Evaluation of the Pharmacokinetic

Home , Antiestrogen, Biotransformation, Catechol estrogen, Glucuronidation, Nicotinamide

Virginia Commonwealth University VCU Scholars Compass

Theses and Dissertations Graduate School

2005

Evaluation of the Pharmacokinetic-Pharmacodynamic relationship, Metabolism and Plasma Protein Binding of the novel antitumor agent, 2-Methoxyestradiol (2ME2), following oral administration in patients with solid tumors.

Nehal Jagdish Lakhani Virginia Commonwealth University

Follow this and additional works at: https://scholarscompass.vcu.edu/etd

Part of the Pharmacy and Pharmaceutical Sciences Commons

Downloaded from https://scholarscompass.vcu.edu/etd/1381

This Dissertation is brought to you for free and open access by the Graduate School at VCU Scholars Compass. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of VCU Scholars Compass. For more information, please contact [email protected].

School of Pharmacy Virginia Commonwealth University

This is to certify that the dissertation prepared by Nehal Jagdish Lakhani entitled “EVALUATION OF THE PHARMACOKINETIC-PHARMACODYNAMIC RELATIONSHIP, METABOLISM AND PLASMA PROTEIN BINDING OF THE NOVEL ANTITUMOR AGENT, 2-METHOXYESTRADIOL (2ME2), FOLLOWING ORAL ADMINISTRATION IN PATIENTS WITH SOLID TUMORS” has been approved by his or her committee as satisfactory completion of the thesis or dissertation requirement for the degree of Doctor of Philosophy.

Dr. Jürgen Venitz, M.D., Ph.D., Co-Director of Dissertation, School of Pharmacy

Dr. William D. Figg, Pharm. D., Co-Director of Dissertation, National Cancer Institute

Dr. Les Edinboro, Ph.D., School of Pharmacy

Dr. Patty Slattum, Pharm. D., Ph.D., School of Pharmacy

Dr. Mohamadi A. Sarkar, Ph.D., School of Pharmacy

Dr. John Roberts, M.D., School of Medicine

Dr. Peter Byron, Ph.D., Department Chair

Dr. Victor Yanchick, Ph.D., Dean, School of Pharmacy

Dr. F. Douglas Boudinot, Dean of the School of Graduate Studies

June 9, 2005

© EVALUATION OF THE PHARMACOKINETIC-PHARMACODYNAMIC RELATIONSHIP, METABOLISM & PLASMA PROTEIN BINDING OF THE NOVEL ANTITUMOR AGENT, 2-METHOXYESTRADIOL (2ME2), FOLLOWING ORAL ADMINISTRATION IN PATIENTS WITH SOLID TUMORS

EVALUATION OF THE PHARMACOKINETIC-PHARMACODYNAMIC RELATIONSHIP, METABOLISM & PLASMA PROTEIN BINDING OF THE NOVEL ANTITUMOR AGENT, 2-METHOXYESTRADIOL (2ME2), FOLLOWING ORAL ADMINISTRATION IN PATIENTS WITH SOLID TUMORS

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at Virginia Commonwealth University.

NEHAL JAGDISH LAKHANI M.B; B.S., Maharaja Sayajirao University, Baroda, India, 2000

Co-Director: Jürgen Venitz, M.D., Ph.D. Associate Professor, Department of Pharmaceutics

Co-Director: William D. Figg, Pharm.D. Head, Clinical Pharmacology Research Core & Molecular Pharmacology Section, National Cancer Institute, National Institutes of Health

Virginia Commonwealth University Richmond, Virginia August 2005

Acknowledgement

I would like to acknowledge the contribution of following people and

organizations for their help and support in the completion of my dissertation project:

Dr. Jürgen Venitz for providing me with the continuous support, guidance and

direction throughout my Ph.D. project. Your hard work and perseverance with my Ph.D.

project will always be appreciated.

Dr. William D. Figg for providing me the wonderful opportunity of pursuing my

education at the National Cancer Institute (NCI). The clinical experience along with research training that you provided was of the highest quality.

Dr. Alex Sparreboom, for sharing his vast knowledge and experience, and providing valuable input in experimental designs, along every step of my Ph.D. project. I am grateful to you for being there when it really mattered.

I am grateful to all the members of my graduate committee, Drs. Mohamadi

Sarkar, Patty Slattum, Les Edinboro and John Roberts for their guidance throughout the project.

Dr. William Dahut for serving as the Principal Investigator of the Phase I clinical trial of 2ME2, Debbie McNally and Miranda Raggio for serving as the research nurses for the 2ME2 clinical trial. Your help, support and input in the clinical trial protocol and numerous manuscripts will always be remembered.

iii

Dr. Anthony Treston from Entremed Inc., for his scientific discussions regarding bioanalytical method development and clinical pharmacology of 2ME2, and for providing guidance every time I needed it, despite your hectic schedule.

Dr. Nicola Smith for helping me with metabolism experiments and Erin Lepper for helping me with the LC/MS/MS method development.

The National Cancer Institute and Department of Pharmaceutics at the Virginia

Commonwealth University (VCU) for completely funding my graduate education.

Thanks to all the members of the Pharmacokinetic-Pharmacodynamic Research

Group and the Figg lab for making the years in graduate school fun!

Finally, thanks to mom, dad and Parth for being so understanding and providing support that only a family can.

Table of Contents

Page

Acknowledgements...... ii

List of Tables ...... x

List of Figures...... xii

Chapter

1 Introduction...... 1

Estrogen Biosynthesis ...... 1

Estrogen Metabolism...... 2

Endogenous 2ME2 production...... 7

Mechanisms of antitumor action of 2ME2...... 9

Estrogenic activity of 2ME2...... 16

Preclinical Pharmacokinetics of 2ME2 ...... 17

Metabolism studies of 2ME2 ...... 18

Plasma Protein Binding of 2ME2...... 19

In-vivo excretion of 2ME2 ...... 19

Pharmacokinetics of orally administered estrogens in humans...... 19

Preliminary Clinical Results...... 22

2 Hypotheses and Objectives...... 24

Hypotheses ...... 24

Objectives...... 25

3 Bioanalytical method for quantitating of 2ME2 in human plasma and urine..26

Introduction ...... 26

Chemicals and reagents...... 27

Stock solutions and standards in plasma ...... 28

Sample pretreatment in plasma ...... 29

Equipment ...... 29

Chromatography and Mass Spectrometry conditions for plasma...... 30

Methodology for Validation in plasma ...... 37

Results of Validation in plasma...... 42

Stock solutions and standards in urine ...... 47

Sample preparation for testing amount of 2ME2 excreted unchanged in

urine...... 48

Sample preparation for testing amount of 2ME2 excreted as glucuronides

in urine...... 48

Sample preparation for testing amount of 2ME2 excreted as sulfates in

urine...... 49

Equipment ...... 50

Chromatography and detection in Urine ...... 50

Acceptability criteria for bioanalytic runs for analysis of 2ME2 in urine...58

Conclusion...... 62

4 Phase I Clinical Trial of Orally administered 2ME2 in patients with solid

tumors ...... 63

Introduction ...... 63

Patients and Methods...... 66

Drug Administration...... 67

Study Design ...... 68

Pharmacokinetic sampling ...... 69

Correlative studies...... 70

Statistical Analysis ...... 71

Results ...... 71

Clinical Outcomes ...... 77

Toxicities...... 80

Pharmacokinetics...... 82

24-hour urinary pharmacokinetics...... 92

Correlative Studies ...... 94

Discussion ...... 98

5 In vitro hepatic metabolism and in vivo elimination pathways of 2ME2 ...... 105

Introduction ...... 105

Chemicals and reagents...... 106

vii

In vitro hepatic phase I metabolism experimental...... 106

HPLC method for metabolite separation...... 107

Metabolism of 2ME2 by human hepatic CYP 450 isoenzymes...... 109

Enzyme Kinetics and estimation of hepatic clearance of 2ME2...... 110

In vitro hepatic phase I metabolism results...... 110

Michaelis-Menten Kinetics of Phase I metabolites...... 114

Phase I metabolite quantification ...... 120

Results of metabolism of 2ME2 by human hepatic CYP 450 isoenzymes122

In vitro hepatic phase II metabolism experimental ...... 133

In vitro phase II metabolism results ...... 134

Michaelis-Menten Kinetics of glucuronides of 2ME2 ...... 137

Results of sulfation experiments ...... 140

Calculation of total hepatic clearance ...... 143

Characterization of in vivo elimination pathways for 2ME2 ...... 144

Calculation of metabolic ratio ...... 144

Discussion ...... 145

6 Plasma Protein Binding studies of 2ME2 in human plasma...... 149

Introduction ...... 149

Chemicals and reagents...... 150

Description of equilibrium dialysis method ...... 150

viii

Method Validation...... 151

Binding to purified human proteins...... 153

Patients and treatment ...... 154

Measurement of total drug concentrations ...... 154

Measurement of unbound drug concentrations ...... 155

Pharmacokinetic analysis ...... 155

Statistical considerations ...... 156

Results of validation of dialysis method ...... 156

In vitro binding of 2ME2...... 157

Estimation of binding parameters...... 162

Displacement interactions on binding sites...... 163

Pharmacokinetics of unbound 2ME2 ...... 164

Discussion ...... 168

7 Summary and Conclusions ...... 171

Formulation ...... 171

Solubility ...... 172

Permeability...... 174

Transporter effects...... 175

Food effects ...... 175

First pass metabolism ...... 176

Oral Bioavailability estimates ...... 177

Plasma protein binding...... 177

Conclusion...... 178

Bibliography ...... 180

Appendices......

A Plasma concentration-time profiles of 2ME2 in patients with solid tumors..197

B LC/MS/MS analysis of 2ME2 in urine of patients with solid tumors ...... 209

C In vitro metabolism of 2ME2...... 216

List of Tables

Page

Table 1: Concentrations of 2-methoxyestrogens in human serum in various physiological conditions...... 8

Table 2: Cell lines inhibited by 2ME2...... 10

Table 3: Summary of pharmacokinetic parameters of 2ME2 in mice...... 18

Table 4: Pharmacokinetic parameters of estrone after a single oral dose of two tablets of

0.625 mg of Premarin®...... 20

Table 5: Pharmacokinetic parameters of 17 β-estradiol after a single oral dose of two

tablets of 0.625 mg of Premarin®...... 20

Table 6: Mobile Phase gradient used for LC separation in LC/MS/MS analytical method for 2ME2 in human plasma...... 30

Table 7: Back Calculated Concentrations from Calibration Curves for LC/MS/MS analytical method for 2ME2 in human plasma...... 44

Table 8: Assessment of Accuracy and Precision from Quality-Control Samples for

LC/MS/MS analytical method for 2ME2 in human plasma...... 45

Table 9: Freeze Thaw Stability of 2ME2...... 47

Table 10: Back Calculated Concentrations from Calibration Curves for LC/MS/MS analytical method for 2ME2 in human urine...... 60

Table 11: Assessment of Accuracy and Precision from Quality-Control Samples for

LC/MS/MS analytical method for 2ME2 in human urine...... 61

Table 12: Summary of clinical trials of 2ME2...... 65

Table 13: Dose Levels for 2ME2 clinical trial...... 69

Table 14: Patient demographics for 2ME2 clinical trial...... 73

Table 15: Patient disposition for 2ME2 clinical trial...... 75

Table 16: Adverse events table for 2ME2 clinical trial...... 81

Table 17: Noncompartment pharmacokinetic parameters of orally administered 2ME2..88

Table 18: Urinary excretion of 2ME2...... 93

Table 19: HPLC gradient for metabolite separation...... 109

Table 20: Kinetic parameters (mean estimate ± standard error) calculated from 2ME2 incubations with HLM protein and NADPH generating system...... 119

Table 21: Kinetic parameters (mean estimate ± standard error) for two glucuronides of

2ME2...... 139

Table 22: Metabolic ratios for glucuronidation and sulfation...... 144

Table 23: Binding of 2ME2 to purified human proteins...... 160

Table 24: Comparison of bound Vs unbound pharmacokinetic parameters of 2ME2. ...167

Table 25: Solubility profile of 2ME2...... 172

Table 26: Permeability of 2ME2 across Caco-2 cell monolayer...... 174

xii

List of Figures

Page

Figure 1: Estrogen Biotransformation Scheme...... 6

Figure 2: Chemical Structure of 2-methoxyestradiol...... 7

Figure 3: Chemical structure of the internal standard 2ME2-d5 ...... 28

Figure 4: LC-APCI-MS-MS chromatogram of blank human plasma sample 2ME2 after liquid-liquid extraction...... 32

Figure 5: LC-APCI-MS-MS chromatogram of standard spiked at a 2ME2 concentration of 100 ng/mL 2ME2 after liquid-liquid extraction ...... 33

Figure 6: LC-APCI-MS-MS chromatogram of standard spiked at a 2ME2 concentration of 1ng/mL 2ME2 after liquid-liquid extraction ...... 34

Figure 7: LC-APCI-MS-MS chromatogram of plasma sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, before receiving the dose of

2ME2; after liquid-liquid extraction ...... 35

Figure 8: LC-APCI-MS-MS chromatogram of plasma sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, 0.5 hour after receiving the dose of 2ME2; after liquid-liquid extraction ...... 36

Figure 9: LC-APCI-MS-MS chromatogram of blank human urine sample 2ME2 after liquid-liquid extraction...... 52

xiii

Figure 10: LC-APCI-MS-MS chromatogram of urine standard spiked at a 2ME2 concentration of 20 µg/mL after liquid-liquid extraction ...... 53

Figure 11: LC-APCI-MS-MS chromatogram of urine standard spiked at a 2ME2 concentration of 10 ng/mL after liquid-liquid extraction ...... 54

Figure 12: LC-APCI-MS-MS chromatogram of an extract of 1mL of urine from a 24 hour collection, obtained from a patient with ovarian cancer who received a single oral dose of

2ME2 at 2200 mg, without addition of β-glucuronidase ...... 55

Figure 13: LC-APCI-MS-MS chromatogram of an extract of 1mL of urine from a 24 hour

collection, obtained from a patient with ovarian cancer who received a single oral dose of

2ME2 at 2200 mg, after incubation of urine with β-glucuronidase...... 56

Figure 14: LC-APCI-MS-MS chromatogram of an extract of 1mL of urine from a 24 hour

collection, obtained from a patient with ovarian cancer who received a single oral dose of

2ME2 at 2200 mg, after incubation of urine with sulfatase...... 57

Figure 15: CA-125 concentrations versus time plot for patient #7...... 78

Figure 16: CT scan of patient #7 showing shrinkage in the left pelvic lymph nodes after

therapy with 2ME2 ...... 79

Figure 17: Plasma concentration-time profile of a patient that received a 400 mg single oral dose of 2ME2...... 83

Figure 18: Plasma concentration-time profile of a patient that received a 800 mg single

oral dose of 2ME2...... 84

Figure 19: Plasma concentration-time profile of a patient that received a 1600 mg single

xiv

oral dose of 2ME2...... 85

Figure 20: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2...... 86

Figure 21: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2...... 87

Figure 22: Box and whisker plots of Cmax after a single oral dose ...... 90

Figure 23: Box and whisker plots of AUClast after a single oral dose ...... 90

Figure 24: Box and whisker plots of CL/F after a single oral dose ...... 91

Figure 25: Changes in Ki67 staining in pretreatment tissue biopsies and after 2 months of

2ME2 treatment ...... 96

Figure 26: Changes in microvessel density staining in pretreatment tissue biopsies and after 2 months of 2ME2 treatment...... 97

Figure 27: Chromatograms of human liver microsomal incubations of 2ME2 optimized for hepatic phase I metabolism with NADPH generating system ...... 112

Figure 28: Chromatograms of human liver microsomal incubations of 2ME2 optimized for hepatic phase I metabolism without NADPH generating system ...... 113

Figure 29: Michaelis-Menten Kinetics for Metabolite A ...... 115

Figure 30: Michaelis-Menten Kinetics for Metabolite B...... 116

Figure 31: Michaelis-Menten Kinetics for Metabolite C...... 117

Figure 32: Michaelis-Menten Kinetics for Metabolite D ...... 118

Figure 33: Metabolites of 2ME2 produced by phase I hepatic microsomal incubations.121

Figure 34: In vitro metabolism of 2ME2 by CYP 1A1 ...... 123

Figure 35: Negative Control for CYP 1A1 incubations...... 124

Figure 36: In vitro metabolism of 2ME2 by CYP 1A2 ...... 125

Figure 37: Negative Control for CYP 1A2 incubations...... 126

Figure 38: In vitro metabolism of 2ME2 by CYP 3A4 ...... 127

Figure 39: Negative Control for CYP 3A4 incubations...... 128

Figure 40: In vitro metabolism of 2ME2 by CYP 3A5 ...... 129

Figure 41: Negative Control for CYP 3A5 incubations...... 130

Figure 42: In vitro metabolism of 2ME2 by CYP 2E1...... 131

Figure 43: Negative Control for CYP 2E1 incubations...... 132

Figure 44: In vitro glucuronidation of 2ME2 in human liver microsomes...... 135

Figure 45: Negative control for in vitro glucuronidation of 2ME2 in human liver microsomes ...... 136

Figure 46: Michaelis-Menten Kinetics for Glucuronide 1...... 137

Figure 47: Michaelis-Menten Kinetics for Glucuronide 2...... 138

Figure 48: In vitro sulfation of 2ME2 with human liver S9 fraction...... 141

Figure 49: Negative control for in vitro sulfation of 2ME2 with human liver S9 fraction ...... 142

Figure 50: Time course to reach equilibrium for the fraction unbound 2ME2...... 157

xvi

Figure 51: Plasma protein binding of 2ME2. Plot of bound 2ME2 (Cb) versus unbound

2ME2 (Cu)...... 158

Figure 52: Plot of 2ME2 Cb / Cu versus Cb in plasma from healthy human volunteers ..159

Figure 53: Histogram depicting binding of 2ME2 to human plasma proteins ...... 161

Figure 54: The mean (symbol) ± SD (error bars) fraction unbound 2ME2 in plasma

versus time profiles...... 165

Figure 55: Concentration-time profiles of total 2ME2 (ng/mL) and unbound 2ME2

(ng/mL) in plasma...... 166

Figure 56: Solubility profile of 2ME2 ...... 173

Figure 57: Plasma concentration-time profile of a patient that received a 400 mg single

oral dose of 2ME2...... 198

Figure 58: Plasma concentration-time profile of a patient that received a 800 mg single

oral dose of 2ME2...... 199

Figure 59: Plasma concentration-time profile of a patient that received a 1600 mg single

oral dose of 2ME2...... 200

Figure 60: Plasma concentration-time profile of a patient that received a 1600 mg single

oral dose of 2ME2...... 201

Figure 61: Plasma concentration-time profile of a patient that received a 1600 mg single

oral dose of 2ME2...... 202

Figure 62: Plasma concentration-time profile of a patient that received a 1600 mg single

xvii

oral dose of 2ME2...... 203

Figure 63: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2...... 204

Figure 64: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2...... 205

Figure 65: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2...... 206

Figure 66: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2...... 207

Figure 67: Plasma concentration-time profile of a patient that received a 3000 mg single oral dose of 2ME2...... 208

Figure 68: Blank human urine after liquid-liquid extraction...... 210

Figure 69: Calibration standard of 2ME2 spiked in human urine at 5 ng/mL...... 210

Figure 70: Calibration standard of 2ME2 spiked in human urine at10 ng/mL...... 210

Figure 71: Calibration standard of 2ME2 spiked in human urine at 20 ng/mL...... 210

Figure 72: Calibration standard of 2ME2 spiked in human urine at 2 µg/mL...... 211

Figure 73: Urine of patient #2 after liquid-liquid extraction ...... 211

Figure 74: Urine of patient #2 after treatment with β-glucuronidase and liquid-liquid extraction...... 211

Figure 75: Urine of patient #2 after treatment with sulfatase and liquid-liquid extraction212

xviii

Figure 76: Urine of patient #3 after liquid-liquid extraction ...... 212

Figure 77: Urine of patient #3 after treatment with β-glucuronidase and liquid-liquid extraction...... 212

Figure 78: Urine of patient #3 after treatment with sulfatase and liquid-liquid extraction213

Figure 79: Urine of patient #5 after liquid-liquid extraction ...... 213

Figure 80: Urine of patient #5 after treatment with β-glucuronidase and liquid-liquid extraction...... 213

Figure 81: Urine of patient #5 after treatment with sulfatase and liquid-liquid extraction214

Figure 82: Urine of patient #6 after liquid-liquid extraction ...... 214

Figure 83: Urine of patient #6 after treatment with β-glucuronidase and liquid-liquid extraction...... 214

Figure 84: Urine of patient #6 after treatment with sulfatase and liquid-liquid extraction215

Figure 85: In vitro metabolism of 2ME2 by CYP 1B1...... 217

Figure 86: Negative control for CYP 1B1 ...... 218

Figure 87: In vitro metabolism of 2ME2 by CYP 2A6 ...... 219

Figure 88: Negative control for CYP 2A6...... 220

Figure 89: In vitro metabolism of 2ME2 by CYP 2B6...... 221

Figure 90: Negative control for CYP 2B6 ...... 222

Figure 91: In vitro metabolism of 2ME2 by CYP 2C19...... 223

Figure 92: Negative control for CYP 2C19 ...... 224

xix

Figure 93: In vitro metabolism of 2ME2 by CYP 2C9...... 225

Figure 94: Negative control for CYP 2C9 ...... 226

Figure 95: In vitro metabolism of 2ME2 by CYP 2D6 ...... 227

Figure 96: Negative control for CYP 2D6...... 228

Abstract

By Nehal J. Lakhani, M.B; B.S., Ph.D.

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at Virginia Commonwealth University.

Virginia Commonwealth University, 2005

Major Director: Jürgen Venitz, M.D., Ph.D. Associate Professor and Vice Chairman, Department of Pharmaceutics

Major Director: William D. Figg, Pharm.D., M.B.A. Head, Clinical Pharmacology Research Core & Molecular Pharmacology Section, National Cancer Institute, National Institutes of Health

The goal of this research was to determine safety, tolerability and pharmacokinetics of

2ME2 in patients with solid tumors and determine maximum tolerated dose (MTD). The following hypotheses were tested: 1) 2ME2 will be well tolerated in clinic when given

orally and will have quantifiable effects on the ex vivo markers of angiogenesis and xxi

apoptosis; 2) 2ME2 will exhibit linear pharmacokinetics; 3) Plasma protein binding will

be extensive and linear; 4) Sulfation will be the major metabolic pathway for 2ME2.

This was a phase I dose escalation study. Twenty patients with refractory solid tumors

were enrolled. 2ME2 was administered orally starting at 400 mg bid with dose escalation

up to 3000 mg bid. Pharmacokinetic sampling was done up to 50 hours after single oral

dose for characterization of pharmacokinetics and plasma drug concentrations which

were determined by liquid chromatography tandem mass-spectrometry [LC/MS/MS,

LOQ: 1ng/mL]. Circulating plasma concentrations were very low at all dose levels with

high interindividual pharmacokinetic variability. Median plasma half-life was about 1-2

days. The unphysiologically high oral CL/F and Vd/F values reflect low oral

bioavailability of 2ME2. There was no dose proportional increase in Cmax or AUClast.

There were no dose limiting toxicities at highest dose level, therefore MTD was not defined. The clinical trial was closed due to the extremely low plasma concentrations of

2ME2 achieved.

Hepatic in vitro metabolism studies showed that 2ME2 was metabolized by CYP 450 enzymes (CYP 1A1, 1A2, 3A4, 3A5 and 2E1) to four major metabolites. Hepatic phase

II metabolism studies revealed two major glucuronide metabolites of 2ME2. Sulfation did not play a major role in metabolism of 2ME2. Total in-vivo hepatic clearance was estimated as 862 mL/min, primarily due to glucuronidation. Less than 0.01 % of total administered dose of 2ME2 was excreted unchanged in urine, and only about 1% was excreted as glucuronides.

Plasma protein binding of 2ME2 was studied using equilibrium dialysis. Mean unbound fraction of 2ME2 (fu) in plasma of patients and healthy human volunteers was

xxii

0.019 ± 0.0043 and 0.027 ± 0.0019 respectively. Binding was concentration-independent

and unaffected by presence of 2-methoxyestrone. 2ME2 binds to albumin, α1-acid glycoprotein (AAG) and sex-hormone binding globulin (SHBG).

CHAPTER ONE

Introduction

Estrogen Biosynthesis

The endogenous sources of estrogens are the ovary and the adrenal glands in the

female during pre-menopausal, non-pregnant life. During the post-menstrual period, the

major sources of estrogen are adipose tissue and the skin 1. Androgens (C19 steroids) are the common chemical precursors of estrogens (C18 steroids); 17β-hydroxysteroid dehydrogenase (17β-HSD) converts androstenedione to testosterone, which in turn is rapidly converted to estradiol (E2) by demethylation at the C19 position followed by aromatization, which can be catalyzed by both aromatase (CYP19) and 17β-HSD 2. Other than estradiol, the ovary also secretes estrone (E1), which serves as an alternative source for estradiol; 17β-HSD type 1 primarily catalyzes the reduction of estrone to estradiol 3, whereas 17β-HSD type 2 primarily catalyzes the reverse metabolic transformation 3. In case of males, the Leydig cells and germ cells of the testes produce estrogen 4.

The other pathway which plays an important role in estrogen biosynthesis is the peripheral conversion of C19 precursors in adipose tissue to free estrogens in obese women and men 5,6. This is the predominant pathway of estrogen biosynthesis in post-

menopausal life. Although these extragonadal tissues have the capacity to convert C19

steroids to C18 steroids, unlike the ovaries, they lack the ability to synthesize C19

precursors. Hence, estrogen production in the extragonadal sites is totally dependent on 2

circulating C19 precursors 1. Estriol (E3) is a peripheral metabolite of estradiol and estrone, and is considered a detoxification (or inactivation) product since it is less active than estrogen. Oral contraceptive pills and estrogen replacement therapy are exogenous sources of estrogens.

Estrogen Metabolism

Phase I Metabolism

The metabolism of endogenous estradiol primarily involves cytochrome P450 (CYP)- dependent hydroxylations at positions C2, C4, or C16, yielding either 2-hydroxy- or 4- hydroxyestradiol, or 16α-hydroxyestradiol, respectively. The 2-hydroxylation of estradiol

and estrone, catalyzed by the CYP1A1/2 and CYP3A isozymes 7-9, is the major metabolic pathway in the liver as compared to the 4-hydroxylation, which is exclusively catalyzed by CYP3A 10. In humans, as much as 50% of estrogen is hydroxylated at the C2 position,

with men exhibiting less extensive metabolism at this position than women 11. The 2-

hydroxyl metabolites of estradiol and estrone are further metabolized by a catechol-o-

methyl transferase (COMT), which converts these products to 2-methoxyestradiol (2-

ME2) and 2-methoxyestrone, respectively. COMT is found in many tissues including the

uterus, liver, kidney, breast, lymphocytes and erythrocytes 12,13.

4-Hydroxyestradiol has been associated with renal cancer in Syrian male hamsters

14. This metabolite also undergoes metabolic redox cycling to generate free radicals such

as superoxides and chemically-reactive (semi-)quinone estrogen intermediates, which

may damage DNA and other cellular components, induce cell transformation and initiate

tumorigenesis 15,16. The (semi-) quinones can also form conjugates with glutathione, the

formation of which is catalyzed by glutathione S-transferase (GST), or they can be

reduced to catechol estrogens by quinone reductases. DNA adducts can result from these

reactive metabolites if the inactivating conjugative pathways are incomplete or absent.

Quinone metabolites formed during redox cycling bind covalently to DNA, and may

produce several forms of DNA damage, including single-strand breaks and/or

hydroxylation of guanine bases of DNA 15.

The role of 16α-hydroxyestradiol in carcinogenesis has not been clearly established. This metabolite is produced in abundance during pregnancy, and women who have been pregnant have a reduced risk of breast cancer. Two epidemiologic studies suggested that the 16α-hydroxylation of estradiol was more pronounced in women with breast cancer 17 and in women with a high familial risk of breast cancer 18, than in controls. However, a third study found no elevation of the 16α-hydroxylation pathway in patients with breast cancer as compared to controls 19.

Phase II metabolism

The phase II metabolism of estrogens involves sulfation, glucuronidation, glutathione conjugation, and o-methylation. Sulfation of steroids is an important biotransformation step involved in the regulation of tissue specific estrogen levels. All of

the hydroxylated metabolites produced by phase I reactions can undergo sulfation via the

estrogen sulfotransferase (SULT) isoform 1A1, a cytosolic enzyme that also conjugates

estradiol and estrone. Estrone sulfatase, a membrane bound enzyme, is responsible for the

desulfation of estrone sulfate and exhibits its highest activity in the liver 20. The sulfated forms of many steroid hormones exhibit half-lives up to ten-fold longer than the

desulfated form in plasma. Biological "cycling" of sulfated/desulfated steroid hormones

has been demonstrated. The intracellular sulfated moiety represents a readily accessible,

yet biologically inactive, storage form for many steroid hormones whereby desulfation

regenerates the biologically active steroid. The transport of estrone sulfate into steroid

hormone responsive cells is not well understood; however some studies have shown that

a human organic anion transporter (Oatp1) has high affinity for both sulfate and

glucuronide estrogen conjugates 21. SULT 1E1 catalyzes the sulfation of 17β-estradiol and SULT1A1 catalyzes the sulfation of 2ME2 22, while SULT2B1 is known to catalyze the sulfation of dihydroxy-epi-androsterone. An enzyme known as steroid sulfatase is involved in the desulfation of various conjugated products, including sulfate conjugates

of estrone, 17β-estradiol, and dihydroxy-epi-androsterone.

Glucuronidation by uridine-5’-diphosphate glucuronosyltransferase (UGT) of

estrogens is generally accepted to play a principal role in elimination pathways followed

by excretion of the conjugates through urine and feces. In a study comparing UGT

activity in matched breast ductal carcinoma and peritumoral tissues, activity was reported

in tissues from only 4 out of 12 individuals studied 23. Furthermore, in those 4 paired samples, the level of activity was 5-fold lower in the tumor tissue than in the peritumoral tissue, suggesting that glucuronidation does play an important role in detoxification of estrogens.

Results of previous studies indicate that the UGT1A1, UGT1A3, and UGT2B family of isoenzymes catalyze catechol estrogen glucuronidation 24,25. When the glucuronidation kinetics of expressed human UGTs with catechol estrogens was investigated, it was found that UGT2B6(Y) reacted with a higher efficiency toward 4-

hydroxyestrogen catechols, whereas UGT1A1 and UGT1A3 showed higher activities

toward 2-hydroxyestrogens 26. It has also been shown that enzymatic de-conjugation of

estrogen glucuronides occurs by β-glucuronidases, and that it may be an important source

of both primary estrogens and their catechol metabolites in preclinical models 27.

COMT catalyzes the o-methylation of catechol estrogens 14,28-30 by transferring a methyl group from the co-factor S-adenosyl methionine to the 2-hydroxyl or 4-hydroxyl groups, and forms 2-ME2 and 4-methoxyestradiol 31. This enzyme is found in many

tissues including the uterus, liver, kidney, breast, lymphocytes and erythrocytes 12,13.

Another important function of COMT is the o-methylation and inactivation of endogenous catecholamines and catechol drugs such as levodopa and methyldopa.32,33 It has been shown that o-methylation of 2-hydroxyestradiol by COMT is protective against its oxidative DNA damage in MCF-7 cells in vitro 34. Figure 1 provides an overview of

the various metabolic routes of estrogens (modified from 35) and Figure 2 depicts the

chemical structure of 2-methoxyestradiol (2ME2).

Figure 1: Estrogen Biotransformation Scheme

OH 16- α -hydroxy E1 (E2)

CYP3A4 CH3O Inhibition of E2-mediated - O3SO Carcinogenesis HO 2-methoxy estradiol s e s 2-OCH 3 E2 (E1) a O r Glutathione Conjugate e f s n a r t o

f T l Estrone (E1) u S COMT n GS e g o H / r t s GS OH E 12 17 11 13 16 - O2 O2 1 9 HO O O 10 14 CYP1A1 / 1A2 2 8 15 Peroxidases

7 CYP3A4 or CYP450 3 5 HO HO HO 4 6 O Estradiol 2-hydroxy estradiol E2-2,3-semiquinone E2-2,3-quinone (E2) 2OH-E2 (E1)

Binding to ER CYP1B1 Stimulation of Cell Proliferation

O 2 O 2 Peroxidases or DNA DNA Adducts CYP450 O HO HO COMT O - O OH E2-3,4-semiquinone E2-3,4-quinone 4-hydroxy estradiol 4OH-E2 (E1) HO GSH GST OCH3 4-methoxy estradiol 4-OCH 3 E2 (E1) Glutathione Conjugate

Figure 2: Chemical Structure of 2-methoxyestradiol

H3CO

2-methoxyestradiol (2-ME2)

Endogenous 2ME2 production:

2-Methoxyestradiol (2ME2) is synthesized in vivo by hydroxylation at the 2-position of estradiol, and subsequent catechol-O-methyltransferase mediated O-methylation.

Endogenous plasma concentrations of 2ME2 are in the picomolar range under normal physiological conditions; however, during late pregnancy the values can increase more than 1000-fold 36. Table 1 provides a detailed synopsis of the levels of 2- methoxyestradiols in humans in various physiological conditions (adapted from 37).

Table 1: Endogenous concentrations of 2-methoxyestrogens in human serum in

various physiological conditions:

Median Range N

(pg/mL) (pg/mL)

Men

19-58 years <10.3 <10.3-35.5 22

Women

Follicular phase 46 18-63 8

Luteal Phase 70 31-138 8

Postmenopausal 10 21-76 10

Pregnant

11-16th week 674 216-1,678 46

37th-40th week 3,768 2,035-10,691 34

In labor 3,580 1,353-9,974 41

Newborn

Cord serum 1,608 575-3,095 41

a Abbreviations: N, total number of replicate observations

Mechanisms of antitumor action of 2ME2

The antitumor activity of 2ME2 is thought to be due to the following components:

1) antiproliferative and antimetastatic activity ; 2) anti-angiogenic activity.

Antiproliferative & Antimetastatic Activity:

2ME2 inhibits the growth of pancreatic cancer cell lines (PaTu 8902, 8988t and

8988s) by 50-90% in a dose- and time-dependent fashion 38. Also, studies in an in vivo murine lung metastasis model, showed that 2ME2 inhibited the number of lung colonies by 60% as compared with controls 38.

Xenograft experiments found that 2ME2 targets both the tumor and the tumor

vasculature. Oral administration of 2ME2 was effective at reducing the rate of xenograft

tumor growth and tumor vascularization in mice in the absence of significant toxicity.

After 2 weeks of treatment with 100 mg/kg/day, the tumor growth of MethA sarcoma and

B16 melanoma cells was inhibited by 66% and 88%, respectively 39. A concomitant reduction in tumor vascularization was also observed 39. Also, interesting is the fact that the antitumor activity is not estrogen receptor dependent 40. The estrogen receptor (ER) negative human breast cancer line, MDA-MB-435 was found to be sensitive to 2ME2

treatment. Binding studies using recombinant proteins showed that 2ME2 has a Ki of 21

nM for ERα and a Ki of 417 nM for ERβ. E2 has Ki of 0.042 nM for ERα and a Ki of 0.132 nM for ERβ. Thus, 2ME2 has 500- and 3200-fold lower affinity than E2 for ERα and ERβ respectively, and the antiproliferative activities of 2ME2 do not require ERs and are not

influenced by ER antagonists or agonists 41. Table 2 shows a comprehensive list of the cell lines that are inhibited by 2ME2:

Table 2: Cell lines inhibited by 2ME2

Cell Type Inhibitory Concentrations References

(µM) (µg/mL)

______

Human Tumor

Lung (HOP-62) 0.7 0.21 42

Lung (H460) 5.0 1.51 43

Lung (A549) 5.0 1.51 43

Colon (HCT-116) 0.47 0.14 43

CNS (SH-SY5Y) 1.3 0.39 42

CNS (SF-539) 0.32 0.10 42

Melanoma (UACC-62) 0.36 0.11 42

Ovarian (OVCAR-3) 0.21 0.06 42

Renal (SN12-C) 0.95 0.29 42

Prostate (DU-145) 1.8 0.54 42

Breast (MDA-MB-435) 0.08-0.6 0.02-0.18 40, 44

Breast (MDA 231) 1.03 0.31 44

Breast (MCF-7) 0.45 0.14 44

Lymphoblast (Jurkat) 0.3 0.09 45

Lymphoblast (TK6) 1-2 0.3-0.6 46

Lymphoblast (WTK1) 1-2 0.3-0.6 46

Human Non Tumor

Skin Fibroblast (HFK2) 2.0 0.6 39

Human Endothelial

HUVEC 0.45 0.14 44

Non-Human-Tumor

Lung (Lewis Lung-Murine) 1.68 0.51 44

Melanoma (B16BL6-Murine) 0.4 0.12 44

Melanoma (B16F10-Murine) 0.3 0.09 44

Endothelial (EOMA-Murine) 0.89 0.27 44

Endothelial(H5V-Murine) 1.0 0.30 47

Non-Human Non Tumor

Lung (V79-Hamster) 3 0.91 48

Ovarian (Granulosa-Porcine) 3 0.91 49

Smooth Muscle (Aorta-Rabbit) 1 0.3 50

Adipocytes (Murine) 1.7 0.51 51

Non-Human Endothelial

Brain Capillary (Bovine) 0.19-0.49 0.06-0.15 40

Pulmonary artery (Bovine) 0.5 0.15 52

Anti-Angiogenic Activity:

Angiogenesis has been shown to be a critical step in the proliferation of cancers with size > 1 mm3 53,54. 2ME2 reduces tumor vasculature in mice injected subcutaneously with Meth A sarcoma and B16 melanoma cells and treated orally with 2ME2 39, 40. In vivo antiangiogenic activity of 2ME2 has been demonstrated in the corneal micropocket

40 and chick chorioallantoic model (CAM) systems 52. In the corneal micropocket VEGF and bFGF induced neovascularization was reduced by 54% and 39%, respectively. In the

CAM model, 2µM of 2ME2 exposure for 3 days completely prevented bFGF-induced angiogenesis. In vitro 2ME2 has been found to inhibit the neovascularization developing

from the rat aortic ring assay 55, 56. 2ME2 appears to affect the angiogenesis cascade at

various steps. It also blocks the tubule formation as well as the invasion through the

collagen matrix 57.

No difference in the number of micro-vessels (CD 31 staining) within MIA PaCa

(Pancreatic Cancer) pulmonary metastases was noted in those animals treated with

1mg/day of 2ME2 versus control 38. Hence, ambiguity exists as regards to the exact mechanism of antiangiogenic properties of 2ME2. Although, 2ME2 has been labeled as an anti-angiogenic agent, endothelial cells are not necessarily more sensitive to its antiproliferative effects. The anti-angiogenic and apoptotic activity of 2ME2 varies with cell type and the regulatory microenvironment 36. For some murine tumor models

administration of oral 2ME2 reduces the rate of tumor growth without signs of toxicity

with a concomitant reduction in tumor vascularization 39. Thus, there is sufficient data to

establish that 2ME2 effectively inhibits angiogenesis, but it is still difficult to isolate the

exact mechanism of action.

Mechanisms of Action of 2ME2:

The above-mentioned antitumor activities of 2ME2 result from the following

mechanisms of action: 1) Apoptotic Activity; 2) Microtubule Activity; 3) Production of

superoxides

Apoptotic Activity

2ME2 appears to induce Caspase-3 activation and causes apoptosis in gastric

carcinoma cell lines (SC-M1 and NUGC-3) through the caspase activation cascade

46 resulting in G2/M arrest . When gastric carcinoma cell lines were treated with 2ME2 for

24 hours, it resulted in a reduction of cell viability in a dose- dependent and a time-

dependent fashion. Moreover, the caspase cascade induced by 2ME2, which is believed

to be the major mechanism causing apoptosis, begins with the activation of caspase-8

followed by caspase-3 and eventually induces DNA fragmentation. 2ME2 induced

apoptosis requires caspase activation and the sequential activation of caspase 8, 9 and 3 is

consistent with the triggering of apoptosis through the membrane bound receptor 58.

Flow cytometry established 2ME2 induced G2/M cell cycle arrest.

In some studies the amount of apoptosis detected was correlated with the antiproliferative activity of 2ME2. Apoptotic cell death induced by 2ME2 has been shown to be p53 mediated 43. Increased activity of functional p53 has been observed after treatment with 2ME2. The number of metastatic lung colonies has been found to

synergistically reduced in mice treated with 2ME2 (50 mg/kg/day) in combination with

adenoviral p53 59. On the other hand, there are several other reports indicating that the induction of apoptosis by 2ME2 may be p53-independent 38, 60. 2ME2 treatment results in upregulation of death receptor 5 (DR5) expression both in vitro and in vivo 58.

Moreover, blocking death receptor signaling by expression of dominant-negative FADD severely attenuates the ability of 2ME2 to induce apoptosis.

Microtubule Activity

A clear increase in the mitotic figures has been demonstrated in cells treated with

2ME2 during the metaphase 45, 48. In endothelial cells, 2ME2 caused selective disruption of microtubules as opposed to other cytoskeletal structures 60. Also, micromolar

concentrations caused disruption of microtubular network in non-synchronized Chinese

hamster V79 cells 48 and multipolar and irregular spindles in synchronized MCF-7 cells.

On the contrary, 2ME2 has been found to inhibit aromatase activity, which is a characteristic observed with drugs that stabilize tubulin and not with agents that inhibit

tubulin polymerization 61.

2ME2, in superstoichiometric concentrations, inhibits the nucleation and propagation of phases of tubulin assembly, but does not affect the reaction extent 62. Similarly, in

substoichiometric concentrations, 2ME2 completely inhibits polymerization. 2ME2

inhibits the rate but not the degree, of polymerization and depolymerization of tubulin

induced by other agents. The inhibition of colchicines binding to tubulin was found to be

competitive in nature (Ki of 22 µM). Thus, it is speculated that 2ME2 binds to the

colchicine site of tubulin, and depending on the reaction conditions, either inhibits

assembly or seems to be incorporated into a polymer with altered stability properties 62.

Production of Superoxides

There have been two contradictory reports over this subject matter. Huang et al 63 have reported that 2ME2 and 2-hydroxyestradiol inhibit both Cu-, Zn- and Mn-

superoxide dismutases and are toxic to human leukemia cells. Kachadourian et al 64 showed that, when human leukemia HL-60 cells were incubated in a medium containing

2ME2, cells showed 53% decrease in aconitase activity (aconitases are very sensitive to superoxide mediated inactivation) as compared to controls. Paraquat (superoxide generating redox cycling agent) decreased aconitase activity in HL-60 cells by 67%.

Kachadourian et al showed that, although 2ME2 supports the increased production of

superoxides, it does not produce that effect by inhibiting superoxide dismutase.

Cardioprotective effects:

It has been suggested for years that estradiol may protect premenopausal women

against coronary artery disease 60. Recent studies hint at the possibility of these

cardioprotective effects being independent of the estrogen receptor status 65, 66. 2- hydroxyestradiol (2-OHE2) and 2ME2 are two metabolites of estrogen, which possess little or no affinity for ERs. Recent studies also indicate that these two metabolites, 2-

OHE2 and 2ME2, are more potent than estradiol in preventing vascular smooth muscle

cell (VSMC) growth 3.

When the local metabolism of estradiol to methoxyestradiols was tested in human

vascular smooth muscle cells (VSMCs) 67 it was reported that methoxyestradiols and their precursors, hydroxyestradiols, are 1) more potent than estradiol in inhibiting VSMC

growth 2) the inhibitory effects of estradiol on VSMC growth are enhanced by CYP450

inducers; and 3) the inhibitory effects of estradiol, both in presence and absence of

CYP450 inducers, are abolished by CYP450 and COMT inhibitors.

Estrogenic Activity of 2ME2

38 2ME2 has 2000 fold lower affinity for the estrogen receptors (ER) than estradiol .

When administered continuously at 1µg/hr to ovariectomized rats, 2ME2 was ineffective in sustaining uterine growth and failed to induce estrus in ovariectomized mice. In contrast, 2-hydroxyestrone, 4-methoxyestrone and estradiol had a strong uterotropic

38 activity . With chronic administration, 2ME2 did not affect the seminal vesicle weights,

while estradiol significantly reduced the seminal vesicle weights in male mice. Similarly,

no tumors were reported in male Syrian hamsters receiving long-term exposure to 2ME2

as compared to estrogens, which caused the induction of renal tumors.

Estradiol, 4-hydroxyestradiol and 2-hydroxyestradiol were found to sustain tumor

growth of an estrogen-dependent hamster kidney tumor cell line, H-301 cells 39. Also,

MCF-7, which is an estrogen dependent cell line, showed sustained proliferation of these cells at low concentration of estradiol (0.1nM) while concentrations up to 10 µM of 2ME2

were not able to sustain the proliferation of these cells 68.

However, in toxicology studies, high doses of 2ME2 administered orally to rats daily

for 14 and 28 days and dogs treated for 28 days showed estrogenic activity 68. It is

speculated that this estrogenic activity may be due to the formation of

2-hydroxyestradiol and possibly from other not yet identified metabolites of 2ME2

interacting with ER.

Preclinical Pharmacokinetics of 2ME2

Preliminary pharmacokinetic studies with 2ME2 were carried out in male and female

CD2F1 mice using oral, intravenous or subcutaneous administration of labeled

[3H-2-ME2] and non-labeled compound. Following oral administration of 100 mg/kg to male mice, Cmax of 190ng/mL was observed in plasma 12 minutes after dosing. The drug

concentrations in plasma fell below 50ng/mL within 1 hour and were undetectable after 2

hours. Similarly, in female mice, the pharmacokinetic profile of 2ME2 after oral

administration demonstrated a rapid absorption and elimination pattern, although values

for Cmax (35ng/ml) and tmax (30 minutes) were different from the values observed in

males. Following intravenous administration of 3mg/kg, the plasma concentrations were

undetectable within 60 minutes for both the male and the female populations. There was

a sizeable difference in the half-lives estimated between the male (t1/2 = 285 min) and female (t1/2 = 114 min) populations. Table 3 provides a summary of pharmacokinetic parameters from preclinical studies:

Table 3: Summary of pharmacokinetic parameters after preclinical studies in mice:

Parameters Oral Intravenous Subcutaneous (100 mg/kg) (3mg/kg) (20mg/kg) Male Female Male Female Male Female

AUC (µM*min) 35.9 53.8 112 84 34.6 232

Cmax (ng/ml) 190 35 >500 > 500 ND ND

tmax (min) 12 30 ND ND ND ND

t1/2 β (min) ND ND 285 114 ND ND

CLtot (mL/min) ND ND 22 29 ND ND

Foral (%) 0.9 1.9 ND ND ND ND

Abbreviations: ND, not determined

Total radioactivity levels could be measured up to 60 hours post administration of

oral, intravenous and subcutaneous doses and the concentrations were at least 10- to 100-

fold higher than the parent compound. Thus, suggesting that 2ME2 gets metabolized

extensively, and the metabolites are detectable in plasma for a long period of time.

Metabolism (in vitro studies)

The in vitro metabolism of 2ME2 by pooled male and female liver microsomes of rats, mice, dogs and humans was investigated, it was found that the normalized half life in mice, rats and humans was less than 10 minutes. The half-life in case of dogs was found to be 29 minutes. 2-methoxyestrone (2-ME1) was identified as a major metabolite of human and mouse liver microsomes. The other metabolites identified in human liver microsomes were 2-hydroxyestradiol (2-OHE2) and 2-hydroxyestrone (2-OHE1). When

2ME2 was incubated with cDNA-expressed cytochrome P450s (CYP 1A1, 1A2, 2A6,

2B6, 2C9, 2C19, 2D6, 2E1 and 3A4) with NADPH generating system, significant

metabolism was observed with CYP1A2 (91% degradation), followed by CYP3A4

(35%), CYP2C9 (33%) and CYP2E1 (31%).

Glucuronidation of 2ME2 has been observed in UDPG-acid fortified mouse and

human microsomal preparation preparations. Evidence of polar conjugates has been

observed in vivo in murine urine after 2ME2 administration.

Plasma protein binding (in vitro studies)

In vitro plasma protein binding of 2ME2 evaluated by using murine and human

plasma has been found to be more than 99% 68.

In-vivo Excretion

Excretion studies in male and female CD2F1 mice given a single dose of radiolabeled

[3H]-2-ME2 were performed and urine and feces were collected for 24 hours following administration. The recovery of total radioactivity in case of males was 42%

(urine=17%, feces=25%) and 54% in case of females (urine=29%, feces=25%).

Pharmacokinetics of orally administered estrogens in humans

The most commonly prescribed estrogen component of Hormone Replacement

Therapy (HRT) is Premarin® (Wyeth-Ayerst Laboratories, Canada) which is a mixture of ten biologically active conjugated equine estrogens 69. Tables 4 and 5 depict the

pharmacokinetic parameters of estrone and 17 β-estradiol, two major estrogens in

Premarin after a single oral dose of two tablets of 0.625 mg.

Table 4 Total Estrone:

Cmax (ng/mL) 4.01 ± 1.7

tmax (h) 8.0 ± 2.3

AUCt (ng*h/mL) 60.5 ± 24.8

AUC∞ (ng*h/mL) 63.8 ± 26.8 t1/2 (h) 14.8 ± 3.9

Table 5 Total 17 β-estradiol:

Cmax (ng/mL) 0.43 ± 0.2

tmax (h) 10.3 ± 5.1

AUCt (ng*h/mL) 6.0 ± 2.5

AUC∞ (ng*h/mL) 6.4 ± 2.4 t1/2 (h) 15.9 ± 5.8

Absorption

Estrogens are generally well absorbed orally 70. Peak plasma concentration of 4 ng/mL occurs approximately 8 hours after a dose of two 0.625 mg tablets of conjugated estrogens 69. The bioavailability is poor because of significant first pass effect by 71, which begins with sulfation of estrogens in the gastrointestinal tract. Estrogens are then metabolized to inactive compounds in the liver (sulfates and glucuronides) and excreted in the urine and bile. Estradiol is excreted in the urine as the sulfate and glucuronide esters along with a small amount of unchanged drug.

Distribution

Dunn et al72 have established that in human females 37 % of estradiol is SHBG bound,

61% is albumin bound and 1.8% is free while in human males 78 % of estradiol is

albumin bound, 20% is SHBG bound and 2.3% is free. Estriol is not bound to plasma sex

hormone-binding globulin (unlike estradiol or estrone) (Padwick 1986, Cardozo 1998).

Metabolism and Excretion

Estradiol is rapidly metabolized in the liver to less active estriol and estrone. It Orally

administered estradiol is excreted in the urine as the sulfate and glucuronide esters along

with a small amount of unchanged drug 70. Circulating estrogens exist in a dynamic

equilibrium of metabolic interconversions. Estradiol is converted reversibly to estrone,

and both can be converted to estriol, which is the major urinary metabolite. Studies of

quantification of biliary and urinary metabolites of exogenous estrogens in human

subjects with biliary drainage after the administration of intravenous E3 found that 79% of the glucuronide conjugate was excreted in urine and 81% of the sulfate conjugate was

73 excreted in bile . Following the administration of intravenous estrone (E1) in humans

with biliary drainage also biliary products appeared to be composed mostly of sulfate

conjugates whereas the urinary products consisted mostly of glucuronidates74.

Preliminary Clinical Results

There are numerous 2ME2 clinical trials underway in the U.S. A clinical trial

being conducted at Indiana University, Indianapolis, IN has reported no grade IV toxicity

and only minor grade III toxicity which could be attributed to disease progression 75, following the administration of 200, 400, 600, 800 and 1000 mg/day to a total of 15 patients. The half-life of the drug in plasma was reported to be 10 hours. Some patients showed clinically significant reduction in bone pain and analgesic use while receiving

2ME2. All patients receiving 1000 mg (N=3) experienced hot flashes.

In a phase II multi-center randomized double blind trial of two doses (400 and 1200 mg/day) of 2ME2 in patients (N=33) with hormone refractory prostate cancer, 2ME2 was very well tolerated and PSA stabilization and declines were observed 76. In a small

number of patients, grade 2 and 3 liver function abnormalities were observed which

normalized rapidly after discontinuation of 2ME2. Analysis of growth ex-vivo factor data

indicates that there is an initial rise in the levels of plasma VEGF & bFGF from baseline

to month 1, followed by decrease of 40% and 55% respectively from month 1 to month 3

on study.

A phase I study of 2ME2 plus docetaxel in patients with metastatic breast cancer

(N=15) was performed at Indiana University 77. Docetaxel at 35 mg/m2 was administered

weekly for 4 of 6 weeks and 2ME2 was given orally (200-1000 mg/day) once daily for 28 days followed by a 13-day observation period in cycle one, continuously thereafter.

After a maximum of 6 cycles of combined therapy, responding or stable patients continued 2ME2 alone until progression. There were no grade 4 toxicities. Grade 3

fatigue, diarrhea and hand-foot syndrome occurred in 5, 4 and 1 patients respectively. 3

patients had grade 3 transaminase elevations that returned to normal with continued

treatment. 2ME2 did not alter docetaxel clearance or dose-normalized AUC. Extensive metabolism to 2-methoxyestrone (2ME1) was observed. 2ME2 trough and peak levels were not altered by concurrent docetaxel administration.

CHAPTER TWO

Hypotheses and Objectives

Overview

This study is a phase I clinical trial of a novel anticancer agent, 2-methoxyestradiol

(2ME2), with the intent of determining the safety, tolerability and pharmacokinetic and pharmacodynamic profile, as well as in vitro metabolism and plasma protein binding of

2ME2, in patients with solid tumors. The hypotheses that will be tested and the objectives of the study are as follows:

Hypotheses

1. 2ME2 will be well tolerated in the clinic when given orally at the proposed doses

and will have quantifiable effects on the ex vivo markers of angiogenesis and

apoptosis in the tumor cells.

2. 2ME2 will exhibit linear pharmacokinetics. Single-dose pharmacokinetics will

predict steady state pharmacokinetics.

3. The plasma protein binding will be extensive and linear in the clinically

achievable concentration range.

4. Hepatic metabolism by sulfation will be the major metabolic pathway responsible

for the metabolism of 2ME2.

Objectives

1. To develop and validate an LC/MS/MS assay that will quantitate 2ME2 in human

plasma, urine, β-glucuronidase and sulfatase treated urine; human liver

microsomes as well as pooled human cytosol and validate it.

2. To characterize the pharmacokinetics of oral 2ME2 in plasma and urine of

patients with solid tumors.

3. To perform pharmacodynamic correlations, if any, with the in-vivo and ex-vivo

markers of biological effect and/toxicity.

4. To assess the in vitro plasma protein binding of 2ME2.

5. To characterize the in vitro metabolism of 2ME2 in humans.

CHAPTER THREE

Bioanalytical method for quantitating 2-methoxyestradiol in Human Plasma and Urine

Introduction

Various methods have been described for the quantitative determination of 2ME2

in biological matrices. A radioimmunoassay using an 125I-labelled ligand was described to quantify 2-methoxyestrogens in human serum 37, but is marred by cross-reactivity with similar estrogenic compounds, i.e., 2-methoxyestrone, 2-methoxyestriol and 2- hydroxyestrone. Chromatographic methods for 2ME2 include an HPLC/UV assay for extracts from the vitreous humor of rabbit eyes78, a GC/MS assay for rat plasma79 and an

HPLC/UV method for human plasma80. Because of problems related to poor specificity at or near lower limit of quantitation of these assays and the need for adequate sensitivity in view of the lower circulating concentrations of 2ME2 in plasma of patients treated with the drug, it was considered essential to improve currently available procedures.

Hence, a simple, specific and rapid assay method was developed and validated for the determination of 2ME2 in human plasma and urine to characterize the clinical pharmacokinetics of 2ME2 was developed. The method discussed in this chapter describes the development and validation of a novel, sensitive and specific analytical method for quantification of 2ME2 in human plasma based on liquid chromatography with atmospheric pressure chemical ionization (APCI)–tandem mass spectrometric detection. 27

Experimental

Chemicals and reagents

2ME2 (HPLC purity, 99.99%) was obtained from Entremed Inc. (Rockville, MD,

USA). HPLC grade methanol and acetonitrile were obtained from Mallinckrodt Baker,

Inc (Phillipsburg, NJ, USA). The internal standard 2ME2-d5 (2-Methoxy-17β-estradiol-

1,4,16,16,17-d5) was purchased from C/D/N Isotopes Inc. (Quebec, Canada) and its chemical structure is depicted in figure 3. Water was filtered and de-ionized using Milli-

Q-UV plus system (Millipore, Milford, MA, USA). Drug-free human plasma was obtained from the Blood Bank at the Warren Grant Magnuson Clinical Center (Bethesda,

MD, USA).

Figure 3: Chemical structure of the internal standard 2ME2-d5

OH D D D

D O

Stock solutions and standards in Plasma

Stock solutions of 2ME2 and 2ME2-d5 (IS) were made in triplicate by dissolving

10 mg of drug in 10 mL of methanol, resulting in primary stock solutions containing 1 mg/mL of 2ME2 and IS. These solutions were stable and stored at -800C for up to 8

weeks. Working solutions of 2ME2 and IS were prepared on each analysis day by serial

dilutions in methanol from the primary stock solution. The difference in drug

concentration in each of the triplicate stock solutions, estimated from the mean peak area

following repeat analysis of a dilution of the stock, was determined to be within ±5%.

Six-point calibration samples containing 2ME2 at concentrations of 1, 5, 10, 20,

50 and 100 ng/mL were prepared daily by addition of aliquots of the working solutions to

drug-free human plasma. Three pools of quality control (QC) samples for 2ME2, used

for the evaluation of accuracy and precision, were prepared in human plasma at

concentrations of 3, 40 and 75 ng/ml. For freeze-thaw stability runs, the aliquots of

prepared quality control samples were stored at -800 C.

Sample preparation for plasma

A volume of 50 µL of a 1 µg/mL solution of 2ME2-d5 (IS) was added to 0.3 mL

of plasma. Next, 3 mL of ethyl acetate were added to the plasma sample spiked with IS.

The mixture was then vortex-mixed for 15 minutes followed by centrifugation at 4000

rpm for 15 minutes. The supernatant was taken off after centrifugation and evaporated to

dryness under a continuous stream of air at 370 C. The extracts were reconstituted in 60

µL of 50% methanol (in water) using vortex mixing, and 20 µL of the reconstitute was injected for LC/MS/MS analysis.

Equipment

Chromatography was performed on a Waters Alliance 2695 LC system (Waters

Corporation, Milford, MA, USA), which included a quarternary pump, with an online degasser, a refrigerated autosampler and column compartment. Separations were performed at ambient temperature on Agilent Zorbax Eclipse® XDB-C18 column (2.1 ×

50 mm i.d.; 5-µm; Waters) using a mobile phase composed of methanol and water. The

mobile phase was delivered as a gradient run over 9 minutes at a flow rate of 0.25

mL/min as shown in Table 6. A Micromass Quattro Micro detector (Micromass

Corporation, Manchester, UK) was used for tandem mass-spectrometry with atmospheric

pressure chemical ionization source operated in the positive ionization mode. The

instrument was operated in the MRM (Multiple Reaction Monitoring) mode monitoring

the ion transitions from m/z 303.1→136.8 (2ME2) and m/z 308.1→138.8 (IS). Data

processing and analysis were performed using Masslynx/Quanlynx software. Detector

settings were as follows: cone voltage, 22 V; extractor voltage, 2V; corona current, 5 µA;

30 source temperature, 1300 C; and desolvation gas flow, 310 L/h. The zero concentration sample (blank) was used to visually verify the purity of the reagents and the lack of other potentially interfering (endogenous) substances, but was not considered for the regression analysis of standards. The goodness-of-fit of various calibration models was evaluated by visual inspection, the correlation coefficient and a lack-of-fit test.

Table 6: Mobile Phase gradient used for LC separation:

Time MeOH Water Flow (min) (%) (%) (mL/min) 0.0 50 50 0.25 0.5 90 10 0.25 5.0 90 10 0.25 5.5 50 50 0.25 7.9 50 50 0.25 9.0 50 50 0.25

Chromatography and detection in Plasma

Typical chromatograms resulting from the LC-APCI-MS-MS analysis of 0.3-mL plasma extracts obtained after liquid-liquid extraction are depicted below, and include a blank plasma sample (Figure 4), a standard spiked at a 2ME2 concentration of 100 ng/mL

(Figure 5), and a standard spiked at a 2ME2 concentration of 1 ng/mL (Figure 6) . Figure

7 depicts a chromatogram of an extract of a plasma sample obtained from a patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, before receiving the

31 dose of 2ME2 and figure 8 shows the chromatogram 0.5 hour after receiving the dose of

2ME2. The retention time of 2ME2 and the IS was about 5.2 minutes with an overall run time of 9 minutes to allow for re-equilibration of mobile phase.

Figure 4: Blank plasma sample (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

Figure 5: Standard spiked at a 2ME2 concentration of 100 ng/mL (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

Figure 6: Standard spiked at a 2ME2 concentration of 1 ng/mL (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

Figure 7: Plasma sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, before receiving the dose of 2ME2 (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

Figure 8: Plasma sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, 0.5 hour after receiving the dose of 2ME2 (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

Methodology for Validation in plasma

Method validation with respect to accuracy and precision was performed

according to procedures described in detail elsewhere 81. The method was validated for plasma in terms of linearity of detector response, accuracy, precision, sensitivity, specificity, freeze-thaw stability, autosampler and re-injection stability, long term stability and short term stability. On each of four validation days, calibration curves were analyzed in duplicate along with quality control samples in replicates of five containing known concentrations of 2ME2. Statistical analysis was performed using the software package NCSS 2001 (J. Hintze, Number Cruncher Statistical Systems, Kaysville, UT,

USA). The extraction recovery was calculated as a percentage by comparing the peak areas of samples prepared at 3 ng/mL and 75 ng/mL in human plasma and the mobile phase (50% acetonitrile in water) in triplicate.

Response function

Calibration curves were constructed by least-squares linear regression analysis of peak area ratio of 2ME2 and the IS versus the drug concentration of the nominal standard

(x) with or without weighting. Calibration curves were fitted by weighted (1/x2) least- squares regression analysis. To establish the optimal quantification method and weight factor, the correlation coefficient of the fitted equation and the accuracy of back- calculated calibration concentrations were taken into consideration.

Accuracy, precision and recovery

Accuracy and precision were determined by analyzing quality control samples

with 2ME2 concentrations in the low, mid and high concentration ranges of the

calibration curve. Accuracy or percent deviation (DEV) was defined as percent difference

between the mean observed concentration and the nominal concentration:

⎛ [nominal] − [observed]⎞ DEV = ⎜ ⎟ ×100% ⎝ [nominal] ⎠

Non-lower limit of quantitation (LLOQ) standards and QC samples were considered reliable if at least two out of every three estimates of accuracy were within the range of 85-115%81.

The precision of the assay was assessed by the between-run and within-run

precision. The between-groups mean square (MSbet), the within-groups mean square

(MSwit), and the grand mean (GM) of the observed concentrations across run days were

obtained by one-way analysis of variance, using the run day as classification variable.

The between-run precision (BRP) was defined as:

(MS − MS ) / n BRP = bet wit ×100% GM where n represents the number of replicates within each validation run. The within-run precision (WRP) was calculated as:

MS WRP = wit ×100% GM

In order to perform the assay with reliable intermediate precision and repeatability,

both the BRP and WRP should not exceed a 15% limit.

In order to establish the extraction efficiency of 2ME2 from plasma, two groups (A and B) of QCs were prepared. In group A, QCs of 3 and 75 ng/mL concentrations were prepared and extracted as described in sections 2.2 and 2.3. In group B, 0.3 mL of blank plasma was spiked with 50 µL of IS, followed by addition of 3 mL of ethyl acetate, vortex-mixed for 15 minutes and centrifuged at 4000 rpm for 15 minutes. The supernatant was separated and, before the drying down step, spiked with 2ME2 to result in concentrations of 3 and 75 ng/mL. Samples were dried and reconstituted as described in section above. Group A was designated as the extracted samples and group B was designated as the unextracted samples. Samples from group A and group B were analyzed as described in section 3.3 and the means of peak areas of 2ME2 obtained at each concentration were compared to obtain % recovery at corresponding concentrations:

% Recovery = E/U × 100 where E is the mean area count of test samples for one concentration level after sample extraction (group A), and U is the mean area count of unextracted samples for one concentration (group B). As a guidance, recovery need not be 100%, but the coefficient of variation should not be more than 15%.

Lower limit of quantification (LLOQ)

The LLOQ was defined as the lowest concentration of 2ME2 that could be reliably and reproducibly measured with values for accuracy, between-run precision, and within-run precision of less than 20%, with concentration determinations performed in quadruplicates on 4 separate occasions. The signal-to-noise ratio ≥ 3 at the limit of

detection, and signal-to-noise ratio ≥ 10 at the limit of quantitation were set as

acceptability criteria.

Specificity

Pooled blank plasma samples were used to determine whether endogenous matrix

constituents co-eluted with 2ME2. Blank samples were also obtained from six different

donors, and were analyzed to determine if there were any interfering peaks around the

retention time of 2ME2. To rule out the specificity issues we faced with a previously

described HPLC/UV method80, a cross-validation of the LC/MS/MS method was

performed with a GC/MS/MS method for determination of 2ME2 in plasma developed by

Entremed Inc. (Rockville, MD, USA). The cross-validation samples were run in

duplicate to determine the 2ME2 concentrations and compared to those determined by the

GC/MS assay.

Matrix Effects

Testing for matrix effects was done by testing the signal of two groups of

samples. The first group of samples group A, consisted of 0.3 mL of unspiked blank

plasma samples which were extracted using the standard liquid-liquid extraction method

as described above. After extraction, the aqueous phase was dried at 370 C under air until dryness. The extracts were reconstituted in 60 µL of 50% methanol (in water) using vortex mixing and 2ME2 from the stock solution was added to it to result in concentrations of 3ng/mL and 75 ng/mL respectively in triplicate, and 20 µL of the reconstitute was injected for LC/MS/MS analysis. The second group of samples B,

consisted of samples of 60 µL of 50% methanol (in water) which were spiked with 2ME2

to result in concentrations of 3 ng/mL and 75 ng/mL in triplicate and 20 µL of the

reconstitute was injected for LC/MS/MS analysis. The peak areas of both groups of

samples were compared to see if there was a significant decline in 2ME2 peak areas.

Matrix effects were considered negligible if the peak area were within ±5% of each other.

Freeze-thaw stability

The stability of 2ME2 in plasma subjected to three consecutive freeze-thaw cycles

(lasting 15 minutes each) was tested by quadruplicate analysis of samples containing

2ME2 at concentrations of 3, 40 and 75 ng/mL of 2ME2. The calculated 2ME2

concentrations were evaluated for accuracy relative to the nominal (spiked) drug concentration. Analytes were considered stable if the concentration deviated less than

±20% from the concentrations of freshly prepared samples.

Short-term and long-term stability

The stability of 2ME2 in plasma samples stored at -800C for two weeks (short- term stability) was tested by analysis, in quadruplicate, of samples containing 2ME2 at 3 and 75 ng/mL. The stability of 2ME2 in plasma samples stored at -800C for over 1 year at -800C (long term stability) was tested by analysis of replicates of five of samples

containing 3 and 100 ng/mL. 2ME2 concentrations were estimated from the stored

samples based on freshly prepared standards and were compared to freshly prepared QCs

at concentrations of 3, 75 and 100 ng/mL. Analytes were considered stable if the

concentration deviated less than ±20% from the concentrations of freshly prepared

samples.

Autosampler and re-injection stability

Samples were prepared in quadruplicate at 3, 40 and 75 ng/mL and left in the

autosampler for 48 hours at 50C. 2ME2 concentrations were estimated from the stored samples based on freshly prepared standards and were compared to freshly prepared QCs at concentrations of 3, 40 and 75 ng/mL. For re-injection stability, the remnant in the vials of 3, 40 and 75 ng/mL, after one injection, was re-injected and 2ME2 concentrations were estimated based on freshly prepared standards and were compared to freshly prepared QCs at concentrations of 3, 40 and 75 ng/mL. Analytes were considered stable

if the concentration deviated less than ±20% from the concentrations of freshly prepared

samples.

Results of validation in plasma

Specificity and cross-validation

Analysis of blank plasma samples obtained from multiple volunteers did not show any

significant chromatographic interference around the retention time of 2ME2 and the IS.

There was a small peak that co-elutes at about the same retention time as 2ME2, but the

signal from that peak was too small to interfere with quantitation of 2ME2 at the LLOQ.

This peak is believed to be due to carry-over of 2ME2 between samples or as a result of

detection of endogenous 2ME2. Seven blinded plasma samples that were previously

quantitated by a GC/MS assay (Entremed, data on file), were analyzed using the

LC/MS/MS assay and the estimated concentrations were found to be comparable to those

determined by the GC/MS assay.

Response Function

The lowest and most constant bias across the concentration range investigated

was obtained following regression analysis of the data to a linear fit with a weighting

factor of 1/x2 for the ratio of the peak area of 2ME2 and the IS (data not shown). For each analytical run, a 6-point plasma standard curve was constructed, and was shown to be linear over the tested range of 1 to 100 ng/mL. The mean (± standard deviation) regression equation obtained during the method validation, obtained in duplicates on 4 separate occasions, showed an intercept of 0.006±0.0034 and a slope of 0.013±0.0003

[Pearson correlation coefficient range (0.9970- 0.9753); n = 4].

Lower limit of quantification (LLOQ)

Using this procedure, the lower limit of quantification was determined to be 1

ng/mL, with a % RSD of 19.8% and a percent deviation from the nominal standard of

+12.1%, which was within acceptable limits.

Accuracy, Precision and Recovery

Over the entire concentration range of the standard curve, the mean observed

percent deviation was between -8.6% and +12.1%, at an imprecision of less than 19.8%

(Table 7).

Table 7

Back Calculated Concentrations from Calibration Curves*

Nominal GMa SD DEV RSD N

(ng/mL) (ng/mL) (ng/mL) (%)

1 1.1 0.2 +12.1 0.2 9

5 5.3 0.7 +5.2 0.1 8

10 10.1 0.8 +1.0 0.1 8

20 20.3 0.9 +1.63 0.04 8

50 51.2 3.8 +2.3 0.07 8

100 91.4 10.5 -8.6 0.11 8

a Abbreviations: GM, grand mean; SD, standard deviation; DEV, percent deviation from

nominal value; RSD, relative standard deviation; N, total number of replicate

observations (total of all validation runs).

The assay performance data for the determination of independently prepared QC

samples of 2ME2 in plasma are presented in Table 3.3. The between-run precision and

within-run precision ranged from 1% to 2.8% and from 1.2% to 3.7%, respectively, for

the various concentrations tested. At the same QC concentrations, the values for

accuracy were always between +5.8% and +7.9%, which is well within the generally

accepted limits for bioanalytical methods 81. The extraction recovery was found to be

86.6% and 80.5% at 3 ng/mL and 75 ng/mL respectively.

Table 8

Assessment of Accuracy and Precision from Quality-Control Samples:

Nominal GMa SD DEV WRP* BRP N

(ng/mL) (ng/mL) (ng/mL) (%) (%) (%)

3 3.2 0.2 +7.9 3.7 2.8 20

40 42 2.4 +4.9 1.2 2 20

75 79.4 4.7 +5.8 2.8 1 20

* - WRP was calculated from five replicates of each QC analyzed within a single day

a Abbreviations: GM, grand mean; SD, standard deviation; DEV, percent deviation from nominal value;

WRP, within-run precision; BRP, between-run precision; N, total number of replicate observations (total of all validation runs).

Stability

Three repeated freeze-thaw cycles of 15 minutes each, had no apparent influence

on the stability of plasma samples containing 2ME2 at concentrations of 3, 40 or 75

ng/mL. After the third freeze-thaw cycle, 2ME2 plasma concentrations had mean

deviations from the nominal values within the range of +11.2%, 12.1%, and 6.6%. There

was no difference in the mean peak areas after one, two, three freeze thaw cycles and

freshly prepared samples when compared as determined using ANOVA (Table 9).

Samples stored for two weeks (short-term stability) at -800C had mean deviations from nominal values within the range of 3.7% at 3 ng/mL, -0.9% at 40 ng/mL and + 0.3% at

75 ng/mL and were considered stable. Samples stored for over a year (long-term stability) at -800C had mean deviations from nominal values of -3.8% at 3 ng/mL and

+1.6% at 100 ng/mL and were also considered sufficiently stable. Processed plasma samples were found to be stable at 50C upon standing in the autosampler tray for at least

48 hours, allowing for overnight analysis of extracted samples. Processed plasma

samples that were re-injected were found to be stable at 50C upon standing in the autosampler tray for at least 48 hours, allowing for re-injection on extracted material, should that be necessary.

Table 9: Freeze Thaw Stability of 2ME2

Concentration p-value F-value df

(ng/mL)

3 0.48 0.918 3

40 0.89 0.206 3

75 0.36 1.219 3

a Abbreviations: df, degrees of freedom

Bioanalytical method for detection of 2ME2 in human urine

Stock solutions and standards in urine

Stock solutions of 2ME2 and 2ME2-d5 (IS) were made in triplicate by dissolving

10 mg of drug in 10 mL of methanol, resulting in primary stock solutions containing 1 mg/mL of 2ME2 and IS. These solutions were stable and stored at -800C for up to 8 weeks. Working solutions of 2ME2 and IS were prepared on each analysis day by serial dilutions in methanol from the primary stock solution. The difference in drug concentration in each of the triplicate stock solutions, estimated from the mean peak area following repeat analysis of a dilution of the stock, was determined to be within ±5%.

For urine, calibration samples containing 2ME2 were prepared by addition of aliquots of the working solutions to drug-free human urine. Calibration standards were prepared within the concentration range of 1 ng/mL to 30 µg/mL. Quality control (QC)

samples for 2ME2 were prepared in human urine at concentrations of 3, 30, 80, 300, 750

ng/mL, and 4 µg/mL. Blank human urine was analyzed alongside to rule out interference

from endogenously produced 2ME2.

Sample preparation for testing amount of 2ME2 excreted unchanged in urine

A volume of 100 µL of a 1 µg/mL solution of 2ME2-d5 was added to 1 mL of

urine. Next, 5 mL of de-ionized water was used to dilute the urine spiked with IS.

Extraction was performed by addition of 5 mL of ethyl acetate. The mixture was then

vortex-mixed for 15 minutes, followed by centrifugation at 4000 rpm for 15 minutes.

The supernatant was taken off after centrifugation and evaporated to dryness under a

continuous stream of air at 370 C. The extracts were reconstituted in 60 µL of 50%

methanol (in water) using vortex mixing, and 20 µL of the reconstitute was injected for

LC/MS/MS analysis. Patient samples were analyzed on two separate occasions and the

mean value of both runs was reported.

Sample preparation for testing amount of 2ME2 excreted as glucuronides in urine

A volume of 100 µL of a 1 µg/mL solution of 2ME2-d5 was added to 0.5 mL of urine.

Next, 4.5 mL of de-ionized water was used to dilute the urine spiked with IS. After that,

1 mL of phosphate buffer (0.1M, pH 6.5) and 80 µL of β-glucuronidase (200 units/mL)

was added. This mixture was incubated for 1 hour at 520 C. Extraction was performed by addition of 5 mL of ethyl acetate. The mixture was then vortex-mixed for 15 minutes, followed by centrifugation at 4000 rpm for 15 minutes. The supernatant was taken off after centrifugation and evaporated to dryness under a continuous stream of air at 370 C.

The extracts were reconstituted in 60 µL of 50% methanol (in water) using vortex

mixing, and 20 µL of the reconstitute was injected for LC/MS/MS analysis. The amount

of 2ME2 excreted unchanged in 24 hour urine was subtracted from the amount of 2ME2

detected in the 24 hour urine after hydrolysis of glucuronides to obtain the absolute

amount of 2ME2 excreted in urine as glucuronides. Patient samples were analyzed on

two separate occasions and the mean value of both runs was reported.

Sample preparation for testing amount of 2ME2 excreted as sulfates in urine

A volume of 100 µL of a 1 µg/mL solution of 2ME2-d5 was added to 1 mL of urine.

Next, 4 mL of de-ionized water was used to dilute the urine spiked with IS. After that,

1 mL of acetate buffer (2M, pH 5.2) and 50 µL of sulfatase (4.9 mg protein/mL) was

added. This mixture was incubated for 1 hour at 520 C. Extraction was performed by addition of 5 mL of ethyl acetate. The mixture was then vortex-mixed for 15 minutes, followed by centrifugation at 4000 rpm for 15 minutes. The supernatant was taken off after centrifugation and evaporated to dryness under a continuous stream of air at 370 C.

The extracts were reconstituted in 60 µL of 50% methanol (in water) using vortex

mixing, and 20 µL of the reconstitute was injected for LC/MS/MS analysis. The amount

of 2ME2 excreted unchanged in 24 hour urine was subtracted from the amount of 2ME2

detected in the 24 hour urine after hydrolysis of sulfates to obtain the absolute amount of

2ME2 excreted in urine as sulfates. Patient samples were analyzed on two separate

occasions and the mean value of both runs was reported.

Equipment

Chromatography was performed on a Waters Alliance 2695 LC system (Waters

50 mm i.d.; 5-µm; Waters) using a mobile phase composed of methanol and water. The mobile phase was delivered as a gradient run over 9 minutes at a flow rate of 0.25 mL/min as shown in Table 1. A Micromass Quattro Micro detector (Micromass

Corporation, Manchester, UK) was used for tandem mass-spectrometry with atmospheric pressure chemical ionization source operated in the positive ionization mode. The instrument was operated in the MRM (Multiple Reaction Monitoring) mode monitoring the ion transitions from m/z 303.1→136.8 (2ME2) and m/z 308.1→138.8 (IS). Data processing and analysis were performed using Masslynx/Quanlynx software. Detector settings were as follows: cone voltage, 22 V; extractor voltage, 2V; corona current, 5 µA; source temperature, 1300 C; and desolvation gas flow, 310 L/h. The zero concentration sample (blank) was used to visually verify the purity of the reagents and the lack of other potentially interfering (endogenous) substances, but was not considered for the regression analysis of standards. The goodness-of-fit of various calibration models was evaluated by visual inspection, the correlation coefficient and a lack-of-fit test.

Chromatography and detection in Urine

Typical chromatograms resulting from the LC-APCI-MS-MS analysis of 1-mL urine extracts obtained after liquid-liquid extraction are depicted here, and include a blank

urine sample (Figure 9), a standard spiked at a 2ME2 concentration of 100 ng/mL (Figure

10), and a standard spiked at a 2ME2 concentration of 1 ng/mL (Figure 11). Figure 12

depicts a chromatogram of an extract of 1mL of urine from a 24 hour collection, obtained

from a patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, without addition of β-glucuronidase, figure 13 shows the chromatogram after incubation of urine with β-glucuronidase and figure 14 depicts the chromatogram after incubation of urine with sulfatase. The retention time of 2ME2 and the IS was about 5.2 minutes with an overall run time of 9 minutes to allow for re-equilibration of mobile phase. Appendix

B contains a more comprehensive collection of the chromatograms obtained after analysis of 24 hour urines.

Figure 9: Blank urine sample (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_015 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ blank 15 5.49 303.1 > 136.8 795.15 4.521e+003 100 0.01 % 3.54 0.63 1.18 1.57 1.86 2.38 3.36 4.16 6.94 7.52 8.20 0 min 2me_042305_015 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ blank 15 5.46 308.2 > 138.8 9.72 1.046e+002 100 5.21 6.066.12 1.74 4.274.31 4.63 % 0.62 0.70 1.33 2.21 2.30 3.00 3.32 3.96 6.38 7.13 7.73 7.84 8.488.62 8.85 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 10: Urine standard spiked at a 2ME2 concentration of 20 µg/mL (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_071 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ 20000 71 5.47 303.1 > 136.8 2152137.00 1.720e+007 100 % 0 min 2me_042305_071 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ 20000 71 5.45 308.2 > 138.8 5665.81 4.002e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 11: Standard spiked at a 2ME2 concentration of 10 ng/mL (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_061 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ 10 61 5.47 303.1 > 136.8 1697.21 1.193e+004 100 % 0 min 2me_042305_061 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ 10 61 5.45 308.2 > 138.8 5701.90 4.252e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 12: Urine sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_026 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P5U 26 5.46 303.1 > 136.8 2884.43 2.444e+004 100 % 0 min 2me_042305_026 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P5U 26 5.43 308.2 > 138.8 4148.67 3.036e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 13: Urine sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, 1 hour after treatment with β-glucuronidase (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_038 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P5B 38 5.47 303.1 > 136.8 773562.94 6.122e+006 100 % 0 min 2me_042305_038 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P5B 38 5.45 308.2 > 138.8 4374.37 3.026e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 14: Urine sample from patient with ovarian cancer who received a single oral dose of 2ME2 at 2200 mg, 1 hour after treatment with sulfatase (2ME2 chromatogram is depicted by 303.1 > 136.8 transition; 2ME2-d5 chromatogram is depicted by 308.2 > 138.8 transition)

2me_042305_050 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P5S 50 5.46 303.1 > 136.8 1644.83 1.262e+004 100 % 2.03 3.33 0 min 2me_042305_050 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P5S 50 5.43 308.2 > 138.8 2592.50 1.816e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Response function

Calibration curves were constructed by least-squares linear regression analysis of peak

area ratio of 2ME2 and the IS versus the drug concentration of the nominal standard (x)

with or without weighting. Calibration curves were fitted by weighted (1/x2) least-squares regression analysis. To establish the optimal quantification method and weight factor, the correlation coefficient of the fitted equation and the accuracy of back-calculated calibration concentrations were taken into consideration.

Specificity

Pooled blank urine samples were used to determine whether endogenous matrix constituents co-eluted with 2ME2.

Acceptability criteria for bioanalytical runs for analysis of 2ME2 in urine

The LLOQ was defined as the lowest concentration of 2ME2 that could be reliably and reproducibly measured on two separate occasions with coefficients of variation of less than 20% with concentration determinations performed in duplicates and a signal-to-noise ratio ≥ 3. On each of two days along with the patient samples, calibration curves were analyzed in duplicate with quality control samples in replicates of three containing known concentrations of 2ME2. The analytic run was considered acceptable only if the percent deviation (DEV) for the quality control samples were within ± 20%. Percent deviation (DEV) was defined as percent difference between the mean observed concentration and the nominal concentration:

⎛ [nominal] − [observed]⎞ DEV = ⎜ ⎟ ×100% ⎝ [nominal] ⎠

Non-lower limit of quantitation (LLOQ) standards and QC samples were considered reliable if at least two out of every three estimates of accuracy were within the range of 80-120%81. The accuracy and precision data as calculated from the analysis of

2ME2 in human urine are depicted in tables 10 and 11.

Table 10

Back Calculated Concentrations from Calibration Curves*

Nominal GMa SD DEV RSD N

(ng/mL) (ng/mL) (ng/mL) (%)

1 1.06 0.27 +6 0.25 4

10 10.8 1.0 +8 0.1 6

20 16.5 1.36 -17.5 0.08 6

50 51.9 5.4 +3.8 0.10 6

100 105.4 7.5 +5.5 0.07 6

200 188.27 16.5 -5.8 0.09 6

500 527.7 92.3 +5.5 0.17 6

1000 935 47.1 -6.5 0.05 6

3000 3047 120.9 +1.5 0.04 6

5000 5003 233 +0.07 0.05 6

10000 10804 187 +8.0 0.01 4

20000 20842 766 4.2 0.04 4

a Abbreviations: GM, grand mean; SD, standard deviation; DEV, percent deviation from

nominal value; RSD, relative standard deviation; N, total number of replicate

observations (total of all validation runs).

Table 11

Assessment of Accuracy and Precision from Quality-Control Samples:

Nominal GMa SD DEV WRP* BRP N

(ng/mL) (ng/mL) (ng/mL) (%) (%) (%)

30 33.2 3.1 +10.9 0.56 § 15

80 85.7 3.1 +7.1 2.3 2.6 15

300 322.3 28.8 +7.4 0.9 1.1 15

800 819.8 49.1 +2.4 0.02 § 15

4000 4099 157 +2.5 0.05 0.05 15

* - WRP was calculated from five replicates of each QC analyzed within a single day

§ - No additional variation was observed as a result of performing the assay on different days

a Abbreviations: GM, grand mean; SD, standard deviation; DEV, percent deviation from nominal value;

WRP, within-run precision; BRP, between-run precision; N, total number of replicate observations (total of all validation runs).

Conclusion

Thus, two methods were developed to for the quantitation of 2ME2 in human plasma and urine. The method in plasma was fully validated and the performance criteria for sensitivity, accuracy, precision, linearity, stability, and specificity were acceptable, indicating that the method can be used for routine determination of 2ME2 in plasma samples obtained from patients treated with the drug. Hence, objective # 1 of this thesis as listed in chapter 2 was achieved.

The method for detection and quantitation of 2ME2 in human urine did not receive a full validation. This method was used, within the set acceptability criteria for bioanalytical runs, for quantitation of 2ME2 in 24-hour urine collection samples obtained from patients treated with the drug.

CHAPTER FOUR

PHASE I Clinical trial of Orally Administered 2-Methoxyestradiol in patients with Solid Tumors

Introduction

The clinical experience with 2ME2 has been extensive. Table 10 summarizes the results from clinical trials of 2ME2 that were completed prior to the beginning of this phase I clinical trial. Because 2ME2 was tolerated well in previous clinical trials and the

DLT had not been reached, a Phase I clinical trial was initiated at 2ME2 doses higher than those administered in previous clinical trials. Following were the primary endpoints of this clinical trial:

1. To characterize the side effect profile of 2ME2

2. To determine the pharmacokinetic profile of 2ME2 in humans

3. To determine if any pharmacodynamic correlations can be made between clinical

activity and/or toxicity

4. To determine the maximum tolerated dose (MTD) of 2ME2

The secondary endpoints of the clinical trial were:

1. To evaluate the correlation between the rat aorta model bioassay with plasma

concentrations of 2ME2, clinical outcome and biological endpoints

2. To assess the changes in PET (O15 & FDG) scan during treatment with 2ME2

3. To assess changes in apoptosis in biopsied endothelial cells and/or tumor cells 64

The tertiary endpoints of the clinical trial were:

1. To determine changes in circulating molecular markers of angiogenesis following

2ME2 therapy (TNF-α, TGF-β, bFGF & VEGF)

2. To determine changes in other molecular markers of angiogenesis in the biopsy

tissues (immunostaining for factor VIII, CD34, bFGF, VEGF, PDGF etc)

Table 12: Summary of clinical trials of 2ME2

Trial Type Doses Indication Patients Comments Phase I 200-1000 Metastatic 31 -MTD not reached (single agent) mg/d Breast Cancer -No grade III/IV toxicities 75 -T1/2 = 10 hours -All (n=3) patients receiving 1000 mg experience hot flashes - Reduction in bone pain and analgesic use noted Phase I 200-1000 Metastatic 15 -MTD not reached (with mg/d Breast Cancer -No grade IV toxicities docetaxel: -Grade 3 fatigue (n=5), diarrhea 35 mg/m2) (n=4) and hand-foot syndrome 77 seen (n=1) -Grade 3 liver transaminase elevations seen in 3 patients, which returned to normal with continued treatment -Extensive metabolism to 2ME1 was observed -2ME2 trough and peak levels were not altered by concurrent docetaxel administration Phase II - 400 mg/d Hormone 33 -Well tolerated (single agent) & refractory -PSA stabilization and declines 76 - 1200 mg/d prostate were observed cancer -Grade 2 and 3 liver function abnormalities observed which normalized after 2ME2 discontinuation -VEGF and bFGF levels showed and 40% and 50% decline respectively after being on therapy for 3 months *Abbreviations: 2ME1, 2-methoxyestrone; bFGF, basic fibroblast growth factor; MTD, maximum tolerated dose; PSA, prostate specific antigen; VEGF, vascular endothelial growth factor

Patients and Methods

Patient enrollment

All patients were required to have histologically-documented solid malignancy that

was metastatic or unresectable and for which standard treatment measures do not exist or

were no longer effective. Also, patients were required to have disease amenable to

biopsy. Patients were required to have an Eastern Cooperative Oncology Group

performance status of 0, 1 or 2 and a life expectancy of more than 3 months. Patients

were excluded if they were receiving active therapy for their cancer (except patients with

prostate cancer receiving LHRH agonists), had brain metastases and if they were

pregnant or lactating.

All patients received, within 4 weeks before enrollment, baseline CT scans and 99Tc- bone scintigraphy scans (if they had bone metastases) along with baseline PET scan (O15 and/or FDG). All patients had undergone a baseline tumor biopsy. A baseline chest X- ray and EKG were obtained at least 2 weeks before starting 2ME2. Baseline laboratory studies were obtained within 7 days before starting therapy. Follow-up laboratory studies were obtained at each clinic visit. Imaging studies were repeated approximately every 2 months.

Complete response was defined as compete disappearance of all tumor lesions for at least two measurement periods separated by 4 or more weeks; soft tissue abnormalities had to be re-biopsied to document absence of disease. Partial response was defined as a decrease of 50% or more in the sum of the products of the longest perpendicular diameters of all measurable lesions or a reduction by 50% or more of the number of areas showing, on bone scan or CT scan lasting at least one month. Progressive disease was

defined as a more than 25% increase in the sum of the products of the longest

perpendicular diameters of all measurable lesions or the appearance of new lesions.

Stable disease was defined a neither responding nor progressing disease for at least 6

months after starting therapy. Patients were removed from the study for progressive

disease or unacceptable toxicity. Clinically progressive disease was defined as

development of new area of malignant disease, progression of soft tissue metastases as

documented by appropriate modalities (imaging/palpation), at least one new metastatic

deposit on 99Tc-bone scintigraphy or increases in tumor marker levels as per consensus guidelines. Toxicity was graded by the National Cancer Institute’s common toxicity criteria, version 2.0. The clinical protocol was reviewed and approved by the National

Cancer Institute’s institutional review board. Written informed consent was obtained from each patient before he or she participated in the study.

Drug Administration

2ME2 was provided by Entremed Inc. (Rockville, MD) through the Clinical Trials

Evaluation Program (CTEP) and was formulated as 200 mg capsules. Patients were administered an initial single oral dose followed by pharmacokinetic sampling.

Approximately one week following pharmacokinetic evaluation, daily oral dosing of

2ME2 began. Twenty-eight days of oral daily dosing was considered one treatment cycle. If patients tolerated 2ME2 well, and did not have progressive disease, patients continued to receive 2ME2.

Study Design

A minimum of three patients were to be accrued at each dose level (DL) for each continuous 28-day cycle. Dose escalation occurred after three patients had been accrued and all 3 patients had undergone one cycle of treatment. The study utilized a dose escalation phase I design (Table 4.1), except that the cohort size was increased to 6 patients at any dose level if biologic activity was suggested. If 0/3 patients experienced dose-limiting toxicity (DLT), and if there is no suggestion of biologic activity, then the subsequent 3 patients were entered at the next higher dose level. If 0/3 experience DLT, but some evidence of biologic activity was suggested at a dose level, then 3 more patients

(total 6) were to be treated at the dose level. If 1/3 patients experienced a DLT, then up to

3 additional patients were to be entered on that dose level. If 1/6 experienced a DLT, then further doses were to be increased by 40%. If 2 patients had a DLT on a dose level, then the dose escalation was complete; 3 additional patients were to be entered onto the dose level immediately below that with 2 patients having DLT, provided that only 3 patients were treated on that dose level. If the dose level below that with two patients experiencing DLT had already entered 6 patients (with 0-1 patients having a DLT), then no further patients are required to establish this as the MTD. Maximum tolerated dose

(MTD) was defined as the highest dose level explored in which less than 2 of 6 patients treated for at least 28 days did not experience Grade 3 or greater non-hematologic toxicity or Grade 4 or greater hematologic toxicity. Toxicity was graded by National

Cancer Institute’s common toxicity criteria v 2.0.

Table 13: Dose Levels for 2ME2 clinical trial

Dosage Levels Dose Patients Accrued 1 400 mg bid 3 2 800 mg bid 3 3 1600 mg bid 6 4 2200 mg bid 5 5 3000 mg bid 3

Pharmacokinetic Analysis

Blood samples were collected for pharmacokinetic analysis before 2ME2

administration and then at 0.5, 1.5, 2, 3, 4, 5, 6, 9, 12, 24, 38, 50 hours after the first dose.

At each clinic visit, at least one blood sample was obtained for pharmacokinetic analysis.

Samples were collected in heparinized tubes and centrifuged at 2400 rpm for 5 minutes;

plasma was aliquoted and stored at -800C until the time of analysis. Samples were

analyzed using a validated method based on liquid chromatography with tandem mass

spectrometric detection (chapter 2). Estimates of pharmacokinetic parameters were

derived from individual concentration-time data sets by non-compartmental analysis

using the software package WinNonlin v 4.0 (Pharsight Corporation, Mountain View,

CA). The peak plasma concentrations and the time to peak concentrations were observed

values. The area under the plasma concentration versus time curve (AUC) was calculated

using the linear trapezoidal method from time zero to the time of the final quantifiable

concentration (AUC[tf]). The AUC was then extrapolated to infinity (AUC[inf]) by dividing the last measured concentration by the rate constant of the terminal phase (k), which was determined by linear-regression analysis of the final three or four time points of the log-linear concentration-time plot. The apparent oral clearance (CL/F) was

calculated by dividing the administered dose by the observed AUC[inf] or AUClast and the

terminal half-life (t1/2) was calculated by dividing 0.693 by k.

Correlative Studies

Frozen core biopsies were thawed in formalin and subsequently embedded in

paraffin. Immunohistochemistry on 4 µm tissue sections of formalin-fixed paraffin- embedded biopsies was performed using a standard avidin-biotin-peroxidase complex indirect immunoperoxidase procedure as described previously 82. Tumor content was

evaluated with the use of hematoxylin and eosin staining. Briefly, after deparaffinization,

rehydration, and antigen retrieval, sections were treated with 1% H2O2 to inactivate the endogenous peroxidase activity and then incubated with the primary antibodies. Mouse

monoclonal antibodies to CD 31, clone JC70A and Ki-67, clone MIB-1 (DAKO Inc.,

Carpinteria, CA) at a dilution of 1: 20 and 1: 50, respectively, were applied to tissue

sections at room temperature for one hour. Binding of antibodies to their antigenic sites

was amplified using Vectastain Elite avidin-biotin-peroxidase complex kits (Vector

Laboratories, Burlingame, CA). Sites of the antigen-antibody reactions were visualized

using 3, 3-diaminobenzidine for 5 minutes and slides were counterstained with Mayer’s

hematoxylin. Normal tonsil known to be positive for Ki67 and breast cancer specimen known to be positive for CD 31 served as positive controls. Negative controls were

performed by omission of the primary antibodies. Immunohistochemial staining signal

was analyzed quantitatively with the assistance of the Automated Cellular Imaging

System (ACIS; ChromaVision Medical Systems, Inc, San Juan Capistrano, CA). Images generated by ACIS were reviewed by an investigator (S.Y.) without knowledge of any

clinical data. Hot spot feature of the ACIS was used to direct to the sites of staining where 3 areas of tumor were scored with an X40 or a free scoring tool. An average labeling percentage was reported for Ki67 and the number of vessels per mm2 for tumor microvessel density.

Statistical Analysis

Spearman rank correlation analysis was performed to determine the relationship between the dose administered and each of the following pharmacokinetic parameters: peak concentration (Cmax), time to peak concentration (Tmax), half-life (T1/2), apparent oral clearance (CL/F), area under the plasma concentration-time curve from time zero to the last time point with measurable concentrations (AUClast), and the apparent volume of

distribution (Vd/F). In the correlation analysis, correlations such that r > 0.7 were

interpreted as strong correlations, those with 0.5 < r < 0.7 were moderately strong

correlations and those with 0.3 < r < 0.5 were weak to moderate, and those with r < 0.3

were interpreted as weak correlations. A Jonckheere-Terpstra trend test was used to test

if there was any trend in CL/F and Vd/F with increasing dose. Statistical analysis was performed using the software package SAS Version 8. A paired t-test was used to establish the statistical significance of pre and post therapy microvascular density (MVD) using CD 31 immunohistochemical staining and Ki67 staining.

Results

Patient Characteristics

Twenty patients with advanced refractory metastatic cancer with biopsiable disease were enrolled onto this study (Table 14, 15). The median age of the patients was 58 years with a range of 37-76 years. Patients had received a median of 4.5 prior therapies for their cancer with a range of 1-21 prior therapies. The two most frequent tumor types were prostate cancer (n=6) and ovarian cancer (n=6). One patient had an ECOG performance status of 2, seventeen were PS 1 and two were PS 0.

Table 14 Patient Demographics

Total no. 20

Age, years

Median 57.5

Range (37-76)

Male:female 11:19

Performance status

0 2

1 17

2 1

Primary tumor site

Prostate 6

Ovary 6

Breast 1

Melanoma 2

Colon 1

Sarcoma 1

Rectal 1

Renal 1

Adrenal 1

No of Prior Therapies (including chemotherapy, hormonal and investigational agents)

0 0

1 1

2 2

≥3 17

Median 4.5

Range 1-21

Table 15: Patient Disposition

Patient No. Cancer Gender Dose Level Off-study Total time (mg/d) Reason on study (months) 1 Melanoma Male 800 PD 4

2 Sarcoma Male 800 PD 1

3 Prostate Male 800 PD 2

4 Colon Male 1600 PD 2

5 Breast Female 1600 PD 2

6 Prostate Male 1600 PD 11

7 Ovarian Female 3200 NA 40 (ongoing) 8 Prostate Male 3200 Tox 2

9 Prostate Male 3200 PD 2

10 Melanoma Female 3200 PD 2

11 Rectal Male 3200 PD 2

12 Ovarian Female 3200 PD 1

13 Ovarian Female 4400 PD 2

14 Ovarian Female 4400 PD 5

15 Ovarian Female 4400 PD 5

16 Prostate Male 4400 PD 3

17 Ovarian Female 4400 PD 2

18 Renal cell Male 6000 PD 0.5

19 Adrenal Female 6000 PD 4

20 Prostate Male 6000 PD 2

*Abbreviations: PD = progressive disease, Tox = toxicity, NA = Not applicable

Clinical Outcomes

Sixteen patients were removed from the study due to progressive disease. One patient

(# 7) is still receiving therapy for over three years duration at the time of this writing.

Twelve patients remained on study for more than 60 days, seven patients remained on study between 31 and 60 days and one patient was taken off study within 30 days.

Patient # 7 was a 51 year old female with history of clear cell carcinoma of the round ligament of the ovary diagnosed in August of 2000. She had multiple relapses of the disease and had undergone total abdominal hysterectomy and bilateral salpingo- oophorectomy and six different chemotherapeutic agents for her disease. The patient was started on 2ME2 at the dose of 1600 mg q 12 h on 4/29/2002 for the treatment of lymph node metastases of the ovarian cancer. The patient continues to tolerate 2ME2 well and continues to be on therapy for over three years at the time of this writing. The patient experienced a partial response with a 78% reduction in the CA125 concentrations (Figure

15) at the 24 month follow up and 67% reduction in the size of her pelvic lymph nodes by

CT scan (Figure 16) and continues to receive the treatment for over 1060 days. Fifteen patients were taken off the trial due to progressive disease within three months from initiation of therapy. Five patients continued on the study for more than three months.

One patient with adenocarcinoma of the prostate had stable disease based on PSA and continued on the treatment for 350 days.

Figure 15: CA-125 concentrations versus time plot for patient #7

CA-125 levels versus months on therapy

30 s l 25 eve 20 25 l 1 -

A 15 C 10

0 0 5 10 15 20 25 30 35 Months on Therapy

Figure 16: CT scan of patient #7 showing shrinkage in the left pelvic lymph nodes after therapy with 2ME2

(A) Pre-Therapy (B) Post-Therapy

Toxicities

2ME2 was very well tolerated at all the administered doses. At dose level 3, one patient (#8) had an episode of a grade 4 angioedema 42 days (cycle 3) after starting

2ME2. Hence, the dose level was expanded to include a total of six patients.

Patient 8 was a 74 year old Caucasian male presenting with prostate cancer and ECOG performance of 1. The patient was started on 2ME2 on 4/24/2002 on a dose of 1600 mg

PO q 12 h. On the evening of 6/5/2002, the patient began to experience tongue swelling and subsequent respiratory difficulty. The patient experienced respiratory failure secondary to angioedema neurotica. The patient was taken to Mercy Hospital of

Pittsburgh Emergency Room where, he was intubated and was treated with diphenhydramine hydrochloride 25 mg IV q12 h, methylprednisolone 40 mg IV q 6 hrs and famotidine 20 mg IV bid and was taken off 2ME2 on the same day. This type of a severe allergic reaction was not an expected event and was considered unrelated to 2ME2 administration because it occurred 38 days after the first dose of 2ME2 was administered and also because the patient was concurrently taking enalapril, a more likely culprit of the adverse event.

Table 16 provides a detailed synopsis of all major toxicities that occurred at each dose level. Three (two male, one female) patients had hot flashes at DL 2 and 3. Fatigue, diarrhea, nausea, hyperglycemia, anemia and edema were seen almost at all dose levels.

Liver transaminases (AST, ALT) and alkaline phosphatases were elevated at higher doses. There was no dose limiting toxicity (DLT) observed and MTD was not reached.

Table 16: Adverse Events Table

Adverse Events Reported Per Patient GRADE 1 2 3 4 5 ALLERGY/IMMUNOLOGY Allergic reaction/hypersensitivity (including drug fever) 1 1 Hemoglobin 3 3 1 BLOOD/BONE MARROW Leukocytes (total WBC) 1 1 CARDIOVASCULAR (ARRHYTHMIA) Sinus tachycardia 1 Edema 1 2 CARDIOVASCULAR (GENERAL) Thrombosis/embolism 2 CONSTITUTIONAL SYMPTOMS Fatigue (lethargy, malaise, asthenia) 4 3 1 Alopecia 1 DERMATOLOGY/SKIN Dry skin 2 Rash/desquamation 1 1 ENDOCRINE Hot flashes/flushes 1 1 Anorexia 3 1 Ascites (non-malignant) 1 Constipation 1 GASTROINTESTINAL Diarrhea patients with a colostomy 1 Diarrhea patients without colostomy 5 2 Nausea 7 2 1 Vomiting 3 4 1 HEMORRHAGE Vaginal bleeding 1 Alkaline phosphatase 3 1 1 Hypoalbuminemia 1 4 HEPATIC SGOT (AST) 3 1 SGPT (ALT) 2 Hypercholesterolemia 1 Hyperglycemia 1 METABOLIC/LABORATORY Hypermagnesemia 1 Hypomagnesemia 3 Hyponatremia 3 NEUROLOGY Dizziness/lightheadedness 4 Neuropathy-sensory 2 PAIN Abdominal pain or cramping 1 2 1 Headache 2 1 PULMONARY Pleural effusion (non-malignant) 1 Pneumothorax 1 RENAL/GENITOURINARY Proteinuria 3 SEXUAL/REPRODUCTIVE FUNCTION Male infertility 1

Pharmacokinetics

At the DL1 (400 mg) two patients had undetectable plasma concentrations of 2ME2.

Only one patient had detectable concentrations and a peak plasma concentration of 3.9

ng/mL was achieved after 30 minutes. At DL2 (800 mg) two patients out of three had

undetectable concentrations of 2ME2. The one patient with detectable concentrations of

2ME2 had a peak plasma concentration of 8.3 ng/mL after 4 hours. At DL3 (1600 mg),

five patients out of six had quantifiable concentrations of 2ME2, with a median peak

plasma concentration of 5.7 ng/mL with a range of 3.0 ng/mL to 18.6 ng/mL, which

occurred between 0.5 to 4 hours. At DL4 (2200 mg), two patients out of 3 had detectable

concentrations of 2ME2, with a median peak plasma concentration of 5.1 ng/mL and

ranged between of 3.4 ng/mL to 7 ng/mL, which occurred between 0.5 to 6 hours. At

DL5 (3000 mg), two patients out of three had quantifiable concentrations of 2ME2 with a

peak plasma concentration of 12.9 ng/mL with a range of 4.3 ng/mL to 21.7 ng/mL,

which occurred between 2 to 6 hours. For those patients with quantifiable

concentrations, noncompartmental PK parameters are summarized in Table 15. Figure

17 through 21 below show representative concentration time profiles (as modeled by

WINNONLIN®) in patients. Individual concentration-time profiles are attached in appendix A.

Figure 17: Plasma concentration-time profile of a patient that received a 400 mg single oral dose of 2ME2

Figure 18: Plasma concentration-time profile of a patient that received a 800 mg single oral dose of 2ME2

Figure 19: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

Figure 20: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

Figure 21: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

Table 17: Summary of Pharmacokinetic parameters of orally administered 2ME2

Dose* 400 mg 800 mg 1600 mg 2200 mg 3000 mg (N=1) (N=1) (N=5) (N=3) (N=2) Tmax (h) 0.5 4.0 1.5 (0.5-4.0) 0.5 (0.5-6.0) 4(2.0-6.0) Cmax 3.9 8.3 5.7 (3.0- 5.1 (3.4-7.0) 12.9 (4.3- (ng/mL) 18.6) 21.4) AUClast 24.6 19.4 138 (86-461) 64 (24-104) 309 (85-534) (ng*h/mL) Half-life (h) NC NC 27(24-486) 20 (1-179) 32 (18-45) CL/F (L/h) 9869 36815 4663 (1405- 17572 11138 8252) (6309- (2490- 81741) 19786) Vd/F (L) 78866 48172 328165 529079 343520 (90554- (171612- (163730- 985782) 1629918) 523311) * Median (range); N = number of patients; NC, not calculated; Tmax, time to peak concentration; Cmax, peak concentration; AUC, area under the curve; CL/F, apparent oral clearance; Vd/F, apparent volume of distribution.

There were no dose proportional increases or correlations with Cmax (Spearman rank

correlation coefficient r = 0.066; p=0.84) or AUClast (r = 0.22; p=0.49). Figure 4.3 and

4.4 depict box and whisker plots of Cmax and AUClast, respectively, versus dose. Except

for a moderately strong correlation between Vd/F and dose (r =0.55; p=0.065),all other

Spearman correlations with dose were weak, including those with Tmax, (r = 0.27; p=

0.39), t1/2 ( r = 0.23; p=0.48) and CL/F (r = 0.048; p= 0.88).

The Jonckheere-Terpstra trend test revealed that there was no trend for CL/F across dose levels and thus this parameter remained statistically constant with increasing dose (p value = 0.94). Given the positive moderately strong correlation with dose (above), Vd/F was found to have a weak (although statistically insignificant) tendency toward increasing across dose levels (p value = 0.097).

The Jonckheere-Terpstra trend test is a nonparametric test used when an ordered categorical parameter, such as dose level, is to be examined with regards to a trend for a continuously measured parameter (e.g., clearance) to increase or decrease to a statistically significant degree as one increases the dose level.

H0: all clearances equal across doses;

HA: there is a trend (up or down) in clearance as dose increases.

This test is a more appropriate test to use in the setting described above than either a

Kruskal-Wallis test (or ANOVA) which merely tests if there is a global difference among groups, and are not sensitive to a trend in values with increasing level of the categorical parameter. It also is somewhat more appropriate to use in the setting described above than correlation analyses, since the ordered categorical values (dose levels) are limited in number and fixed into a small set of labeled values, while for a correlation analysis, normally it better to assume that both parameters being correlated are continuous and free to range over a wide number of values 83.

Figure 22: Box and whisker plots of Cmax after a single oral dose. The box and whiskers represent the 10th,25th, 75th and 90th percentile of the data.

Cmax Vs Dose 25.00

18.75

Cmax 12.50

6.25

0.00 400 800 1600 2200 3000 Dose

Figure 23: Box and whisker plots of AUClast after a single oral dose. The box and whiskers represent the 10th, 25th, 75th and 90th percentile of the data.

AUC Vs Dose 600.00

450.00

AUC 300.00

150.00

0.00 400 800 1600 2200 3000 Dose

The Jonckheere-Terpstra trend test revealed that there was no trend for CL/F across

dose levels and was thus remaining constant (p value = 0.94). Figure 24 depicts box and

whisker plots of CL/F respectively versus dose. Given the positive moderately strong

correlation with dose (above), Vd/F was found to have a weak (although statistically

insignificant) tendency toward increasing across dose levels (p value = 0.097). Overall,

the high extent of variability prevented a formal conclusion with respect to linear/nonlinear pharmacokinetic assessment.

Figure 24: Box and whisker plots of CL/F after a single oral dose. The box and whiskers represent the 10th, 25th, 75th and 90th percentile of the data.

CL/F Vs Dose 40000.00

30000.00

CL/F 20000.00

10000.00

0.00 400 800 1600 2200 3000 Dose

2ME2 was undetectable in plasma of two patients each at DL 1 and 2, and one

patient each at DL 3, 4 and 5. Circulating plasma concentrations were very low at all

dose levels. There was extremely high interindividual pharmacokinetic variability of

2ME2. The drug appears to be absorbed rapidly, but incompletely. The median plasma

half-life was about 1-2 days, but highly variable. The unphysiologically high values for

oral CL/F and Vd/F reflect low oral bioavailability of 2ME2 presumably due to

dissolution rate-limited absorption and/or extensive first-pass elimination (see chapter 5

and 7).

Urinary Pharmacokinetics of 2ME2

24-Hour urine collections from four patients were analyzed for the amount of 2ME2

excreted unchanged in urine, glucuronides and sulfates (chapter 3). 2ME2 was detected

in all four patients whose 24-hour urines were analyzed. The median amount of 2ME2

excreted unchanged in urine was 0.002% (0.0015-0.0075 %). There was a significant

increase in 2ME2 concentration detected after the treatment of patient urines with β- glucuronidase. The median absolute amount of 2ME2 excreted in urine as glucuronides was 0.81% (0.013-1.32 %). Following table 18 provides a more detailed description of the above data and also provides a comparison with the Cmax obtained in each patient.

Table 18: Comparison of the Cmax achieved in the plasma versus the breakdown of excretion in urine.

Patient Cmax % Dose % Dose % Dose Dose (ng/mL) administered excreted excreted as excreted as (mg)

unchanged glucuronides sulfates

in urine

# 13 4.6 0.007 1.21 0.009 2200

# 6 1.2 0.002 0.01 NC 800

# 16 3.4 0.002 1.32 NC 2200

# 15 3.2 0.002 0.39 NC 2200

*Abbreviations: NC, not calculated; Cmax, peak concentration

The % dose excreted as sulfates of 2ME2 was not calculated for patients # 6, 15 and

16, because there was no increase in the 2ME2 peak area as measured in the patient urine

post-treatment with sulfatase (after de-sulfation). At the most, about 1.4 % of the total

administered dose of 2ME2 is accounted for in the urine. This suggests a lack of oral

absorption and/or excretion in the urine/bile as other unidentified metabolites. Majority

of the 2ME2 excreted in the urine is in the form of glucuronide metabolites and very little

is excreted as parent compound or sulfated metabolite. Although, the number of data

points available here are insufficient to make this conclusion, there appears no correlation

between the concentrations achieved in the plasma and the fraction of 2ME2 that is

excreted in the urine as glucuronides or unchanged. A larger sample size is required to

make this conclusion.

Correlative Studies

Cellular proliferation occurs through a defined process in which several phases can be recognized. From the resting (G0) phase they join the active cycling population after

appropriate stimuli and enter the first gap (G1) phase. Both phases have a highly variable

duration. In G1, the cell prepares for the synthesis (S) phase, in which DNA synthesis and doubling of the genome take place. The S phase is followed by a period of apparent

inactivity known as the second gap (G2) phase, in which the cell prepares for further

separation of chromatids during the mitotic (M) phase84. Ki67 labeling correlates with

the S phase fraction85 and mitotic index.85,86 Using frozen sections, the Ki67 labeling

index was prognostically relevant in several studies in invasive breast cancer 87,88.

Expressed on hematopoietic stem cells, CD31 is also expressed on endothelial cells and belongs to the immunoglobulin superfamily. It mediates adhesion between cells

(platelets, monocytes, polymorphonuclear cells, endothelial cells, and discrete populations of circulating lymphocytes) that express CD31. It is an adhesion receptor molecule, responsible for transmitting signals through the adhesion cascade, which involves the integrin family of proteins. CD 31 is generally regarded as the most sensitive and specific endothelial marker in paraffin sections89.

We performed CD 31 staining (marker of microvessel density) before and after therapy for seven patients, and Ki67 staining for eight patients. Figure 25 and 26 depict the absolute changes in Ki67 staining and microvascular density from the baseline after being on therapy for at least two months. Ki67 is a marker for cell proliferation, hence after treatment with a true antiproliferative agent there would be a decline in Ki67

staining in serial tissue biopsies. Figure 25 shows that there was a minor decline in Ki67

staining in three patients (# 1, 6 and 5) after treatment with 2ME2, but statistically 2ME2

had no effect on cell proliferation by Ki67 staining (p value = 0.47; paired t-test).

Microvessel density is a marker for angiogenesis; hence after treatment with a true

antiangiogenic agent, there would be decrease in microvessel density in serial tissue biopsy staining. Figure 26 shows that there is no noticeable decline in microvessel density, except for one patient (patient #9), but statistically 2ME2 had no significant effect on microvascular density (p value = 0.096; paired t-test) by CD 31 staining. Thus,

2ME2 did not produce any antiangiogenic or antiproliferative effects in the tissue biopsies of patients that were examined in this clinical trial.

Figure 25: Changes in Ki67 staining in pretreatment tissue biopsies and after 2 months of 2ME2 treatment

) 40 % ( ng

ni 30 ai t s 67

Ki 20

0 Pre Therapy ≥ 2 months of therapy

Figure 26: Changes in microvessel density staining in pretreatment tissue biopsies and after 2 months of 2ME2 treatment

400

350 c mm) i 300 b cu

er 250 p y ( t i s 200 n e D

r a

l 150 scu a

v 100 o cr i

M 50

0 Pre Therapy ≥ 2 months of therapy

Pharmacodynamic correlation

There were no significant adverse effects seen with 2ME2, nor there was any evidence

of any biological activity at any dose level except for the patient with ovarian cancer who

showed a partial response. 2ME2 also had no effect on the microvessel density and cell

proliferation as seen with immunohistochemistry studies. Thus, pharmacodynamic

correlation with any drug effect was not attempted.

Discussion

Ovarian cancer is the fifth most common form of cancer in women in the United states, accounting for 4% of the total number of cases and 25% of those cases that occur in the female genital tract and it is responsible for 5% of all cancer deaths in women 90.

Of the various histological subtypes of ovarian cancer, clear cell carcinoma, constitutes 5-

10% of ovarian cancer. Clear cell carcinoma is usually more resistant to systemic chemotherapy than other types, and has a worse prognosis 91,92. Elevated vascular endothelial growth factor (VEGF) levels have been reported with epithelial ovarian cancer 93 and hypoxia inducible factor-1α (HIF-1α) is a transcription factor that

modulates the transcription of VEGF 94. 2ME2 downregulates HIF-1 at the

posttranscriptional level and inhibits HIF-1 induced transcriptional activation of VEGF

expression 95.

The current study was performed to explore the safety, feasibility and pharmacokinetics of 2ME2 administered orally to patients with solid tumors. In this study, we found that 2ME2 was very well tolerated orally at all dose levels. There were

no dose limiting toxicities (DLT), and maximum tolerated dose (MTD) was not reached.

However, there were some patients that showed evidence of hot flashes, fluid retention

and thrombosis, typical of estrogenic compounds. Nonetheless, there were no significant

toxicities reported even at the highest dose levels. 2ME2 showed a very promising

response in one patient with clear cell carcinoma of the ovary. Four patients with ovarian

cancer (adenocarcinomas) treated with 2ME2 at equal or higher were taken off the trial

within less than two months because of progressive disease. The perceived clinical

efficacy of 2ME2 in one patient with ovarian cancer is speculated to be due to its

histological subtype, although the clinical efficacy of 2ME2 was not reflected in the

immunohistochemistry studies performed. The potential role of 2ME2 in the treatment of

ovarian clear cell carcinoma needs to be further investigated.

Nuclear receptors for estrogen, progesterone, and androgens are expressed in

ovarian epithelial tumor cells. The majority of ovarian cancers express ER whereas far

fewer tumors express PR 96. Despite high expression of ERα, the response rate to hormonal therapy, such as tamoxifen, in ovarian cancer is low and the lack of PR expression in ER-positive ovarian tumors has been suggested to indicate impaired estrogen action 97. It has also been suggested that expression of estrogen receptor β (ERβ) may contribute to determination of sensitivity to hormonal agents. ERα and ERβ have distinct patterns of tissue expression and can exist as homo- or heterodimers that may have different responses to estrogens and antiestrogens. ERβ is expressed in the normal ovary and most studies have found expression to either be similar or to decrease in ovarian tumors. By comparing receptor mRNA levels in ovarian cancer cells to those found in ovarian epithelial cells, disruptions of ERα, PR, and AR mRNA expression in

100

cancer cells have been noted96. It is suggested that progesterone and/or androgen actions

may protect ovarian epithelial cells from neoplastic transformation 96. Gene expression

patterns have revealed that estrogen receptor 1 gene is expressed at relatively low levels

in clear cell cancers, compared with other ovarian cancers98. Overexpression of PR predicts a favorable outcome in patients with ovarian cancer99. 2ME2 has been recently shown to bind significantly to ER when used at concentrations that are cytotoxic and antiangiogenic 100, which may help explain the partial response seen in the patient with

clear cell carcinoma of the ovary as compared to other ovarian cancer subtypes, which

did not respond to 2ME2 therapy.

Large interpatient variability in the derived pharmacokinetic parameters exists which

need to be further explored. The apparent lack of dose proportionality with increasing

doses of 2ME2 and extremely low circulating plasma concentrations point to either poor

solubility and/or gastrointestinal permeability and/or extensive first pass metabolism

resulting in poor and variable bioavailability. There was no statistically significant

increase in Cmax or AUClast of 2ME2 after administration of single oral doses ranging

from 400 mg to 3000 mg.

Sulfation of estrogens is catalyzed by SULT 1E1. Sequencing of SULT 1E1 exons

revealed three nonsynonymous coding SNPs (cSNPs) that altered the following encoded

amino acids: Asp22Tyr, Ala32Val and Pro253His. Among these, 12 pairs of SNPs were

tightly linked. In addition, 12 unambiguous SULT1E1 haplotypes were identified,

including six that were common to Caucasian and African Americans 100. There is a two-

to three-fold increase in K(m) values for the His253 allozyme and a greater than five-fold

increase for the Tyr22 allozyme, thus raising the possibility of pharmacogenetic variation

101

of estrogen sulfation in different populations, and thus be a potential source of

pharmacokinetic variability of 2ME2.

UGT1A1, UGT1A3, and UGT2B family of isoenzymes catalyze catechol

estrogen glucuronidation 24,25. When the glucuronidation kinetics of expressed human

UGTs with catechol estrogens was tested, it was found that UGT2B6(Y) reacted with a

higher efficiency toward 4-hydroxyestrogen catechols, whereas UGT1A1 and UGT1A3

showed higher activities toward 2-hydroxyestrogens 26. It has also been shown that

enzymatic de-conjugation of estrogen glucuronides occurs by β-glucuronidases, and that it may be an important source of both primary estrogens and their catechol metabolites in preclinical models 27. Pharmacogenetic variants of UGT1A1*1 and *28 and *37 have

been shown to have altering effects on the degree of glucuronidation 100. Thus, pharmacogenetic variation of different UGTs can potentially be a source of pharmacokinetic variability of 2ME2.

The homozygous COMTLL allele is about 4- to 5-fold less effective in

methylating catechol substrates in vitro as compared to the wild-type allele 101. Serum levels of estradiol are significantly higher in postmenopausal women carrying the homozygous variant or the heterozygous (COMTHL) genotype as compared to those with

the wild-type genotype after the oral administration of estradiol valerate tablets (dose, 2

mg) 102. Since this COMT polymorphism correlates with decreased enzyme activity, and

because it is a major pathway responsible for the metabolism of carcinogenic catechol

estrogens, it is believed that the presence of the COMTLL polymorphism is associated

with an increased risk of breast, endometrial, and ovarian cancer (see below). Thus

102

COMT pharmacogenetics may also be a potential source of pharmacokinetic variability

of 2ME2.

The bioavailability of other orally administered estrogens is known to be poor because

of significant first pass effect 71. There has been a widespread recognition of the

variability in bioavailability of orally formulated cancer chemotherapeutic drugs 103. An ideal chemotherapeutic drug would have little interpatient and intrapatient variability with successive doses. A significant inverse correlation between decreasing absolute bioavailability and intersubject variability in absolute bioavailability has been established

104. Hence, drugs with poor bioavailability are more likely to show high interpatient and

intrapatient variability.

Studies performed over the last several years have shown that the ATP-binding

cassette transporter, ABCG2 [also known as breast cancer resistance protein (BCRP), or

mitoxantrone resistance protein (MXR)], may be involved in transport or efflux of

estrogens. Biing et al 105 showed that estrogens are capable of induction of ABCB1 gene

(previously known as MDR1), which codes for P-glycoprotein, one of the major factors

determining multidrug resistance in cancers. But, recent studies suggest that estradiol

may not be a substrate for P-glycoprotein 106. The ABCG2 gene is also highly expressed

in several healthy organs such as the intestine 107, and it has been proposed that the use of specific inhibitors of this protein could be useful in making substrates orally available 108.

Certain estrogen agonists (e.g., estriol and diethylstilbestrol) as well as antagonists (e.g., tamoxifen and its analogs) have been shown to reverse the ABCG2-mediated resistance to SN-38, mitoxantrone, and topotecan in cell culture 109. It has been hypothesized that

this effect of estrogens is related to a competitive inhibition of ABCG2 activity 110. In

103

addition, estrogens have been implicated in the regulation of ABCG2 protein expression

by producing a dose-dependent decrease in mRNA levels 111. More recently, investigations have pointed to the notion that ABCG2 may be involved in the transport of unconjugated estradiol 112 as well as the sulfate-conjugates of estrone and 17β-estradiol

113.

Two drug properties are of paramount importance that control a drug’s rate and extent

of oral absorption, are solubility and gastrointestinal permeability. 2ME2 is structurally a

steroid and falls in the Class 2 of the Biopharmaceutical Classification System

(BCS)114,115 (see chapter 7). 2ME2 possesses characteristics of a Class 2 compound and exhibits poor solubility and high permeability. The poor solubility, high permeability argument is reinforced by the results that have been observed in this clinical trial. As indicated by very low circulating plasma concentrations of 2ME2, absence of any major adverse events and the results of urinary excretion pharmacokinetics, it is evident that

2ME2 is poorly absorbed from the gastrointestinal tract. About 1% of the total dose administered is recovered in the urine after 24 hours (chapter 5), which reflects the poor

bioavailability of this compound. 2ME2 also undergoes a high hepatic first-pass effect

(see chapter 5), which is also partly responsible for low plasma circulating concentrations. High hepatic first pass effect, low plasma concentrations of 2ME2 and the fact that majority of 2ME2 recovered in the urine is glucuronidated, lead to the conclusion that high circulating concentrations of metabolites (mostly glucuronides) of

2ME2 are very likely. Factors responsible for the poor oral bioavailability and low circulating concentrations of 2ME2 are discussed in further detail in chapter 7.

104

Collectively, the results of pharmacokinetic data supported the decision to terminate the study prior to reaching the primary objective of determining the MTD. This once more strengthened the importance of performing pharmacokinetic analyses during a phase I study in order to guide the decision making process. From an ethical point of view, unfavorable pharmacokinetic results should prevent unnecessary exposure of patients to potential harmful and less effective cytotoxic drugs. New formulations of

2ME2 are currently being developed with the intent of increasing oral bioavailability.

Thus, objective #2 and #3 of this thesis as listed in chapter 2 was achieved. Hypothesis

#1 was established to be partly true as 2ME2 was well tolerated in the clinic but did not reach concentrations high enough in the plasma to have quantifiable effects on the markers of angiogenesis. Hypothesis #2 proved wrong as 2ME2 did not exhibit linear pharmacokinetics.

CHAPTER 5

In vitro hepatic metabolism and in vivo elimination pathways of 2ME2

Introduction

Interindividual variability in 2ME2 pharmacokinetics, toxicity and response as seen in

Chapter 4 was extensive, and largely unexplained. There has been a widespread

recognition of the variability in bioavailability of orally formulated cancer

chemotherapeutic drugs 103. An ideal chemotherapeutic drug would have little interpatient and intra-patient variability with successive doses. A significant inverse correlation between decreasing absolute bioavailability and intersubject variability in absolute bioavailability has been established 104. Hence, drugs with poor bioavailability are more likely to show high interpatient and intrapatient variability. One of the reasons proposed for poor bioavailability and high inter-individual variability in 2ME2 pharmacokinetics is extensive metabolism and inter-individual variability in the rate and extent of 2ME2 metabolism. The bioavailability of orally administered estrogens is known to be poor because of significant first-pass effect 71. The aim of this study was to

define the in vitro and in vivo metabolic fate of 2ME2, characterize the enzyme kinetics

of the metabolites of 2ME2 and to estimate the total hepatic clearance of 2ME2 (CLh). 106

Materials for metabolism experiments

Chemicals and reagents

2ME2 (HPLC purity, 100%) was obtained from Entremed Inc. (Rockville, MD,

USA). Methanol (HPLC grade) and acetonitrile (HPLC grade) were obtained from

Mallinckrodt Baker, Inc (Phillipsburg, NJ, USA). Water was filtered and de-ionized

using Milli-Q-UV plus system (Millipore, Milford, MA, USA). Pooled human liver

microsomal protein (HLM), pooled human liver S9 fraction and recombinant CYP

isoenzymes were obtained from Xenotech (Lenexa, KS, USA). Nicotinamide adenine

dinucleotide (NADP+), D-glucose-6-phosphate (G6P), glucose-6-phosphate

dehydrogenase (G6PDH), magnesium chloride (MgCl2), uridine diphosphate glucuronic

acid (UDPGA), alamethicin (channel-forming antibiotic known to increase the

permeability of biological membranes through the creation of transmembrane pores up to

20 A in diameter), saccharolactone (an inhibitor of β-glucuronidases), adenosine-3’- phosphate-5’-phosphosulfate (PAPS), β-glucuronidase and sulfatase were obtained from

Sigma-Aldrich Co. (St Louis, MO, USA). 1M Tris buffer was obtained from Quality

Biological Inc. (Gaithersburg, MD, USA). 2,6 Dichloro-4-nitrophenol (DCNP) was obtained from Alfa Aesar (Ward Hill, MA, USA).

In vitro hepatic phase I metabolism experimental

Microsomal incubations were preparing using pooled HLM protein (1 mg/mL),

Tris-HCl buffer (100 mM, pH 7.4), 2ME2 and an NADPH-generating system containing magnesium chloride (4 mM). NADPH generating system was prepared by adding

0.5 mL of NADP+ (10 mM), 32 units of glucose-6-phosphate dehydrogenase (1300

107

U/mL), 1 mL of D-glucuose-6-phosphate (0.1 M), 900 µL de-ionized water, and 100 µL magnesium chloride (1 M). Microsomal incubations without NADPH-generating system were used as negative controls. Incubations were performed at 370C in a shaking water bath. Reactions were terminated with the addition of twice the reaction volume of acetonitrile. The reaction mixture was centrifuged at 3000 RPM for 15 minutes and the supernatant was analyzed by liquid chromatography method described below. Typical range of substrate concentrations in in vitro metabolism experiments ranges from 0.1-100

µM. The linearity of metabolite formation with human liver microsomes was established by incubating 2ME2 at concentrations ranging from 50 nM – 1000 µM. Metabolite peaks were observed only at substrate concentrations 100 µM and above due to the analytic limitations of the HPLC/UV method of detection. The linear range of metabolite formation for substrate concentration was established between 50-300 µM. All metabolite kinetics experiments were performed with 2ME2 concentration of 100 µM.

After establishing the optimal substrate concentration, the linearity of metabolite formation for protein concentrations was determined using 0.2, 0.4, 0.6, 0.8 and 1 mg/mL concentrations of human liver microsomes. After determining the optimal protein concentration, the linearity of metabolite formation for various incubation times was established by incubation 2ME2 with HLM protein for 0, 15, 30, 45, 60, 75, 90 and 120 minutes.

HPLC method used for metabolite quantification

Chromatography was performed on an HP 1100 system (Agilent Technology,

Palo Alto, CA, USA), which included a binary pump, a refrigerated autosampler, a

108 degasser and a photodiode-array detector. Chromatographic separations were performed at ambient temperature on a Phenomenex Prodigy (250 × 4.6 mm i.d.; 5µ PS;

Phenomenex, Torrance, CA, USA) using a mobile phase composed of acetonitrile and water. The mobile phase was delivered as a gradient at 1mL/min flow rate for 40 minutes

(Table 19), and the effluent was monitored at a UV wavelength of 205 nm. Injection volume was 150 µL. Chromatographic data were collected and analyzed using the

Chemstation software (Agilent). Calibration graphs were calculated by least-squares linear regression analysis of the peak area of 2ME2 versus the drug concentration of the nominal standard (x). The concentration of metabolites produced was estimated from a standard curve of the parent compound 2ME2, assuming equal molar absorptivities for the parent 2ME2 and the metabolites. A peak separated on HPLC was considered a metabolite peak if all the below criteria were satisfied:

1. The retention time of the metabolite peak was different from the parent compound or any other peaks in the background.

2. Peak area of the peak increased with increase in substrate concentration and or/time of incubation.

3. The peak could be successfully reproduced in at least 3 different individual incubations.

4. The metabolite peak was not produced in the negative control samples and in the samples of the standard curve.

109

Table 19: HPLC gradient for metabolite quantification

Time (min) % Acetonitrile % Water

0 10 90

3 10 90

20 80 20

25 80 20

30 10 90

40 10 90

Metabolism of 2ME2 by human hepatic CYP 450 isoenzymes

Microsomal incubations were prepared using recombinant human CYP P450

enzymes (0.22 mg/mL), Tris-HCl buffer (100 mM, pH 7.4), 2ME2 (300 µM), and an

NADPH-generating system (prepared as described above). Recombinant CYP isozyme

incubations without NADPH-generating system were used as negative controls.

Incubations were performed at 370C in a shaking water bath. Reactions were terminated

with the addition of twice the reaction volume of acetonitrile. The reaction mixture was

centrifuged at 3000 RPM for 15 minutes and the supernatant was analyzed by liquid

chromatography method described above. The CYP isozymes tested were CYP1A1,

CYP 1A2, CYP 1B1, CYP 2A6, CYP 2B6, CYP 2C9, CYP 2C19, CYP 2D6, CYP 3A4,

CYP 3A5 and CYP 2E1. The linear range of metabolite formation for substrate

concentration was established between 100-700 µM. After establishing the optimal

110

substrate concentration, the linearity of metabolite formation for protein concentrations was determined using 0.05, 0.11, 0.22, 0.44, 0.66 and 1 mg/mL concentrations recombinant human CYP isozymes. After determining the optimal protein concentration, the linearity of metabolite formation for various incubation times was established by incubation 2ME2 with HLM protein for 0, 15, 30, 45, 60, 75, 90 and 120 minutes.

Enzyme Kinetics and estimation of hepatic clearance of 2ME2

The results obtained from in vitro metabolism experiments were model-fitted using the Michaelis-Menten equation and the kinetic parameters were calculated by nonlinear regression analysis using SigmaPlot7 (SPSS Inc, USA). The in vitro intrinsic clearance

(CLint) was estimated by the Vmax/Km ratio. The predicted CLint was scaled up to in vivo

extrapolated CLint assuming a human liver of 1.5 kg, and the conversion factor of 100 mg liver tissue contains 1 mg HLM protein 116. The total hepatic clearance was calculated using the following formula from the well-stirred model:

CLh = Q × fu×CLint/(Q + fu × CLint)

Where CLh is the total hepatic clearance, fu is the free fraction of 2ME2 in plasma, CLint being the total in vitro intrinsic clearance (sum of all pathways). The human hepatic blood flow was assumed to be 1500 mL/min and the fu of 0.27 (from chapter 6) was used

in these calculations.

Results

In vitro hepatic phase I metabolism results

111

After incubation of 2ME2 with HLM and NADPH generating system, there was

appearance of four major metabolite peaks, i.e. A, B, C, D (Figure 27, 2ME2

concentration = 200 µM) as seen after HPLC separation. The metabolites were named A

(TR = 14.7 minutes), B (TR = 16.3 minutes), C (TR = 17.7 minutes) and D (TR = 25.1 minutes). These metabolites were not produced in the negative control (Figure 5.1 B).

112

Figure 27: Chromatograms of human liver microsomal incubations of 1000 µM 2ME2 optimized for hepatic phase I metabolism with NADPH generating system

400

D 300 2ME2

200 mAU C B 100 A

0 5 10 15 20 25 Time (min)

113

Figure 28: Chromatograms of human liver microsomal incubations of 1000 µM 2ME2 optimized for hepatic phase I metabolism without NADPH generating system

400

300 2ME2

200 mAU

100

0 5 10 15 20 25 Time (min)

114

Michaelis-Menten Kinetics of Phase I metabolites

The metabolite data yielded good fits for Michaelis-Menten kinetics with a single binding site. Figures 29-32 depict the Michaelis-Menten fitting of data using SigmaPlot® for phase I metabolites of 2ME2.

115

Figure 29: Michaelis-Menten Kinetics for Metabolite A (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Metabolite A Kinetics

0.006

0.005

0.004

0.003

0.002 V [nmol/min/µg]

0.001

0.000 0 200 400 600 800 1000 1200

S [µM]

116

Figure 30: Michaelis-Menten Kinetics for Metabolite B (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Metabolite B Kinetics

0.0035

0.0030

0.0025 ] µg / 0.0020 l/min 0.0015 mo n

V [ 0.0010

0.0005

0.0000 0 200 400 600 800 1000 1200

S [µM]

117

Figure 31: Michaelis-Menten Kinetics for Metabolite C (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Metabolite C Kinetics

0.0035

0.0030

0.0025

0.0020

0.0015

V [nmol/min/µg] 0.0010

0.0005

0.0000 0 200 400 600 800 1000 1200

S [µM]

118

Figure 32: Michaelis-Menten Kinetics for Metabolite D (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Metabolite D Kinetics

0.010

0.008 ]

/µg 0.006

mol/min 0.004 n V [

0.002

0.000 0 200 400 600 800 1000 1200

S [µM]

119

Table 20 provides the apparent Km, Vmax, CLint and in vivo extrapolated CLint values for all the four phase I metabolites produced.

Table 20: Kinetic parameters (mean estimate ± standard error) calculated from 2ME2 incubations with HLM protein and NADPH generating system

______

2 Metabolite Km Vmax R of fit CLint in vivo extrapolated CLint

(µM) (pmol/min*µg) (µL/min*µg) (mL/min) ______

A 90.0 ± 11.3 6.0±0.0002 0.9312 0.09 1350

B 60.0 ± 5.9 3.1 ± 0.0001 0.9504 0.077 1161

C 59.8 ± 4.4 3.0 ± 0.0 0.9707 0.075 1127

D 103.0 ± 14.1 9.0 ± 0.0003 0.9228 0.131 1966

______

120

Phase I metabolite identification

The structural elucidation of the major phase I metabolites of 2ME2 was performed by

a collaborating laboratory using LC/MS/MS and were found to be 2-methoxyestrone

(2ME1), 2-hydroxyestradiol (2OHE2), 2OHE1 (2-hydroxyestrone), 2, 3- dimethoxy-17β estradiol and 2, 3-dimethoxyestrone. Figure 33 depicts the structures of phase I metabolites.

121

Figure 33: Metabolic scheme of phase I metabolism of 2ME2

O OH

HO HO

HO HO 2OHE2 2OHE1

OH O

O O

HO HO 2ME1 2ME2

OH O

O O

O 2,3 dimethoxy E1 2,3 dimethoxy-17β E2

122

Metabolism of 2ME2 by human hepatic CYP 450 isoenzymes

After incubation of 2ME2 with recombinant CYP isoenzymes and NADPH generating

system, metabolite peaks were seen in CYP 1A1 (Figures 34, 35), 1A2 (Figures 36, 37),

3A4 (Figures 38, 39), 3A5 (Figures 40, 41) and 2E1 (Figures 42, 43) seen after HPLC

separation. Incubation of 2ME2 with CYP 1A1 produced a metabolite peak at TR = 15.1

minutes. Incubation of 2ME2 with CYP 1A2 produced a metabolite peak at TR = 18.7

minutes. Incubation of 2ME2 with CYP 3A4 produced three metabolite peaks at TR =

16.4, 17.7 and 18.5 minutes. Incubation of 2ME2 with CYP 3A5 produced three metabolite peaks at TR = 15.7, 16.4 and 17.7 minutes. Incubation of 2ME2 with CYP

2E1 produced two metabolite peak at TR = 17.7 and 19.5 minutes. It was observed that multiple isoforms produced metabolites with the same retention times. If we compare retention times of the produced metabolites, metabolite A (TR = 14.7 minutes) was not

produced by any isoforms, metabolites B (TR = 16.3 minutes) and C (TR = 17.7 minutes) were produced by CYP 3A4, 3A5 and 2E1 and metabolite D (TR = 25.1 minutes) was not produced by any isoform Incubation of 2ME2 with recombinant CYP 1B1, 2A6, 2B6,

2C9, 2C19 and 2D6 did not produce any metabolite (chromatograms in appendix C).

123

Figure 34: In vitro metabolism of 2ME2 by CYP 1A1

1A1

400

300 2ME2

metabolite 200 mAU

100

0 5 10 15 20 Time (min)

124

Figure 35: Negative Control for CYP 1A1 incubations

1A1 negative control

400

300

2ME2

200 mAU

100

0 5 10 15 20 Time (min)

125

Figure 36: In vitro metabolism of 2ME2 by CYP 1A2

CYP 1A2

200

150

2ME2

100 Metabolite mAU

0 5 10 15 20 Time (min)

126

Figure 37: Negative Control for CYP 1A2 incubations

1A2 negative control

200

150

2ME2 100 mAU

0 5 10 15 20 Time (min)

127

Figure 38: In vitro metabolism of 2ME2 by CYP 3A4

CYP 3A4

200

2ME2

150

Metabolites 100 mAU

0 5 10 15 20

Time (min)

128

Figure 39: Negative Control for CYP 3A4 incubations

Negative Control for CYP 3A4

200

150

2ME2

100 mAU

0 5 10 15 20 Time (min)

129

Figure 40: In vitro metabolism of 2ME2 by CYP 3A5

CYP 3A5

200

2ME2

150

Metabolites 100 mAU

0 5 10 15 20

Time (min)

130

Figure 41: Negative Control for CYP 3A5 incubations

Negative Control for CYP 3A5

200

150 2ME2

100 mAU

0 5 10 15 20

Time (min)

131

Figure 42: In vitro metabolism of 2ME2 by CYP 2E1

CYP 2E1

200

2ME2 150 Metabolites

100 mAU

0 5 10 15 20

Time (min)

132

Figure 43: Negative Control for CYP 2E1 incubations

Negative Control for CYP 2E1

200

150 2ME2

100 mAU

0 5 10 15 20 Time (min)

133

In vitro hepatic phase II metabolism experimental

Uridine diphosphate glucuronosyltransferase (UGT) is localized to the

endoplasmic reticulum (ER). The major portion of the protein is located in the ER

lumen, including the proposed substrate-binding domains and the catalytic site. The

microsomal membrane impedes the access of UDPGA to the active site, resulting in

latency of UGT activity in intact ER-derived microsomes. Active transport of UDPGA is

believed to occur in hepatocytes, but the transport system has not been fully

characterized. Alamethicin, channel-forming antibiotic known to increase the

permeability of biological membranes through the creation of transmembrane pores, has

been shown to increase the accessibility of UDPGA to the active site.

Microsomal incubations for glucuronidation experiments were prepared using

pooled HLM protein (0.5 mg/mL), Tris buffer (100mM, pH 7.4), UDPGA (5mM),

alamethicin (50 µg/mL), 1-4 saccharolactone (5mM), MgCl2 (5mM), 2ME2 (200 µM)

and de-ionized water. Incubations were performed at 370C in a shaking water bath.

Microsomal incubations without UDPGA were used as negative controls. Reactions were terminated with the addition of twice the reaction volume of acetonitrile. The reaction mixture was centrifuged at 3000 RPM for 15 minutes and the supernatant was analyzed by liquid chromatography method described previously.

The linearity of metabolite formation with human liver microsomes was established by incubating 2ME2 at concentrations of 50, 75, 100, 200, 300, 500, 700 and

900 µM. Quantifiable metabolite peaks were observed at all substrate concentrations.

The linear range of metabolite formation for substrate concentration was established between 50-900 µM. All metabolite kinetics experiments were performed with 2ME2

134

concentration of 200 µM. After establishing the optimal substrate concentration, the linearity of metabolite formation for protein concentrations was determined using 0.2,

0.4, 0.6, 0.8, 1 and 1.2 mg/mL concentrations of human liver microsomes. After determining the optimal protein concentration, the linearity of metabolite formation for various incubation times was established by incubation 2ME2 with HLM protein for 0,

15, 30, 45, 60, 75, 90, 120 and 150 minutes.

Microsomal incubations for sulfation experiments were prepared using pooled human liver S9 fraction protein (0.5 mg/mL), Tris buffer (100mM, pH 7.4), PAPS (400

µM), 2ME2 and de-ionized water. Microsomal incubations with DCNP (10 µM) were

used as negative controls. Incubations were performed at 370C in a shaking water bath.

Reactions were terminated with the addition of twice the reaction volume of acetonitrile.

The reaction mixture was centrifuged at 3000 RPM for 15 minutes and the supernatant

was analyzed by liquid chromatography method described previously.

In vitro phase II metabolism results

After incubation of 2ME2 with the system optimized for glucuronidation, there was

appearance of two major metabolites, i.e. G 1 and G 2 (Figure 44, 2ME2 concentration

200 µM) as seen after HPLC separation. These metabolites were not produced in the

negative control (Figure 45).

135

Figure 44: In vitro glucuronidation of 2ME2 in human liver microsomes

Glucuronides of 2ME2

400

300

2ME2 200 mAU

G1 G2

100

0 0 5 10 15 20 Time (min)

136

Figure 45: Negative control for in vitro glucuronidation of 2ME2 in human liver microsomes

Negative Control for Glucuronidation

400

300 2ME2

200 mAU 100

-100 0 5 10 15 20 Time (min)

137

Michaelis-Menten Kinetics of glucuronides of 2ME2

The metabolite data yielded good fits for Michaelis-Menten kinetics with a single binding site. Figures 46 and 47 depict the Michaelis-Menten fitting of data using

SigmaPlot® for glucuronides of 2ME2.

Figure 46: Michaelis-Menten Kinetics for Glucuronide 1 produced after incubation of 200 µM 2ME2 in human liver microsomes (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Glucuronide 1 Kinetics

0.3 ] g m

n/ 0.2 mi / mol

µ 0.1 V [

0.0 0 2500 5000 7500 10000 12500 S [µM]

138

Figure 47: Michaelis-Menten Kinetics for Glucuronide 2 produced after incubation of 200 µM 2ME2 in human liver microsomes (symbols represent measured observations; curve as predicted by Michaelis-Menten Kinetics)

Glucuronide 2 Kinetics

0.100 ] g 0.075 m n/ mi / 0.050 mol µ

V [ 0.025

0.000 0 2500 5000 7500 10000 12500 S [µM]

139

Table 20 provides the apparent Km, Vmax, CLint and in vivo extrapolated CLint values for the two glucuronides of 2ME2.

Table 20: Kinetic parameters (mean ± standard error) for two glucuronides of 2ME2 ______

2 Glucuronide Km Vmax R of fit CLint in vivo extrapolated CLint

(µM) (µmol/min*mg) (mL/min*mg) (mL/min) ______

1 90.3± 22.9 0.215 ± 0.012 0.8428 3.58 53700

2 92.8 ± 30.2 0.065 ± 0.004 0.6002 1.06 15898

______

140

Results of sulfation experiments

After incubation of 2ME2 with system optimized for sulfation, there was no

appearance of new metabolite peaks at the end of one hour incubation period. There was no significant difference between peak area of 2ME2 measured in incubations with

PAPS, with DCNP (inhibitor of sulfation) and without PAPS. These results were reproduced at three separate occasions where each sample was analyzed in duplicates.

Figure 5.11 depicts the chromatogram obtained after incubation of 2ME2 for sulfation with PAPS (Figure 48) and without PAPS (Figure 49).

141

Figure 48: In vitro sulfation of 2ME2 with human liver S9 fraction

400

300

2ME2

200 mAU

100

0 5 10 15 20 Time (min)

142

Figure 49: Negative control for in vitro sulfation of 2ME2 with human liver S9 Fraction

400

300 2ME2

200 mAU

100

0 5 10 15 20 Time (min)

143

Estimation of total hepatic clearance

All in vitro intrinsic clearances (CLint) reported above in Table 18 and 19 for the

metabolites A, B, C, D, G 1 and G 2 were extrapolated to estimate the in vivo CLint

(mL/min) for a human liver of 1.5 kg. The assumption used for scaling was 100 mg

human liver contains 1 mg of human liver microsomal protein. The total in vivo,

extrapolated CLint of 2ME2 for a human liver was calculated as the sum of all individual

in vivo extrapolated CLint. The total in vivo extrapolated CLint thus obtained was used in the equation of the well-stirred model to obtain total hepatic clearance (CLh). The total

hepatic clearance was reported to be 862 mL/min. This number is more than half of

hepatic blood flow which is assumed to be 1500 mL/min or 750 mL/min of plasma. The

CLh is calculated in relation to plasma. Hence, CLh of 2ME2 exceeds the flow of plasma

through the liver. This suggests that 2ME2 is a high-extraction ratio drug. This number

is believed to be an underestimation because biliary excretion after hepatic metabolism is

assumed to be negligible. Also, glucuronidation of phase I metabolites is not taken into

account. We have reported low circulating concentrations of 2ME2 in the human plasma

after administration of high doses orally (chapter 4). This finding corroborates the

observation of poor circulating concentrations of 2ME2 after oral absorption, thus

indicating that 2ME2 has poor oral bioavailability and high first-pass effect is one of the

major factors determining its oral bioavailability.

144

Characterization of in vivo elimination pathways for 2ME2

Calculation of metabolic ratio

The metabolic ratio for glucuronides and sulfates in the urine was calculated using the

following formula:

⎛ Amount of 2ME2 excreted as conjugate (glucuronide/sulfate) ⎞ Metabolic Ratio = ⎜ ⎟ ⎝ Amount of 2ME2 excreted unchanged ⎠

Following table 22 provides the metabolic ratios as calculated from the data obtained after analysis of 24-hour urine samples:

Table 22: Metabolic ratios

Patient Metabolic Ratio for Metabolic Ratio for glucuronidation sulfation

#13 162 1

#6 7 NC

#16 634 NC

#15 251 NC

*Abbreviations: NC, not calculable

Thus, 2ME2 was excreted in the urine predominantly as glucuronides. Very little

amount was excreted unchanged or as sulfates.

145

Discussion

Glucuronidation by uridine-5’-diphosphate glucuronosyltransferase (UGT) of

estrogens is generally accepted to play a principal role in elimination pathways followed

by excretion of the conjugates through urine and feces. Results of previous studies

indicate that the UGT1A1, UGT1A3, and UGT2B family of isoenzymes catalyze

catechol estrogen glucuronidation 24,25. The two major glucuronides of 2ME2 are G1 and

G2. The total CLint for glucuronidation pathways is at least 12,000 fold the total CLint for the hepatic phase I metabolism pathways, and the total in vivo extrapolated CLint for all glucuronidation pathways is at least 12 fold the total in vivo extrapolated CLint for all hepatic phase I metabolism pathways. Hence, we conclude that the predominant route of metabolism for 2ME2 is through glucuronidation and subsequent excretion in urine and through biliary excretion in feces. The reported apparent Km values for glucuronidation of estradiol in human liver microsomes was 17 µM for the 3-glucuronide of estradiol and

117 6 µM for the 17β-glucuronide of estradiol . The apparent Km values reported for

118 human hepatic glucuronidation of ethinyl estradiol is 200 ± 120 µM . The Km obtained

for glucuronidation of 2ME2 was significantly higher than the previously reported values

for estradiol glucuronidation, but was more akin to the apparent Km values obtained from

ethinyl estradiol glucuronidation, taking the standard deviation into account. Hence,

2ME2 seems to have lesser affinity for hepatic glucuronosyltransferases as compared to

estradiol. The reported apparent Vmax values for glucuronidation of estradiol in human liver microsomes was 0.40 nmol/mg/min for the 3-glucuronide of estradiol and 0.32

117 nmol/mg/min for the 17β-glucuronide of estradiol . The Vmax values that we report for glucuronidation of estradiol in human liver microsomes are about 150-fold higher for G2

146

and 500 fold higher for G1. The differences in Vmax are attributed to the higher amounts

of substrate and microsomal protein used in our experiments along with the differences in

the analytical method used for quantification of metabolite concentrations, higher

amounts of cofactors (Mg+2, alamethicin, saccharolactone) used in our experiments. But, the calculated Clint values for 3-glucuronide of estradiol by was 23.5 ml/min/mcg and for

17-glucuronide of estradiol was 53.5 ml/min/mcg which are 20,000-50,000 fold higher than that reported by us for 2ME2. Thus, 2ME2 has a very low Clint as compared to

estradiol.

Hepatic phase I metabolism also plays a major role in the metabolism of 2ME2

resulting in the production of 2ME1, 2OHE2, 2OHE1 and 2, 3- dimethoxy-17β estradiol

and 2, 3-dimethoxyestrone. Much like other estrogens, the major CYP isoenzymes

responsible for the metabolism of 2ME2 are CYP 1A1, 1A2, 3A4, 3A5 and 2E1. The 2-

hydroxylation of estradiol and estrone, catalyzed by the CYP1A1/2 and CYP3A isozymes

7-9, is the major metabolic pathway in the liver as compared to the 4-hydroxylation, which

is exclusively catalyzed by CYP3A 10. The 2-hydroxyl metabolites of estradiol and estrone are further metabolized by a catechol-o-methyl transferase (COMT), which converts these products to 2-methoxyestradiol (2ME2) and 2-methoxyestrone, respectively. The reported apparent Km values for E2 hydroxylation catalyzed by the

119 human cytochrome P450 isoenzymes range from 20-156 µM . All Km values that we report for phase I reactions fall within the range of previously reported values for estradiol hydroxylation.

Sulfation of steroids is an important biotransformation step involved in the regulation of tissue specific estrogen levels, majority of which is believed to take place in the

147

gastrointestinal tract. Sulfation is known to be a high affinity, low capacity type reaction.

SULT 1E1 catalyzes the sulfation of 17β-estradiol. Unlike other estrogens, 2ME2 was

not metabolized through sulfation in the in vitro system, which was further corroborated

by the results obtained in human urine after administration of 2ME2 orally.

- Sulfotransferase enzymes catalyze the transfer of SO3 from 3’-phosphoadenosine-5’- phosphosulfate, the enzymatic cofactor, to phenolic acceptor groups on estrogens. 17β-

Estradiol and estrone are sulfated at the 3-position to 17β-estradiol 3-sulfate and estrone

3-sulfate 120. Estrogens are sulfated predominantly at the 3-position and not at 17β

position which is adjacent to a methyl group; unlike glucuronidation which can occur

either at the 3 or 17β hydroxyl group of steroid hormones 21. 2ME2 has a methoxy group

at C-2 position, which might be responsible for producing steric hindrance large enough

to hamper sulfation. The half-life of sulfates is much longer in the plasma as compared

to glucuronides, and a larger proportion of sulfates are excreted into the bile as compared

to urine, which might have been responsible for undetectable sulfate conjugates in vivo in

the urine of three patients.

The CLh was calculated as 862 mL/min, which is more than half of the liver blood flow. The liver plasma flow is about 750 mL/min, since blood consists of essentially

50% plasma. The actual CLh may approximate the hepatic blood flow making 2ME2 a high extraction ratio drug, thus partly explaining the poor bioavailability of 2ME2. The calculated CLh is believed to be an underestimation of the actual CLh of 2ME2, because our calculation does not account for biliary excretion of 2ME2 in feces and the glucuronidation of phase I metabolites of 2ME2, which might be significant considering that there are at least four major metabolites produced by hepatic phase I metabolism.

148

Thus, it is concluded from this study that the major route of metabolism of 2ME2 is

through glucuronidation and excretion in urine and feces. Glucuronidation of orally

administered 2ME2 most likely occurs during the gastrointestinal/hepatic first pass and

during enterohepatic recirculation, if it exists. The high first pass effect for 2ME2 may be one of the reasons responsible for the very low circulating concentrations of 2ME2 observed in patients (chapter 4). The glucuronides of 2ME2 may also be responsible for the long half life of 2ME2 in plasma. The glucuronides are excreted in the bile thus subjecting 2ME2 to the possibility of enterohepatic recirculation (as seen with other estrogens), and also to ‘cycling’ of estrogens as described in chapter 1, thus prolonging its half life. The patient with ovarian cancer who showed a partial response on treatment with 2ME2 had very low circulating concentrations of 2ME2. One of the reasons proposed for this observation is that an active metabolite of 2ME2 may be responsible for its antitumor activity. Further studies need to be done to identify the antitumor activity of the metabolites of 2ME2. Sulfation does not seem to play a major role in the renal excretion of 2ME2, thus disproving the hypothesis #4 as listed in chapter 2. The metabolic fate of orally administered 2ME2 was established and thus the objective #5 as listed in chapter 2 was achieved.

CHAPTER SIX

Plasma Protein Binding studies of 2-Methoxyestradiol in Human Plasma

Introduction

The pharmacokinetic evaluation of 2ME2 given orally to cancer patients has shown a

considerably long terminal half-life of 2ME2 in plasma (approximately 1-2 days), in spite of having low circulating plasma concentration (chapter 4). The basis for this long half- life in humans may possibly be related to enterohepatic recirculation processes.

However, a variety of other factors may influence the prolonged circulation of 2ME2 in

humans, including binding of the compound to plasma proteins. Indeed, drugs with high

affinity for plasma proteins often demonstrate a relatively slow distribution and

elimination of drug from the central compartment, which may prolong the apparent half-

life. Estrogens have been known to bind primarily to albumin and sex hormone binding

globulin (SHBG). Dunn et al72 have established that in human females 37 % of estradiol

is SHBG bound, 61% is albumin bound and 1.8% is free while in human males 78 % of

estradiol is albumin bound, 20% is SHBG bound and 2.3% is free.

The purpose of this study was to characterize the binding properties of 2ME2 to

human plasma and individual proteins using a novel microequilibrium dialysis method

that might help better explain the pharmacokinetics of this compound.

Materials and methods

Chemicals and reagents 150

2-methoxyestradiol (chromatographic purity, 99.99%) and [3H]-2ME2 (specific

activity, 5 Ci/mmol) were kindly supplied by Entremed Inc. (Rockville, MD, USA).

HPLC-grade methanol, ethanol and acetonitrile were obtained from J. T. Baker

(Phillipsburg, NJ, USA). De-ionized water was generated with a Hydro-Reverse Osmosis

system (Durham, NC, USA) connected to a Milli-Q UV Plus purifying system

(Marlborough, MA, USA). Bio-Safe II scintillation fluid was obtained from Research

Products International (Mount Prospect, IL, USA). Albumin, α1-acid glycoprotein

(AAG), sex-hormone binding globulin (SHBG) and 2-methoxyestrone (2ME1) were

obtained from Sigma-Aldrich Inc, USA. Other chemicals were of reagent grade or better.

Pure protein solutions at respective physiological concentrations were prepared in 0.01 M

phosphate buffer (pH 7.4). The stock solutions of all test substances were made in

ethanol. Drug-free human plasma was obtained pre-dose samples of cancer patients

receiving 2ME2, and the plasma fraction was separated by centrifugation (3000 × g for 5

min at 37 °C), and frozen within 1 hour after collection. Drug-free heparinized human

plasma from healthy volunteers was obtained from the National Institutes of Health

Clinical Center Blood Bank (Bethesda, MD, USA).

Development and validation of dialysis method

Equilibrium dialysis was performed on a plate rotator (Model # 74-2334, Harvard

Apparatus, Holliston, MA, USA) at 37°C in a humidified atmosphere of 5% CO2 using

96-wells micro-dialysis plates (Harvard Apparatus) 121. The dialysis compartments in

each well are separated by a regenerated cellulose membrane with a 5-kDa cut-off.

Experiments were carried out with 250-µl aliquots of plasma containing a tracer amount

151

of [3H]-2ME2 (5µL of 0.02 mCi/mL) against an equal volume of 0.01 M phosphate buffer (pH 7.4). Drug concentrations in 125 µl-aliquots of both compartments were

measured by liquid scintillation counting (LSC) for 1 minute following the addition of

Bio-Safe II scintillation fluid on a Model LS6000IC counter (Beckman Instruments, Inc.,

Columbia, MD).

Method Validation

Precision

The precision of the assay was assessed by calculation of the between-run and within-run precision. The between-groups mean square (MSbet), the within-groups mean square (MSwit), and the grand mean (GM) of the observed concentrations across run days

were obtained by one-way analysis of variance, using the run day as classification

variable. The between-run precision (BRP) was defined as:

(MS − MS ) / n BRP = bet wit ×100% GM

where n represents the number of replicates within each validation run. The within-run

precision (WRP) was calculated as:

MS WRP = wit ×100% GM

The precision was assessed by analyzing quadruplicate samples prepared from 4 different plasma sources in quadruplicate on 4 separate occasions. Within- and between- assay precision estimates were obtained by one-way analysis of variance, and reported as relative standard deviation.

152

Specificity

To evaluate the specificity of this procedure and check for displacement effect of

2-methoxyestrone (2ME1) on protein binding of 2ME2, blank human plasma was spiked

with 10 ng/mL, 100 ng/mL, 500 ng/mL, 800 ng/mL, 1 µg/mL, 10 µg/mL, 40 µg/mL and

analyzed in quadruplicate for changes in the fraction unbound drug (fu).

Stability and Recovery

The stability of [3H]-2ME2 at 370C was assessed in human plasma and 0.01 M phosphate buffer (pH 7.4) by comparing in quadruplicates the liquid scintillation counts obtained after 24 hours to freshly prepared samples. The stability of 2ME2 in plasma at

370C was analyzed by comparing peak area in quadruplicates of 1ng/mL (LLOQ) of

2ME2 in plasma kept for 24 hours using an LC/MS/MS method 122 to freshly prepared samples. Freeze-thaw stability of [3H]-2ME2 was established by comparing in

quadruplicate, the liquid scintillation counts obtained by plasma samples spiked with

[3H]-2ME2 after one freeze-thaw cycle, with freshly prepared plasma samples with equal amount of [3H]-2ME2. Freeze-thaw stability of 2ME2 in plasma has been established

elsewhere 122. Samples were considered stable if the difference between peak areas (for

2ME2 in plasma) and liquid scintillation counts (for [3H]-2ME2 in Plasma or buffer) was within ±20%.

In order to calculate the recovery, two groups of samples (group A and B) were prepared in replicates of eight. In group A, plasma and 0.01 M phosphate buffer (pH 7.4) from both sides of the equilibrium dialysis well (~500 µL), were completely emptied into

153 the scintillation counting fluid. In group B, were freshly prepared samples with 250 µL of PBS and 250 µL of plasma with 5 µL of [3H]-2ME2 (0.02 mCi/mL). Recovery was calculated using the following formula:

Scintillation Count Group B % Recovery = ×100 Scintillation Count Group A

Time to Equilibrium and Saturation of PPB in healthy human plasma

The time course of equilibrium was assessed in quadruplicate at 0.5, 1, 1.5, 2, 3, 4, 5,

6, 18, 22, 24 and 30 and 48 hours after start of the experiment. The determination of fu was also performed in plasma samples from healthy human volunteers over the anticipated clinically relevant concentration range of 2ME2 (0, 1, 5, 10, 15, 20, 30 ng/mL) and at concentrations higher than that at 40, 100 and 400 ng/mL to test saturability of plasma protein binding.

In vitro binding to purified human plasma proteins

The binding of 2ME2 to purified human proteins was studied by testing various protein concentrations in triplicate in 0.01 M phosphate buffer (pH 7.4). Albumin binding was studied within the concentration range of 1-5 g/dl (physiological range: 3.5-

4.5 g/dL), AAG binding was studied within the concentration range of 0.01-0.25 g/dL

(physiological range: 0.1-0.25 g/dL) and SHBG binding was studied within the concentration range of 10-200 nM (physiological range: 13-130 nM).

154

Ex-vivo plasma protein binding experiments

Patients and treatment

Plasma samples from 5 patients, who received single oral doses of 2ME2 orally, were used to determine the in vivo unbound fraction of 2ME2 in cancer patients. From each patient, serial plasma samples were obtained during the first course of treatment at the following time points: (i) immediately before drug administration (pre-dose), and (ii) at

0.5, 1.5, 2, 3, 4, 5, 6, 9, 12, 24, 38 and 50 hours after the first drug administration. All blood samples were immediately placed in an ice-water bath, centrifuged within 30 minutes of collection at 1000 × g for 10 min at 4°C, and were stored at or below –70 °C until analysis.

Measurement of total drug concentrations

Total 2ME2 concentrations were determined using a validated analytical method based on liquid chromatography coupled tandem mass spectrometric detection as described in chapter 2 122.

155

Measurement of unbound drug concentrations

The fraction unbound (fu) 2ME2 in each individual patient plasma sample was determined using equilibrium dialysis, and samples were analyzed for total radioactivity

(i.e., [3H]-2ME2) by liquid-scintillation counting as described above. The unbound drug concentrations (Cu) were calculated from the fraction unbound drug (fu) and the total drug concentration in plasma (Cp) (i.e., the total of unbound and protein bound), as Cu = fu ×

Cp.

Pharmacokinetic analysis

Estimates of pharmacokinetic parameters for total and unbound 2ME2 in plasma were

derived from individual concentration-time data sets by noncompartmental analysis using

the software package WinNonlin v4.0 (Pharsight Corporation, Mountain View, CA,

USA). The area under the plasma concentration-time curve (AUC) was calculated using

the linear trapezoidal method from time zero to the time of the final quantifiable

concentration (AUC[tf]). The AUC was extrapolated to infinity by dividing the last

measured concentration by the rate constant of the terminal phase (k), determined by log-

linear regression analysis. The apparent oral clearance (CL/F) was calculated by dividing

the administered dose by the observed AUC [inf], and the terminal half-life was

calculated as ln2/ k.

156

Statistical considerations

All binding experiments were performed at least in triplicate on at least 3 separate

occasions, and statistical analyses were carried out using NCSS v2001 (J. L. Hintze,

Kaysville, UT, USA). The effects of 2ME2 concentration, protein concentrations on drug

binding were estimated by a one-way ANOVA. All data are presented as mean values ±

standard deviation (SD), unless stated otherwise, and for all tests the a priori cutoff for

statistical significance was taken at p-value < 0.05.

Results

Validation of dialysis method for 2ME2

Preliminary experiments revealed that equilibrium was attained approximately 24

hours (Fig 50). The data were modeled using Graphpad Prism version 2.0. The mean

coefficient of variation of all sample values was less than 20%, assuring high

discriminatory power in the detection of changes in fu for 2ME2 in patient samples. With

the final method, the within-run and between-run variability were 7.1% and 8.5%

respectively. 2ME1 (10 ng/mL - 10µg/mL) did not change the fraction unbound for

2ME2 (p value > 0.05). [3H]-2ME2 was found to be stable in plasma and PBS at 370 C

for 24 hours. [3H]-2ME2 was found to be stable in plasma after one freeze thaw cycle.

The % recovery for [3H]-2ME2 was found to be 92%. Hence, we can conclude that there is negligible amount of nonspecific binding of 2ME2.

157

Figure 50: Time course to reach equilibrium for the fraction unbound 2ME2. Data are presented as means of individual observations (symbols) and a predicted model fit according to a modified Hill function (R2 = 0.85) with standard deviations (error bars).

Time to Equilibrium

4 ) %

d ( 3 un bo n 2 U n o i t 1 ac r F

0 0 10 20 30 40 50 60 Time (h)

In vitro binding interactions with 2ME2

2ME2 was found to bind highly to human plasma (mean, 97.4 ± 0.2 %), with a free drug fraction of (2.7 ± 0.2 %). There was no significant source difference in fu when plasma was used from different healthy individuals (mean fu, 2.68; p-value > 0.05; one- way ANOVA). The fu obtained in previously frozen plasma from healthy volunteers was found to be slightly higher than that observed in the plasma from six cancer patient (mean fu, 1.9 % versus 2.6%; p value = 0.045).

At clinically relevant concentrations of 2ME2 (1 to 30 ng/mL), the binding was concentration-independent (P > 0.05), indicating a low-affinity, possibly non-specific and non-saturable process. Figure 51 depicts the relationship between bound and unbound concentrations of 2ME2.

158

Figure 51: Binding of 2ME2 to human plasma

400 ) M

n 300 ( E2 200 und 2M

o 100 B

0 0.0 2.5 5.0 7.5 10.0 Unbound 2ME2 (nM)

There is a linear relationship between bound and unbound 2ME2 concentrations

(R2 = 0.99). The binding characteristics of 2ME2 in human plasma were further examined by plotting the ratio of Cb/Cu versus Cb in plasma.

159

Figure 52: Plot of 2ME2 Cb/Cu versus Cb in plasma from healthy human volunteers 75 u

/C 50 b E2 C

2M 25

0 0 100 200 300 400 2ME2 C (nM) b

Above figure 52 shows the plot of the ratio of bound and unbound 2ME2 concentrations versus bound 2ME2 concentration. The slope of linear regression in the above not significantly different from zero (p-value = 0.57). Hence, the binding of 2ME2 is non-saturable in the therapeutically achievable range.

2ME2 binding to albumin (1 – 5 g/dL), α1-acid glycoprotein (0.1-0.25 mg/dL), and sex-hormone binding globulin (1-20 µg/mL) was found to be protein concentration- dependent, i.e., fu decreased with increase in protein concentration (p-value < 0.05; one- way ANOVA). Table 23 provides a detailed synopsis of the effect of concentration of different proteins on fu (%) of 2ME2. Figure 53 provides a histogram comparing the binding of 2ME2 with HSA (53A), AAG (53B) and SHBG (53C) at different concentrations.

160

Table 23 Binding of 2ME2 to purified human proteins

Protein Concentration Mean fu (%) N

Albumin 50 mg/mL 3.1 3

Albumin 40 mg/mL 3.8 3

Albumin 25 mg/mL 5.3 3

Albumin 10 mg/mL 10.5 3 ______

AAG 2.5 mg/mL 27.6 3

AAG 1 mg/mL 48.8 3

AAG 0.5 mg/mL 63.4 3

AAG 0.1 mg/mL 81.5 3 ______

SHBG 20 µg/mL 73.3 2

SHBG 5 µg/mL 85.5 3

SHBG 1 µg/mL 92.6 3

*Abbreviations: AAG, α-1 acid glycoprotein; SHBG, sex-hormone binding globulin; fu, fraction unbound ; N, number of replicate observations

161

Fig 53: Binding of 2ME2 to human plasma proteins. Data are presented as mean values (bars) ± SD (error bars).

Figure 53 (A): Binding of 2ME2 with human serum albumin (1-5 g/dL)

100 d n u

o 75 n b ) o i

(% 50 act r f 2

E 25 2M

0 50.0 40.0 25.0 10.0 Albumin concentration (mg/mL)

(B) Binding of 2ME2 with human serum α1-acid glycoprotein (0.01-0.25 g/dL)

75 d n u o 50 n b ) o i (% act r f

2 25 E 2M

0 2.51.00.50.1 AAG concentration (mg/mL)

162

30 d n u o 20 n b ) o i (% act r f

2 10 E 2M

0 20.0 5.0 1.0 SHBG concentration (µg/mL)

Estimation of binding parameters

Estimation of binding parameters to individual proteins was performed using the data

obtained from buffer solutions of purified human proteins. The drug concentration ratio

in the buffer protein solution after dialysis was calculated for each paired observation,

and was taken as an estimate of the unbound drug fraction (fu). The bound drug fraction

(fbd) was calculated as fbd = (1 – fu). The total binding constants (nK) for the drug interactions with isolated proteins (HSA, AAG and SHBG) was calculated using the following equation, assuming a single class of nonsaturable binding sites and no allosteric effects:

Cb = nK ⋅ P ⋅Cu

1 fu = 1+ nK ⋅ P

where P is the protein concentration, (Cu) is unbound drug concentration, (Cb) is bound

163 drug concentration, nK (product of number of binding sites by affinity constant) denotes the total binding constant. Here nK*P is the slope of the regression line, Cbd, Cu and P are expressed as molar concentrations (M-1).

1 nKP fb = 1-fu = 1− = 1+ nK ⋅ P 1+ nK ⋅ P

1 fu / fb = nK ⋅ P nK*P = fu/fb

Thus, nK = fu/(fb*P)

Here, n represents the number of saturable binding sites per mole of protein, P the molar concentration of protein binding, K the association constant, and nK the contribution constant of nonspecific, non-saturable binding on one site (per molar concentration of protein).

The association constants for binding were 4.29 × 104 (± 4.2× 103) M-1, 4.41 × 104

(1.39× 103) M-1 and 1.5 × 106 (2.6 × 105) M-1 for albumin, AAG and SHBG respectively.

The significance of these binding coefficients are detailed in the ‘Discussion’ section of this chapter.

Displacement interactions on binding sites

2ME1 concentrations measured in the plasma are 100-200 fold higher than 2ME2 concentrations after administration of oral 2ME2 123. Hence, we examined the effect of

2ME1 on the binding of 2ME2. There was no change observed in fu of 2ME2 in the presence of 2ME1 (10 ng/mL - 10 µg/mL). (p-value = 0.78, t-test).

164

Pharmacokinetics of unbound 2ME2

The equilibrium dialysis method was next applied ex-vivo to define the concentration- time profiles of total and unbound 2ME2 in 5 patients with cancer that received single oral doses of 2ME2. One patient received 2ME2 at dose of 800 mg, two patients received 2ME2 at 1600 mg and two patients received 2ME2 at 2200 mg each. Total

2ME2 levels detected in these patients were very low (range: 1.1 ng/mL – 18.6 ng/mL).

The fu was calculated for all five patients at all the pharmacokinetic time points. Ex vivo, the fu was constant at all the pharmacokinetic time points for all five patients with cancer

(Figure 54). The combination of very low/undetectable concentrations of 2ME2 and tremendous pharmacokinetic variability in the measured concentrations for 2ME2 prevents us from presenting a mean plasma concentration-time profile for all five patients. Figure 55 presents a combined concentration-time profile of total and unbound

2ME2 as measured in one patient who received a single oral dose of 1600 mg 2ME2.

Table 22 provides a comparison of the pharmacokinetic parameters obtained after modeling the bound and total 2ME2 concentrations using WINNONLIN.

165

Figure 54: The mean (symbol) ± SD (error bars) fraction unbound 2ME2 in plasma versus time profiles. Data were obtained from five patients with cancer given a single oral dose of 2ME2.

8 ) % (

d 7 n

bou 6 n

5 on U i t c a 4 Fr

E2 3 2M

0 0 10 20 30 40 50 60

Time (h)

166

Figure 55: Concentration-time profiles of total 2ME2 (ng/mL) and unbound 2ME2 (ng/mL) in plasma (panel A; ‘▲’ represent unbound 2ME2 concentrations and; ‘■’ represent total 2ME2 concentrations; note logarithmic scale on Y axis and points connected using straight lines)

100 ) L

m 10 ng/ on ( i t a r 1 ent onc

E2 C 0.1 M 2

0.01 0 102030405060 Time (h)

167

Table 24: Comparison of 2ME2 bound versus unbound pharmacokinetic parameters*

Parameter Unbound 2ME2 Total 2ME2

Cmax (ng/mL) 0.2 18.6

Tmax (h) 0.5 0.5

AUC (ng·h/mL) 6.0 704.23

CL/F (L/h/m2) 213 2272000

Vd/F (L) 5309 90554210

T1/2 (h) 17.24 27.63

R2 of fit 0.99 0.80

* Data were obtained from 1 patient who received 2ME2 orally at a dose of 1600 mg

Abbreviations: Cmax, peak plasma concentration; Tmax, time to peak concentration; AUC, area under the plasma concentration-time curve; CL/F, apparent oral clearance; T1/2, half-life of the terminal phase; Cu, unbound drug concentration; Cp, total drug concentration in plasma.

Thus, from the above data it appears that unbound 2ME2 exhibits a shorter half-life in plasma as compared to total 2ME2. It is premature to make this conclusion based on the above data, because of a small sample size. Further, studies are warranted to determine whether plasma protein binding enhances the plasma half-life of 2ME2.

168

Discussion

In the present study we have described the in vitro and ex vivo plasma protein binding of 2ME2. The binding of 2ME2 to human plasma was approximately 98 % and independent of drug concentration over the clinically relevant range, thus proving right the hypothesis #3 of this thesis as listed in chapter 2. When binding studies were extended to individual proteins, it was found that 2ME2 binds extensively to human albumin, AAG and SHBG. The association constant for binding was 4.29 × 104 (± 4.2×

103) M-1, 4.41 × 104 (1.39× 103) M-1 and 1.5 × 106 (2.6 × 105) M-1 for albumin, AAG and

SHBG respectively. Thus, 2ME2 shows low affinity, high capacity type binding to albumin and AAG and it shows high affinity, low capacity type binding characteristics with SHBG. It has previously been reported that 2ME2 and 2ME1 bind very strongly to

SHBG 124. This strong binding to SHBG has been attributed to the addition of the methoxy group at the C-2 position. The previously reported association constant for

2ME2 binding to SHBG at 370 C was 15.6 × 108 M-1 125. The differences between the values of our and previously reported association constants of 2ME2 for SHBG binding is attributed to the differences in analytical methods, where we utilized direct measurement of radiolabeled 2ME2 whereas the previously reported method used competitive displacement with testosterone, as well as differences in the source of protein. Both the reported values fall within the range of high affinity binding. The association constant for estradiol to SHBG at 370 C is 0.29 × 109 M-1126. Thus, 2ME2 appears to bind less strongly to SHBG than estradiol. The previously reported values for the affinity constants of estrone sulfates and ethinyl estradiol for albumin range between

169

1.8 × 104 M-1 to 0.5 × 106 M-1 to 127. Our reported values for 2ME2 association constant for albumin falls within this range previously reported for other steroids.

There was a slight but statistically significant increase in the fraction unbound 2ME2 observed in plasma obtained from healthy volunteers compared to plasma from cancer patients. This type of effect of disease state on drug binding has been described previously for various agents, including fluconazole 128, methadone 129 and propranolol

129 due to increase in plasma AAG levels and a concomitant decline in plasma albumin.

Although there was very low inter-individual variability in the fraction unbound

2ME2 in the plasma of cancer patients at the 2ME2 dose levels tested, there was very high inter-individual variability in the total 2ME2 concentrations measured. Consistent with the in vitro data, almost 98% of drug was bound within the circulation without any trend over time. Although, there was no change in the fraction unbound drug over the therapeutically achieved concentrations of 2ME2, there was statistically significant difference in fu due to variability in albumin and AAG concentrations in cancer patients.

But since there was no change in total unbound fraction, a drastic effect of changes in protein binding on the pharmacological response to 2ME2 may be unlikely. The high binding to plasma proteins of 2ME2 along with enterohepatic recirculation and the metabolic ‘cycling’ of 2ME2 may be a significant contributing factor to its long half life in the plasma.

In conclusion, a reliable and reproducible equilibrium dialysis method for the determination of the fraction unbound 2ME2 in plasma was developed and validated, thus achieving objective #4 of this thesis as listed in chapter 2. The extent of binding of

2ME2 to albumin and AAG was higher as compared to estradiol, while the extent of

170 binding to SHBG was comparable to estradiol as reported previously 72. 2ME2 was found to bind with a high degree of affinity to several human plasma proteins, including albumin, AAG and SHBG. This clearly signifies the importance to account for differences in the fraction unbound drug when attempting to extrapolate data obtained in in vitro model systems in protein-free media to the clinical situation.

CHAPTER SEVEN

Summary and Conclusions

This chapter will try to put together all that had been learned in the previous chapters and present a complete picture of the clinical pharmacology of 2ME2. Moreover, we will try to explore the reasons for poor bioavailability and low circulating concentrations of

2ME2 observed after oral dosing even at very high doses. The major barriers to oral drug delivery and systemic exposure are formulation, solubility, permeability, transporter effects, first pass metabolism. We will discuss each of these factors in this chapter. The probable future of drug development of 2ME2 will also be discussed.

Formulation

2ME2 was formulated as 200 mg capsules with lactose, sodium starch glycolate, colloidal silicon dioxide and magnesium stearate. The number of capsules ingested by a patient as a single dose ranged from two (400 mg q 12 h/DL1) to fifteen (3000 mg q 12 h/DL5). Since the number of capsules being ingested is quite high, there has been a speculation that the diarrhea resulting in patients on 2ME2 may be because of lactose intolerance. In conjunction, the resulting diarrhea can potentially hamper the absorption of the drug. We had eight episodes of diarrhea (grade 1 or 2) on our clinical trial. Thus, the formulation of 2ME2 as a lactose based capsule was not the most optimal but was not likely the most important factor hampering bioavailability. 172

Solubility

2ME2 is a poorly soluble compound. Following table 25 and figure 56 represents the solubility profile of 2ME2 from pH 2 to pH 11 130.

Table 25: Solubility profile of 2ME2

Buffer pH Solubility (µg/mL) HCl 2 1.9 Citrate 3 3 Citrate 4 2.8 Citrate 5 3 Phosphate 6 2.3 Phosphate 7 3.3 Phosphate 8 2.8 Borate 9 3.4 Carbonate 10 4.8 Diamine-Glycine 11 46.25

173

Figure 56: Solubility profile of 2ME2

Solubility Profile of 2ME2

30 (ug/mL) 20

10 Solubility

024681012

2ME2 possesses poor aqueous solubility spanning the complete pH profile of the gastrointestinal tract. In addition, the solubility of 2ME2 in water is < 0.6 µg/mL 130.

Thus, 2ME2 when administered along with a 250 mL glass of water will dissolve at the most 150 µg of the drug in water 114, not taking the pH into consideration. Taking the pH into consideration the solubility of 2ME2 in 250 mL fluid the amount of 2ME2 dissolved will range from 475 µg (pH 1.9) to 700 µg (pH 8). Thus, 2ME2 can be classified as a poorly soluble compound. Solubility is hence believed to be one of the major factors affecting oral bioavailability of 2ME2.

174

Permeability

The apparent permeability (Papp) of 2ME2 using monolayers of intestinal Caco-2 cells cultured in Transwell inserts 130, measuring one-way transport from the apical to the basolateral side, in comparison with propranolol is depicted in the Table 26 below:

Table 26: Permeability of 2ME2 across Caco-2 cell monolayer

-6 Compound Mean Papp (X 10 cm/s)

14C-Mannitol 0.89 (0.10)

3H-Propranolol 13.9 (3.6)

2ME2 28.8 (4.9)

Comparing the Papp values of 2ME2 and propranolol, which is a marker for high permeability, we can conclude that 2ME2 is a highly permeable drug. Thus, 2ME2 falls within the class 2 compounds in the Biopharmaceutics Classification System (BCS) 114,115 as defined by the FDA. A high fraction of 2ME2 that dissolves into gastrointestinal fluid will permeate from the apical to the basolateral side and get through to the portal circulation. Permeability of 2ME2 is thus not believed to be a major factor affecting the bioavailability.

Transporter Effects Since 2ME2 is a class 2 compound, high permeability will allow ready access into the gastrointestinal tract, and thus uptake transporters will have no effect on absorption. But,

175 the low solubility will limit the concentrations entering the enterocytes, thereby preventing the saturation of efflux transporters. Thus, efflux transporters can potentially play a major role in determining the extent of oral bioavailability (Fextent) and the rate of absorption. The effect of efflux transporters needs to be examined in future studies to further test this hypothesis. Although, the efflux transporter effect seems very unlikely since there is no increase in the Cmax or AUC with increase in dose as seen in chapter 4.

Food Effects (High-Fat Meals) Food effects result from changes in drug solubility, delay in gastric emptying, stimulation of bile flow, change in gastrointestinal pH, increase in splanchnic blood flow change in luminal metabolism of drug, and physical or chemical interaction with the dosage form 115. The solubility of 2ME2 will be increased with a high fat meal. High-fat meals are also suspected to inhibit both influx and efflux drug transporters 115. Since uptake transporters are not expected to play a major role in 2ME2 gastrointestinal absorption, the inhibition of efflux transporters could potentially increase the extent of oral bioavailability (Fextent). Hence, all the above-mentioned factors combined are expected to cause an increase in Fextent. The time to maximal concentration (tmax) may be delayed because of prolonged gastric emptying time.

176

First-pass metabolism First-pass metabolism of a drug starts with the metabolism in the gastrointestinal tract.

As seen in chapter 5, 2ME2 is highly glucuronidated and is a substrate for CYP 3A enzymes, which are present in the gastrointestinal tract. Due to poor solubility and low gastrointestinal concentration of 2ME2, the CYP 3A enzymes and UDP- glucuronosyltransferases in the gastrointestinal tract will not be saturated and the first pass metabolism of 2ME2 will start right from the gastrointestinal tract.

Sulfonyltransferases are another group of enzymes which are located predominantly in the gastrointestinal tract and also play a major role in the metabolism of steroid compounds as discussed in chapter 1. But, we were unable to establish sulfation of 2ME2 in vivo or in vitro. Thus, we can conclude that 2ME2 arguably undergoes phase I metabolism by CYP 3A enzymes and glucuronidation in the gastrointestinal tract. 2ME2 undergoes extensive metabolism in the liver as seen in chapter 5, glucuronidation being the predominant pathway for metabolism. The predicted in vivo plasma hepatic clearance

(CLh) of 2ME2 was 862 mL/min, which is more than half the blood flow of the liver.

This number is believed to be an underestimation of the actual CLh since this calculation did not take into consideration the glucuronidation of the phase I metabolites and the biliary excretion of the glucuronidated metabolites. Hence, the actual CLh is believed to approach and possibly exceed liver blood flow, thus making 2ME2 a high extraction ratio drug. It can be concluded that first pass effect is also a major player in determining the oral bioavailability of 2ME2.

177

Oral Bioavailability Estimates Only about 1% of the total 2ME2 dose administered orally was recovered from the urine as glucuronides after 24 hours. In the preclinical studies of excretion in mice

(chapter 1), approximately 50% of orally administered [3H]-2ME2 was recovered in the urine (~25%) and feces (~25%) combined after 24 hours (one half life, chapter 5).

Hence, if we assume that an equal amount of 2ME2 is being excreted in feces and urine, then we can make a rough estimate of 2ME2 oral bioavailability in humans as being about 2%. This observation is in corroboration with the estimation of 1-2% oral bioavailability in mice (chapter 1).

Plasma Protein Binding

2ME2 is heavily bound to plasma proteins (approximately 98%), thus decreasing the unbound fraction available for the biological activity. This complements the observation that there was no DLT and the MTD was not achieved. Also, only one patient showed a partial response out of twenty that were treated. The extremely high Vd/F as mentioned in chapter 4, is counterintuitive to the observation of extensive plasma protein binding.

But, the extremely high Vd/F is most probably a reflection of the poor bioavailability of the compound and not the distribution characteristics. 2ME2 was found to bind with a high degree of affinity to several human plasma proteins, including albumin, AAG and

SHBG. The extent of binding of 2ME2 to albumin and AAG was higher as compared to estradiol, while the extent of binding to SHBG was comparable to estradiol as reported previously 72. Thus, extensive plasma protein binding is also a major barrier to the systemic exposure of 2ME2.

178

Conclusion

2ME2 showed exciting preclinical activity in the form of in vitro cell lines studies and xenograft studies in terms of biological activity. But, the solubility and oral bioavailability data that were available from the preclinical studies were overlooked, and the development of this agent was hurried into the clinic at a premature stage.

Pharmacokinetic data from previous Phase I clinical trials of 2ME2 should have been obtained and studied before embarking upon the clinical trial at NCI with higher doses of

2ME2. From an ethical point of view, unfavorable pharmacokinetic results should prevent unnecessary exposure of patients to potential harmful and less effective cytotoxic drugs. 2ME2 does possess promising antitumor activity as demonstrated by the partial response in the ovarian cancer patient (patient #7) and corroborating data from other clinical trials, but the low circulating concentrations, tremendous interpatient variability and lack of dose proportionality seen in our patients demands a better formulation of

2ME2 to be developed. The fact that patient #7 showed had a partial response even in the presence of low circulation concentrations of 2ME2, leads us to believe that an active metabolite of 2ME2 may be responsible for its biological activity. Entremed has currently developed a new formulation of 2ME2 which is a Nanocrystal Colloidal

Dispersion (NCD), which achieves significantly higher concentrations as compared to the capsular formulation of 2ME2. However, it still remains to be seen, how much improvement in oral bioavailability this approach will achieve. Another approach that could theoretically be tried to improve the bioavailability of 2ME2 is to change the route of administration from oral to intravenous. But, this approach would run into problems as it would not be a practically feasible option for patients with cancer receiving long term

179

2ME2, since the role of 2ME2 will more than likely be in combination with more aggressive therapy to keep the cancer in ‘check’ or in a palliative setting.

Bibliography

1. Simpson ER, Clyne C, Speed C, et al: Tissue-Specific Estrogen

Biosynthesis and Metabolism. Ann NY Acad Sci 949:58-67, 2001

2. Nelson LR, Bulun SE: Estrogen production and action. J Am Acad

Dermatol 45:S116-24, 2001

3. Dubey RK, Gillespie DG, Zacharia LC, et al: Methoxyestradiols mediate the antimitogenic effects of estradiol on vascular smooth muscle cells via estrogen receptor-independent mechanisms. Biochem Biophys Res Commun 278:27-33, 2000

4. Simpson E, Rubin G, Clyne C, et al: Local estrogen biosynthesis in males and females. Endocr Relat Cancer 6:131-7, 1999

5. Grodin JM, Siiteri PK, MacDonald PC: Source of estrogen production in postmenopausal women. J Clin Endocrinol Metab 36:207-14, 1973

6. MacDonald PC, Madden JD, Brenner PF, et al: Origin of estrogen in normal men and in women with testicular feminization. J Clin Endocrinol Metab 49:905-

16, 1979

7. Shou M, Korzekwa KR, Brooks EN, et al: Role of human hepatic cytochrome P450 1A2 and 3A4 in the metabolic activation of estrone. Carcinogenesis

18:207-14, 1997 181

8. Aoyama T, Korzekwa K, Nagata K, et al: Estradiol metabolism by complementary deoxyribonucleic acid-expressed human cytochrome P450s.

Endocrinology 126:3101-6, 1990

9. Powell WA, Mitchell BF, Challis JR: Effects of steroids on progesterone output by explants of human chorion. Gynecol Obstet Invest 22:64-72, 1986

10. Kerlan V, Dreano Y, Bercovici JP, et al: Nature of cytochromes P450 involved in the 2-/4-hydroxylations of estradiol in human liver microsomes. Biochem

Pharmacol 44:1745-56, 1992

11. Martucci CP, Fishman J: P450 enzymes of estrogen metabolism. Pharmac

Ther Vol 57:237-257, 1993

12. Vadlamuri SV, Glover DD, Turner T, et al: Regiospecific expression of cytochrome P4501A1 and 1B1 in human uterine tissue. Cancer Lett 122:143-50, 1998

13. Weisz J: Biogenesis of catecholestrogens: Metabolic activation of estrogens by phase I enzymes. Polycyclic Aromat Compd 6:241-251, 1994

14. Liehr JG, Roy D: Free radical generation by redox cycling of estrogens.

Free Radic Biol Med 8:415-23, 1990

15. Liehr JG: Mechanisms of metabolic activation and inactivation of catecholestrogens: A basis of genotoxicity. Polycyclic Aromat Compd 6:229-239, 1994

16. Liehr JG, Ulubelen AA, Strobel HW: Cytochrome P-450-mediated redox cycling of estrogens. J Biol Chem 261:16865-70, 1986

17. Schneider J, Kinne D, Fracchia A, et al: Abnormal oxidative metabolism of estradiol in women with breast cancer. Proc Natl Acad Sci U S A 79:3047-51, 1982

182

18. Osborne MP, Karmali RA, Hershcopf RJ: Omega-3 fatty acids: modulation of estrogen metabolism and potential for breast cancer prevention. Cancer

Invest 8:629-631, 1988

19. Adlercreutz H, Fotsis T, Hockerstedt K, et al: Diet and urinary estrogen profile in premenopausal omnivorous and vegetarian women and in premenopausal women with breast cancer. J Steroid Biochem 34:527-30, 1989

20. Milewich L, Garcia RL, Gerrity LW: Steroid sulfatase and 17 beta- hydroxysteroid oxidoreductase activities in mouse tissues. J Steroid Biochem 21:529-38,

1984

21. Raftogianis R, Creveling C, Weinshilboum R, et al: Estrogen metabolism by conjugation. J Natl Cancer Inst Monogr:113-24, 2000

22. Spink BC, Katz BH, Hussain MM, et al: SULT1A1 catalyzes 2- methoxyestradiol sulfonation in MCF-7 breast cancer cells. Carcinogenesis 21:1947-

1957, 2000

23. Belanger A, Brochu M, Lacoste D, et al: Steroid flucuronides: human circulatory levels and formation by LNCaP cells. J Steroid Biochem 40:593-598, 1991

24. Senafi SB, Clarke DJ, Burchell B: Investigation of the substrate specificity of a cloned expressed human bilirubin UDP-glucuronosyltransferase: UDP-sugar specificity and involvement in steroid and xenobiotic glucuronidation. Biochem J 303 (

Pt 1):233-40, 1994

25. Mojarrabi B, Butler R, Mackenzie PI: cDNA cloning and characterization of the human UDP glucuronosyltransferase, UGT1A3. Biochem Biophys Res Commun

225:785-90, 1996

183

26. Cheng Z, Rios GR, King CD, et al: Glucuronidation of catechol estrogens by expressed human UDP-glucuronosyltransferases (UGTs) 1A1, 1A3, and 2B7. Toxicol

Sci 45:52-7, 1998

27. Zhu BT, Evaristus EN, Antoniak SK, et al: Metabolic deglucuronidation and demethylation of estrogen conjugates as a source of parent estrogens and catecholestrogen metabolites in Syrian hamster kidney, a target organ of estrogen- induced tumorigenesis. Toxicol Appl Pharmacol 136:186-93, 1996

28. Ball P, Knuppen R: Formation, metabolism, and physiologic importance of catecholestrogens. Am J Obstet Gynecol 163:2163-70, 1990

29. Schneider J, Huh MM, Bradlow HL, et al: Antiestrogen action of 2- hydroxyestrone on MCF-7 human breast cancer cells. J Biol Chem 259:4840-5, 1984

30. Guldberg HC, C.A. M: Catechol-o-methyltransferase: pharmacological aspects and physiological role. Pharmacol. Rev. 27:135-206, 1975

31. Cavalieri EL, Stack DE, Devanesan PD, et al: Molecular origin of cancer: catechol estrogen-3,4-quinones as endogenous tumor initiators. Proc Natl Acad Sci U S A

94:10937-42, 1997

32. Axelrod JT, R.: Enzymatic o-methylation of epinephrine and other catechols. J Biol Chem 233:702-705, 1958

33. Campbell NR, Dunnette JH, Mwaluko G, et al: Platelet phenol sulfotransferase and erythrocyte catechol-O-methyltransferase activities: correlation with methyldopa metabolism. Clin Pharmacol Ther 35:55-63, 1984

184

34. Lavigne JA, Goodman JE, Fonong T, et al: The effects of catechol-O- methyltransferase inhibition on estrogen metabolite and oxidative DNA damage levels in estradiol-treated MCF-7 cells. Cancer Res 61:7488-94, 2001

35. Lakhani NJ, Sarkar MA, Venitz J, et al: 2-methoxyestradiol, a promising anticancer agent. Pharmacotherapy 23:165-172, 2003

36. Shang W, Konidari I, Schomberg DW: 2-Methoxyestradiol, an endogenous estradiol metabolite, differentially inhibits granulosa and endothelial cell mitosis: a potential follicular antiangiogenic regulator. Biol Reprod 65:622-7, 2001

37. Berg D, Sonsalla R, Kuss E: Concentrations of 2-methoxyestrogens in human serum measured by a heterologous immunoassay with an 125I-labelled ligand.

Acta Endocrinol (Copenh) 103:282-8, 1983

38. Schumacher G, Kataoka M, Roth JA, et al: Potent antitumor activity of 2- methoxyestradiol in human pancreatic cancer cell lines. Clin Cancer Res 5:493-9, 1999

39. Fotsis T, Zhang Y, Pepper MS, et al: The endogenous oestrogen metabolite 2-methoxyoestradiol inhibits angiogenesis and suppresses tumour growth.

Nature 368:237-9, 1994

40. Klauber N, Parangi S, Flynn E, et al: Inhibition of angiogenesis and breast cancer in mice by the microtubule inhibitors 2-methoxyestradiol and taxol. Cancer Res

57:81-6, 1997

41. LaVallee TM, Zhan XH, Herbstritt CJ, et al: 2-Methoxyestradiol inhibits proliferation and induces apoptosis independently of estrogen receptors alpha and beta.

Cancer Res 62:3691-7, 2002

185

42. Cushman M, He HM, Katzenellenbogen JA, et al: Synthesis, antitubulin and antimitotic activity, and cytotoxicity of analogs of 2-methoxyestradiol, an endogenous mammalian metabolite of estradiol that inhibits tubulin polymerization by binding to the colchicine binding site. J Med Chem 38:2041-9, 1995

43. Mukhopadhyay T, Roth JA: Induction of apoptosis in human lung cancer cells after wild-type p53 activation by methoxyestradiol. Oncogene 14:379-84, 1997

44. Pribluda VS, Gubish ER, Jr., Lavallee TM, et al: 2-Methoxyestradiol: an endogenous antiangiogenic and antiproliferative drug candidate. Cancer Metastasis Rev

19:173-9, 2000

45. Atalla H, Makela TP, Adlercreutz H, et al: 2-methoxyestradiol arrests cells in mitosis without depolymerizing tubulin. Biochem Biophys Res Commun 228:467-473,

1996

46. Seegers JC, Lottering ML, Grobler CJ, et al: The mammalian metabolite,

2-methoxyestradiol, affects P53 levels and apoptosis induction in transformed cells but not in normal cells. J Steroid Biochem Mol Biol 62:253-67, 1997

47. Josefsson E, Tarkowski A: Suppression of type II collagen-induced arthritis by the endogenous estrogen metabolite 2-methoxyestradiol. Arthritis Rheum

40:154-63, 1997

48. Aizu-Yokota E, Susaki A, Sato Y: Natural estrogens induce modulation of microtubules in Chinese hamster V79 cells in culture. Cancer Res 55:1863-8, 1995

49. Spicer LJ, Hammond JM: Catecholestrogens inhibit proliferation and

DNA synthesis of porcine granulosa cells in vitro: comparison with estradiol, 5 alpha-

186 dihydrotestosterone, gonadotropins and catecholamines. Mol Cell Endocrinol 64:119-26,

1989

50. Nishigaki I, Sasaguri Y, Yagi K: Anti-proliferative effect of 2- methoxyestradiol on cultured smooth muscle cells from rabbit aorta. Atherosclerosis

113:167-70, 1995

51. Pico C, Puigserver P, Oliver P, et al: 2-Methoxyestradiol, an endogenous metabolite of 17beta-estradiol, inhibits adipocyte proliferation. Mol Cell Biochem 189:1-

7, 1998

52. Yue TL, Wang X, Louden CS, et al: 2-Methoxyestradiol, an endogenous estrogen metabolite, induces apoptosis in endothelial cells and inhibits angiogenesis: possible role for stress-activated protein kinase signaling pathway and Fas expression.

Mol Pharmacol 51:951-62, 1997

53. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1:27-31, 1995

54. Hanahan D, Folkman J: Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 86:353-64, 1996

55. Nicosia RF, Ottinetti A: Growth of microvessels in serum-free matrix culture of rat aorta. A quantitative assay of angiogenesis in vitro. Lab Invest 63:115-22,

1990

56. Bauer KS, Dixon SC, Figg WD: Inhibition of angiogenesis by thalidomide requires metabolic activation, which is species-dependent. Biochem Pharmacol 55:1827-

34, 1998

187

57. James VH, Reed MJ, Lai LC, et al: Regulation of estrogen concentrations in human breast tissues. Ann N Y Acad Sci 595:227-35, 1990

58. La Valee TM, Zhan XH, Herbstritt CJ, et al: 2-methoxyestradiol induces apoptosis through activation of the intrinsic pathway. Proceeding of the American

Association for Cancer Research 43:703, 2002

59. Kataoka M, Schumacher G, Cristiano RJ, et al: An agent that increases tumor suppressor transgene product coupled with systemic transgene delivery inhibits growth of metastatic lung cancer in vivo. Cancer Res 58:4761-5, 1998

60. Zacharia LC, Jackson EK, Gillespie DG, et al: Increased 2- methoxyestradiol production in human coronary versus aortic vascular cells.

Hypertension 37:658-62, 2001

61. Purohit A, Singh A, Ghilchik MW, et al: Inhibition of tumor necrosis factor alpha-stimulated aromatase activity by microtubule-stabilizing agents, paclitaxel and 2-methoxyestradiol. Biochem Biophys Res Commun 261:214-7, 1999

62. D'Amato RJ, Lin CM, Flynn E, et al: 2-Methoxyestradiol, an endogenous mammalian metabolite, inhibits tubulin polymerization by interacting at the colchicine site. Proc Natl Acad Sci U S A 91:3964-8, 1994

63. Huang P, Feng L, Oldham EA, et al: Superoxide dismutase as a target for the selective killing of cancer cells. Nature 407:390-5, 2000

64. Kachadourian R, Liochev SI, Cabelli DE, et al: 2-methoxyestradiol does not inhibit superoxide dismutase. Arch Biochem Biophys 392:349-53, 2001

65. Iafrati MD, Karas RH, Aronovitz M, et al: Estrogen inhibits the vascular injury response in estrogen receptor alpha-deficient mice. Nat Med 3:545-8, 1997

188

66. Karas RH, Hodgin JB, Kwoun M, et al: Estrogen inhibits the vascular injury response in estrogen receptor beta-deficient female mice. Proc Natl Acad Sci U S

A 96:15133-6, 1999

67. Barchiesi F, Jackson EK, Gillespie DG, et al: Methoxyestradiols mediate estradiol-induced antimitogenesis in human aortic SMCs. Hypertension 39:874-9, 2002

68. Entremed: Investigator's Brochure: 2-methoxyestradiol, initial investigational new drug application. Rockville, MD

69. Bhavnani BR, Nisker JA, Martin J, et al: Comparison of pharmacokinetics of a conjugated equine estrogen preparation (premarin) and a synthetic mixture of estrogens (C.E.S.) in postmenopausal women. J Soc Gynecol Investig 7:175-83, 2000

70. Gilman AG, Goodman LS, Rall TW: The pharmacological basis of therapeutics. Macmillan Publishing, New York, NY 7th ed, 1985

71. Takahashi K, Okada M, Ozaki T, et al: Safety and efficacy of oestriol for symptoms of natural or surgically induced menopause. Hum Reprod 15:1028-36, 2000

72. Dunn JF, Nisula BC, Rodbard D: Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid- binding globulin in human plasma. J Clin Endocrinol Metab 53:58-68, 1981

73. Stoa KF, Levitz M: Comparison of the conjugated metabolites of intravenously and intraduodenally administered oestriol. Acta Endocrinol (Copenh)

57:657-68, 1968

74. Miyazaki T, Kirdani RY, Slaunwhite WR, Jr., et al: Studies on phenolic steroids in human suvjects. XV. Biliary and urinary excretion patterns of estrone. J Clin

Endocrinol Metab 33:128-37, 1971

189

75. Miller KD, Haney LG, Pribluda VS, et al: A phase I safety, pharmacokinetic and pharmacodynamic study of 2-methoxyestradiol (2ME2) in patients with refractory metastatic breast cancer. Proceedings of The American Society of Clinical

Oncology 20:43a, 2001

76. Wilding G, Sweeney C, King DM, et al: Phase 2, multicenter, randomized, double blind, safety, pharmacokinetic, pharmacodynamic and efficacy study of two doses of 2-methoxyestradiol administered orally in patients with hormone refractory prostate cancer [abstr]. Proceedings of American Society of Clinical Oncology Presented at the

38th annual meeting of the American Society of Clinical Oncology, Orlando, May 18-21,

2002., 2003

77. Miller KD, Murry DJ, Curry E, et al: A phase I study of 2- methoxyestradiol (2ME2) plus docetaxel in patients with metastatic breast cancer.

Proceedings of The American Society of Clinical Oncology 21:111a, 2002

78. Robinson MR, Baffi J, Yuan P, et al: Safety and pharmacokinetics of intravitreal 2-methoxyestradiol implants in normal rabbit and pharmacodynamics in a rat model of choroidal neovascularization. Exp. Eye Res. 74:309-317, 2002

79. Zacharia LC, Dubey RK, Jackson EK: A gas chromatography/mass spectrometry assay to measure estradiol, catecholestradiols, and methoxyestradiols in plasma. Steroids 69:255-261, 2004

80. Lakhani NJ, Sparreboom A, Dahut WL, et al: Determination of the antiangiogenesis agent 2-methoxyestradiol in human plasma by liquid chromatography with ultraviolet detection. Journal of Chromatography B 806:289-293, 2004

190

81. Shah V, Midha K, Findlay J, et al: Bioanalytical method validation--a revisit with a decade of progress. Pharm Res 17:1551-7, 2000

82. Tan AR, Yang X, Hewitt SM, et al: Evaluation of biologic end points and pharmacokinetics in patients with metastatic breast cancer after treatment with erlotinib, an epidermal growth factor receptor tyrosine kinase inhibitor. J Clin Oncol 22:3080-90,

2004

83. Hollander M, Wolfe DA: Nonparametric Statistical Methods. John Wiley and Sons, Inc.:202-211, 1999

84. van Diest PJ, van der Wall E, Baak JPA: Prognostic value of proliferation in invasive breast cancer: a review. J Clin Pathol 57:675-681, 2004

85. Spyratos F, Ferrero-Pous M, Trassard M, et al: Correlation between MIB-

1 and other proliferation markers: clinical implications of the MIB-1 cutoff value. Cancer

94:2151-9, 2002

86. Weidner N, Moore DH, 2nd, Vartanian R: Correlation of Ki-67 antigen expression with mitotic figure index and tumor grade in breast carcinomas using the novel "paraffin"-reactive MIB1 antibody. Hum Pathol 25:337-42, 1994

87. Veronese SM, Gambacorta M, Gottardi O, et al: Proliferation index as a prognostic marker in breast cancer. Cancer 71:3926-31, 1993

88. Sahin AA, Ro J, Ro JY, et al: Ki-67 immunostaining in node-negative stage I/II breast carcinoma. Significant correlation with prognosis. Cancer 68:549-57,

1991

191

89. Berger R, Albelda SM, Berd D, et al: Expression of platelet-endothelial cell adhesion molecule-1 (PECAM-1) during melanoma-induced angiogenesis in vivo. J

Cutan Pathol 20:399-406, 1993

90. Tsuda H, Bandera CA, Birrer MJ, et al: Cyclin E amplification and overexpression in clear cell adenocarcinoma of the ovary. Oncology 67:291-9, 2004

91. Goff BA, Sainz de la Cuesta R, Muntz HG, et al: Clear cell carcinoma of the ovary: a distinct histologic type with poor prognosis and resistance to platinum-based chemotherapy in stage III disease. Gynecol Oncol 60:412-7, 1996

92. Behbakht K, Randall TC, Benjamin I, et al: Clinical characteristics of clear cell carcinoma of the ovary. Gynecol Oncol 70:255-8, 1998

93. Yamamoto S, Konishi I, Mandai M, et al: Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels. Br J Cancer

76:1221-7, 1997

94. Shweiki D, Itin A, Soffer D, et al: Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 359:843-5, 1992

95. Mabjeesh NJ, Escuin D, La Valee TM, et al: 2ME2 inhibits tumor growth and angiogenesis by disrupting microtubules and dysregulating HIF. Cancer Cell 3:363-

375, 2003

96. Lau K-M, Mok SC, Ho S-M: Expression of human estrogen receptor- alpha and -beta , progesterone receptor, and androgen receptor mRNA in normal and malignant ovarian epithelial cells. PNAS 96:5722-5727, 1999

192

97. Aghmesheh M, Edwards L, Clarke CL, et al: Expression of steroid hormone receptors in BRCA1-associated ovarian carcinomas. Gynecol Oncol 97:16-25,

2005

98. Schaner ME, Ross DT, Ciaravino G, et al: Gene Expression Patterns in

Ovarian Carcinomas. Mol. Biol. Cell 14:4376-4386, 2003

99. Lee P, Rosen DG, Zhu C, et al: Expression of progesterone receptor is a favorable prognostic marker in ovarian cancer. Gynecol Oncol 96:671-7, 2005

100. Sutherland TE, Schuliga M, Harris T, et al: 2-Methoxyestradiol Is an

Estrogen Receptor Agonist That Supports Tumor Growth in Murine Xenograft Models of

Breast Cancer. Clin Cancer Res 11:1722-1732, 2005

101. Scanlon PD, Raymond FA, Weinshilboum RM: Catechol-o- methyltransferase: thermolabile enzyme in erythrocytes of subjects homozygous for allele for low activity. Science 5:63-65, 1979

102. Worda C, Sator MO, Schneeberger C, et al: Influence of the catechol-O- methyltransferase (COMT) codon 158 polymorphism on estrogen levels in women. Hum

Reprod 18:262-6, 2003

103. DeMario MD, Ratain MJ: Oral chemotherapy: rationale and future directions. J Clin Oncol 16:2557-67, 1998

104. Hellriegel ET, Bjornsson TD, Hauck WW: Interpatient variability in bioavailability is related to the extent of absorption: implications for bioavailability and bioequivalence studies. Clin Pharmacol Ther 60:601-7, 1996

193

105. Biing JT, Yang YF, Liu HS, et al: The induction of multidrug resistance in human cervical carcinoma cell lines by estrogenic hormones. Proc Natl Sci Counc Repub

China B 18:64-70, 1994

106. Fujise H, Annoura T, Sasawatari S, et al: Transepithelial transport and cellular accumulation of steroid hormones and polychlorobiphenyl in porcine kidney cells expressed with human P-glycoprotein. Chemosphere 46:1505-11, 2002

107. Maliepaard M, Scheffer GL, Faneyte IF, et al: Subcellular Localization and Distribution of the Breast Cancer Resistance Protein Transporter in Normal Human

Tissues. Cancer Res 61:3458-3464, 2001

108. Kruijtzer CMF, Beijnen JH, Rosing H, et al: Increased Oral

Bioavailability of Topotecan in Combination With the Breast Cancer Resistance Protein and P-Glycoprotein Inhibitor GF120918. J Clin Oncol 20:2943-2950, 2002

109. Sugimoto Y, Tsukahara S, Imai Y, et al: Reversal of Breast Cancer

Resistance Protein-mediated Drug Resistance by Estrogen Antagonists and Agonists. Mol

Cancer Ther 2:105-112, 2003

110. Imai Y, Tsukahara S, Ishikawa E, et al: Estrone and 17b-estradiol reverse breast cancer resistance protein-mediated multidrug resistance. Jpn. J. Cancer Res.

93:231-235, 2002

111. Ee RPL, Kamalakaran S, Beck WT: Regulation of human breast cancer resistance protein (BCRP/ABCG2) by estrogen. Proceedings of the American

Association for Cancer Research 44:113, 2003

194

112. Janvilisri T, Venter H, Shahi S, et al: Sterol transport by the human breast cancer resistance protein (ABCG2) expressed in Lactococcus lactis. J. Biol.

Chem.:20645-51, 2003

113. Imai Y, Asada S, Tsukahara S, et al: Breast cancer resistance protein transports sulfated conjugates of estrone and 17b-estradiol. Proceedings of the American

Association for Cancer Research 44:112, 2003

114. Amidon GL, Lennernas H, Shah VP, et al: A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm Res 12:413-20, 1995

115. Wu CY, Benet LZ: Predicting drug disposition via application of BCS: transport/absorption/ elimination interplay and development of a biopharmaceutics drug disposition classification system. Pharm Res 22:11-23, 2005

116. Houston JB: Utility of in vitro drug metabolism data in predicting in vivo metabolic clearance. Biochem Pharmacol 47:1469-79, 1994

117. Fisher MB, Campanale K, Ackermann BL, et al: In vitro glucuronidation using human liver microsomes and the pore-forming peptide alamethicin. Drug Metab

Dispos 28:560-6, 2000

118. Pacifici GM, Back DJ: Sulphation and glucuronidation of ethinyloestradiol in human liver in vitro. J Steroid Biochem 31:345-9, 1988

119. Jefcoate CR, Liehr JG, Santen RJ, et al: Tissue-specific synthesis and oxidative metabolism of estrogens. J Natl Cancer Inst Monogr:95-112, 2000

195

120. Hernandez JS, Watson RW, Wood TC, et al: Sulfation of estrone and 17 beta-estradiol in human liver. Catalysis by thermostable phenol sulfotransferase and by dehydroepiandrosterone sulfotransferase. Drug Metab Dispos 20:413-22, 1992

121. Acharya MR, Baker SD, Verweij J, et al: Determination of fraction unbound docetaxel using microequilibrium dialysis. Anal Biochem 331:192-4, 2004

122. Lakhani NJ, Lepper ER, Sparreboom A, et al: Determination of 2- methoxyestradiol in human plasma, using liquid chromatography-tandem mass spectrometry. Rapid Comm Mass Spec 19:1176-1182, 2005

123. Sledge GW, Miller KD, Haney LG, et al: A phase I study of 2- methoxyestradiol (2ME2) in patients with refractory metastatic breast cancer.

Proceedings of The American Society of Clinical Oncology 21:111a, 2002

124. Philip A, Murphy BE: Relative binding of certain steroids of low polarity to human sex hormone-binding globulin: strong binding of 2-methoxyestrone, a steroid lacking the 17 beta-OH group. Steroids 47:373-9, 1986

125. Dunn JF, Merriam GR, Eil C, et al: Testosterone-estradiol binding globulin binds to 2-methoxyestradiol with greater affinity than to testosterone. J Clin

Endocrinol Metab 51:404-6, 1980

126. Pan CC, Woolever CA, Bhavnani BR: Transport of equine estrogens: binding of conjugated and unconjugated equine estrogens with human serum proteins. J

Clin Endocrinol Metab 61:499-507, 1985

127. Reed MJ, Fotherby K: Interaction of ethynyloestradiol sulphate with plasma proteins. J Steroid Biochem 13:495-7, 1980

196

128. Arredondo G, Calvo R, Marcos F, et al: Protein binding of itraconazole and fluconazole in patients with cancer. Int J Clin Pharmacol Ther 33:449-52, 1995

129. Abramson FP: Methadone plasma protein binding: alterations in cancer and displacement from alpha 1-acid glycoprotein. Clin Pharmacol Ther 32:652-8, 1982

130. Entremed: Unpublished Data on file. Rockville, MD

APPENDIX A

PLASMA CONCENTRATION TIME PROFILES OF 2ME2 IN PATIENTS WITH SOLID TUMORS

198

Figure 57: Plasma concentration-time profile of a patient that received a 400 mg single oral dose of 2ME2

199

Figure 58: Plasma concentration-time profile of a patient that received a 800 mg single oral dose of 2ME2

200

Figure 59: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

201

Figure 60: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

202

Figure 61: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

203

Figure 62: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

204

Figure 63: Plasma concentration-time profile of a patient that received a 1600 mg single oral dose of 2ME2

205

Figure 64: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2

206

Figure 65: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2

207

Figure 66: Plasma concentration-time profile of a patient that received a 2200 mg single oral dose of 2ME2

208

Figure 67: Plasma concentration-time profile of a patient that received a 3000 mg single oral dose of 2ME2

209

APPENDIX B

LC/MS/MS ANALYSIS OF 2ME2 IN URINE OF PATIENTS WITH SOLID TUMORS

210

Figure 68: Blank human urine after liquid-liquid extraction

Figure 69: Calibration standard of 2ME2 spiked in human urine at 5 ng/mL

2me_042305_060 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ 5 60 5.46 303.1 > 136.8 1499.89 1.014e+004 100 % 0 min 2me_042305_060 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ 5 60 5.45 308.2 > 138.8 5731.95 4.242e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 70: Calibration standard of 2ME2 spiked in human urine at10 ng/mL

Figure 71: Calibration standard of 2ME2 spiked in human urine at20 ng/mL

2me_042305_062 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ 20 62 5.47 303.1 > 136.8 2291.47 1.906e+004 100 % 0 min 2me_042305_062 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ 20 62 5.43 308.2 > 138.8 5802.98 4.229e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

211

Figure 72: Calibration standard of 2ME2 spiked in human urine at 2 µg/mL

Figure 73: Urine of patient #2 after liquid-liquid extraction

2me_042305_023 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P2U 23 5.47 303.1 > 136.8 8233.87 6.714e+004 100 % 0 min 2me_042305_023 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P2U 23 5.45 308.2 > 138.8 4042.67 2.903e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 74: Urine of patient #2 after treatment with β-glucuronidase and liquid- liquid extraction

2me_042305_035 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P2B 35 5.47 303.1 > 136.8 554993.19 4.459e+006 100 % 0 min 2me_042305_035 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P2B 35 5.45 308.2 > 138.8 4324.06 3.133e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

212

Figure 75: Urine of patient #2 after treatment with sulfatase and liquid-liquid extraction

2me_042305_047 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P2S 47 5.46 303.1 > 136.8 14033.13 1.107e+005 100 % 2.48 0 min 2me_042305_047 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P2S 47 5.43 308.2 > 138.8 3286.27 2.418e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 76: Urine of patient #3 after liquid-liquid extraction

2me_042305_024 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P3U 24 5.47 303.1 > 136.8 1620.30 1.035e+004 100 % 0 min 2me_042305_024 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P3U 24 5.45 308.2 > 138.8 3571.42 2.580e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 77: Urine of patient #3 after treatment with β-glucuronidase and liquid- liquid extraction

2me_042305_036 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P3B 36 5.47 303.1 > 136.8 4002.78 3.088e+004 100 % 0 min 2me_042305_036 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P3B 36 5.43 308.2 > 138.8 4331.35 3.269e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

213

Figure 78: Urine of patient #3 after treatment with sulfatase and liquid-liquid extraction

2me_042305_048 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P3S 48 5.46 303.1 > 136.8 734.33 3.559e+003 100 % 0.01 0.49 1.19 1.86 2.72 4.47 4.79 7.22 7.59 8.40 0 min 2me_042305_048 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P3S 48 5.43 308.2 > 138.8 4200.27 2.962e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 79: Urine of patient #5 after liquid-liquid extraction

Figure 80: Urine of patient #5 after treatment with β-glucuronidase and liquid- liquid extraction

214

Figure 81: Urine of patient #5 after treatment with sulfatase and liquid-liquid extraction

Figure 82: Urine of patient #6 after liquid-liquid extraction

2me_042305_027 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P6U 27 5.46 303.1 > 136.8 2663.74 2.248e+004 100 % 1.32 0 min 2me_042305_027 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P6U 27 5.43 308.2 > 138.8 4259.85 3.132e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Figure 83: Urine of patient #6 after treatment with β-glucuronidase and liquid- liquid extraction

2me_042305_039 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P6B 39 5.46 303.1 > 136.8 268935.94 2.195e+006 100 % 0 min 2me_042305_039 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P6B 39 5.43 308.2 > 138.8 5046.96 3.472e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

215

Figure 84: Urine of patient #6 after treatment with sulfatase and liquid-liquid extraction

2me_042305_051 Smooth(SG,2x1) 2ME MRM of 2 channels,AP+ P6S 51 5.47 303.1 > 136.8 1103.12 7.295e+003 100 % 1.32 3.40 0 min 2me_042305_051 Smooth(SG,2x1) 2ME-d5 MRM of 2 channels,AP+ P6S 51 5.45 308.2 > 138.8 2856.95 2.084e+004 100 % 0 min 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

APPENDIX C

IN VITRO METABOLISM OF 2ME2

217

Figure 85: In vitro metabolism of 2ME2 by CYP 1B1

CYP 1B1

400

300

2ME2 200 mAU

100

0 5 10 15 20 Time (min)

218

Figure 86: Negative control for CYP 1B1

Negative control for CYP1B1

400

300

2ME2

200 mAU

100

-5 0 5 10 15 20 Time (min)

219

Figure 87: In vitro metabolism of 2ME2 by CYP 2A6

CYP 2A6

200

150

100 2ME2 mAU

0 5 10 15 20

Time (min)

220

Figure 88: Negative control for CYP 2A6

Negative Control for CYP 2A6

200

150

2ME2 100 mAU

0 5 10 15 20 Time (min)

221

Figure 89: In vitro metabolism of 2ME2 by CYP 2B6

CYP 2B6

400

300 2ME2

200 mAU

100

0 5 10 15 20

Time (min)

222

Figure 90: Negative control for CYP 2B6

Negative Control for CYP 2B6

400

300

2ME2 200 mAU

100

0 5 10 15 20

Time (min)

223

Figure 91: In vitro metabolism of 2ME2 by CYP 2C19

CYP 2C19

400

300

2ME2 200 mAU

100

0 5 10 15 20 Time (min)

224

Figure 92: Negative control for CYP 2C19

Negative Control for CYP 2C19

400

300

2ME2 200 mAU

100

0 5 10 15 20

Time (min)

225

Figure 93: In vitro metabolism of 2ME2 by CYP 2C9

CYP 2C9

400

300

2ME2 200 mAU

100

0 5 10 15 20

Time (min)

226

Figure 94: Negative control for CYP 2C9

Negative Control for CYP 2C9

400

300

200 2ME2 mAU

100

0 5 10 15 20

Time (min)

227

Figure 95: In vitro metabolism of 2ME2 by CYP 2D6

CYP 2D6

400

300

2ME2

U 200 mA

100

0 5 10 15 20 Time (min)

228

Figure 96: Negative control for CYP 2D6

Negative Control for CYP 2D6

400

300

200 mAU

100

0 5 10 15 20

Time (min)