Pulse TP‟ (15 mg/Kg topotecan).

The criteria for determining the end point were tumor sizes exceeding 2.0 cm in diameter or animals showing signs of morbidity. The tumor sizes were measured on a daily basis until the end point or sacrifice. The long (D) and short diameters (d) were measured with calipers. Tumor volume (cm3) was calculated as V = 0.5 x D x

92 d2. When the end point was reached or at the end of the treatment, the animals were sacrificed by cervical dislocation.

Metastatic mouse model: 1 x 106 BE(2)-c cells or NUB-7 cells were injected into lateral tail veins of NOD/SCID mice to generate „experimental‟ metastases as previously described [338]. Fourteen days after injection, the mice were randomized into four groups and treated in same way as the inguinal xenograft model. The treatment was continued until death or end point for BE(2)-c model and till fourteen days for NUB-7 model.

Protocol and endpoints for both xenograft and IV metastatic models were approved by Sickkids animal committee facility.

4.1.3 Immunohistochemistry and histopathology Formalin fixed tissues were paraffin embedded and sections cut at 7um. These sections were deparaffinized through xylene and ethanol, rehydrated in Phosphate- buffered Saline (PBS) (# 311-010-CL, Wisent Bioproducts, St. Bruno, Quebec,

Canada) and incubated overnight with primary antibodies for vWF at 4°C. After the primary antibody treatment, all the slides were washed three times with PBS and incubated with a broad spectrum poly-horseradish peroxidase (HRP) conjugated secondary antibody (# 87-9663, Invitrogen, Camarillo, CA) for 1h at room temperature. After washing three times with PBS, slides were stained with diaminobenzidine (DAB) and counterstained with hematoxylin. Microscopic images were captured by Olympus UTV1-X microscope mounted with Qimaging Retiga

2000R camera.

Frozen sections from the SH-SY5Y tumor model were fixed with 4% paraformaldehyde, permeabilized with 0.05% Triton X-100. After blocking with 5% bovine serum albumin in PBS for one hour, the sections were incubated overnight with rabbit polyclonal anti-CD31 antibody (#ab28364, Abcam, Cambridge, MA; dilution 1:50). The sections were washed three times with PBS containing 0.1% tween 20 (PBST) and incubated with Alexa fluor 594 donkey anti-rabbit IgG

(#A21207, Invitrogen, Camarillo, CA; dilution 1:300) for one hour. After washing with

PBST, the slides were mounted with Vectashield mounting medium (#H-1200,

Vector, with DAPI). The microscopic images of the stained sections were captured by Nikon ECLIPSE Ti series fluorescence microscope, using NIS Elements (BR

3.10) software.

Microscopic images of six fields of high vascular density were digitally captured and the pixel values for stained areas were quantified using ImageJ software. Tumor angiogenesis was quantified as the number of pixels of regions positive for vWF or

CD31.

4.1.4 Analysis of CEPs and CECs by flow cytometry

Approximately 160µl of mouse blood was collected in K2-EDTA tubes by saphenous vein puncture in SH-SY5Y, KHOS and RH30 models after twenty, twenty eight and thirty one days respectively. Blood was immediately stored at 4°C until analysis. The

CEP/CECs were measured by flow cytometry within 48hrs of blood collection as previously described [221]. CEPs were defined as CD45-, VEGFR-2+, CD117+, and

CD13+, while CECs were defined as CD45-, VEGFR-2+, CD117-, and CD13+. 7-

Aminoactinomycin D was employed to exclude the apoptotic cells. CD45+ cells

94 represented the hematopoietic cells. The absolute number of CEPs was calculated as the percentage of events collected in CEP enumeration gates multiplied by the total WBC count. For WBC counts, 10µl of blood was mixed with 90µl of Turk‟s solution. The cells were counted using hemocytometer under the microscope.

4.1.5 Bone marrow progenitor assay

The bone marrow progenitor culture was performed as previously published by our group [390]. After the sacrifice of animals belonging to RH30 model, the femur was isolated. Bone marrow was flushed out of the femur with AMEM containing 2%FBS.

Bone marrow cells of mice belonging to each group were pooled and 200,000 cells were cultured in methylcellulose media (Methocult #3434, Stem cell technologies) in

35mm culture dish, in triplicate. Bone marrow cells of non tumor bearing mice (n=3), pooled and cultured (in triplicate), concurrently with those of each group of mice, were used as reference culture plates. After fourteen days, Colony Forming Units of granulocytes and macrophages (CFU-GM) were counted under the optical microscope.

4.1.6 PK of topotecan and pazopanib

Non-tumor bearing animals were randomized into four groups (n=3): Control, PZ, TP and TP+PZ. The doses of the drugs were the same as for the inguinal xenograft and metastatic models described above.

After single drug administration, the saphenous vein blood samples (30µl) were collected in heparinized microcentrifuge tubes at 30 min, 1h, 2h, 4h, 8h, 12h, 18h and 24 h as per the Laboratory Animal Services protocol. Plasma was immediately isolated after blood collection by centrifugation. For the topotecan assay, 10µl

95 plasma was immediately precipitated with 20µl methanol and centrifuged. The supernatant and rest of the plasma were stored at -80µC until analysis.

Assay of pazopanib: 5µl of plasma was precipitated with 40µl methanol and centrifuged. 30 µl of supernatant was injected into the HPLC system, which consisted of Phenomenox C18 column (Luna; 150 x 4.6 mm; particle size 5 µ), UV detector (267 nm). The mobile phase was a 50:50 mixture of 10mM potassium phosphate and methanol, with flow rate 1.0 ml/min. The concentrations of calibration standards were 5.0, 10, 50,100 and 200µg/ml.

Assay of topotecan: Prior to analysis, the previously prepared 20µl methanolic extract was mixed with 50µl of internal standard solution (5ng/ml d6 topotecan dissolved 0.1% formic acid in acetonitrile). The mixture was then centrifuged and the supernatant was transferred to auto- sampler vials.

The LC/MS system consisted of an HPLC (Agilent Infinity 1290), Column (Kinetex

HILIC, 2.6u, 100A, 50x4.6mm) and a mass spectrometer (Sciex 5500-QTrap). The analytes were eluted by gradient flow. Mobile phase A was water: acetonitrile (10:90) and mobile phase B was 10mM ammonium acetate (pH 3.2). The mobile phase ratio was 5% A for 0-2 min, 20% A for 4-6min and 5% A for 8-10min at a flow rate 0.5 ml/min. The samples were analyzed by positive ion electrospray ionization technique in multiple reaction monitoring modes. The following mass transitions were monitored: 422.2 to 377.0 m/z, (topotecan M+H) and 428.2 to 377.0 m/z (topotecan d6 M+H). The concentrations of calibration standards were 0.5, 1.0, 5.0, 10 and 100 ng/ml.

4.1.7 Statistical analysis

In vitro dose-response (IC50), in vivo tumor growth curves and the number of pixels for immunohistochemistry are presented as mean ± SD. Statistical significance was assessed by student‟s t-test. Prism 5 (Version 5.04) for Windows, GraphPad

Software, San Diego California USA, was used for the calculation of IC50 and P values.

4.2 Results

4.2.1 Drug-induced in vitro cytotoxities

Both topotecan and pazopanib caused a dose-dependent reduction in viability of

HUVEC with IC50 of 4.87 ng/ml and 398.0 ng/ml respectively [Figure 5]. Pazopanib did not affect the viabilities of the any tumor cell line at any of the concentrations tested

[Figure 6]. Topotecan demonstrated a dose-dependent reduction in the viability of all the tumor cell lines. Among neuroblastoma cell lines, SH-SY5Y cells (IC50 = 5.3 ng/ml) was more sensitive to topotecan than BE(2)-c (IC50= 45.6 ng/ml) and SK-N-BE(2) cells

(IC50= 65.0 ng/ml) [Figure 7].

Among sarcoma cell lines, the IC50 of topotecan on RH30, RD and KHOS cell lines were 7.4 ng/ml, 7.5 ng/ml and 4.9 ng/ml respectively [Figure 8]. Among all tumor cell lines tested, addition of 5000 ng/ml pazopanib only caused a significant reduction of

IC50 in SK-N-BE(2) cells (IC50=35.1ng/ml, P=0.046) [Figure 7].

Figure 4: In-vitro dose- response of topotecan and pazopanib on HUVEC

1 2 0 1 2 0

]

C C

E 1 0 0

1 0 0 E

V U

8 0 U 8 0

[ [

y 6 0

6 0 y

l i

i 4 0 b

4 0 b

i v

v 2 0

2 0

% 0 % 0

1 1 0 0 0 0 0 1 .1 1 0 0 0 . 1 0 0 0 0 .0 0 1 0 0 0 1 0 0 0 0 1 0 1 0 0 1 1 0 1 [T o p o te c a n ], n g /m l [P a z o p a n ib ], n g /m l

Figure 5: In-vitro dose-response of pazopanib on tumor cell lines

140 120

y 100

i l

i 80

b a

i 60 v SK-N-BE2

40 SH-SY5Y % 20 KHOS RH30 0 0 4000 8000 12000 16000 20000 [Pazopanib], ng/ml

Figure 6: In-vitro dose-response study of topotecan alone and in combination with 5000 ng/ml pazopanib on neuroblastoma cell lines

] 1 2 0

T o p o te c a n Y

5 1 0 0

Y C o m b in a tio n S

- 8 0

H S

[ 6 0

i l

i 4 0

b a

i 2 0

0 % 0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0

[T o p o te c a n ], n g /m l

] )

2 1 2 0

( T o p o te c a n E

B 1 0 0

- C o m b in a tio n N

- 8 0

K S

[ 6 0

t i

l 4 0

b a

i 2 0

0 % 0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0 [T o p o te c a n ], n g /m l

1 2 0 ]

c T o p o te c a n - ) 1 0 0

2 C o m b in a tio n (

E 8 0

B [

y 6 0

l i

b 4 0

a i

v 2 0

% 0 0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0 [T o p o te c a n ], n g /m l

100

Figure 7: In-vitro dose-response study of topotecan alone and in combination with 5000 ng/ml pazopanib on sarcoma cell lines

1 2 0

] 1 0 0 T o p o te c a n 0

3 C o m b in a tio n

H 8 0

[

y 6 0

i b

a 4 0

i v

% 2 0

0 0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0 [T o p o te c a n ], n g /m l

1 2 0

T o p o te c a n

1 0 0 ]

D C o m b in a tio n

R 8 0

[

i l

i 6 0

b a

i 4 0

% 2 0

0 0 .0 1 0 .1 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0 [T o p o te c a n ], n g /m l

1 0 0

] T o p o te c a n S

O 8 0 C o m b in a tio n

K [

6 0

l i

b 4 0

2 0 %

0 1 1 0 1 0 0 1 0 0 0 1 0 0 0 0 [T o p o te c a n ], n g /m l

101

4.2.2 LDM topotecan and pazopanib in neuroblastoma mouse models

The criteria for effectiveness of the treatments in SK-N-BE(2) xenograft model were tumors regression and enhancement of survival times. Drugs were administered daily over a period of fifty six days. In addition, the animals belonging to the TP+PZ group, which remained alive, were also retreated from the 103rd day to evaluate the impact of resuming treatment on reversing drug resistance [Figure 9]. Retreatment was continued until the 125th day, after which the mice were sacrificed. In this model, no significant difference was observed between the tumor growth rates of the TP and PZ treated groups. Compared to control, both the single agents significantly enhanced survival of animals (P<0.05). The survival in TP+PZ group was significantly higher compared to both control (P<0.005) and the single agents‟ groups (P<0.005). Retreatment was associated by transient tumor growth delay up to 120 days. All the animals were sacrificed by the 125th day.

In SH-SY5Y xenograft model, animals belonging to all the four groups were sacrificed after twenty days treatment, when the tumor end point was reached. Tumor growth delay and the difference in tumor weights at the end of the treatment were the criteria for assessment of treatment effectiveness. The treatments caused tumor growth delay in the order PZ

BE(2)-c and NUB-7 are N-Myc amplified, I-type malignant neuroblastoma cells which have high potential to migrate and metastasize [391, 392]. In the NUB-7 metastatic

102 model, the animals belonging to all the four groups were sacrificed after fourteen days treatment [Figure 11]. Compared to the control, TP and TP+PZ liver weights were significantly lower in TP+PZ treated animals, compared to PZ [Figure 11B]. Microscopic tumors were visible in the livers of mice belonging to all the groups except TP+PZ confirming the ability of TP+PZ to control liver metastasis [Figure 11C].

Survival time was used as the parameter to assess the efficacy of treatments in our

BE(2)-c metastatic model. All the treatment groups demonstrated a statistically significant enhanced survival [Figure 12A]. Survival of TP treated animals was higher than PZ treated animals (P<0.05). The mean survival span of animals in TP+PZ group was approximately two fold (100.8 days) compared to the TP group (52.4 days),

P<0.005. At the time of death or end point, the animals belonging to control, PZ and TP groups had macroscopically detectable tumors in liver. Animals belonging to TP+PZ group did not reveal any evidence of liver metastasis [Figure 12B]. Animals in all the four groups of BE(2)-c model had evidence of tumors present in kidney, adrenal gland and bone marrow.

In all the graphs belonging to chapter-4, P values between the mean observations of two treatment groups have only been mentioned in cases where statistically significant

(P<0.05) differences were observed.

103

Figure 8: In-vivo efficacy of metronomic topotecan and pazopanib in SK-N- BE(2) subcutaneous xenograft model (n=4). † in the survival curve indicates that animal in combination group died due to gavaging error on 9th day of treatment

) 6

c (

5 e

m 4

l o

v 3

o 2

m u

T 1

0 0 2 5 5 0 7 5 1 0 0 1 2 5 D a y s

1 0 0 n = 4

l 8 0

a v i †

v 6 0

u s

4 0 %

2 0

0 0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 D a y s

C o n tro l PZ TP T P + P Z

104

Figure 9: In-vivo efficacy TP and PZ in SH-SY5Y subcutaneous xenograft model. (A) The photographs of tumors isolated after 20 days treatment. (B) Histogram depicting the comparison of tumor weights after treatment. (C) The tumor growth curve during the treatment.

A C o n tro l

P Z

T P

T P + P Z

B P < 0 .0 0 0 1

P < 0 .0 0 0 1 4 P = 0 .0 4 n=5

) P < 0 .0 0 0 1

g (

t P < 0 .0 0 0 1

g i

e 2

o P = 0 .0 0 8

m 1

u T

0 C o n tro l P Z T P T P + P Z

) 6 c C o n tro l

c n = 5

(

e T P

m 4

u l

o P Z

o 2 T P + P Z

u T

0 0 4 8 1 2 1 6 2 0 D a y s

105

Figure 10: In-vivo efficacy of TP and PZ in NUB-7 IV metastatic neuroblastoma model. (A) Photographs of liver isolated from mice after completion of the treatment. (B) Comparison of liver weights after treatment. (C) The micrographs of hematoxyline and eosine stained sections of liver (magnification, x10)

A C o n tro l

P Z

T P

T P + P Z

B P = 0 .0 0 2

4 P = 0 .0 1

) n=5

g (

3 r

e P = 0 .0 4

f 2

g i

e 1 W

0 C o n tro l P Z T P T P + P Z

C o n tro l P Z T P T P + P Z

106

Figure 11: In-vivo efficacy of TP and PZ in BE(2)-c IV metastatic neuroblastoma model. (A) The Kaplan-Meier survival curve of animals. (B) Micrographs of hematoxylene and eosine stained organs isolated after sacrifice of animals

A n=5 100 Control

l 80 PZ

v i

v 60 TP

r u

s TP+PZ

40 % 20

0 0 20 40 60 80 100 120 140 Days

B Control PZ TP TP+PZ

B M

107

4.2.3 Effect of LDM topotecan and pazopanib on the tumor growth in sarcoma models

Since PZ had demonstrated limited efficacy in neuroblastoma models, we decided to evaluate the anti-tumor activity of Pulse TP and compare it with TP in KHOS osteosarcoma model, in which the mice were sacrificed after twenty eight days treatment. Here, both pulseTP and TP delayed the tumor growth, with significantly lower tumor weight at the end of the treatment [Figure 13A and B]. The tumor growth rate curve [Figure 13C] reveals that the single agents caused tumor growth delay, but not tumor size reduction, while TP+PZ, induced tumor growth delay until twenty two days, after which tumor size reduction was observed. The TP+PZ group had significantly lower tumor weights compared to the control, PulseTP and TP [Figure

13B].

In rhabdomyosarcoma RH30 xenograft model, the animals were treated for fifty six days. The animals belonging to control and TP reached the end point before this period, while those in PZ and TP+PZ treated groups remained alive after the discontinuation of treatment [Figure 14]. TP was ineffective in controlling the tumor growth. In view of activity of PZ in soft tissue sarcoma we decided to test PZ. PZ as a single agent as well as the combination TP+PZ delayed the tumor growth and enhanced the survival by two fold, compared to both control and TP. TP+PZ group had significantly lower tumor size (P=0.03), compared to those of PZ group.

108

Figure 12: Efficacy of TP and PZ in osteosarcoma KHOS subcutaneous xenograft model. (A) Photographs of tumors collected after 28 days treatment. (B) Histogram depicting comparison of tumor weights after treatment. (C) The tumor growth rate during treatment.

A C o n tro l

P u ls e T P

T P

T P + P Z

B * * * *

* * * n = 5 * *

)

g (

t h

g * * i

e 2

r * * o

m 1

u T

0 C o n tro l P u ls e T P T P T P + P Z

7 C o n tro l )

m 6 P u lse T P

c .

u TP

c 5

(

e T P + P Z

m u

l 3

r 2 o

m 1

u T 0 0 1 0 2 0 3 0 D a y s

109

Figure 13: Efficacy of TP and PZ in rhabdomyosarcoma RH30 subcutaneous xenograft model.

6 n = 5

C o n tro l

)

c c

( PZ

e 4

m TP

o v

T P + P Z r

o 2

u T

0 0 2 0 4 0 6 0 8 0 D a y s

110

4.2.4 Effect of treatment on tumor microvessel densities

Comparison of the pixel counts of six fields of highly vascularized regions of tumor sections stained for CD31 and vWF revealed that TP+PZ significantly reduced the microvessel density of the tumors, compared to the control in SH-SY5Y, RH30 and

KHOS models. [Figure 15, 16]. In SH-SY5Y model, PZ but not TP caused significant reduction in microvessel densities compared to the control [Figure 15]. In RH30 model, none of the single agents caused reduction in microvessel densities, compared to the control [Figure 16A], while in KHOS model, both pulse TP and TP caused reduction in microvessel densities [Figure 16B].

111

Figure 14: Effect of treatment with TP and PZ on tumor microvessel density in neuroblastoma mice model.

Microscopic images of highly vascularized areas, stained for CD31 in tumor sections from SH-SY5Y xenograft (original magnification x 10). The red color represents regions stained for CD31. The histogram represents comparison of pixels of areas positive for CD31 as measured using imageJ software.

112

Figure 15: Effect of treatment with TP and PZ on tumor microvessel density of tumors in sarcoma models.

(A) and (B) are the sections from RH30 and KHOS xenografts respectively, stained for vWF (original magnification x10). Arrows point toward regions stained for vWF. The histogram represents comparison of pixels of areas positive for CD31 as measured using imageJ software.

113

4.2.5 Effect of the treatments on CECs and CEPs In our SH-SY5Y neuroblastoma model, after twenty days treatment, TP+PZ significantly reduced both viable CEC and CEP levels compared to the control and single agents groups [Figure 17A, 18A]. Though the single agents caused reduction in both CEC and / or CEP levels, compared to the control, the differences were not statistically significant.

In RH30 rhabdomyosarcoma model, after thirty one days treatment, TP+PZ caused significant reductions in both viable CECs and CEPs levels compared to the control and

TP [Figure 17B, 18C]. Compared to PZ, TP+PZ caused significant viable CEP reduction. PZ treated group had significantly lower viable CEP levels compared to the control.

In KHOS osetosarcoma model, CEP and CEC levels were measured after twenty eight days treatment [Figure 17C, 18B]. TP+PZ caused significant reduction in viable CEC and CEP levels compared to the control and pulse TP. Also, TP caused a significant reduction in viable CEP levels compared to pulse TP.

114

Figure 16: The effect of the treatment regimens on CEP levels in blood

The histograms indicate the comparison of CEP levels in (A) SH-SY5Y xenograft model after twenty days treatment; (B) RH30 model after thirty one days treatment; (C)

KHOS xenograft model after twenty eight days treatment.

A . S H -S Y 5 Y B . R H 3 0

P = 0 .0 2 P = 0 .0 0 0 8

P = 0 .0 2 d 0 .1 5 d

o P = 0 .0 2 P = 0 .0 1 o

0 .0 6 o

b l

P = 0 .0 4 l P = 0 .0 0 3

µ 0 .1 0

0 .0 4

0 .0 2 0 .0 5

i V 0 .0 0 V 0 .0 0 C o n tro l P Z T P T P + P Z C o n tro l P Z T P T P + P Z

C . K H O S

P = 0 .0 0 0 7

P = 0 .0 0 1

d o

o P = 0 .0 3 l P = 0 .0 0 8

b 1 .5

l µ

P = 0 .0 2

0 1

1 .0

P E

C 0 .5

b a i 0 .0 V C o n tro l P u ls e T P T P T P + P Z

115

Figure 17: The effect of the treatment regimens on CEC levels in blood.

The histograms indicate the comparison of CEC levels in (A) SH-SY5Y xenograft model after twenty days treatment; (B) RH30 model after thirty one days treatment; (C)

KHOS xenograft model after twenty eight days treatment.

A . S H -S Y 5 Y B . K H O S

P = 0 .0 0 3 P = 0 .0 2

0 .5 d 0 .5

d o

o P = 0 .0 0 6

o l

0 .4 b 0 .4 b

P = 0 .0 0 2

/ /

0 .3 0 .3 C

P = 0 .0 0 9 C E 0 .2 E

C 0 .2

e l

b 0 .1 0 .1

i V 0 .0 V 0 .0 C o n tro l P Z T P T P + P Z C o n tro l P Z T P T P + P Z

C . R H 3 0

P = 0 .0 0 1

d P = 0 .0 0 3

o 4 o

l P = 0 .0 1

µ 3

0 1

P = 0 .0 1 /

E C

b a i 0 V C o n tro l P u ls e T P T P T P + P Z

116

4.2.6 Safety of TP and PZ in mice In SH-SY5Y and KHOS model model, all the three treatment regimens significantly reduced WBC [Figure 19 A and B]. TP+PZ treated group had significantly lower WBC levels compared to both the single agents [Figure 19 A and B]. Surprisingly, TP had significantly lower WBC count compared to pulse TP. In RH30 model, PZ and TP+PZ reduced WBC levels significantly, while TP did not [Figure 19C]. Despite the significant lowering of WBC induced by the combination, compared to the control and the single agents, the animals belonging to this group in all the models were active and showed no signs of illness during or after this period until tumor sizes reaching the end point.

CFU-GM were counted in RH30 model, where the mice were sacrificed at different times i.e. day 32, day 35, day 71 and day 73 for control, TP, PZ and TP+PZ respectively. Percentage CFU-GM count for each plate was calculated as the percentage of CFU-GM number in that plate to the average CFU-GM number in reference plates [Figure 19C]. TP treated group had significantly lower CFU-GM counts compared to the control. TP+PZ treated group had signficantly lower CFU-GM number compared to the control, but not compared to the single agent groups.

117

Figure 18: Toxic effect of the treatment regimens bone marrow

(A) and (B) are the histograms indicating the comparison of WBC levels in SH-SY5Y xenograft model after twenty days treatment; (B) KHOS model after thirty one days treatment; respectively. (C) Histogram indicating the comparison of WBC and % CFU-

GM, as measured by bone marrow progenitor assay in the RH30 model after thirty one days of treatment.

A . S H -S Y 5 Y B . K H O S

P < 0 .0 0 0 1 P = 0 .0 0 7

P = 0 .0 0 0 5 P = 0 .0 2 3 0 0 P = 0 .0 3 d P = 0 .0 0 4 3 0 o

P < 0 .0 0 0 1

b o

P = 0 .0 2

l P = 0 .0 0 4

l b

µ 2 0 0

2 0 P < 0 .0 0 0 1 0

µ P = 0 .0 1

/ 1 0 0

1 0

0 0 C o n tro l P Z T P T P + P Z C o n tro l P u ls e T P T P T P + P Z

C . R H 3 0

P = 0 .0 0 4 P = 0 .0 0 0 7

1 5 0 2 0 P = 0 .0 4 6

P = 0 .0 0 2

o o

l P = 0 .0 0 1

1 5 M

b P = 0 .0 0 6 1 0 0

0 F

0 1 0

/ 5 0

C 5

B W 0 0 C o n tro l P Z T P T P + P Z C o n tro l P Z T P T P + P Z

118

4.2.7 PK did not reveal drug interaction between topotecan and pazopanib in TP+PZ group In the present study, the PK of topotecan and pazopanib was conducted to detect any pharmacokinetic interaction between topotecan and pazopanib in the TP+PZ group.

Peak plasma concentrations of topotecan in TP and TP+PZ groups were 19.75ng/ml and 33.05ng/ml, respectively, while the trough concentration was 0.77ng/ml and 2.79 ng/ml [Figure 20 A,B]. The peak plasma concentration of pazopanib was reached in 2h in both PZ and TP+PZ groups [Figure 21 A,B]. The Cmax of pazopanib was 133.5 ng/ml and 122.4 ng/ml in PZ and TP+PZ groups respectively, while the trough concentration was 9.46 ng/ml and 14.56 ng/ml respectively. For both drugs, no significant difference was observed between plasma concentrations of single agent and combination treated animals at any time point. A significant inter-animal drug concentration variability was detected and larger group studies may be necessary to detect drug-drug interactions and changes in trough concentration. The previously reported optimal plasma concentration of pazopanib effectiveness (40 µM or ≈ 18

µg/ml) [393] was maintained until at least 18 h in both PZ and TP+PZ groups.

119

Figure 19: Plasma concentration-time profiles of topotecan

(A) single agent TP treated group and (B) in TP+PZ treated group.

A . T P tre a te d g r o u p

A U C = 9 3 .5 5 n g /m l.h r 40 0 - 2 4

10 [Topotecan], ng/ml [Topotecan], 0 0 5 10 15 20 25 Time (hours)

B . T P + P Z tr e a te d g r o u p 60

30 A U C 0 - 2 4 = 1 2 2 .0 n g /m l.h r 20

10 [Topotecan], ng/ml [Topotecan], 0 0 5 10 15 20 25 Time (hours)

120

Figure 20: Plasma concentration-time profiles of pazopanib

(A) single agent PZ treated group and (B) in TP+PZ treated group.

A . P Z tre a te d g ro u p 150 A U C = 1 8 0 2 µ g /m l.h r 125 0 -2 4

100

25 [Pazopanib], µg/ml[Pazopanib], 0 0 5 10 15 20 25 Time (hours)

B . T P + P Z tre a te d g ro u p 150

125 A U C = 1 7 3 3 µ g /m l.h r 100 0 -2 4

25 [Pazopanib], µg/ml[Pazopanib], 0 0 5 10 15 20 25 Time (hours)

121

4.3 Discussion

Angiogenesis plays important roles in cancer growth, metastasis, and response to therapy. In pediatric tumors such as neuroblastoma, osteosarcoma and rhabdomyosarcoma, in situ tumor angiogenesis and the levels of circulating angiogenic factors correlate with metastatic disease and poor prognosis [67, 78, 88].

Considering the previously reports regarding superior efficacy of the combination of

LDM chemotherapy and anti-VEGF therapy, we evaluated the effectiveness of LDM regimen of oral topotecan and its combination with one of the clinically approved

RTKI, pazopanib, in the murine models of three pediatric solid tumors, with particular emphasis on the antiangiogenic mechanism and their potential bone marrow toxicity.

The doses of drugs were selected on the basis of previous studies. The daily oral doses of 1.0 mg/Kg topotecan and 150 mg/Kg pazopanib have been previously found to be effective in ovarian cancer mouse models [170]. Shaked et al has previously defined the Optimal Biologic Dose (OBD) of LDM chemotherapy as the dose causing maximum reduction in CEPs with minimal or no toxicity after daily treatment for one week; this dose is associated with maximum antiangiogenic efficacy [330]. In a previous dose-response study, the daily dose of oral metronomic topotecan (0.5, 1.0 and 1.5 mg/Kg) caused greater reduction in microvascular density compared with weekly MTD regimen (7.5 and 15mg/Kg) in an ovarian cancer model, but the mice treated with 1.5 mg/Kg daily, oral topotecan showed decreased food intake, indicative of adverse effect [278]. By applying the aforementioned definition of OBD, we postulated that 1.0 mg/Kg oral topotecan administered daily, would be the OBD, or within the range of the OBD. The antiangiogenic efficacy of

122 weekly pulse topotecan and daily LDM topotecan has also been compared in our osteosarcoma model.

The effectiveness of our treatment(s) was tested in four neuroblastoma mice models taking into consideration the genetic and phenotypic heterogeneity of neuroblastoma.

There are multiple oncogenic pathways that drive neuroblastoma. The major genetic driving abnormalities in neuroblastoma are MYCN amplification (20%), allelic loss of chromosome material from 1p36 and 11q23 and gain of genetic material on chromosome 17q [394].

Within the same type of neuroblastoma tumor, cells can be classified as N-type

(neuroblastic) which show adrenergic neuronal phenotype, S-type (substrate adherant) which demostrate schwanian features and I-type (intermediate) [395].

These phenotypes are interconvertible. N-type cells are small, rounded, loosely adherant cells with numerous neurite-like processes. S-type cells are flat, large cells with more epithelial-like phenotype and are highly substrate adherant, which resemble melanocytic, schwanian or meningeal phenotype. I-type cells are comparatively undifferentiated and can differentiate into N-type or S-type lineages.

In neuroblastoma,N-Myc over-expression is associated with worse prognosis [396].

In-vivo, N-Myc over-expression is associated with higher angiogenic activity in neuroblastoma [397]. N-Myc stimulates PI3K/Akt pathway, which in turn upregulates

VEGF expression and also downregulates antiangiogenic factors [398].

N-Myc also downregulates endogenous angiogenesis inhibitor IL-6 and therefore antiangiogenic therapy has been proposed to be effective against MYCN amplified

123 neuroblastoma [399]. P53 is another determinant of response to antiangiogenic therapy. P53 mediates the apoptotic effects of hypoxia in tumor cells. However, in the absence of functioning P53, the apoptotic effects of antiangiogenesis-induced hypoxia is lost [400].

Our first neuroblastoma model was developed using the SK-N-BE(2) cell line. SK-N-

BE(2) is a MYCN amplified cell line and predominantly comprises of N-type cells

[386]. It was obtained from bone marrow metastasis of a relapsed, stage-IV, cisplatin resistant neuroblastoma patient [76]. It is noteworthy that a major distinguishing feature of SK-N-BE(2) from SK-N-BE(1) (collected from the bone marrow aspirate of same patient during diagnosis, before therapy) is P53 mutation, which might have rendered it resistant to vincristine, doxorubicin, cyclophosphamide and radiotherapy

[385]. Expression of HIF-1 alpha and VEGF has been reported in hypoxic SK-N-

BE(2) cells. Previously, in our laboratory, it has been proved that VEGF/Flt-1 autocrine loop is necessary for HIF-1 alpha mediated activation ERK-1/2 pathway which leads to survival, drug resistance (cisplatin, melphalan and etoposide) and in- vivo angiogenesis [76]. Therefore, we considered this robust cell line, which is resistant to conventional chemotherapeutics and hypoxia-mediated apoptosis, to be an appropriate representative of aggressive neuroblastoma.

SH-SY5Y, a non-MYCN amplified cell line, is more sensitive to therapy than SK-N-

BE(2) [387]. High VEGF concentration has been detected in cultured SH-SY5Y cells

[71]. It has previously reported to be sensitive to therapies by anti-VEGF antibody as well as combination of bevacizumab and nutlin-3a (MDM2 inhibitor) [401, 402].

124

We developed IV metastatic models to simulate the minimal residual disease in neuroblastoma. Currently, 13-cis retinoic acid is an example of a non-cytotoxic drug which is active against drug resistant cell lines without severe toxicity [403]. Likewise,,

LDM chemotherapy and targeted antiangiogenic agents which are proven to be effective against drug resistant tumor cells, are promising candidates for targeting minimal residual disease in neuroblastoma [99].

For rhabomyosarcoma we used RH30, which is an alveolar rhabdomyosarcoma cell line possessing PAX3/FKHR translocation [389]. Alveolar rhabdomyosarcoma is the most aggressive type of pediatric rhabdomyosarcoma [404]. Translocations between

PAX3 or PAX7 and FKHR (also known as FOXO1) are major the oncogenic drivers in alveolar rhabdomyosarcoma. PAX3/FKHR chimeric oncogenic product is a more potent transcriptional activator than PAX3 gene product, which leads to malignant transformation of cells [405]. PAX3/FKHR translocation occur in ≈59% of alveolar rhabdomyosarcoma and is associated with increased risk of therapy failure and death compared to PAX7/FKHR translocation [404]. PAX3/FKHR, along with IGF-II, upregulate the expression of VEGF and PDGF-B [406].

The downstream targets of PAX3/FKHR are proto-oncogenes like FGFR4, MET and

MYCN, all of which are over-expressed in RH30 cell line [407]. RH30 was used to study the effect of FGFR4 inhibitor, where it inhibited the cell proliferation Also, high levels of

MYCN is associated with PAX3/FKHR translocation in rhabdomyosarcoma. Therefore, we considered RH30 to be a suitable representative of PAX3/FKHR translocated alveolar rhabdomyosarcoma.

125

In this study, tumor microvessel density and CECs and CEPs were employed as antiangiogenic markers. Tumor microvessel density in vascular hotspots is a widely employed parameter of assessing tumor angiogenesis, staging patients and evaluating antiangiogenic response [408]. Microvessel density has been correlated with higher proliferative index in neuroblastoma [409]. In another study, tumor angiogenesis measured by vascular index correlated with MYCN amplification and poor outcome in neuroblastoma [67]. In osteosarcoma, tumor microvessel density has been correlated with higher angiogenic expression and worse prognosis [88]. CEPs have demonstrated usefulness as antiangiogenic markers in variety of preclinical and clinical studies. In mice, the recruitment of fluorescent bone marrow derived endothelial cells were observed in orthotopic neuroblastoma xenograft, where it constituted 13-14% of total

CD31+ blood vessels [410]. VEGF stimulates the recruitment of CEPs into the tumor neovasculature and thus contributes to the endothelial lining [59]. Therefore blockade of the VEGF-signalling pathway is expected to reduce the CEP level in blood and thus inhibit angiogenesis. Prevention of CEP mobilization is also a mechanism of action of

LDM chemotherapy [155]. Therefore, the combination of both agents is expected to have additive or synergistic effect on the prevention of CEP mobilization. Our data confirm these findings in all pediatric tumor models with various degrees of responses.

WBC levels in peripheral blood and CFU-GM were employed as markers of safety to

LDM topotecan and pazopanib. Both topotecan and pazopanib are known to be myelosuppressive [257, 411]. Though, acute neutropenia can be avoided with lower dose of topotecan, the possibility of delayed myelosuppression upon its prolonged use must be considered. Also, even though the toxicities of LDM chemotherapy and antiangiogenic agents are manageable, their additive toxic effects on bone marrow can

126 lead to serious chronic side effects. Here, we employed the White Blood Cell (WBC) count and colony forming potential of granulocyte and macrophage progenitors as parameters to assess the bone marrow toxicities of our treatments. Bone marrow colony forming unit assay has been used to assess the toxicity of camptothecins and to compare the toxicities of camptothecins between mice and humans [412, 413].

PK drug interactions between co-administered drugs eventually affects the plasma, tissue or intercellular concentration of drugs. Anticancer drugs have shown to interact with other drugs. PK interaction of taxanes, vinca alkalloids and irinotecan with psychotropic drugs are common occurances [414]. As far as interactions between anticancer drugs are concerned, topotecan is reported to lower the bioavailability of docetaxel in solid tumor patients due to the involvement of the same enzyme CYP3A4

[260].

A number of anticancer drugs are the substrates of ABC transporters. In gene knockout mice models, ABCG2 (BCRP) was found to reduce the biliary and fecal excretion and to increase the renal excretion of topotecan [415]. Polymorphism in ABCG2 significantly affects the pharmacokinetics of 9-aminocamptothecins. In a study, solid tumor patients heterozygous for ABCG2 achieved higher AUC than those with wild type ABCG2 [416].

Tyrosine kinase inhibitors have also demonstrated drug-drug interactions. Gefitinib, erlotinib, imatinib and nilotinib are the substrates of multidrug resistant transporters such as P-glycoprotein and ABCG2 [417]. Gefitinib is reported to enhance the oral absorption and hence the bioavailability of irinotecan in mice. In another study, sunitinib has shown the ability to inhibit P-glycoprotein and ABCG2 [417]. Lapatinib is an inhibitor of BCRP due to which it increased the intracellular concentrations of

127 substrates like SN38 and miloxantrone [418]. Pazopanib is a substrate of PgP and

BCRP, which limits the pazopanib distribution in brain. Dual inhibition of PgP and

BCRP by elacridar resulted in a 5-fold increase in pazopanib brain penetration. Co- administration with PgP and BCRP inhibiting RTKIs canertinib and erlotinib caused 2-

2.5 flod increase in pazopanib brain accumulation [419]. Likewise, everolimus is also reported to enhance the brain penetration of vandetanib in mice [420].

Pazopanib is metabolized by CYP3A4, therefore, its co-administration with a strong

CYP3A4 inhibitor ketoconazole has increased its bioavailability [421]. Proton pump inhibitor esomeprazole, which raises the gastric pH reduces the bioavailability of pazopanib [421]. Since, pazopanib is a highly lipophilic weak base, its solubility is affected by higher pH, which reduces the soluble fraction of drug capable of being absorbed by gastrontestinal tract.

In this study, the fact that both topotecan and pazopanib are the substrates of CYP3A4 and MDR transporters raises the possibility of competitive metabolic or efflux inhibition of one drug by another. Also, pazopanib is reported to be a weak inhibitor of CYP3A4 and CYP2D6 [422]. Another objective behind measuring the plasma concentration profiles of the drugs was to acquire information regarding the plasma concentration required for the superior activity of LDM topotecan and/or pazopanib observed in this study. Interspecies dose-translation of these concentrations can be used to achieve similar effect in human subjects.

In-vitro, pazopanib neither had any effect on the viability of any of the cell lines, nor did it enhance the cytotoxicity of topotecan on any of the cell lines except SK-N-BE(2) but was active on HUVEC cell lines. In agreement with our hypothesis, in-vivo, LDM

128 topotecan and its combination with pazopanib delayed the tumor growth and significantly enhanced the animal survival in all the models, TP+PZ showing higher anti-tumor efficacy compared to TP and PZ or Pulse TP. TP was more effective than

PZ in neuroblastoma models, while in RH30 model, PZ was more effective in delaying tumor growth than TP.

In neuroblastoma, TP+PZ delayed tumor growth in SK-N-BE(2) and SH-SY5Y models, and reduced micrometastasis in BE(2)-c and NUB-7 models. The superiority of the combination over the single agents could be partially explained by its antiangiogenic activity, as observed by the significant reduction of all the three markers: viable CECs, viable CEPs and tumor microvessel density, by TP+PZ, compared to the both TP and PZ in SH-SY5Y neuroblastoma models. However, among the single agents, only PZ demonstrates antiangiogenic activity, as observed by the significant reduction in the microvessel density. The delay of tumor growth at metastatic sites by TP+PZ in NUB-7 and BE(2)-c metastatic models indicates that the combination of LDM topotecan and pazopanib can potentially control minimal residual disease and enhance the survival in high risk neuroblastoma. Furthermore, even though animals in BE(2)c metastatic model eventually succumbed to the disease after prolonged treatment with TP+PZ, there was no tumor burden in liver of these animals. Animals died due to tumor burden in other organs such as kidney and bone marrow, whereas in control as well as in single agent groups, liver was the most affected organ. This is an indication that the organ or tissue environment can have a significant impact on the response to antiangiogenic therapy.

129

In KHOS osteosarcoma model, all the regimens tested caused significant reduction in the levels of viable CECs and CEPs and microvessel densities after twenty eight days treatment. Though there was no significant difference between the tumor weights of Pulse TP and TP upon sacrifice, the viable CEP levels in TP treated group were significantly lower than those in Pulse TP treated group, indicating that metronomic topotecan is more antiangiogenic than the pulse dosing of topotecan.

In RH30 rhabdomyosarcoma model, TP+PZ caused significant reduction in viable

CEC and CEP levels and microvessel density compared to both control and TP. In addition, significant reduction in viable CEP level was demonstrated with PZ alone after thirty one days treatment, thus correlating with its tumor response. After exposure to single agent PZ, the microvessel densities of tumor xenografts, isolated at the time of tumor progression two weeks after discontinuation of treatment, were not different from those of control group. TP+PZ had significantly low viable CEPs than PZ. By analyzing the observations from tumor growth rate, circulating biomarker levels and microvessel density experiment, we are postulating that in the rhabdomyosarcoma model, PZ and TP+PZ are more effective than TP and that the antiangiogenic effectiveness of TP+PZ is more sustained than PZ after the discontinuation of the treatment.

Our PK study did not reveal any significant differences in the plasma concentrations of TP or PZ between single agent and the combination groups, at any of the time points examined. However, a significant inter-animal variability was detected at the trough level of TP in the TP+PZ group, though it did not reach statistical significance; it was higher in the TP+PZ group than in the TP group. For

130 pazopanib, 40µM (≈ 18µg/ml) has been reported to be the optimum plasma concentration for the inhibition of VEGFR2 phosphorylation in mice [393]. Since the plasma concentration of pazopanib was above this limit until 18h, it can be concluded 150 mg/Kg pazopanib can inhibit VEGFR2 phosphorylation for at least

18h after oral drug administration. Also, the possibility of involvement of pharmacokinetic drug interaction on the superior efficacy of TP+PZ can be excluded.

One challenge which can be anticipated from this study is the additive toxicity of

TP+PZ compared to the single agents. In rhabdomyosarcoma model, the CFU-GM counts were significantly lower in TP+PZ group compared to the untreated control.

Since, in this group the bone marrows were collected after discontinuation of treatment, the significantly lower colony forming potential of bone marrow progenitors indicate delayed myelotoxicity due to prolonged therapy with the combination of LDM topotecan and pazopanib.

131

5 To conduct a time–response study to investigate the changes in tumor xenograft behavior in response to prolonged therapy with LDM topotecan and pazopanib in a neuroblastoma mice model.

This chapter represents the work which has been published: “Kumar S, Mokhtari RB,

Oliveira ID, Islam S, Toledo SRC, Yeger H, Baruchel S (2013). Tumor Dynamics in

Response to Antiangiogenic Therapy with Oral Metronomic Topotecan and Pazopanib in Neuroblastoma Xenografts. Transl Oncol. 6(4):493-503”.

In the previous chapter, in BE(2)-c metastatic model, the animals reached the end point even after continued therapy with TP+PZ . This finding reflects a commonly observed drawback of antiangiogenic therapy. Compared to the large number of clinical trials conducted, the successes with antiangiogenic therapy is rather modest, despite the initial enthusiasm and promising preclinical results [423].

The earlier theory that antiangiogenic therapy will eradicate the tumor by starving tumor cells was a simplistic assumption. The fact is that, though it does not directly target the tumor cells, antiangiogenic therapy can lead to phenotypic changes in tumor cells by altering the tumor microenvironment due to its effect on tumor vasculature [423].

Antiangiogenic therapy affects the tumor-stroma interaction by disrupting the tumor vasculature, not only by reduced delivery of oxygen and nutrients, but also by interfering with the survival signals between endothelial cells, stromal cells and tumor cells [424]. Initially, antiangiogenic agents induce a state of transient dormancy, which is followed by either elimination of the tumor or by development of resistant tumors. In most of the cases, the dormancy is followed by tumor growth.

132

One of the biggest drawbacks concerning proper scheduling of antiangiogenic therapy is dynamic response of tumor to antiangiogenic therapy and incomplete understanding about the mechanisms of antiangiogenic agents. Anti-VEGF drugs have different mechanisms of action, each of which has a different time-frame. Initially, antiangiogenic drugs induce vascular normalization due to vascular pruning and inhibition of VEGF pathway [115, 423]. This effect occurs within a limited time-frame and dose-range referred to as “normalization window”. Administration of cytotoxic agents within this window offers maximum efficacy due to proper intratumoral distribution, which is otherwise not possible due to abnormal tumor vasculature [115]. However, identification of this normalization window is difficult because it depends upon the drug, tumor type and disease stage. Moreover, this mechanism would be advantageous only in case of chemotherapy intended to have cytotoxic effect, which is often administered as induction regimen. Maintenance therapy would require the administration of antiangiogenic drugs beyond normalization window. Continuation of anti-VEGF therapy beyond the normalization window has a different mechanism of action, i.e. vascular disruption, which is expected to starve the tumor cells [115]. However, vascular disruption results in hypoxia and the dynamic response of tumor cells to changing microenvironment and hypoxia are not properly understood.

Therefore, preclinical studies have a great value in predicting the duration of antiangiogenic activity and understanding the mechanisms of resistance. In this chapter, we report our findings of a time-response study with SK-N-BE(2) neuroblastoma mice model. Our purpose was to investigate the possible resistance mechanism to TP+PZ therapy, where the behavior of tumors after different durations of our antiangiogenic therapies was studied.

133

5.1 Methods

5.1.1 In-vivo Tumor treatment

The pre-clinical SK-N-BE(2) neuroblastoma xenograft model has been previously described in Chapter-4. When the tumors reached 0.5 cm, mice were randomized into four treatment groups: control (untreated) n=4; PZ (150 mg pazopanib) n=8; TP

(1.0 mg/Kg topotecan) n=8; TP+PZ (combination of topotecan and pazopanib, same doses as single agents) n=12. Durations of treatment planned for each regimen were

28 days, 56 days and 80 days with end points for control, TP and PZ not expected to go beyond 56 days, as per our observation in the previous chapter. Four animals from each treatment group were sacrificed either after treatment for the entire planned duration or upon reaching the end point, whichever occurred first. The criteria for end point termination were either tumor size exceeding 2.0 cm in diameter or animals showing signs of morbidity.

5.1.2 Immunohistochemistry

Harvested tumors were either fixed in 10% formalin or rapidly frozen in OCT.

Formalin-fixed and paraffin embedded tissues sections were cut at 5-7um, deparaffinized by xylene and ethanol and rehydrated in phosphate buffered saline (#

311-010-CL,Wisent Bioproducts, St. Bruno, Quebec, Canada). Following antigen retrieval with citrate buffer (pH 6.0), sections were blocked for one hour in 2% bovine serum albumin. Sections were subsequently incubated with primary antibodies. The primary antibodies were anti-phospho histone H3 (Millipore, Billerica, MA #06-570; dilution 1:500), anti-hexokinase-II (Cell Signalling, Danvers, MA #2867; dilution

1:100), anti-carbonic anhydrase-IX (Novus Biologicals, Littleton, CO, #NB100-417; dilution 1:500). This was followed by HRP conjugated, and then staining (Super

134

PicTureTM kit ; Invitrogen #87-8963). The sections were counterstained with hematoxylin. Immunohistochemistry for Ki67 and Glut-1 were performed in the

Ventana Benchmark Ultra automated machine. The primary antibodies were anti-

Ki67 (Ventana, Tucson, AZ; #790-4286, 16 min incubation at 37°C) and anti-Glut1

(Spring Biosciences, Fremont, CA, #E2840; dilution 1:500, 24 min incubation at

37°C). Mounted slides were examined under Olympus UTV1-X microscope mounted with Qimaging Retiga 2000R camera.

Frozen sections were fixed with 4% paraformaldehyde and permeabilized with

0.05% Triton X-100. After blocking with 5% bovine serum albumin in PBS for 1 hour, the sections were incubated overnight with primary antibodies: rabbit polyclonal anti-

CD31 antibody (Abcam, ab28364, dilution 1:50); rabbit anti-cleaved caspase-3 antibody (Cell Signalling 9711s; 1:400); HIF-1 alpha (BD Transduction Laboratory);

FITC conjugated alpha smooth muscle actin (SMA) (Sigma F3777, 1:500). The sections were then incubated with secondary antibodies conjugated with Alexafluor

488 or Alexafluor 594 for one hour. Following washing, sections were incubated with

DAPI and the slides were mounted with Vectashield mounting medium (H1000). The examination of tissue sections were done under Nikon ECLIPSE Ti series fluorescence microscope, using NIS Elements (BR 3.10) software.

5.1.3 Western blot

Frozen tissue portions were put in ice cold lysis buffer and homogenized. The proteins were resolved by SDS-PAGE electrophoresis and then transferred on to

PVDF membrane. After blocking for one hour with 5% milk in 0.1% Tween 20 in Tris buffer saline (TBST), the membrane was incubated overnight with primary antibodies

135 for HIF-1 alpha (BD, 610958, 1:1000), VEGF (SantaCruz, SC-152, 1:500), beta actin

(Abcam, ab8226, 1:10,000). The membranes were subsequently washed with TBST and incubated with the HRP conjugated anti-rabbit or anti-mouse secondary

(1:10,000) antibodies for one hour. The bands were detected by chemiluminescence using Amersham‟s ECL plus western blot detection system.

5.1.4 Real Time PCR

Total RNA was extracted from tumor samples using TRIzol reagent (#15596-026,

Invitrogen-Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. RNA quality and quantity were determined by gel electrophoresis and spectrophotometry, respectively. Genomic DNA contamination was removed with

Deoxyribonuclease I Amplification Grade (#18068-015, Invitrogen-Life Technologies,

Carlsbad, CA, USA). Synthesis of complementary DNA (cDNA) was performed using

1 μg of total RNA according to the manufacturer's protocol of ImProm-IITM Reverse

Transcription System (#A3800, Promega, Madison, WI, USA). The primers for

PDGF-C and the endogenous gene b-actin (ACTB) were designed using Primer

Express (3.0) Software from Applied Biosystems (Foster City, CA, USA) taking care that the forward and reverse sequences were in different exons. The sequences of primers for PDGF-C are 5'-GGG CTT GAA GAC CCA GAA GAT -3' (forward) and 5'-

CCA TCA CTG GGT TCC TCA ACT T-3' (reverse) and those for ACTB are 5‟-

AAGGCCAACCGCGAGAAG-3‟ (forward) and 5‟-

ACAGCCTGGATAGCAACGTACA-3‟ (reverse). Expression levels of PDGFC were determined by quantitative real-time PCR. This analysis was performed in a thermocycler Applied Biosystems Prism 7500 Sequence Detection System (PE

Applied Biosystems, Inc, Foster City, CA) using relative quantification. The reaction

136 mixture combined 6 μL SYBR®Green PCR Master Mix (#4309155, Applied

Biosystems-Life Technologies, Foster City, CA, USA), 3 μL sense/antisense primers, and 3 μL cDNA. The cycling conditions were: 95°C for 10 minutes, 40 cycles of 95°C for 15 seconds, and 60°C for 1 minute. Reactions were done in triplicate. For each sequence, a standard curve was constructed to determine the sensitivity and efficiency of assays. For each sample, the cycle threshold (Ct) was determined

(mean of the 3 reactions) for both the target gene and the endogenous control gene and was normalized for cDNA quantity. Subtracting the Ct of the endogenous control gene from the Ct of the target gene yields the ΔCt. The ΔCt of the control reference was then subtracted from the ΔCt of the tumor sample, yielding the ΔΔCt, and the relative quantification value was expressed as 2−ΔΔCt.

5.1.5 Statistics

Statistical significance for the difference of mean values for any markers between two treatment groups was assessed by two-tailed t test using Graphpad Prism 5.2.

For comparing the levels of markers in immunohistochemistry and immunofluorescence experiments, mean pixels indicate the mean readings of four xenografts in each group.

137

5.2 Results

5.2.1 Treatment with TP and PZ

The treatments were started two weeks after the subcutaneous injection of tumor cells, when tumor sizes ranged from 0.4 cm to 0.7 cm. End points for control, PZ and TP treated groups were reached after 23 days, 28 days and 46 days [Figure 22 A, B], respectively, and the tumors were harvested at these times. For TP+PZ, tumor growth was inhibited and remained unchanged until ≈50 days, after which they started growing gradually. However, the tumors in TP+PZ treated animals did not reach the 2cc end point until 80 days. Therefore we were able to examine the effect of continuous treatment with TP+PZ for 28 days, 56 days and 80 days.

138

Figure 21: Effect of treatment on tumor growth and survival.

The graph (A) represents the effect of treatments on the tumor growth rate. The graph (B) is the survival curve. For both a and b, n= 4, 8, 8 and 12 for control, TP, PZ and TP+PZ, respectively. In graph (B) the line representing survival for TP+PZ treated animals is not extrapolated to zero of y-axis because these animals did not reach the end point. In the other three groups, all the animals belonging to each group were sacrificed when at least one animal in that group reached the end point.

A 5 Control PZ 4 TP TP + PZ 3

1 Tumor volumeTumor(cc) 0 0 10 20 30 40 50 60 70 80 90 Days

B 100

60 Control PZ 40 % survival % TP 20 TP+PZ

0 0 20 40 60 80 100 Days

139

5.2.2 The effect of treatments on the apoptosis

Cleaved caspase-3 was used as a marker of apoptosis in our experiment.

Immunofluorescence labeling for cleaved caspase-3 (marker of apoptosis) revealed that in TP+PZ treated tumors, at all the three durations, had significantly increased numbers of cleaved caspase-3 positive cells relative to the untreated animals [Figure

23]. The apoptotic cells were evenly distributed throughout the tumor section rather than being confined to particular areas.

140

Figure 22: Apoptosis in the tumor tissues after treatment

Microscopic images (magnification, x20) of tumor sections showing hypoxic areas (HIF- 1 alpha positive, green) and cleaved caspase 3 (red). The graph below represents the pixels of caspase-3 stained areas, as measured by ImageJ software.

141

5.2.3 Effect of treatments on hypoxia and angiogenic gene expression

The tumors showed intensely stained areas for CAIX, in the tumors treated with TP for

28 days and those treated with TP+PZ at all three durations [Figure 24 A]. Western blot analysis for HIF-1 alpha revealed intense bands in all treatment groups except for untreated animals and TP treated animals sacrificed after 46 days treatment [Figure 24

B]. The tumors treated with TP+PZ for 56 days and 80 days showed the highest expression of HIF-1 alpha and VEGF [Figure 24 B]. Real time PCR revealed that the

PDGF-C expression was higher in all the treatment groups, at all durations, compared to the untreated control [Figure 24 C]. It was highest in the later durations (56 days and

80 days) of TP+PZ therapy than in other treatment groups. These results highlighted a strong effect of treatment induced hypoxia likely correlating with an anti-angiogenic effect.

142

Figure 23: Hypoxia and angiogenic gene expression in tumors after treatment

(A) Microscopic images (magnification, x40) of tumor sections on which immunohistochemistry for CAIX were performed. (B) Comparison of expression of HIF- 1 alpha and VEGF in the tumors of various treatment groups by western blot. (C) The relative gene expression of PDGF-C in treated tumors with respect to control tumors by 2−ΔΔCt method. The P value indicates the statistical significance of difference of PDGF-C expression between control groups and the mean gene expression in all treatment groups combined. TP+PZ (28d) was not included because this sample demonstrated RNA degradation (absence of bands for subunits 18 and 28S of RNA by electrophoresis of agarose gel) due to which the CT value for ACTB gene (endogenous gene used in this methodology) was extremely high. For both western blot and real- time PCR (B and C), equal volumes of tumor homogenate from all replicate tumor xenografts belonging to each treatment sub-group were mixed before the experiment.

143

144

5.2.4 The effect of treatments on tumor vessel density and pericyte coverage

Given the above observations we next examined the vascularization of these tumors.

The markers for endothelial cells and pericytes were CD31 and alpha smooth muscle actin (SMA), respectively. Immunofluorescence staining revealed that compared to control, none of the single agents, at any of the durations reduced the microvessel density, whereas TP+PZ significantly reduced microvessel densities at all the three durations [Figure 25 A, B]. SMA expression increased with time in the TP+PZ group.

The mean ratio of SMA to CD31 positive areas was used to compare the extent of pericytic coverage of tumor vasculature [Figure 25 A, C]. This ratio was significantly higher and ≥ 1.0 in tumors treated with TP+PZ for 56 and 80 days, while significantly lower in tumors treated with TP+PZ for 28 days, compared to the control tumors. The

SMA:CD31 ratio did not vary significantly from control in any of the single agent treated groups. Control tumors showed both mature blood vessels characterized by larger diameter and thick pericyte coverage (left); and immature blood vessels characterized by smaller diameter and fewer pericyte coverage (right). Although, the TP+PZ treated tumors had higher pericyte coverage, vessels here were smaller than the mature vessels in control tumors. These observations suggested a loss of tumor vascularization but maturation of the few remaining vessels in the TP+PZ tumors.

145

Figure 24: The effect of treatments on tumor vasculature

(A) Microscopic images (magnification, x20) of tumor sections showing CD31 positive tumor vasculature (red) and SMA positive pericytes (green). The two images for control sections indicate the tumor vasculature with a high pericyte coverage (left) and those with low pericyte coverage (right). (B) The comparison of microvessel densities, measured in terms of pixels of CD31 positive areas in different groups of mice. (C) The comparison of extent of pericyte coverage, measured in terms of pixel ratio for SMA to CD31 positive areas in different groups of mice. The pixels were measured by ImageJ software.

Control Control PZ TP (28d) A

TP (46d) TP+PZ (28d) TP+PZ (56d) TP+PZ (80d)

B 7 C * 1.2 6 ** 1.0 5 0.8 4 * 3 * ** 0.6 *

2 0.4 Pixels Pixels (CD31)

1 0.2 Pixel Pixel ratio (SMA:CD31) 0 0.0

PZ PZ

Control TP (28d) TP (46d) Control TP (28d) TP (46d) TP+PZ (28d)TP+PZ (56d)TP+PZ (80d) TP+PZ (28d)TP+PZ (56d)TP+PZ (80d)

146

5.2.5 The effect of treatments on proliferative index and mitotic index of tumor cells

The observation that tumor growth was significantly reduced by TP and PZ treatments led us to examine the proliferative and mitotic indices. The number of proliferating cells, as indicated by Ki67 staining was significantly higher in groups TP (28days)

(P=0.0009) and TP+PZ (28days) (P=0.0007) relative to the control [Figure 26A].

However, in all other groups, it did not show a significant difference over control tumors.

These contrary results i.e. stable tumor but high proliferative index indicated a possible cell cycle arrest. Therefore, mitotic index of the tumors were measured by phosphohistone H3 (PH3) immunohistochemistry. However, compared to control tumors, the mitotic index was significantly higher in treatment groups TP (28days) (0.005), TP+PZ

(28days) (P=0.0002) and TP+PZ (80days) (P=0.01) [Figure 26B].

147

Figure 25: The effect of treatments on proliferation and mitotic index of tumors.

Immunohistochemistry on paraffin embedded tumor sections stained with antibodies for (A) proliferation marker Ki67 (magnification, x20) and (B) mitotic marker PH3 (magnification, x40). The brown areas indicate cells or areas expressing Ki67 or PH3. Tonsil was used as positive control for both Ki67 and PH3. The graphs below the images represent the comparison of proliferative and mitotic indices measured as the pixels for areas positive for Ki67 and PH3 respectively.

Control PZ TP (28d) TP (46d) A

TP+PZ(28d) TP+PZ(56d) TP+PZ(80d) positive control

148

B Control PZ TP (28d) TP (46d)

TP+PZ(28d) TP+PZ(56d) TP+PZ(80d) positive control

149

5.2.6 Effect of treatments on indicators of elevated glycolysis in the tumor cells

In order to decipher what could be happening in the tumors as vascularization decreased and remaining vessels stabilized, yet with a more hypoxic microenvironment, we looked at other hypoxia and metabolic relevant markers [425].

Immunohistochemistry revealed that expression of Glut-1 and hexokinase-II were significantly higher, compared to the control, in the tumors treated with TP+PZ for all three durations. [Figure 27 A, B]. In control and single agent treated tumors, the cells overexpressing Glut-1 were scattered uniformly among those lacking Glut-1 expression. In contrast, in TP+PZ treated tumors, cells overexpressing Glut-1 were present in groups, separated from those lacking Glut-1 expression.

150

Figure 26: The effect of treatments on markers of glycolysis

(A) The microscopic images (magnification, x20) of tissue sections stained for antibody for Glut-1. Placenta was used as positive control tissue. (B) The microscopic images (magnification, x40) of tissue sections stained with antibody for hexokinase-II. Paraffin embedded section of a human glioblastoma was used as a positive control. Arrows indicate the cells overexpressing Glut-1 or hexokinase-II.

A Control PZ TP (28d) TP (46d)

TP+PZ(28d) TP+PZ(56d) TP+PZ(80d) Positive control

6 0

) *

- t

u 4 0 *

(

s * * l

e 2 0

i P

0 l ) ) ) ) ) o d Z d d d d r 8 P 6 8 6 0 t 2 5 8 n (2 (4 ( ( ( o P P Z Z Z C T T P P P + + + P P P T T T

151

B Control PZ TP (28d) TP (46d)

TP+PZ(28d) TP+PZ(56d) TP+PZ(80d) Positive control

) 4 0

I I

- * *

e s

a 3 0

k o

x 2 0

e * H

( *

l 1 0

x i P 0 l ) ) ) ) ) o d Z d d d d r 8 P 6 8 6 0 t 2 5 8 n (2 (4 ( ( ( o P P Z Z Z C T T P P P + + + P P P T T T

152

5.3 Discussion

A major challenge with the discovery of an effective therapy for cancers despite promising pre-clinical efficacy is the loss of efficacy or refractoriness of cancers or the relapse of cancers after remission. This limitation applies to antiangiogenic therapies as well [423]. Unlike conventional chemotherapy and therapies that directly targeted tumor cells, antiangiogenic therapy was expected to sustain the antitumor effect because the expected target was genetically stable endothelial cells. However, even antiangiogenic therapy was found to lose efficacy after an initial phase of tumor regression or stabilization. In these cases, resistance mechanisms such as upregulation of angiogenic factors, vascular maturation, hypoxic adaptation of tumor cells and enhanced metastatic potential of tumor cells have been reported [423, 425,

426].

In the Chapter-4 of this study, our treatment with LDM topotecan and pazopanib delayed tumor growth but was unable to stop tumor growth in a neuroblastoma metastatic model even after continued treatment. Therefore, in this part of the study we attempted to gain insight into the changes in tumor behavior in response to prolonged treatment with LDM topotecan and pazopanib in SK-N-BE(2) subcutaneous xenograft model.

We used MYCN amplified and P53 mutated cell line SK-N-BE(2) to investigate the mechanism resistance to antiangiogenic therapy. It is a widely studied cell line with regards to undesirable effects such as invasiveness and stemness. Previously, the effect of CXCR4 mediated enhancement of invasiveness of neuroblastoma cells and the effect of tumor microenvironment on CXCR4 was confirmed using SK-N-BE(2) cells

153

[427]. The effect of hypoxia on tumor stemness has been studied in our laboratory using SK-N-BE(2) where the highly tumorigenic side population cells migrated towards injured conditioned medium migrated in Boyden‟s chamber [61]. It possesses molecular features associated with resistance to hypoxia or antiangiogenic therapy; MYCN amplification and P53 mutation [76, 385] . Previously, xenografts derived from P53-null colorectal cancer cells has been found to be significantly less effective to DC101 and vinblastine compared to those derived from wild type cells [400]. N-myc promotes hypoxic progression of neuroblastoma cells by countering HIF-1 alpha mediated inhibition of cell cycle progression [428]. Though, in Chapter-1 the disease progression during continued therapy was observed in experimental metastatic model, we investigated the tumor dynamics in response to our therapy in subcutaneous xenograft model because subcutaneous xenografts allow uniform tumor growth and frequent tumor monitoring.

The immunofluorescence labeling of the xenografts revealed that the tumors treated with the TP+PZ maintained a significantly higher level of hypoxia and higher number of apoptotic cells at all the three durations. However, the existing vasculature now exhibited robust pericyte coverage as compared to the control treated tumors. The cells in tumors treated with TP+PZ for prolonged durations (56 and 80 days) demonstrated comparable proliferative and mitotic indices with that of untreated tumors. However, unlike untreated tumors and other treated groups, tumors in these two groups had higher levels of Glut-1 and hexokinase-II, indicative of an alteration in metabolism, which may facilitate survival and proliferation under hypoxia.

154

In the present experiment, tumor vessel densities in tumors treated with single agents were not significantly different from those in the control when they reached the end point. Hence, the loss of efficacy of single agent LDM topotecan and pazopanib can be attributed to the fact that single agents are unable to sustain antiangiogenic effect. Here, angiogenesis may occur due to the expression of factors including VEGF and PDGF-C. In contrast, the tumors in the combination group maintained a low microvessel density even after 56 days and 80 days of treatment, the stages when they were growing gradually, despite having higher expressions of

VEGF and PDGF-C. Notably, these tumors had higher expression of hypoxia markers, HIF-1 alpha and CAIX.

The tumor hypoxia was assessed by western blot of HIF-1 alpha and immunohistochemistry of CAIX. CAIX is a downstream target of HIF-1 alpha and is also used as an indicator of tumor hypoxia [429]. CAs are ubiquitously expressed in cells under both physiological and pathological conditions. It catalyzes the reversible reaction of water and carbon dioxide (produced by cellular respiration) into bicarbonate, thus maintain acid-base balance in the tissues. CAIX isoform is widely expressed in cancers. In tumor cells, it transports the hydrogen ion to extracellular matrix, thereby maintaining a physiological intracellular pH and a low pH in extracellular matrix. It has been co-localized with pimonidazole at a distance of 80µm from CD34+ blood vessels in head and neck tumors, which makes it a suitable indicator of tumor hypoxia. .In neuroblastoma, it is associated with adverse pathobiology [430].

Though, HIF-1 alpha is known to be stabilized in normoxic cells by oncogene induced expression and stabilization of HIF-1 alpha in several cancers [431], in this

155 study it is a definitive marker of hypoxia. Previously, it has been reported that at 21% or 5% O2, HIF-2 alpha predominated in the cytoplasm and nuclei of SK-N-BE(2)c cells with little or no HIF-1 alpha, whereas at 1% O2, stabilization and nuclear localization of both HIF-1 alpha an HIF-2 alpha were observed [432]. Therefore, in the present experiment higher HIF-1 alpha stabilization is indicative of high tumor hypoxia. High HIF-1 alpha is conceivable considering the low microvessel densities in the xenografts treated with TP+PZ. Also high VEGF and CAIX can be expected because they are the downstream targets of HIF-1 alpha [433]. HIF-1 alpha, resulting from hypoxia, is a major cause of antiangiogenic resistance [33]. HIF-1 alpha stabilized by hypoxia acts as a transcription factor for number of genes involved in angiogenesis, cell survival, hypoxic adaptation, invasiveness and stemness [433]. The list of genes regulated by HIF-1 alpha is shown in Figure 28.

The increased hypoxia, as indicated by HIF-1 alpha expression, can have several implications. Previously in our laboratory, hypoxia is proved to impart stem cell phenotype to SK-N-BE(2) neuroblastoma cells [61]. Another study conducted in our laboratory reports that CXCR4 expression in neuroblastoma cells, including SK-N-

BE(2), is associated with an invasive phenotype [427]; while a study conducted elsewhere confirms that CXCR4 expression in monocytes, tumor-associated macrophages, HUVEC, fibroblasts and cancers cells is up-regulated by hypoxia

[434]. Chronic hypoxia imparts drug resistance to vincristine and etoposide in neuroblastoma cells (SH-EP1 and SH-SY5Y) accompanied by increase in HIF-1 alpha and CAIX expression, whereas, down-regulation of HIF-1 alpha reversed this resistance [435].

156

Here, the up-regulation of HIF-1 alpha and VEGF is contradictory to previous evidences regarding the anti-HIF-1 alpha effects of topotecan and VEGF inhibitors

(sunitinib and sorafenib). Topotecan is reported to reduce the expression of HIF-1 alpha and VEGF production, induced by IGF-1/PI3K/AKT pathway in several MYCN and non-MYCN neuroblastoma cell lines including SK-N-BE(2) [279, 280]. Ligand- induced activation of tyrosine kinases such as EGFR and PDGFR-β enhanced the expression of HIF-1 alpha in SK-N-AS neuroblastoma cells, whereas RTKIs, sunitinib and sorafenib were able to inhibit this downstream effect of tyrosine kinases

[436]. Also combining bevacizumab or sunitinib with low dose topotecan reduced down-stream targets of HIF-1 alpha such as VEGF and Glut-3, simultaneously maintaining a low vessel density in neuroblastoma xenografts [8]. However, it needs to be noted that in that study the animals were treated with the combination for only two weeks, whereas in our study the earliest sacrifice of animals treated with LDM topotecan and combination was done after four weeks (28 days), during which the tumor phenotype may change. Prolonged observation may be a key to revealing phenotypic changes that are informative about the malignant potential.

157

Figure 27: The downstream targets of HIF-1 alpha. (reproduced, with permission, from Semenza, G.L., Targeting HIF-1 for cancer therapy. Nat Rev Cancer, 2003. 3(10): 721-32)

158

We entertained the idea that a cause of resistance to the combination therapy could be higher pericyte coverage. Though, in single agent groups, there were pericytes around blood vessels, in the TP+PZ groups treated for 56 days and 80 days the fraction of SMA:CD31 was ≥ 1 indicating that, despite loss of vascularization, all the existing vasculature had pericyte coverage. Tumor vascular maturation characterized by increased pericyte coverage of endothelium is reported to render endothelial cells resistant to apoptotic effects of antiangiogenic therapy [423, 426].

PDGF-B and C have been implicated in causing vascular maturation [361, 363].

Platelet Derived Growth Factor-C (PDGF-C) is an angiogenic factor involved in VEGF- independent angiogenesis [437] . PDGF-C has been linked to higher pericyte coverage in glioblastoma [361]. Therefore, our observation of higher expression of PDGF-C suggests that PDGF-C may have contributed to the vascular maturation of TP+PZ treated tumors. PDGF-C is known to impart endothelial resistance to anti-VEGF therapy. It has been reported to be upregulated in tumors refractory to anti-VEGF treatment in mice tumor models and its inhibition by PDGF-C neutralizing antibodies enhanced the efficacy of anti-VEGF therapy [438].

Next we considered the various mechanisms by which tumors treated with TP+PZ for

56 days and 80 days tumors might grow despite the low vasculature and higher hypoxia. This can occur either due to lower apoptosis or higher proliferation rate compared to TP+PZ tumors treated for 28 days. We used Ki67 to compare the proliferative index, which is a widely used proliferation marker in neuroblastoma [439].

Ki67 assessed by immunohistochemistry is positively correlated with progressive disease and worse outcome in neuroblastoma. Ki67 is a non-cell cycle specific

159 proliferation marker. Phosphohistone H3 is an indicator of mitosis. Histone H3 is associated with the structure of chromatin [440]. During mitosis, it is phosphorylated at serine residue, whereas, it is dephosphorylated upon cell‟s exit from mitosis. Inhibition of aurora kinase, involved in the phosphorylation of histone H3, has been reported to induce apoptosis in SH-SY5Y neuroblastoma cells [441]. Therefore, we considered

Ki67 and phosphohistone H3 as appropriate indicator of proliferation and mitosis, respectively, in our neuroblastoma xenografts.

We observed that although the TP+PZ treated tumors, at all three durations, contained a significantly higher number of apoptotic cells, these tumors also had an appreciably higher proliferative index comparable to untreated and single agent treated tumors. Here we conclude that though some tumor cells succumbed to antiangiogenic therapy, others had adapted to the hypoxic environment. Another conflicting observation is that tumors treated with TP+PZ for 28 days had a higher proliferative index than any other groups. Considering the stable tumor sizes at this stage of therapy, these tumors would be expected to have relatively lower proliferative indices. Hence, we postulated that this could be an indication of cell cycle arrest. However, significantly higher levels of phospho histone H3 were observed in TP+PZ treated tumors after 28 days treatment, which indicates that the number of cells reaching M-phase is higher in this group. Therefore, this finding is an indication of higher proliferative index, unless there is mitotic arrest. However, all the reports of hypoxia induced cell cycle arrest suggest that hypoxia causes cell cycle arrest of cancer cells in G1 or G2/M phase [442-444]. There are no reports of hypoxia causing cell cycle arrest in mitotic phase alone. Also, if direct cytotoxic effect of topotecan on tumor cells is considered, it causes cell cycle arrest before the cells

160 reach mitosis [445]. Therefore, we surmised that the higher Ki67 expression in tumors treated with TP+PZ for 28 days might have indicated a higher proliferative index in this group. On another point, although this might seem improbable, considering the low tumor sizes at this stage of treatment, the possibility is that, though, TP+PZ therapy demonstrated sustained antiangiogenic efficacy till 80 days, the overall antitumor efficacy was lost before 28 days as the tumor cells started to proliferate. Since we do not have TP+PZ tumors collected before 28 days, the actual time limit till when TP+PZ therapy can inhibit the proliferation of tumor cells could not be established in this study.

Our next query was how tumor cells could proliferate in an environment characterized by hypoxia and low vasculature as in the TP+PZ treated tumors. We postulated that the SK-N-BE(2) tumor xenografts treated with TP+PZ for prolonged durations (56 days and 80 days), being comprised of heterogeneous population of cells, included some of which were undergoing higher apoptosis and others which had acquired the potential to survive and proliferate under hypoxia. One mechanism by which tumor cells acquire the ability to survive and proliferate under hypoxia is by depending on glycolysis [446]. This can be expected considering higher levels of

HIF-1 alpha in TP+PZ treated tumors.

Elevation of glycolysis as an adaptive mechanism is a critical step in somatic evolution of tumor cells. The altered glucose metabolism in cancer cells was first discovered by

Otto Warburg in 1920s [446]. He attributed this phenomenon to impaired mitochondrial activity and therefore hypothesized that cancers arise due to impaired mitochondrial metabolism. However, later it was described that elevated glycolysis is the result, not

161 the cause of malignant transformation of tumor cells and that this adaptive mechanism occurs even in the presence of oxygen in cells with intact mitochondria (aerobic glycolysis) and also in the absence of oxygen (anaerobic glycolysis).

Anaerobic glycolysis: Tumors contain areas of intermittent and chronic hypoxia due to inefficient vasculature, rapid growth of tumor cells and revascularization [446]. Tumors farther than 100µm are unable to get adequate supply of oxygen and nutrients. This scarcity of oxygen and nutrients exerts evolutionary pressure on tumor cells.

Consequently, cells undergo changes that equip them to survive and proliferate under hypoxia. This includes dependence of tumor cells on glycolytic pathway, which is brought about by up-regulation of glucose transporters and glycolytic phosphorylating enzymes.

Under physiological condition of normoxia, cells rely on glycolysis, TCA or Kreb‟s cycle and oxidative phosphorylation for glucose metabolism [446]. Glycolysis is the anaerobic phase which converts glucose into two molecules of pyruvic acid. Pyruvic acid enters

TCA cycle where it is broken down into CO2 and H2O through a series of steps. ATP yield from glycolysis, Kreb‟s cycle and oxidative phosphorylation are 2, 2 and 36 respectively. Under hypoxia, due to lack of oxygen, tumor cells rely totally upon the anaerobic stage of cellular respiration i.e. glycolysis. However, since the yield of ATP is considerably lower in glycolysis, the cells compensate by increasing the number of glycolytic cycles by upregulating glucose intake and upregulating glycolytic enzymes.

Therefore, upregulation of glucose transporter (Glut-1) and glycolytic enzymes are indicators of elevated glycolysis. The pyruvic acid produced in glycolysis is converted to lactic acid by lactate dehydrogenase and is transported to extracellular matrix which

162 renders the tumor microenvironment acidic while maintaining a favourable intracellular pH [446]. The metabolic behavior of cells in normoxia and hypoxia has been illustrated in Figure 29.

In anaerobic glycolysis, upregulation of glucose transporters and glycolytic enzymes are mediated by HIF-1 alpha [447]. The down-stream targets of HIF-1 alpha relevant to glycolysis are aldolase A, phosphoglycerate kinase, hexokinases and glucose transporters (Glut-1 and Glut-3) [448]. In neuroblastoma, HIF-1 alpha and MYCN combinatorially regulate Glut-1 and hexokinase-II [428].

Anaerobic glycolysis has been reported in several cancers. In pancreatic cancer cells glycolytic genes such as hexokinase-II, Phosphoglycerokinase-1, pyruvate dehydrogenase and lactate dehrdrogenase as well as Glut-1 were up-regulated by 7- fold when exposed to hypoxia [449]. Hypoxia is known to induce multidrug resistant and glycolytic genes (Glut-1, hexokinase-II, GAPDH and lactate dehydrogenase) and co- localized with pimonidazole in ovarian and breast cancer xenografts [425].

There are reports of anaerobic glycolysis in response to antiangiogenic therapy. In glioblastoma patients, anti-VEGF treatment with bevacizumab reduced vessel density, increased tumor hypoxia and also elevated the glycolysis which was accompanied by enhanced invasiveness [376]. In HCC cells, lactate dehydrogenase inhibitor oxamic acid potentiated the effects of sunitinib, sorafenib and imatinib [450]. In colorectal cancer cells, bevacizumab resistance was associated with elevated glycolysis and glycolysis inhibitor 3-BrPA demonstrated efficacy against xenografts derived from bevacizumab resistant cells [451].

163

Figure 28: Diagrammatic representation of glucose metabolism under normoxia and hypoxia (reproduced, with permission, from Gatenby, R.A. and R.J. Gillies, Why do cancers have high aerobic glycolysis? Nat Rev Cancer, 2004. 4(11): 891-9).

Aerobic glycolysis: Elevated glycolysis in normoxic cancer cells may seem counter- intuitive provided the low output of ATP in this phase of cellular respiration compared to

TCA cycle and oxidative phosphorylation [452]. However, deeper investigations have established that aerobic glycolysis is a beneficial phenomenon which functions to fulfill the cellular necessities beyond ATP production. In fact, glycolytic intermediates provide raw materials for macromolecule synthesis (lipids, nucleotides and amino acids) in rapidly proliferating tumor cells. Aerobic glycolysis is triggered by oncogenes such as

164 mutated ras, receptor tyrosine kinases and AKT which upregulate HIF-1 alpha, glycolytic enzymes and glucose transporters.

Glut-1 is a member of Glut family of glucose transporters [453]. Gluts are present on the cell membrane and are involved in the bidirectional transport of glucose. Glut-1 is the most abundantly expressed among Gluts and is mostly expressed in erythrocytes, endothelial cells and fibroblasts. Since glucose is a polar molecule unable to transport across the lipophilic cell membrane by passive diffusion, Glut-1 greatly enhances the glucose transport by facilitated diffusion. Abundant Glut-1 expression is a common feature of many cancers. In rectal cancer, high Glut-1 expression, as detected by immunohistochemistry, along with HIF-1 alpha and glycolytic enzymes, is an indicator of anaerobic glycolysis [454].

Hexokinase-II is a glycolytic enzyme that catalyzes the conversion of glucose to glucose-6-phosphate. However, it has another function i.e. inhibition of apoptosis which is effected by its binding to Voltage Dependent Anion Channel (VDAC) [455]. VDAC and pro-apoptotic Bax form the channel in mitochondrial membrane for release of cytochrome-c which initiates apoptosis. Hexokinase-II binding to VDAC inhibits this process and provides survival advantage to cells. Therefore, hexokinase-II confers dual survival advantage to cancer cells, first by facilitating elevated glycolysis and second by its anti-apoptotic mechanism. Hence it is commonly detected in most cancers [456].

Other indicators of elevated glycolysis in cancers are phosphofructokinase, glyceraldehyde-3-phosphate dehydrogenase, phosphoglycerate kinase, pyruvate kinase and lactate dehydrogenase [457].

165

HIF-1 alpha is known to elevate glycolysis in cancers [433]. Combinatorial effects of

MYCN and HIF-1 alpha, is reported to promote the proliferation and expression of

Glut-1 and hexokinase-II in MYCN amplified neuroblastoma cells [428]. Glut-1 and hexokinase-II have been used previously as markers of elevated glycolysis [425].

Glut-1 is upregulated in order to facilitate the transport of glucose molecules. Glut-1 has been proposed to be a marker for hypoxia in solid tumors [458-460].

Hexokinases are another class of enzymes that catalyse the first step of glycolysis i.e. phosphorylation of glucose. Hexokinase-II is up-regulated in cancers to facilitate enhanced glycolysis; therefore it has been used as a definitive marker of glycolysis in many cancers, including neuroblastoma [428, 456].

These findings led us to perform immunohistochemistry for Glut-1 and hexokinase-II in the tumor xenografts. Whereas the control and single agent groups revealed few cells with high Glut-1 expression, uniformly scattered within the tumor; TP+PZ treated tumors had cells showing intense staining for Glut-1, present as clusters.

Hexokinase-II was also highly expressed in TP+PZ treated tumors. This indicates that, in this study, TP+PZ treated tumors become dependent on the glycolytic pathway to meet the energy requirement. We therefore surmise that the hypoxia challenged tumor cells upregulated HIF1alpha which then led to upregulation of Glut-

1 and hexokinase-II, needed for resumption of strong tumor growth. Interestingly, we observed foci of such cells suggesting a possible clonogenic effect, that is, outgrowth of hypoxia resistant clones.

The elevated glycolysis has several implications in cancers. Glut-1 up-regulation is associated with increased proliferation and worse prognosis in neuroblastoma and

166 breast cancers [461, 462]. Apart from its role in glycolysis, hexokinase-II also imparts resistance to chemotherapy-induced apoptosis by binding to the mitochondrial membrane, thereby reducing its permeability [456]. Not surprisingly, clotrimazole, which displaces hexokinase-II from mitochondria, enhances the cytotoxicity of cisplatin in ovarian cancer [463].

Before establishing a general link between HIF-1 alpha, MYCN amplification and tumor adaptation to hypoxia, other conflicting evidences need to be considered. In an ovarian cancer study, HIF-1 alpha expression was positively associated with patient survival because here it enhanced the apoptosis [433]. One study reports positive correlation between high HIF-1 alpha expression and favorable outcome in advanced stage neuroblastoma [464]. As a matter of fact, before drawing a conclusion regarding the effect of prolonged antiangiogenic therapy on a particular cancer, it is important to have a more extensive knowledge about the molecular features of that cancer. One of the transcriptional targets of MYCN is tumor suppressor gene P53 [465]. The net effect of HIF-1 alpha and MYCN amplification i.e. whether they are likely to cause apoptosis or cell survival, also depends upon the status of tumor suppressor gene P53 [466]. Wild-type P53, once induced by HIF-1 alpha, acts as a transrepressor for HIF-1 alpha target genes involved in angiogenesis and glycolysis, and activates the transcription of pro-apoptotic genes.

However, if P53 is inactive or mutated, the effect of HIF-1 alpha on pathways that promote cell survival predominates. Here, SK-N-BE(2) used for developing xenograft carries mutation in the P53 gene [385], therefore, it is possible that HIF-1 alpha and

MYCN were able to induce cell survival, proliferation and pro-angiogenic gene expression in response to antiangiogenesis induced hypoxia. Considering the above

167 facts, the tumor response to prolonged TP+PZ therapy observed in this study, using a MYCN amplified / P53 mutated cell line, cannot be generalized for all cases of neuroblastoma or any other pediatric solid tumors.

168

6 Summary and Conclusions

6.1 Thesis Summary

In Chapter-4, we have proved that the combination of LDM topotecan and pazopanib has higher antitumor efficacy compared to single agents in neuroblastoma, rhabdomyosarcoma and osteosarcoma. This combination caused significant lowering of CAFs, compared to control and single agents. Since no other synergism or additive effect have been observed between topotecan and pazopanib in in-vitro and PK experiments, the mechanism behind the efficacy of the combination can be attributed to antiangiogenic activity in all three solid tumor models. Direct interaction of pazopanib with VEGF signaling pathway within tumor cells cannot be excluded

[38]. Potential bone marrow toxicity may be expected with the combination of LDM topotecan and pazopanib, therefore, caution should be taken before claiming that such a combination is not myelotoxic.

In Chapter-5, we report that LDM topotecan and pazopanib, in a neuroblastoma xenograft model developed from a MYCN amplified/P53 mutated cell line, loses antitumor efficacy after some time even if used continuously. This mechanism of resistance to the combination therapy is different from that to single agent therapy.

Whereas the tumors evade antiangiogenic effect of single agent therapy, the combination of these agents demonstrated sustained antiangiogenic effect. The tumors treated with TP+PZ for prolonged duration have a limited but more mature vasculature.

Despite sustained antiangiogenic effect, overall antitumor efficacy of TP+PZ is lost and the resistance is associated with the capacity of tumor cells to survive and proliferate under antiangiogenesis-induced hypoxia. These tumor cells rely heavily upon the

169 glycolytic pathway for maintenance of tumor growth as revealed by higher staining of

Glut-1 and hexokinase-II in TP+PZ treated tumors. Interestingly, our time study reveals that the tumor cells acquire this capacity long before the tumors start to show an actual increase in size, i.e., resumption of growth.

6.2 Thesis conclusions

The findings of Chapter-1 of this study support the Phase-I trial of this combination in pediatric solid tumors with a potential, if proven to be safe, to be integrated into a post-stem cell transplantation regimen. The evidence suggests that TP+PZ can be an effective strategy to prolong the survival in patients with pediatric extracranial solid tumors. An international phase-I study is currently under development in Canada and Europe under the leadership of C17 pediatric developmental clinical trial network and the European Phase-I Consortium ITCC.

However, there are potential concerns with prolonged therapy using TP+PZ. First, apart from significantly higher efficacy, TP+PZ also had significantly higher toxicity, compared to the single agents. In order to conclude that the combination of oral metronomic topotecan and pazopanib widens the therapeutic window, its benefit must outweigh its risk, which remains to be confirmed in the clinical trial. Second concern regarding

TP+PZ therapy is the development of resistance. Our time-response study with LDM topotecan and pazopanib in a MYCN amplified/ P53 mutated neuroblastoma model proves that though the antiangiogenic activity of TP+PZ is maintained during prolonged therapy, it cannot be translated to overall antitumor efficacy due to the acquired resistance of the tumor cells, which is associated with the elevated glycolysis. Also, we have observed that tumor cells start proliferation long before notable tumor size

170 increase was observed. Hence, from a clinical perspective, any sign of a stabilized tumor burden under such treatments should be taken with caution. Instead, further investigations into markers of antiangiogenic resistance would enable clinicians to decide the duration after which the antiangiogenic therapy needs to be discontinued or modified. Elevated plasma levels of proangiogenic factors have been suggested as marker of antiangiogenic resistance [54]. Based upon the present study, we also suggest that indicators of proliferation, cell cycle and elevated glycolysis in circulating tumor cells can be potential biomarkers of antiangiogenic resistance. Furthermore, the contrary results between this study and previous short-term studies, with respect to

HIF-1 alpha and downstream targets in neuroblastoma, supports the need for long-term preclinical investigation of antiangiogenic therapies if they are to be considered as maintenance regimen [8]. In effect, combining antiangiogenic therapies with drugs targeting tumor metabolism could be a more effective strategy to overcome antiangiogenic resistance [33].

However, the dynamic response of tumor to prolonged TP+PZ therapy could be specific to this molecular feature. Considering the previous evidences regarding the influence of P53 status on HIF-1 alpha mediated tumor response, these findings regarding the antiangiogenic resistance to TP+PZ therapy cannot be generalized for wider range of neuroblastoma or other solid tumors. Therefore, such time-response studies need to be conducted on preclinical models representing a wider range of pediatric cancers with different molecular phenotypes.

171

7 Future Directions

In the present studies we have investigated the efficacy of LDM topotecan and pazopanib in pediatric solid tumor mice models and the resistance mechanism to these drugs during prolonged administration in a neuroblastoma model. The findings of this study have given an idea about the potential strategies to reverse the drug resistance to TP+PZ therapy during prolonged treatment. Also, this study opens the rationale for several future investigations. Following are some strategies that can be adopted to reverse the resistance to antiangiogenic therapy.

1) Use of inhibitors of glycolytic pathway: Considering the dependence of tumor cells on glycolytic pathway due to elevated hypoxia, glycolytic inhibitors hold promise for avoiding or reversing the drug resistance to antiangiogenic therapies such as

TP+PZ. Glycolytic inhibitors have the potential to reverse the hypoxia-induced drug resistance and also the resistance mediated by ABC transporters [467]. 3- bromopyruvate acid (3-BrPA), an inhibitor of hexokinase, is reported to reduce Glut-1 expression in neuroblastoma cell lines [461]. Another hexokinase inhibitor lonidamine has demonstrated efficacy in the clinical trials of glioblastoma multiforme, ovarian cancer and NSCLC in combination with other agents [467]. 2-deoxy-D-glucose has demonstrated anti-clonogenic effect on N-type, S-type and I-type neuroblastoma cell lines [468]. Inhibitors of glycolytic enzymes such as phosphofructokinase, triosphosphate isomerase, GAPDH, enolase, pyruvate kinase and lactate dehydrogenase inhibitors are also under preclinical investigation for cancer therapy

[467].

172

2) Hypoxia Activated Prodrugs: Hypoxia Activated Prodrugs (HAPs) are a class of pro-drugs of cytotoxic agents, which are activated under the hypoxic conditions of tissues, especially inside the tumors. Since these agents can only be activated under hypoxic conditions and exist as inactive prodrugs under normoxia, they can specifically target the malignant cells residing in hypoxic zones, while having little or no cytotoxic effect on the normal cells. The examples of HAPs which are under investigation are

AQ4N, PR104 and TH302. PR-104 has been tested by PPTP and the Phase-I trial has been conducted in adult solid tumor patients [469]. AQN4, a prodrug of

Topoisomerase-II inhibitor, has been reported to enhance the efficacy of other cytotoxic agents like cisplatin, cyclophosphamide and thiotepa [470, 471] while its safety has been established in a phase-1 trial [472].

TH-302 is a prodrug of active phosphoramidate bis-alkylator derived from ifosfamide, which can be converted to active nitrosoamine derivative under hypoxic condition. It has been preclinically tested as a monotherapy and in combination with cytotoxic drugs

([473, 474]. Its Phase-I trials have been conducted in solid malignancy patients as single agent and in soft tissue sarcoma patients in combination with doxorubicin [475,

476]. Our preliminary studies have indicated the preclinical efficacy of this drug in osteosarcoma (as single agent) and neuroblastoma (in combination with sunitinib).

173

8 Limitations of this study

1) Though this study has proved the superior efficacy of LDM topotecan and pazopanib in four models of neuroblastoma, one model of osteosarcoma and one model of rhabdomyosarcoma, the response of tumor to prolonged therapy has been studied only in two models of neuroblastoma SK-N-BE(2) and BE(2)-c metastatic model, which are MYCN amplified/P53 mutated cell lines. Since the effect of hypoxia induced HIF-1 alpha depends upon multiple factors, including MYCN amplification and

P53 abnormality, the findings of this study may not be applicable to a wider range of neuroblastomas, e.g. those harboring single MYCN copy or wild-type P53 . Since this therapy is intended to be used as a prolonged maintenance regimen for pediatric solid tumors in general, the response of other tumor models representing wider molecular features of neuroblastoma and sarcoma is pending.

2) The changes in tumor behavior in response to prolonged TP+PZ therapy was studied in a subcutaneous xenograft model. This may present a drawback because the therapeutic response is influenced by tissue microenvironment. As evidence, in the chapter-1 of this study, we observed that prolonged TP+PZ therapy was able to prevent the tumor development in liver, but not in kidney and bone marrow (unlike the control and single agent therapies) in BE(2)-c metastatic model. An orthotopic or i.v. metastatic model can be more appropriate to study the tumor behavior in response to prolonged antiangiogenic therapy.

3) This study proves an association, not a causal relationship, between elevated glyolysis and increased proliferation of the tumor cells in response to prolonged therapy with TP+PZ. Glycolysis can be a cause or an effect of increased proliferative capacity

174 of the tumor cells under antiangiogenesis-induced hypoxia. Further investigation is required to decipher the actual molecular mechanisms which lead to the resistance to this combination.

175

9 References

1. Canadian Cancer Society website [cited 2013]. Available at http://www.cancer.ca/about-us/our-research.

2. SEER, author. Surveillance, epidemiology, and end results (SEER) program web site. [Cited 2003]. Available at: http://www.seer.cancer.gov.

3. Bayo, J., et al., Chemotherapy-induced nausea and vomiting: pathophysiology and therapeutic principles. Clin Transl Oncol, 2012. 14(6): p. 413-22.

4. Arnon, J., et al., Genetic and teratogenic effects of cancer treatments on gametes and embryos. Hum Reprod Update, 2001. 7(4): p. 394-403.

5. Nunez, S.B., et al., Risk-based health monitoring of childhood cancer survivors: a report from the Children's Oncology Group. Curr Oncol Rep, 2007. 9(6): p. 440-52.

6. Gillet, J.P. and M.M. Gottesman, Mechanisms of multidrug resistance in cancer. Methods Mol Biol, 2010. 596: p. 47-76.

7. Kuttesch, J.F., Multidrug resistance in pediatric oncology. Invest New Drugs, 1996. 14(1): p. 55-67.

8. Kucerová, H., et al., Significance of P-glycoprotein expression in childhood malignant tumors. Neoplasma, 2001. 48(6): p. 472-8.

9. Goto, H., et al., Multidrug resistance-associated protein 1 (MRP1) expression in neuroblastoma cell lines and primary tumors. Med Pediatr Oncol, 2000. 35(6): p. 619-22.

10. Komdeur, R., et al., Multidrug resistance proteins in rhabdomyosarcomas: comparison between children and adults. Cancer, 2003. 97(8): p. 1999-2005.

11. Klunder, J.W., et al., Expression of multidrug resistance-associated proteins in rhabdomyosarcomas before and after chemotherapy: the relationship between lung resistance-related protein (LRP) and differentiation. Hum Pathol, 2003. 34(2): p. 150-5.

12. Plasschaert, S.L., et al., The role of breast cancer resistance protein in acute lymphoblastic leukemia. Clin Cancer Res, 2003. 9(14): p. 5171-7.

13. van den Heuvel-Eibrink, M.M., et al., Increased expression of the breast cancer resistance protein (BCRP) in relapsed or refractory acute myeloid leukemia (AML). Leukemia, 2002. 16(5): p. 833-9.

176

14. Bertino, J.R., et al., Resistance Mechanisms to Methotrexate in Tumors. Oncologist, 1996. 1(4): p. 223-226.

15. Awad, O., et al., High ALDH activity identifies chemotherapy-resistant Ewing's sarcoma stem cells that retain sensitivity to EWS-FLI1 inhibition. PLoS One, 2010. 5(11): p. e13943.

16. Vinogradov, S. and X. Wei, Cancer stem cells and drug resistance: the potential of nanomedicine. Nanomedicine (Lond), 2012. 7(4): p. 597-615.

17. Zunino, F., et al., Role of apoptotic response in cellular resistance to cytotoxic agents. Pharmacol Ther, 1997. 76(1-3): p. 177-85.

18. Viktorsson, K., R. Lewensohn, and B. Zhivotovsky, Apoptotic pathways and therapy resistance in human malignancies. Adv Cancer Res, 2005. 94: p. 143- 96.

19. Shah, M.A. and G.K. Schwartz, Cell cycle-mediated drug resistance: an emerging concept in cancer therapy. Clin Cancer Res, 2001. 7(8): p. 2168-81.

20. Garraway, L.A. and P.A. Jänne, Circumventing cancer drug resistance in the era of personalized medicine. Cancer Discov, 2012. 2(3): p. 214-26.

21. Guo, T., et al., Mechanisms of sunitinib resistance in gastrointestinal stromal tumors harboring KITAY502-3ins mutation: an in vitro mutagenesis screen for drug resistance. Clin Cancer Res, 2009. 15(22): p. 6862-70.

22. London, W.B., et al., Clinical and biologic features predictive of survival after relapse of neuroblastoma: a report from the International Neuroblastoma Risk Group project. J Clin Oncol, 2011. 29(24): p. 3286-92.

23. Pappo, A.S., et al., Survival after relapse in children and adolescents with rhabdomyosarcoma: A report from the Intergroup Rhabdomyosarcoma Study Group. J Clin Oncol, 1999. 17(11): p. 3487-93.

24. Wasilewski-Masker, K., et al., Late recurrence in pediatric cancer: a report from the Childhood Cancer Survivor Study. J Natl Cancer Inst, 2009. 101(24): p. 1709-20.

25. Kågedal, B., Detecting minimal residual disease in neuroblastoma: still a ways to go. Clin Chem, 2009. 55(7): p. 1268-70.

26. Chertin, B., et al., Post-chemotherapy microscopic residual prostate rhabdomyosarcoma: long-term conservative follow-up. Pediatr Surg Int, 2002. 18(1): p. 68-9.

27. Paietta, E., Assessing minimal residual disease (MRD) in leukemia: a changing definition and concept? Bone Marrow Transplant, 2002. 29(6): p. 459-65.

177

28. Reynolds, C.P., Detection and treatment of minimal residual disease in high-risk neuroblastoma. Pediatr Transplant, 2004. 8 Suppl 5: p. 56-66.

29. Weichselbaum, R.R. and S. Hellman, Oligometastases revisited. Nat Rev Clin Oncol, 2011. 8(6): p. 378-82.

30. Cai, J.Y., et al., Minimal residual disease is a prognostic marker for neuroblastoma with bone marrow infiltration. Am J Clin Oncol, 2012. 35(3): p. 275-8.

31. Schleiermacher, G., et al., Increased risk of systemic relapses associated with bone marrow micrometastasis and circulating tumor cells in localized ewing tumor. J Clin Oncol, 2003. 21(1): p. 85-91.

32. Gallego, S., et al., Detection of bone marrow micrometastasis and microcirculating disease in rhabdomyosarcoma by a real-time RT-PCR assay. J Cancer Res Clin Oncol, 2006. 132(6): p. 356-62.

33. Borowitz, M.J., et al., Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia and its relationship to other prognostic factors: a Children's Oncology Group study. Blood, 2008. 111(12): p. 5477-85.

34. Coustan-Smith, E., et al., Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia after first relapse. Leukemia, 2004. 18(3): p. 499-504.

35. Yu, A.L., et al., Anti-GD2 antibody with GM-CSF, interleukin-2, and isotretinoin for neuroblastoma. N Engl J Med, 2010. 363(14): p. 1324-34.

36. Klingebiel, T., et al., Treatment of children with metastatic soft tissue sarcoma with oral maintenance compared to high dose chemotherapy: report of the HD CWS-96 trial. Pediatr Blood Cancer, 2008. 50(4): p. 739-45.

37. Casanova, M., et al., Vinorelbine and low-dose cyclophosphamide in the treatment of pediatric sarcomas: pilot study for the upcoming European Rhabdomyosarcoma Protocol. Cancer, 2004. 101(7): p. 1664-71.

38. Weis, S.M. and D.A. Cheresh, Tumor angiogenesis: molecular pathways and therapeutic targets. Nat Med, 2011. 17(11): p. 1359-70.

39. Folkman, J., Tumor angiogenesis: therapeutic implications. N Engl J Med, 1971. 285(21): p. 1182-6.

40. Hanahan, D. and R.A. Weinberg, The hallmarks of cancer. Cell, 2000. 100(1): p. 57-70.

41. Ferrara, N. and R.S. Kerbel, Angiogenesis as a therapeutic target. Nature, 2005. 438(7070): p. 967-74.

178

42. Asahara, T., et al., Isolation of putative progenitor endothelial cells for angiogenesis. Science, 1997. 275(5302): p. 964-7.

43. Song, S., et al., PDGFRbeta+ perivascular progenitor cells in tumours regulate pericyte differentiation and vascular survival. Nat Cell Biol, 2005. 7(9): p. 870-9.

44. Carmeliet, P. and R.K. Jain, Molecular mechanisms and clinical applications of angiogenesis. Nature, 2011. 473(7347): p. 298-307.

45. Salceda, S. and J. Caro, Hypoxia-inducible factor 1alpha (HIF-1alpha) protein is rapidly degraded by the ubiquitin-proteasome system under normoxic conditions. Its stabilization by hypoxia depends on redox-induced changes. J Biol Chem, 1997. 272(36): p. 22642-7.

46. Ke, Q. and M. Costa, Hypoxia-inducible factor-1 (HIF-1). Mol Pharmacol, 2006. 70(5): p. 1469-80.

47. Ivan, M., et al., HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science, 2001. 292(5516): p. 464-8.

48. Liao, D. and R.S. Johnson, Hypoxia: a key regulator of angiogenesis in cancer. Cancer Metastasis Rev, 2007. 26(2): p. 281-90.

49. Forsythe, J.A., et al., Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol, 1996. 16(9): p. 4604- 13.

50. Ferrara, N., H.P. Gerber, and J. LeCouter, The biology of VEGF and its receptors. Nat Med, 2003. 9(6): p. 669-76.

51. Senger, D.R., et al., Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science, 1983. 219(4587): p. 983-5.

52. Ferrara, N., Vascular endothelial growth factor. Arterioscler Thromb Vasc Biol, 2009. 29(6): p. 789-91.

53. Gerhardt, H., et al., VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol, 2003. 161(6): p. 1163-77.

54. Cascone, T. and J.V. Heymach, Targeting the angiopoietin/Tie2 pathway: cutting tumor vessels with a double-edged sword? J Clin Oncol, 2012. 30(4): p. 441-4.

55. Holmes, K., et al., Vascular endothelial growth factor receptor-2: structure, function, intracellular signalling and therapeutic inhibition. Cell Signal, 2007. 19(10): p. 2003-12.

56. Karar, J. and A. Maity, PI3K/AKT/mTOR Pathway in Angiogenesis. Front Mol Neurosci, 2011. 4: p. 51.

179

57. Rundhaug, J.E., Matrix metalloproteinases and angiogenesis. J Cell Mol Med, 2005. 9(2): p. 267-85.

58. Stetler-Stevenson, W.G., Matrix metalloproteinases in angiogenesis: a moving target for therapeutic intervention. J Clin Invest, 1999. 103(9): p. 1237-41.

59. Li, B., et al., VEGF and PlGF promote adult vasculogenesis by enhancing EPC recruitment and vessel formation at the site of tumor neovascularization. FASEB J, 2006. 20(9): p. 1495-7.

60. Heissig, B., et al., Recruitment of stem and progenitor cells from the bone marrow niche requires MMP-9 mediated release of kit-ligand. Cell, 2002. 109(5): p. 625-37.

61. Das, B., et al., Hypoxia enhances tumor stemness by increasing the invasive and tumorigenic side population fraction. Stem Cells, 2008. 26(7): p. 1818-30.

62. Folkman, J., Role of angiogenesis in tumor growth and metastasis. Semin Oncol, 2002. 29(6 Suppl 16): p. 15-8.

63. Naumov, G.N., L.A. Akslen, and J. Folkman, Role of angiogenesis in human tumor dormancy: animal models of the angiogenic switch. Cell Cycle, 2006. 5(16): p. 1779-87.

64. Craft, P.S. and A.L. Harris, Clinical prognostic significance of tumour angiogenesis. Ann Oncol, 1994. 5(4): p. 305-11.

65. Schneider, P., et al., What role for angiogenesis in childhood acute lymphoblastic leukaemia? Adv Hematol, 2011. 2011: p. 274628.

66. Kieran, M.W., Anti-angiogenic therapy in pediatric neuro-oncology. J Neurooncol, 2005. 75(3): p. 327-34.

67. Meitar, D., et al., Tumor angiogenesis correlates with metastatic disease, N-myc amplification, and poor outcome in human neuroblastoma. J Clin Oncol, 1996. 14(2): p. 405-14.

68. Sugiura, Y., et al., Matrix metalloproteinases-2 and -9 are expressed in human neuroblastoma: contribution of stromal cells to their production and correlation with metastasis. Cancer Res, 1998. 58(10): p. 2209-16.

69. Ribatti, D., et al., Angiogenesis extent and expression of matrix metalloproteinase-2 and -9 correlate with progression in human neuroblastoma. Life Sci, 2001. 68(10): p. 1161-8.

70. Eggert, A., et al., High-level expression of angiogenic factors is associated with advanced tumor stage in human neuroblastomas. Clin Cancer Res, 2000. 6(5): p. 1900-8.

180

71. Meister, B., et al., Expression of vascular endothelial growth factor (VEGF) and its receptors in human neuroblastoma. Eur J Cancer, 1999. 35(3): p. 445-9.

72. Fakhari, M., et al., Upregulation of vascular endothelial growth factor receptors is associated with advanced neuroblastoma. J Pediatr Surg, 2002. 37(4): p. 582-7.

73. Rössler, J., et al., Vascular endothelial growth factor expression in human neuroblastoma: up-regulation by hypoxia. Int J Cancer, 1999. 81(1): p. 113-7.

74. Bäckman, U., A. Svensson, and R. Christofferson, Importance of vascular endothelial growth factor A in the progression of experimental neuroblastoma. Angiogenesis, 2002. 5(4): p. 267-74.

75. Ribatti, D., et al., Angiogenesis in neuroblastoma. Ann N Y Acad Sci, 2004. 1028: p. 133-42.

76. Das, B., et al., A hypoxia-driven vascular endothelial growth factor/Flt1 autocrine loop interacts with hypoxia-inducible factor-1alpha through mitogen-activated protein kinase/extracellular signal-regulated kinase 1/2 pathway in neuroblastoma. Cancer Res, 2005. 65(16): p. 7267-75.

77. Sköldenberg, E.G., et al., The angiogenic growth factors HGF and VEGF in serum and plasma from neuroblastoma patients. Anticancer Res, 2009. 29(8): p. 3311-9.

78. Yudoh, K., et al., Concentration of vascular endothelial growth factor in the tumour tissue as a prognostic factor of soft tissue sarcomas. Br J Cancer, 2001. 84(12): p. 1610-5.

79. Diniz, G., et al., Alternative prognostic factors in pediatric embryonal rhabdomyosarcoma: Nm23 expression, proliferative activity and angiogenesis. Turk J Pediatr, 2004. 46(3): p. 239-44.

80. Gee, M.F., et al., Vascular endothelial growth factor acts in an autocrine manner in rhabdomyosarcoma cell lines and can be inhibited with all-trans-retinoic acid. Oncogene, 2005. 24(54): p. 8025-37.

81. Pavlakovic, H., W. Havers, and L. Schweigerer, Multiple angiogenesis stimulators in a single malignancy: implications for anti-angiogenic tumour therapy. Angiogenesis, 2001. 4(4): p. 259-62.

82. Holzer, G., et al., Concentration of vascular endothelial growth factor (VEGF) in the serum of patients with malignant bone tumors. Med Pediatr Oncol, 2001. 36(6): p. 601-4.

83. van der Schaft, D.W., et al., Tumor cell plasticity in Ewing sarcoma, an alternative circulatory system stimulated by hypoxia. Cancer Res, 2005. 65(24): p. 11520-8.

181

84. Lee, T.H., et al., Production of VEGF165 by Ewing's sarcoma cells induces vasculogenesis and the incorporation of CD34+ stem cells into the expanding tumor vasculature. Int J Cancer, 2006. 119(4): p. 839-46.

85. Zhou, Z., et al., VEGF(165), but not VEGF(189), stimulates vasculogenesis and bone marrow cell migration into Ewing's sarcoma tumors in vivo. Mol Cancer Res, 2007. 5(11): p. 1125-32.

86. Reddy, K., et al., Stromal cell-derived factor-1 stimulates vasculogenesis and enhances Ewing's sarcoma tumor growth in the absence of vascular endothelial growth factor. Int J Cancer, 2008. 123(4): p. 831-7.

87. Zhou, Z., et al., Suppression of Ewing's sarcoma tumor growth, tumor vessel formation, and vasculogenesis following anti vascular endothelial growth factor receptor-2 therapy. Clin Cancer Res, 2007. 13(16): p. 4867-73.

88. Kreuter, M., et al., Prognostic relevance of increased angiogenesis in osteosarcoma. Clin Cancer Res, 2004. 10(24): p. 8531-7.

89. Mikulić, D., et al., Tumor angiogenesis and outcome in osteosarcoma. Pediatr Hematol Oncol, 2004. 21(7): p. 611-9.

90. Mantadakis, E., et al., Lack of prognostic significance of intratumoral angiogenesis in nonmetastatic osteosarcoma. J Pediatr Hematol Oncol, 2001. 23(5): p. 286-9.

91. Hardee, M.E. and D. Zagzag, Mechanisms of glioma-associated neovascularization. Am J Pathol, 2012. 181(4): p. 1126-41.

92. Robles Irizarry, L., et al., Therapeutic targeting of VEGF in the treatment of glioblastoma. Expert Opin Ther Targets, 2012. 16(10): p. 973-84.

93. Kaur, B., et al., Hypoxia and the hypoxia-inducible-factor pathway in glioma growth and angiogenesis. Neuro Oncol, 2005. 7(2): p. 134-53.

94. Bartels, U., et al., Vascularity and angiogenesis as predictors of growth in optic pathway/hypothalamic gliomas. J Neurosurg, 2006. 104(5 Suppl): p. 314-20.

95. Bao, S., et al., Stem cell-like glioma cells promote tumor angiogenesis through vascular endothelial growth factor. Cancer Res, 2006. 66(16): p. 7843-8.

96. Ribatti, D., Judah Folkman, a pioneer in the study of angiogenesis. Angiogenesis, 2008. 11(1): p. 3-10.

97. Gasparini, G., The rationale and future potential of angiogenesis inhibitors in neoplasia. Drugs, 1999. 58(1): p. 17-38.

182

98. Klement, G., et al., Continuous low-dose therapy with vinblastine and VEGF receptor-2 antibody induces sustained tumor regression without overt toxicity. J Clin Invest, 2000. 105(8): p. R15-24.

99. Browder, T., et al., Antiangiogenic scheduling of chemotherapy improves efficacy against experimental drug-resistant cancer. Cancer Res, 2000. 60(7): p. 1878-86.

100. Wilson, P.M., M.J. LaBonte, and H.J. Lenz, Assessing the in vivo efficacy of biologic antiangiogenic therapies. Cancer Chemother Pharmacol, 2013. 71(1): p. 1-12.

101. Arvizo, R.R., et al., Mechanism of anti-angiogenic property of gold nanoparticles: role of nanoparticle size and surface charge. Nanomedicine, 2011. 7(5): p. 580- 7.

102. Todorova, M., et al., Antitumor effects of combining metronomic chemotherapy with the antivascular action of ultrasound stimulated microbubbles. Int J Cancer, 2013. 132(12): p. 2956-66.

103. Petrillo, M., G. Scambia, and G. Ferrandina, Novel targets for VEGF- independent anti-angiogenic drugs. Expert Opin Investig Drugs, 2012. 21(4): p. 451-72.

104. Kaya, M., et al., Concomitant tumour resistance in patients with osteosarcoma. A clue to a new therapeutic strategy. J Bone Joint Surg Br, 2004. 86(1): p. 143-7.

105. Kaya, M., et al., Prevention of postoperative progression of pulmonary metastases in osteosarcoma by antiangiogenic therapy using endostatin. J Orthop Sci, 2007. 12(6): p. 562-7.

106. Gaballah, K., et al., The antiangiogenic agent ZD4190 prevents tumour outgrowth in a model of minimal residual carcinoma in deep tissues. Br J Cancer, 2009. 101(3): p. 418-23.

107. Sims, T.L., et al., Bevacizumab suppresses neuroblastoma progression in the setting of minimal disease. Surgery, 2008. 144(2): p. 269-75.

108. Rofstad, E.K., et al., Antiangiogenic treatment with thrombospondin-1 enhances primary tumor radiation response and prevents growth of dormant pulmonary micrometastases after curative radiation therapy in human melanoma xenografts. Cancer Res, 2003. 63(14): p. 4055-61.

109. Kerbel, R.S., Tumor angiogenesis: past, present and the near future. Carcinogenesis, 2000. 21(3): p. 505-15.

110. Gordon, M.S., D.S. Mendelson, and G. Kato, Tumor angiogenesis and novel antiangiogenic strategies. Int J Cancer, 2010. 126(8): p. 1777-87.

183

111. Cook, K.M. and W.D. Figg, Angiogenesis inhibitors: current strategies and future prospects. CA Cancer J Clin, 2010. 60(4): p. 222-43.

112. Gately, S. and W.W. Li, Multiple roles of COX-2 in tumor angiogenesis: a target for antiangiogenic therapy. Semin Oncol, 2004. 31(2 Suppl 7): p. 2-11.

113. Oudard, S., et al., Everolimus (RAD001): an mTOR inhibitor for the treatment of metastatic renal cell carcinoma. Expert Rev Anticancer Ther, 2009. 9(6): p. 705- 17.

114. Del Bufalo, D., et al., Antiangiogenic potential of the Mammalian target of rapamycin inhibitor temsirolimus. Cancer Res, 2006. 66(11): p. 5549-54.

115. Jain, R.K., Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science, 2005. 307(5706): p. 58-62.

116. Kerbel, R.S., Tumor angiogenesis. N Engl J Med, 2008. 358(19): p. 2039-49.

117. Gan, H.K., B. Seruga, and J.J. Knox, Sunitinib in solid tumors. Expert Opin Investig Drugs, 2009. 18(6): p. 821-34.

118. Burstein, H.J., et al., Phase II study of sunitinib malate, an oral multitargeted tyrosine kinase inhibitor, in patients with metastatic breast cancer previously treated with an anthracycline and a taxane. J Clin Oncol, 2008. 26(11): p. 1810- 6.

119. Socinski, M.A., et al., Multicenter, phase II trial of sunitinib in previously treated, advanced non-small-cell lung cancer. J Clin Oncol, 2008. 26(4): p. 650-6.

120. Demetri, G.D., et al., Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib: a randomised controlled trial. Lancet, 2006. 368(9544): p. 1329-38.

121. Wilhelm, S.M., et al., Preclinical overview of sorafenib, a multikinase inhibitor that targets both Raf and VEGF and PDGF receptor tyrosine kinase signaling. Mol Cancer Ther, 2008. 7(10): p. 3129-40.

122. Figlin, R., C. Sternberg, and C.G. Wood, Novel agents and approaches for advanced renal cell carcinoma. J Urol, 2012. 188(3): p. 707-15.

123. Llovet, J.M., et al., Sorafenib in advanced hepatocellular carcinoma. N Engl J Med, 2008. 359(4): p. 378-90.

124. Sonpavde, G. and T.E. Hutson, Pazopanib: a novel multitargeted tyrosine kinase inhibitor. Curr Oncol Rep, 2007. 9(2): p. 115-9.

125. Hamberg, P., J. Verweij, and S. Sleijfer, (Pre-)clinical pharmacology and activity of pazopanib, a novel multikinase angiogenesis inhibitor. Oncologist, 2010. 15(6): p. 539-47.

184

126. Sternberg, C.N., et al., Pazopanib in locally advanced or metastatic renal cell carcinoma: results of a randomized phase III trial. J Clin Oncol, 2010. 28(6): p. 1061-8.

127. Kessler, E.R., et al., Axitinib, a new therapeutic option in renal cell carcinoma. Drugs Today (Barc), 2012. 48(10): p. 633-44.

128. Grabowski, P., et al., Vandetanib therapy in medullary thyroid cancer. Drugs Today (Barc), 2012. 48(11): p. 723-33.

129. Sahade, M., F. Caparelli, and P.M. Hoff, Cediranib: a VEGF receptor tyrosine kinase inhibitor. Future Oncol, 2012. 8(7): p. 775-81.

130. Ye, C., et al., Distribution, metabolism, and excretion of the anti-angiogenic compound SU5416. Toxicol In Vitro, 2006. 20(2): p. 154-62.

131. Glade Bender, J.L., et al., Phase I trial and pharmacokinetic study of bevacizumab in pediatric patients with refractory solid tumors: a Children's Oncology Group Study. J Clin Oncol, 2008. 26(3): p. 399-405.

132. Narayana, A., et al., Bevacizumab in recurrent high-grade pediatric gliomas. Neuro Oncol, 2010. 12(9): p. 985-90.

133. Hwang, E.I., et al., Long-term efficacy and toxicity of bevacizumab-based therapy in children with recurrent low-grade gliomas. Pediatr Blood Cancer, 2013. 60(5): p. 776-82.

134. Reismüller, B., et al., Feasibility and tolerability of bevacizumab in children with primary CNS tumors. Pediatr Blood Cancer, 2010. 54(5): p. 681-6.

135. Dubois, S.G., et al., Phase I and pharmacokinetic study of sunitinib in pediatric patients with refractory solid tumors: a children's oncology group study. Clin Cancer Res, 2011. 17(15): p. 5113-22.

136. DuBois, S.G., et al., Tolerability and pharmacokinetic profile of a sunitinib powder formulation in pediatric patients with refractory solid tumors: a Children's Oncology Group study. Cancer Chemother Pharmacol, 2012. 69(4): p. 1021-7.

137. Widemann, B.C., et al., A phase I trial and pharmacokinetic study of sorafenib in children with refractory solid tumors or leukemias: a Children's Oncology Group Phase I Consortium report. Clin Cancer Res, 2012. 18(21): p. 6011-22.

138. Glade Bender, J., et al., A phase I trial and pharmacokinetic study of aflibercept (VEGF Trap) in children with refractory solid tumors: a children's oncology group phase I consortium report. Clin Cancer Res, 2012. 18(18): p. 5081-9.

139. Keir, S.T., et al., Initial testing of the multitargeted kinase inhibitor pazopanib by the Pediatric Preclinical Testing Program. Pediatr Blood Cancer, 2012. 59(3): p. 586-8.

185

140. Stempak, D., D. Seely, and S. Baruchel, Metronomic dosing of chemotherapy: applications in pediatric oncology. Cancer Invest, 2006. 24(4): p. 432-43.

141. Duration and intensity of maintenance chemotherapy in acute lymphoblastic leukaemia: overview of 42 trials involving 12 000 randomised children. Childhood ALL Collaborative Group. Lancet, 1996. 347(9018): p. 1783-8.

142. Rubie, H., et al., Localised and unresectable neuroblastoma in infants: excellent outcome with low-dose primary chemotherapy. Br J Cancer, 2003. 89(9): p. 1605-9.

143. Houghton, P.J., et al., Efficacy of topoisomerase I inhibitors, topotecan and irinotecan, administered at low dose levels in protracted schedules to mice bearing xenografts of human tumors. Cancer Chemother Pharmacol, 1995. 36(5): p. 393-403.

144. Santana, V.M., et al., A pilot study of protracted topotecan dosing using a pharmacokinetically guided dosing approach in children with solid tumors. Clin Cancer Res, 2003. 9(2): p. 633-40.

145. Stewart, C.F., et al., Results of a phase II upfront window of pharmacokinetically guided topotecan in high-risk medulloblastoma and supratentorial primitive neuroectodermal tumor. J Clin Oncol, 2004. 22(16): p. 3357-65.

146. Santana, V.M., et al., Improved response in high-risk neuroblastoma with protracted topotecan administration using a pharmacokinetically guided dosing approach. J Clin Oncol, 2005. 23(18): p. 4039-47.

147. Pasquier, E., et al., Concentration- and schedule-dependent effects of chemotherapy on the angiogenic potential and drug sensitivity of vascular endothelial cells. Angiogenesis, 2013. 16(2): p. 373-86.

148. Ghiringhelli, F., et al., Metronomic cyclophosphamide regimen selectively depletes CD4+CD25+ regulatory T cells and restores T and NK effector functions in end stage cancer patients. Cancer Immunol Immunother, 2007. 56(5): p. 641-8.

149. Kosmaczewska, A., et al., The significance of Treg cells in defective tumor immunity. Arch Immunol Ther Exp (Warsz), 2008. 56(3): p. 181-91.

150. Loeffler, M., J.A. Krüger, and R.A. Reisfeld, Immunostimulatory effects of low- dose cyclophosphamide are controlled by inducible nitric oxide synthase. Cancer Res, 2005. 65(12): p. 5027-30.

151. Banissi, C., et al., Treg depletion with a low-dose metronomic temozolomide regimen in a rat glioma model. Cancer Immunol Immunother, 2009. 58(10): p. 1627-34.

186

152. Tanaka, H., et al., Classification of chemotherapeutic agents based on their differential in vitro effects on dendritic cells. Cancer Res, 2009. 69(17): p. 6978- 86.

153. Tanaka, H., et al., Dual therapeutic efficacy of vinblastine as a unique chemotherapeutic agent capable of inducing dendritic cell maturation. Cancer Res, 2009. 69(17): p. 6987-94.

154. Stoelting, S., et al., Low-dose oral metronomic chemotherapy prevents mobilization of endothelial progenitor cells into the blood of cancer patients. In Vivo, 2008. 22(6): p. 831-6.

155. Bertolini, F., et al., Maximum tolerable dose and low-dose metronomic chemotherapy have opposite effects on the mobilization and viability of circulating endothelial progenitor cells. Cancer Res, 2003. 63(15): p. 4342-6.

156. Murakami, H., et al., Circulating endothelial progenitor cells in metronomic chemotherapy using irinotecan and/or bevacizumab for colon carcinoma: Study of their clinical significance. Exp Ther Med, 2011. 2(4): p. 595-600.

157. Bocci, G., et al., Thrombospondin 1, a mediator of the antiangiogenic effects of low-dose metronomic chemotherapy. Proc Natl Acad Sci U S A, 2003. 100(22): p. 12917-22.

158. Bocci, G., et al., Metronomic ceramide analogs inhibit angiogenesis in pancreatic cancer through up-regulation of caveolin-1 and thrombospondin-1 and downregulation of cyclin D1. Neoplasia, 2012. 14(9): p. 833-45.

159. Nör, J.E., et al., Thrombospondin-1 induces endothelial cell apoptosis and inhibits angiogenesis by activating the caspase death pathway. J Vasc Res, 2000. 37(3): p. 209-18.

160. Emmenegger, U., et al., Tumors that acquire resistance to low-dose metronomic cyclophosphamide retain sensitivity to maximum tolerated dose cyclophosphamide. Neoplasia, 2011. 13(1): p. 40-8.

161. Aljuffali, I.A., et al., Enhanced antitumor activity of low-dose continuous administration schedules of topotecan in prostate cancer. Cancer Biol Ther, 2011. 12(5): p. 407-20.

162. Jang, J.W., et al., Suppression of hepatic tumor growth and metastasis by metronomic therapy in a rat model of hepatocellular carcinoma. Exp Mol Med, 2011. 43(5): p. 305-12.

163. Ko, K.K., et al., Metronomic treatment of temozolomide increases anti- angiogenicity accompanied by down-regulated O(6)-methylguanine-DNA methyltransferase expression in endothelial cells. Exp Ther Med, 2011. 2(2): p. 343-348.

187

164. Wu, H., et al., Metronomic docetaxel chemotherapy inhibits angiogenesis and tumor growth in a gastric cancer model. Cancer Chemother Pharmacol, 2011. 68(4): p. 879-87.

165. Todorova, V.K., Y. Kaufmann, and V.S. Klimberg, Increased efficacy and reduced cardiotoxicity of metronomic treatment with cyclophosphamide in rat breast cancer. Anticancer Res, 2011. 31(1): p. 215-20.

166. Rozados, V.R., et al., The immune response and the therapeutic effect of metronomic chemotherapy with cyclophosphamide. Oncol Res, 2010. 18(11-12): p. 601-5.

167. Panigrahy, D., et al., Inhibition of tumor angiogenesis by oral etoposide. Exp Ther Med, 2010. 1(5): p. 739-746.

168. Zeitlin, B.D., et al., Metronomic small molecule inhibitor of Bcl-2 (TW-37) is antiangiogenic and potentiates the antitumor effect of ionizing radiation. Int J Radiat Oncol Biol Phys, 2010. 78(3): p. 879-87.

169. Cham, K.K., et al., Metronomic gemcitabine suppresses tumour growth, improves perfusion, and reduces hypoxia in human pancreatic ductal adenocarcinoma. Br J Cancer, 2010. 103(1): p. 52-60.

170. Hashimoto, K., et al., Potent preclinical impact of metronomic low-dose oral topotecan combined with the antiangiogenic drug pazopanib for the treatment of ovarian cancer. Mol Cancer Ther, 2010. 9(4): p. 996-1006.

171. Merritt, W.M., et al., Bridging the gap between cytotoxic and biologic therapy with metronomic topotecan and pazopanib in ovarian cancer. Mol Cancer Ther, 2010. 9(4): p. 985-95.

172. Shen, F.Z., et al., Low-dose metronomic chemotherapy with cisplatin: can it suppress angiogenesis in H22 hepatocarcinoma cells? Int J Exp Pathol, 2010. 91(1): p. 10-6.

173. Soffer, S.Z., et al., Novel use of an established agent: Topotecan is antiangiogenic in experimental Wilms tumor. J Pediatr Surg, 2001. 36(12): p. 1781- 4.

174. McCrudden, K.W., et al., Topotecan is anti-angiogenic in experimental hepatoblastoma. J Pediatr Surg, 2002. 37(6): p. 857-61.

175. Zhu, X.Z., H.M. Yin, and J. Mei, Inhibition of tumors cell growth in osteosarcoma- bearing SD rats through a combination of conventional and metronomic scheduling of neoadjuvant chemotherapy. Acta Pharmacol Sin, 2010. 31(8): p. 970-6.

176. Brandi, G., et al., Durable complete response of hepatocellular carcinoma after metronomic capecitabine. Tumori, 2010. 96(6): p. 1028-30.

188

177. Rajdev, L., et al., Phase I trial of metronomic oral vinorelbine in patients with advanced cancer. Cancer Chemother Pharmacol, 2011. 68(5): p. 1119-24.

178. Dantonello, T.M., et al., Embryonal rhabdomyosarcoma with metastases confined to the lungs: report from the CWS Study Group. Pediatr Blood Cancer, 2011. 56(5): p. 725-32.

179. Sumikawa, Y., et al., Long-term metronomic docetaxel chemotherapy for inoperative angiosarcoma of the scalp. J Dermatol, 2011. 38(4): p. 393-5.

180. Ballardini, P., et al., Long-lasting response with metronomic capecitabine in advanced hepatocellular carcinoma. Tumori, 2010. 96(5): p. 768-70.

181. Nelius, T., K. Rinard, and S. Filleur, Oral/metronomic cyclophosphamide-based chemotherapy as option for patients with castration-refractory prostate cancer: review of the literature. Cancer Treat Rev, 2011. 37(6): p. 444-55.

182. He, S., et al., Capecitabine "metronomic" chemotherapy for palliative treatment of elderly patients with advanced gastric cancer after fluoropyrimidine-based chemotherapy. Med Oncol, 2012. 29(1): p. 100-6.

183. Simon, T., et al., Long term outcome of high-risk neuroblastoma patients after immunotherapy with antibody ch14.18 or oral metronomic chemotherapy. BMC Cancer, 2011. 11: p. 21.

184. Fousseyni, T., et al., Children treated with metronomic chemotherapy in a low- income country: METRO-MALI-01. J Pediatr Hematol Oncol, 2011. 33(1): p. 31- 4.

185. Massaccesi, M., et al., Multiple tumor marker elevation in androgen ablation- refractory prostate cancer with long-term response to metronomic chemotherapy: a case report. Int J Biol Markers, 2010. 25(4): p. 243-7.

186. Minturn, J.E., et al., A phase II study of metronomic oral topotecan for recurrent childhood brain tumors. Pediatr Blood Cancer, 2011. 56(1): p. 39-44.

187. Zhao, X., et al., Biomarker alterations with metronomic use of low-dose zoledronic acid for breast cancer patients with bone metastases and potential clinical significance. Breast Cancer Res Treat, 2010. 124(3): p. 733-43.

188. Addeo, R., et al., Low-dose metronomic oral administration of vinorelbine in the first-line treatment of elderly patients with metastatic breast cancer. Clin Breast Cancer, 2010. 10(4): p. 301-6.

189. Greiner, J., et al., Metronomic treatment with low-dose trofosfamide leads to a long-term remission in a patient with docetaxel-refractory advanced metastatic prostate cancer. Case Rep Med, 2010. 2010: p. 395720.

189

190. Penel, N., et al., Megestrol acetate versus metronomic cyclophosphamide in patients having exhausted all effective therapies under standard care. Br J Cancer, 2010. 102(8): p. 1207-12.

191. Italiano, A., et al., "Metronomic" chemotherapy in advanced soft tissue sarcomas. Cancer Chemother Pharmacol, 2010. 66(1): p. 197-202.

192. Kong, D.S., et al., Phase II trial of low-dose continuous (metronomic) treatment of temozolomide for recurrent glioblastoma. Neuro Oncol, 2010. 12(3): p. 289- 96.

193. Visa, L., M. Caballero, and J.J. Grau, [Response to metronomic chemotherapy in a metastatic adenoid cystic carcinoma of the parotid gland]. Acta Otorrinolaringol Esp, 2010. 61(6): p. 452-4.

194. Sharp, J.R., et al., A multi-centre Canadian pilot study of metronomic temozolomide combined with radiotherapy for newly diagnosed paediatric brainstem glioma. Eur J Cancer, 2010. 46(18): p. 3271-9.

195. Baruchel, S., et al., Safety and pharmacokinetics of temozolomide using a dose- escalation, metronomic schedule in recurrent paediatric brain tumours. Eur J Cancer, 2006. 42(14): p. 2335-42.

196. Kerbel, R.S., et al., Possible mechanisms of acquired resistance to antiangiogenic drugs: implications for the use of combination therapy approaches. Cancer Metastasis Rev, 2001. 20(1-2): p. 79-86.

197. Sweeney, C.J., et al., The antiangiogenic property of docetaxel is synergistic with a recombinant humanized monoclonal antibody against vascular endothelial growth factor or 2-methoxyestradiol but antagonized by endothelial growth factors. Cancer Res, 2001. 61(8): p. 3369-72.

198. Tran, J., et al., A role for survivin in chemoresistance of endothelial cells mediated by VEGF. Proc Natl Acad Sci U S A, 2002. 99(7): p. 4349-54.

199. Gerber, H.P., V. Dixit, and N. Ferrara, Vascular endothelial growth factor induces expression of the antiapoptotic proteins Bcl-2 and A1 in vascular endothelial cells. J Biol Chem, 1998. 273(21): p. 13313-6.

200. Gerber, H.P., et al., Vascular endothelial growth factor regulates endothelial cell survival through the phosphatidylinositol 3'-kinase/Akt signal transduction pathway. Requirement for Flk-1/KDR activation. J Biol Chem, 1998. 273(46): p. 30336-43.

201. Mandriota, S.J., et al., Vascular endothelial growth factor increases urokinase receptor expression in vascular endothelial cells. J Biol Chem, 1995. 270(17): p. 9709-16.

190

202. Kakeji, Y. and B.A. Teicher, Preclinical studies of the combination of angiogenic inhibitors with cytotoxic agents. Invest New Drugs, 1997. 15(1): p. 39-48.

203. Teicher, B.A., E.A. Sotomayor, and Z.D. Huang, Antiangiogenic agents potentiate cytotoxic cancer therapies against primary and metastatic disease. Cancer Res, 1992. 52(23): p. 6702-4.

204. Soffer, S.Z., et al., Combination antiangiogenic therapy: increased efficacy in a murine model of Wilms tumor. J Pediatr Surg, 2001. 36(8): p. 1177-81.

205. Kim, E.S., et al., Distinct response of experimental neuroblastoma to combination antiangiogenic strategies. J Pediatr Surg, 2002. 37(3): p. 518-22.

206. Zhang, L., et al., Combined anti-fetal liver kinase 1 monoclonal antibody and continuous low-dose doxorubicin inhibits angiogenesis and growth of human soft tissue sarcoma xenografts by induction of endothelial cell apoptosis. Cancer Res, 2002. 62(7): p. 2034-42.

207. Zhou, F., et al., Metronomic chemotherapy in combination with antiangiogenic treatment induces mosaic vascular reduction and tumor growth inhibition in hepatocellular carcinoma xenografts. J Cancer Res Clin Oncol, 2012. 138(11): p. 1879-90.

208. Abraham, D., et al., Low dose carboplatin combined with angiostatic agents prevents metastasis in human testicular germ cell tumor xenografts. J Urol, 2003. 170(4 Pt 1): p. 1388-93.

209. Svensson, A., et al., CHS 828 inhibits neuroblastoma growth in mice alone and in combination with antiangiogenic drugs. Pediatr Res, 2002. 51(5): p. 607-11.

210. Foy, K.C., et al., Immunotherapy with HER-2 and VEGF peptide mimics plus metronomic paclitaxel causes superior antineoplastic effects in transplantable and transgenic mouse models of human breast cancer. Oncoimmunology, 2012. 1(7): p. 1004-1016.

211. Malvicini, M., et al., Single low-dose cyclophosphamide combined with interleukin-12 gene therapy is superior to a metronomic schedule in inducing immunity against colorectal carcinoma in mice. Oncoimmunology, 2012. 1(7): p. 1038-1047.

212. Jia, L. and D.J. Waxman, Thrombospondin-1 and pigment epithelium-derived factor enhance responsiveness of KM12 colon tumor to metronomic cyclophosphamide but have disparate effects on tumor metastasis. Cancer Lett, 2013. 330(2): p. 241-9.

213. Snoeks, T.J., et al., 2-methoxyestradiol analogue ENMD-1198 reduces breast cancer-induced osteolysis and tumor burden both in vitro and in vivo. Mol Cancer Ther, 2011. 10(5): p. 874-82.

191

214. Wu, H., et al., Capecitabine combined with (-)-epigallocatechin-3-gallate inhibits angiogenesis and tumor growth in nude mice with gastric cancer xenografts. Exp Ther Med, 2012. 3(4): p. 650-654.

215. Adhim, Z., et al., E10A, an adenovirus-carrying endostatin gene, dramatically increased the tumor drug concentration of metronomic chemotherapy with low- dose cisplatin in a xenograft mouse model for head and neck squamous-cell carcinoma. Cancer Gene Ther, 2012. 19(2): p. 144-52.

216. Naganuma, Y., et al., Metronomic doxifluridine chemotherapy combined with the anti-angiogenic agent TNP-470 inhibits the growth of human uterine carcinosarcoma xenografts. Cancer Sci, 2011. 102(8): p. 1545-52.

217. Iwamoto, H., et al., Metronomic S-1 chemotherapy and vandetanib: an efficacious and nontoxic treatment for hepatocellular carcinoma. Neoplasia, 2011. 13(3): p. 187-97.

218. Tang, T.C., et al., Development of a resistance-like phenotype to sorafenib by human hepatocellular carcinoma cells is reversible and can be delayed by metronomic UFT chemotherapy. Neoplasia, 2010. 12(11): p. 928-40.

219. Tran Cao, H.S., et al., Metronomic gemcitabine in combination with sunitinib inhibits multisite metastasis and increases survival in an orthotopic model of pancreatic cancer. Mol Cancer Ther, 2010. 9(7): p. 2068-78.

220. Mainetti, L.E., et al., Antitumoral and antimetastatic effects of metronomic chemotherapy with cyclophosphamide combined with celecoxib on murine mammary adenocarcinomas. J Cancer Res Clin Oncol, 2011. 137(1): p. 151-63.

221. Daenen, L.G., et al., Low-dose metronomic cyclophosphamide combined with vascular disrupting therapy induces potent antitumor activity in preclinical human tumor xenograft models. Mol Cancer Ther, 2009. 8(10): p. 2872-81.

222. Ma, J. and D.J. Waxman, Modulation of the antitumor activity of metronomic cyclophosphamide by the angiogenesis inhibitor axitinib. Mol Cancer Ther, 2008. 7(1): p. 79-89.

223. Lorgis, V., O. Varbedian, and F. Ghiringhelli, Metronomic cyclophosphamide with cisplatin and bevacizumab: a new chemotherapeutic regimen for refractory anaplastic ependymoma. Anticancer Res, 2012. 32(11): p. 5067-70.

224. Mayer, E.L., et al., Combination antiangiogenic therapy in advanced breast cancer: a phase 1 trial of vandetanib, a VEGFR inhibitor, and metronomic chemotherapy, with correlative platelet proteomics. Breast Cancer Res Treat, 2012. 136(1): p. 169-78.

225. Montagna, E., et al., Metronomic chemotherapy combined with bevacizumab and erlotinib in patients with metastatic HER2-negative breast cancer: clinical and biological activity. Clin Breast Cancer, 2012. 12(3): p. 207-14.

192

226. Dellapasqua, S., et al., Increased mean corpuscular volume of red blood cells predicts response to metronomic capecitabine and cyclophosphamide in combination with bevacizumab. Breast, 2012. 21(3): p. 309-13.

227. Kummar, S., et al., A phase I study of veliparib in combination with metronomic cyclophosphamide in adults with refractory solid tumors and lymphomas. Clin Cancer Res, 2012. 18(6): p. 1726-34.

228. Koumarianou, A., et al., Combination treatment with metronomic temozolomide, bevacizumab and long-acting octreotide for malignant neuroendocrine tumours. Endocr Relat Cancer, 2012. 19(1): p. L1-4.

229. Eichbaum, M., et al., The PACOVAR-trial: a phase I/II study of pazopanib (GW786034) and cyclophosphamide in patients with platinum-resistant recurrent, pre-treated ovarian cancer. BMC Cancer, 2011. 11: p. 453.

230. Aigner, J., et al., Long-term remission in a patient with heavily pretreated, advanced ovarian cancer achieved by bevacizumab and metronomic cyclophosphamide treatment. Anticancer Drugs, 2011. 22(10): p. 1030-3.

231. Saloustros, E., et al., Metronomic vinorelbine plus bevacizumab as salvage therapy for patients with metastatic breast cancer. J BUON, 2011. 16(2): p. 215- 8.

232. Chen, Y.M., et al., A phase II randomized trial of gefitinib alone or with tegafur/uracil treatment in patients with pulmonary adenocarcinoma who had failed previous chemotherapy. J Thorac Oncol, 2011. 6(6): p. 1110-6.

233. Reinhardt, J., et al., Long-term remission of an advanced recurrent endometrial cancer in a heavily pretreated patient using a combined regimen with bevacizumab and metronomic cyclophosphamide. Anticancer Drugs, 2011. 22(8): p. 822-4.

234. Sánchez-Muñoz, A., et al., Bevacizumab plus low-dose metronomic oral cyclophosphamide in heavily pretreated patients with recurrent ovarian cancer. Oncology, 2010. 79(1-2): p. 98-104.

235. Hsu, C.H., et al., Phase II study of combining sorafenib with metronomic tegafur/uracil for advanced hepatocellular carcinoma. J Hepatol, 2010. 53(1): p. 126-31.

236. Buckstein, R., et al., High-Dose celecoxib and metronomic "low-dose" cyclophosphamide is an effective and safe therapy in patients with relapsed and refractory aggressive histology non-Hodgkin's lymphoma. Clin Cancer Res, 2006. 12(17): p. 5190-8.

237. Stempak, D., et al., A pilot pharmacokinetic and antiangiogenic biomarker study of celecoxib and low-dose metronomic vinblastine or cyclophosphamide in

193

pediatric recurrent solid tumors. J Pediatr Hematol Oncol, 2006. 28(11): p. 720- 8.

238. Russell, H.V., et al., A phase I study of zoledronic acid and low-dose cyclophosphamide in recurrent/refractory neuroblastoma: a new approaches to neuroblastoma therapy (NANT) study. Pediatr Blood Cancer, 2011. 57(2): p. 275-82.

239. André, N., et al., Pilot study of a pediatric metronomic 4-drug regimen. Oncotarget, 2011. 2(12): p. 960-5.

240. Kieran, M.W., et al., A feasibility trial of antiangiogenic (metronomic) chemotherapy in pediatric patients with recurrent or progressive cancer. J Pediatr Hematol Oncol, 2005. 27(11): p. 573-81.

241. Choi, L.M., et al., Feasibility of metronomic maintenance chemotherapy following high-dose chemotherapy for malignant central nervous system tumors. Pediatr Blood Cancer, 2008. 50(5): p. 970-5.

242. Sterba, J., et al., Combined biodifferentiating and antiangiogenic oral metronomic therapy is feasible and effective in relapsed solid tumors in children: single-center pilot study. Onkologie, 2006. 29(7): p. 308-13.

243. Zapletalova, D., et al., Metronomic chemotherapy with the COMBAT regimen in advanced pediatric malignancies: a multicenter experience. Oncology, 2012. 82(5): p. 249-60.

244. Sondhi, V., et al., Complete remission and long-term survival in a child with relapsed medulloblastoma with extensive osteosclerotic bony metastasis with a novel metronomic chemobiological approach. J Pediatr Hematol Oncol, 2012. 34(5): p. e195-8.

245. Peyrl, A., et al., Antiangiogenic metronomic therapy for children with recurrent embryonal brain tumors. Pediatr Blood Cancer, 2012. 59(3): p. 511-7.

246. André, N., et al., Metronomic etoposide/cyclophosphamide/celecoxib regimen given to children and adolescents with refractory cancer: a preliminary monocentric study. Clin Ther, 2008. 30(7): p. 1336-40.

247. Padovani, L., et al., Reirradiation and concomitant metronomic temozolomide: an efficient combination for local control in medulloblastoma disease? J Pediatr Hematol Oncol, 2011. 33(8): p. 600-4.

248. Chu, T.F., et al., Cardiotoxicity associated with tyrosine kinase inhibitor sunitinib. Lancet, 2007. 370(9604): p. 2011-9.

249. Force, T., D.S. Krause, and R.A. Van Etten, Molecular mechanisms of cardiotoxicity of tyrosine kinase inhibition. Nat Rev Cancer, 2007. 7(5): p. 332- 44.

194

250. Chen, H.X. and J.N. Cleck, Adverse effects of anticancer agents that target the VEGF pathway. Nat Rev Clin Oncol, 2009. 6(8): p. 465-77.

251. Reardon, D.A., et al., Phase I study of sunitinib and irinotecan for patients with recurrent malignant glioma. J Neurooncol, 2011. 105(3): p. 621-7.

252. Iwamoto, F.M., et al., Phase II trial of pazopanib (GW786034), an oral multitargeted angiogenesis inhibitor, for adults with recurrent glioblastoma (North American Brain Tumor Consortium Study 06-02). Neuro Oncol, 2010. 12(8): p. 855-61.

253. Jellvert, A., et al., Effective oral combination metronomic chemotherapy with low toxicity for the management of castration-resistant prostate cancer. Exp Ther Med, 2011. 2(4): p. 579-584.

254. Azad, N.S., et al., Combination targeted therapy with sorafenib and bevacizumab results in enhanced toxicity and antitumor activity. J Clin Oncol, 2008. 26(22): p. 3709-14.

255. Batchelor, T.T., et al., AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell, 2007. 11(1): p. 83-95.

256. Burris, H.A., et al., Phase I study of pazopanib in combination with paclitaxel and carboplatin given every 21 days in patients with advanced solid tumors. Mol Cancer Ther, 2012. 11(8): p. 1820-8.

257. Kollmannsberger, C., et al., Topotecan - A novel topoisomerase I inhibitor: pharmacology and clinical experience. Oncology, 1999. 56(1): p. 1-12.

258. Kruijtzer, C.M., et al., Increased oral bioavailability of topotecan in combination with the breast cancer resistance protein and P-glycoprotein inhibitor GF120918. J Clin Oncol, 2002. 20(13): p. 2943-50.

259. Zamboni, W.C., et al., Phenytoin alters the disposition of topotecan and N- desmethyl topotecan in a patient with medulloblastoma. Clin Cancer Res, 1998. 4(3): p. 783-9.

260. Zamboni, W.C., et al., Pharmacokinetic and pharmacodynamic study of the combination of docetaxel and topotecan in patients with solid tumors. J Clin Oncol, 2000. 18(18): p. 3288-94.

261. Herben, V.M., et al., Oral topotecan: bioavailablity and effect of food co- administration. Br J Cancer, 1999. 80(9): p. 1380-6.

262. Schellens, J.H., et al., Bioavailability and pharmacokinetics of oral topotecan: a new topoisomerase I inhibitor. Br J Cancer, 1996. 73(10): p. 1268-71.

195

263. Jia, P., et al., Breast cancer resistance protein-mediated topotecan resistance in ovarian cancer cells. Int J Gynecol Cancer, 2005. 15(6): p. 1042-8.

264. Zander, S.A., et al., Sensitivity and acquired resistance of BRCA1;p53-deficient mouse mammary tumors to the topoisomerase I inhibitor topotecan. Cancer Res, 2010. 70(4): p. 1700-10.

265. Hendricks, C.B., et al., Effect of P-glycoprotein expression on the accumulation and cytotoxicity of topotecan (SK&F 104864), a new camptothecin analogue. Cancer Res, 1992. 52(8): p. 2268-78.

266. Dhooge, C.R., et al., Expression of the MDR1 gene product P-glycoprotein in childhood neuroblastoma. Cancer, 1997. 80(7): p. 1250-7.

267. Hirschmann-Jax, C., et al., A distinct "side population" of cells with high drug efflux capacity in human tumor cells. Proc Natl Acad Sci U S A, 2004. 101(39): p. 14228-33.

268. Carol, H., et al., Initial testing of topotecan by the pediatric preclinical testing program. Pediatr Blood Cancer, 2010. 54(5): p. 707-15.

269. Zamboni, W.C., et al., Relationship between topotecan systemic exposure and tumor response in human neuroblastoma xenografts. J Natl Cancer Inst, 1998. 90(7): p. 505-11.

270. Zamboni, W.C., et al., Relationship between tumor extracellular fluid exposure to topotecan and tumor response in human neuroblastoma xenograft and cell lines. Cancer Chemother Pharmacol, 1999. 43(4): p. 269-76.

271. Metzger, M.L., et al., Topotecan is active against Wilms' tumor: results of a multiinstitutional phase II study. J Clin Oncol, 2007. 25(21): p. 3130-6.

272. Hawkins, D.S., et al., Topotecan by 21-day continuous infusion in children with relapsed or refractory solid tumors: a Children's Oncology Group study. Pediatr Blood Cancer, 2006. 47(6): p. 790-4.

273. Wagner, S., et al., Oral topotecan in children with recurrent or progressive high- grade glioma: a Phase I/II study by the German Society for Pediatric Oncology and Hematology. Cancer, 2004. 100(8): p. 1750-7.

274. Daw, N.C., et al., Phase I and pharmacokinetic study of topotecan administered orally once daily for 5 days for 2 consecutive weeks to pediatric patients with refractory solid tumors. J Clin Oncol, 2004. 22(5): p. 829-37.

275. Furman, W.L., et al., Protracted intermittent schedule of topotecan in children with refractory acute leukemia: a pediatric oncology group study. J Clin Oncol, 2002. 20(6): p. 1617-24.

196

276. Clements, M.K., et al., Antiangiogenic potential of camptothecin and topotecan. Cancer Chemother Pharmacol, 1999. 44(5): p. 411-6.

277. O'Leary, J.J., et al., Antiangiogenic effects of camptothecin analogues 9-amino- 20(S)-camptothecin, topotecan, and CPT-11 studied in the mouse cornea model. Clin Cancer Res, 1999. 5(1): p. 181-7.

278. Merritt, W.M., et al., Anti-angiogenic properties of metronomic topotecan in ovarian carcinoma. Cancer Biol Ther, 2009. 8(16): p. 1596-603.

279. Beppu, K., et al., Topotecan blocks hypoxia-inducible factor-1alpha and vascular endothelial growth factor expression induced by insulin-like growth factor-I in neuroblastoma cells. Cancer Res, 2005. 65(11): p. 4775-81.

280. Puppo, M., et al., Topotecan inhibits vascular endothelial growth factor production and angiogenic activity induced by hypoxia in human neuroblastoma by targeting hypoxia-inducible factor-1alpha and -2alpha. Mol Cancer Ther, 2008. 7(7): p. 1974-84.

281. Rapisarda, A., R.H. Shoemaker, and G. Melillo, Targeting topoisomerase I to inhibit hypoxia inducible factor 1. Cell Cycle, 2004. 3(2): p. 172-5.

282. Kummar, S., et al., Multihistology, target-driven pilot trial of oral topotecan as an inhibitor of hypoxia-inducible factor-1α in advanced solid tumors. Clin Cancer Res, 2011. 17(15): p. 5123-31.

283. Hackl, C., et al., Metronomic oral topotecan prolongs survival and reduces liver metastasis in improved preclinical orthotopic and adjuvant therapy colon cancer models. Gut, 2013. 62(2): p. 259-71.

284. Heath, E.I., et al., A phase I study of the pharmacokinetic and safety profiles of oral pazopanib with a high-fat or low-fat meal in patients with advanced solid tumors. Clin Pharmacol Ther, 2010. 88(6): p. 818-23.

285. Bukowski, R.M., U. Yasothan, and P. Kirkpatrick, Pazopanib. Nat Rev Drug Discov, 2010. 9(1): p. 17-8.

286. Sleijfer, S., et al., Pazopanib, a multikinase angiogenesis inhibitor, in patients with relapsed or refractory advanced soft tissue sarcoma: a phase II study from the European organisation for research and treatment of cancer-soft tissue and bone sarcoma group (EORTC study 62043). J Clin Oncol, 2009. 27(19): p. 3126- 32.

287. Sonpavde, G., T.E. Hutson, and C.N. Sternberg, Pazopanib, a potent orally administered small-molecule multitargeted tyrosine kinase inhibitor for renal cell carcinoma. Expert Opin Investig Drugs, 2008. 17(2): p. 253-61.

197

288. Plummer, R., et al., A Phase I study of pazopanib in combination with gemcitabine in patients with advanced solid tumors. Cancer Chemother Pharmacol, 2013. 71(1): p. 93-101.

289. du Bois, A., et al., Open-label feasibility study of pazopanib, carboplatin, and paclitaxel in women with newly diagnosed, untreated, gynaecologic tumours: a phase I/II trial of the AGO study group. Br J Cancer, 2012. 106(4): p. 629-32.

290. Necchi, A., et al., Pazopanib in advanced and platinum-resistant urothelial cancer: an open-label, single group, phase 2 trial. Lancet Oncol, 2012. 13(8): p. 810-6.

291. Bible, K.C., et al., A multiinstitutional phase 2 trial of pazopanib monotherapy in advanced anaplastic thyroid cancer. J Clin Endocrinol Metab, 2012. 97(9): p. 3179-84.

292. van der Graaf, W.T., et al., Pazopanib for metastatic soft-tissue sarcoma (PALETTE): a randomised, double-blind, placebo-controlled phase 3 trial. Lancet, 2012. 379(9829): p. 1879-86.

293. Ward, J.E., et al., A randomized, phase II study of pazopanib in castrate- sensitive prostate cancer: a University of Chicago Phase II Consortium/Department of Defense Prostate Cancer Clinical Trials Consortium study. Prostate Cancer Prostatic Dis, 2012. 15(1): p. 87-92.

294. Lim, W.T., et al., A Phase II study of pazopanib in Asian patients with recurrent/metastatic nasopharyngeal carcinoma. Clin Cancer Res, 2011. 17(16): p. 5481-9.

295. Yau, T., et al., Phase I dose-finding study of pazopanib in hepatocellular carcinoma: evaluation of early efficacy, pharmacokinetics, and pharmacodynamics. Clin Cancer Res, 2011. 17(21): p. 6914-23.

296. Bible, K.C., et al., Efficacy of pazopanib in progressive, radioiodine-refractory, metastatic differentiated thyroid cancers: results of a phase 2 consortium study. Lancet Oncol, 2010. 11(10): p. 962-72.

297. Taylor, S.K., et al., A phase II study of pazopanib in patients with recurrent or metastatic invasive breast carcinoma: a trial of the Princess Margaret Hospital phase II consortium. Oncologist, 2010. 15(8): p. 810-8.

298. Friedlander, M., et al., A Phase II, open-label study evaluating pazopanib in patients with recurrent ovarian cancer. Gynecol Oncol, 2010. 119(1): p. 32-7.

299. Altorki, N., et al., Phase II proof-of-concept study of pazopanib monotherapy in treatment-naive patients with stage I/II resectable non-small-cell lung cancer. J Clin Oncol, 2010. 28(19): p. 3131-7.

198

300. Prince, H.M., et al., Vascular endothelial growth factor inhibition is not an effective therapeutic strategy for relapsed or refractory multiple myeloma: a phase 2 study of pazopanib (GW786034). Blood, 2009. 113(19): p. 4819-20.

301. Powles, T., et al., An indirect comparison of the toxicity of sunitinib and pazopanib in metastatic clear cell renal cancer. Eur J Cancer, 2012. 48(17): p. 3171-6.

302. Monk, B.J., et al., Phase II, open-label study of pazopanib or lapatinib monotherapy compared with pazopanib plus lapatinib combination therapy in patients with advanced and recurrent cervical cancer. J Clin Oncol, 2010. 28(22): p. 3562-9.

303. Jain, R.K., et al., Biomarkers of response and resistance to antiangiogenic therapy. Nat Rev Clin Oncol, 2009. 6(6): p. 327-38.

304. Sessa, C., et al., Biomarkers of angiogenesis for the development of antiangiogenic therapies in oncology: tools or decorations? Nat Clin Pract Oncol, 2008. 5(7): p. 378-91.

305. Jubb, A.M., et al., Predicting benefit from anti-angiogenic agents in malignancy. Nat Rev Cancer, 2006. 6(8): p. 626-35.

306. DePrimo, S.E. and C. Bello, Surrogate biomarkers in evaluating response to anti-angiogenic agents: focus on sunitinib. Ann Oncol, 2007. 18 Suppl 10: p. x11-9.

307. Duda, D.G., M. Ancukiewicz, and R.K. Jain, Biomarkers of antiangiogenic therapy: how do we move from candidate biomarkers to valid biomarkers? J Clin Oncol, 2010. 28(2): p. 183-5.

308. Rini, B.I., et al., Diastolic blood pressure as a biomarker of axitinib efficacy in solid tumors. Clin Cancer Res, 2011. 17(11): p. 3841-9.

309. Schiavetti, A., et al., Vascular endothelial growth factor serum levels in children with newly diagnosed rhabdomyosarcoma. Pediatr Blood Cancer, 2012. 59(4): p. 627-30.

310. Kontovinis, L.F., et al., Sunitinib treatment for patients with clear-cell metastatic renal cell carcinoma: clinical outcomes and plasma angiogenesis markers. BMC Cancer, 2009. 9: p. 82.

311. Bando, H., et al., Association between intratumoral free and total VEGF, soluble VEGFR-1, VEGFR-2 and prognosis in breast cancer. Br J Cancer, 2005. 92(3): p. 553-61.

312. Yang, J.C., et al., A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N Engl J Med, 2003. 349(5): p. 427-34.

199

313. Harmon, C.S., et al., Mechanism-related circulating proteins as biomarkers for clinical outcome in patients with unresectable hepatocellular carcinoma receiving sunitinib. J Transl Med, 2011. 9: p. 120.

314. Willett, C.G., et al., Surrogate markers for antiangiogenic therapy and dose- limiting toxicities for bevacizumab with radiation and chemotherapy: continued experience of a phase I trial in rectal cancer patients. J Clin Oncol, 2005. 23(31): p. 8136-9.

315. Volm, M., R. Koomägi, and J. Mattern, Prognostic value of vascular endothelial growth factor and its receptor Flt-1 in squamous cell lung cancer. Int J Cancer, 1997. 74(1): p. 64-8.

316. Motzer, R.J., et al., Activity of SU11248, a multitargeted inhibitor of vascular endothelial growth factor receptor and platelet-derived growth factor receptor, in patients with metastatic renal cell carcinoma. J Clin Oncol, 2006. 24(1): p. 16-24.

317. Mross, K., et al., Phase I clinical and pharmacokinetic study of PTK/ZK, a multiple VEGF receptor inhibitor, in patients with liver metastases from solid tumours. Eur J Cancer, 2005. 41(9): p. 1291-9.

318. Escudier, B., et al., Sorafenib for treatment of renal cell carcinoma: Final efficacy and safety results of the phase III treatment approaches in renal cancer global evaluation trial. J Clin Oncol, 2009. 27(20): p. 3312-8.

319. Lansiaux, A., et al., Circulating thrombospondin 1 level as a surrogate marker in patients receiving cyclophosphamide-based metronomic chemotherapy. Invest New Drugs, 2012. 30(1): p. 403-4.

320. Zhu, A.X., et al., Efficacy, safety, and potential biomarkers of sunitinib monotherapy in advanced hepatocellular carcinoma: a phase II study. J Clin Oncol, 2009. 27(18): p. 3027-35.

321. Goon, P.K., et al., Circulating endothelial cells, endothelial progenitor cells, and endothelial microparticles in cancer. Neoplasia, 2006. 8(2): p. 79-88.

322. Hristov, M., W. Erl, and P.C. Weber, Endothelial progenitor cells: mobilization, differentiation, and homing. Arterioscler Thromb Vasc Biol, 2003. 23(7): p. 1185- 9.

323. Mancuso, P., et al., Circulating endothelial-cell kinetics and viability predict survival in breast cancer patients receiving metronomic chemotherapy. Blood, 2006. 108(2): p. 452-9.

324. Matsusaka, S., et al., Circulating endothelial progenitors and CXCR4-positive circulating endothelial cells are predictive markers for bevacizumab. Cancer, 2011. 117(17): p. 4026-32.

200

325. Taylor, M., et al., High levels of circulating VEGFR2+ Bone marrow-derived progenitor cells correlate with metastatic disease in patients with pediatric solid malignancies. Clin Cancer Res, 2009. 15(14): p. 4561-71.

326. Pradhan, K.R., et al., Polychromatic flow cytometry identifies novel subsets of circulating cells with angiogenic potential in pediatric solid tumors. Cytometry B Clin Cytom, 2011. 80(5): p. 335-8.

327. Kleinman, M.E., et al., Increased circulating AC133+ CD34+ endothelial progenitor cells in children with hemangioma. Lymphat Res Biol, 2003. 1(4): p. 301-7.

328. Taddeo, A., et al., Circulating endothelial progenitor cells are increased in patients with classic Kaposi's sarcoma. J Invest Dermatol, 2008. 128(8): p. 2125- 8.

329. Shao, Y.Y., et al., High circulating endothelial progenitor levels associated with poor survival of advanced hepatocellular carcinoma patients receiving sorafenib combined with metronomic chemotherapy. Oncology, 2011. 81(2): p. 98-103.

330. Shaked, Y., et al., Optimal biologic dose of metronomic chemotherapy regimens is associated with maximum antiangiogenic activity. Blood, 2005. 106(9): p. 3058-61.

331. Rössler, J., et al., The selective VEGFR1-3 inhibitor axitinib (AG-013736) shows antitumor activity in human neuroblastoma xenografts. Int J Cancer, 2011. 128(11): p. 2748-58.

332. Farace, F., et al., Levels of circulating CD45(dim)CD34(+)VEGFR2(+) progenitor cells correlate with outcome in metastatic renal cell carcinoma patients treated with tyrosine kinase inhibitors. Br J Cancer, 2011. 104(7): p. 1144-50.

333. DuBois, S.G., et al., Circulating endothelial cells and circulating endothelial precursor cells in patients with osteosarcoma. Pediatr Blood Cancer, 2012. 58(2): p. 181-4.

334. Marković, O., et al., Expression of VEGF and microvessel density in patients with multiple myeloma: clinical and prognostic significance. Med Oncol, 2008. 25(4): p. 451-7.

335. Abramson, L.P., et al., Increased microvascular density predicts relapse in Wilms' tumor. J Pediatr Surg, 2003. 38(3): p. 325-30; discussion 325-30.

336. Marback, E.F., et al., Tumour angiogenesis as a prognostic factor for disease dissemination in retinoblastoma. Br J Ophthalmol, 2003. 87(10): p. 1224-8.

337. Zhao, Y.Y., et al., Predictive value of intratumoral microvascular density in patients with advanced non-small cell lung cancer receiving chemotherapy plus bevacizumab. J Thorac Oncol, 2012. 7(1): p. 71-5.

201

338. Zhang, L., et al., In vivo antitumor and antimetastatic activity of sunitinib in preclinical neuroblastoma mouse model. Neoplasia, 2009. 11(5): p. 426-35.

339. Castillo-Avila, W., et al., Sunitinib inhibits tumor growth and synergizes with cisplatin in orthotopic models of cisplatin-sensitive and cisplatin-resistant human testicular germ cell tumors. Clin Cancer Res, 2009. 15(10): p. 3384-95.

340. Wang, X., et al., High dose intermittent sorafenib shows improved efficacy over conventional continuous dose in renal cell carcinoma. J Transl Med, 2011. 9: p. 220.

341. Kamat, A.A., et al., Metronomic chemotherapy enhances the efficacy of antivascular therapy in ovarian cancer. Cancer Res, 2007. 67(1): p. 281-8.

342. Hsu, C.Y., et al., Dynamic contrast-enhanced magnetic resonance imaging biomarkers predict survival and response in hepatocellular carcinoma patients treated with sorafenib and metronomic tegafur/uracil. J Hepatol, 2011. 55(4): p. 858-65.

343. Stockhammer, F., et al., Continuous low-dose temozolomide and celecoxib in recurrent glioblastoma. J Neurooncol, 2010. 100(3): p. 407-15.

344. Blankenberg, F.G., et al., Noninvasive assessment of tumor VEGF receptors in response to treatment with pazopanib: a molecular imaging study. Transl Oncol, 2010. 3(1): p. 56-64.

345. Mupparaju, S., et al., Tumor pO₂ as a surrogate marker to identify therapeutic window during metronomic chemotherapy of 9L gliomas. Adv Exp Med Biol, 2011. 701: p. 107-13.

346. Shao, Y.Y., et al., Early alpha-fetoprotein response predicts treatment efficacy of antiangiogenic systemic therapy in patients with advanced hepatocellular carcinoma. Cancer, 2010. 116(19): p. 4590-6.

347. Kesari, S., et al., Phase II study of metronomic chemotherapy for recurrent malignant gliomas in adults. Neuro Oncol, 2007. 9(3): p. 354-63.

348. Verhoeff, J.J., et al., Concerns about anti-angiogenic treatment in patients with glioblastoma multiforme. BMC Cancer, 2009. 9: p. 444.

349. Reardon, D.A., et al., Phase II study of metronomic chemotherapy with bevacizumab for recurrent glioblastoma after progression on bevacizumab therapy. J Neurooncol, 2011. 103(2): p. 371-9.

350. Tamaskar, I., J. Dhillon, and R. Pili, Resistance to angiogenesis inhibitors in renal cell carcinoma. Clin Adv Hematol Oncol, 2011. 9(2): p. 101-10.

202

351. Reardon, D.A., et al., Metronomic chemotherapy with daily, oral etoposide plus bevacizumab for recurrent malignant glioma: a phase II study. Br J Cancer, 2009. 101(12): p. 1986-94.

352. Kerbel, R.S., Inhibition of tumor angiogenesis as a strategy to circumvent acquired resistance to anti-cancer therapeutic agents. Bioessays, 1991. 13(1): p. 31-6.

353. Kubisch, R., et al., A Comprehensive Gene Expression Analysis of Resistance Formation upon Metronomic Cyclophosphamide Therapy. Transl Oncol, 2013. 6(1): p. 1-9.

354. Casanovas, O., et al., Drug resistance by evasion of antiangiogenic targeting of VEGF signaling in late-stage pancreatic islet tumors. Cancer Cell, 2005. 8(4): p. 299-309.

355. Mizukami, Y., et al., Induction of interleukin-8 preserves the angiogenic response in HIF-1alpha-deficient colon cancer cells. Nat Med, 2005. 11(9): p. 992-7.

356. Fernando, N.T., et al., Tumor escape from endogenous, extracellular matrix- associated angiogenesis inhibitors by up-regulation of multiple proangiogenic factors. Clin Cancer Res, 2008. 14(5): p. 1529-39.

357. Ebos, J.M., et al., Multiple circulating proangiogenic factors induced by sunitinib malate are tumor-independent and correlate with antitumor efficacy. Proc Natl Acad Sci U S A, 2007. 104(43): p. 17069-74.

358. Ceradini, D.J., et al., Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1. Nat Med, 2004. 10(8): p. 858-64.

359. Arbab, A.S., Activation of alternative pathways of angiogenesis and involvement of stem cells following anti-angiogenesis treatment in glioma. Histol Histopathol, 2012. 27(5): p. 549-57.

360. Du, R., et al., HIF1alpha induces the recruitment of bone marrow-derived vascular modulatory cells to regulate tumor angiogenesis and invasion. Cancer Cell, 2008. 13(3): p. 206-20.

361. di Tomaso, E., et al., PDGF-C induces maturation of blood vessels in a model of glioblastoma and attenuates the response to anti-VEGF treatment. PLoS One, 2009. 4(4): p. e5123.

362. Benjamin, L.E., I. Hemo, and E. Keshet, A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF. Development, 1998. 125(9): p. 1591-8.

363. Abramsson, A., P. Lindblom, and C. Betsholtz, Endothelial and nonendothelial sources of PDGF-B regulate pericyte recruitment and influence vascular pattern formation in tumors. J Clin Invest, 2003. 112(8): p. 1142-51.

203

364. Chantrain, C.F., et al., Stromal matrix metalloproteinase-9 regulates the vascular architecture in neuroblastoma by promoting pericyte recruitment. Cancer Res, 2004. 64(5): p. 1675-86.

365. Franco, M., et al., Pericytes promote endothelial cell survival through induction of autocrine VEGF-A signaling and Bcl-w expression. Blood, 2011. 118(10): p. 2906-17.

366. Helfrich, I., et al., Resistance to antiangiogenic therapy is directed by vascular phenotype, vessel stabilization, and maturation in malignant melanoma. J Exp Med, 2010. 207(3): p. 491-503.

367. Jockovich, M.E., et al., Retinoblastoma tumor vessel maturation impacts efficacy of vessel targeting in the LH(BETA)T(AG) mouse model. Invest Ophthalmol Vis Sci, 2007. 48(6): p. 2476-82.

368. Rubenstein, J.L., et al., Anti-VEGF antibody treatment of glioblastoma prolongs survival but results in increased vascular cooption. Neoplasia, 2000. 2(4): p. 306-14.

369. Pàez-Ribes, M., et al., Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis. Cancer Cell, 2009. 15(3): p. 220-31.

370. Ebos, J.M., et al., Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis. Cancer Cell, 2009. 15(3): p. 232-9.

371. Seiz, M., et al., Far-distant metastases along the CSF pathway of glioblastoma multiforme during continuous low-dose chemotherapy with temozolomide and celecoxib. Neurosurg Rev, 2010. 33(3): p. 375-81; discussion 381.

372. Jakovljević, G., et al., Vascular endothelial growth factor in children with neuroblastoma: a retrospective analysis. J Exp Clin Cancer Res, 2009. 28: p. 143.

373. Zhang, L., et al., Vascular endothelial growth factor overexpression by soft tissue sarcoma cells: implications for tumor growth, metastasis, and chemoresistance. Cancer Res, 2006. 66(17): p. 8770-8.

374. Peles, E., et al., Angiogenic factors in the cerebrospinal fluid of patients with astrocytic brain tumors. Neurosurgery, 2004. 55(3): p. 562-7; discussion 567-8.

375. Nardo, G., et al., Glycolytic phenotype and AMP kinase modify the pathologic response of tumor xenografts to VEGF neutralization. Cancer Res, 2011. 71(12): p. 4214-25.

376. Keunen, O., et al., Anti-VEGF treatment reduces blood supply and increases tumor cell invasion in glioblastoma. Proc Natl Acad Sci U S A, 2011. 108(9): p. 3749-54.

204

377. Vander Heiden, M.G., L.C. Cantley, and C.B. Thompson, Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science, 2009. 324(5930): p. 1029-33.

378. Martin-Padura, I., et al., Residual dormant cancer stem-cell foci are responsible for tumor relapse after antiangiogenic metronomic therapy in hepatocellular carcinoma xenografts. Lab Invest, 2012. 92(7): p. 952-66.

379. Conley, S.J., et al., Antiangiogenic agents increase breast cancer stem cells via the generation of tumor hypoxia. Proc Natl Acad Sci U S A, 2012. 109(8): p. 2784-9.

380. Rapisarda, A. and G. Melillo, Overcoming disappointing results with antiangiogenic therapy by targeting hypoxia. Nat Rev Clin Oncol, 2012. 9(7): p. 378-90.

381. Keith, B. and M.C. Simon, Hypoxia-inducible factors, stem cells, and cancer. Cell, 2007. 129(3): p. 465-72.

382. Keshelava, N., S. Groshen, and C.P. Reynolds, Cross-resistance of topoisomerase I and II inhibitors in neuroblastoma cell lines. Cancer Chemother Pharmacol, 2000. 45(1): p. 1-8.

383. Kerbel, R.S. and B.A. Kamen, The anti-angiogenic basis of metronomic chemotherapy. Nat Rev Cancer, 2004. 4(6): p. 423-36.

384. Garcia, A.A., et al., Phase II clinical trial of bevacizumab and low-dose metronomic oral cyclophosphamide in recurrent ovarian cancer: a trial of the California, Chicago, and Princess Margaret Hospital phase II consortia. J Clin Oncol, 2008. 26(1): p. 76-82.

385. Tweddle, D.A., et al., Evidence for the development of p53 mutations after cytotoxic therapy in a neuroblastoma cell line. Cancer Res, 2001. 61(1): p. 8-13.

386. Walton, J.D., et al., Characteristics of stem cells from human neuroblastoma cell lines and in tumors. Neoplasia, 2004. 6(6): p. 838-45.

387. Livingstone, A., et al., N-myc gene copy number in neuroblastoma cell lines and resistance to experimental treatment. Eur J Cancer, 1994. 30A(3): p. 382-9.

388. Carloni, G., et al., Integration and loss of a single v-Ki-ras gene affects tumorigenic potential of human osteosarcoma cells. FEBS Lett, 1988. 229(2): p. 333-9.

389. Zeng, F.Y., et al., PAX3-FKHR sensitizes human alveolar rhabdomyosarcoma cells to camptothecin-mediated growth inhibition and apoptosis. Cancer Lett, 2009. 284(2): p. 157-64.

205

390. Das, B., et al., Squalene selectively protects mouse bone marrow progenitors against cisplatin and carboplatin-induced cytotoxicity in vivo without protecting tumor growth. Neoplasia, 2008. 10(10): p. 1105-19.

391. Ross, R.A., et al., Human neuroblastoma I-type cells are malignant neural crest stem cells. Cell Growth Differ, 1995. 6(4): p. 449-56.

392. Dimitroulakos, J., et al., NUB-7: a stable I-type human neuroblastoma cell line inducible along N- and S-type cell lineages. Cell Growth Differ, 1994. 5(4): p. 373-84.

393. Limvorasak, S. and E.M. Posadas, Pazopanib: therapeutic developments. Expert Opin Pharmacother, 2009. 10(18): p. 3091-102.

394. Oberthuer, A., et al., Molecular characterization and classification of neuroblastoma. Future Oncol, 2009. 5(5): p. 625-39.

395. Ciccarone, V., et al., Phenotypic diversification in human neuroblastoma cells: expression of distinct neural crest lineages. Cancer Res, 1989. 49(1): p. 219-25.

396. Chan, H.S., et al., MYCN protein expression as a predictor of neuroblastoma prognosis. Clin Cancer Res, 1997. 3(10): p. 1699-706.

397. Ribatti, D., et al., In vivo angiogenic activity of neuroblastoma correlates with MYCN oncogene overexpression. Int J Cancer, 2002. 102(4): p. 351-4.

398. Kang, J., et al., N-myc is a novel regulator of PI3K-mediated VEGF expression in neuroblastoma. Oncogene, 2008. 27(28): p. 3999-4007.

399. Hatzi, E., et al., N-myc oncogene overexpression down-regulates IL-6; evidence that IL-6 inhibits angiogenesis and suppresses neuroblastoma tumor growth. Oncogene, 2002. 21(22): p. 3552-61.

400. Yu, J.L., et al., Effect of p53 status on tumor response to antiangiogenic therapy. Science, 2002. 295(5559): p. 1526-8.

401. Segerström, L., et al., The anti-VEGF antibody bevacizumab potently reduces the growth rate of high-risk neuroblastoma xenografts. Pediatr Res, 2006. 60(5): p. 576-81.

402. Patterson, D.M., et al., Effect of MDM2 and vascular endothelial growth factor inhibition on tumor angiogenesis and metastasis in neuroblastoma. Angiogenesis, 2011. 14(3): p. 255-66.

403. Matthay, K.K., et al., Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13- cis-retinoic acid. Children's Cancer Group. N Engl J Med, 1999. 341(16): p. 1165-73.

206

404. Sorensen, P.H., et al., PAX3-FKHR and PAX7-FKHR gene fusions are prognostic indicators in alveolar rhabdomyosarcoma: a report from the children's oncology group. J Clin Oncol, 2002. 20(11): p. 2672-9.

405. Scheidler, S., et al., The hybrid PAX3-FKHR fusion protein of alveolar rhabdomyosarcoma transforms fibroblasts in culture. Proc Natl Acad Sci U S A, 1996. 93(18): p. 9805-9.

406. Wang, W., et al., The cooperative transforming effects of PAX3-FKHR and IGF-II on mouse myoblasts. Int J Oncol, 2005. 27(4): p. 1087-96.

407. Olanich, M.E. and F.G. Barr, A call to ARMS: targeting the PAX3-FOXO1 gene in alveolar rhabdomyosarcoma. Expert Opin Ther Targets, 2013. 17(5): p. 607- 23.

408. Hlatky, L., P. Hahnfeldt, and J. Folkman, Clinical application of antiangiogenic therapy: microvessel density, what it does and doesn't tell us. J Natl Cancer Inst, 2002. 94(12): p. 883-93.

409. Drozynska, E., et al., A correlation of microvascular density and proliferative activity to clinical and histological characteristics in neuroblastoma. Neoplasma, 2006. 53(3): p. 247-52.

410. Jodele, S., et al., The contribution of bone marrow-derived cells to the tumor vasculature in neuroblastoma is matrix metalloproteinase-9 dependent. Cancer Res, 2005. 65(8): p. 3200-8.

411. Kumar, R., et al., Myelosuppression and kinase selectivity of multikinase angiogenesis inhibitors. Br J Cancer, 2009. 101(10): p. 1717-23.

412. Erickson-Miller, C.L., et al., Differential toxicity of camptothecin, topotecan and 9- aminocamptothecin to human, canine, and murine myeloid progenitors (CFU- GM) in vitro. Cancer Chemother Pharmacol, 1997. 39(5): p. 467-72.

413. Masubuchi, N., R.D. May, and R. Atsumi, A predictive model of human myelotoxicity using five camptothecin derivatives and the in vitro colony-forming unit granulocyte/macrophage assay. Clin Cancer Res, 2004. 10(19): p. 6722-31.

414. Yap, K.Y., et al., Clinically relevant drug interactions between anticancer drugs and psychotropic agents. Eur J Cancer Care (Engl), 2011. 20(1): p. 6-32.

415. Tiwari, A.K., R. Zhang, and J.M. Gallo, Overlapping functions of ABC transporters in topotecan disposition as determined in gene knockout mouse models. Mol Cancer Ther, 2013.

416. Zamboni, W.C., et al., Disposition of 9-nitrocamptothecin and its 9- aminocamptothecin metabolite in relation to ABC transporter genotypes. Invest New Drugs, 2006. 24(5): p. 393-401.

207

417. Shukla, S., et al., Sunitinib (Sutent, SU11248), a small-molecule receptor tyrosine kinase inhibitor, blocks function of the ATP-binding cassette (ABC) transporters P-glycoprotein (ABCB1) and ABCG2. Drug Metab Dispos, 2009. 37(2): p. 359-65.

418. Perry, J., et al., A synergistic interaction between lapatinib and chemotherapy agents in a panel of cell lines is due to the inhibition of the efflux pump BCRP. Mol Cancer Ther, 2010. 9(12): p. 3322-9.

419. Minocha, M., et al., Enhanced brain accumulation of pazopanib by modulating P- gp and Bcrp1 mediated efflux with canertinib or erlotinib. Int J Pharm, 2012. 436(1-2): p. 127-34.

420. Minocha, M., et al., Co-administration strategy to enhance brain accumulation of vandetanib by modulating P-glycoprotein (P-gp/Abcb1) and breast cancer resistance protein (Bcrp1/Abcg2) mediated efflux with m-TOR inhibitors. Int J Pharm, 2012. 434(1-2): p. 306-14.

421. Tan, A.R., et al., Effects of ketoconazole and esomeprazole on the pharmacokinetics of pazopanib in patients with solid tumors. Cancer Chemother Pharmacol, 2013.

422. Goh, B.C., et al., An evaluation of the drug interaction potential of pazopanib, an oral vascular endothelial growth factor receptor tyrosine kinase inhibitor, using a modified Cooperstown 5+1 cocktail in patients with advanced solid tumors. Clin Pharmacol Ther, 2010. 88(5): p. 652-9.

423. Bergers, G. and D. Hanahan, Modes of resistance to anti-angiogenic therapy. Nat Rev Cancer, 2008. 8(8): p. 592-603.

424. Crawford, Y. and N. Ferrara, Tumor and stromal pathways mediating refractoriness/resistance to anti-angiogenic therapies. Trends Pharmacol Sci, 2009. 30(12): p. 624-30.

425. Milane, L., Z. Duan, and M. Amiji, Role of hypoxia and glycolysis in the development of multi-drug resistance in human tumor cells and the establishment of an orthotopic multi-drug resistant tumor model in nude mice using hypoxic pre-conditioning. Cancer Cell Int, 2011. 11: p. 3.

426. Loges, S., T. Schmidt, and P. Carmeliet, Mechanisms of resistance to antiangiogenic therapy and development of third-generation anti-angiogenic drug candidates. Genes Cancer, 2010. 1(1): p. 12-25.

427. Zhang, L., et al., Tissue microenvironment modulates CXCR4 expression and tumor metastasis in neuroblastoma. Neoplasia, 2007. 9(1): p. 36-46.

428. Qing, G., et al., Combinatorial regulation of neuroblastoma tumor progression by N-Myc and hypoxia inducible factor HIF-1alpha. Cancer Res, 2010. 70(24): p. 10351-61.

208

429. Potter, C. and A.L. Harris, Hypoxia inducible carbonic anhydrase IX, marker of tumour hypoxia, survival pathway and therapy target. Cell Cycle, 2004. 3(2): p. 164-7.

430. Dungwa, J.V., L.P. Hunt, and P. Ramani, Carbonic anhydrase IX up-regulation is associated with adverse clinicopathologic and biologic factors in neuroblastomas. Hum Pathol, 2012. 43(10): p. 1651-60.

431. Kuschel, A., P. Simon, and S. Tug, Functional regulation of HIF-1α under normoxia--is there more than post-translational regulation? J Cell Physiol, 2012. 227(2): p. 514-24.

432. Holmquist-Mengelbier, L., et al., Recruitment of HIF-1alpha and HIF-2alpha to common target genes is differentially regulated in neuroblastoma: HIF-2alpha promotes an aggressive phenotype. Cancer Cell, 2006. 10(5): p. 413-23.

433. Semenza, G.L., Targeting HIF-1 for cancer therapy. Nat Rev Cancer, 2003. 3(10): p. 721-32.

434. Schioppa, T., et al., Regulation of the chemokine receptor CXCR4 by hypoxia. J Exp Med, 2003. 198(9): p. 1391-402.

435. Hussein, D., et al., Chronic hypoxia promotes hypoxia-inducible factor-1alpha- dependent resistance to etoposide and vincristine in neuroblastoma cells. Mol Cancer Ther, 2006. 5(9): p. 2241-50.

436. Nilsson, M.B., et al., Multiple receptor tyrosine kinases regulate HIF-1alpha and HIF-2alpha in normoxia and hypoxia in neuroblastoma: implications for antiangiogenic mechanisms of multikinase inhibitors. Oncogene, 2010. 29(20): p. 2938-49.

437. Li, X., et al., VEGF-independent angiogenic pathways induced by PDGF-C. Oncotarget, 2010. 1(4): p. 309-314.

438. Crawford, Y., et al., PDGF-C mediates the angiogenic and tumorigenic properties of fibroblasts associated with tumors refractory to anti-VEGF treatment. Cancer Cell, 2009. 15(1): p. 21-34.

439. Mejía, C., et al., Prognostic significance of cell proliferation in human neuroblastoma: comparison with other prognostic factors. Oncol Rep, 2003. 10(1): p. 243-7.

440. Hans, F. and S. Dimitrov, Histone H3 phosphorylation and cell division. Oncogene, 2001. 20(24): p. 3021-7.

441. Faisal, A., et al., The aurora kinase inhibitor CCT137690 downregulates MYCN and sensitizes MYCN-amplified neuroblastoma in vivo. Mol Cancer Ther, 2011. 10(11): p. 2115-23.

209

442. Box, A.H. and D.J. Demetrick, Cell cycle kinase inhibitor expression and hypoxia-induced cell cycle arrest in human cancer cell lines. Carcinogenesis, 2004. 25(12): p. 2325-35.

443. Manohar, S.M., et al., Cyclin-dependent kinase inhibitor, P276-00, inhibits HIF- 1α and induces G2/M arrest under hypoxia in prostate cancer cells. Prostate Cancer Prostatic Dis, 2012. 15(1): p. 15-27.

444. Lou, J., et al., XQ2, a novel TPZ derivative, induced G2/M phase arrest and apoptosis under hypoxia in non-small cell lung cancer cells. Biosci Biotechnol Biochem, 2010. 74(6): p. 1181-7.

445. Feeney, G.P., et al., Tracking the cell cycle origins for escape from topotecan action by breast cancer cells. Br J Cancer, 2003. 88(8): p. 1310-7.

446. Gatenby, R.A. and R.J. Gillies, Why do cancers have high aerobic glycolysis? Nat Rev Cancer, 2004. 4(11): p. 891-9.

447. Lum, J.J., et al., The transcription factor HIF-1alpha plays a critical role in the growth factor-dependent regulation of both aerobic and anaerobic glycolysis. Genes Dev, 2007. 21(9): p. 1037-49.

448. Marín-Hernández, A., et al., HIF-1alpha modulates energy metabolism in cancer cells by inducing over-expression of specific glycolytic isoforms. Mini Rev Med Chem, 2009. 9(9): p. 1084-101.

449. Guillaumond, F., et al., Strengthened glycolysis under hypoxia supports tumor symbiosis and hexosamine biosynthesis in pancreatic adenocarcinoma. Proc Natl Acad Sci U S A, 2013. 110(10): p. 3919-24.

450. Fiume, L., et al., Inhibition of lactic dehydrogenase as a way to increase the anti- proliferative effect of multi-targeted kinase inhibitors. Pharmacol Res, 2011. 63(4): p. 328-34.

451. Xu, J., et al., Colorectal Cancer Cells Refractory to Anti-VEGF Treatment Are Vulnerable to Glycolytic Blockade due to Persistent Impairment of Mitochondria. Mol Cancer Ther, 2013. 12(5): p. 717-24.

452. Lunt, S.Y. and M.G. Vander Heiden, Aerobic glycolysis: meeting the metabolic requirements of cell proliferation. Annu Rev Cell Dev Biol, 2011. 27: p. 441-64.

453. Olson, A.L. and J.E. Pessin, Structure, function, and regulation of the mammalian facilitative glucose transporter gene family. Annu Rev Nutr, 1996. 16: p. 235-56.

454. Shim, B.Y., et al., Glucose transporter 1 (GLUT1) of anaerobic glycolysis as predictive and prognostic values in neoadjuvant chemoradiotherapy and laparoscopic surgery for locally advanced rectal cancer. Int J Colorectal Dis, 2013. 28(3): p. 375-83.

210

455. Pastorino, J.G., N. Shulga, and J.B. Hoek, Mitochondrial binding of hexokinase II inhibits Bax-induced cytochrome c release and apoptosis. J Biol Chem, 2002. 277(9): p. 7610-8.

456. Mathupala, S.P., Y.H. Ko, and P.L. Pedersen, Hexokinase II: cancer's double- edged sword acting as both facilitator and gatekeeper of malignancy when bound to mitochondria. Oncogene, 2006. 25(34): p. 4777-86.

457. Muñoz-Pinedo, C., N. El Mjiyad, and J.E. Ricci, Cancer metabolism: current perspectives and future directions. Cell Death Dis, 2012. 3: p. e248.

458. Airley, R., et al., Glucose transporter glut-1 expression correlates with tumor hypoxia and predicts metastasis-free survival in advanced carcinoma of the cervix. Clin Cancer Res, 2001. 7(4): p. 928-34.

459. Chung, F.Y., et al., GLUT1 gene is a potential hypoxic marker in colorectal cancer patients. BMC Cancer, 2009. 9: p. 241.

460. Rademakers, S.E., et al., Metabolic markers in relation to hypoxia; staining patterns and colocalization of pimonidazole, HIF-1α, CAIX, LDH-5, GLUT-1, MCT1 and MCT4. BMC Cancer, 2011. 11: p. 167.

461. Matsushita, K., et al., Glycolysis inhibitors as a potential therapeutic option to treat aggressive neuroblastoma expressing GLUT1. J Pediatr Surg, 2012. 47(7): p. 1323-30.

462. Kang, S.S., et al., Clinical significance of glucose transporter 1 (GLUT1) expression in human breast carcinoma. Jpn J Cancer Res, 2002. 93(10): p. 1123-8.

463. Shulga, N., R. Wilson-Smith, and J.G. Pastorino, Hexokinase II detachment from the mitochondria potentiates cisplatin induced cytotoxicity through a caspase-2 dependent mechanism. Cell Cycle, 2009. 8(20): p. 3355-64.

464. Noguera, R., et al., HIF-1alpha and HIF-2alpha are differentially regulated in vivo in neuroblastoma: high HIF-1alpha correlates negatively to advanced clinical stage and tumor vascularization. Clin Cancer Res, 2009. 15(23): p. 7130-6.

465. Chen, L., et al., p53 is a direct transcriptional target of MYCN in neuroblastoma. Cancer Res, 2010. 70(4): p. 1377-88.

466. Hammond, E.M. and A.J. Giaccia, Hypoxia-inducible factor-1 and p53: friends, acquaintances, or strangers? Clin Cancer Res, 2006. 12(17): p. 5007-9.

467. Scatena, R., et al., Glycolytic enzyme inhibitors in cancer treatment. Expert Opin Investig Drugs, 2008. 17(10): p. 1533-45.

468. Shutt, D.C., et al., 2-deoxy-D-glucose induces oxidative stress and cell killing in human neuroblastoma cells. Cancer Biol Ther, 2010. 9(11): p. 853-61.

211

469. McKeage, M.J., et al., A phase I trial of PR-104, a pre-prodrug of the bioreductive prodrug PR-104A, given weekly to solid tumour patients. BMC Cancer, 2011. 11: p. 432.

470. Friery, O.P., et al., Enhancement of the anti-tumour effect of cyclophosphamide by the bioreductive drugs AQ4N and tirapazamine. Br J Cancer, 2000. 82(8): p. 1469-73.

471. Gallagher, R., et al., The chemopotentiation of cisplatin by the novel bioreductive drug AQ4N. Br J Cancer, 2001. 85(4): p. 625-9.

472. Papadopoulos, K.P., et al., A phase 1 open-label, accelerated dose-escalation study of the hypoxia-activated prodrug AQ4N in patients with advanced malignancies. Clin Cancer Res, 2008. 14(21): p. 7110-5.

473. Liu, Q., et al., TH-302, a hypoxia-activated prodrug with broad in vivo preclinical combination therapy efficacy: optimization of dosing regimens and schedules. Cancer Chemother Pharmacol, 2012. 69(6): p. 1487-98.

474. Sun, J.D., et al., Selective tumor hypoxia targeting by hypoxia-activated prodrug TH-302 inhibits tumor growth in preclinical models of cancer. Clin Cancer Res, 2012. 18(3): p. 758-70.

475. Weiss, G.J., et al., Phase 1 study of the safety, tolerability, and pharmacokinetics of TH-302, a hypoxia-activated prodrug, in patients with advanced solid malignancies. Clin Cancer Res, 2011. 17(9): p. 2997-3004.

476. Ganjoo, K.N., et al., A phase I study of the safety and pharmacokinetics of the hypoxia-activated prodrug TH-302 in combination with doxorubicin in patients with advanced soft tissue sarcoma. Oncology, 2011. 80(1-2): p. 50-6.