Early Drug Development E-Book

• Herausgeber: Wiley-VCH Verlag GmbH & Co. KGaA
• Kategorie: Wissenschaft und neue Technologien
• Serie: Methods and Principles in Medicinal Chemistry
• Sprache: Englisch

Seitenzahl: 1546

Veröffentlichungsjahr: 2018

Beschreibung

This one-stop reference systematically covers key aspects in early drug development that are directly relevant to the discovery phase and are required for first-in-human studies.
Its broad scope brings together critical knowledge from many disciplines, ranging from process technology to pharmacology to intellectual property issues.
After introducing the overall early development workflow, the critical steps of early drug development are described in a sequential and enabling order: the availability of the drug substance and that of the drug product, the prediction of pharmacokinetics and -dynamics, as well as that of drug safety. The final section focuses on intellectual property aspects during early clinical development. The emphasis throughout is on recent case studies to exemplify salient points, resulting in an abundance of practice-oriented information that is usually not available from other sources.
Aimed at medicinal chemists in industry as well as academia, this invaluable reference enables readers to understand and navigate the challenges in developing clinical candidate molecules that can be successfully used in phase one clinical trials.

Leseprobe

Table of Contents

Cover

Dedication

Preface

References

A Personal Foreword

A Personal Foreword

Chapter 1: Early Drug Development: Progressing a Candidate Compound to the Clinics

References

Part I: Drug Substance

Chapter 2: Early Phase API Process Development Overview

2.1 Introduction

2.2 API Process Development Overview

2.3 The Transition from Discovery to Development

2.4 Process Development Organizational Construct

2.5 Process Development Equipment

2.6 Summary

References

Chapter 3: The Discovery/Development Transition

3.1 Introduction

3.2 Discovery‐to‐development Transition Before 1980

3.3 Discovery‐to‐development Transition in the 1980s

3.4 Discovery‐to‐development Transition in the 1990s

3.5 Present Practice at BMS

3.6 Application in Small Biotechnology Companies Today

3.7 Application in CROs

3.8 Conclusions

References

Chapter 4: Active Pharmaceutical Ingredient Cost of Goods: Discovery to Early Development

4.1 Introduction

4.2 Stages of Research

4.3 Synthetic Route Translatability and Scalability: Strategy

4.4 Raw Material Considerations

4.5 Continual Assessment of Alternative Routes and Technologies, Including Preparative Chromatography

4.6 Initial CoG Projections

4.7 CoG Versus Campaign Time Cycle

4.8 Synthetic Route Translatability and Scalability: Tactics

4.9 Preparing a CoG Estimate

4.10 Ancillary Expenses

4.11 Long‐Term Considerations

4.12 Summary

Acknowledgments

References

Chapter 5: New Technologies in Process Development

5.1 Introduction

5.2 Synthetic Biochemistry

5.3 Chemical Catalysis

5.4 Continuous Chemistry

5.5 Conclusion

Acknowledgments

References

Chapter 6: Vortioxetine and Early Drug Development Considerations at the Interface of R&D

6.1 Introduction

6.2 Synthesis of Vortioxetine

6.3 Metabolites of Vortioxetine

6.4 Conclusion

Nil Abbreviations

References

Chapter 7: Development of a Practical Synthesis of 4′‐Azido‐2′β‐Methyl‐2′‐Desoxycytosine and Its Prodrugs as HCV Chemotherapeutic Agents

7.1 Introduction

7.2 New Synthesis of (2′

R

)‐2′‐deoxy‐2′‐

C

‐methyl uridine (10)

7.3 Dehydration and Iodoazidation Steps

7.4 Functionalization at C‐4′

7.5 Synthesis of the API

7.6 Solid Form Selection

7.7 Process Safety

7.8 Impurity Strategy

7.9 Conclusion

References

Part II: Drug Product

Chapter 8: Solubility, Permeability, and Their Interplay

8.1 Introduction

8.2 Solubility

8.3 Permeability

8.4 The Solubility–Permeability Interplay

8.5 Summary

Nil List of Abbreviations

References

Chapter 9: Solid‐State Properties

9.1 Introduction

9.2 Amorphous and Crystalline States: Basic Concepts

9.3 Physical Properties of Drug Substance

9.4 Summary

Nil List of Abbreviations

References

Chapter 10: Salt and Cocrystal Screening

10.1 Introduction

10.2 Screening

10.3 Salt/Cocrystal Selection

10.4 Scale‐Up

10.5 Formulation Considerations

10.6 Regulatory Aspects

10.7 Case Studies

10.8 Summary

Nil List of Abbreviations

References

Chapter 11: Particle Size Reduction: From Microsizing to Nanosizing

11.1 Strategic Plans and Risk Management of Particle Size

11.2 Particle Size Reduction Techniques

11.3 Particle Size Analysis

11.4 Bioavailability and the Desired Particle Size

11.5 Enabling Formulation Approach by Particle Size Reduction in Early Drug Development

11.6 Benefits of Commercial Products Using Nanosized Crystalline Particles

11.7 Perspectives in Nanosizing Crystalline Particles

11.8 Conclusions

References

Chapter 12: Early Drug Development: From a Drug Candidate to the Clinic

12.1 Preclinical Formulation Selection

12.2 Formulation Selection for FiH

12.3 Conclusion

Acknowledgments

References

Chapter 13: A Practical Guide for the Preparation of Drug Nanosuspensions for Preclinical Studies: Including

In Vivo

Case Studies

13.1 Introduction

13.2 Selecting the Appropriate Type of Formulation Based on Compound Properties and Type of Study

13.3 Microsuspensions

13.4 Nanosuspensions

13.5 Manufacturing Methods

13.6 Additional Characterizations and Considerations Before

In Vivo

Dose Decisions and Administration Route Selection

13.7 Case Studies

13.8 Conclusions

References

Part III: Pharmacokinetics and Pharmacodynamics

Chapter 14: Integration of Pharmacokinetic and Pharmacodynamic Reasoning and Its Importance in Drug Discovery

14.1 Introduction

14.2 Understand Your Target Biology

14.3 Understand Your Concentration–Response Relationship and Time Delays

14.4 Understand Temporal Differences Between Concentration and Response

14.5 Understand Your Translational Context and Options

14.6 Communication Across Discovery Disciplines

14.7 Final Remarks

References

Chapter 15: Prediction of Human Pharmacokinetics and Pharmacodynamics

15.1 General Introduction

15.2 Predicting Human Pharmacokinetics

15.3 Predicting Human PKPD

15.4 Dose Predictions

15.5 Estimating and Conveying Uncertainty in PK and PKPD Predictions

15.6 Future Perspectives

References

Chapter 16: Translational Modeling and Simulation for Molecularly Targeted Small Molecule Anticancer Agents: Case Studies of Multiple Tyrosine Kinase Inhibitors, Crizotinib and Lorlatinib

16.1 Introduction

16.2 Translational Pharmacology in Oncology

16.3 Quantitative M&S Approach

16.4 Case Study: Crizotinib (PF02341066)

16.5 Case Study: Lorlatinib (PF06463922)

16.6 Closing Remarks

Acknowledgment

Nil List of Abbreviations

References

Chapter 17: Informing Decisions in Discovery and Early Development Research Through Quantitative and Translational Modeling

17.1 Introduction

17.2 Neuroscience: Prediction of the Clinically Efficacious Exposure and Dose Regimen for a Novel Target

17.3 Diabetes: Leveraging a Platform Approach for Two‐way Translation and Integration of Knowledge Between Clinical Lead and Backup Discovery Compounds

17.4 Antibacterials: Semi‐mechanistic Translational PK/PD Approach to Inform Optimal Dose Selection in Human Clinical Trials for Drug Combinations

17.5 Anti‐inflammation: Early Go/No‐Go Based on Differentiation Potential Compared with Competitors

17.6 Summary

Nil List of Abbreviations

References

Part IV: Toxicology

Chapter 18: Preclinical Toxicology Evaluation

18.1 Introduction

18.2 Conclusions

Nil List of Abbreviations

References

Chapter 19: Nonclinical Safety Pharmacology

19.1 Introduction

19.2 Historical Background

19.3 Regulatory Framework

19.4 Role in Discovery and Candidate Selection

19.5 Preparation for First‐in‐human Studies

19.6 Translation from Nonclinical Safety Pharmacology to the Clinic

19.7 Future Directions and Current Discussions

19.8 Summary

Nil List of Abbreviations

References

Chapter 20: Early Drug Development

20.1 Introduction

20.2 Predictive Toxicology

20.3 Predictive Modeling

20.4 Industry Perspectives

20.5 Regulatory Perspectives

20.6 Conclusion

Nil List of Abbreviations

References

Chapter 21: Addressing Genotoxicity Risk in Lead Optimization: A PDE10A Inhibitor Case Study

21.1 Introduction

21.2 Lead Optimization Project: Searching for PDE10A Inhibitors

21.3 Transcriptional Profiling to Capture Polypharmacology

21.4 High Content Imaging as an Independent Confirmation

21.5

In Vitro

Micronucleus Testing to Validate Transcriptional Signature

21.6 Data Integration

21.7 Hypothesis for a Potential Structure–Activity Relationship

21.8 Conclusion

Nil List of Abbreviations

References

Chapter 22: The Integrated Optimization of Safety and DMPK Properties Enabling Preclinical Development: A Case History with S1P

1

Agonists

22.1 Introduction to the S1P

1

Agonist Lead Optimization Program

22.2 Early Attention to Preclinical Safety

22.3 Aryl Hydrocarbon Receptor Activation Observed in Rat

22.4 CYP1A (Auto) Induction Observed in Non‐rodent Species

22.5 Introduction to the Biology and Function of the Aryl Hydrocarbon Receptor

22.6 Considerations of AhR Binding and CYP1A Induction on Compound Progression

22.7 Reacting to Data: Strategy Modification in Lead Optimization

22.8 Iterative Experimentation Identifies Molecules for Progression

22.9 Delivery of Human AhR Agonist Assay

22.10 Minimizing Cardiovascular Safety Risk Through S1P Receptor Selectivity

22.11 Positioning Dose as the Focus of Lead Optimization

22.12 Delivery of Multiple Candidates for Development

22.13 Conclusions

Acknowledgments

References

Chapter 23: From TRAIL to ONC201: Case Study on the Safety Benefit of Developing Targeted Agents Against Cancer‐selective Pathways

23.1 Introduction: Toxicity, a Reason Behind Failed Clinical Trials

23.2 Addressing Safety at the Onset: Targeting a Cancer‐specific Pathway

23.3 Maximizing Efficacy and Minimizing Toxicity at the Bench

23.4 Leveraging Preclinical Animal Studies to Predict Clinical Performance

23.5 Applying Lessons from

In Vitro

and

In Vivo

Studies to Clinical Trials

23.6 Summary

References

Part V: Intellectural Property

Chapter 24: Patent Law Relevant to Early Drug Development

24.1 Introduction

24.2 Overview of Patent Protection

24.3 Requirements for Patent Protection

24.4 Patent Infringement

24.5 Overview of Drug Development

24.6 Extending the Life of a Product

24.7 Summary

References

Chapter 25: Patent Protection Strategy

25.1 Benefits of Patent Protection

25.2 Requirements for Patentability

25.3 The Significance of a Patent Portfolio

25.4 Planning a Patent Portfolio

25.5 Timing of Patent Applications

25.6 Prosecution of Patent Applications

25.7 Extending Patent Coverage Through Additional Applications

25.8 Modifications to Issued Patents

25.9 Conclusion

Chapter 26: Intellectual Property: The Patent Landscape Viewed from Generic and Originator Perspectives

26.1 Introduction

26.2 Market Exclusivities That Protect Branded Drugs

26.3 The Patent Cliff

26.4 Paragraph IV Issues

26.5 Injunctions

26.6 The Generic Company's Goals

26.7 Strategies Adopted by Innovators

26.8 Strategies Adopted by Generic Companies

26.9 Conclusion

Chapter 27: Patent Considerations in Collaborative Drug Development

27.1 Introduction

27.2 What is Intellectual Property?

27.3 Before the Research Begins

27.4 After the Research Ends and the Patent Issues

27.5 Termination of the Relationship (Death and Divorce)

27.6 Conclusion

Nil List of Abbreviations

References

Index

End User License Agreement

List of Tables

Chapter 1

Table 1.1 Target product profile (TPP) example as an essential early drug development tool.

Chapter 3

Table 3.1 Time to the first dose in humans.

Table 3.2 Exclusivity times.

Table 3.3 Early development delay factors.

Table 3.4

FS

‐to‐

FIM

times.

Chapter 4

Table 4.1 CoG calculation, with variables.

Table 4.2 Costs of raw materials and chromatographies.

Table 4.3 Labor costs.

Table 4.4 CoGs for various scenarios.

Chapter 7

Table 7.1 Selected properties of different API solid forms.

Chapter 8

Table 8.1 Traditional definitions for estimating drug solubility.

Chapter 9

Table 9.1 Overview of marketed drug products using amorphous solid dispersions [51, 52].

Chapter 10

Table 10.1 Partial list of pharmaceutically acceptable counterions; coformers would be neutral counterparts.

Table 10.2 An example of maximum daily dose identified for counterions.

Table 10.3 Crystallization composition and processing variables.

Table 10.4 General guidelines for API properties.

Table 10.5 Form matrix for salt selection.

Table 10.6 Examples of salt selections.

Table 10.7 Cocrystal formulations reported for early development studies.

Table 10.8 Examples of process‐induced transformations of salts and cocrystals.

Table 10.9 Examples of commercial products containing different crystalline forms in a variety of dosage forms.

Chapter 11

Table 11.1 Particle size risk assessments for particle size on bioavailability of high and low solubility drugs (may be different based on specific new drug substance and corporate capabilities).

Table 11.2 Relevant guidelines for particle size sample preparation, data analysis, and types of products requiring particle size and/or particulate analysis.

Table 11.3 Particle size measurement and particulate detection techniques common in the pharmaceutical industry.

Table 11.4 Examples of particle size method questions and possible goals.

Table 11.5 Chronology of fenofibrate formulation development.

Table 11.6 Claimed sizes of fenestrations of the vasculature in different organs and selected pathological states.

Chapter 12

Table 12.1 Aims and formulation requirements of different preclinical studies.

Table 12.2 Example of the type of formulation characterization work conducted at the different stages of preclinical formulation development.

Table 12.3 Criteria used to calculate score for clinical performance of PIC formulations.

Table 12.4 Compound properties.

Table 12.5 Physiologically relevant solubility.

Table 12.6 Summary of Caco‐2 permeability data.

Chapter 13

Table 13.1 Scales of milling in Fritsch Planetary micro mill Pulverisette 7 classic line using beads of zirconium oxide.

Table 13.2 Solubility in different media. All measurements after 24 h, at 37 °C. The measurements were not performed with the fluorescence method described, but as described elsewhere [53, 54].

Table 13.3 Solubility in different vehicles. The measurements were not performed with the fluorescence method described, but with a common LC approach [54].

Chapter 14

Table 14.1 Comparison of half‐lives and rate‐limiting processes for two drug case studies.

Table 14.2 Comparisons of the basic distributional, turnover, and on/off‐binding response models with respect to the conceptual structure, parameters, auxiliary parameters, equations, baseline value, and determinants of pharmacodynamic steady state.

Table 14.3 Some typical features in response–time data interpreted by medicinal chemists, pharmacologists, and kineticists.

Chapter 15

Table 15.1 Physiological values from various animal species and humans to be used in the Øie–Tozer equation to predict the volume of distribution (

V

ss

).

Chapter 16

Table 16.1 Outlines of

in vivo

nonclinical tumor studies with crizotinib.

Table 16.2 PK parameter estimates of crizotinib in ALK

WT

– and MET– models.

Table 16.3 PKPD parameter estimates of crizotinib in ALK

WT

– and MET– models.

Table 16.4 PKDZ parameter estimates of crizotinib in ALK

WT

– and MET– models.

Table 16.5 Summary of crizotinib PK–PDDZ parameter estimates in ALK

WT

– and MET– models.

Table 16.6 Outlines of

in vivo

nonclinical tumor studies with lorlatinib.

Table 16.7 PK parameter estimates of lorlatinib in ALK

MT

– and ROS1– models.

Table 16.8 PKPD parameter estimates of lorlatinib in ALK

MT

– models.

Table 16.9 PKDZ parameter estimates of lorlatinib in ALK

MT

– and ROS1– models.

Chapter 17

Table 17.1 Summary of

in vitro

and

in vivo

data utilized in the semi‐mechanistic translational PK/PD model of MK‐7655.

Chapter 18

Table 18.1 A generic drug toxicity development program to support first‐time‐in‐man Phase I clinical studies.

Chapter 19

Table 19.1 Key elements of the core battery assessment (ICH S7A).

Table 19.2 Ventricular repolarization assessments (ICH S7B).

Table 19.3 Test parameters and domains for cerebral function evaluation.

Chapter 20

Table 20.1 Summary of representative performance metrics from testing computational toxicology predictive models applied to bacterial (

Salmonella typhimurium

) mutagenicity data sets to predict DNA‐reactive chemicals and drugs.

Table 20.2 Common statistical measures of performance for predictive computational toxicology classification models.

Table 20.3 Simple confusion matrix to describe the performance of a classification model.

Chapter 22

Table 22.1 The overall profile of exemplar molecules entering early safety studies.

Table 22.2 Comparison of the CYP1A1, CYP1A2 (rat), or CYP1A2 (human) induction observed across various

in vivo

and

in vitro

assay formats.

Table 22.3 The overall profile of

5

having optimal potency, receptor selectivity, PK, human dose, developability, and CYP1A properties suitable for further preclinical development.

Table 22.4 S1P Receptor selectivity comparing fingolimod with

5,

a novel compound from lead optimization.

List of Illustrations

Chapter 2

Figure 2.1 Typical API supply quantities and Process Development stages for a standard drug filing pathway.

Figure 2.2 Typical drug development timeline.

Figure 2.3 Process Development Organization and partner functions.

Figure 2.4 Evolution of equipment for process chemistry.

Chapter 3

Figure 3.1 Chemical structures of propranolol, 5,8‐dihydropropranolol, and Corgard.

Figure 3.2 Chemical structure of BMS‐911543. ()

Figure 3.3 Development route devised for BMS‐911543. ()

Figure 3.4 Process development timeline for BMS‐911543.

Figure 3.5 Schematic representation of the early discovery route and the optimized sequence used in development.

Chapter 4

Figure 4.1 Synthetic route focus by stage.

Figure 4.2 API drug substance cost of good attributes.

Figure 4.3 Cost of goods: iterative projections.

Figure 4.4 Cost of goods: iterative projections – front‐loading process safety.

Figure 4.5 CoG versus time cycle.

Scheme 4.1 Preparation of GSK1292263A.

Figure 4.6 Additional considerations contributing to CoGs.

Chapter 5

Scheme 5.1 Retrosynthetic steps involving some well‐known industrial biocatalytic steps. Enzyme acronyms: P450, cytochrome P450 monooxygenase; URDP, uridine phosphorylase; PNP, purine nucleoside phosphorylase; DERA, 2‐deoxyribose‐5‐phosphate aldolase.

Figure 5.1 Graph prepared from the CHEM21 EFPIA members portfolio review of synthetic routes to prepare active pharmaceutical ingredients (see footnote 1).

Scheme 5.2 Status of various biotransformations (not exhaustive).

Scheme 5.3 Transaminase‐catalyzed preparation of sitagliptin.

Scheme 5.4 Preparative scale IRED‐catalyzed reductive amination of hexan‐2‐one.

Scheme 5.5

In vitro

biohydroxylation of α‐isophorone using wt‐P450

BM3.

Scheme 5.6 Cyclopropanation reactions catalyzed by P450

BM3

lysates.

Scheme 5.7 CAR/ATA/IRED cascade for the preparation of chiral 2‐aryl piperidines.

Scheme 5.8 ω‐Transaminase/acyltransferase cascade for the conversion of aldehydes to amides.

Scheme 5.9 Example of hydrogen borrowing used in pharmaceutical synthesis.

Scheme 5.10 Amide formation by carbonylation in synthesis of lotrafiban.

Scheme 5.11 Typical classes of transition‐metal‐catalyzed C–C bond formation.

Scheme 5.12 Example of direct reductive coupling of an aryl with alkyl bromide.

Scheme 5.13 Further examples of catalytic C–C bond formation applied to pharmaceutical synthesis.

Scheme 5.14 Application of asymmetric hydrogenation with DKR in synthesis of MK‐3102.

Scheme 5.15 Application of catalytic oxidation to a Rosuvastatin precursor.

Scheme 5.16 Application of Chan–Lam coupling on 100 g scale.

Scheme 5.17 Use of asymmetric hydrogenation in synthesis of AZD2716.

Scheme 5.18 Original and improved routes to PDE4 inhibitor.

Scheme 5.19 Intramolecular C–H functionalizationin synthesis of BMS‐911543.

Scheme 5.20 Synthesis of intermediate

1

.

Scheme 5.21 Direct C–H activation using Xantphos as ligand.

Scheme 5.22 Synthesis of a nitrile intermediate in the synthesis of Doravirine.

Scheme 5.23 New conditions for synthesis of a nitrile intermediate in the synthesis of Doravirine.

Scheme 5.24 Synthesis of Elbasvir.

Scheme 5.25 Photocatalytic indoline dehydrogenation to access intermediate

12

.

Scheme 5.26 Microreactor generation of an unstable aryl lithium species.

Scheme 5.27 High temperature Diels–Alder reaction in flow.

Scheme 5.28 Application of flow to a biphasic alcohol oxidation reaction.

Scheme 5.29 Hydroformylation of a (

S

)‐naproxen intermediate.

Scheme 5.30 Use of sodium azide in flow.

Scheme 5.31 Photochemical ring opening of propellane.

Scheme 5.32 Generation and reaction of singlet oxygen in a continuous photoreactor in order to prepare an artemisinin intermediate.

Scheme 5.33 Debromination of a geminal dibromocyclopropane using a split electrochemical flow cell.

Scheme 5.34 Conversion of aldehydes to amides by anodic oxidation.

Scheme 5.35 Methoxylation of

N

‐pyrrolidine using a microelectrolysis flow cell.

Scheme 5.36 Access to five different APIs through alternative configuration of a chemical assembly system.

Scheme 5.37 Access to four different APIs through alternative configuration of a chemical assembly system.

Scheme 5.38 Continuous manufacture of Aliskiren hemifumarate.

Chapter 6

Figure 6.1 Structure of vortioxetine (

1

, Lu AA21004, Brintellix® or Trintellix®).

Figure 6.2 The three main synthesis strategies employed in the medicinal chemistry program leading to vortioxetine.

Scheme 6.1 Foundation of the iron‐mediated synthetic route.

Scheme 6.2 Preparation of vortioxetine using the iron‐mediated synthetic route on solid support.

Scheme 6.3 Synthesis of vortioxetine analog

1a

using parallel chemistry.

Scheme 6.4 Improved synthesis of vortioxetine analog

1a

using the mustard route.

Scheme 6.5 Initial version of the palladium‐mediated synthetic route.

Scheme 6.6 First synthesis of Boc‐protected vortioxetine via the palladium‐mediated synthetic route.

Scheme 6.7 Synthesis of vortioxetine via the “reversed” palladium‐mediated synthetic route.

Figure 6.3 Key building blocks for the palladium‐mediated synthetic strategy.

Scheme 6.8 Synthesis of

11

C‐ and

3

H/tritium (T)‐labeled vortioxetine.

Figure 6.4 Overview of metabolic pathways for vortioxetine.

Scheme 6.9 Synthesis of the proposed mono‐hydroxylated vortioxetine metabolite

29

.

Scheme 6.10 Synthesis of the actual mono‐hydroxylated vortioxetine metabolite

33

.

Scheme 6.11 Synthesis of major vortioxetine metabolite

24

or Lu AA34443.

Scheme 6.12 Synthesis of vortioxetine metabolite

25

.

Scheme 6.13 Failed attempt to synthesize hydroxylamine glucuronide

27

, via direct coupling of hydroxylamine

38

.

Scheme 6.14 Synthesis of

27

via a double reductive amination strategy.

Chapter 7

Figure 7.1 Targets of synthetic interest and a reasonable precursor (

1

).

Scheme 7.1 Original synthesis of (2′

R

)‐2′‐deoxy‐2′‐

C

‐methyl uridine

10

.

Scheme 7.2 Original synthesis of MV075379.

Scheme 7.3 Scalable synthesis of compound

10

.

Figure 7.2 Proposed structures of impurities in uridine protection.

Scheme 7.4 Diastereoselective reduction of compound

18

.

Scheme 7.5 5′‐Dehydration of compound

10

.

Scheme 7.6 Azidation/oxygenation sequence.

Scheme 7.7 Focus on the iodoazidation step.

Figure 7.3 Reference spectra in neat THF.

Scheme 7.8 Alternative azidation sequence via epoxide opening.

Scheme 7.9 Alternative synthesis of compound

24

.

Scheme 7.10 Overview of the end game.

Figure 7.4 Main API process impurities.

Chapter 8

Figure 8.1 The biopharmaceutics classification system as was developed.

Figure 8.2 The correlation between human fraction of dose absorbed (

F

abs

) and the effective permeability (

P

eff

) across the human jejunal membrane.

Figure 8.3 The different mechanisms for intestinal permeability. A, paracellular diffusion over tight junctions; B, transcellular simple passive diffusion; C, carrier‐mediated transcellular transport; D, carrier‐mediated efflux transport; and E, transcellular vesicular transport.

Figure 8.4 Illustration of the unstirred water layer (UWL) adjacent to the intestinal membrane.

Figure 8.5 Etoposide's theoretical (dashed line) and experimental (markers) permeability as a function of the solubility enhancement afforded by the formulation.

Chapter 9

Figure 9.1 Timescales of different crystallization methods used to screen for different physical forms.

Source

: Anderton (2004) [59]. Reproduced with permission of Russell Publishing Ltd.

Figure 9.2 Particle shape descriptors. ()

Figure 9.3 Microscopic images of (a) metformin HCl (as is); metformin HCl crystallized from (b) water, (c) formamide, (d) ethanol, (e) methanol, and (f)

n

‐propanol. ()

Chapter 10

Figure 10.1 Schematic of different crystalline forms. The red box indicates that polymorphs are possible for all the forms listed.

Figure 10.2 Screening strategies in early and late development.

Figure 10.3 Examples of homo‐ and heterosynthons.

Source

: Thakuria et al. 2013 [10]. Reproduced with permission of Elsevier.

Figure 10.4 Schematic diagram of high‐throughput screening.

Source

: Morissette et al. 2004 [21]. Reproduced with permission of Elsevier.

Figure 10.5 Example of a salt screening work flow.

Source

: Gross et al. 2007 [38]. Reproduced with permission of American Chemical Society.

Figure 10.6 Schematic representation of the pH solubility profile of a basic drug indicating that the solubilities may be expressed by two independent curves and that the point where two curves meet is the pH

max

. S

T

is the total solubility; BH

+

and B represent protonated (salt) and free base forms, respectively; the subscript “s” represents the saturation species.

Source

: Adapted from Serajuddin and Pudipeddi 2002 [77].

Figure 10.7 Example of a decision tree for salt/cocrystal selection.

Figure 10.8 (a) Phase diagram of caffeine–glutaric acid–acetonitrile in the temperature range of 10–35 °C; (b) ideal operational region in the phase diagram.

Source

: Adapted from Yu et al. 2010 and 2011 [89, 90].

Figure 10.9 Illustration of ternary phase diagram (API, active pharmaceutical ingredient; CCF, cocrystal conformer).

Source

: Aitipamula et al. 2014 [69]. Reproduced with permission of Royal Society of Chemistry.

Figure 10.10 Formulation processes used in solid oral dosage forms. Highlighted boxes indicate processes that could result in a form change for API or excipients.

Source

: Zhang et al. 2004 [127]. Reproduced with permission of Elsevier.

Chapter 11

Figure 11.1 Spherical equivalent diameters for spherical polystyrene beads (a) and microcrystalline cellulose (b).

Source

: DiMemmo et al. 2011 [18]. Reproduced with permission of Cambridge University Press.

Figure 11.2 Activities and decision points associated with particle size method development and validation.

Source

: Hubert et al. 2008 [19]. Reproduced with permission of American Pharmaceutical Review.

Figure 11.3 Case study for modeling with Bristol‐Myers Squibb Company compound A. (a) Surface response plot of simulated

C

max

change with respect to mean particle diameter and pH change. (b) Surface response plot with simulated AUC change with respect to mean particle diameter and pH change.

Source

: Mathias and Crison 2012 [33]. Reproduced with Permission of Springer.

Figure 11.4 SEM images of the reference batch and the most similar and dissimilar materials using particle size, particle morphology, and surface area as the critical factors: (a) Bx‐100 (API‐07) (reference batch); (b) Bx‐091 (API‐07) (similar batch); (c) Bx‐085 (API‐01) (similar batch); (d) Bx‐022 (API‐15) (least similar batch).

Source

: Ferreira et al. 2016 [34]. Reproduced with Permission of Springer.

Figure 11.5

In vivo

drug exposure of 5 μmol kg

−1

AC88 and BA99 nanosuspension ((a), black bar) and microsuspension ((b), gray bar). Three

in vivo

exposure metrics: (1) Maximum plasma concentration

C

max

(μmol l

−1

). (2) Area under the plasma concentration–time profile AUC(h × kg l

−1

)/dose. (3) Bioavailability (F) was determined by AUCoral/AUCiv F(%). Column charts was drawn based on original data.

Source

: Sigfridsson et al. 2011 [36]. Reproduced with the Permission of Taylor & Francis.

Chapter 12

Figure 12.1 An example of a solubilization technology application map linking formulation selection to API properties and resource required for development at the authors' lab.

Figure 12.2 Wet milling process for preparation of nanosuspension formulations at different volumes with the setup for the smaller volume (1–25 ml) shown on the left and larger volumes (50–2000 ml) shown on the right.

Figure 12.3 Factors to be considered prior to nomination for preclinical formulation work.

Figure 12.4 ASD profiles for free base and HCl salt forms of a drug candidate molecule.

Figure 12.5 Influence of permeability and gastric solubility on (a) clinically successful and (b) clinically unsuccessful PIC formulations.

Figure 12.6 Influence of permeability and intestinal solubility on (a) clinically successful and (b) clinically unsuccessful PIC formulations.

Figure 12.7 Depiction of dose numbers for PIC formulations – influence of permeability and the lowest gastrointestinal solubility measurement on (a) clinically successful and (b) clinically unsuccessful PIC formulations.

Figure 12.8 Comparison of compound T pharmacokinetic profiles for a tablet and PIC formulation.

Chapter 13

Figure 13.1 pH‐solubility profile for a simple salt (circles) of a weak basic (triangles) drug with an intrinsic solubility of 1 μM and a p

K

a

of 4.5 and where the salt form has a solubility of 1 mM. The line indicates the pH

max

, where the thermodynamic solubility of the drug is at its maximum. Below this pH the salt is the stable form and at higher pH the free base is stable.

Figure 13.2 Amorphous nanoparticles of felodipine show Ostwald ripening (filled triangles). With Miglyol present, the increase in particle size was inhibited (open circles).

Figure 13.3 Experimental results for the fraction of crystalline drug nanoparticles of felodipine versus time in dissolution experiments performed at 25 °C where milled crystalline nanoparticles are dissolved in pure water (filled circles) and in water/DMA 9/1 (v/v) (open circles). The total concentration of felodipine was 0.2 μM (water) and 2 μM (water/DMA). The solid lines are the results of theoretical predictions.

Source

: Adapted from Lindfors et al. 2007 [46]

Figure 13.4 The mean plasma concentration (±SEM) of the compound versus time after oral administration of the drug as crystalline microsuspensions (open circles) and as nanocrystals (filled triangles) to rats at 3 μmol kg

−1

(a), 30 μmol kg

−1

(b), and 300 μmol kg

−1

(c).

N

 = 3 for each formulation.

Figure 13.5 (a) Aqueous solubility of ultrasonically crystallized nanoparticles of the compound, stabilized by DPPE‐PEG2000, (b) Aqueous solubility of amorphous nanoparticles of the drug in 1% DMA (v/v) at room temperature (drug/Miglyol/Pluronic L121 3 : 1 : 2, w/w/w). Pluronic was included to inhibit Ostwald ripening. Miglyol as the only additive did not inhibit particle growth. In this experiment the presented light scattering of the nanoparticles was background corrected for the light scattering from the inhibitors.

Figure 13.6 The amorphous particles of the compound increased in particle size without an Ostwald ripening inhibitor present (filled triangles). By including Miglyol/Pluronic L121, the particle growth was avoided (open circles).

Figure 13.7 The mean plasma concentration (±SEM) of the compound versus time after oral administration of 50 μmol kg

−1

drug as crystalline microsuspensions (filled squares), nanocrystals (open squares), and amorphous nanoparticles (open circles) to rats,

n

 = 2. The two crystalline suspensions are superimposed.

Figure 13.8 The amorphous particles of the compound increased in particle size without an Ostwald ripening inhibitor present (filled triangles). By including Miglyol (open circles) or Miglyol/Pluronic L121 (open squares), the particle growth was avoided.

Figure 13.9 The mean plasma concentration (±SEM) of the compound versus time after oral administration of the drug as nanocrystals (open circles) and amorphous nanoparticles (filled triangles) to rats,

n

 = 3. The administered dose was 100 μmol kg

−1

.

Chapter 14

Figure 14.1 Concentration–response–time relationships and how they may be utilized. (a, d) Available information about the

in vivo

pharmacokinetic parameters and dose are combined with, e.g.

in vitro

concentration–response (binding) parameters to simulate the

in vivo

response–time course prior to the execution of an

in vivo

study. (b, e) Acute concentration–time data are used to “drive” the response–time data in a kinetic/dynamic model. From the regression of acute

in vivo

response–time data, the equilibrium concentration–response is predicted. (c, f)

In vivo

response–time data are deconvoluted with an

in vivo

concentration–response model (or

in vitro

binding data) in order to obtain the exposure–time profile behind the response–time course.

Figure 14.2 Schematic presentation of data patterns from three data sets (Examples 1–3) of which the underlying mechanism of action is known (

known target biology

), and three (Examples 4–6) where the exact relation to target mechanism of action is less easy to pinpoint (

unknown target biology

) and therefore a mathematical description may suffice [4, 9].

C, R

,

S,

and

I

denote the drug concentration, the pharmacological response, and the symbols for stimulatory and inhibitory drug action, respectively. MM,

D

ip

, and irrev denote Michaelis–Menten loss, intraperitoneal dose, and irreversible action, respectively.

Figure 14.3 (a) Schematic illustration of the concentration–response relationship at equilibrium. (b) The pharmacodynamic model of drug action as stimulation of the production (turnover rate) or buildup of response. (c) Response–time courses that then show the onset, intensity, and duration of response as well as the peak shift with increasing doses. The time delay between concentration (red bar indicates time of maximum concentration in plasma) and response–time courses will manifest itself as a counterclockwise hysteresis plot (not shown). A peak shift is also seen in the response–time course with increasing doses.

Figure 14.4 (a) Schematic presentation of the

in vivo

concentration–response relationship. The red and blue double arrows depict the concentration intervals covered (graph b) and the corresponding response–time courses (graph c). (b) The red line curve demonstrates the concentration–time course and its relative position to the potency values (a low numerical value of EC

50

equals high potency; a high numerical value equals low potency). (c) The response–time courses corresponding to plasma exposure and potency is indicated in graph b. When potency (EC

50

) is 10, plasma exposure is insufficient to elicit a full response; in this case the response–time course by and large mimics the concentration–time course. For comparison, when plasma exposure exceeds the target potency (EC

50

 = 0.1), the response–time course substantially differs from the shape of the concentration–time course [3].

Figure 14.5 Anti‐atherosclerotic effect of compound X across an exposure range covering more than three orders of magnitude. PD response readout dots represent single observations. The suggested anti‐atherosclerotic concentration falls within the 0.1–2 µM range.

Figure 14.6 Furosemide and schedule dependence. A single 120‐mg dose of furosemide results in less total (Na

+

‐)diuresis over 12 h than does three 40‐mg doses. The 40‐mg area under the effect curve AUC

e

 = 600 mmol Na

+

/12 h; the 120‐mg AUC

e

 = 430 mmol Na

+

/12 h.

Source

: Adapted from Wright et al. (2011) [19].

Figure 14.7 Unbound concentration vs free fraction of test compound X in mouse, gerbil, rat, guinea pig, rabbit, dog, and human plasma. Note the large interspecies differences at pharmacological concentrations (

C

u

 < 0.1 µM), and the species‐dependent nonlinear increase in free fraction

f

u

with increasing unbound concentrations [3].

Figure 14.8 (a) Unbound (red line) concentration is equal in plasma,

C

up

, and tissue (biophase)

C

uT

at steady state due to simple diffusion. (b) Unbound concentration in plasma

C

up

is higher than in tissue

C

uT

due to different

sink

conditions in tissue such as transporters, clearance, irreversible binding, ionization (ion trapping), or bulk flows (CSF). (c) Unbound concentration in plasma

C

up

is lower than in the tissue biophase due to transporters or ionization.

Figure 14.9 (a, d) Schematic illustration of a situation with an instantaneous equilibrium between plasma concentration (red curve) and the pharmacological response (blue curve). The upper graph shows the concentration– and response–time courses which peak at the same time (➁ and ➂;

t

max

) together with two time points with the same exposure and response values, respectively (➀ and ➃). The corresponding lower graph shows the concentration–response relationship derived from the upper two time courses. Note that the rise in response when concentrations increase superimposes the decline in response when concentrations decrease. (b, e) This illustrates a small delay between concentration– and response–time curves, with a shift in their

t

max

values. Plotting the matching concentration–response relationship now results in a loop (hysteresis) for increasing and decreasing concentration–response values. (c, f) A substantial time delay is found between concentration– and response–time courses, carrying noticeably different

t

max

values and terminal slopes. The resultant lower plot also shows a large loop (large hysteresis) with the equilibrium concentration–response relationship occurring within the loop. The two time points with equal exposure (similar

C

p

; ➀ and ➁) demonstrate very different response values due to the time it takes for the pharmacological response to develop [3].

Figure 14.10 (a) Semilogarithmic plot of the concentration–time (red symbols) and suppression of response–time (blue symbols) courses after a 3‐h constant rate drug (Case Study PD21) [4] infusion. Drug action is by inhibition of the turnover rate (buildup) of response. (b) Clockwise concentration–response data from the left graph plotted in time order. Note the differences in the down and upswing of the hysteresis plot. Small gray arrows show the time order. The dotted gray line shows the equilibrium concentration–response relationship. The potency (concentration resulting in half‐maximal response, IC

50

, is approximately 30 nM) and efficacy (intensity, ∼90 response units) are also shown. Note that an initial (∼20–30 min) plasma concentration of 100 nM gives a mild suppression of the response (➀) because the onset of response lags after the rapid rise in the plasma concentration. At the same concentration but in the decay part of the exposure curve (➁), the response is much more suppressed. With a half‐life of 0.4 h for the response, 17 half‐lives (400 min) during which the response has developed will have elapsed at this stage. The response is at each time point a consequence of the prior history of drug exposure, not only total exposure but also the rise, fall, and duration of exposure.

Figure 14.11 (a) Body weight change in control (baseline), benchmark (red), and test compound A‐(magenta) and test compound B (blue)‐treated animals. Note the drift in baseline, the maximal weight gain (upper limit), and the maximum weight loss (lower limit). Drug treatment occurred between days 0 and 18. (b) Exposure vs time data of test compound B. Test compounds were given as a µmol kg

−1

body weight dose.

Figure 14.12 Schematic illustration of concentration vs time data. (a) Assuming data are only AUC

driven

gives the same effects in graph ➀ and different in graph ➁. Situation ➀ may work for strictly irreversible effects, but is questionable for reversible systems displaying saturation. Slowly developing pharmacological responses are often erroneously portrayed as being AUC

driven

. (b) Assuming data are only

C

max

driven

agrees with the differentiation in ➂ but creates a biologically implausible situation in graph ➃. Instances with a rapid equilibrium between plasma concentration and pharmacological response are often erroneously portrayed as being

C

max

driven

.

Figure 14.13 Illustration schematic depicting the iterative interdisciplinary collaboration work mode toward optimization and integration of drug pharmacokinetic and pharmacodynamic properties and a desired candidate drug profile.

Chapter 15

Figure 15.1 Illustration of the “regression‐offset” method for correcting CL

int,u

generated in human hepatocytes due to its systematic underprediction of

in vivo

CL. A regression correlates the measured CL

int,u

in hepatocytes to the observed CL

int,u

estimated from PK data in humans.

Source

: Sohlenius‐Sternbeck et al. 2012 [24]. Reproduced with permission of Taylor & Francis.

Figure 15.2 Prediction of renal clearance (CL

r

) using the KBF method, corrected for species differences in plasma protein binding (

f

u

) in rat (a) and dog (b).

Source

: Reproduced with the permission of Paine et al. 2011 [36].

Figure 15.3 Schematic illustration of a PKPD prediction model‐building framework. This captures both compound and system properties (upper panel) and the key quantitative relationships one should strive to establish, relating the level of efficacy to target engagement to drug concentration time profile (lower panel). The shaded areas in this hypothetical example illustrate the desired level of efficacy (lower panel, right) and how that translates to level of target engagement (lower panel, middle) and drug exposure (lower panel, left).

Figure 15.4 Schematic illustration of the wide range of target engagement required to elicit a meaningful pharmacological effect for different pharmacological target classes.

Figure 15.5 Illustration of the steady state (a) and temporal relationship (b) between PK and the target engagement that the PKPD model should quantify.

Figure 15.6 PKPD model diagram for the absorption and disposition of rHuEpo and its effects on reticulocytes (RET), red blood cells (RBC), and hemoglobin concentrations (Hb) (a). The model was developed in rodents and then translated to humans using a combination of allometry and known species differences in, e.g. cell life span and baseline values for blood cells. The humanized model was used to simulate the pharmacodynamic response to rHuEpo for reticulocytes (RET), red blood cells (RBC), and hemoglobin (HB) (b). Median (solid line) and 90% CI (shaded area) of predictions recapitulate well the observed data (filled circles).

Source

: Mager et al. 2009 [66]. Reproduced with permission of Elsevier.

Figure 15.7 Simulations showing the relationship between predicted daily dose (relative fold change) and the predicted CL in a case where the PKPD relationship suggests that the plasma concentration needs to be maintained over a minimal concentration,

C

e,min

, during the entire dosing interval. Two cases are presented here for an approximate fivefold range of predicted CL: A short half‐life compound ((a): midpoint half‐life estimate

ca

. 7 h) and a long half‐life compound ((b): midpoint estimate for the half‐life

ca

. 30 h). These simulations illustrate how the predicted daily dose can be very sensitive (almost exponentially related) to the predicted CL for the short half‐life compound but less so (close to linearly related) for a compound with a longer half‐life.

Figure 15.8 Examples of cumulative probability distributions of predicted daily doses from Monte Carlo simulations integrating uncertainty in the predicted individual PK and PKPD parameters. The panel (a) compares the predicted total daily dose for once daily, twice daily dosing, and an extended release (ER) formulation. The panel (b) illustrates the comparison between two competing molecules for once daily (UID) versus twice daily (BID) dosing. In both examples, the PKPD model suggested that concentration in plasma needed to be above a certain threshold.

Source

: Sundqvist et al. 2015 [54]. Reproduced with permission of John Wiley and Sons.

Chapter 16

Figure 16.1 Two key aspects for translational understanding of pharmacokinetic–pharmacodynamic‐disease relationships from nonclinical models to the clinic. MTAs, molecularly targeted agent; PK–PDDZ, pharmacokinetics–pharmacodynamics‐disease.

Figure 16.2 Main work stream of quantitative modeling and simulation approaches to characterize

in vivo

pharmacokinetic–pharmacodynamic–disease relationships of molecularly targeted agents. MTAs, molecularly targeted agents; PKPD, pharmacokinetic–pharmacodynamic response; PKDZ, pharmacokinetic–disease response.

Figure 16.3 Observed and one‐compartment PK model‐fitted plasma concentrations of crizotinib in ALK

WT

– and MET– Models. The x‐axis represents the time after dosing in hours, and the y‐axis represents the observed crizotinib plasma concentrations (Obs) with the model‐fitted typical crizotinib plasma concentration–time courses (Pred) in nanograms per milliliter on a logarithmic scale in ALK– and MET– studies.

Figure 16.4 Observed and model‐fitted ALK and MET inhibition by crizotinib in ALK

WT

– and MET–models. The x‐axis represents the time after dosing in hours, the left side of the y‐axis represents the model‐fitted typical crizotinib concentrations in plasma (Cp Pred) and the effect compartment (Ce Pred) in nanograms per milliliter on a logarithmic scale, and the right side of y‐axis represents the observed and model‐fitted typical PD responses (PD Obs and PD Pred, respectively) in the ratio to the mean value of control animal data in ALK– and MET– studies.

Figure 16.5 Observed tumor volumes and model‐fitted TGI curves by crizotinib in ALK

WT

– and MET– models. The x‐axis represents the treatment period in days, and the y‐axis represents the observed individual tumor volumes (Obs) with the model‐fitted typical tumor growth curves (Pred) in cubic millimeters in ALK (A)– and MET (B)– studies.

Figure 16.6 Comparison of crizotinib exposure–response curves for target modulation and TGI in ALK

WT

– and MET– models. Crizotinib exposure–response curves for target modulation (ALK and MET) and tumor growth inhibition (TGI) were simulated at the concentration range of 1–1000 ng ml

−1

with sigmoidal

E

max

model using the estimated PK–PDDZ parameters (EC

50

,

E

max

, and

γ

) obtained from ALK (a)– and MET (b)– studies. The

x

‐axis represents the plasma concentration of crizotinib in nanograms per milliliter on a logarithmic scale, and the

y

‐axis represents the ratios of PD responses from 0 to 1, i.e. target modulation (ALK and MET inhibition) and tumor growth inhibition

Figure 16.7 Summary of quantitative characterization of crizotinib PK–PDDZ modeling for target modulation and antitumor efficacy in ALK

WT

– and MET– models.

C

p

, plasma concentration;

t

, time after dosing;

C

e

, effect‐site concentration;

T,

tumor volume.

Figure 16.8 Prediction of crizotinib‐mediated ALK and MET inhibition in patients following oral administration of crizotinib at the doses of 250 mg twice daily. Crizotinib‐mediated ALK (a) and MET (b) inhibition in patients following 14‐day multiple‐dose oral administration of crizotinib at the clinically recommended doses of 250 mg twice daily were simulated by the link model with the estimated PD parameters obtained from nonclinical ALK– and MET– studies

Figure 16.9 Observed and one‐compartment PK model‐fitted plasma concentrations of lorlatinib in ALK

MT

– models. The

x

‐axis represents the time after dosing in hours, and the

y

‐axis represents the observed lorlatinib plasma concentrations (Obs) with the model‐fitted typical plasma concentration–time courses (Pred) in nanograms per milliliter on a logarithmic scale.

Figure 16.10 Observed and model‐fitted ALK inhibition by lorlatinib in ALK

MT

– models. The x‐axis represents the time after dosing in hours, the left side of the

y

‐axis represents the observed and model‐fitted typical PD responses (PD Obs and PD Pred, respectively) in the ratio to the mean value of control animal data, and the right side of

y

‐axis represents the model‐fitted typical plasma concentrations of lorlatinib (Cp Pred) in nanograms per milliliter on a logarithmic scale.

Figure 16.11 Observed tumor volumes and model‐fitted TGI curves by lorlatinib in ALK

M

– and ROS1– models. The x‐axis represents the treatment period in days, and the

y

‐axis represents the observed individual tumor volumes (Obs) with the model‐fitted typical tumor growth curves (Pred) in cubic millimeters in ALK (a)‐ and ROS1 (b)‐studies.

Figure 16.12 Comparison of lorlatinib exposure–response curves for target modulation and TGI in ALK

MT

– and ROS1– models. Lorlatinib exposure–response curves for ALK inhibition and tumor growth inhibition (TGI) were simulated at the concentration range of 0.01–10 000 ng ml

−1

with sigmoidal

E

max

model using the estimated PK–PDDZ parameters (EC

50

,

E

max

. and

γ

) obtained from ALK (a)‐ and MET (b)‐studies. The

x

‐axis represents the plasma concentration in nanograms per milliliter on a logarithmic scale, the left side of the

y

‐axis represents the tumor growth inhibition from 0% to 120%, and the right side of y‐axis in represents the ALK inhibition from 0% to 100% in ALK‐models.

Figure 16.13 Summary of quantitative characterization of lorlatinib PK–PDDZ modeling for target modulation and antitumor efficacy in ALK

MT

–models.

Figure 16.14 Main work stream to make a Go/No‐Go decision based upon PK–PDDZ understanding in nonclinical studies to increase confidence in drug and target in the clinic.

Chapter 17

Figure 17.1 Key application areas in which translational PK/PD can be applied in discovery and clinical development. The colored blocks indicate the different phases in which addressing these key questions is most likely to be impactful.

Figure 17.2 The translational biomarker scheme representing an overview and mapping of potential biomarkers and assays that can be used to establish PK/PD relationships of a target of interest within a single species and the opportunities for translation across species. This particular scheme shows the availability of assays and data relevant for translational modeling for MK‐1. In the nonhuman primate (NHP), data was generated for MK‐1 and standards of care; data in human was collected for standards of care alone. More generally, this translational biomarker scheme facilitates transparent communication of the perceived translational opportunities and knowledge gaps within a discovery program and promotes alignment on objectives when developing translational modeling plans.

Figure 17.3 Schematic representation of the translational PK/PD framework for MK‐1.

Figure 17.4 Concentration versus receptor occupancy relationship for MK‐1. The red solid line represents the predicted PK–TE profile of MK‐1 in humans using the developed translational modeling framework. The shaded area indicates the 90% confidence interval of the predicted curve calculated from the experimental uncertainty. The blue circles are the actual observed TE data in human, and the blue solid line is the modeled PK–TE profile using Equation 17.1. The dashed line identifies the target occupancy of MK‐1 that is hypothesized to be clinically effective.

Figure 17.5 Predicted human MK‐1 dose versus TE relationship including an explicit expression of modeling uncertainties (shaded area). These simulations graphically show the likelihood of a specific dose of MK‐1 to achieve the desired level of target occupancy hypothesized to be clinically effective.

Figure 17.6 Schematic representation of the translational integrated glucose–insulin model for GPR40 agonists.

Figure 17.7 Illustration of data incorporated in the translational analysis, including PK and temporal GK rat glucose and insulin data from a typical PK/PD study (a), human glucose and insulin data for TAK‐875 (b) [34], and the IVIVC developed to predict

in vivo

EC

50

driving insulin secretion from the

in vitro

IP1 assay EC

50

values across different compounds (middle bottom). The panel at middle top illustrates graphically how the IVIVC is used in the model to relate how increasing plasma drug concentration further enhances glucose‐dependent insulin secretion.

Figure 17.8

(

a) Simulation by the modified IGI model of the relationship between the dose of GPR40 agonist MK‐8666 and the reduction in fasting glucose after two weeks of dosing in diabetes type 2 patients. The solid line represents the mean predicted response and the shaded area the 90% confidence intervals. The black circles show the observed clinical results. (b) Simulations of potential candidates in a similar two‐week proof‐of‐concept study, with reference simulated glucose levels for 150 and 300 mg MK‐8666 shown in light purple and dark blue, respectively, to capture the corresponding projected clinical dose range for the potential candidate molecules.

Figure 17.9 Schematic representation of the semi‐mechanistic PK/PD model structure describing antibacterial effects of the MK‐7655 and imipenem (IPM) combination.

Figure 17.10 Visual predictive check upon fitting the translational model to

in vitro

hollow fiber time–kill data of the MK‐7655 and imipenem (IPM) combination. The open circles are the observed data, the solid line represents the median predicted response, and the shaded area represents the 95% prediction intervals. CFU = colony forming unit.

Figure 17.11

In vivo

model‐based simulations of the bacterial growth–kill response by the MK‐7655 and IPM combination in mouse. The open circles are the observed data, the solid line represents the median predicted response, and the shaded area represents the 95% prediction intervals.

Figure 17.12 Clinical model‐based simulations of the dose–response relationship of MK‐7655 24 h after administration. The solid reference line indicates static growth, and the dashed reference line indicates a one hundred‐fold reduction in CFU. A variety of

P. aeruginosa

strains were simulated including constitutive strains with MICs of 4–64 μg ml

−1

(a) and inducible strains with MICs of 4–64 μg ml

−1

(b).

Figure 17.13 (a) Effect sizes for doses 2 and 3 of MK‐2 from the MAD study, compared with the competitor Phase II data and the current SOC (dotted line). (b) The calculated likelihood that MK‐2 performs worse, similar, or better than either the comparator drug (left) or the SOC (right).

Chapter 18

Figure 18.1

The relation between clinical development stages and nonclinical safety activities

. To enable First Time In Man (FTIM) clinical trials, repeat dose, safety pharmacology and genetic toxicity studies are needed. The duration and timing of the toxicology studies should consider the clinical phases. Human metabolite assessments are needed to ensure the appropriateness of toxicology species, carcinogenicity assessment and reprotoxicity potential. Juvenile toxicity studies may be needed depending on the indication, the toxicological profile and the target ages of paediatric patients. Approval of the drug product may be contingent on post‐approval studies. Throughout development and the product life cycle, additional toxicological support is needed to address emerging impurities.

Chapter 19

Figure 19.1 Temporal relationship between surface ECG and cardiac action potential. Reducing repolarization of the cardiac action potential by inhibiting the

I

kr

current results in an increase in the action potential duration (APD), which, in turn, increases the surface ECG QT interval.

Chapter 21

Figure 21.1 1,3‐Alkoxy‐pyrrolidines substituted with different heteroaromatic ring systems: Q

1

is (substituted) benzofuran, phenyl, quinoxaline, pyridine, quinoline, or pyrazine. Q

2

is (substituted) quinazoline, quinoline, imidazopyridazine, or triazolopyridine.

Figure 21.2 Exploratory analysis of transcriptional effects of all quality filtered informative genes induced by 58 compounds. (a) Spectral map analysis showing on the

y

‐axis a clustering of four compounds based on a subset of tubulin genes indicated in red. (b) Gene profile plot of the tubulin genes together with a summarization of the tubulin genes at the bottom. The summarization reduces the random noise and clearly shows a subgroup of four compounds downregulating the tubulin genes. DMSO samples are indicated in red.

Figure 21.3 Boxplot of the fold changes of all quality filtered informative genes in four different cell lines for compound 8148 with tubulin genes colored in red. The tubulin genes are in general among the most downregulated genes.

Figure 21.4 High content images of U2OS cells expressing an endogenous green fluorescent TUBA1B at 8 h. (a) Nocodazole, an aneugenic MNT positive compound, showing microtubule aggregate formation at a concentration of 25 μM. (b) DMSO controls do not show the typical microtubule aggregates. (c) Compound 8148, showing tubulin downregulation, shows aggregate formation at 10 μM. (d) Compound 0558, within the same chemotype as 8148 showing neither tubulin downregulation nor aggregate formation at none of the concentrations (picture at 30 nM).

Figure 21.5 Scatterplot of the maximum microtubule aggregate score, summarizing three features, observed across the different concentrations at time point 8 h versus the summarization of the quality filtered informative tubulin genes. Only a subset of compounds, the ones profiled in high content imaging, was used to calculate the summarization score of the tubulin genes. A positive microtubule aggregate score corresponds with microtubule aggregate formation. The four compounds downregulating tubulins also show microtubule aggregate formation.

Figure 21.6 Concentration profile plot of the microtubule aggregate score at 8 h for six compounds (vinblastine, colchicine, nocodazole, 4735, 8148, and 0558) indicated with different colors. The concentration at which the image is shown in Figure 21.4 is indicated with the corresponding letter.

Figure 21.7 Scatterplot of the maximum microtubule aggregate score observed across the different concentrations at time point 8 h versus the percentage increase in cell count over 23 h at the corresponding concentration. A positive microtubule aggregate score suggests formation of microtubule aggregates. The horizontal line indicates the minimum increase in cell count over the replicates of the DMSO controls, whereas the average increase in cell count is plotted for DMSO.

Figure 21.8 Micronucleated TK6 cells induced by (a) clastogen mitomycin C and (b) compound 8148 that clearly show formation of large‐sized micronuclei, suggesting an aneugenic mode of action.

Figure 21.9 Ordered Zhang scores of all instances in CMAP based on the tubulin signature of compound 8148. A positive score indicates a similar ranking of the tubulin genes. The top scores of the three reference compounds present in CMAP (vinblastine, nocodazole, and colchicine) are indicated with red dots and annotated.

Figure 21.10 New compounds were synthesized in the quinazoline series, without substitution at R

2

and R

3

and small or electron‐with‐drawing groups at R

1

.

Figure 21.11 Gene expression profiles of all quality filtered informative tubulin genes for a new transcriptional experiment including the newly synthesized compounds, reference compounds, and some positives (colored green) and negatives (colored orange) from earlier profiling experiment. DMSO samples are indicated in red, and their variation is indicated with gray bars. The three compounds, indicated with gray vertical bars, showing different levels of tubulin downregulation are subsequently tested in MNT.

Chapter 22

Figure 22.1 Structure of

f

ingolimod (FTY‐720) and FTY‐720 phosphate.

Figure 22.2 Core elements of the screening cascade for the design and selection of S1P

1

agonists.

Figure 22.3 Upregulation of genes associated with the AhR gene panel for

2, 3

, and

4

following 7‐day oral administration to rats. The response to prototypical inducers β‐naphthoflavone (BNF) and 3‐methylcholanthrene (3MC) after 4‐day administration is included for reference. Cyp: cytochrome P450. NQO1: NAD(P)H quinone oxidoreductase. Ephx1: epoxide hydrolase.

Figure 22.4 The relationship between systemic exposure (AUC) and hepatic CYP1A1 mRNA induction for compounds 2 (squares), 3 (diamonds), and 4 (triangles) following 7‐day oral administration to the rat. Daily doses were 30 and 100 mg kg

−1

for 2 and 30 and 100 and 300 mg kg

−1

for 3 and 4, respectively. The figure illustrates the alignment of a relationship when the AUC is adjusted to represent the unbound exposure (open shapes) compared with the total exposure (closed shapes).

Figure 22.5 Compound

2

is an auto‐inducer in cynomolgus monkeys. Toxicokinetic data (exposure, as area under the curve and Cmax) for

2

on days 1 and 7 following oral dosing at 30 mg kg

−1

 day

−1

to monkeys.

Figure 22.6 Upregulation of hepatic CYP1A1 and CYP1A2 mRNA in cynomolgus monkeys by

2

following 7‐day oral dosing at 30 mg kg

−1

.

Figure 22.7 The relationship between unbound systemic exposure (AUC) and hepatic CYP1A1 mRNA for the inducers

2

(squares),

3

(diamonds),

4

(triangles) and the non‐inducer

5

(circles) following 7‐day administration to the rat. Daily oral doses were 1, 30, 100, and 300 mg kg

−1

for

5

and as previously stated for

2–4

.

Figure 22.8 Structures of compounds

6–21

progressed to the 4‐day rat

in vivo

induction assay. THIQ: Tetrahydroisoquinoline.

Figure 22.9 Summary of the upregulation of CYP1A1 and CYP1A2 mRNA following 4–7‐daily oral administrations of various S1P

1

agonists to the rat. The shaded area represents the boundaries of compounds considered non‐inducers at a given dose. The compounds are colored by template: THIQ (red), aza‐THIQ (pink), indazoles (yellow), benzazepines (blue), and benzoxazepines (green), and shaped by class: acid (square), amine (circle), and zwitterions (diamond).

Figure 22.10 Consistent exposure (as shown by dose‐normalized AUC) was observed with all compounds throughout the 4‐day rat induction study, indicating the absence of auto‐induction in the rat.

Figure 22.11 Examples of reported AhR agonists of a planar hydrophobic nature. TCDD,

22,

and BNF,

23

.

Figure 22.12 Crystallographic data for

5

(a) and

3

(b) illustrating the dihedral angles.

Figure 22.13 Data comparison for 15 S1P

1

agonists comparing agonist activity in the human AhR reporter gene assay and the rat

in vivo

. The compounds are colored by template and symbols separated for clarity: THIQ (red), indazoles (yellow), benzazepines (blue), and benzoxazepines (green), and shaped by class: acid (square), amine (circle), and zwitterions (diamond).

Figure 22.14 The absence of bradycardia in telemetered rats with

5

at 1, 30, or 100 mg kg

−1

compared with

1a (

FTY720).

Chapter 23

Figure 23.1

Schematic diagram of the TRAIL signaling pathway leading to cell death

. TRAIL receptor trimerization occurs upon pathway engagement leading to caspase activation. The diagram depicts the potential for signal amplification with cross talk between the extrinsic and intrinsic pathways of cell death, and negative regulation of cell death by anti‐apoptotic molecules.

Figure 23.2 Chemical structures of some initial TRAIL‐inducing compounds identified in phenotypic screen.

Figure 23.3 Early experience with TRAIL‐inducing compounds and path toward further development of ONC201.

Chapter 26

Figure 26.1 Patentable subject matter in a new drug portfolio.

Figure 26.2 Relationship between time of filing and scope of claim for a new drug patent portfolio.

Chapter 27

Figure 27.1 Prior art: Previous toy design.

Figure 27.2 Modified toy design of invention.

Guide

Cover

Table of Contents

Begin Reading

Pages

C1

vi

vi

xv

xvi

xvii

xix

xx

xix

xx

1

2

3

4

5

6

7

8

9

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

125

126

127

128

129

130

131

132

133

134

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

169

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

367

368

369

370

371

372

373

374

376

377

378

379

380

381

382

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

647

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

721

722

723

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

742

743

744

745

746

747

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765