pharmacy - Principles and Practice of Pharmaceutical Medicine

Document Sample
pharmacy - Principles and Practice of Pharmaceutical Medicine Powered By Docstoc
					Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                       Copyright # 2002 John Wiley & Sons, Ltd
                                                                    ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                       Principles and Practice of
                       Pharmaceutical Medicine
 Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                        Copyright # 2002 John Wiley & Sons, Ltd
                                                                     ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

Principles and Practice of
Pharmaceutical Medicine
                                     Andrew J Fletcher
                             Temple University, Pennsylvania, USA
                                     Lionel D Edwards
                 Novartis and Pharma Pro Plus Inc, New Jersey, USA

                                       Anthony W Fox
                         EBD Group Inc, Carlsbad, California, USA
                                           Peter Stonier
                  Consultant in Pharmaceutical Medicine, Surrey, UK
       Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                              Copyright # 2002 John Wiley & Sons, Ltd
                                                                           ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

Copyright # 2002 by John Wiley & Sons, Ltd.,
                    Baffins Lane, Chichester,
                    West Sussex PO19 1UD, UK

                      National 01243 779777
                      International (‡44) 1243 779777
                      e-mail (for orders and customer service enquiries):
                      Visit our Home Page on: or

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical, photocopying,
recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act
1988 or under the terms of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court
Road, London W1P 0LP, UK, without the permission in writing of the publisher.

Other Wiley Editorial Offices

John Wiley & Sons, Inc., 605 Third Avenue,
New York, NY 10158±0012, USA

WILEY-VCH Verlag GmbH, Pappelallee 3,
D-69469 Weinheim, Germany

John Wiley & Sons Australia, Ltd., 33 Park Road, Milton,
Queensland 4064, Australia

John Wiley & Sons (Asia) Pte, Ltd., 2 Clementi Loop #02±01,
Jin Xing Distripark, Singapore 129809

John Wiley & Sons (Canada), Ltd., 22 Worcester Road,
Rexdale, Ontario M9W 1L1, Canada

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 0-471-98655-0

Typeset in 10/11.5 pt Times from the author's disks by Kolam Information Services Pvt. Ltd., Pondicherry, India
Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham
This book is printed on acid-free paper responsibly manufactured from sustainable forestry,
in which at least two trees are planted for each one used for paper production.
     Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                            Copyright # 2002 John Wiley & Sons, Ltd
                                                                         ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)


About the Editors . . . . . . . . . . . . . . . . . . . . . . . . ix               8   Good Clinical Practices . . . . . . . . . . . . . . . 69
                                                                                        Wendy Bohaychuk and Graham Ball
List of Contributors . . . . . . . . . . . . . . . . . . . . . . xi
Preface      . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii    9   Quality Assurance, Quality Control
                                                                                       and Audit . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Section I:           Overview of                                                         Donna Cullen
                     Pharmaceutical Medicine
                                                                                  10   Phase I: The First Opportunity for
 1    Pharmaceutical Medicine as a Medical                                             Extrapolation from Animal Data
      Speciality . . . . . . . . . . . . . . . . . . . . . . . . . . . 3               to Human Experience . . . . . . . . . . . . . . . . 95
        Michael D. Young                                                                  Stephen Curry, Dennis McCarthy,
                                                                                          Heleen H. DeCory Matthew Marler
 2    What Pharmaceutical Medicine Is and                                                 and Johan Gabrielsson
      Who Does It . . . . . . . . . . . . . . . . . . . . . . . . 13
       W. Wardell and Susan Toland                                                11   Phase II and Phase III Clinical
       and Anthony W. Fox                                                              Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
                                                                                         Anthony W. Fox
 3    Competency-based Training System for
      Clinical Research Staff . . . . . . . . . . . . . . . 17                    12   Phase IV Drug Development:
        Jay D. Miller                                                                  Post-marketing Studies . . . . . . . . . . . . . . 133
                                                                                         Lisa R. Johnson-Pratt
Section II:            Drug Discovery and
                       Development                                                13   Over-the-counter Medicines                 . . . . . . . . . . 141
                                                                                        Paul Starkey
      Introduction . . . . . . . . . . . . . . . . . . . . . . . . 30
        Lionel D. Edwards and                                                     Section III:           Special Populations
        Anthony W. Fox
                                                                                       Introduction . . . . . . . . . . . . . . . . . . . . . . . 154
 4    Drug Discovery: Design and                                                         Lionel D. Edwards
      Serendipity . . . . . . . . . . . . . . . . . . . . . . . . . 31
        Leslie J. Molony
                                                                                  14   Drug Research in Older Patients                   . . . . . . 157
 5    Pharmaceutics . . . . . . . . . . . . . . . . . . . . . . 45                       Lionel D. Edwards
        Anthony W. Fox
                                                                                  15   Drug Development Research
 6    Non-clinical Toxicology                  . . . . . . . . . . . . . . 55          in Women . . . . . . . . . . . . . . . . . . . . . . . . . 165
       Frederick Reno                                                                     Lionel D. Edwards

 7    Informed Consent . . . . . . . . . . . . . . . . . . . . 65                 16   Clinical Research in Children . . . . . . . . . 183
        Anthony W. Fox                                                                   Lionel D. Edwards
vi                                                              CONTENTS

Section IV: Applied Aspects of Drug                                     26   Emergency and Compassionate-use
                                                                             INDs and Accelerated NDA or ANDA
                                                                             ApprovalsÐProcedures, Benefits
     Introduction . . . . . . . . . . . . . . . . . . . . . . . 190          and Pitfalls . . . . . . . . . . . . . . . . . . . . . . . . 299
       Anthony W. Fox                                                          Anthony W. Fox

17   Biotechnology Products and Their                                   27   Japanese Regulations             . . . . . . . . . . . . . . . 307
     Development . . . . . . . . . . . . . . . . . . . . . . 191                           Â
                                                                               Etienne Labbe
       David Shapiro and
       Anthony W. Fox                                                   28   The Development of Human Medicines
                                                                             Control in Europe from Classical
18   Orphan Drugs . . . . . . . . . . . . . . . . . . . . . 203              Times to the Year 2000 . . . . . . . . . . . . . . 325
      Bert Spilker                                                             John Griffin

19   Pharmacoeconomics: Economic                                        29   Ethnic Issues in Drug Registration . . . . 347
     and Humanistic Outcomes . . . . . . . . . . . 211                         Lionel D. Edwards, J.M. Husson
       Raymond J. Townsend, Jane                                                                    Â
                                                                               A. Kumagai, E. Labbe, C. Naito,
       T. Osterhaus and                                                        M. Papaluca, S. Walker, R. Williams
       J. Gregory Boyer                                                        M. Weintraub and H. Yasurhara

20   Pharmacoepidemiology and the
     Pharmaceutical Physician . . . . . . . . . . . 223                 Section VI: Medical Services
       Hugh Tilson
                                                                             Introduction . . . . . . . . . . . . . . . . . . . . . . . 364
21   Statistical Principles and Their                                          Anthony W. Fox
     Application in Biopharmaceutical
     Research . . . . . . . . . . . . . . . . . . . . . . . . . . 231   30   An Introduction to Medical Affairs . . . . 365
       Dan Anbar                                                               Gill Price
22   Data Management . . . . . . . . . . . . . . . . . 259              31   Drug Labeling . . . . . . . . . . . . . . . . . . . . . 371
      T.Y. Lee and Michael Minor                                               Anthony W. Fox
23   Patient Compliance . . . . . . . . . . . . . . . . . 269
                                                                        32   Organizing and Planning Local,
       Jean-Michel Metry
                                                                             Regional, National and International
                                                                             Meetings and Conferences . . . . . . . . . . . 375
24   Complementary Medicines . . . . . . . . . . . 281
                                                                               Zofia E. Dziewanowska,
       Anthony W. Fox
                                                                               Linda Packard and
                                                                               Lionel D. Edwards

Section V: Drug Registration                                            33   Drug Surveillance . . . . . . . . . . . . . . . . . . 379
                                                                               Howard J. Dreskin and
     Introduction . . . . . . . . . . . . . . . . . . . . . . . 288            Win M. Castle
       Anthony W. Fox
                                                                        34   Disease ManagementÐWhat Does
25   United States Regulations              . . . . . . . . . . . 289        It Mean? . . . . . . . . . . . . . . . . . . . . . . . . . . 389
      William Kennedy                                                           Roy Lilley
                                                                 CONTENTS                                                                       vii

35   Publishing Clinical Studies . . . . . . . . . . . 403               40    Outsourcing Clinical Drug
       Anthony W. Fox                                                          Development Activities to Contract
                                                                               Research Organizations (CROs):
                                                                               Critical Success Factors . . . . . . . . . . . . . 461
Section VII:            Legal and Ethical                                        John R. Vogel
                                                                         41    The Third World . . . . . . . . . . . . . . . . . . . 483
     Introduction . . . . . . . . . . . . . . . . . . . . . . . 414              Gamal Hammad
       Sara Croft and Tim Pratt
                                                                         42    Financial Aspects of Clinical Trials . . . . 501
36   Pharmaceutical Product Liability                  . . . . . 421             R. Graham Hughes and
       Han W. Choi and                                                           N. Turner
       Howard B. Yeon
                                                                         43    The Impact of Managed Care on
37   Patents . . . . . . . . . . . . . . . . . . . . . . . . . . . 431         the US Pharmaceutical Industry . . . . . . . 513
       Gabriel Lopez                                                             Robert Chaponis,
                                                                                 Christine Hanson-Divers and
38   Fraud and Misconduct in                                                     Marilyn J. Wells
     Clinical Research . . . . . . . . . . . . . . . . . . . 441
       Frank Wells                                                       AppendixÐUseful Internet Links . . . . . . . . . . 529
                                                                         Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
Section VIII:             Business Aspects
39   The Multinational Corporations: Cultural
     Challenges, the Legal/Regulatory
     Framework and the Medico-commercial
     Environment . . . . . . . . . . . . . . . . . . . . . . . 453
       R. Drucker and
       R. Graham Hughes
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                      About the Editors

ANDREW J. FLETCHER, MB, BChir, (Cantab),                        the efficacy subcommittee Topic 5 (Acceptability
MS (Columbia), FFPM, DipPharmMedRCP, is                         of Foreign Clinical Data) of the International
Senior Assistant Editor of The Merck Manuals,                   Committee on Harmonization (ICH).
Merck & Co. Inc. and Adjunct Professor of                          He is a Fellow of the Faculty of Pharmaceutical
Pharmaceutical Health Care at Temple University                 Medicine and an Adjunct Professor at Temple Uni-
School of Pharmacy. He graduated from Cam-                      versity Graduate School of Pharmacology. He has
bridge University and St. Bartholomew's Hospital,               taught `Drug Development' for PERI for over ten
London, briefly trained in Neurosurgery, joined                 years and is on the teaching faculty of the National
CIBA-GEIGY in the UK as Medical Advisor,                        Association of Physicians. He is a founder member
then European Medical Director, for Syntex, then                of the American Academy of Pharmaceutical Phys-
joined Merck, first in the international division after         icians.
graduating in business from Columbia University,
New York City, he joined the Merck Manual as                    ANTHONY (`Tony') W. FOX, BSc, MBBS,
Assistant Editor. He teaches pharmaceutical medi-               FFPM, MD (Lond), DipPharmMedRCP, CBiol,
cine, bioethics, and medical and scientific writing at          FIBiol, is President of EBD Group, San Diego.
Temple University's School of Pharmacy. He is a                 From The Royal London Hospital, after general
founder member and ex-trustee of the American                   clinical training he was Rotary International
Academy of Pharmaceutical Physicians.                           Fellow at Emory University (Atlanta), and CIBA-
                                                                Geigy Fellow at Harvard. Industrial positions at
LIONEL D. EDWARDS, MB, BS, LRCP, MRCS,                          Procter and Gamble and Glaxo came next. He was
Dip RCOG and FFPM., is President of Pharma Pro                  then Vice-President of a small pharmaceutical com-
Plus Inc., a drug development consulting company                pany. Among many societies, Tony is Charter
and Director Medical Affairs Novartis USA. Previ-               Member, Trustee, and Education VP of the Ameri-
ously, he was Vice President of Clinical Research at            can Academy of Pharmaceutical Physicians. Publi-
Bio-Technology Pharmaceutical Corporation, a                    cations span several areas of pharmaceutical
small Biotech firm making a profit with operations              medicine, e.g., regulation, pharmacology, clinical
in the US and International marketplaces. Prior to              trials, pharmacovigilance, analgesics, migraine,
this he worked at Noven, Inc., a small Skin Patch               genotoxicology, and metabolism. He has four
Technology firm with large internationally licensed             patents, and five journals use his reviews.
partnersÐCiba and Rhone Poulenc Rorer. He was
Assistant Vice-President, International Clinical Re-            PETER D. STONIER, BA, BSc, PhD, MBChB,
search at Hoffman-La Roche, and Senior Director                 MRCPsych, FRCP, FRCPE, FFPM has 24 years
of Schering-Plough International Research, and                  experience in pharmaceutical medicine. He was
Director of US Domestic Gastrointestinal, Hormo-                Medical and Board Director of the UK Hoechst
nal and OTC Research Departments. Dr. Edwards                   Group of companies until he became a consultant
has been involved in all aspects of clinical trials over        in 2000. He is immediate past-President of the Fac-
the years on many different research drug devices in            ulty of Pharmaceutical Medicine of the Royal
10 therapeutic areas.                                           Colleges of Physicians UK. Formerly he was Presi-
   He served as Chairman of the PMA Special                     dent of the International Federation of Associ-
Population committee for 5 years, also he was on                ations of Pharmaceutical Physicians and
the Institute of Medicine Committee for Research                Chairman of the British Association of Pharma-
in Women, sponsored by the NIH. He served on                    ceutical Physicians. He is Visiting Professor in
x                                      ABOUT THE EDITORS

pharmaceutical medicine at Kings College,         marketing and careers in the pharmaceutical
London and at the University of Surrey, which     industry. He is a graduate of Manchester Medi-
under his direction introduced the first MSc      cal School, qualifying in 1974, following a BSc
degree in Pharmaceutical Medicine in 1993. His    degree in physiology and a PhD in protein
publications include edited works in human psy-   chemistry.
chopharmacology, clinical research, medical
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                   List of Contributors

Anbar, Dan Millennium Biostatistics Inc., Bound                 Hanson Divers, Christine US Scientific Initiatives
Brook, NJ, USA                                                  and Customer Support, Health Economics and
                                                                Outcomes Research, AstraZeneca, Apex, NC, USA
Bohaychuk, Wendy Good Clinical Research Prac-
tices Consultants, Lakehurst, Ontario, Canada                   Hammad, G.        Watford, UK
Boyer, Gregory J. Pharmacia Corporation, Stokie,                Hughes, Graham R.             Technomark        Consulting
Illinois, USA                                                   Services, London, UK
Castle, Win M.       Glaxo Smithkline, Philadelphia,            Husson, J.M. Paris, France
                                                                Johnson-Pratt, Lisa R.        Merck & Co. Inc., North
Chaponis, Robert J. Global Medical Affairs, Phar-               Wales, PA, USA
macia Corporation, Peapack, NJ, USA
                                                                Kennedy, William Consultant Delaware, USA
Choi, Han W.                                                    former V.P. Regulation Affairs
Croft, Sara     Shook, Hardy and Bacon, MNP                                  Â
                                                                Labbe, Etienne      Sanofi-Synthelabo, Paris, France
London, UK
                                                                Lee, T.Y.     ACER/EXCEL Inc., USA
Cullen, Donna      Auditrial, Fairlawn, NJ, USA
                                                                Lilley, Roy Independent Health Analyst, former
Curry, Stephen H. President Stephen H. Curry,                   NHS Trust Chairman, Camberley, Surrey, UK
Consulting, Professor of Pharmacology and Physi-
                                                                Linda, Packaid       La Jolla, CA, USA
ology, University of Rochester, NY, USA
                                                                Lopez, Gabriel      Basking Ridge, NJ, USA
DeCory, Heleen H.        Astra Arcus USA Inc., Ro-
chester, USA                                                    Marler, Matthew        Astra Arcus USA Inc., Roches-
Dreskin, Howard J.         Glaxo SmithKline Phila-              ter, USA
delphia, PA, USA                                                McCarthy, Dennis J. Drug Metabolism and Phar-
Drucker, R.      Technomark Consulting Services,                macokinetics, AstraZeneca Pharmaceuticals LP,
London, UK                                                      Wilmington, Delaware, USA

Dziewanowska, Zofia        La Jolla, CA, USA                       Â
                                                                Metry, Jean-Michel        AARDEX Ltd, Zug, Switzer-
Edwards, Lionel D.      Novartis, East Hanover, USA
                                                                Miller, Jay D.      Amgen Inc., Thousand Oaks, Cali-
Fox, Anthony W. EBD Group Inc, 6120 Paseo                       fornia, USA
del Norte, Suites 52±L2, Carlsbad CA 92009, USA
                                                                Minor, Michael       ACER/EXCEL Inc., USA
Gabrielsson, Johan Pharmacokinetics and Phar-
macodynamics Section, AstraZeneca R+D Soderta-                  Molony, Leslie J. Biotechnology Business Strat-
lie, Sweden                                                     egies, Pleasant Hill, CA, USA
Graham, Ball Good Clinical Research Practices                   Naito, C. Teikyo University, Japan
Consultants, Lakehurst, Ontario, Canada
                                                                Ostechaus, Jane T. Wasateh Health Outcomes,
Griffin, John P.    Quartermans, Welwyn, UK                     Park City, Utah, USA
xii                                      LIST OF CONTRIBUTORS

Papaluca, Amati M. EMEA, London, UK                   Turner, Nadia    AstraZeneca, Macclesfield, Chesh-
                                                      ire, UK
Pratt, Timothy Shook, Hardy and Bacon LLP,
Kansas City, USA                                      Vogel, John R.    John R. Vogel Associates, Kihei,
                                                      HI, USA
Price, Gill   VP MedImmune Inc., USA
                                                      Walker, S. Centre of Medicine Research, Carshal-
Reno, Frederick    Merritt Island, FL, USA
                                                      ton, UK
Spilker, Bert    Pharmaceutical Research and
                                                      Wardell, William Wardell Associates           Inter-
Manufacturers of America, 1100 fifteenth street
                                                      national LLC, Princeton, NJ, USA
NW, Washington DC 20005, USA
                                                      Wells, Frank    Medicolegal Investigations Ltd, Ips-
Starkey, Paul Former Vice President Smithkline
                                                      wich, UK
Beecham Consumer Healthcare, Morris Plains, NJ,
USA                                                   Wells, Marilyn J. Department of Health, Physical
                                                      Education, and Recreation, Hampton University,
Shapiro, David    Scripps Clinic, La Jolla, CA, USA
                                                      Virginia, USA
Tilson, Hugh H. University of North Carolina
                                                      Williams, R. US Pharmacopia, Rockville, USA
School of Public Health, Chapel Hill, NC, USA
                                                      Yasurhara, H. Teikyo University, Japan
Toland, Susan Wardell Associates International,
LLC Princeton, NJ, USA                                Yeon, Howard B.
Townsend, Raymond J. Wasatch Health Out-              Young, Michael D. Strategic Healthcare Develop-
comes, Park city, Utah, USA                           ment, Wayne, PA, USA
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)


Pharmaceutical medicine is a relatively new,                       As editors, we would like to thank our contribu-
but rapidly growing, academic discipline in the                 tors for their expertise, their dedication, and their
USA. The American Academy of Pharmaceutical                     vision. We would like to thank and acknowledge
Physicians (AAPP) was founded in 1993 and                       the work and counsel of our colleague Robert Bell,
hosted, in 1999, a meeting of the International                 MD, MRPharmS, who helped us greatly during the
Federation of Associations of Pharmaceutical                    early part of this project. We would also like to
Physicians (IFAPP). The birth of AAPP coincided                 thank and acknowledge the enormous help, en-
with many ongoing changes in the pharmaceutical                 couragement, and patience of the team at John
industry in the USA, as health care delivery began              Wiley & Sons, Inc., UK, with whom we have
to move more towards managed care, and large                    worked closely over these past few years, among
corporations began to amalgamate and downsize.                  whom we have particularly stressed (!) Michael
As these trends continue into the 21st century,                 Davis, Deborah Reece, Hannah Bradley, Lewis
pharmaceutical physicians are increasingly regard-              Derrick, and Hilary Rowe.
ing consultancy work and contract research organ-                  Lastly, we would like to thank our families, and
ization (CRO) affiliation as good career oppor-                 friends, who have withstood the frequent telephone
tunities, and now recognize the need for continuing             calls, e-mails, and meetings, often late into the
education and training in this broad spectrum dis-              night. Indeed, to all who made this project possible,
cipline.                                                        both authors and non-authors, we thank you. We
   This textbook, which represents a collaborative              are certain that this specialty, and our patients,
effort of international experts, is dedicated to the            even though we may help them vicariously, will
more than 3,500 pharmaceutical physicians and all               benefit because of your contributions.
the other professionals working in the US pharma-
ceutical industry and allied fields. It is also                                                          Andrew Fletcher
intended to be useful for those outside the USA                                                           Lionel Edwards
because the basic tenets of the specialty have, for a                                                          Tony Fox
long time, become global.                                                                                   Peter Stonier
     Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                            Copyright # 2002 John Wiley & Sons, Ltd
                                                                         ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)


abbreviated new drug application             safety data 316                           biopharmaceutical research 231±47
      (ANDA) 304±5                         antibodies 195                              biotechnology 441
accelerated approvals 302±5                antisense drugs 196±7                       biotechnology companies 191
acute toxicity animal studies 58           anxiety neurosis 174                        biotechnology industry
additives 355                              archiving 79±80                                impact 10
adeno-associated viruses 197               area under the concentration time curve        patents 200±1
adenosine deaminase (ADA) 198                   (AUC) 47, 48, 60, 63, 101, 351         biotechnology products 191±201
adverse drug experiences (ADEs)            Armitage technique 126                         classes 193±8
   handling 25±6                           arthritis 161                                  clinical trials issues 193±8
adverse drug reactions (ADRs) 367,         ascending dose-ranging cohort                  definition 191
      371, 374, 381, 382, 384, 386              design 123                                ethical issues 199
   database 387                            aspirin 158                                    industry statistics 200
   in children 386                         Association of British Pharmaceutical          informed consent 200
   reporting 337±8, 387, 492                    Industries (ABPI) 335, 420, 445           manufacturing issues 193
adverse events (AEs) 74, 77, 384           ATP 41                                         regulatory considerations 192, 199
   gender differences 174                  audiovisual presentations 410±11               safety issues 198±9
   reporting, ethnic effects 356±7         audit document 263, 266                        sales by market segment 200
age differences 355                        audit result 267                               types 191
agency laws                                audit sheet 263, 266                           vs. conventional drug products
   Middle East 493                         audit summary 263                                 192
aging population                           audits 90±2                                 Black Cohosh 283
   impact on society 157±8                   future trends 92±3                        blastocyst 170
agonist±antagonists 105                    autonomy concept 65                         blinding (or masking) 81, 236±7
Agouron Pharmaceuticals, Inc. 42                                                       blocking 239
AIDS 4, 183, 197, 357                      B. subtilis 34                              blood pressure 161
AIDS-associated infective retinitis 52     Bacille Calmet±Guerrin vaccine              British Association for Lung
AIDS Coalition to Unleash Power                 (BCG) 196                                    Research 49
      (ACTUP) 5                            baseline burden of illness 215±16           British Association of Pharmaceutical
alcohol effects 354                        basic chemical or structural research 6           Physicians (BrAPP) 13
allometric scaling 101±2, 104              Bayesian trial designs 129±30               British Medical Association (BMA)
Alzheimer's disease 158, 161±2, 525        Bendectin 444                                     332
American Academy of Pharmaceutical         benefit±risk analysis 131±2                 British Pharmacopoeia 332
      Physicians 13, 15                    benefit±risk evaluation 383                 Budapest Treaty on the International
American College of Epidemiology           benzocaine 50                                     Recognition of the Deposit of
      (ACE) 226                            benzodiazepines 111, 174                          Micro-organisms for the Purposes
American Heart Association 166             bias 236±7, 239                                   of Patent Procedure 439
analgesic effect-time data 108               concepts 118                              budgets see financial aspects
analysis of variance model                   defences against 119                      Bugbane 283
      (ANOVA) 242                            in clinical trials 118                    bulking agents 47
angina in women 166                          sources 124
angiotensin-converting enzyme (ACE)        binders 47                                  C. elegans 34
      inhibitors 167, 193, 386             binomial model 241                          calcitonin gene-related peptide
animal±human extrapolation 95±115          bioequivalence studies 48                        (CGRP) 194
animal pharmacokinetics 316±17             biologic licence application (BLA) 19,      calendar displays 272
animal pharmacology 316                         20, 25, 27, 28                         Cancer Act 1939 333
animal species 57                          biological products 457                     cancers
animal studies 121                           history 191                                 childhood 183
   acute toxicity 58                       Biological Standards Act 1975 334           Candida albicans 51
   in vitro and in vivo 96                 biological therapeutics 43                  carcinogenicity studies 62±3
   prediction to humans in vivo 101± 4     biomolecular technology 8                   cardiac fatigue 357
532                                                        INDEX
case-control study designs 224             clinical protocols see protocols            competency-based training
case report data and databases 26          clinical research                                 program 17±28
case report forms see CRFs                    competencies associated with 23±6        competitor intelligence (CI) 369±76
causality 233                                 competencies associated with                information source available 370
   assessment 385                                planning 19                           complementary medicines 281±6
   determination 384                          GCP-compliant 69±70                         commonly used 282±5
CBER (Center for Biologicals                  standards 443                               miscellaneous 285
      Evaluation and Research) 290, 297,   clinical research and development plans        regulatory aspects 285
      457                                        (CRDPs) 19, 22                           terminology 281±2
CDER (Center for Drug Evalutaiton          clinical research fraud 443                 complementary therapies 282
      and Research) 297±8, 457             clinical studies                               adverse effects 285
cell-based assays 41                          design 19±20                             compliance see patient compliance
cell products 199                             reports 27±8                             composite hypothesis 232
cellular mechanisms of disease 32             setting up 70±2                          computer-assisted design (CAD) of
center effect 249±50                          sites 71±2                                     drugs 42±3
Center for Biologics Evaluations and       clinical trial certificate (CTC) 336, 344   computer-assisted NDAs
      Research see CBER                    clinical trials                                   (CANDAs) 63, 169, 298
Center for Disease Control and                analysis 256±7                           conferences
      Prevention (CDC) 161                    basic designs and factors suited and        organization and planning 377±80
Central and Eatern Europe (CEE)                  unsuited 123                          confidence 245±7
      459                                     controls on conduct 336                  confidence intervals 245±7
central nervous system drugs 174              enrolment criteria 382                   confidence level 246
Central Pharmaceutical Affairs Council        exemption (CTX) 95, 336, 344, 458,       confidence limits 246
      (CPAC) 309                                 460                                   confirmatory experiments 231
Centre for Medicines Research                 historical 119±20                        consent forms 448
      (CMR) 353, 358                          limitations 120±1                        Consumer Protection Act of 1987 424
chemical diversity 42                         monitoring 24                            contemporaneous independent
chemical libraries 39                         regulations 317±18                             treatment allocation 126±7
childbearing population 177±8                 regulatory governance of 121             contract investigational sites (CISs) 461,
children                                      reporting 256±7                                462
   ADRs in 386                                sample size 381±2                        contract law 416±18, 423, 424
   clinical research 183±8                    see also financial aspects; publishing   contract research organizations
   pharmacokinetic studies 186                   clinical trials                             (CROs) 91±2, 463±84
   pharmacological action of drugs 386     clinicians                                  contributory negligence 426
   see also pediactric studies                role in industry 14 ±15                  controlled clinical trial (CCT) 235±9
China see People's Republic of China       Clostridium welchii 333                     controlling factors in adoption of new
chlordiazepoxide 173                       clotting factors 109±10                           therapeutic agents 5
Chondroitin 283±4                             vitamin K-dependent 110                  copyright 433
chronic lymphocytic leukemias 199          Cochrane Collaboration 281                     in publishing clinical trials 412±13
chronic toxicity studies 61                Code of Federal Regulations (CFR) 457       corporate culture 456±7
chronology diagrams 272                    codes of practice 420                       cost-containment strategies 525
cimetidine 203                             cogenital abnormalities 167                 cost-effectiveness 369
Cimicifuga racemosa 283                    Color Additives Amendments 86               cost-to-benefit consideration 4
CIOMS I 382±3                              combinatorial chemistry 39, 40, 43          costs see financial aspects
CIOMS II 382±3                             Commission for Health Improvement           coumadin 158
CIOMS IV 383                                     (CHImp) 392                           Council for International Organizations
civil law 416±18                           Committee for Orphan Medical                      of Medical Sciences
class A drugs 339                                Products (COMP) 209±10                      (CIOMS) 382±3
class B drugs 339                          Committee for Proprietary Medicinal         Cox II inhibitors 161
class C drugs 339                                Products (CPMP) 209, 340, 341,        CRFs 24, 25, 71, 256, 259
clinical competencies 18±23                      345, 458                                 cover sheet 262
clinical data coordinator (CDC) 260        common technical document (CTD) 345            design of format and content 21±2
clinical development 6, 317                communication to prescribers and               image review process 262
clinical development plan 55±6, 121              patients 388                             managing 26
clinical investigators                     comparative superiority trials 134             preparation 260
   identification and selection 22±3       compassionate use 301±2                     criminal law 416
   sources 23                              competencies associated with clinical       crisis management 9
clinical±legal interface 137                     research 23±6                         critical research organizations see CROs
clinical/marketing interface 138           competency-based education and              CROs
clinical outcomes 212, 213                       training system (CBETS) 17±18            capability 471
                                                       INDEX                                                     533
  choice of contract 512±13             diagnostic procedures 243               drug assessment process
  compatibility 473                     diazepam 51, 173                          role of 361
  critical success steps 466±83         diet 355                                drug case studies 353±4
  early warnings 483                    diethylstilbestrol (DES) 429±30         drug clearance
  financial aspects 509±13              diets                                     elementary aspects 96±7
  financial stability 511±12               high-or low-fat 355                    prediction of human 97±9
  frequent causes of problems 465±6     differential diagnosis 143              drug constituents 45±6
  identifying problems 483              direct-to-consumer advertising (DTCA)   drug delivery targeting 46
  instructions to bidders 474 ±5              campaigns 256, 523                drug development 7
  leveraging experience 473±4           disaster recovery plan 266±7              epidemiology 226±7
  modern view 464                       disclaimers 425±6                         legal/regulatory framework 457±61
  obtaining and comparing bids 510±11   disease                                   procedures 315±20
  pharmaceutical industry views 463±4      cellular mechanisms of 32              process 18
  prequalifying 473                        in vivo models of 36                   regulations, Japan 311
  proposals from 474 ±9                 disease management (DM) 391±406           regulations governing 19
  responsibilities 481                     areas on concern 393±4                 role of 361
  roles 464, 481                           benefits to health service             specific guidelines 315
  selection criteria 471±4                    providers 396±7                   drug differences
  strategies for using 464                 benefits to patients 395±7             influences on 354±5
  traditional view 463±4                   benefits to pharmaceutical           drug discovery 7
  see also sponsor±CRO                        companies 397                       allometric approaches 102±4
cross-over experiments 239                 company requirements 404±5             design and serendipity 31±44
cross-over studies 124 ±5                  contractual framework 392              process 38
cultural challenges 455±7                  definition 391, 392                    project design 31
cultural responsiveness 457                effectiveness 398±9                  drug evaluation 4
CYP2C19 351                                guidelines and protocols 392         drug formulations 45±52, 56
CYP2D6 351                                 implementation in practice 402±4       choice 46
cytokines 32±3, 195±6                      ineffectiveness 399±400                pediatric studies 186
cytomegalovirus retinitis 197              key issues 403                         safety and efficacy trials 136
                                           overview 391±3                       drug holidays 273±4
Dalcon shield 9                            patient perspectives 404             Drug Information Association (DIA) 18
data analysis 254±6                        politics and history 393±4           drug interactions 137
data collection 119                        prospective companies 404              adverse 158
  statistician's role 256                  relationship with pharmaceutical       potential for 382
  with integrity 78                           industry 394±5                    drug labeling 373±6, 386
data confidence 217                        staffing 403                           European Union (EU) 376
data creation flow chart 261            disease-modifying drugs (DMDs) 31         investigational drugs 25
data entry 262                          disproving liability 169                  Japan 373±4
data integrity assurance 78±9           Dixon up±down technique 125               USA 374±5
data management 259±67                  DNA 33, 35±6, 52, 195, 197, 351         drug licensing 227±9
  project material 259                  DNA viruses 197                         drug management
data package 359±60                     donezil 161                               training 9
data process status 263                 dosage                                  drug metabolism 60
data processing 260±6                      pediatric studies 185±6              drug monitoring see drug surveillance;
data provision                          dosage formulations 136, 137                 safety monitoring
  pediatric studies 185                 dosage selection                        drug registration 227±9
data validity 217                          OTC 144                                ethnic issues in 349±63
database release memorandum 266, 267    dose determination 57                        future 362
database update 262                     dose-response date 359                       medical practice 355±6
datasets 263                            dose±response relationship 254               subjective factors 355±8
Debendox 444                            dose ranges                                  terminology, diagnosis and other
decision-making 4, 211, 213±14             gender differences 172                    subjective factors 357±8
Declaration of Helsinki 65, 86          dose size 121                           drug requirements
Dedrick plot 104                        double-blind 237, 241                     non-clinical studies 57
demographics                            double-dummy method 237                 drug research in older patients 157±64
  elderly population 157                double entry 262                          regulatory response 159
Department of Health, Education and     Down's syndrome 167                     drug risk as epidemiologic problem
     Welfare (DHEW) 86±7                downsizing 388                               225
depression 161, 174                     drop-outs 254±5                         drug safety 9
developed countries 157                 drug accountability management 25       drug screening flowchart 39
534                                                        INDEX
drug screening process 41±2                essential hypertension 353                        243, 289±94, 298, 299, 358, 368, 370,
drug surveillance 381±9                    ethical issues 72±3                               371, 374, 419, 420, 445, 457, 460
  see also safety monitoring                 biotechnology products 199                   audits 90±2
drug utilization                             drug research in older                       background 85±7
  elderly population 158                        populations 158±9                         economic considerations 294
  restrictions 521                           pediatric studies 187                        meetings with 296
drugs see investigational drugs;             publishing clinical trials 407               organizational aspects 294±5
     investigational new drug              ethics committee 56, 72±5                   FDA Modernization Act
     application (IND) and specific          financial aspects 507                           (FDAMA) 184
     drugs                                 ethnic differences                          fee-for-service (FFS) health insurance
Durham±Humphrey                              pharmacokinetics/                               policies 515±18
     Amendment 1951 86, 370                     pharmacodynamics 351±2                 fees see financial aspects
                                             prescribing differences 352               female population see gender differences;
E. coli 34                                 ethnic effects                                    women
ECU50u 105, 106                              adverse event reporting                   fertility testing 170
ECG 122                                         (ADRs) 356±7                           fetal damage liabilities 178
ECHO model 213                             ethnic factors                              financial aspects 503±14
economic outcomes 213                        and population extrapolation 359             charges 506
EDU50u 125                                   classification of intrinsic and              clinical trial medication 507
EEG 110±11, 122                                 extrinsic 360                             consultancy 507
effect-compartment model 107, 109            pharmacologic implications 354               CROs 509±13
effect-distribution model 107                potential sensitivity to 359                 equipment costs 508
efficacy 21, 105                           ethnic issues in drug registration 349±63      ethics committees 507
   OTC 143±4                                 future 362                                   external costs 503
   stopping a clinical trial 129             medical practice 355±6                       institutional review boards
Egypt                                        subjective factors 355±8                        (IRBs) 507
   pharmaceutical market 493±4               terminology, diagnosis and other             internal costs 508±10
   private pharmaceutical                       subjective factors 357±8                  investigator fees 503±5
      companies 494                        ethnic susceptibility 352±3                    laboratory charges 506
elderly population                         Europe                                         meetings 508
   clinical studies 162                      Orphan Drug legislation 209±10               multinational trials 508
   compliance in 162                       European Agency for the Evaluation of          patient fees 507
   demographics 157                             Medicines (EMEA) 344, 346, 347            Phase III project 509
   disease issues 161                      European Federation Pharmaceutical             price ratios for common research
   drug utilization 158                         Association (EFFPA) 353                      procedures 504
   informed consent 162                    European Medicines Evaluation Agency           printing and copying 508
   investigative sites 163                      (EMEA) 70, 341±2, 376, 419                regulatory fees 507
   medication regimens 162                 European Patent Convention (EPC) 438           relative costs by country 505
   recruitment 163                         European Patent Office (EPO) 438               relative costs by therapeutic area
   screening 163                           European Union (EU) 335, 336, 346±7,              504
elimination rate 96±7                           458, 459                                  travel 506±7
ELISA (enzyme-linked                         Directives 339±41, 418±19, 424            Fine Chemical Database 40
      immunoabsorbent assay) 40±1            drug labeling 376                         Fine Chemical Directory 42
embryonic malformation 166                   future clinical trial legislation 344     Food and Drug Administration
embryonic stem cells 35                    evergreening process 9                            Modernization Act
Emergency INDs 299±301                     evolutionary designs 125                          (FDAMA) 293±4
enzymes 195                                expectedness 384                            Food and Drug Administration. See
Ephedra spp. 284                           expert reports 63±4                               FDA
epidemiology 223±4                         exploratory experiments 231                 Food and Drug Agency (FDA) 49
   drug development 226±7                  expressed sequence tags (ESTs) 34           Food and Drugs Act 1906 370
   drug registration and licensing 227±9   extrapyramidal reactions 173                Food, Drug and Cosmetic Act 85, 184,
   methodologies 224                                                                         291±2, 370, 419
epidermal growth factor (EGF)              FACS 37                                     formulary enforcement policies
      receptor 32                          falsifiability condition 231, 232                 521
epilepsy                                   Far East                                    formulary status 137
   outcomes 216                               pharmaceutical market 495                fraud 443±51
equipoise concept 65                          registration dossier compilation 498        animal research 448
equivalence trials 135                        trading with 498                            British cases 444±5
error analysis 259                         FDA 19, 27, 47, 58, 67, 85, 134, 145±7,        definition 444
error probabilities 232±3                        149, 150, 160, 175, 192, 206, 212,       detection 446±7
                                                           INDEX                                                           535
  examples 444                             Good Clinical Practice (GCP) 17, 19,         human genomic map 44
  historical aspects 443±5                      69±84, 349, 367, 443, 445±6, 25, 314    human medicines control 327
  investigation 446±7                        basic tenets 69                              centralized procedure 341±2
  literature 446                             compliant clinical research 69±70            classical times to end of 18th
  prevention 445±6                           general regulatory framework 69±70              century 327
  prosecution 447                            implementation 70                            decentralized or mutual recognition
  recent cases 448                           investigator responsibilities 72                procedure 343
                                             sources of international guidelines/       19th and 20th century to Medicines
gall bladder disease 387                        regulations 82±4                             Act 1968 332±4
gases 48±9                                 good laboratory practice (GLP) 55, 57,         recommendations 333
gastric emptying time                           314                                     humanistic outcomes 213
  women 172                                good manufacturing practice                  Hypericum perforatum 283
Gaucher's disease 194, 195, 198                 (GMP) 314                               hypertension 161
gender data 165                              principles and guidelines 338±9            hypothesis 232
gender data collection 178                   standards 338                                testing 243
gender differences 166, 170                Good post-marketing Surveillance
  adverse events 174                            Practive (GPMSP) 315                    ibuprofen 173
  dose ranges 172                          Guidance E5 359                              ICh guidelines 57, 59, 62, 250
  in drug handling 172±4                   Guide to Clinical Trials 18                        Â
                                                                                        IL-1a 33
  metabolism 173±4                         Guideline on Drug Development in the         imipramine 173
  pharmaceutical industry                       Elderly 159                             immigration 158
     practice 175±6                        Guidelines for the Study and Evaluation of   immune adjuvants 196
  weight-for-height tables 172                  Gender Differences in the Clinical      impurities and stability 56±7
  see also women                                Evaluation of Drugs 170                 incidence of a disease 208
gender-related research 174±6                                                           indomethacin 158
gene therapy 43, 197±8                     HU2u blockers 358                            Indonesia
  pharmacokinetics 198                     Haemophilus influenzae 34                       clinical trials 501
gene therapy agents                        harmonization of data requirements 345          health centres 501
  pharmacokinetic properties 193           headache                                        pharmaceutical industry 501±2
General Agreement on Tariffs and             diagnosis 142                              inflammatory bowel disease 50
     Trades (GATT) 440                     health economics 212                         information systems (IS) 397±8
General Consideration for the Clinical     health maintenance organizations             information technology (IT) 397±8, 403,
     Evaluation of Drugs 174                    (HMO) 516                                     527±8
General Considerations for Clinical          network model 517                          informed consent 65±7, 73, 121
     Trials 458                              staff model 517                               benefit±risk assessment 131
General Medical Council (GMC) 445,         Health Maintenance Organizations                biotechnology products 200
     447, 448                                   (HMO) Act 1973 516                         children 66
  disciplinary powers 450±1                health outcomes 212                             elderly population 162
general pharmacovigilance 228              healthcare 212                                  emergency patients 67
  see also pharmacovigilance               healthcare budget 4                             essential elements 65±6
generalizability 239±40, 249±50            healthcare provision 4                          ethical basis 65
generic drugs 304±5                        heart attack in women 166                       pediatric studies 187
  scandal 9                                heart disease in women 166                      principles for conducting 75
genetic counseling 170                     heart failure 161                               responsibility of parties to 67
genetic susceptibility 352±3               height differences 355                          substitute 67
genetic variation 351                      helium/oxygen mixtures 48                       surrogate 66
genomics and new target                    hepatic clearance                               unwritten 66
     identification 34±5                     comparison of predicted and                   written 65±6
geriatric population see aging                  actual 99                               inhalational toxicology 49
     population; older patients              equations 98                               inhalers 49
Gingko biloba 283                          herpes simplex virus 1 (HSV1) 197            initial clinical studies 55
ginseng 284                                hexachlorophene toxicity 9                   initial non-clinical considerations 56±7
Glasgow Coma Score 67                      Hill factor 106                              initial proof of principle 56
global utilization of streptokinase and    HIV 196                                      injectates 51
     t-PA for occluded coronary arteries   homeopathic drugs 285                        innovation in pharmaceutics 52
     (GUSTO) 135                           homeopathy 282                               in situ studies 35
glomerular filtration rates (GFR) 185      homogeneity 240                              institutional review boards (IRBs) 56,
glucosamine/chondroitin                    hormones 195                                       67, 89, 459, 461, 507
     combinations 283±4                    horseradish peroxidase (HRP) 41                 pediatric studies 187
Goldenthal guidelines 62                   Human Genome Project 34                         review 72±5
536                                                            INDEX
integrated summary 61, 63                        drug labeling 373±4                        multiple player influence 522±3
integrated summary of efficacy                   regulations 307                            population and demographics 524±5
      (ISE) 257                                Japanese health authorities 307±10           terminology 516
integrated summary of safety (ISS) 257         Japanese pharmaceutical laws               managed care organization
intellectual property rights 433±4                 310±11                                      (MCO) 516±28
intent-to-treat (ITT) analysis 255±6                                                        types or models 516±17
interference practice 440±1                    Kaplan±Meier analysis 128                  marketing approval application 343±4
interim analysis 253±4                         Kava 284                                   maximum outsourcing 465
interim efficacy analyses 129                  Kefauver±Harris Act 192                    MCA 342
International Clearing House for Birth         Kefauver±Harris Drug                       Meat Inspection Act 290
      Defects Monitoring 179                       Amendements 1962 86, 184, 292,         Med-DRA (Medical Dictionary for
International Conference on                        371                                         Regulatory Activities) 385
      Harmonization (ICH) 13, 17, 55,                                                     Medicaid 158, 461, 522
      57, 59, 62, 87, 159±60, 243, 250, 345,   labeling see drug labeling                 medical affairs 367±72
      349, 358, 383±4, 387, 458, 462, 463      labeling of investigational products 22,     organization 367±71
International Federation of Associations             25                                     overview 367
      of Pharmaceutical Physicians             laboratory charges 506                       personnel 368
      (IFAPP) 13±14                            large, automated, multipurpose,              phaseline 371
international regulatory guidelines for              population-based systems             medical communications 368
      safety pharmacology studies 61                 (LAMPS) 225±7                        medical culture 456, 457
International Society for                      large, simple study 127±8                  medical loss ratio (MLRs) 518±19
      Pharmacoepidemiology (ISPE)              last-observation-carried-forward           medical science liaisons 367
      226                                            method 255                           Medicare 158, 461, 492, 522
international treaties 437±40                  law of contract see contract law           Medicare Prospective Payment System
Internet 4                                     law of tort 416±18                              (PPS) 516
interpretation of measurements 119             LDU50u 58                                  medications and devices
intranasal administration 50                   legal framework for regulating               control at study sites 80±1
investigational drugs                                pharmaceutical products 418±20         management 80
   labeling of 25                              legal principles 416                         overall accountability 81
   packaging of 25                             legal/regulatory framework for drug          preparation 80
investigational new drug application                 development 457±61                     shipment 80
      (IND) 48, 56, 87, 95, 128, 146, 292,     licensing in 388                           medicines
      295±6, 419, 459, 460                     licensing of new medicines 335±7             establishing differences among 208
   see also emergency INDs; treatment          licensing out 388                          Medicines Act 1968 335, 336, 339, 344,
      INDs                                     licensing requirements 130                      418
investigational products                       life-threatening diseases 302±4            medico-commercial environment 461±2
   labeling of 22, 25                          linear models 242                          meetings
   packaging of 22                             link-model 107±9                             audiovisual presentations 410±11
investigators                                  linkage analysis 34                        megatrials 135
   choosing 138                                lipophilic drugs 172                       melatonin 46
   fees 503±5                                  liposomes 197                              menstrual cycle
   responsibilities 72                         lithium 173                                  standardizing for 171
   selection 71±2                              log-in 260±2                               Merck Index 46
investigator's brochure (IB)                   lubricants 47                              meta-analysis 257
   new clinical data 21                        lyophilizates 51                           metabolism
   preparation 20                                                                           differences in 386
in vitro assay development 37±43               Malaysia                                     gender differences 173±4
   primary vs. secondary screens 37±9           product registration 499±500              methaqualone 173
   technological considerations 37              trading with 498±500                      metronidazole 50
   throughput and assay cycles 40±1            managed care 515±29                        Michaelis±Menten equation 106
in vitro data collection 95                     basic concepts 515                        Middle East 492±5
in vitro functional assays 95                   cost containment strategies 520±2           agency laws 493
in vitro/in vivo prediction 96±101              emerging trends 527±8                       company registration
in vivo data 110                                historical perspectives 515±18                 requirements 494±5
in vivo models of disease 36                    impact on pharmaceutical                    pharmaceutical and healthcare
in vivo studies 35                                 industry 519±27                             market 492±3
                                                impact on pharmaceutical product life     migraine syndrome 118
Japan                                              cycle 526                              minimization trials 125±7
  clinical fees 506                             key principles 518±19                     misconduct 443±51
  drug development regulations 311              market competition 523±4                  missing data 255
                                                            INDEX                                                       537
Misuse of Drugs                                drug research in 157±64                 Phase IV studies 147
    Regulations 1985 339±40                       regulatory response 159              prescription-to-OTC switch 146±9
Misuse of Drugs Regulations                  open-label trials 135                     proposed labeling 148
    Act 1971 339                             Oraflex 9                                 safety evaluation 143
model parameters 242                         oral bioavailability                      special study designs 149±50
molecular modeling 40                          elementary aspects 100                  specific clinical testing 145±6
molecular target identification 32±3           prediction 101                          tolerability 143
monitoring 73±6                              oral contraceptives 173, 177, 387         unique characteristics 144±6
 objectives 77                               oral formulations                       over-the-counter (OTC)
More Secret Remedies 333                       optimal design features 47                pharmaceuticals 3
mRNA 34                                      oral suspensions 46±8                   oxazepam 173
multicenter trials 248±50                    oral transmucosal administration 50     oxygen/nitrous oxide 48
multinational corporations 455±62            organ transplantation 199
 languages 457                               orphan disease 203                      p-value 243±5
multiplicity 251±3                           Orphan Drug Act 203, 207±8, 293         P450 cytochrome enzyme systems 351
mutagenicity studies 59±60                     unintended consequences 208           packaging 51±2
mutagenicity testing 170                     Orphan Drug legislation                   of investigational drugs 25
mutations 36                                   Asian countries 210                     of investigational products 22
Mycoplasma genetalium 34                       Europe 209±10                         package insert 322
                                               USA 209                               Palmito caroliniensis 283
National Committee for Quality               orphan drugs 203, 312                   paraldehyde 50
     Assurance (NCQA) 519                      benefits from development             parallel-group studies 122±3
National Heart and Lung Blood                     perspective 207                    Paris Convention for the Protection of
     Institute (NHLBI) 163                     classification 204                         Industrial Property 1883 437±8
National Health Insurance Drug Price           defintion 203                         Parkinson's disease 161
     List 322                                  development 206                       partial agonists 105
National Institute for Clinical Excellence     development disincentives and         patent application 441±2
     (NICE) 392                                   obstacles 207                      Patent Cooperation Treaty
national reimbursement policies 357            discovery 206                              (PCT) 438±40, 442
nausea 150                                     distribution 206                        international phase 439
nebulized drugs 49±50                          economic±medical interface 204±5        national phase 439
negligence 417±18, 425                         heterogeneous group 203               patent protection 435
new biological entities (NBEs) 336             interested parties 205                patentable subject matter 435±6
new chemical entities (NCEs) 42, 55,           marketing 206                         patents 433±42
     159, 336                                  marketing benefits in selling 206±7     biotechnology industry 200±1
new drug applications (NDAs) 45, 63±4,         principles 203±4                        criteria for obtaining 436±7
     92, 117, 131, 175, 208, 214, 292,         regulatory processes 204                history 434±5
     296±7, 303, 304, 313, 460                 specific sources of information 206     role of 434
   content 320                               osteoarthritis 161                      patient compliance 269±79
new drug approval process 320±3              outcomes                                  actions to enhance 274±5
new medicines                                  research 528                            classification 273
   sponsors of 361                             three-dimensional assessment            common patterns 273
New Medicines in Women 177                        212±13                               cost-effectiveness 277
new molecular entities (NMEs) 159, 184         see also pharmacoeconomics and          definition 269±70
new uses for old drugs 43                         specific areas                       direct compliance evaluation 270
NHS and pharmaceutical                       outsourcing 464±5                         during clinical trials 270±2
     companies 394±5                         over-the-counter (OTC) 351                full compliance 273
nitrous oxide/oxygen 48                      over-the-counter (OTC) medication 5,      improving 275
noise and noise reduction 234, 243, 249           141±51, 289                          in elderly population 162
non-clinical summary documents 61              advertising and marketing 150±1         indirect compliance evaluation 270
non-clinical toxicology 55±64                  criteria for use 142                    interactive packaging 276±7
non-compliance see patient non-                development cycle 145                   methods of evaluating 270
     compliance                                dosage selection 144                    monitoring with feedback 275
Normal distribution 245                        efficacy 143±4                          non-compliance 273
NSAIDS 158, 161, 358                           labeling 149±50                         over-compliance 273
null hypothesis 232                            market 148                              partial compliance 273
Nuremberg Code 86                              market support studies 150±1            Phase I clinical studies 271
                                               marketing 145                           Phase II clinical studies 271
Office of Research Integrity (ORI) 445         pharmaceutical physician's role         Phase III clinical studies 271±2
older patients                                    in 144±6                             Phase IV clinical studies 272
538                                                         INDEX
patient compliance (cont.)                    definition 3                              patient compliance 271
  potential players involved in real-time     function 5                                regulatory practice 350
     compliance 275±6                         industrial roles 14±15                    representative female
  reports 272±3                               medical specialty 3±11, 13                   population 171±2
  standards for analyzing real-time           role of 13                                toxicological support 57±61
     compliance data 272                      sub-specializations 14                    women 171±2
  timing non-compliance 274                 pharmaceutical products                   Phase II clinical studies 56, 58, 112, 113,
patient compliance/satisfaction 136±7         legal framework for                          117±32, 458, 462
patient±consumers 4                              regulating 418±20                      common designs 122
patient fees 507±8                          pharmaceutical research and                 industry practice 170
Patient Management Strategies 398                development 525±6                      objectives and prerequisites 121±5
patient needs 5                             Pharmaceutical Research and                 patient compliance 271
patient non-compliance 254                       Manufacturers of America               pharmacoeconomics 214
patient package insert 138                       (PhRMA) 160, 183                       regulatory practice 350
patient participation 124                   pharmacoeconomic outcomes                   small-scale 130
patient rights 4                              current and future uses 219±20            toxicological support 61±3
payer±providers 4                           pharmacoeconomic research 528             Phase III clincal studies 58, 117±32,
pediatric studies 183±8                     pharmacoeconomic trials                        248±9, 354, 458, 462
  data provision 185                          Phase IV 138                            Phase III clinical studies
  dosing 185±6                              pharmacoeconomics 369                       common designs 122
  drug formulations 186                       baseline burden of illness 215±16         financial aspects 509
  ethical concerns 187                        economic and humanistic                   patient compliance 271±2
  final rules 184±5                              outcomes 211±21                        pharmacoeconomics 214
  informed consent 187                        in development programs 213±14            regulatory practice 350
  IRBs 187                                    information demands 213                   toxicological support 61±3
  physiological variations 185                outcomes research 211±12                Phase IV clinical studies 117±39
  placebo control 188                         Phase II studies 214                      goals and tactics 133
  recruitment 186±7                           Phase III studies 214                     investigators 138
  toxicology 186                              reporting and publications 218±19         objectives 133
  vulnerability 187                           studies within clinical trials 216±18     OTC 147
pentamidine 219                               traditional clinical development          patient compliance 272
People's Republic of China 486±92                programme 214±15                       pharmacoeconomic trials 138
  clinical trials 486±7                     pharmacoepidemiology 223±30                 practical aspects and problems 134
     procedures 490±2                         definitions 226                           safety surveillance 136
  pharmaceutical joint ventures 487           future 229±30                             searching for new indications 136
  requirements for authorization of           training 229                              tactical aspects 137±8
     clinical trials 487±90                 pharmacognosy 42±3, 282                     types 133±5
peptides 194                                pharmacokinetic parameters 59, 243        Philippines
performance metrics 481±2                   pharmacokinetic properties 121              clinical trials 500±1
perpherazine 351                            pharmacokinetic studies 60                  health care and pharmaceuticals 500
pessary 51                                    children 186                              regulatory affairs and registration 500
Petasites hybridus 46                       pharmacokinetic/pharmacodynamic           Phiso-Hex (hexachlorophene) 9
phantom fetus 167, 169                           (PK/PD) model/computer-              phocomelia 387
Pharmaceutical Affairs Bureau                    generated feedback 112               Physicians' Desk Reference 386
     (PAB) 308±9                            pharmacokinetic/pharmacodynamic           physicochemical properties 46, 48
pharmaceutical benefit manager                   (PK/PD) models 192, 198              physiological substance models 109
     (PBM) 517, 518, 528                    pharmacokinetic/pharmocodynamic           physiological systems 35
pharmaceutical companies 191                     (PK/PD) models                       pivotal clinical trials 130
Pharmaceutical Education and Research         complex and time-dependent 106±10       placebo comparisons 194
     Institute Inc. (PERI) 18                 elementary aspects 104±5                placebo control
pharmaceutical industry                       single-compartment time-                  pediatric studies 188
  basic cycle 6                                  independent 105±6                    placebo effect 237, 241
  individual or corporate                   pharmacophores 37±8                       placebo treatment 123
     responsibility 420±1                   pharmacovigilance 228, 337±8, 371         plasma concentration 107
  voluntary codes 420                         and problems currently facing           plasma determination 57
Pharmaceutical Manufacturers'                    industry 387±8                       plasma kinetics 107
     Association (PMA) 253                    postmarketing surveillance 492          plasma protein binding 100
  Commission on Drug Safety 368             Phase I clinical studies 56, 95±115,      Pneumocystis carinii pneumonia 219
pharmaceutical medicine                          117±32, 458, 462                     population demographics 350
  coverage 3±5                                industry practice 170                   population derivation 265
                                                              INDEX                                                         539
population extrapolation and ethnic           Proprietary Association of Great             definition 203
     factors 359                                   Britain 335                             prevalence 203
population inclusion and exclusion            proprietary medicinal products            RAS 32
     criteria 239±41                            controls of 340±4                       receptor/ligand assays 40
populations and subpopulations 386            prospective cohort epidemiologic          receptor science 8
postapproval activities 323±5                      studies 224                          receptors 32
post-marketing adverse drug experiences       protein chemistry 8                          sensitivity 353
     (AE) 370                                 protein pump inhibitor 358                recruitment
post-marketing approval medical               proteins                                     pediatric studies 186±7
     affairs 7                                  recombinant vs. natural 41              rectal administration 50 ±1
post-marketing safety                         prothrombin complex activity              re-evaluation system 324 ±5
     surveillance 370±1                            (PCA) 109±10                         re-examination system 324
post-marketing studies 133±9                  protocols 21, 71, 121, 256, 259           regulatory compliance 425
post-marketing surveillance studies 228         design and preparation 20               regulatory development 6
postmarketing surveillance                      elements 20                             regulatory governance
     organization 323±4                       pseudoephidrine 158                          clinical trials 121
postural hypotension 357                      publishing clinical trials 407±14            drug development 19
potency 105                                     audiovisual presentations at academic   regulatory issues
potential sensitivity to ethnic factors 359        meetings 410±11                         biotechnology products 192, 199
precision 245±7                                 CD-ROM vs. textbook 411                 regulatory processes
preclinical research and development 6          classic components in peer-reviewed        orphan drugs 204
preclinical studies 316                            journal 409±10                       regulatory strategy 56
preferred provider organization                 copyright 412±13                        renal perfusion 185
     (PPO) 517                                  desirability of, and biases in 408±9    repeated-dose toxicity studies 58±9
pregnancy 167±70, 177, 387                      electronic facsimiles 411               reporting
  clinical studies 168±9                        electronic journals 412                    clinical research 27±8
  contraindications 386                         ethics 407                                 clinical trials 256±7
  registries 228±9                              isolated abstracts 410                     criteria 384
prescribing differences                         newer forms 411±12                         patient compliance 272±3
  ethnic differences 352                        posters 410                             reproduction studies 62
Prescription Drug Users Fee Act               Pure Food and Drug Act (PFDA) 290±1       request for proposal (RFP) 465, 474
     (PDUFA) 293                                                                        research fraud 443
prescription drugs 4, 5                       quality assurance 87±8                    resource allocations worksheet 475±9
prescription-event monitoring                   imported drugs and medical devices      retroviruses 197
     (PEM) 228                                    (GMPI) 314                            review of products on the market
Prescription Medicines Code of Practice         sampling 259                                  pre-1971 337
     Authority (PMCPA) 420                    quality control 88±90                     review process 321±2
Prescription Only Medicines List                rules 259                               Revised Policy on Inclusion of Women of
     (POM) 339                                quality of life (QOL) 217, 218                  Childbearing Potential in Clinical
press releases 412                            quality protocols 21                            Trials 175
prestudy visits 23                            quality standards 313±14                  Reye's syndrome 9
prevalence of a disease 208                   quantitative interaction 249              rheumatoid arthritis 161
prion-mediated disease 196                    quantitative structure±activity           routes of sale and supply 339±40
probability 231, 234, 239                         relationships (QSARs) 42              Royal College of Physicians (RCP) 11,
proctitis 50                                  query resolution 262                            13, 445
product defects 425                           quinidine 158                             Rx-to-OTC switch 146±8
product labeling see drug labeling
product liability 137, 423±30                 R&D 10                                    Saccharomyces cervisiae 34
  international issues 426±7                    expenditure 7, 8                        safety assessment 57
  landmark cases 427±30                         process 6                               safety committees 124
  legal defenses 425±6                        racial groups                             safety evaluation 55
  principles of law 423±5                       definitions 351                           OTC 143
product licence application (PLA) 117         random error 242                          safety events
product licences 45                           randomization 81, 129, 235±7, 239           reporting and recording 76±8
Product Licences of Right (PLRs) 337          randomized blocks 236                     safety issues 9, 25
product life cycles 526±7                     randomized control trial (RCT) 217          biotechnology products 198±9
pro-glidant 47                                randomized controlled clinical trial        stopping a clinical trial 128±9
project setup 260                                  (RCCT) 241                           safety measures 240
project team formation 260                    rare diseases                             safety monitoring 381
propranolol 173                                 case reports 130                          postmarketing 381±2
540                                                      INDEX
safety parameters 21                     special study designs                        study specifications 466±71
safety pharmacology 60                      OTC 149±50                                   importance of accuracy 466
   international regulatory guidelines   specialized designs 128                         preparation 471
      61                                 spironolactone 161                              worksheet 466±71
safety reports                           sponsor roles and responsibilities           study subjects
   preparing 27                                479                                       information to be provided 75±6
safety studies 360                       sponsor±CRO                                  subgroups 238
safety surveillance                         communication/decision-making             subspecializations in pharmaceutical
   Phase IV clinical trials 136                model 482                                    medicine 14
St John's Wort 283                          end-of-study meeting 483                  summary basis of approval (SBA) 323
salt content 355                            periodic oversight meetings 482±3         summary of product characteristics
SAM-e 284                                   problems 465±6                                  (SPC) 346
sample size 247±8                           relationship 479±83                       suppositories 50
Saudi Arabia                                study initiation meeting 482              surveillances studies
   healthcare structure 493              sponsors of new medicines 361                   postmarketing 386±7
Saw Palmetto 283                         spontaneous case reports 384±5               syrups 46±8
scan process 260±2                       spontaneously arising phenotypic             Systolic Hypertension in the Elderly
schedule 1 339                                 models 36                                    Person (SHEP) 163
schedule 2 339                           SRC 32
schedule 3 339                           stability and impurities 56±7                tablets 46±8
schedule 4 339                           stability studies 316                        tactical outsourcing 464
schedule 5 340                           stability testing 52                         target identification 38
Scientific Committee on Medicinal        standard deviation 242                          methods 33±5
      Products and Medical               standard deviation (SD) 235, 237±8,             molecular 32±3
      Devices 344±5                            243, 247, 248                          target validation 36±7
scientific experiment 231                standard error (SE) 243                      teratogenic issues 167±9
scientific meetings                      standard error of the mean (SEM) 235         teratogenicity 167
   organization and planning 377±80      Standard Normal distribution 245                testing 170
scientific method 231                    standard operating procedures                teratology studies 62
scintillation proximity assay (SPA) 40         (SOPs) 87                              test statistic 243
Secret Remedies 332                      statistical bias 236±7                       Thailand
seed protection 433                      statistical inference 243±7                     health care and pharmaceutical
selection bias 236±7                     statistical method 231±2                           industry 496±7
selection of subjects 240±1              statistical model 241±3                      thalidomide 56, 62, 167, 334±5, 371, 387,
self-diagnosis 142                       statistical power 232±3                            427±9, 444
self-medication 141±2                    statistical principles 231±47                theophylline 50
sequential analysis study design 126     statistical test 232±3                       therapeutic coverage 272±3
sequential cohort designs 123            statistical trial design 248±54              therapeutic index 143
serious adverse event (SAE)                 overdesign 248                            Therapeutic Substances Act
      procedures 24                      statistics 118±19                                  (TSA) 333±4
serious diseases 302±4                   status reporting 263                         Third World 158, 485±502
signal 234                               stopping a clinical trial                       overview 485
signal-to-noise ratio 234, 240±1            efficacy issues 129                       three-dimensional outcome
signal transduction enzymes 41              safety issues 128±9                             assessment 212±13
significance level 233                   strata 238                                   time course of effect 105
simple hypothesis 232                    strategic outsourcing 465                    tissue products 199
Singapore                                stratification 127, 171, 237±9               TNF 32±3
   health care and pharmaceutical        strict liability 418, 423±4                  topical drugs 50
      industry 497±8                     stroke 161                                   tort law 423, 424
skewed dosing 274                        structure±activity relationship (SAR) 8,     toxic shock syndrome 9
skipped dosing 274                             42                                     toxicokinetic assays 57
societal culture 455±7                   Student's t-distribution 244                 toxicokinetic data 60
societal development 6                   Studies in Support of Special Populations:   toxicokinetics 59
socioeconomic influences 357                   Geriatrics 159                         toxicological coverage 121±2
source data verification 78±9            study closure 26                             toxicological support
South Korea                              study design                                    Phase II and Phase III studies 61±3
   health care and                          overdesign 248                               pre-IND and Phase I clinical
      pharmaceuticals 495±6                 sample size 247±8                               studies 57±61
special populations 137                  study documentation 23±4                     toxicological testing 46
special studies 63                       study initiation 23±4                        toxicology
                                                           INDEX                                                         541
   non-clinical 55±64                        Orphan Drug legislation 209              Waxman±Hatch Amendment 293
   pediatric studies 186                     regulations 289±98                       weight differences 355
trade secrets 433                                                                     weight-for-height tables for males and
trademarks 433                             vaccines 196                                    females 172
training                                   vaginal administration 51                  wellness management 400±2
   drug management 9                       validation 119                             wholesale dealers' licences 339
   pharmacoepidemiology 229                value-added clinical development           within-patient dose titration designs
training program                                 program 214±15                            124
   competency-based 17±28                  variability                                women
transdermal drugs 50                          source of 233±5                           disease presentatons 166±7
transgenic technology 35±6                 variables measured before and after drug     drug development research
treatment-by-center interaction 249              (or placebo) administration 119           165±81
Treatment INDs 301±2                       variance 234                                 drug research subjects 176
treatment withdrawal study designs 128     Venereal Disease Act 1917 333                gastric emptying time 172
Trichomonas 51                             vitamin K-dependent clotting                 in clinical studies 56
Tylenol 9                                        factors 110                            Phase I studies 171±2
type A and B reactions 384±5               volumes of distribution                      representative population 167, 177
type I and II errors 118, 232, 233, 242,      elementary concepts 100                   response to medications 166
     244                                      prediction of human 100                   see also gender differences
                                           volunteer studies 171                      World Intellectual Property
uncertainty                                vulnerability                                   Organization (WIPO) 438, 439
  source of 233±5                             pediatric studies 187                   Written Opinion 439
uncertainty conditions 231±2
unmet clinical needs 31                    wafers 46±8                                Zollinger±Ellison syndrome 203
USA                                        warfarin 109, 158                          Zomax 9
  drug labeling 374±5                      warranty 424±5
Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                       Copyright # 2002 John Wiley & Sons, Ltd
                                                                    ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                               Section I
     Overview of Pharmaceutical Medicine
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

         Pharmaceutical Medicine as a Medical
                                               Michael D. Young
                                      Celltech Chiroscience PLC, Slough, UK

Medicine is an art that has been practiced since                   The exponential growth in scientific knowledge,
time immemorial. The use of herbs and natural                   particularly over the last 100 years, has brought
medicaments to relieve pain or to aid the sick in               about a paradigm shift in our approach to pharma-
coping with their afflictions has been a part of all            ceuticals. Until the twentieth century, the sale and
societies. In the Western world, medicine has de-               use of medicines and medical devices was almost
veloped at least since the time of the Greeks and               entirely unregulated by governments. It was a
RomansÐthe Hippocratic oath reminds us of this                  case of caveat emptor, with only the drug-taker's
nearly 2500 year history. However, the progress of              common sense to protect against the dangers of the
medicine has been very different from that of many              so-called patent medicines and `snake oils'. The
other arts within society. It has come of age after an          obvious abuses in these situations led eventually
incredibly long maturation period. As a function                to government intervention, to professional regula-
capable of offering a successful treatment for an               tion, and to requirements that drugs be pure and
human ailment, medicine is very much a develop-                 unadulterated. With advances in science and in the
ment of the last 100±150 years. Indeed, the major               ability to define and establish drug efficacy came a
advances have come in the last 50±75 years.                     requirement to demonstrate that drugs were also
   The role of physicians in society has changed                safe. Finally, as late as the second half of the twen-
over the centuries. It may have reached its nadir               tieth century came the legel requirement to estab-
during the early renaissance, when the general                  lish that pharmaceuticals were effective before they
attitude was, as Shakespeare said, `Trust not the               were marketed. These legal requirements reflected
physician; his antidotes are poison'. From the nine-            changes in social attitudes and expectations
teenth century onward, as their diagnostic under-               grounded in the questions that the development
standing has grown and as their therapeutic agents              of biological and basic sciences had made it pos-
have become increasingly effective, physicians have             sible to ask and to answer. The response to these
come to be increasingly valued. Today, much of                  changes has led to the development of the specialty
the practice of medicine in all of its subspecialties           of pharmaceutical medicine.
is based on a physician's diagnosis and treatment                  Pharmaceutical medicine can be defined as: `the
with drugs, devices, or surgery. This radical change            discipline of medicine that is devoted to the discov-
to an era of focused treatments, after aeons of using           ery, research, development, and support of ethical
homespun remedies and then watching hopefully                   promotion and safe use of pharmaceuticals, vac-
for the crisis or the fever to pass, has accompanied            cines, medical devices, and diagnostics'. (By-laws
the recent revolutions in the understanding                     of the American Academy of Pharmaceutical Phys-
of biological processes and in technical and bio-               icians). Pharmaceutical medicine covers all medic-
technical capabilities. These developments have                 ally active agents from neutraceuticals, through
allowed us to produce pure therapeutic agents                   cosmeceuticals and over-the-counter (OTC)
and to establish how to use them safely and effect-             pharmaceuticals, to prescription drugs. Further-
ively.                                                          more, the specialty is not confined to those
4                                             PRINCIPLES AND PRACTICE OF PHARMACEUTICAL MEDICINE

physicians working within what is classically con-                        with what, for some, may be seen as the rationing
sidered the pharmaceutical industry, but includes                         and/or the means-testing of access to the totality of
those involved in the clinical management or regu-                        healthcare options. These are significant ethical
lation of all healthcare products. It is the basic                        and social issues and physicians within the pharma-
specialty for physicians within the cosmetics and                         ceutical industry or the health regulatory agencies
nutrition industry for those in the device industry                       will inevitably be required to provide a perspective,
and for those in `not-for-profit' companies, such as                      both internally and to those outside.
those responsible for the national blood supplies                            The second new decision maker in the provision
and/or for specialized blood products. Further-                           of healthcare has arrived even more recently as a
more, it is the fundamental discipline for physicians                     crucial component. These are the end-user or pa-
who are in government health ministries, insurance                        tient groups. The rising status of the physician since
companies, National Health Trusts or HMO man-                             the nineteenth century encouraged a paternalistic
agement, drug regulatory agencies or any other                            doctor±patient relationship, with the physician
oversight or regulatory function for healthcare.                          clearly in the lead. In recent times the nature of
   In the early part of this quarter-century, for a                       this relationship has come under question. The
medicine to be adopted and for it to sell, it was                         advent of holistic medical concepts focused on the
sufficient that science could conceive of a new                           whole patient, and taking into account the entirety
treatment, that technology could deliver that treat-                      of an individual patient's life has forced changes in
ment, and that clinical research could prove it ef-                       the focusing of any therapeutic interaction. The
fective and safe for the physician to use. This is no                     general increase in educational standards within
longer the case.                                                          the developed world and the massive increase in
   Over the past three decades we have seen the                           available information culminating today with the
emergence of two major influences in decisions                            electronic media and the Internet has inevitably
about new advances in healthcare. These are the                           produced a more informed patient. This has em-
payer±providers and the patient±consumers. Their                          powered the patient and led to the formation of all
role in the decision-making process has increased                         kinds of public interest and patient groups. Fur-
rapidly in the last 25 years, as can be seen in                           thermore, the ability in this century to think in
Figure 1.1                                                                terms of the maintenance of good health and even
   With an increasing proportion of society's                             of the abolition of disease (e.g. smallpox and polio)
healthcare budget going on pharmaceuticals, even                          has changed the patient's and society's attitude to
a growth in the percentage of the gross national                          what they can and should expect of physicians.
product that governments are willing to allocate to                       Today we are very much moving towards a balance
healthcare has been unable to meet the demands of                         in the therapeutic interaction, if not to a patient±
unbridled development. This has made the payer/                           doctor relationship. This change is a seminal one
provider a major determiner of the use of pharma-                         for the delivery of healthcare and for the develop-
ceuticals. All possible treatments cannot be freely                       ment of new therapeutic agents.
available to all and a cost-to-benefit consideration                         For prescription drugs, the major factor bringing
has had to be introduced. This in turn has ensured                        about the involvement of patient groups was prob-
that pharmaceutical medicine involves pharmacoe-                          ably the revolution in the new drug evaluation
conomics training and even media training to deal                         process caused by the AIDS epidemic. This terrible
                                                                          affliction occurred at a time when groups within
                                                                          society were forming to fight for their recognition
    Relative influences (%)

                                                                          and/or rights quite independent of the occurrence
                                                                          of a life-threatening disease. Nonetheless, within
                                                                          the Western world, it is clear that these groups
                                                                          rapidly came to form a vanguard for patients rights
                                                                          with respect to AIDS. They challenged the pater-
                                                                          nalism within medicine and insisted on access and
                              1970s   1980s   1990s   2000s
                                                                          full disclosure of what was going on in pharmaceut-
Figure 1.1 The influencers of healthcare provision                        ical medicine and within academic medical politics.
                           PHARMACEUTICAL MEDICINE AS A MEDICAL SPECIALTY                                            5

Without this openness such patients had lost confi-       Table 1.1 Controlling factors in the adoption of new thera-
dence in pharmaceutical companies, in academia,           peutic agents
and in the medical and regulatory establishment.
                                                          Influences                 Controllers/`GateKeepers'
Having forced a re-evaluation and a greater respect
for patients' needs, AIDS Coalition to Unleash            Medical science            Regulatory agencies
Power (ACTUP) and others have brought patient                                        Physicians
representatives into the drug development process.                                   Health professionals
Such educated and involved patients have, in their        Healthcare providers       Politicians
turn, come to understand the scientific methodology                                  National health services/HMOs
and the requirement for the adequate testing of new                                  Insurance companies
drugs. Indeed, the requirements have consequently         Consumers                  Patient groups
become much more acceptable to patients in gen-                                      Pharmacists
eral. Nevertheless, there is no doubt that these pro-
active patient representative groups have changed
forever the role of the patient in the development of     diseases or to health areas considered today an
therapeutics and of healthcare within society.            inevitability of life or a condition for which the
   Pharmaceutical medicine is the discipline that         patient should `just take charge'. Typical examples
specializes within medicine in overseeing the pro-        will be, on the one hand, an increased focus on
cess of developing new therapeutics to improve the        quality of life or on the effects of aging (such as
standard of health and the quality of life within         cognitive dysfunction, the menopause, osteopor-
society. Inevitably, then, it was one of the first        osis, and waning immunological function, with
medical specialties to feel this change in patients'      consequent increasing vulnerability to disease)
view of the quality of their care. An integral part of    and, on the other hand, disorders such as obesity,
all progress in healthcare is evaluating the needs of     attention deficit, hyperactivity, and even anorexia/
patients and society and the gaps in the present          bulimia. As the patients or their representatives
provisions for those needs. In order to oversee           respond and `take charge', we should not be sur-
this progress, pharmaceutical medicine involves           prised to see a change in what are considered thera-
the combination of first, the medical sciences to         peutic modalities and how they are made available.
evaluate disease; second, the economic sciences to        We might expect a demand for products that do
evaluate the value with respect to costs; and third,      not need prescriptions (e.g. minerals, neutraceuti-
the ethical and social sciences to evaluate the utility   cals, and cosmeceuticals) or for patients to be able
of any new drug to patients and to society as a           to self-diagnose and use prescription drugs moved
whole.                                                    to a `pharmacy only' or to a full OTC status. Some
   As with all products, truly successful therapeutic     of these moves may well fit within one or more
agents are those that meet all the customers' needs.      governments' desire to reduce the national pharma-
In today's and tomorrow's world the concept that          ceutical bill and hence may be something that has
all that is needed is for medicines to meet the scien-    both patient and provider endorsement.
tific requirements of being effective and safe is            Those seeking to develop therapeutic products
essentially an anachronism. It is not just the scien-     will need to understand these dynamic interactions
tific factors and customers that must be satisfied.       and the consequent potential changes in one or
Table 1.1 shows that the two other critical factors       more society's approach to its healthcare. Indeed,
or influences outlined in Figure 1.1 produce many         this is another opportunity for pharmaceutical
more customers to be served.                              medicine to broaden. The speciality should cover
   As members of the public become generally more         all pharmacologically active treatments, all disease
and more informed, it is inevitable that they will        preventions and all health maintenance modalities.
want to take more of a role in deciding on their own      The objective is to maximize patient benefits and
health and how any disease they may have is to be         extend product lifecycles, as well as company sales.
treated. It is important to realize that this is likely   Clearly, pharmaceutical medicine requires an abil-
to change the demand for healthcare. Some of the          ity to read the direction society is taking and
focus will move to areas not classically considered       an understanding that, on a global basis, various

                                                                    More satisfied
                                              Better             customer (patients,
                                             products          health professionals etc) Improved quality
                                                                                              of life

                                      Development                      More product use
                                       innovation                            and
                                                                        more profits
                                                       Stronger stock
                                                         with ability
                                                to raise money/reinvest and
                                              build a self-sustaining company

Figure 1.2 The cycle that drives the pharmaceutical industry

societies can take different attitudes to how they                  the National Institutes of Health in the USA (Fig-
will regulate and/or classify a therapeutic agent.                  ures 1.3, 1.4).
However they are classified or regulated, new thera-                   Similar growth in R&D investment has been seen
peutic agents will continue to be needed, health                    outside America, e.g. in the UK. With such a mas-
benefits to deliver now, and to be potentially signifi-             sive R&D effort, the process has inevitably become
cant revenue generators for a business, allowing                    subdivided into several functional sections, the
investment in future therapeutics. This is the basic                most obvious being:
cycle (Figure 1.2) that drives the pharmaceutical
industry.                                                           . Basic chemical or structural researchÐexplor-
   The R&D process is moving forward as biomed-                       ing the genetic basic of a disease or the micro-
ical science progresses and disease processes are                     structure of a receptor or enzyme active site,
better understood. The process of developing a                        and from that, developing tailored molecules
therapeutic agent is much more than the better                        to provide specific interactions and potential
understanding of a disease leading to a new ap-                       therapeutic outcomes.
proach to its management. The process includes:                     . Preclinical research and developmentÐusing
first, state-of-the-art technical manufacturing sci-                  biological systems, up to and including animal
ences to ensure a drug substance is pure; second,                     models, to explore the causes of diseases and
appropriate and innovative pre-clinical science to                    the potential safety and efficacy of new thera-
ensure that a new chemical entity is as safe as                       peutic agents.
possible before being used by humans; third, the                    . Clinical developmentÐusing humans, both the
most sophisticated clinical evaluation method-                        healthy and those with a disease, to evaluate the
ology. This methodology must establish the effi-                      safety and efficacy of a new drug. This section
cacy and safety of a new treatment in humans and                      is itself, by convention, subdivided into three
include a multidisciplinary approach to the med-                      phases.
ical, social, and economic issues of quality of life                . Regulatory and societal developmentÐensuring
and cost±benefit. Finally, the process includes the                   that the entire development of each new thera-
business management of the social and political                       peutic is seen in the context of its need to meet
issues inherent in establishing, communicating                        governmental requirements and that the appro-
and assuring the value of the new drug within a                       priate value-added components (e.g. quality of
global economy.                                                       life, cost±benefit, evidence-based medicine, rela-
   The amount spent on R&D by the pharmaceut-                         tive competitive positioning) over and above the
ical industry has grown logarithmically over the                      basic demonstration of safety and efficacy are
last few decades and now the industry outspends                       integrated into the product's database.
                                                    PHARMACEUTICAL MEDICINE AS A MEDICAL SPECIALTY                                                            7



              R&D expenditure ($m)







                                              '82   '83   '84   '85   '86    '87   '88   '89  '90      '91     '92   '93   '94     '95   '96
                                                                                         Year                                             *

Figure 1.3 Total UK pharmaceutical R&D expenditure (including capital), 1982±1996 (*estimated for 1996). (From Centre for
Medicines Research Report, 1996)

. Post-market approval medical affairsÐthis in-                                          the level of being tested in humans, and over 80%
  volves the promotion of each product bymar-                                            of products that are tested in man never become
  keting and sales functions and the oversight of                                        licensed drugs. Of course, all the many failed re-
  this process by pharmaceutical physicians. Two                                         search and development efforts must be paid for, as
  other critical postmarketing components are                                            well as the relatively few successful projects. As
  first, continued learning about the safety and                                         Figure 1.3 shows, this can only be done from the
  efficacy of the product in normal medical prac-                                        earnings on the new treatments that are developed.
  tice, as opposed to clinical trials; and second,                                       This, and the need to return to shareholders a
  the development of new or improved uses of the                                         profit on their long-term investment in the R&D
  product as more is learned about it and as                                             process, are the basic factors in the cost of new
  medical science moves on. The former of these                                          drug. A major role of pharmaceutical medicine is
  two functions is termed `pharmacovigilance'                                            to ensure that the value of new therapies is clearly
  and the latter `product evergreening'.                                                 demonstrated so that society can see the cost±bene-
                                                                                         fit of new medicines.
   So the whole process of developing a new drug is                                         Overall, the process of moving from a research
extremely expensive and time-consuming. It is a                                          concept through development to a marketed drug
also very difficult and risky process. Indeed, the                                       and then further refining the drug's value through-
majority of initial new product leads never reach                                        out what marketing would call the product's life

                                                                                                        Process Chemistry
 Molecular Biology                                        Pharmacology & Biology                                                            Marketing

                                       Medicinal chemistry                     Clinical Pharmacology                               Regulatory

      Genomics                                                  Toxicology                               Clinical Research               Medical Affairs

 Concept testing                            Drug Design                       Clinical Validation            Product Positioning               Market

Figure 1.4 Integrated drug discovery and development. (Adapted from Taylor, 1993)
8                                           PRINCIPLES AND PRACTICE OF PHARMACEUTICAL MEDICINE



R&D expenditure ($ millions)




                                       80           85           90              95              97               00
                                                                Year                                               *

Figure 1.5 R&D expenditures, ethical pharmaceuticals, research-based pharmaceutical companies, 1980±2000. (From PhRMA
Annual Survey 1997; 2000 expenditure from Ernst and Young, 2001)

cycle involves many disciplines. It can be seen in the                 gies, such as combinatorial chemistry, genomics
terms shown in Figure 1.5. The basic responsibility                    and high-throughput screening, and computer-
for establishing and maintaining the safety and                        assisted drug design, and so in the 1990s we have
efficacy of a drug involves knowing where all of                       basic pharmaceutical discovery being carried out at
these differing functions can impact on the risks                      the molecular and disease mechanism level. As
and benefits of medicines for patients.                                such, we now have many models to evaluate and
   In the 1950s and 1960s random screening and                         have probably reversed the development paradigm
serendipity was the basis of the approach to new                       to one that Dr Stanley Crooke, the Chief Executive
drug discovery. The structure±activity relationships                   Officer of Isis, has described as `target-rich [but]
were rudimentary and used simplistic pharmaco-                         chemical-poor'.
phores and animal `models of diseases'. This ap-                          Inevitably, in today's world, where science seems
proach had essentially thousands of chemicals                          to be producing amazing advances almost weekly,
chasing a few models to hopefully find a new drug.                     the focus is on R&D and on further improvements
The 1970s and 1980s have seen the impact of recep-                     in healthcare in the future. This should not cause us
tor science. They have seen the development of                         to take our eye off the needs of today and the
protein chemistry and elucidation of many enzymes                      ability of today's medicines to be used most effect-
and cell surface structures. Finally, the 1990s have                   ively. The value of a new therapeutic agent is not
seen the impact of enabling biomolecular technolo-                     maximal at the time of its first approval. Much can
                           PHARMACEUTICAL MEDICINE AS A MEDICAL SPECIALTY                                      9

be done after market approval to ensure that a new        adequately for the care providers and/or for those
drug's utility is both fully understood and is actu-      who have to manage the healthcare resources of our
ally realized. The physicians within pharmaceutical       societies. Even when well-developed and appropri-
medicine need to oversee and lead this process.           ately used for their approved indication, many
This requires that they are trained in economics          drugs take on a new lease of life as the medical
and business as well as medicine. Indeed, some            sciences change and new therapeutic uses become
may well go on to specialized courses in those            possible; e.g. lidocaine was a very well known local
areas leading to diplomas and even university             anesthetic and was decades old when it found a new
degrees.                                                  role as an antiarrhythmic within the new context of
   The rapid advances in the biosciences and our          cardiac resuscitation and coronary care units.
gains in the understanding of diseases offer the             By the same token, as medicine moves forward,
opportunity of new benefits or uses for drugs to          the acceptability and safety of a drug can change. It
be developed after they have been marketed. Con-          is a basic axiom of pharmaceutical medicine that
sequently, there is a real and ongoing role for those     no drug can ever be considered completely safe.
in pharmaceutical medicine to follow the advances         This is true no matter how much human use data
of medical sciences and to improve the value of the       is available. For example, Phiso-Hex (hexachloro-
drugs of today within the medical and healthcare          phene) gained broad usage as a skin wash and
practices of tomorrow. This `evergreening' process        scrub to combat the spread of infection. It was
is analogous to physicians in their practice learning     used in pediatric and neonatal units in hospitals,
about a therapy and, as they come to know more            by nurses and surgeons as a scrub and was even
about the use of the treatment, and as their practice     sold over the counter as a teenage acne remedy.
dynamics change, modifying the use of that ther-          Notwithstanding all this, it became a safety issue.
apy to the maximal benefit of patients.                   This was because, as medical science advanced,
   The management of a drug on the market is a            more and more premature babies were able to sur-
professional challenge for which no medical school        vive. The skin of those babies was more permeable
trains its physicians. The overall process and skill is   than that of full-term babies, children or adults.
an important part of the training within the speci-       There was therefore a new potentially `at-risk'
ality of pharmaceutical medicine. This effort may         group. Hexachlorophene toxicity in humans was
include the issues of quality-of-life evaluations, to-    considered to have resulted and this led to the
gether with the appropriate development of evi-           product being modified or removed in many
dence-based medicine, of outcomes research and            markets worldwide.
of cost±utility sciences. All of these are techniques        The scale of the response to this issue provides a
needed within pharmaceutical medicine. Used ap-           case history that highlights another skill and
propriately, they can help not only to establish the      training required within pharmaceutical medicine,
curative value of a new medicine but also to ensure       namely crisis management. This is a very important
that the therapy gets delivered optimally.                technique which is critical in addressing substan-
   Just as is one's personal practice of medicine,        tive health issues. In the relatively recent history of
there is no more rewarding experience than the            healthcare, there have been several such issues, e.g.
optimal use of a treatment modality in a complex          Zomax, Oraflex, Tylenol tampering, toxic shock
clinical case with a successful outcome and a happy       syndrome, Reye's syndrome, the Dalcon shield,
patient; there is an equivalent reward in pharma-         contaminated blood supply, silicon implants and
ceutical medicine for a physician who: positions a        the so-called `generic drug scandal', to mention but
product to deliver the best benefit for all patients;     a few.
convinces all those delivering the care to use the           Today, as much as being a leader in R&D, it is
product; and sees a consequent real improvement           part of the role of the pharmaceutical physician to
in society's level of healthcare. In the past, many       recognize new opportunities and to be alert for any
good therapeutic agents have not been used as or          emerging evidence of potential added benefits and/
when they should be. This was not because patients        or new safety issues, as products and those of com-
in trials have not been benefited; rather, it was         petitors are used more broadly outside the confines
because the value message had not been positioned         of clinical trials.

   Many of the areas of expertise needed in pharma-      industry that are in use today. As such, they are
ceutical medicine overlap with the expertise of other    influencing the health of far more people globally
medical disciplines. The most obvious overlap has        than they ever could in the context of their own
perhaps seemed to be with clinical pharmacology.         individual clinical practice.
Indeed, clinical pharmacologists have a real interest       Any discussion of the discipline of pharmaceut-
in the R&D of the pharmaceutical industry and            ical medicine today would be incomplete without a
their training is a good one for entry into the indus-   comment on the impact of biotechnology and the
try. However, clinical pharmacology is by no means       burgeoning biotechnology revolution. This is a
the entirety of pharmaceutical medicine. Indeed,         revolution that is driven in a very different way
some pharmaceutical physicians will work in even         than that in which the pharmaceutical industry
more basic and theoretical science settings, whilst      has classically been run. The prime drivers are a
others will be in more commercial settings. Of           multitude of small venture capital companies
course, many within the specialty can and do             which are espousing the very cutting edges of re-
focus on the development of disease models and           search in biologics, genetics, and technology. They
the evaluation of new chemical entities in these         are largely managed by a combination of bioscien-
diseases. The most modern methods in such areas          tists and financiers. In this context the role of
are vital to the successful development of new           pharmaceutical medicine takes on its most extreme
drugs, and the continued and continuous inter-           variants. At one end are physician/scientists, who
action between the industry and academia is abso-        are the research brain of the venture, and at the
lutely necessary. Indeed, the distinction between        other end are physicians/businessmen, who are the
academia and pharmaceutical medicine is becom-           money-raising voice of the venture. In either of
ing blurred. The pharmaceutical industry R&D             these settings, pharmaceutical medicine is needed
effort is now leading to Nobel prizes being awarded      and the specialist will apply all of the training
to those in the industry for pioneering work on          components that, as I have already indicated, com-
subjects as diverse as prostaglandins, anti-infec-       pose this new discipline. The biotechnology indus-
tives, and pharmacological receptors such as the         try is carrying forward some of the best and
histamine and the b-adrenergic receptor. The direct      brightest projects of the world's leading academic
interaction within a company between those in-           institutions. It is moving pure research concepts
volved in basic research on receptors, active sites      through applied research into development and
or genetic code reading sites; those synthesizing new    finally to the production of remarkable new thera-
molecules; and those testing them in the clinic, leads   peutic products. This industry has already created
to the potential for a very fruitful research effort.    two or three new companies of substance, with
Naturally, the industry as a prime inventor has the      sales of over $1 billion/year and a capitalization
opportunity to carry out seminal work with entirely      measured in billions. More than these obvious
unique concepts, even if many of them do not             and huge successes, the industry has spawned liter-
become therapies for humans. The human is a              ally thousands of venture capital efforts and new
unique animal which can, and does, exhibit unique        companies developing drugs, devices, diagnostics
responses to a new chemical entity. No pre-clinical      and all manner of medical technologies. Amaz-
work can be entirely predictive of a successful re-      ingly, this is an industry which has come into
sponse in the clinic, and there can, in the end, be no   being in the last decade or two. Like the PC and
substitute for human testing. Some products fail         software industry, it is revolutionizing society's ap-
because of safety problems specific to man, and          proach to new product development and even to
some because the early promise of efficacy in            what a new therapeutic agent actually is. Already,
model systems is not realized in man.                    companies are finding that the major transition
   Those who join this new specialty may come from       points in the therapeutic product development pro-
many medical backgrounds and can well spend              cess, from molecular to biochemical system, to cel-
much of their time doing things other than pharma-       lular system, to organ model, to intact organism, to
cology. In a very real way, those in pharmaceutical      mammalian model, to humans, are all real water-
medicine are practicing medicine. They are respon-       sheds. Pharmaceutical medicine provides the re-
sible for the products of the pharmaceutical             quired understanding of each of these processes
                          PHARMACEUTICAL MEDICINE AS A MEDICAL SPECIALTY                                           11

and particularly of the transition points. In a very    function in the twenty-first century. The speciality
real sense, the success of these emerging companies     lies at the conjunction of changing societal needs
will be determined by the quality of their pharma-      for healthcare, the burgeoning biosciences and the
ceutical medicine efforts.                              understandings of how to provide improved qual-
   The new discipline of pharmaceutical medicine is     ity of life and cost±utility for patients today. The
a specialty which has only very recently become         expertise it contains and provides includes basic
recognized in its own right as a specialty within       sciences such as chemistry and mathematics, ap-
medicine. Indeed, the Faculty of Pharmaceutical         plied sciences such as engineering, economics and
Medicine of the Royal College of Physicians was         business, biological sciences such as pharmacology
only founded in 1990 in the UK and the Academy          and toxicology, and the medical sciences from
in the USA even more recently in 1994. Like many        paediatrics to geriatrics and from family medicine
new ventures, this new medical speciality is not        to the individual subspecialties. As such, pharma-
seen by all today as one of the premiere medical        ceutical medicine is one of the most challenging,
roles. However, there is a growing involvement of       exciting and rewarding areas of medicine. It is a
academics within the pharmaceutical industry and        career for those who wish to be in the vanguard of
Nobel prize-winning work is being done within the       research on multiple fronts.
industry. Furthermore, there is a growing under-
standing within academia that in the past someone
else was capitalizing on their intellectual endeav-                           REFERENCES
ours, so we are seeing more medical and bioscience
academics patenting their discoveries and going         Centre for Medicines Research (1996). UK Pharmaceutical R & D
into business. As this progress continues, the two        expenditure 1982±1986. Monograph, London.
                                                        Taylor JB (1993) In: D'Arcy RF et al (Eds) Textbook of Pharma-
disciplines of research and business are coming to        ceutical Medicine. Queen's University: Belfast.
realize that neither can do the other's work.           Erast & Young (2001). Biotechnology Annual Report. http://
Pharmaceutical medicine is the natural common   
pathway and the integrating specialty which will        Pharmaceutical Research Manufacturers Association (1997)
                                                          PhRMA Annual Report 1999±2000. Monograph. Washington
fill this need and will deliver the healthcare ad-
vances of the future. If this is so, then pharmaceut-
ical medicine will become a leadership medical
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

  What Pharmaceutical Medicine Is, and Who
                  Does It
                  William Wardell1, Susan Toland1, and Anthony W. Fox2
       Wardell Associates International, Princeton, NJ, USA, and2 EBD Group Inc, Carlsbad, CA, USA

Pharmaceutical medicine deals with the discovery,                  Pharmaceutical medicine is a young specialty, in
development, evaluation, and monitoring of drugs                its organized form. It is true that physicians have
and devices, as well as with medical aspects of their           worked with pharmaceutical companies for many
commercial promotion. Although the main focus                   decades, but during the first two-thirds of the twen-
of this chapter is on the activities of pharmaceutical          tieth century they were often viewed as either a
physicians in pharmaceutical companies, it should               necessary evil or a window dressing for respectabil-
be remembered that other arenas occupied by this                ity. In the early 1970s, the British Association of
discipline include regulatory authorities, univer-              Pharmaceutical Physicians (BrAPP) was formed.
sities, and clinical investigators in their clinical            In 1975, BrAPP introduced a 2 year program of
practices. Overlapping disciplines include clinical             part-time study, and this may be viewed as the
pharmacology, pharmacoeconomics, and biostatis-                 beginning of organized pharmaceutical medicine.
tics. In the USA, it could be further argued that               The Royal Colleges of Physicians (RCP; London,
those who operate managed care organizations                    Edinburgh, and Glasgow) then promptly sup-
also require many of the same skills as pharmaceut-             ported this initiative by providing examinations
ical physicians. In most countries, public health               for a diploma in pharmaceutical medicine, based
physicians and pharmaceutical physicians share at               upon the BrAPP syllabus.
least some concerns and methodologies. If there is                 Elsewhere, many countries are now following
any specialty that has been practicing evidence-                where the UK has led. In 1999, Belgium introduced
based medicine for decades, then it is pharmaceut-              a diploma modeled closely on that of the RCP. At
ical medicine (it is nice to watch the rest of the              the time of writing, Switzerland will probably next
profession catching up with us!). However, it is                to introduce its diploma. Elsewhere, the American
typically within pharmaceutical companies where                 Academy of Pharmaceutical Physicians (chartered
one finds the full spectrum of pharmaceutical medi-             in 1994) has become the largest single group of
cine being practiced.                                           doctors organized in this specialty, and its Board
   Pharmaceutical medicine is a medical specialty in            has approved an educational syllabus that is also
which both physicians and non-physicians take                   compatible with the RCP model; learning resources
part. This is not unusual. For example, various med-            for this American syllabus are currently being mar-
ical specialties require specialist nurses, nuclear             shaled.
medicine requires physicists, pathology requires                   The large number of national associations
histologists, and venereology requires contact                  of pharmaceutical physicians worldwide are co-
tracers. In our case, pharmaceutical medicine re-               ordinated by the International Federation of Asso-
quires pharmacists, clinical research associates, stat-         ciations of Pharmaceutical Physicians (IFAPP).
isticians, administrators, and financiers. Interaction          This Federation concerns itself with various global
with a diverse set of other types of professional can           aspects of the specialty, e.g. the International
be one of the most rewarding (and educational)                  Conference on Harmonization. A conference on
aspects of a career in pharmaceutical medicine.                 pharmaceutical medicine is held every 2 years,

being co-sponsored by IFAPP and the national             Table 2.1 Some sub-specializations in pharmaceutical
association where the conference takes place.            medicine.
   Given the global, and cross-professional, aspects
of pharmaceutical medicine, what, then, can a med-
ically-trained person specifically contribute? A         Biostatistics
degree in medicine, and some years as a junior           Clinical pharmacology
doctor, require training that is of unusual breadth      Clinical trialist
among academic pursuits. This breadth of training,       Consulting
perhaps fairly termed a `jack-of-all-trades but
master of none', leads to the capability to contrib-     Genomics
ute thought to the overall scope of development          Industrial intelligence
programs. An important part of interacting with          Informatics
the diversity of other professionals is to be able to    Information technology
see, and explain when necessary, how one aspect          Legal affairs
can impact on others. Furthermore, in drug devel-        Licensing
opment and postmarketing surveillance, it is the         Lobbying and Politics
physician who is most likely to be able to estimate      Marketing
the clinical hazards and anticipate the clinical con-    Medical monitor
                                                         Medical writing
dition of the patients (or normal subjects) who are      Patent law
exposed to the drugs. Most important of all, it is the   Pharmacology
physician who is the guardian of the patient from an     Pharmacoeconomics
ethical standpoint.                                      Pharmacoepidemiology
                                                         Regulation (governmental)
                                                         Regulatory affairs
             SUBSPECIALIZATION                           Toxicology

There are any number of ro that pharmaceutical
physicians can play (see Table 2.1). Later chapters
in this book consider drug discovery, clinical trials,    WHY DO CLINICIANS JOIN INDUSTRY?
regulatory affairs, marketing, etc. Suffice it to say
that pharmaceutical physicians are found in all          Industry, or government employment, are not for
these subspecialties. They are also found at the         everyone. Doctors, pharmacists, nurses and others
next higher level of complexity, the coordination        who enjoy diversity should be especially attracted
and leadership of all these individual activities,       to this specialty. Versatility and adaptability, vice
using integrated development plans. Physicians           versa, are required for a successful career in this
usually find it easy to move between such subspe-        discipline.
cialties and, indeed, between different companies:          Another personality trait that must be considered
this is further evidence of the value of the breadth     is the capability to cooperate and listen. The typical
of training in medicine.                                 clinician is usually in sole charge of the patient, and
   The number of subspecializations has increased        needs brook little disagreement in decision-making
in recent years. Examples include the recent             from his/her staff. The bombastic surgeon is an
tremendous advances in pharmacoeconomics                 international phenomenon; the pharmaceutical
and information science. At least one pharmaceut-        physician has to be the exact opposite in his/her
ical physician is now both a part-time regulator and     approach. The really successful doctor in industry
a part-time private barrister in the UK. The in-         will be someone who actually enjoys receiving opin-
creasing interaction between pharmaceutical com-         ion from people who are not medically qualified.
panies and health maintenance organizations in the          For pharmaceutical physicians involved in drug
USA has led some pharmaceutical physicians into          development (a large subset of the specialty), there
the worlds of finance, contract, and epidemi-            are also product-related satisfactions. These can be
ology.                                                   relatively vicarious (e.g. a reduction in spontaneous
                        WHAT PHARMACEUTICAL MEDICINE IS, AND WHO DOES IT                                              15

adverse event report frequency after a labeling          cine as a career. Lest one suspect that this was
change). More directly, occasionally, we have            seeking the opinions of the already converted, it
friends or relatives who benefit from the drugs          should be pointed out that other specialist colleges
that we develop. One of the authors of this chapter      have conducted similar surveys. No other medical
has been hugged to the point of assault by a com-        specialty in the USA contains such a large propor-
plete stranger in Georgia (USA), a patient who           tion of doctors with such positive views.
judged him responsible for `curing' her migraine.
Neither protests that the reality is teamwork, nor
perfume-induced sneezing, deterred this extra-
verted Southern lady!                                                     FURTHER READING
   Some join the industry for only a short period,
                                                         Berde B (1985) Physicians as employees of the pharmaceutical
and then decide that they would prefer to return to        industry. Eur J Clin Pharmacol 28: 363±5.
their original clinical callings. This, too, can be of   Bootman JL, Townsend RJ, McGhan WF (eds) (1996) Prin-
professional benefit. One's clinical skills do not         ciples of Pharmacoeconomics, 2nd edn. Harvey Whitney
decline much in the first 2 or 3 years away from           Books: Cincinatti, OH.
                                                         Dollery C (1994) Medicine and the pharmacological revolution
patients, and they can, in any case, be maintained
                                                           (the Harveian oration of 1993). J R Coll Physicians (Lond)
with part-time or locum tenens clinical employment         28: 59±69.
while one holds a position in a pharmaceutical           Dziewanowska ZE (1990) Globalization of the pharmaceutical
company. Upon leaving industry, it is likely that          industry; opportunities for physicians in clinical research.
the physician will take back to the clinic some new        J Clin Pharmacol 30: 890±92.
                                                         Fox AW (2001) What is pharmaceutical medicine? Clin Res 1:
skills, e.g. better management techniques and how
to approach clinical data with scientific scepticism;    Gabbay FJ (1987) Consolidating the discipline of pharmaceut-
this experience is not usually available in the ordin-     ical medicine in the United Kingdom. In Burley D, Haasard
ary clinical situation. Such a physician will also         C, Mullinger B (eds), The Focus for Pharmaceutical Know-
have inside knowledge of how to increase the prob-         ledge: The Proceedings of the Sixth International Meeting of
                                                           Pharmaceutical Physicians. Macmillan: London; 286±92.
ability of industrial sponsorship of his/her clinical    Jefferis JJ (1985) The pharmaceutical industry and academic
research project! One author of this chapter, when         medicine: opportunities for physician collaboration. Circula-
approached by a doctor with trepidation or vacil-          tion 72 (suppl 1): 21±4.
lating on whether to accept an industry position,        Sampson MC (1984) Career Opportunities in Industrial Clinical
has a standard response: `Now is not make-your-            Research. In Matosen GM (ed.), The Clinical Research Pro-
                                                           cess in the Pharmaceutical Industry. Marcel Dekker: New
mind-up time. That will be in 2 or 3 years because         York; 479±505.
then you must probably decide whether to stay in         Shaw L (1991) A misunderstood specialty: a survey of physicians
the industry'.                                             in the pharmaceutical industry. J Clin Pharmacol 31:
   In 2000, the American Academy of Pharmaceut-            419±22.
ical Physicians surveyed its members. One of the         Spilker B (1989) Career opportunities for physicians in the
                                                           pharmaceutical industry. J Clin Pharmacol 29: 1069±76.
questions asked about personal fulfillment, and          Young MD (1990) Globalization of the pharmaceutical indus-
more than 90% of respondents indicated high                try: the physician's role in optimizing drug use. J Clin Phar-
degrees of satisfaction with pharmaceutical medi-          macol 30: 990±93.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

         Competency-based Training System for
               Clinical Research Staff
                                                   Jay D. Miller
                                      Amgen Inc., Thousand Oaks, California

The pharmaceutical industry is a highly regulated               education programs to comply with current GCP
industry in which many of the activities and tasks              requirements.
performed by staff are defined by regulations and
guidelines issued by various regulatory authorities
around the world. The training requirements for                       WHAT IS A COMPETENCY-BASED
sponsor companies can therefore be fairly well de-                       TRAINING PROGRAM?
fined. In addition, international initiatives by regu-
latory authorities and trade organizations have                 Few people come to the pharmaceutical industry
further defined the role of staff involved in clinical          from academia with the requisite knowledge and
research.                                                       skills necessary to plan, conduct, and report clinical
   The International Conference on Harmoniza-                   research to regulatory authority standards. This
tion (ICH) guideline for Good Clinical Practice                 knowledge and skill usually needs to be provided
(GCP; 1997), for example, describes a minimum                   to the new staff by the way of in-house training.
standard for the ethical and scientific standards                  One approach to education and training in the
for designing, conducting, and reporting clinical               industry is what is called `competency-based
research. The ICH GCP guideline is the unified                  training'. A competency is a skill, knowledge or
standard for the European Union (EU), Japan                     behavior required to undertake effectively the
and the USA to facilitate mutual acceptance of                  tasks and responsibilities for which an individual
clinical data. The ICH GCP guideline, together                  is responsible.
with other ICH guidelines, provides an operational                 A competency-based education and training
definition of the core competencies needed by clin-             system (CBETS) details the essential knowledge
ical staff to conduct world-class clinical research.            and skills needed by sponsor's staff to complete the
   One of the principles of ICH GCP is that `each               requirements of GCP. The concept of a CBETS is
individual involved in conducting a trial should be             different from traditional educational and training
qualified by education, training, and experience to             approaches. Traditional approaches tend to ad-
perform his or her respective task(s)'. Specifically            dress the training needs of individuals based on
regarding the selection and qualifications of moni-             their job descriptions, e.g. within a sponsor com-
tors, the ICH GCP guideline states that `monitors               pany, a monitor will receive training on how to
should be appropriately trained and should have                 monitor a clinical trial, and a physician will receive
the scientific and/or clinical knowledge needed to              training in protocol development. In this trad-
monitor the trial adequately'. Most major pharma-               itional education and training model, the required
ceutical firms have always had varying degrees of               tasks are functionally defined. The monitor may
in-house education and training for staff, supple-              not learn much about preparing protocols and the
mented (as appropriate) by external workshops,                  physician may not learn much about monitoring.
courses and training meetings. The ICH GCP                      However, each may be intimately involved in both
guideline formalizes the requirements for credible              tasks.

   The CBETS asks what tasks the sponsor needs          urgent needs in providing education and training
to have done to meet its drug development goals         around the knowledge and skills.
and objectives. The primary tasks of clinical             While many readers will be thoroughly familiar
research and good clinical practice can be described    with the tasks described below and described in
rather precisely. Once one knows what the major         more detail throughout the chapters of this book,
tasks are and what activities are needed to accom-      a brief explanation of the task is provided, together
plish these tasks, one can define the knowledge         with the success factor in completing the task.
and skills needed by staff to complete the tasks
and, finally, what education and training should
be provided to understand the knowledge and                     General Clinical Competencies
   When the tasks and activities are fully defined, a   Understanding of the drug development process
CBETS will ask who is going to do these tasks and       All new clinical staff need to understand the overall
how competent (e.g. expert, fundamental skills)         drug development process. Before new investiga-
they need to be to complete the tasks. In the           tional products can be introduced into man, exten-
example provided above, it is useful for the phys-      sive preclinical and toxicological studies are
ician to have a fundamental knowledge of the            performed. Staff who will be responsible for the
monitoring process, even though he she will not         clinical portion of drug development need to have
be performing the tasks. The physician may, how-        an appreciation of the work that has been under-
ever, be supervising the monitors. It is appropriate    taken to progress the compound through to the
for the monitor to receive advanced training in the     clinical phases.
requirements of monitoring, since this is one of his/      This includes understanding the vision, mission
her major functions. In terms of protocol develop-      and objectives of the sponsor's clinical develop-
ment, the physician and monitor each need expert        ment strategy. Most sponsors have a unique inves-
competency to develop the protocol, since the           tigational product development system. Being
sponsor may be investing hundreds or millions of        familiar with this system is important in under-
dollars in the program. The monitor only needs          standing the sponsor's decision-making approach,
fundamental skills in protocols, since his/her re-      its internal milestones, and its methods for
sponsibility may be limited to implementing the         budgeting and resourcing clinical programs. Indi-
protocol at the clinical site.                          viduals new to industry should identify and under-
   The CBETS can apply to behavioral and                stand the function of the major departments
management training as well as technical training.      comprising the clinical research and development
Education and training programs in the pharma-          process, as well as understanding the operation of
ceutical industry can be designed to provide the        sponsor management bodies.
competencies necessary to prevent or remove obs-           To gain this knowledge, new staff members
tacles to staff performance. These obstacles can be     should attend the sponsor's orientation program
well defined in advance in a CBETS.                     on drug development and the appropriate Pharma-
                                                        ceutical Education and Research Institute Inc.
                                                        (PERI) or Drug Information Association (DIA)
      COMPETENCY-BASED TRAINING                         overview courses on investigational drug develop-
      PROGRAM FOR CLINICAL STAFF                        ment. There is considerable literature available
                                                        that discusses the drug development process, such
The following is a description of the typical essen-    as the Guide to Clinical Trials by Dr Bert Spilker.
tial competencies needed to plan, conduct, and          Regulatory authorities provide important guide-
report a clinical research program in a regulated       lines on registration expectations. In addition,
environment. Each competency is described, as           many sponsors have internal documentation ex-
well as the knowledge and skills a sponsor's repre-     plaining the company's systems and processes.
sentative would need to be successful in completing     Senior-level staff can also attend the advanced
the task. Judgment can then be applied to the most      course on international investigational product de-
                 COMPETENCY-BASED TRAINING SYSTEM FOR CLINICAL RESEARCH STAFF                               19

velopment and regulatory issues sponsored by              Competencies Associated with Planning
Tufts University.                                                  Clinical Research

     Understanding the US FDA regulations               Conceptualization and Development of Clinical
                                                          Research and Development Plans (CRDPs)
The Food and Drug Administration (FDA) in the
USA remains the dominant regulatory authority in        Developing an international CRDP to answer
the world. An understanding of the FDA regula-          questions defined by the investigational product
tory structure, operations and functions is very        target profile is a key activity of senior-level indus-
important to individuals new to the pharmaceutical      try personnel. This competency requires an under-
industry or new to the industry within the USA.         standing of toxicology and clinical pharmacology
The different approaches of the Drug and Biologics      to identify clinical target profile criteria. The
Divisions should be understood. Staff need to iden-     CRDP defines the critical path for the clinical pro-
tify and understand FDA investigational drug IND        gram and the clinical budget. The CRDP also de-
and biologic regulations and FDA guidelines and         fines investigational drug development assessment
regulations.                                            and decision points, and the project resource (per-
   Knowledge and skills are required for the spe-       sonnel and budget) estimates.
cific regulations governing drug development in the        CRDPs will cover:
USA. These include Title 21, Code of Federal
Regulations Parts 50,56,312 and 314.                    . Preparing the clinical section of IND/CTX sub-
   Senior-level staff need to understand how the          mission.
communication process works with the FDA, for           . Preparing clinical reports needed to support
example, End-of-Phase II Meetings, IND Annual             IND/CTX submissions.
Report, FDA Advisory Committee meetings, Pre-           . Clinical research and scientific methodology.
New Drug or Biologic Licence Applications               . Phase I studies.
(NDA/BLA) Meetings, clinical holds, IND termin-         . Phase II studies.
ation, and FDA audits.                                  . Phase III studies.
                                                        . Phase IV studies and safety surveillance studies.
                                                        . Pilot efficacy studies.
          Understanding Good Clinical                   . Pharmacokinetic and bioavailability studies.
                Practices (GCP)                         . Dose-ranging studies.
Understanding the responsibilities and obligations      . Dose-titration studies.
of sponsors in terms of GCP is fundamental know-        . Registration studies.
ledge, essential to conduct clinical research. Cur-     . Marketing and safety surveillance studies.
rently, most pharmaceutical firms reference the
ICH GCP guideline as the minimum standard for           The goal of these plans is to provide an efficient
conducting clinical trials.                             NDA/BLA with the minimum studies needed for
   The responsibilities and obligations include         registration and approval in the world markets.
knowledge of the elements of informed consent,          The medical, scientific, regulatory and marketing
the role and responsibilities of institutional review   opinions must be weighed and balanced in the
boards/independent ethics committees (IRB/IEC)          plans.
and the importance of clinical study quality assur-
                                                                  Understand and Conceptualize
   All sponsor staff have a specific and direct re-
                                                                     Clinical Study Design
sponsibility for the safety and welfare of subjects
participating in clinical trials. A full and complete   To successfully create a CRDP, the individual must
understanding of GCP is requisite for all sponsor       know the basic concepts of research design and
staff. Most sponsors provide internal training on       statistics, the concepts of clinical research and
these issues. There are excellent PERI or DIA           investigational drug development; possess an in-
overview courses covering GCP.                          depth understanding of the concepts of clinical

pharmacology, pharmacokinetics, pharmacody-             ing a protocol include: defining clear protocol ob-
namics, toxicology, state-of-the-art therapeutic        jectives, identifying primary efficacy and safety
medicine and methodology, and FDA/EU/ICH                parameters and appropriate subject selection cri-
therapeutic research guidelines and regulatory          teria, and identifying correct dosages and route.
issues; and understand basic concepts of project           To prepare appropriate protocols, staff must
planning and scheduling.                                understand: research design and statistical infer-
                                                        ences for clinical research; state-of-the-art research
                                                        designs and trials; therapeutic area guidelines;
 Preparation of the Investigator's Brochure (IB)
                                                        GCP; regulatory requirements, guidelines and
The IB is a summary of clinical and preclinical         country-specific issues; national and international
data on the investigational product which is rele-      medical practices; sponsor protocol review and ap-
vant to the study of the investigational product in     proval procedures; and possess in-depth investiga-
human subjects and the investigator's assessment        tional product-disease knowledge.
of risk in participating in the study. The sponsor         Clinical protocols are the building blocks of the
compiles clinical information for the preparation       clinical research plan and the NDA/BLA. Proto-
of the IB.                                              cols specify the conditions that permit and lead to
   The preparation of an IB may be performed by         meaningful and credible results in clinical pro-
clinical staff or a medical writing group. The activ-   grams. Operationally, protocols provide a written
ities included in preparing the IB include: coordin-    agreement between the sponsor and the investiga-
ation of the compilation of clinical and preclinical    tor on how the trial is going to be conducted. This
data from contributing departments (e.g. clinical       agreement allows the sponsor to ensure that the
pharmacology, toxicology); describing the phys-         study will be done to the highest ethical and med-
ical, chemical and pharmaceutical properties and        ical standards and that the quality of the data can
formulation; preparing a clear, concise summary of      relied upon as credible and accurate.
the information relating to the safety and effective-      All clinical protocols and supporting documents
ness of the investigational product; providing a        are reviewed and approved internally by a group of
detailed description of possible risks and benefits     senior clinical research managers. This group
of the investigational product; and a clear rationale   assesses the overall study design and ability of the
for the dosage and dosing interval.                     study to meet its objectives, as well as the quantity
   To prepare an IB, the sponsor's representative       and quality of the data. In addition, the group
must understand: the fundamental purpose and            reviews the procedures for the safety and welfare
uses of the IB; the basic format and content of         of the subjects, to ensure compliance to GCP and
sponsor IBs; the clinical pharmacology and toxicol-     ethical principles. The quality of a clinical protocol
ogy findings; the investigational product±disease       can be assessed by how well the elements of the
relationships; the international regulatory require-    protocol are prepared. The elements of clinical
ments governing IBs; and the indications and safety     protocols are described in Table 3.1.
profile of the investigational product.
                                                                 Table 3.1 Elements of clinical protocols

  Design and Preparation of Clinical Protocols                   Background and rationale
                                                                 Study objectives
The clinical protocol describes the objectives,                  Experimental design and methods
design, methodology, statistical considerations,                 Schedule of assessments
and organization of the trial. The sponsor is usually            Subjection selection criteria
responsible for developing the protocol in industry-             Screening procedures for entry
                                                                 Study parameters
sponsored clinical trials. However, internal and                 Trial medication
external content experts are frequently consulted.               Premature withdrawal
Protocols must be written ensuring medical sound-                Subject replacement policy
ness and clinical practicality.                                  Criteria for excluding data
                                                                 Statistical analysis plans
  Often the sponsor uses a template to complete
the sections of the protocol. The tasks of develop-
                  COMPETENCY-BASED TRAINING SYSTEM FOR CLINICAL RESEARCH STAFF                                21

   The extent of a background section will vary              The efficacy and safety parameters describe how
with the drug's stage of development. New clinical        and when the variables are going to be recorded,
data not already included in the IB should be em-         usually in relation to drug administration and
phasized. The rationale provides a concise state-         follow-up periods. How adverse events are managed
ment of the reasons for conducting the study and          and recorded are particularly important to the spon-
the basis for the dosage selection and duration that      sor and regulatory authorities. Protocol authors
will be used in the trial. Quality protocols should       should ensure that the study defines the criteria for
target relevant information in the background and         success or failure of treatment. End-points should
convincing rationale for the study.                       be clear and defined. Since many clinical phenom-
   Every protocol must state a primary, quantifi-         ena are open to interpretation, protocols should
able, study objective. Secondary objectives should        provide definitions of variables and time windows
be limited in scope and related to the primary            for their collection. If the assessments are purely
question. Objectives must be specific and capable         subjective, then methods to prevent observer bias
of answering a key clinical question required by the      (so-called ``observer truing'') must be employed.
clinical research plan.                                   Addressing these issues will improve the quality
   The study design is an important element in            and meaningfulness of the results of the study.
assessment of quality protocols. The overall                 The description of the management of trial medi-
purpose of the study design is to reduce the vari-        cation is often a source of confusion. Protocols
ability or bias inherent in all research. Good study      must include clear directions for dosing intervals
design will always address control methods that           and adjustments. Since patients will never follow a
reduce experimental bias. These control methods           protocol precisely in all cases, provisions for miss-
will often include treatment blinding, randomiza-         ing doses or ``what-if ?'' situations should be antici-
tion and between- or within-patient study designs.        pated. Good protocols always include, in addition,
The schedule of assessments describes a schedule of       adequate compliance checks of drug consumption
time and events and provides a complete profile           by the subjects of the study.
of the overall trial design. Good quality schedule           Protocols should predetermine how subjects will
of assessments sections also include acceptable           be replaced following dropping out of the study.
time windows around the variables being col-              This is important, because the means by which
lected.                                                   subjects are replaced can adversely affect the stat-
   The inclusion and exclusion criteria are de-           istical analysis. Similarly, a decision concerning the
scribed in the subject selection part of the protocol.    conditions under which a subject would not be
To a large extent, the success or failure of a par-       evaluable must be stated explicitly before the
ticular clinical trial can often be traced back to how    study starts. This is intended to minimize inten-
well the criteria were developed. Protocol authors        tional or unintentional data manipulation.
strive to include the most appropriate patient
population to satisfy the study objective and still
                                                                Design of the Format and Content of
include those kinds of patients that will ultimately
                                                                     Case Report Forms (CRFs)
receive the drug. Therefore, selection criteria can be
unreasonable and unnecessary in some cases and            The CRF is the document used to record all of the
vague and non-specific in other cases. The manage-        protocol-specified data used to describe individual
ment of concomitant medications is particularly           subject results. Many sponsors use standard
problematic. The protocol must attempt to define          modules to prepare the CRF.
those medications that are permitted for intercur-          To prepare successful CRFs, the sponsor's staff
rent illnesses and those that are prohibited, since       must know: typical clinical practices; therapeutic
they will interfere with the interpretation of the test   conventions; investigator and staff needs; data
medication. While there are no easy answers, qual-        management and analysis plans; project-specific
ity protocols are able to justify with some precision     definitions and procedures; CRF completion prob-
the rationale for each inclusion and criteria. How        lem areas; remote-data entry and review; and ap-
these criteria are applied is handled in the screening    proval procedures for CRFs. Ideally, CRFs should
for study entry section.                                  be pre-tested with internal and external experts.

   The quality of a clinical trial can be influenced by    necessary to estimate, from the CRDP, the bulk
how well the CRF is designed. If the investigator's        investigational product supply needs.
staff cannot enter the protocol data as required, the         To handle drug supplies successfully, the spon-
sponsor will have a considerable challenge in trying       sor's representative must know: the procedures for
to interpret the results.                                  ordering bulk investigational product supplies;
   There are a number of design principles that            models for bulk investigational product quantity
facilitate the use of CRFs in clinical trials. These       estimation; investigational product packaging
principles include the concepts of standardization         time frames; protocol-specific and country-specific
and minimization. The sponsor standardizes the             requirements for packaging and shipping investiga-
design of CRFs in one consistent international             tional product supplies; procedures for pack-
format. This permits uniform databases, consist-           aging international investigational product
ency in collection and more rapid data entry. In           supplies; investigational product supply tracking
addition, standardization facilitates the monitoring       systems; investigational product ordering and
process and therefore increases accuracy of the            packaging processes; general investigational prod-
data. While efficiency is an important variable in         uct formulation and packaging processes and con-
the design process, the systems must also be suffi-        figurations; protocol design; randomization
ciently flexible to account for the variances between      procedures; and investigational product dispensing
projects. Finally, an important principle of both          and accountability.
protocol and CRF design is to collect only the
data needed to satisfy the objectives of the proto-
                                                                Identification and Selection of Clinical
col. The inherent temptation to collect more data
                                                                Investigators for Study Placement and
must be resisted.
                                                                Conducting Prestudy Evaluation Visits
   There are several CRF design characteristics
which define quality CRFs. Some of these in-               Selecting investigators. The proper selection of
clude:                                                     clinical investigators is one of the key success factors
                                                           for any clinical program. The principal investigator
. Limiting the amount of space for free text.              has the primary responsibility for the success of
. Providing instructions on the CRF for its com-           the trial. His/her leadership and direction of co-
  pletion.                                                 investigators and study staff is critical in performing
. Consistent layout of information within the              the requirements of today's trials. Time spent in
  CRF.                                                     learning who are the best investigators is well
. Simple, unambiguous language.                            spent and pays significant dividends in the end.
. Collecting only raw data, letting the computer              To successfully identify and select clinical investi-
  do transformation calculations.                          gators, the sponsor's representatives need to: iden-
. Intensive monitor training in the use of the             tify internal and external sources for potential
  CRFs.                                                    investigators; define investigator selection criteria,
. Use of the project-specific procedure manual to          protocol requirements, expected cost of the study,
  document conventions and CRF decisions.                  and investigator and facility qualifications; inter-
                                                           view potential investigators; and, finally, schedule
High-quality CRF design is probably the cheapest           and conduct prestudy site evaluation visits.
investment in big returns on a clinical trial.                The clinical team has an important role in deter-
                                                           mining the quality selection of clinical investiga-
                                                           tors. Selection criteria will be based upon the
     Packaging and Labeling of Investigational
                                                           needs of the clinical research and development
                                                           plan (CRDP) and the individual protocols. Quality
The investigational product is the active ingredient       investigators can be identified by:
or placebo being tested in a clinical trial. Forecast-
ing investigational drug supplies is important, in         . Previous clinical research experience.
that it must be done well in advance of the start          . Previous performance on sponsor and other
date of the clinical trial. To make this forecast, it is     company trials.
                   COMPETENCY-BASED TRAINING SYSTEM FOR CLINICAL RESEARCH STAFF                                          23

        Table 3.2 Sources of quality investigators       Table 3.3 Prestudy visit questions

        Clinical leaders/therapeutic area heads          How will the protocol specifically operate at the prospective
        Country company heads                             center?
        Consultants                                      How will informed consent be obtained? By whom?
        Colleague recommendations                        How will source documents be managed?
        Investigator recommendations                     How will adverse events be handled and followed up
        Scientific and medical literature                 (serious and non-serious events)?
        Physician directories                            How many studies is the investigator conducting
        Speakers at professional meetings                 currently?

. Their reputation among peers and the quality           drug supplies. The nature and form of informed
  of their publications.                                 consent is reviewed. In these discussions, the moni-
. The experience and training of their support           tor is attempting to identify aspects of the study
  staff.                                                 which present difficulties or problems for the inves-
. The quality and reputation of their research           tigator. Quality investigators usually have clear
  facilities.                                            understanding and strategy for the above activities.
                                                         Examples of the questions that require answering
Potential sources of quality investigators are shown     during prestudy visits are shown in Table 3.3.
in Table 3.2.                                               Some objective measure of the availability of the
   Many physicians may need to be considered             correct patient population is important during a
before the best investigators can be identified. Pre-    prestudy visit. This can often be best accomplished
liminary contact should be done by telephone.            through a chart or hospital census review by the
Only those investigators who satisfy the primary         monitor. The time spent doing this aspect of a
selection criteria need to be visited.                   clinical trial will invariability result in better and
                                                         more timely results in clinical programs.
Prestudy visits. The purpose of the prestudy visit          Assuming that the outcome of the prestudy
is to evaluate the investigator's interest and ability   visit(s) is successful, the sponsor's representative
to conduct the study to the required sponsor             will need to develop and negotiate study contracts
standards. Special attention is paid to the quality      and secure essential documents.
of the investigator's staff and facilities, as well as
to the availability of the required patient popula-
tion. In conducting the prestudy site evaluation
                                                          Competencies Associated with Conducting
visit, the sponsor's representative determines
whether or not the investigator is qualified by                      Clinical Research
training and experience to conduct the trial.
                                                                       Conducting Study Initiation
   The prestudy visit is a professional exchange of
information. The investigator is informed of the         The study initiation visit is sometimes confused with
preclinical and clinical background of the drug.         the prestudy visit. The purpose of the study initi-
Of primary importance to the investigator is the         ation meeting is to orientate the study staff to the
rationale for use of the drug and the expected safety    requirements of the protocol. At the point of the
profile. Much can be inferred from the investiga-        study initiation visit, the study site should be fully
tor's preparation and questions about the investi-       ready to begin all aspects of the trial. The monitor
gational drug. The protocol should be explained,         must ensure that the study medication and mater-
including the requirements for the patient popula-       ials are available at the site. In addition, all essential
tion, the study design, and a description of the         documentation must be completed and available.
safety and efficacy variables.                           Key study documentation is shown in Table 3.4.
   Other aspects of the study are also discussed with       All study staff who will have direct involvement
the investigator, such as the completion of the CRF,     in the trial should participate in the study initiation
access to source documents, and management of            visit. This usually includes: the investigator and

     Table 3.4 Key Study Documentation                     that require discussion with the investigator or
                                                           staff and any safety questions or issues. The moni-
     Approved protocol and CRF                             tor compares CRFs with source documents, con-
     Informed Consent Form and Subject Information Sheet
                                                           firming that source data are consistent with CRF
     Investigator's Curriculum Vitae
     Written IRB/IEC approval                              entries, identifies all serious adverse events, re-
     Local regulatory approval                             solves previous and current CRF queries, and con-
     Signed study contract                                 firms completeness of investigator records and
     Laboratory ranges and accreditation                   files.
                                                              To be a successful monitor, the sponsor repre-
                                                           sentative should know: how to interpret hospital/
subinvestigator(s), the study coordinator or re-           clinic records/charts, laboratory tests, and inter-
search nurse, pharmacist, and laboratory personnel         pretations; has to query resolution procedures;
or specialists as needed.                                  protocol and CRF data requirements; medical no-
   During the meeting, all major points and                menclature; Serious adverse event (SAE) proced-
requirements of the protocol are reviewed and              ures; and health authority requirements. In
discussed. Procedures for subject enrollment are           addition, a monitor needs to have excellent inter-
particularly important, since this is the area which       personal communication and problem-solving
may cause the most problems for the site. During           skills.
the presentation, participants may raise important            Clinical monitoring requires clinical, interpret-
medical or logistical issues that have or have             ive, and administrative skills. The monitor needs
not been anticipated by the protocol authors.              to confirm subject selection and patient enrollment
It is important to note these concerns and com-            compliance. Quality monitoring will always in-
municate them to the protocol authors, as                  clude and confirm the following activities:
   The sponsor's representative should be compe-           . Properly obtained informed consent.
tent in the basic medical and scientific issues of the     . Adherence to the protocol procedures and in-
investigational product and protocol, know the               clusion/exclusion criteria.
target disease or symptoms, be able to train the           . Transcription of data from source docu-
investigative staff on the conduct of the study, con-        ments to the CRF that is both consistent and
firm facility capabilities, conduct the site initiation      logical.
meeting, describe adverse event reporting require-         . Identification of any safety issues, including
ments, and be able to resolve protocol issues during         serious adverse events.
and after meeting.                                         . Proper accountability and reconciliation of
                                                             drug supplies.
                                                           . Continued adequacy of facilities and staffing.
        Conducting Clinical Trial Monitoring
Clinical trial monitoring includes those activities        The frequency of clinical monitoring depends on
that ensure that the study is being conducted              the actual accrual rate of the patients. Complex
according to the protocol. Monitoring permits an           studies may need to be visited more frequently,
in-process assessment of the quality of the data           depending on the accrual rate of subjects, the
being collected. The first alert to safety issues is       amount of data, and the number of visits. Gener-
often revealed during the process of monitoring            ally, most investigators should be monitored every
the clinical trial.                                        4 ±6 weeks. Sufficient time for good monitoring
   Monitoring clinical studies involves the act of         practices should be anticipated by the monitors.
overseeing the progress of a clinical trial. Monitors      Following a monitoring visit, the monitor will
ensure that the study is conducted, recorded and           prepare a monitoring report for sponsor records
reported in accordance with the protocol. This is          and follow-up correspondence to the trial site.
accomplished by the review of CRFs on-site for             The monitor may need to plan intervention and
possible errors, inconsistencies, and omissions.           possible replacement of non-performing or non-
The monitor identifies errors and discrepancies            compliant trial centers.
                 COMPETENCY-BASED TRAINING SYSTEM FOR CLINICAL RESEARCH STAFF                                      25

         Managing Drug Accountability                     Table 3.5 Typical labeling requirements for investiga-
                                                          tional drug
The sponsor is responsible for providing the inves-
tigator with investigational drug(s). Both the spon-      Local language             Route of administration
sor and the investigator have a role in drug              Name of investigator       Dosage
accountability. The sponsor's representative in-          Study number               Dosage form
                                                          Bottle number              Quantity or volume
spects storage of investigational product supplies;
                                                          Lot number                 Storage precautions
checks study site investigational product dispens-        Drug name or code          Directions for use
ing records; checks randomization and blinding;           Manufacturer name          Note: `For Clinical Trial'
and maintains records of investigational product          Manufacturer address       Caution statement
shipments. The monitor reconciles investigational         Local affiliate name       Expiry date
product shipped, dispensed and returned; arranges
for shipment of investigational product to core
country or investigative sites; checks investiga-       supplies are being kept under the required storage
tional product supplies at site against enrollment      conditions.
and withdrawals; maintains investigational prod-           Study drug must be dispensed according to the
uct accountability records; resolves investigational    randomization schedule. Failure to do so can result
product inventory problems; implements tracking         in some of the data having to be discarded during
system for investigational product management on        statistical analysis. This issue can prove to be prob-
a study and project level; arranges for the return      lematic when a single site is studying patients at
and/or destruction of unused investigational prod-      different locations. Finally, the double-blind code
uct supplies; and ensures final reconciliation of       must not be broken, except when essential for the
investigational product supplies.                       management of adverse events. The breaking of
   GCPs require sponsors to be able to account for      treatment codes can make that patient's data un-
the drug supplies prepared and shipped to the in-       usable for efficacy analyses.
vestigator, the investigator's use of those supplies,
and the return and destruction of remaining drug
                                                          Handling Adverse Drug Experiences (ADEs)
supplies. Planning drug supplies is a detailed and
complex activity. Bulk and formulated drug re-          Safety concerns are present throughout the drug
quests must typically be made at least 6 months in      development process. From the filing of INDs,
advance of the need for those supplies. This is to      through the conduct of clinical trials, to the ap-
account for the ordering of intermediates or fin-       proval process of the NDA/BLA and the marketing
ished drug, purchasing of comparator agents, and        of the drug, safety is the primary concern of any
for quality control testing.                            clinical program.
   Drug packaging should follow as consistent a            Management of safety is a principal responsibil-
format as possible within a project and must be         ity of the sponsor monitor. The monitor has re-
identical within multicenter trials. Regulatory docu-   sponsibility for informing the investigator about
ments required for investigational drug use in the      the safety requirements of the study. This will in-
core countries must be anticipated and made avail-      clude a discussion of expected and unexpected ad-
able when needed, e.g. methods of analysis, stability   verse events, how to report adverse events should
data, customs declarations. The typical require-        they occur, and how to characterize the adverse
ments for drug labels is described in Table 3.5.        events in terms of project-specific definitions.
   Once the study is under way, the use of the             Monitors are expected to review CRF and
investigational drug must be accounted for by the       source documents with particular attention to po-
investigator's staff. Subjects should return unused     tential safety problems. On the CRF, the adverse
medication and empty containers to the investiga-       events section and laboratory results section are
tor. The amount of drug dispensed and the amount        reviewed for important findings. Often, relevant
used by the patients are compared for discrepan-        notes are made by the investigator in the comment
cies. This provides a measure of compliance by the      section of the CRF. In source documents, safety
study subjects. Monitors must also check that drug      issues may be uncovered in the progress notes of

hospital charts or the interpretative reports of vari-   obtain the data required in the trial. The study
ous diagnostic tests, e.g. chest X-rays, ECGs.           close-down visit usually occurs after the last subject
Safety problems can manifest themselves in many          has completed the trial, including any post-treat-
ways. Monitors must be alert to exaggerated              ment follow-up visits. Drug supplies should be rec-
changes from baseline with expected pharmaco-            onciled and the integrity of the double-blind
logical effects, acute and chronic effects and mul-      treatment codes should be confirmed. Any out-
tiple drug treatment reactions.                          standing queries should be resolved and docu-
   Monitors are often the first company representa-      mented.
tives to learn about an adverse event. The timeliness       Arrangements for retaining source data should
of reporting the event to management is important        be confirmed with the investigator. In addition, the
in satisfying regulatory reporting requirements.         investigator should notify the IRB/IEC of the com-
In general, the expectation is that the sponsor          pletion of the study. When the final draft of the
will learn of the event within 24 h of its occurrence.   clinical study report is available, it should be given
The sponsor monitor will transmit the information        to the investigator for signature. In multicenter
about the event to headquarters within 24 h and          trials, a single lead investigator may sign a pooled
headquarters will get the information to the drug        study report.
safety department within 24 h. The monitor should
immediately notify appropriate senior managers
                                                            Reviewing, Editing and Verifying In-house
of serious ADEs that are unexpectedly discovered.
                                                                Case Report Data and Databases
These strict timelines are designed to keep us
in compliance with the regulatory authorities.           While the goal of monitoring is to provide com-
Failure to adhere to the reporting timelines requi-      plete and accurate (``clean'') CRFs, it is necessary
red for regulatory authorities is evidence of negli-     to review CRFs for consistency and unrecognized
gence on the part of the sponsor. The sponsor            errors once they are received in-house. The use of
monitor is responsible for assuring adherence to         computer edit and logic checks supports this effort,
reporting systems for managing serious adverse           where computer output is verified against CRF
events.                                                  data. Discrepancies are identified and CRF queries
   The sponsor monitor is responsible for the timely     are generated for resolution.
follow-up of all serious adverse events. The cases          The goal of managing CRFs is to get the data
must be followed to completion. The monitor needs        from the CRFs to a clean database in the fastest
to collect all required follow-up information on         time possible, while maintaining the highest level of
ADEs.                                                    quality. To accomplish this task, CRFs must be
   To be successful, monitors need to be competent       ready for data entry at the site. CRFs must
in:                                                      be cleaned on an ongoing basis during the study.
                                                         To do this, efficient systems must be incorporated
. Basic medicine and therapeutics.                       to simplify the query process. The approach used
. Recognizing clinical signs and symptoms.               by some sponsors permits electronic exchange of
. Interpretation of laboratory findings.                 CRF data between the investigator, monitor, and
. Medical practice, nomenclature, and termin-            data entry personnel. Computerized checking pro-
  ology.                                                 grams and edit checks make the process more
. Relevant regulatory requirements.                      value-added for the monitors.
. Protocol requirements.                                    Clinical teams should design database before the
                                                         trial begins, reduce the amount of data collected,
The sponsor needs to provide on-going review of          use standardized CRFs, and complete the review
safety data for investigational products.                process on an on-going basis. The philosophy is,
                                                         `Do it right, first time' at the source.
                                                            To be successful, staff must know how to prepare
            Closing Down the Center
                                                         CRFs for data entry, be able to verify database
Closing down a study is important, because it may        consistency with original records and CRFs, and
represent the sponsor's last and best chance to          assure that queries are handled effectively.
                 COMPETENCY-BASED TRAINING SYSTEM FOR CLINICAL RESEARCH STAFF                            27

        Competencies Associated with                   . Clarification of outstanding issues regarding
         Reporting Clinical Research                     data analysis and presentation.
                                                       . Drafting of assigned study report sections
             Preparing Clinical Study                    according to the clinical study report prototype.
                     Reports                           . Interpretation of adverse events.
                                                       . Interpretation of laboratory findings.
The requirements for reporting clinical trials to      . Interpretation of efficacy findings.
international regulatory authorities are similar in    . Ensuring that data support the conclusions.
intent but differ in detail. Sponsors approach prep-   . Ensuring that reports satisfy regulatory re-
aration of NDA/BLA documentation in a modular            quirements.
format. Each module satisfies a specific documen-      . Developing clear, simple graphs, tables, and
tation need. The modules are generally organized         figures to illustrate and support findings.
as follows:                                            . Ability to write a clear, concise report that ac-
                                                         curately summarizes and interprets the results.
. Module I. Includes a basic summary of the
  study, not unlike a publication. It includes:
  study rationale, objectives, methods, results,               Preparing Annual Safety Reports
  and conclusions. Module I also has a large           Sponsors in the USA are required to submit annu-
  appendix, which includes: list of investigators;     ally to the FDA a summary of safety findings of
  drug lot numbers; concomitant diseases and           investigational products. This involves verification
  medications; intent-to-treat analysis; patient       of AE tabulations against computer data listings
  listings of adverse events and relevant labora-      and the preparation of safety tables. The current
  tory abnormalities; and publications on the          findings are reviewed and compared with AE data
  study.                                               from the past reporting period.
. Module II. Includes the protocol and any                The sponsor's representatives must be able to
  modifications, case report form, detailed meth-      clarify any outstanding issues regarding safety in-
  odology, and the glossary of original termin-        terpretation and presentation of the data. Since this
  ology and preferred terms.                           information is of critical importance to the FDA,
. Module III. Presents the detailed efficacy           the annual report must be written in a clear, concise
  findings, including the intent-to-treat analysis     manner that accurately summarizes and interprets
  population and the efficacy data listings.           the safety results. The annual report should pro-
. Module IV. Presents the detailed safety find-        vide clear, simple graphs, tables, and figures to
  ings, including the intent-to-treat analysis         illustrate and support safety findings. Following
  population and the safety data listings.             the submission of the annual report, safety findings
. Module V. Includes individual center sum-            are usually integrated into an updated version of
  mary reports, quality assurance measures, stat-      the IB.
  istical methods and analyses, and randomization         To be able to prepare annual reports, the spon-
  lists.                                               sor's representative should know how the reports
                                                       satisfy FDA requirements and those of other regu-
The skills necessary to prepare a clinical study       latory authorities. The clinical representative
report include:                                        should be able to interpret clinical safety and la-
                                                       boratory findings. The ability to understand
. Advanced research design, methodology, and           computer-generated clinical output and the organi-
  statistics.                                          zation and structure of the NDA/BLA safety data-
. Preparation and review of study tabulations.         base is important.
. Ability to confirm that study tabulations con-          The annual report and NDA/BLA safety update
  form to protocol design.                             review and approval procedures must be under-
. Ability to verify study tabulations against com-     stood as well as the procedures for the preparation
  puter data listings.                                 of the IB.

     Preparing Clinical Sections of NDA/BLA           The expert report provides an independent assess-
                                                      ment of the risk±benefit of the drug and its use. The
The knowledge and skill needed to prepare an
                                                      text is limited to 25 pages, but may include an
NDA/BLA/PLA includes the ability to:
                                                      `unlimited' number of attachments. Many com-
                                                      panies have been creative in font size and two-
. Verify individual study tabulations against
                                                      sided preparation of the document.
  overall summary computer listings.
                                                         Certain trends and directions can be recognized
. Prepare brief descriptions of the studies.
                                                      in the preparation of NDA/BLAs. The ICH has the
. Interpret critical clinical safety and efficacy
                                                      long-term goal of harmonizing the content of
                                                      European, US and Japanese NDA/BLAs. EU
. Interpret laboratory findings.
                                                      registration dossiers are becoming more detailed
. Develop clear tables and figures to illustrate
                                                      in the process and are expected to include inte-
  and support clinical findings.
                                                      grated summaries in the future. The US FDA will
. Summarize, interpret, and integrate the overall
                                                      accept more non-US data for drug approval as
  safety and efficacy results.
                                                      common high standards for clinical trials become
. Prepare NDA/BLA clinical study summaries,
                                                      well established in the world. Finally, electronic
  benefit±risk summary, expert reports, and
                                                      NDA/BLAs will be the norm and are already re-
  package insert.
                                                      quired in the USA.
In addition, an understanding of electronic NDA/
BLAs (CANDA/BLAs) and EU/FDA data presen-
tation requirements are useful.
   The expert report usually generates considerable   International Conference on Harmonization (1997) Good
discussion within a project. The document is most       clinical practice: consolidated guideline. Fed Reg 62 (90).
often prepared by sponsor under the guidance of       International Conference on Harmonization (1997) Draft guide-
an external expert. Whilst internal experts are ac-     line on general considerations for clinical trials. Fed Reg 62
ceptable, it should be remembered that the regula-
                                                      International Conference on Harmonization (1996) Guideline
tory authorities are looking for an individual who      on the structure and content of clinical study reports. Fed Reg
knows the drug thoroughly and can express an            61 (138).
unbiased opinion of its medical importance. The       CFR 21 (50, 56, 312, 314).
expert report is not just a summary but a critical    Spilker B (1991) Guide to Clinical Trials. Raven: New York.
assessment of the clinical evaluation of the drug.
Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                       Copyright # 2002 John Wiley & Sons, Ltd
                                                                    ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                              Section II
           Drug Discovery and Development
      Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                             Copyright # 2002 John Wiley & Sons, Ltd
                                                                          ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                            Introduction to Section II
                              Lionel D. Edwards1 and Anthony W. Fox2
                           Basking Ridge, NJ, and 2EBD Group Inc, Carlsbad, CA, USA

This section of the book is concerned about the                 artificial in comparison to how real drugs get de-
events that must take place in the conversion of a              veloped, which requires intimate coordination of
drug into a medicine that can be approved by                    clinical development with regulatory compliance.
regulatory authorities. It is important to know                 Similarly, the financial and legal underpinnings of
what does and does not comprise these events.                   the drug development process are described in later
   First, the chapters here may be regarded loosely             sections of this book, but are constraints that con-
as the premarketing phases of a successful drug's               stantly govern the thinking of the good pharma-
life cycle. This not congruent with the whole of                ceutical physician during clinical development.
drug development. What is needed to get a drug                     Fourth, the chapters on Phase I clinical trials and
approved is not the same as what is needed to make              pharmacoeconomic research have been written by
that drug into a commercial success.                            experts in these fields. These are very rapidly-de-
   Second, the overall emphasis here is clinical de-            veloping disciplines. The typical pharmaceutical
velopment, how this is done, and what preclinical               physician has usually paid little attention to these
information is needed in order to carry it out.                 aspects of clinical development, but it is our belief
Necessarily, we have here had to be general, and                that these will dictate his/her clinical development
these chapters are mostly written to describe, and              plans to an ever-increasing degree in the future.
assist with, typical development issues. However,               Phase I studies can shorten overall clinical develop-
for some disciplines, this is impossible; for example,          ment time, and the pharmacoeconomic leveraging
there is little preclinical pharmacology in this                of (especially) Phase III and Phase IV studies (with
section, since this discipline is really product-               preparatory Phase II work) are now essential in the
specific. In contrast, the general principles of toxi-          modern competitive environment.
cology (for example), and what the pharmaceutical                  The integration of the activities described in
physician needs to know about toxicology before                 these chapters is essential to good clinical develop-
starting any clinical trial, may be usefully stated in          ment. Some cross-referencing and overlap between
the general case.                                               these chapters is deliberately included to emphasize
   Third, while some clinical developers erroneously            such integration. Pharmaceutical physicians, spe-
view regulatory affairs as the implementation of                cialized though they may be in one discipline or
their clinical development plans (and probably                  another, are well-advised to keep an observant eye
vice versa), the proper constraints imposed by regu-            on the interactions between their own and other
latory authorities are so fundamental that they de-             company departments. Experience is probably the
serve a section to themselves (see Section III).                best teacher of integration. But we hope these chap-
However, we would emphasize that this division is               ters convey some idea.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

       Drug Discovery: Design and Serendipity
                                                J. Leslie Molony
                                  ProPharma Partners, Inc., Hayward CA USA

How is it that medicines are discovered? In ancient                          UNMET CLINICAL NEED
times, and even today, tribal peoples knew the
healing or hallucinogenic properties of indigenous              Usually, scientists are directed to research new
plants and animals. The knowledge was accumu-                   targets in specific therapeutic areas based on
lated through generations, recorded by chant and                unmet clinical needs. Once a need is identified and
living memory, and was derived largely from human               a particular therapeutic area chosen, the biological
experience. Although many of the drugs in use                   research begins. It is during this first stage of drug
today were discovered by chance, most drug discov-              discovery that anecdotal clinical observations, em-
ery scientists engage in directed research, based on a          pirical outcomes, and `data' collected from folk
series of steps, each requiring substantial scientific          medicine are often employedÐif only as direc-
input. While available facilities, resources, technol-          tion-finding tools.
ogy focus, or even corporate culture can define the                Once a direction is chosen, it must be validated
procedures followed by researchers at particular                scientifically, within a defined biological system.
institutions, there are some obvious, generally ap-             Because human disease or pathology is usually
plicable milestones in this process leading to the              multifactorial, the first task of the researcher is to
discovery of therapeutics.                                      narrow down the search to better defined mechan-
                                                                isms, preferably a small number of pathophysiolo-
                                                                gically observable processes, e.g. pinpointing one
     DESIGNING A DRUG DISCOVERY                                 or two types of cells which can be considered causes
               PROJECT                                          of the pathology. From the cellular stage, the re-
                                                                searcher next defines specific molecular targets,
An outline of the thought processes involved in                 such as receptors or cellular enzymes that comprise
designing and implementing a Drug Discovery pro-                the destructive phenotype.
gram is diagrammed in Figure 4.1. This chapter                     Researchers will target systems which are
will discuss the process, and give practical examples           affected by, or may be directly involved with, a
from contemporary drug discovery scenarios.                     particular disease. The treatments arising from
   All drug discovery projects depend on luck to                these types of approaches can be palliative, or
be successful, but research and careful planning can            may find a market or need as disease-modifying
improve chances of success and lower the cost.                  drugs. Prime examples of palliative therapies are
Project teams can streamline the discovery process              drugs designed to alleviate side effects of treatment
by mapping the most direct methods that will yield              with toxic chemotherapeutic drugs, such as nausea
a discovery. Using the tools available from modern              and wasting. In these cases, drug discovery scien-
biology, chemistry, robotics, and computer simula-              tists search for drugs which alleviate each symptom
tions, years can be eliminated from the search for              as if it were an isolated pathology.
new drugs. The costs of getting a new therapeutic                  Disease-modifying drugs are those which directly
into the marketplace in 1997 were estimated at                  affect the primary disease. Examples of DMDs are
$300±400 million and the average discovery and                  the chemotherapeutic agents themselves, which
development time is still 7 years.                              destroy tumors by preventing their growth.

     CELLULAR MECHANISMS OF DISEASE                          Common types of target identification experi-
                                                          ments involve raising monoclonal antibodies to
Drug discovery biologists and pharmacologists re-         proteins (receptors) on the surface of cells derived
search and evaluate the physiological mechanisms          from diseased tissue, then screening the hybridoma
of the disease to be addressed by a therapeutic sub-      supernatants for activity in preventing a cellular
stance. Initially, research takes the form of litera-     manifestation of the disease. Referring back to
ture searches and intense reading. The first step of      the cancer example, tumor cells often contain over-
the biologists is to identify potential physiological,    expressed, mutated, or absent `oncogenes'. Onco-
genetic or cellular mechanisms which could explain        genes are proteins which regulate the signaling
the pathological outcomes. In many cases, prelimin-       from a particular receptor in normal cells, but are
ary research has been published on several aspects        mutated, and thus are constitutively active or con-
of the pathology. It is then the task of the biologists   stitutively inactive, in different tumor cells. They
to design and implement experiments which further         are known as `oncogenes' because their modifica-
define or eliminate potential drug targets.               tion is often the cause of the abnormal behavior of
   This process can begin with the identification of      tumor cells. Examples of oncogenes are RAS and
a particular type of cell that can be used for cell       SRC. The normal RAS is known to regulate cellu-
biological, molecular and biochemical investiga-          lar division and link the nuclear changes to alter-
tions. Is there an anomaly in a cell derived from a       ations in the cellular architecture required for
tumor, to use a cancer example, which renders that        mitosis (cytoskeleton and cell motility). SRC is
tumor cell unique from normal cells derived from          known as a key signaling molecule which can affect
the same tissue? If the difference is significant and     cell growth by modulating the responses of the
can be reproducibly observed in the laboratory, it        epidermal growth factor (EGF) receptor to its
can be exploited for drug discovery. In other dis-        ligand. Many recent cancer drug discovery efforts
eases, the cell which is identified can be normal, but    in the cancer area have targeted SRC, RAS or the
activated to a destructive state by stimulation with      EGF receptor. The molecular targets chosen for
disease pathogens. In rheumatoid arthritis, for           these discovery efforts ranged from the EGF recep-
example, the normal T lymphocyte is stimulated            tor itself, the EGF receptor enzymatic signaling
to react to antigens present in the joint, thus de-       activity, the SRC molecule itself or its enzymatic
veloping a destructive phenotype.                         activity, or any of the enzymes which regulate key
   Once a target cell is identified and the patho-        interactions in this signaling cascade. RAS inhibi-
logical mechanism described, molecular targets            tor discovery projects have focused on the bio-
can be evaluated. Modern drug discovery usually           logical requirement that RAS be translocated
requires identification of a particular molecule          from the cytosol to the membrane before its enzym-
whose function can be modified by a chemical sub-         atic activity is functional. Prevention of the enzym-
stance. The choice of the molecular target is             atic event which allows translocation of RAS to the
an important milestone in the drug discovery pro-         plasma membrane in cancer cells will prevent acti-
cess.                                                     vation of events mediated by RAS.
                                                             Other examples of molecular target identifica-
                                                          tion can be derived from following some thought
  MOLECULAR TARGET IDENTIFICATION                         processes that a discovery scientist would follow to
                                                          design a novel approach to treating inflammatory
     Combining Basic and Applied Research                 diseases. A cell or molecular biologist beginning a
                                                          program in arthritis in 1997 would have access
Molecular target identification is critical to the        to literature from clinical trials of anti-TNF anti-
discovery process. Often, the molecular targets           bodies, and data describing transgenic mice which,
are not obvious, although cellular and histologi-         when genetically engineered to cause monocytes to
cal disease pathologies have been described in the        express constant levels of the cytokine (TNF), will
literature. At this point, the researcher returns         develop arthritis. Also available would be large
to the laboratory bench to design critical experi-        volumes of data on the cellular infiltrates in the
ments.                                                    inflamed joint: monocytes and T lymphocytes are
                               DRUG DISCOVERY: DESIGN AND SERENDIPITY                                     33

the most prevalent, and data regarding the eleva-       formulated into a pharmaceutical product, i.e., an
tion in levels of other cytokines, such as IL-1b, and   orally bioavailable TNF antagonist.
other mediators of inflammation, such as leuko-            The next question to be answered is what indi-
trienes and phospholipases. After an exhaustive         vidual regulatory enzyme will be the most effective
review of the recent literature, a drug discovery       and most specific target. To answer this question,
scientist might conclude that inhibitors of TNF         biochemists studying the transcription regulation
receptors might provide a significant benefit to        of inflammatory mediators may be consulted. Lit-
those with arthritis, and to those in the develop-      erature on the subject is available. However, part of
mental stages of the disease. At that point, the        the goal of drug discovery researchers is to take a
scientists then decide which target to approach:        unique, less obvious angle. This involves exhaust-
the TNF receptor itself, the ligand TNF, the signal     ive literature searches, attendance at meetings, and
transduction enzymes regulating the function of         working with competitive intelligence teams to de-
the receptor or, perhaps, expression of TNF itself      termine what approaches are being taken by the
via regulating the TNF gene.                            competition. Unless a company can work faster or
   Hypothetically, because we also know that ster-      better than the competition by taking a direct ap-
oids, a currently used therapeutic, works by inhib-     proach, it is, perhaps, a better strategy to approach
ition of gene expression of many of these mediators,    the target identification issue less directly.
it will probably be successful to target the gene          Each individual laboratory working on TNF as a
transcription pathways leading to expression of         therapeutic target is approaching the problem from
TNF in the joint. However, other considerations         a different direction. For example, one group may
exist. Are the transcription mechanisms known,          seek to inhibit transcription factor activation by
and if so, are there any other required enzymes or      phosphorylation or proteolysis, while another
proteins which are regulated by those same mech-        group seeks to inhibit the binding of the transcrip-
anisms? As it turns out, the mechanisms which           tion regulatory complex to the DNA. Key decision
regulate TNF transcription are common to path-          points will be made when targets are cloned, and
ways of inflammation, but are not common to pro-        assays are set up for screening. Those assays which
teins and enzymes required for cellular functions.      can be related directly to cellular events, which
   Because anti-TNF treatments and TNF-                 allow screening of the required number and type
overexpressing transgenic mice have already been        of chemical compounds, and which are predictive
developed, it is known that TNF itself can cause        of in vivo responses, will be chosen for follow-up by
destructive arthritis. Preventing TNF-receptor ac-      the screening team. Other assays that are developed
tivation by using antibodies to TNF will prevent        during this exploratory stage may be used as sec-
arthritis, and even reverse the destruction.            ondary screens, or will be used for validation of
   The investigators may seek the counsel of            targets. Figure 4.2 generally outlines the process
marketing experts and physicians regarding the          that has been developed in the scenario in this
use of the antibodies, and any clinical trial data      section. Target Identification requires further elab-
available through the literature on the anti-TNF        oration, and will be discussed next.
antibodies. These antibodies will be competing
against any product which comes out of a TNF
antagonist drug discovery program. Antibodies                   METHODS OF NEW TARGET
may present significant delivery problems because                   IDENTIFICATION
they are not orally bioavailable, i.e. they cannot be
taken in pill form, but must be injected. Other         The ideal molecular drug target is one which has
problems associated with non-human antibodies           been discovered in-house and is proprietary. Alter-
result from elicitation of immune responses to anti-    natively, a receptor or enzyme discovered by a
bodies themselves after several doses, thus limiting    researcher at a university, the cDNA of which
the number of times the antibody can be adminis-        is available for licensing by the drug company,
tered. The goal of the discovery group would be to      could also provide an effective and proprietary
discover a therapeutic with better properties than      target. These targets are often discovered serendip-
the existing antibodies, such as one that could be      itously, or could be the result of extensive `fishing

expeditions' using protein biochemistry, immun-         ological studies and linkage analysis, will provide
ology (antibody targeting), pharmacology, or mo-        molecular targets for the few diseases which are
lecular biology methods.                                derived from hereditary alterations in the human
   Cell and molecular biologists can identify a gene    genome, often the result of inbreeding in isolated
or genes responsible for the cellular variations        populations. An example of new target identifica-
seen in certain techniques employing subtractive        tion using these methods was the identification of
cloning. Total mRNA from normal cells are com-          ApoE as an important causative factor in Alzhei-
pared with total mRNA from diseased or modified         mer's disease (Pericak-Vance et al 1991).
cells (called normalization). Usually, numerous            The pharmaceutical industry can further exploit
genes are found to be modified or alternatively         this genomic technology by choosing specific dis-
expressed, and the mRNA will be detected in the         eases, performing epidemiological research to find
hybridization reactions. The real work comes in         families with patterns of hereditary disease, and
sequencing and identifying these genes, and ascer-      mapping the transmission of the disease to find
taining which are the unknown, novel sequences.         the specific genes that cause the pathology. Genes
Regarding the newly discovered genes, e.g. poten-       are then identified using positional cloning.
tial causes of cell transformation into cancerous          However, other researchers are not closing in on
tissue, many biochemical and cell biological experi-    suspected mutations, but identifying all the genes
ments will be required to determine the function of     expressed by human cells, and then sequencing
the protein coded for by the new gene.                  short segments of each one. These `expressed se-
   Another method of target identification is to        quence tags' (ESTs) can then used as starting
raise monoclonal antibodies to cell-surface recep-      points to derive full-length cDNA for a newly iden-
tors derived from a cell suspected to be involved in    tified gene.
the pathology targeted. Antibodies then identified         Genomics may, with the completion of the
by FACS (fluorescence activated cell sorting, or        Human Genome Project, become a mainstay of
labeling techniques) as binding to the target cell      new target identification for drug discovery. The
are then used in functional assays to prevent a         Human Genome Project is a consortium of govern-
response thought to be critical for disease develop-    ment and industry-funded laboratories, which se-
ment. If an antibody can inhibit this response, the     quenced the entire human genome using both EST
antibody is then used by protein biochemists to         and positional cloning methods. This project iden-
isolate the culpable receptor.                          tified and sequenced all of the estimated 60, 000 to
                                                        30 000 ±100 000 human genes (3 billion nucleotides)
                                                        by the year 2001.
  Genomics and New Target Identification                   At the time of writing complete genomic se-
                                                        quences are available for atleast 141 viruses, 51
Drug companies have recently become involved in         organelles, two eubacteria, one eukaryote and
the use of genomics to identify new genes which         most mammalian mitochondria. Several organisms
might have some role in disease pathology. Mo-          for which complete genome data are available in-
lecular biologists can seek mutations or alterations    clude Haemophilus influenzae, Mycoplasma geneta-
in genetic signatures which are predictive of the       lium, and Saccharomyces cervisiae. The volume of
targeted disease in a large population. If this muta-   this information may be illustrated by almost the
tion is always associated with the disease, they can    simplest case: the yeast, S. cerevisiae contains ap-
then map this gene to the disease. This is known as     proximately 6000 genes (Schuler et al 1999), the
`linkage analysis'. When linkage analysis can pre-      sequences of which are available to scientists at
dict that a disease-regulatory gene falls within a      FTP sites on the internet. It is expected that by
certain region of a chromosome bounded by other         maps for the common bacteria E. coli and B. sub-
known genes, scientists can determine the genetic       tilis, and the C. elegans will also be complete and
sequence of the causative gene. This process is         soon publicly available.
known as `positional cloning' because the cloning          The sequencing of unknown genes will not dir-
begins and ends at certain chromosomal locations.       ectly identify new molecular targets for disease.
Positional cloning, coupled with global epidemi-        However, the availability of sequences will permit
                               DRUG DISCOVERY: DESIGN AND SERENDIPITY                                       35

rapid identification of genes once a target protein is   allow a tissue or organ-specific understanding of
identified, without having to sequence more than a       the effects of potential new drugs. Cardiovascular
few peptides of the protein. The access to gene          pharmacologists often study isolated arteries,
sequence information should shave months off of          which are maintained in a physiological salt solu-
the discovery process, allowing rapid cloning of         tion. Electric stimulation can induce contraction of
new targets for assay development.                       the vascular smooth muscle, and the effects of
   Genomics and epidemiology will identify genes         hypertensive drugs on vascular contraction can
that are altered or mutated in certain diseases.         then be measured. Historically, these systems were
However, this molecular biological breakthrough          often used as primary drug screening tools. Be-
will not permit the mapping of a function to the         cause these methods are much less direct than mo-
genes. That task is assigned to the cell biologists,     lecular screening, they are now relegated to
who will have to study basic cellular mechanisms         secondary or tertiary roles as validation of the
with normal and diseased genes, or study trans-          targets or drugs discovered.
genic animals expressing the disease-specific gene
in certain tissues, before a function can be attrib-
uted to the newly identified gene.                                    Transgenic Technology
   Yet another group of biotechnology companies
is exploiting this niche, to provide `functional         Recently, the pharmaceutical industry has adopted
genomics' services which allow identification            transgenic technology, allowing researchers to en-
of biological functions associated with their clients'   gineer rats and mice whose physiology mimics a
newly-identified disease genes. Functional genomics      disease on a molecular and tissue-specific level.
takes advantage of cell or developmental biology,        Transgenic animals are used in the pharmaceutical
and measures the effects of modified vs. normal          industry for two major purposes: target validation
gene sequences in functional assays. These func-         and protein expression. The expression of recom-
tional assays are not, themselves, drug screens.         binant proteins in large quantities by genetically
They are methods of validating new targets, which        engineered animals is a common method of
must then be developed further for the discovery of      obtaining recombinant proteins or enzymes to be
drugs.                                                   used as therapeutics or nutritional supplements.
                                                         Sheep, goats, and cattle have been engineered to
                                                         produce recombinant proteins in their milk, for
             Physiological Systems                       example. These types of transgenic animals will not
                                                         be discussed here.
Drug discovery scientists must bear in mind what            In the pharmaceutical industry, as in basic re-
the broader effects of inhibiting, modifying, or         search, the cause of a disease can be identified by
eliminating this new target would be on the organ-       introducing a suspect gene into the genome of a
ism. The perfect target is organ-, tissue- or cell-      rodent. This technology utilizes the plasticity of the
specific, thereby limiting effects to the system         embryonic stem cells derived from fertilized ova.
involved in the disease. The choice of a target for      Embryos are removed at the 4 ±8 cell-stage, micro-
a disease will be critical to the outcome and per-       injected or transfected with a plasmid containing the
formance of the drug, and will determine what            gene to be expressed, probably engineered to be
organs or tissues will be susceptible to side effects.   dependent upon a promoter which limits expression
                                                         to specific organs or tissues, and engineered to con-
                                                         fer upon those ES cells carrying the plasmid resist-
CRITICAL IMPORTANCE OF IN SITU AND                       ance to the selectable marker. After selection of the
           IN VIVO STUDIES                               population, genetically altered embryonic stem cells
                                                         are grown to the 6±8 cell stage and placed in utero.
Pharmacologists are often able to develop tissue         The resultant rodents will express the introduced
and whole animal models of human disease. In             gene if the plasmid was incorporated into the host
some instances, studies on isolated tissues, such as     DNA. Using this method, researchers can then
blood vessels, heart muscle or brain slices, will        determine if an overabundance or restriction of a

specific factor will result in a disease, or will prevent   induce a pathological state by introduction of a
one.                                                        pathogen or stimulant directly into a healthy
   In contrast, the `knockout mouse' was engin-             animal.
eered to determine the functions of unknown                    Whether the disease is induced endogenously
genes. The test gene is targeted for deletion in the        (genetically) or exogenously by manipulation or
mouse embryo, and its DNA is replaced with plas-            administration of a disease-producing agent, the
mid containing a disrupted gene which cannot be             development of new animal models is a time-con-
expressed, and a marker gene which confers resist-          suming process. Pharmaceutical researchers are
ance. After positive selection, embryos are then            supervised by veterinary medical doctors, and by
reintroduced into mothers and carried to term.              committees which review the protocols to be used.
However, in most cases, homozygous knockout                 These veterinary professionals and committees
mice (mice which lack the gene in both sets of              oversee the issues related to humane use, as well as
chromosomes) develop abnormalities that are in-             updating researchers on more efficient methods
compatible with life beyond an early embryonic              which might allow reduction in the number of
stage. In some cases, however, the knockout                 animals used.
mouse which lacks specific genes, thus does not                Why are in vivo (whole animal) studies still im-
express the resultant receptor or enzyme, and will          portant to drug discovery? While new technologies
give unique insight into the function of the previ-         are becoming available to develop computer-gen-
ously unknown gene.                                         erated models of whole organisms, this has not
                                                            eliminated the need for animal use. Until reliable
                                                            computer models accurately predict the effect of
 Spontaneously Arising Phenotypic Models                    each chemical compound on the cell, the tissue
                                                            and the organism as a whole, physiological studies
Mutations which cause disease can arise spontan-            will remain critical. New chemical or biological
eously. Genetic mapping methods utilizing pos-              therapeutics are often broad-acting, affecting per-
itional cloning can help identify disease-causative         haps unrelated physiological systems. The effects
genes and their proteins in animals which have              of new compounds on one system may negate or
spontaneously developed diseases similar to those           enhance its therapeutic effect, thereby eliminating
of humans. An example of this type of technology            or identifying a drug discovery lead candidate. In
is the ob/ob genetic mouse, which is obese, and has         addition, results from testing novel drugs in histor-
mutations in a gene for a peptide hormone known             ical animal models of disease are often required by
as leptin. This mouse, and its counterpart the db/db        regulatory agencies prior to allowing testing in
mouse, which has mutations in the leptin receptor,          humans.
can be used as animal models for obesity. A similar
mouse, the Agouti strain, is also obese and has
defects in melanocortin receptors, which regulate                       TARGET VALIDATION
hair color as well as appetite, and the expression or
release of leptin by adipose tissue. The Agouti             The next logical step, once a molecular target
mouse develops type II diabetes, and therefore              has been identified, is the validation of the molecu-
can be used as an animal model of that disease in           lar target and the evaluation of this target for pos-
humans. Of course, human disease is rarely as               sible effects in other systems. This stage of the
simple as a single genetic misread, so these models         project may involve developing antibodies which
must be used with some caution when testing drugs           inhibit actions of the target molecule, and testing
or when identifying the causative genes.                    these antibodies in animal models of disease. In
                                                            experimental systems where a receptor is expressed
                                                            in the mouse or rat, and its elimination will not be
           In Vivo Models of Disease                        fatal, the genetic knockout technology can be
Human pathology is inevitably more complex than                There may be fortunate scenarios where drugs,
those of rats and mice. Thus, it is often necessary to      often developed for other uses, have broad-acting
                               DRUG DISCOVERY: DESIGN AND SERENDIPITY                                      37

effects which include preventing a newly identified     drug screening was performed manually. These
molecular interaction causing a known pathology.        assays were competitive, because drug companies
This drug, while not an ideal candidate to treat        had limited numbers of proprietary substances in
the pathological state, may be used as a positive       their libraries, and medicinal chemists were limited
control for in vitro and in vivo assay validation.      in the number of new chemical entities they could
Alternatively, antibodies which inhibit the func-       design and synthesize manually.
tion of a targeted receptor or ligand may be used          For example, a biotechnology division has estab-
in vivo to evaluate the pharmacological and physio-     lished a novel high-throughput robotics system
logical effects of inhibiting that particular inter-    which is integrated with data analysis and data
action.                                                 storage computers. There are usually limits to the
   If the outcome of target validation experiments      types of assays that can be run using this integrated
is promising, drug screens will be established to       technology. If, for example, a novel assay has been
identify new chemical entities which will serve as      developed using measurements of fluorescently
specific therapeutics in place of antibodies or the     labeled cells flowing through a detector (FACS),
broader-acting drugs. Usually, these target valid-      the equipment must be redesigned or the assay
ation models will be used to validate the drug          must be modified to use existing technologyÐa
screening assays as they are developed, and can be      fluorescent plate reader using a 96-well format,
used as decision-making points for any new drugs        for example. In smaller, more technology-driven
identified using the in vitro assays.                   companies, assay developers will construct their
                                                        assays specifically to make use of certain propri-
                                                        etary platforms, e.g. some companies use propri-
      IN VITRO ASSAY DEVELOPMENT                        etary cell lines and cell-based screening, while
                                                        others use robotics and high throughput screening,
An assay, or screen, is the primary tool used to        but restrict their targets to certain classes of mol-
integrate a biological system with chemical com-        ecules, such as 7 transmembrane domain G-protein
pounds that will become drugs. Assays are experi-       linked receptors.
mental systems which allow rapid and reproducible
measurements of the effects of chemicals on the
molecular target. These can take many forms,                    Primary vs. Secondary Screens
from whole cell-based assays to one-enzyme col-
orometric-readout assays. Critical components of        The keys to a successful drug discovery program
any assay system are the availability of positive and   are the assays used to evaluate chemical or
negative controls, and quality control parameters       biological compounds. Assays will be prioritized
obtained from experimenting with multiple pos-          based on several scientific considerations. If the
sible parameters. Assays or screens are, in them-       chosen target is a receptor whose ligand is known,
selves, experimental systems, therefore good            and we are seeking a drug which will inhibit the
scientific method must be applied. Data generated       receptor from binding to its ligand, a simple recep-
by screens must meet criteria of statistical signifi-   tor/ligand screen may be established. If the target is
cance. A drug screening decision tree is dia-           a signal transduction enzyme which modifies an-
grammed in Figure 4.3. The importance of those          other intracellular protein, an enzyme-substrate
decision points is described further in this section.   screen can be established using a simple colorimet-
                                                        ric readout. These are often designed to be primary,
                                                        or first round, screens.
         Technological Considerations                      Before any novel chemical entities are screened
                                                        in the primary assay, known compounds and con-
Assay development for drug screening requires the       trols from many classes of drugs, known as phar-
creative flexibility to integrate the assay or assays   macophores, are often tested in the assay. This will
into existing technology. Prior to the technological    provide a set of controls, both negative and posi-
breakthroughs resulting in robotic high through-        tive, for developing new compounds based on
put screening and combinatorial chemistry, most         chemical structures. Pharmacophores are often

                                         Designing a Drug Discovery Project
         Defining the Medical Need and how to address it

                                Unmet Medical Need (Disease)

                       Tissue or Organ

                    Cellular Mechanisms

                   Molecular Mechanisms
                                                                      Defining how the project will be organized
  New Molecular Target for Drug Discovery
                                                                                     Assay Development
                                                                                       Decision Tree
                    Target Validation

                                    Technology (cells,               Chemistry             Required Throughput              Data
                                 gene therapy, antibodies?)        considerations     (now many compounds to screen)    considerations

                                        Structure-based, natural products,       Secondary, validating                 Database design,
                                             antisense combinatorial              assay requirements                    Information flow

                                                           Application of drug      Toxicity Assays to derive
                                                             likeness rules            Therapeutic Index

Figure 4.1 The drug discovery process

Figure 4.2 Target identification

used to find chemical backbones with some activ-                                Most primary assays are used in high-throughput
ity. These backbones can then be optimized using                             screening systems, and can be adapted for use with
rational design or methodical medicinal chemistry.                           robotics. The assay should then be highly reprodu-
Often, antibodies are used as controls because no                            cible, and contain both negative and positive
existing chemical inhibits the interaction.                                  controls. Because the reagents used for assays
                                  DRUG DISCOVERY: DESIGN AND SERENDIPITY                                   39

Figure 4.3 Drug screening flowchart

may present some variability, quality control par-     receptors other than the target would be elimin-
ameters are often used to validate reagents before     ated. In most cases, the drugs will have broad
employing them in screens.                             activity at very high doses, but will be quite specific
   Primary assays are used for the first round         at pharmacological doses.
of testing drugs which may or may not have activ-         In addition to evaluating and validating in
ity. It is important that secondary, or validation,    vitro activity of a new compound, the drug discov-
assays be established. These secondary screens         ery program must validate that newly discovered
will allow a second criterion (or more) for            chemicals are not toxic. Cytotoxicity assays are
accepting a compound as active. These assays are       usually included in a drug discovery flow chart.
often related to the primary assay, but contain        Drugs which have good activity in the in vitro
more of the entire cellular system. Often, secondary   assays, but are cytotoxic, may prove difficult
assays utilize live cells in which the enzyme and/     to develop further. Often, medicinal chemists can
or receptor are crucial for a specific cellular re-    modify them to eliminate the toxic side effects.
sponse.                                                Once the toxic effects are ``engineered away,''
   Secondary assays can also be used to evaluate a     the active compounds can then be tested in
molecular interaction which the researcher does, or    animals.
does not, want to inhibit with a drug. This can
select for compounds that are more specific.
Examples where `elimination' type of assays are        Considerations Regarding Specific Types of
used are receptor systems which have very similar                     Compounds
structures but have unique functions. G-protein-
coupled receptors, such as the b-adrenergic or sero-   The four major sources of starting material for
tonin receptors, have similar structures but quite     drug screens are chemical libraries, natural prod-
different functions. In these and many other           ucts, specifically designed medicinal chemistry-de-
examples, specificity can be engineered into the       rived drugsÐoften modified and synthesized using
discovery process by the addition of secondary         directed combinatorial chemistryÐand computa-
screens. Compounds that are active against any         tionally designed drugs.

   Large pharmaceutical companies often have             active ligand to a known amount of receptor. Re-
their own libraries of compounds, collected after        ceptors can be presented as components of a
decades of manual synthetic chemistry, which are         membrane preparation, or can be purified from
used as starting points for random screening. In         natural or recombinant sources. Ligands can be
addition to these proprietary compounds, test com-       either purified natural or recombinant ones, or
pounds can be obtained commercially and are de-          can be prepared from synthetic ligands, such as
rived from the Fine Chemical Database. Usually,          peptide sequences or drugs known to bind to the
these chemicals are organicÐonly partially soluble       receptor. There are numerous formats available for
in water. The amount of organic solvent required         assay set-up, such as filters, multiwell plates, or
to solubilize the chemical will often determine          beads, to which the receptor is bound. Radiola-
whether or not it can be evaluated in an aqueous-        belled ligand is incubated with the receptor in the
based assay. Random screening of broad com-              presence of test compounds. Unbound material is
pound collections or mixtures of natural products        removed by washing plates or filtering wash solu-
is a common method of seeking new chemical               tions through filters, then radioactivity is quanti-
leads.                                                   fied, either in solution, using scintillation counters,
   For companies which do not use large com-             or using the filter format and scintillation counters
pound collections or sources of natural products,        made to detect radioactivity in that format.
combinatorial chemistry and molecular modeling              An assay technology which combines radioactiv-
are both used to discover leads. Medicinal chemists      ity with bead technology is known as the scintilla-
use ``drug-likeness'' rules, or physical properties of   tion proximity assay (SPA) system, which works by
the starting components, to define what types of         binding a receptor to a bead which contains a
chemicals are tested and progressed through devel-       scintillant. Membrane preparations or purified re-
opment. These rules, developed by C.A. Lipinski at       ceptors are linked to beads, often using wheat-germ
Pfizer, provide guidelines to assist with selection of   agglutinin lectin. The receptor±bead complex is
compounds that possess physical properties re            then incubated in a solution containing radio-
similar to existing drugs.                               actively labeled ligand. When ligand binds, radio-
                                                         activity is in proximity to the scintillant, and thus
                                                         stimulates a response. If the ligand is inhibited
 Considerations Regarding Throughput and                 from binding, no response is observed. This tech-
               Assay Cycles                              nology is also effective for enzymatic assays, in
                                                         which the radioactively labeled substrate is bound
There are four basic types of classical assays: recep-   to the bead and is then cleaved, removing the radi-
tor/ligand assays, enzyme/substrate assays, anti-        ation and the signal if the enzyme is active. In this
body/antigen assays, and cell-based assays, which        case, however, a positive signal indicates inhibition
use live cells and measure a cellular response.          of the enzyme. For SPA assays, the readout is
Assays which measure the association of two mol-         quantified using a scintillation counter that can
ecules are the simplest types of assays to develop,      measure filters or multiwell plates. The SPA
because the binding reaction can be followed by          screening format is often used for high-throughput
radioactivity or colorimetric or fluorescent read-       screening, since it is easily scaled up.
outs. There are several technical developments              ELISAs (enzyme-linked immunoabsorbent
which reduce or eliminate the need for radioactivity.    assays) are another common framework used for
   Classical receptor/ligand assays measure the          drug screening. An enzyme-linked-antibody takes
binding of a radio-labeled ligand to its receptor.       the place of a ligand, whose receptor is bound to a
The type of radioactivity will depend on factors         plate or filter. The mixture of drug and enzyme-
such as the structure of the ligand, the affinity of     linked antibody is incubated in the well with the
the binding reaction, and environmental factors          receptor. After a series of washes to remove
such as disposal of radioactive waste. 125 I, 3 H,       unbound material, the substrate for the enzyme is
   S, 32 P and, infrequently, 14 C, have been used       added to the well. A common enzyme/substrate pair
for receptor/ligand screening. Briefly, the assay        is alkaline phosphatase and pNPP ( p-nitrophenyl
measures the binding of a known amount of radio-         phosphate), which results in a yellow color. Another
                               DRUG DISCOVERY: DESIGN AND SERENDIPITY                                      41

enzyme commonly used is horseradish peroxidase           of drug screening, since all drugs will eventually
(HRP). Substrates for HRP include 2,20 -azino-           have to be tested with whole cells. However, be-
bis(3-ethylbenzthiazoline-6-sulfonic acid (ABTS:         cause there are multiple cellular systems that could
green), tetramethylbenzidine (blue), and o-dianisi-      be affected, it becomes more difficult to define a
dine (yellow-orange).                                    mechanism of action for a compound discovered
   Signal transduction enzymes such as tyrosine          using a cell-based assay.
kinases can be assayed in high-throughput screens.
These screens often measure the binding of a sub-
strate, either the ATP at that binding site, or a more            The Drug Screening Process
specific site where the enzyme recognizes its sub-
strate and transfers the phosphate. In these assays,     Once assays are developed and characterized and
a synthetic peptide substrate is often synthesized.      sources of new compounds chosen, the screening
To detect whether or not the substrate has bound,        process begins. This is the most routine and least
one can use either 32 PATP or enzyme-linked anti-        creative task of drug discovery. At the same time,
bodies recognizing the phosphorylated amino acid.        the more efficient the screening process, the more
Some assays are designed for fluorescence or lumi-       likely it will be to discover new chemical entities
nescence readouts using probes with those proper-        that have therapeutic value. Screening can be per-
ties.                                                    formed manually or can be totally automated;
                                                         most screening groups are somewhere in the middle
                                                         of that spectrum. Often, drugs are weighed out by
      Recombinant vs. Natural Proteins                   hand and put into solution, then loaded into liquid
                                                         sample handlers for dilution and delivery. Robots
Molecular target enzymes or receptors must be            can also be used to set up 96-well plates to which
highly purified, and must be obtained in sufficient      receptors are bound, or cells attached. Depending
quantities and with consistent activities. Proteins      upon the format of the assay, the screening and
used in assays may be derived from recombinant           data collection process may also be performed
expression systems or from natural sources. The          automatically.
choice will be made according to the amount of              Spectrophotometric devices, called microplate
material needed and the ease of extraction from          readers, collect raw data resulting from colorimet-
the source. Usually, recombinant proteins are over-      ric or fluorescent screens. Similarly, scintillation
expressed in bacterial, yeast insect or mammalian        devices measure the amount of radioactivity in
systems, resulting in much greater yields of pure        samples from drug screens. The computer format
protein per gram of starting material than can be        of the data will then allow it to be exported into a
obtained from natural tissues. Care must be taken        spreadsheet or statistical analysis computer pro-
to normalize the activity or binding affinities for      gram for analysis.
each preparation, requiring biochemical quality             First-round screens based on random com-
control monitoring That will permit comparison           pounds or natural products are usually performed
between assays run, and thus relative activities of      by testing one concentration of a new compound or
compounds tested, using different batches of target      natural product mixture. In some cases, the test is
protein.                                                 done in duplicate or triplicate, depending on factors
                                                         such as space and time. Active compounds are then
                                                         titrated for dose±response analysis. If a dose±re-
                Cell-based Assays                        sponse is observed, the active compounds are tested
                                                         in the secondary assays, and tested for toxicity.
A more direct method of screening compounds for          Once a compound has proven active, not cytotoxic
cellular activity, while simultaneously determining      (if required), and dose±responsive, it can be ana-
that the compound can access the receptor or             loged using combinatorial chemistry or medicinal
enzyme in its cellular environment, is by assaying       chemistry methods. Small modifications in the side
the effects of drugs on whole cells. This type of        chains of active molecules will be produced, and
assay will eliminate at least one step in the process    these modified compounds tested in the assays.

   Chemical structure modifications which enhance        with similar charge distributions, but with unique
activity and refine specificity will be followed up      structural features, that can be considered
using medicinal chemistry methods, relating              new chemical entities (NCEs). This traditional me-
structures of compounds to activity in certain           dicinal chemistry represents the scientific basis
assays. These exercises will result in the definition    of many pharmaceutical breakthroughs in the
of a quantitative structure±activity relationship        last several decades. Novel methods for discovery
(SAR) for the series of compounds active in the          or identification of NCEs from computer-
assays.                                                  generated models were not developed until the
   The development of quantitative structure±            1980s.
activity relationships (QSARs) is made more effi-           Because random screens will generate leads with
cient by the use of computer databases into which        quite varied structures, chemists then turn to com-
are integrated chemical structures. Compounds            puter analyses to help them narrow down the most
synthesized for other discovery programs can be          critical chemical components of a potential new
related to activities from new assays to help re-        drug. Many statistical and analytical programs
searchers build specificity and activity profiles for    are available.
novel drugs.                                                Once a significant number of structures is
   The screening process itself can take between 2       obtained, chemists can employ QSARs to map
weeks and over a year, depending upon the through-       chemical structures with the most desirable activ-
put available. The most rapid technology screened        ities. These algorithms can also be used to eliminate
10 000 compounds or more each week. At that rate,        undesirable activities, such as cytotoxicity or lack
new assays can be cycled through screening rapidly       of specificity. QSAR analysis is a very powerful
for lead identification. New technology is being de-     tool in drug discovery. These methods have signifi-
veloped which will increase that screening potential     cantly reduced the number of compounds a chemist
to 100 000 compounds/month. Therefore, screening         must synthesize, by directing and focusing chem-
of one molecular target with a large collection of       ical efforts.
compounds will take on the order of a few weeks.
Optimization and QSAR analyses for `hits' can be
                                                                       Designing Novel Drugs
performed using the manual screening methods, or
robotic screens with lower throughput ranges.            Computer-assisted design (CAD) of drugs is a simi-
                                                         lar method of making use of the structural attri-
                                                         butes of the receptor or enzyme target. Using data
                                                         obtained from X-ray diffraction of crystalline
      Sources of New Chemical Entities                   arrays of the receptor (usually bound to a ligand
                                                         or antibody), the computational chemist can con-
                   Existing Drugs
                                                         struct a three-dimensional model of the target, com-
Chemical diversity is a critical variable in drug dis-   plete with charge distributions and conformational
covery. To obtain diversity, chemists have turned to     variations. This model can then be used to predict
natural products, have used random screening of          which chemical structures will have the desired
broad chemical collections such as the Fine Chem-        properties to fit into the active sites on the molecule.
ical Directory and their corporations' own chemical      One biotechnology company that has been success-
directories, and have developed methods of com-          ful using this CAD method is Agouron Pharma-
binatorial chemistry to deliver more compounds.          ceuticals, Inc. in San Diego (Now a division of
Random screening is performed when there are no          Pfizer, Inc.) California. Interested readers should
starting pointsÐno peptide or natural product            refer to work performed by Agouron, and to
leads, and no drugs known to bind to the receptor        reviews of CAD provided in the reference list.
or enzyme target.
   Medicinal chemists often begin by finding             Pharmacognosy is not yet a redundant source of
a chemical structure, either from a previous patent      molecules. The Pacific Yew has recently yielded
or literature report, that is an antagonist or agonist   the lead for successful therapies for ovarian
of a receptor target. They can design compounds          cancer. One large pharmaceutical company has
                                    DRUG DISCOVERY: DESIGN AND SERENDIPITY                                           43

recently concluded an agreement with a Central                    New Uses for Old Drugs. Lastly, opportunities
American country to preserve its entire flora and                 still exist for astute clinicians to find new uses for
give that company exclusive rights to any pharma-                 old drugs, and for these newly-discovered uses to
cophores within it.                                               lead to new and unexpected drugs. The recent ap-
                                                                  proval of bupropion as a smoking cessation agent is
Combinatorial Chemistry. The breakthroughs in                     a good example of a chance observation while the
technology that have allowed sequencing of genes                  drug was being used for its initial indication, which
`on a chip' and high throughput screening of com-                 was as an antidepressant. This has led to realization
pounds in microtiter plate format have also caused                of the influence of nicotine on depression, and
a revolution in chemical synthesis, known as com-                 investigational drugs of a new class, based on this
binatorial chemistry.                                             alkaloid molecule, are now being designed.

Biological Therapeutics. The chapter on biotech-
nology drugs enlarges on this subject in more                                            SUMMARY
detail, but suffice it to say here that vaccines,
antibodies, proteins, peptides, and gene therapies                This chapter began with a survey of the modern
all now exist. These biological drugs bring with                  methods of drug discovery. Figure 4.4 describes the
them specific, regulatory, clinical trials and manu-              relationship between screening of compounds,
facturing difficulties. Gene therapy, in particular,              chemical modifications, drug formulations
carries human safety risks that do not apply to                   (pharmaceutics), preclinical testing, and clinical
other classes of therapy, e.g. the infective nature of            trials. Pharmaceutical physicians should be aware
some types of vector that are employed, and the                   of some of the techniques employed and the rapid
potential for incorporation of the test genetic ma-               rate at which genetic information is becoming
terial into the genome in males, leading to expres-               available. It should be noted that this modern
sion of gene products in offspring.                               revolution has not quite completely swept away

Figure 4.4 CAD, computer-assisted design; ID identification; SAR, structure±activity relationship

the occasional new drug found by serendipidity or              Beeley LJ, Duckworth DM (1996) The impact of genomics on
astute clinical observation.                                     drug design. Drug Discovery Today 1(11): 474 ±80.
                                                               Chapman D (1996) The measurement of molecular diversity: a
   As the human genomic map has been de-                         three dimensional approach. J Computer-Aided Mol Design
ciphered, and target discovery streamlined, the de-              10: 501±12.
cision-making process now involves additional                  Goffeau A et al (1996) Life with 6000 Genes. Science 274:
hurdles. There is a wealth of newly identified mol-              546±67.
                                                               Kozlowski MR (1996) Problem solving in laboratory automa-
ecules available from the genome project, and few
                                                                 tion. Drug Discovery Today 1(11): 481±8.
novel methods to effectively sort them into ``drug-            Lipinski CA et al (1997) Experimental and Computational ap-
gable'' targets. And, the discovery methods for                  proaches to estimate solubility and permeability in drug dis-
identifying new, novel chemical backbones with                   covery and development settings. Advanced Drug Deliv. Rev.
drug-like properties are still being developed.                  23: 3±29.
                                                               Lipper RA (1999) How can we optimize selection of drug devel-
   One thing to note is that the field has moved on              opment candidates from many compounds at the discovery
since the genome has been sequenced. Chemistry is                stage? Modern Drug Discovery 2(1): 55±60.
still the rate-limiting step. In other words, new              Pericak-Vance MA, Bebout JL, Gaskell PC et al (1991) Linkage
chemical structures that have unique activities are              studies in familial Alzheimer disease: evidence for chromo-
still hard to come by.                                           some 19 linkage. Am J Hum Genet 48: 1034±1050.
                                                               Schuler GD et al (1996) A gene map of the human genome.
                                                                 Science 274: 540 ±46.
                                                               Van Drie JH (1997) Strategies for the determination of pharma-
                     REFERENCES                                  cophoric 3D database queries. J Computer-Aided Mol Design
                                                                 11: 39±52.
Amersham Life Science (1993) Brochure on scintillation Drug    Wainer IW (1993) Stereoisomers in clinical oncology: why it is
 Discovery proximity assays. Publication No. S 593/657/4/93/     important to know what the right and left hands are doing.
 09. Amersham International:                                     Ann Oncol 4 (Supp 2): 7±13.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                                 Anthony W. Fox
                                       EBD Group Inc., Carlsbad, CA, USA

It is a triumph of modern pharmaceutics that most               least as many questions about `chemistry and
physicians think, both when conducting clinical                 manufacturing' (for which read `pharmaceutics')
research and in ordinary clinical practice, that `a             as about clinical efficacy and safety.
drug is a drug is a drug'. The typical pharmaceut-
ical physician never makes a pharmaceutical (or
galenical) drug formulation. Nevertheless, under-                   THE CONSTITUENTS OF A MEDICINE
standing at least some elements of this science is
important because:                                              `A drug is not a drug is not a drug' because, when
                                                                administered to a human being, in the general case,
. A suitable formulation permits the conduct of                 it contains:
  clinical trials.
. Formulations constrain clinical trial design, e.g.            . Active compound at a dose within a limited
  whether a well-matched placebo is likely to be                  range, sometimes as a racemic mixture.
  available, or whether special procedures to                   . Manufacturing impurities.
  mask an infused, colored solution might be                    . One or more excipients.
  needed.                                                       . Degradants of the active compound.
. Packaged white powders are probably not mar-                  . Degradants of the impurities.
  ketable, and overcoming galenical problems is a               . Degradants of the excipients.
  sine qua non for product feasibility.
. Formulation can strongly influence patient ac-                An impurity is defined as a compound which is the
  ceptability and thus probability of commercial                by-product of the manufacturing process used for
  success.                                                      the active compound, and which is not removed
. Product storage and stability (or lack thereof )              prior to formulation. Impurities can have their
  can bias clinical trials results.                             own toxic potential, and control of impurity con-
                                                                tent is therefore a highly important feature in any
For all these reasons and more, marketing and                   New Drug Application (NDA). An excipient is de-
clinical input on suitable formulations should be               fined as a material that is incorporated into the
included in the earliest considerations of project              formulation to aid some physicochemical process,
feasibility, and it behooves the pharmaceutical                 e.g. tablet integrity, dissolution, or taste; excipients
physician to be able to provide such input in an                are typically chosen from among many compounds
informed manner. Equally, the pharmaceutical                    without pharmacological properties (e.g. lactose),
physician should understand the constraints and                 although there are examples where pharmacoki-
difficulties that his colleagues work under, even               netics change with the excipient used. There are
though the physician is not him/herself qualified               specialized examples of excipients; e.g. propellants
for a position in the company's pharmaceutics de-               are excipients that assist in the delivery of inhaled
partment. At the end of the day, product licences               drugs to the respiratory tract. A degradant is de-
are awarded and new drug applications (NDAs)                    fined as a compound that accumulates during the
are typically approved after the resolution of at               storage of bulk drug or finished formulation.

   Not even this list, however, comprehensively de-       by using drug removed from the production pro-
scribes a formulation. Many tablets carry printed         cess before the last step, for example, before the last
identification markings or are color-coated; dye-         recrystallization. This usually guarantees that a
stuffs are outside the definition of an excipient,        lower purity, i.e. mixture with greater molecular
and allergies to them have been documented. Dif-          diversity than the drug of interest, will be tested
ferential efficacy exists among differently colored       toxicologically than that to which patients will ac-
placebos, and this should therefore also be expected      tually be exposed.
for active formulations. Furthermore, formulations           The evasion of formulation and toxicological
have more subtle but nonetheless differential char-       testing by `herbal medicine' manufacturers is com-
acteristics, such as whether the tablet was com-          pletely illogical in this context. For example, Peta-
pressed at a higher or lower pressure.                    sites hybridus (the butterbur or bog rhubarb)
                                                          contains well-characterized carcinogens. Butterbur
                                                          extract tables are sold as chronic oral therapies for
        THE FORMULATION CHOICE                            bladder dysfunction and migraine prevention, and
                                                          claimed to be innocuous on chemical purity
Clearly, the formulation chosen for particular            grounds. Similarly, oral melatonin has an absolute
drugs is not random. Furthermore, the degree to           bioavailability of about 15% maximum, and has
which it is critical varies from drug to drug. For        been withdrawn in the UK and Japan due to con-
example, hydrocortisone is available for at least         cerns about safety (DeMuro et al 2000).
seven routes of administration, as tablets, several          Various physicochemical properties of bulk drug
creams and ointments, intraocular solutions, sup-         can be measured. Some will be reasonably familiar
positories, intra-rectal foams, injections, and ear-      from medical school biochemistry, e.g. the pH of
drops. Even newer drugs, with fewer indications           drug solutions, and the pK of particular molecules
than hydrocortisone, seek greater market accept-          in aqueous solutions. Log P is a measure of lipo-
ability by providing a variety of alternative formu-      philicity, usually being measured as the octanol/
lations (e.g. sumatriptan is available as an injection,   water distribution coefficient when the aqueous
intranasal spray, suppository, and tablets).              phase is buffered at pH 7.4. Powder density is the
   One commonly-used principle is to target drug          ratio of weight to volume occupied by a powder;
delivery to the organ where beneficial effects are        some powder particles pack together more effi-
likely to occur. This can achieve:                        ciently, and a comparison between table salt and
                                                          talcum powder is an illustration. Particle size and
. Relatively fast onset of effect.                        distribution is measured typically using an infra-red
. Locally high drug concentrations.                       device. Solubilities in various solvents are also
. Relatively low systemic drug concentration,             helpful to those whose task is to make drugs into
  avoiding toxicity.                                      prescribable pharmaceutical formulations. Hygro-
                                                          scopicity is a measure of the capability of a drug to
Probably the most common applications of this             absorb water from the atmosphere; such drugs gain
principle are the administration of b-adrenergic          weight with time, and are often less stable than
agonists bronchodilators by inhalation, and the           drugs which do not have this capability. Standard
use of topical hydrocortisone creams.                     manuals such as the Merck Index provide many of
   Impurities and degradants may possess their            these data.
own toxicological properties. Early in develop-
ment, the structures of these impurities and degra-
dants may be poorly characterized. Typically, both                  SPECIFIC FORMULATIONS
bulk drug and finished product become more re-
fined as clinical development proceeds. Thus, in                 Tablets, Syrups, Wafers and Oral
order to preclude any new toxicology problems                               Suspensions
developing later during clinical development, it is
common practice to use the least pure bulk drug for       The excipients of oral formulations vary according
toxicology studies. This is commonly accomplished         to the physical state of the desired finished product,
                                               PHARMACEUTICS                                                 47

as well as how it must be manufactured. Binders are       . When there are major metabolites, then collec-
used to hold the various components together, and           tions should accommodate at least three half-
include starches and polyvinylpyrrolidine (to which         times of their elimination.
many dogs exhibit a species-specific allergy).
Bulking agents (sometimes called dilutants, or, con-      In this case, the Tmax , Cmax , AUC, and half-time of
fusingly for a solid formulation, diluents) include       elimination for parent drug and principal metabol-
lactose and cellulose, and increase tablet weight,        ites become the end-points of the study. For com-
which may improve production uniformity. Coat-            bination therapies, these end-points have to be
ings are often sugar- or cellulose-based and may be       measured and fulfilled for all active components,
employed when a drug tastes foul, or to create a          and they should not be administered separately.
particular color scheme. Silica and starch may also          The regulation does not define what a significant
be used to improve the flow of powder in mass             difference might be, although a commonly applied
production, when it is known as a pro-glidant, and        standard seems to be a formulation whose mean
stearic acid salts are used to enable tablets to escape   Tmax , Cmax , and AUC is within 20% of the reference
the press when finished, this being an unusual use        material, and is also within the 95% confidence
of the term lubricant.                                    interval. However, these limits are tightened when:
   The large majority of generic drugs are tablets
designed for systemic delivery of drug. The regula-       . The therapeutic ratio of the drug is low.
tion does not require that innovative and generic         . The drug has solubility < 5 mg/ml.
drugs have identical excipients. Exemption from           . Tablet dissolution in vitro is slower than 50% in
demonstration of efficacy may be obtained, pro-             30 min.
vided that bioequivalence can be demonstrated             . The absolute bioavailability is < 50%.
with another drug that is the subject of an ordinary      . There is extensive first-pass metabolism that
NDA. The regulation Title 21, Code of Federal               makes rate of absorption, as well as extent, a
Regulations (21 CFR) 320.1 ± 320.63 defines bioe-           factor governing exposure.
quivalence as: `. . . demonstrated if the product's       . There are special physicochemical constraints,
rate and extent of absorption, as determined by             such as chelation, complex formation, or crys-
comparison of measured parameters, e.g. concen-             tallization to consider (see 21CFR320.33).
tration of active drug ingredient in the blood, urin-
ary excretion rates, or pharmacological effects, do          It should be noted that there are other ways to
not indicate a significant difference from the refer-     demonstrate bioequivalence, and in the United
ence material's rate and extent of absorption'. Al-       States of the Food and Drug Administration
though, theoretically, it may be possible to              (FDA) will advise, in advance, on adaptations to
demonstrate bioequivalence using a well-validated         `standard' measures of bioequivalence, in the inter-
in vitro or animal method, per 21CFR320.24 (ii)±          ests of promoting the right clinical trial design. For
(iii), the data that has traditionally been most per-     example, two oral formulations can be compared
suasive has been a pharmacokinetic comparison of          with an intravenous dose. If the drug is concen-
the generic and reference drugs in man. The com-          trated in the urine, but has negligible concentration
monest study design is to compare two oral formu-         in the blood (e.g. nitrofurantoin antibiotics), then
lations with the following optimal design features        urine sampling with a frequency that matches the
(21CFR320.26):                                            blood samples above can be employed. Multiple
                                                          dose bioequivalence study designs are also avail-
. Normal volunteers in the fasting state.                 able. Rarely, the testing of bioequivalence at
. Single-dose, randomized, cross-over, with well-         steady-state is required, because patients cannot
  defined reference material.                             be withdrawn from therapy, and normal volunteers
. Collection of blood samples for at least three          would face an undue hazard. The regulation also
  half-times of elimination, and at a frequency           permits bioequivalence to be demonstrated using
  that captures distribution phase, Cmax and              chronopharmacological effect, i.e. pharmacody-
  Tmax , and at identical times for the two formu-        namic data, and specifies the frequency and timing
  lations.                                                of endpoints in much the same way as for blood

samples (see above). This can be useful for drugs         with a variety of infusional agents. In most situ-
that are not intended to be absorbed systemically,        ations, the gas is made by vaporizing a liquid (usu-
e.g. the rate of onset and offset of topical anesthesia   ally a halogenated hydrocarbon or ether).
to a standardized experimental injury.                    Validated vaporizers, usually designated for use
   Bioequivalence studies do not always require the       with a single liquid, are required. In addition,
filing of an Investigational New Drug application         many patients are paralyzed during surgery and
(IND). An IND is needed always if the generic drug        mechanical ventilators (including a hand-squeezed
is without an approved innovator in the USA, is           bag) must consider drug economy, occupational
radioactive, or is a cytotoxic. However, when single-     exposure of the staff, carbon dioxide scrubbing,
or multiple-dose studies do not exceed the approved       and other pharmacokinetic features that are rarely
clinical dose sizes, and when there will be retention     encountered elsewhere. This is quite apart from the
samples available for inspection, then an IND need        usual considerations of tidal and minute volumes,
not be filed. An IND is needed for a multiple dose        oxygen supply, blood pressure management, etc.
bioequivalence study, when a single-dose study has        Gas flow can be measured with various devices,
not preceded it. The usual protections for human          and exhaled gas concentrations (including for
subjects are required, and, of course, these include      carbon dioxide) can now be measured real-time.
an Institutional Review Board approval.                   Malignant hyperthermia is an adverse event that
   By definition, sustained release formulations          is almost always associated with the inhalation of a
differ pharmaceutically and pharmacokinetically           halogenated hydrocarbon, and which can be
from the innovator drug. Delayed or sustained re-         treated with dantrolene (Strazis and Fox, 1993).
lease oral formulations are used for chronic ther-           The theory relating physicochemical properties
apy, and may have two principal advantages: (a)           of gases and the partial pressure at which they can
reduction in dose frequency (and thus, hopefully,         achieve anesthesia is beyond the scope of this chap-
improved compliance; see Chapter 21); and (b) re-         ter. Indeed, this question hinges on how the state of
duction of Cmax for a standard AUC, which can             anesthesia can itself be measured, one of the more
improve tolerability when adverse events are              difficult pharmacodynamic endpoints in pharma-
plasma concentration-related. The demonstration           ceutical medicine. One wit, also a famous cardi-
of bioequivalence usually hinges on the following         othoracic anesthesiologist, has commented: `If
factors: (a) equivalence of AUC to an innovator           you can tell me what consciousness is, then I will
drug at steady state; (b) the absence of any chance       tell you what anesthesia is!'.
of `dose dumping'; (c) consistency of performance            There are some uses for gaseous drugs outside of
from dose to dose [see 21CFR320.25(f )].                  surgery. Nitrous oxide and oxygen mixtures are
   Various tactics can be employed in the pursuit of      sometimes used as analgesics during labour, or
the sustained-release strategy, including mixtures        when transferring patients in pain by road or heli-
of granules with different thicknesses of polymer         copter. In very cold weather, nitrous oxide can
coating, all contained within a single capsule; os-       liquify, reducing the delivered dose; shaking the
motically driven tablets, which slowly release drug       container helps.
through a small aperture during the entire traverse          Helium/oxygen mixtures are used to improve
of the small bowel; layered tablets with deliberately     oxygenation in patients with subtotal airways ob-
delayed dissolution; and tablets designed to release      struction, exploiting the superior flow characteris-
their contents only in relatively alkaline environ-       tics of the lighter gas. The use of this mixture as
ments (i.e. beyond the ampulla of Vater). It is           prophylaxis against nitrogen narcosis in the deep
illogical to seek sustained-release formulations for      sea, while minimizing fire hazard, is also well-de-
drugs with relatively long half-times of elimination.     scribed. Fire hazard from oxygen (arguably a gas-
                                                          eous drug under some circumstances) is important;
                                                          the disastrous fire inside the command capsule of
                       Gases                              Apollo 3, during a lift-off rehearsal on the pad at
                                                          Cape Kennedy, took place within a pure oxygen
Gases are usually administered in the context of          atmosphere. Reduction in total atmospheric pres-
general anesthesia, either alone or in combination        sure, to reduce fire hazard, has since been employed
                                             PHARMACEUTICS                                                49

in all pressurized American space vehicles, but they   man. Furthermore, in vitro studies use apparati
still contain supra-atmospheric partial pressures of   that do not model the anatomy of the human re-
oxygen.                                                spiratory tree, let alone the diseased respiratory
                                                       tree. The British Association for Lung Research
                                                       have recognized this complexity and issued a con-
Metered Dose Inhalers and Nebulized Drugs              sensus statement (Snell and Ganderton 1999)
                                                       which recommends, at a minimum, a five-stage
In general, and with a few rare exceptions (see        collection apparatus, examination of a range of
below), the inhaled route of administration is the     particle sizes (0:05 À 5 mm), a range of flow rates
most difficult that is commonly encountered.           and patterns to mimic the various physiological
Metered dose inhalers and nebulizers are considered    states, the development of an apparatus modeled
together here because they have many features in       on the shape of the human pharynx, the concomi-
common. These formulations, in both cases, are         tant use of swallowed activated charcoal in clinical
administered as aerosols of drug solution.             studies to minimize absorption by swallowing drug
   It is customary, in textbooks for a general audi-   impacting on the oropharynx, regional lung assess-
ence, to insert at this point a graph that relates     ments in three dimensions, and further develop-
aerosol particle size to the various levels of the     ment of useful statistics to describe such findings.
airway where drugs can impact. Particles                  The metered dose inhaler has been in use for
> 10 mm are stated to be commonly impacted in          about 50 years and doubtless forms the mainstay
the pharynx; < 5 mm particles are assumed to be        for the treatment of asthma, as well as for patients
ideal for alveolar delivery, and < 0:05 mm particles   with chronic bronchitis with a reversible compon-
are said not to impact at all, being liable to be      ent. Great technical challenge has been experienced
exhaled. This is an oversimplification. Particle de-   in the last few years, due to the need to change their
position is dependent on a large number of other       propellants into non-fluorohydrocarbon materials,
factors, attested to by a vast literature that has     as part of the global effort to protect the atmos-
accumulated over at least the last 25 years, strad-    pheric ozone layer. The contribution of metered
dling the border with the disciplines of pulmonol-     dose inhalers to this problem, in comparison
ogy and industrial hygiene. Other factors governing    to vapour escaping from refrigerators and car air
particle deposition (and example studies) include:     conditioners, must have been negligible. Nonethe-
                                                       less, indirectly, these huge costs are now being
. Coughing (Camner et al 1979)                         borne by healthcare systems worldwide. The clin-
. Mucociliary action (Lippmann et al 1980)             ical studies had to rely on efficacy parameters be-
. Exercise and minute ventilation (Bennett et al       cause of the inability to quantitate lung deposition
  1985)                                                and the general aim of avoiding systemic drug ab-
. Mucous production and ability to expectorate         sorption.
  (Agnew et al 1985)                                      A wide variety of nebulizers are now available.
. Apnoeic pause at the end of inhalation (Legath       They all have their own physicochemical proper-
  et al 1988)                                          ties. In the absence of the ability to quantitate lung
. Whether a patient is actually having an asthma       deposition, the Food and Drug Agency (FDA) has
  attack (Patel et al 1990)                            now said that it will approve only combinations of
. Breathing pattern, airway calibre, spacers and       new drugs with specified nebulizers; labeling for a-
  reservoirs (Bennett 1991)                            dornase is the first to exhibit this change in policy.
. The physicochemical properties of the drug(s)        This requires that the clinical development plan be
  (Zanen et al 1996)                                   implemented, as early as possible, with the nebu-
. Lung morphometry (Hofmann 1996)                      lizer that is intended to be marketed.
. Sampling techniques, on which exposure calcu-           Inhalational toxicology is generally required to
  lations are based (Cherrie and Aitken 1999).         support inhalational clinical trials and product ap-
                                                       provals. This highly specialized field requires the
The truth is that it is practically impossible to      validation of the nebulizing system for each drug
measure the lung deposition of inhaled drugs in        and species separately. The scaling from animals to

man in the selection of initial doses is something of   and aerosols for local absorption. This intent of
an imponderable, given the complexities in meas-        this guideline is to facilitate the development
uring lung deposition described above.                  of generic products for use by this route of adminis-
                                                        tration. This guideline has been challenged on
                                                        several scientific and technical grounds (e.g. Harri-
               Oral Transmucosal                        son 2000). While this situation is not yet resolved,
                                                        many guidelines in the USA remain in draft status
The best drugs for oral transmucosal administra-        for long periods of time, and are treated as defini-
tion are those that do not taste bad, and which have    tive.
high potency. For example, among opioids, the two
drugs that have been successfully developed using
this type of formulation are buprenorphine and                      Transdermal vs. Topical
fentanyl. Formulations vary, but include sublin-
gual pellets, chewable gums, and solid formulations     The principal distinction between transdermal and
that are held on a stick, somewhat like a lollipop.     topical drugs is that only the former are intended
Most wafer formulations dissolve in the mouth and       for systemic delivery. Both formulations are, how-
are actually converted into a solution for swallow-     ever, subject to the same skin irritancy testing prior
ing and gastrointestinal absorption (e.g. rizatriptan   to human exposures.
wafer). Benzocaine lozenges are intended for the           The skin is biologically intended to be a barrier.
same purpose but to dissolve more slowly, thus          Evading this barrier is not easy, because drugs must
bathing the oesophagus as a symptomatic treat-          traverse dead epithelium and live dermis; the
ment (e.g. for radiation oesophagitis); a similar       former is hydrophobic, while lipophilic drugs
approach is used with anti-fungal drugs.                tends to form a reservoir in the latter. As in oral
                                                        transmucosal administration, potent drugs, with
                                                        modest requirements for mass absorbed and rea-
                    Intranasal                          sonable lipophilicity, are the best candidates for
                                                        transdermal delivery. Fentanyl, nicotine, and sco-
The absorptive capacity of the nasal mucosa has         polamine are good examples.
been known for centuries. Even if unexploited by
pharmaceutical scientists, the abuse of cocaine (in-
cluding by primitive peoples), and nicotine (snuff )                           Rectal
has routinely used this route of administration for
systemic delivery. Vast annual tonnages of anti-        The use of suppositories is probably one of the
allergy and decongestant drugs are now adminis-         clearest examples of cross-cultural differences in
tered to the noses of the developed world. These are    the approach to pharmaceuticals. A surgeon on a
intended to treat local symptoms, and avoidance of      famous ocean liner has commented that: `Part of
systemic absorption is a favourable feature. a-         the problem of stocking one's pharmacy is that one
Adrenergic agonists, antihistamines, and steroids       needs three times as many drugs as when working
probably lead the list for this topical route of ad-    on land: tablets for the Brits, shots (injectables) for
ministration.                                           the Yanks, and suppositories for the French!'
   Therapeutically, interest in the nasal mucosa for    However, the route of administration is eminently
systemic absorption of drugs initially centered on      logical, e.g. for the acute treatment of migraine,
its capability to absorb small to moderately sized      where drugs are often vomited.
polypeptides. For example, vasopressin-like drugs          In the UK and USA, peri-operative antibiotics
(nonapeptides) may be used to treat diabetes insi-      (especially metronidazole), theophylline for pediat-
pidus in patients with panhypopituitarism. This         ric nocturnal asthma, and topical treatments for
avoids repeated parenteral injections, and avoids       proctitis and inflammatory bowel disease are the
the digestive capacity of the gut.                      most commonly used rectally-administered drugs.
   In the USA, the FDA has issued a draft guideline        Paraldehyde remains an effective way to abort
on establishing the bioequivalence of nasal sprays      seizure in pediatric patients in the emergency room,
                                               PHARMACEUTICS                                                 51

without the need to find a vein. Use a glass syringe.    the formulation. If a drug cannot be dissolved in
In the USA, diazepam can be administered by the          a concentrated manner in a suitable vehicle, then
same route, for the same purpose.                        often dose size must increase. Intravenous injec-
                                                         tions of penicillin-type antibiotics are much more
                                                         comfortable than when the same dose is adminis-
                      Vaginal                            tered intramuscularly.
                                                            Organic solvents are often used to enhance the
An intravaginal suppository is more accurately           rate of absorption from the subcutaneous or intra-
termed a pessary. Most are designed for topical or       muscular site of administration. For example,
local use, including for Candida albicans and Tri-       benzyl alcohol and sodium benzoate are used to
chomonas infections, as well as for preparation of       dissolve diazepam, and extravasation of this for-
the cervix prior to induction of labour. Contracep-      mulation is not as serious a problem as for thio-
tive devices are outside of the scope of a chapter on    pental.
pharmaceutics, although the nonoxyl-containing              Rarely, adverse events are reported when an ap-
sponge pessary is a unique formulation.                  parently innocuous formulation is administered by
                                                         the wrong route. Usually these problems arise be-
                                                         cause of excipients that the typical physician takes
            Injectates (s.c., i.m, i.v.)                 little interest in. Intravenous remifentanil is formu-
                                                         lated with glycine, and hence it is not well-suited for
The solubility of a drug, and the compatibility of a     epidural administration.
particular solvent with the site of injection, are          The development of an injectate is often one
inter-related factors governing the suitability of       tactic used for obtaining a patent. Even though a
this route of administration, and the pharmaceut-        composition of matter patent (i.e. the structure of
ical formulation that is employed.                       the drug molecule itself) may be old, the develop-
   Water-soluble drugs are usually also hygro-           ment of a non-obvious injectate, and its method of
scopic, and need to be stored in an anhydrous            use for a new indication, may be sufficient to
environment to maintain their stability. These are       obtain a further patent and thus extend effective
most easily supplied as lyophilized powders, which       proprietary coverage. Such patents are usually
can be reconstituted with water or saline immedi-        stronger in North American than in European jur-
ately prior to injection. This applies to early devel-   isdictions.
opment as well, when the filtration of drug solution
through micropore nylon filters prior to lyophiliza-
tion can be a convenient way to ensure sterility.                             Packaging
Lyophilizates in stoppered vials can also be sub-
jected to g-irradiation to ensure sterility. Stability   The selection of an inert package is an essential part
studies should include not only the range of tem-        of the pharmaceutical development of a drug. There
peratures and humidities (see below), but also with      are many standard stoppers, plastic and glass
the vials inverted.                                      bottles, etc., with which regulatory authorities are
   Intravenous formulations are probably the least       very familiar. Stability studies must be conducted,
demanding of all injectates. Solutions of thiopental     of course, in the same sorts of packaging.
sodium are routinely administered at the induction          Packaging nonetheless varies, and over a period
of anesthesia, but are extremely alkaline and would      of months or years an apparently impervious ma-
be very damaging if administered subcutaneously          terial may, permit the ingress of water. Foil wraps
or intramuscularly. Unfortunately, this occurs oc-       are generally available for all tablets, although they
casionally as an iatrogenic adverse event when the       are inconvenient to arthritic hands; these are usu-
injection (pH 9) extravasates; serious injury can        ally the most impervious of all materials. PVC blis-
easily occur to the structures in the cubital fossa      ter packs are at the other end of the spectrum;
(including the median nerve).                            Padfield (1985) has provided one example where a
   The route of administration may also be               0.8% increase in tablet weight within a PVC pack-
governed by tolerability aspects associated with         age occurred within 12 weeks.

   Drugs, both investigational and prescription,          achievements, even if the drugs involved are very
are today transported over great distances. Airlines      familiar.
often advertise their cargo holds as pressurized             What are we likely to see in the future? Novel
and temperature-controlled, but even so require           pharmaceutical formulations seem to fall into two
special arrangements for the conveyance of                groups, those being used for gene therapy and
livestock. The potential for condensation in the          those being used elsewhere.
air, after degradation when the pallet sat for several       Investigational gene therapies are now common-
hours on the unshaded tarmac in Dakkar, is                place, and are comprised of two components: the
great.                                                    DNA itself (the `construct') and usually a method
                                                          of delivery (`the vector'). Naked DNA can be
                                                          injected but its expression is inefficient. Vectors
              STABILITY TESTING                           may include viruses. However, such viruses have
                                                          to be human, and their attenuation sometimes is
Stability testing of drugs is an entire subspecialty      lost after administration, leading to very serious
within the pharmaceutical professions and cannot          adverse events. Non-viral vectors can include
be covered in depth here. However, suffice it to say      targeted liposomes, microspheres, and emulsions.
that it is the pharmacist's duty to ensure that drugs        There are several other examples of truly unique
are being used when they have been tested with            formulations or routes of administration that we
various challenges, which may be considered as            may expect to be further exploited in the future.
factorial combinations of:                                AIDS-associated infective retinitis is treated with a
                                                          drug administered by intraocular injection, and the
. Low and high temperatures.                              current parlous state of retinal detachment treat-
. Low and high humidity.                                  ments suggests that this route of administration
. Exceeding the labeled drug shelf-life.                  may find wider use. It turns out that cell mem-
. In contact with all feasible components of the          branes become leaky when exposed to high volt-
  packaging (e.g. both the glass and the stopper          ages: otherwise insoluble or excluded drugs can
  of a vial, the latter by inverted storage).             enter the cell under these conditions, and this uses
. Exposure to bright and subdued light (in some           a multi-tined stimulator, known as an electropora-
  case clear and amber glass bottles).                    tor. Needleless injectors have been available for
                                                          decades, yet still seem to be under-used (the needle-
It is these data that justifies approval and con-         less injector used by Dr `Bones' McCoy of the `USS
tinued marketing of a drug that complies with the         Enterprise' is clockwork, develops several thousand
`quality' criterion of the oft-quoted triad, `safety,     pounds pressure per square inch, and feels like a
efficacy, quality'. This is usually not a trivial exer-   mild middle-finger percussion when used over the
cise.                                                     deltoid).

     INNOVATION IN PHARMACEUTICS                                             SUMMARY

Innovation has always been a very visible activity        The objective of this chapter has been to provide
in pharmaceutics. As noted above, we very rarely          the pharmaceutical physician with some appreci-
administer powders out of paper cones today. The          ation of the complexity of pharmaceutical develop-
dry-powder inhaler used for sodium cromoglycate           ment. Understanding the vocabulary will help
was developed because that drug is almost insol-          participation in team meetings, where pharmaceut-
uble, but is also now being revived in the post-          ical and clinical development must be coordinated.
hydrofluorocarbon era. The intravenous emulsion           A chapter on this scale will never equip a pharma-
of propofol is also unique, again being invented out      ceutical physician to conduct pharmaceutical de-
of necessity. Pharmaceutical physicians should            velopment. At the very least, it should now be clear
not underestimate the scale of these technological        that a drug is not a drug is not a drug.
                                                        PHARMACEUTICS                                                            53

                       REFERENCES                                   Legath L, Naus A, Halik J (1988) Determining the basic
                                                                      characteristics of aerosols suitable for studies of deposition
                                                                      in the respiratory tract. J Hyg Epidemiol Microbiol Immunol
Agnew JE, Pavia D, Clarke SW (1985) Factors affecting the             32: 287±97.
  `alveolar deposition' of 5 microns inhaled particles in healthy   Lippmann M, Yeates DB, Albert RE (1980) Deposition, reten-
  subjects. Clin Phys Physiol Meas 6: 27±36.                          tion, and clearance of inhaled particles. Br J Indust Med 37:
Bennett WD (1991) Aerosolized drug delivery: fractional depos-        337±62.
  ition of inhaled particles. J Aerosol Med 4: 223±7.               Padfield JM (1985) Making drugs into medicines. In Burley
Bennett WD, Messina MS, Smaldone GC (1985) Effect of exer-
                                                                      DM, Binns TB (eds), Pharmaceutical Medicine. Arnold:
  cise on deposition and subsequent retention of inhaled
                                                                      London and New York; 51.
  particles. J Appl Physiol 59: 1046±54.
                                                                    Patel P, Mukai D, Wilson AF (1990) Dose±response effects of
Camner P, Mossberg B, Philipson K, Strandberg K (1979)
                                                                      two sizes of monodispersed isoproterenol in mild asthma. Am
  Elimination of test particles from the human tracheobron-
                                                                      Rev Resp Dis 141: 357±60.
  chial tract by voluntary coughing. Scand J Resp Dis 60: 56±62.
                                                                    Snell NJ, Ganderton D (1999) Assessing lung deposition of
Cherrie JW, Aitken RJ (1999) Measurement of human exposure            inhaled medications. Consensus statement from a workshop
  to biologically relevant fractions of inhaled aerosols. Occup       of the British Association for Lung Research, held at the
  Environ Med 56: 747±52.                                             Institute of Biology, London, UK, April 17 1998. Resp Med
DeMuro RL, Nafziger AN, Blask DE et al (2000) The absolute            93: 123±33.
  bioavailability of oral melatonin. J Clin Pharmacol 40: 781±4.    Strazis KP, Fox AW (1993) Malignant hyperthermia: A review
Harrison LI (2000) Commentary on the FDA draft guidance for           of published cases. Anesth Analg 77: 297±304.
  bioequivalence studies for nasal aerosols and nasal sprays for    Zanen P, Go LT, Lammers JW (1996) Optimal particle size for -
  local action: an industry view. J Clin Pharmacol 40: 701±7.         2 agonist and anticholinergic aerosols in patients with severe
Hoffmann W (1996) Modeling techniques for inhaled particle            airflow obstruction. Thorax 51: 977±80.
  deposition: the state of the art. J Aerosol Med 9: 369±88.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                             Non-clinical Toxicology
                                                  Frederick Reno
                                               Merritt Island, FL, USA

The evaluation of the safety of new pharmaceutical              addition, all non-clinical toxicology studies that are
agents through non-clinical studies is a critical               intended to support clinical trials or marketing ap-
aspect of any development program. Usually in the               plications must be conducted in compliance with
discovery stage, or what can be considered the `re-             good laboratory practices (GLP; Federal Register
search' phase of research and development, either in            December 1978). The volume edited by Williams
vivo or in vitro studies have established the pharma-           and Hottenderf (1997) provides more information
cological profile of the new drug and a rationale for           on the subjects that are discussed below.
its potential clinical efficacy. At this stage, the po-
tential agent can be considered a new chemical entity
(NCE) or perhaps an analog or metabolite of an                      CONSIDERATIONS RELATED TO THE
existing one. Preliminary studies are also made                       CLINICAL DEVELOPMENT PLAN
with respect to drug absorption, metabolism, and
excretion. In many companies, drug metabolism is a              The nature, timing, and extent of the initial non-
separate entity from the toxicology function but, for           clinical toxicology program depend on the clinical
the sake of completeness of this chapter, a discussion          development plan that it must support. The ICH
of this important research area will be included. At            guidelines further specify the extent and duration
some point, a decision is made to move the agent                of non-clinical studies that are required to initiate
into the `development' phase, and the initiation of             or continue clinical studies (Federal Register No-
non-clinical toxicology studies necessary to estab-             vember 1997, and see below). Therefore, it is im-
lish safety for initial clinical trials is begun.               portant that the clinical development plan, at least
   Over the past two decades, separate regulatory               the initial stages, be clearly delineated.
authorities in the USA, Europe, and Japan have
established their own guidelines for the types and
                                                                               Initial Clinical Studies
extent of preclinical studies that are necessary. Al-
though often quite detailed, these jurisdictions were           Usually, the initial clinical goals are to study toler-
rarely similar, and designing a non-clinical toxicol-           ability and to provide initial pharmacokinetic as-
ogy program that would be universally accepted                  sessments. These studies may only involve single
was difficult, if not impossible. The International             doses of the drug administered to normal volun-
Conference on Harmonization (ICH), a tripartite                 teers. Such a clinical study would require a re-
group that consists of regulators and pharmaceut-               stricted set of toxicity studies to support the safe
ical company representatives from the three geo-                use of the drug in this situation. On the other hand,
graphical areas, has been meeting for several years             some companies achieve economies by having the
to address the harmonization of many aspects of the             initial toxicology program be sufficient to support
drug development process. These meetings have                   not only initial clinical studies but also Phase II.
resulted in the issue of many guidelines (in either             The toxicology studies may then involve repeated
draft or final form) for non-clinical studies and               doses over a period of weeks. Thus, the initial
applicable in all three jurisdictions. These guide-             clinical studies must be determined before the
lines will be identified throughout this chapter. In            non-clinical program can be designed.

            Initial Proof of Principle                    the filing of an investigational new drug applica-
                                                          tion (IND) for the first study in man.
In most cases, a proof of principle (i.e. initial indi-
cation of clinical efficacy) during early Phase II
clinical studies will require clinical treatment for                  INITIAL NON-CLINICAL
some period of time, ranging from days (diagnostic                      CONSIDERATIONS
agents, etc.) to weeks or months (for other types of
drug). Since exposure of patients in clinical trials      Of equal importance to the successful initiation of a
(in most cases) cannot last beyond the duration of        non-clinical program are several factors that can
the animal studies, careful consideration of the          have a great impact on the rapidity with which a
development schedule must be made, so that no             program can be implemented. Experience has
delays are caused through lack of toxicological           shown that overlooking the importance of these
coverage. This requires that the appropriate pre-         factors can result in unanticipated delays, costing
clinical reports are available prior to the planned       time and money.
initiation of the clinical trial.

                                                                        Formulation Aspects
             Enrollment of Women
                                                          It is desirable for the pivotal non-clinical studies to
Most regulatory agencies now request that women           be carried out using the proposed clinical route of
be enrolled into the clinical studies as early in Phase   administration and with a formulation that best
II as possible. Since thalidomide, reproduction,          approximates that anticipated for initial clinical
and teratology studies have been required prior to        usage, although, of course, this is unlikely to be
enrollment of large numbers of women in clinical          the formulation that is eventually marketed when
studies. In some cases, depending upon the pro-           the program is successful. Factors such as method
posed indication for the drug, postmenopausal or          of synthesis, excipients and appropriate vehicles
otherwise reproductively incapable women can be           usually evolve from bench-scale drug supplies and
used. However, the timing of the enrollment of            simple vehicles, to more sophisticated galenicals as
women needs to be understood well in advance so           the program proceeds. Scale-up of manufacturing
that the lack of appropriate non-clinical reports         processes can result in bulk drug with different
does not hinder clinical development.                     impurities, and adverse effects may be due to
                                                          parent drug, metabolites or impurities. Further-
                                                          more, tablets or capsules cannot be given to most
     CONSIDERATION OF REGULATORY                          animal species, and the non-clinical studies are
              STRATEGY                                    therefore carried out using dosing solutions or sus-
                                                          pensions. The type of formulation can affect the
It is common, particularly for American com-              pharmacokinetics of the drug, thus altering the
panies, to carry out initial Phase I studies abroad.      toxicological profile, making comparison of animal
This is because these studies can often be initiated      and human pharmacokinetics, in the context of the
outside the USA with little regulatory involvement,       formulations used, into a critical element in the
and with safety considerations being reviewed by          evaluation of human safety.
an Institutional Review Board or Ethics Commit-
tee. Consequently, the company accepts the entire
responsibility for any hazards or risks to the study                     Impurities/Stability
subjects. This has the effect of allowing Phase I
studies to be initiated more rapidly and thus             Early-stage small-scale synthesis methods will often
obtaining information on preliminary safety and           create a different profile of impurities or degra-
pharmacokinetic data earlier. Awareness of such           dants than drug supplies produced by scaled-up
a strategy dictates a different scenario to the pre-      processes. Every batch of drug used in non-clinical
clinical manager than would a strategy requiring          studies must have a certificate of analysis that
                                        NON-CLINICAL TOXICOLOGY                                            57

clearly specifies the purity levels and the quantities   method for the determination of parent drug (and
of impurities (which may include residual solvents,      possible major metabolites) in solvents and plasma,
unreacted starting materials or degradants). The         usually validated for multiple non-clinical species.
impurities must be reviewed in terms of the poten-
tial contribution that they can make to toxic effects
that may be manifested in the non-clinical studies.                    Appropriate Species
There are ICH guidelines that pertain to impurities,
and the extent to which additional toxicity studies      In the early stages of the development of any drug,
need to be performed in impurities (Federal Regis-       there is little, if any, information on which to make
ter January 4 1996, March 19 1996).                      a scientific judgment relative to the most appropri-
   Of equal importance is the stability of the drug in   ate animal species for non-clinical studies that will
the non-clinical formulation. This can determine         best predict responses in the human. In these cases,
whether the non-clinical formulations must be pre-       since regulatory agencies require the use of both a
pared daily or can be prepared weekly. If drugs are      rodent and a non-rodent species, the typical ap-
to be given orally, it is obvious that they must be      proach would be to use the rat and the dog for the
resistant to degradation of gastric acids and must       toxicity studies, and mice or rabbits for other more
be stable in the formulation itself (water, carbox-      specialized studies. Primates may be needed when
ymethylcellulose suspensions, etc.). As will be dis-     there is availability of considerable background
cussed in more detail later, this requires the           data in these species in terms of the parameters of
availability of an analytical method at the earliest     interest (hematology, blood chemistry, histopath-
stages of development.                                   ology, etc.). When candidate drugs are proteins
                                                         (e.g. animal-derived monoclonal antibodies), then
                                                         antibody formation may be major issue and may
               Drug Requirements                         dictate the choice of species. For example, it may be
                                                         known that only the chimpanzee does not develop
The amount of bulk drug that is typically required       neutralizing antibodies to the drug, which would
to carry out the non-clinical studies may be a big       lead one to select that species as the non-clinical
surprise, in comparison to that needed for initial       model. Topical formulations are another special
clinical studies. While many biologically derived        case, and the rabbit is commonly employed. The
drugs may require relatively small quantities, due       selection of the animal species for the non-clinical
to the potency of the material or the limited            program is often not straightforward.
number of non-clinical studies that are possible
(see below), a typical program need for `first time
in man' drugs that are relatively non-toxic may          TOXICOLOGICAL SUPPORT PRE-IND AND
require 2±3 kg of active drug. For many companies,           FOR PHASE I CLINICAL STUDIES
this can be difficult from either a manufacturing
standpoint (small quantities synthesized prior to        The preliminary evaluation of the safety assessment
scale-up) or cost.                                       of any new drug requires multiple studies, some of
                                                         which evaluate general and multiple endpoints
                                                         (such as toxicity studies). Other studies evaluate
  Analytical Methods for Dose and Plasma                 more specific and defined endpoints (such as muta-
              Determinations                             genicity studies and safety pharmacology studies).
                                                         Drugs that are derived from a biological origin,
GLP regulations require confirmation of the po-          such as proteins, monoclonal antibodies, or drugs
tency of all formulations used in non-clinical           produced by biological vectors (or what are gener-
studies. Furthermore, current ICH guidelines also        ally referred to as `biotechnology products'),
require toxicokinetic data (i.e. animal pharmacoki-      present additional problems that require a signifi-
netics determined at one or more time points during      cantly modified approach. The ICH guidelines
a non-clinical toxicology study). Both the potency       recognize that unique approaches may be needed,
and toxicokinetic assays require an analytical           has addressed this in a further guideline (ICH 1997),

and poses additional problems for the toxicologist                               Table 6.2 Duration of repeated-dose toxicity studies to
(Terrell and Green 1994). This section will elabor-                              support Phase III clinical trials in the EU, and product
                                                                                 marketing in all jurisdictionsa
ate on those studies needed to support the safety of
a typical xenobiotic agent, although the same gen-
                                                                                 Duration of clinical trial    Minimum duration of repeated-dose
eral principles follow for biotechnology products,                                                             toxicity studies
often being necessary but not sufficient. There are
two types of guidelines that must be considered in                                                             Rodents             Non-rodents
initiating the non-clinical program. The first relates
                                                                                  2 weeks                     1 month             1 month
to the types of studies required; the second relates to                           1 month                     3 months            3 months
protocol requirements for the studies themselves.                                 3 months                    6 months            3 months
   The types of studies needed are dictated by na-                               > 3 months                    6 months            Chronic
tional regulatory requirement, although the ICH                                  a
                                                                                    The above table reflects the marketing recommendations in all three
has promulgated a international guideline (Federal                               ICH regions, except that a chronic non-rodent study is recommended for
Register November 25 1997) that is progressing                                   clinical use > 1 month in Japan.
through the final review stage at the present time.
These studies, outlined in Tables 6.1 and 6.2, vary                              initial studies in man. Additional specialized stud-
somewhat by the phase of the clinical trial, and                                 ies might be needed in order to study the potential
may still vary among countries where the trial is                                for an effect that might be characteristic of drugs in
being conducted. The US Food and Drug Admin-                                     the particular class in question (e.g. antibody de-
istration (FDA) has also published guidelines that                               terminations for some biological products; neuro-
outline the requirements necessary to initiate initial                           toxicity studies for drugs acting on the central
clinical studies (FDA 1995). This latter document                                nervous system, etc).
focuses more on the extent of study documentation
required than the study types, and allows for data
to be submitted that is not in final report form.                                                 Acute Toxicity Studies
   The following sections briefly describe the stud-
ies that would typically be performed to support                                 Single-dose studies in animals are an important
                                                                                 first step in establishing a safety profile. Note,
Table 6.1 Duration of repeated-dose toxicity studies to                          however, that the calculation of an LD50 is no
support Phase I and Phase II clinical trials in the EU, and Phase                longer required or scientifically necessary. The
I, II, and III clinical trials in the USA and Japana
                                                                                 aim of single-dose studies is to explore a range of
Duration of clinical trial       Minimum duration of repeated-dose
                                                                                 doses. Identification of doses without drug-related
                                 toxicity studies                                effects, a dose that produces some level of exagger-
                                                                                 ated effect (not necessarily death) that helps iden-
                                 Rodents                Non-rodents              tify potential side effects, and other doses in
                                                                                 between helps all further toxicological (and clin-
Single dose                      2±4 weeksb             2 weeks
 2 weeks                        2±4 weeksb             2 weeks                  ical) tolerability assessments. These studies can be
 1 month                        1 month                1 month                  designed using `up-and-down' or other tactics to
 3 months                       3 months               3 months                 reduce the time and number of animals required.
 6 months                       6 months               6 monthsc                These studies may then guide dose selection for the
> 6 months                       6 months               Chronicc
                                                                                 first repeated-dose studies. Various guidelines for
  In Japan, if there are no Phase II clinical trials of equivalent duration to   the performance of these studies are available, and
the proposed Phase III trials, then non-clinical toxicology studies of the       the ICH has also published its own guideline (Fed-
durations shown in Table 6.2 should be considered.
  In the EU and USA, 2 week studies are the minimum duration. In                 eral Register August 26 1996).
Japan, 2 week non-rodent and 4 week rodent studies are needed. In the
USA, with FDA concurrence, single-dose toxicity studies with extended
examinations can support single-dose human exposures.
  Data from 6 months of administration in non-rodents should be avail-                     Repeated-dose Toxicity Studies
able before clinical exposures of more than 3 months. Alternatively, if
applicable, data from a 9 month non-rodent study should be available
before clinical treatment duration exceeds that supported by other toxi-         Repeated-dose studies are designed to identify safe
cology studies.                                                                  levels of the drug following treatment regimens
                                         NON-CLINICAL TOXICOLOGY                                           59

that are designed to provide continuous exposure          parison with human data as it becomes available
of the animals to the test drug. Ideally, the route       later.
of administration should mimic that planned
in man, and the animal studies should involve
longer durations of exposure and higher doses                          Mutagenicity Studies
than those planned clinically. The type and dur-
ation of specific studies, and which ones are needed      Mutagenicity studies are highly specialized. There
relative to different stages of clinical development,     are multiple hereditary components in both som-
were mentioned previously (Federal Register No-           atic and germinal cells that may be affected by
vember 25 1997). Protocols must specify the               drugs. During the 1970s, it was thought (somewhat
number of animals per group, numbers of groups               È
                                                          naõvely) that these studies may be replacements
and experimental procedures to be carried out, and        for the long and costly carcinogenicity studies
standard versions of these have been available            that are required for many drugs. Although this
for some time. In general, for initial repeated-          goal was never realized, mutagenicity studies none-
dose studies, protocols require the use of three          theless provide useful indications of the ability of
dose groups plus a control, and a minimum of              a drug to alter genetic material, which may later
10 rodents and three non-rodents per sex per              be manifested in studies of carcinogenic or terato-
group. Doses must be selected that will allow for         genic effects (Kowalski, 2001). Genotoxicity stud-
the identification of toxic effects at the highest dose   ies are relatively inexpensive and may also serve,
as well as a no-effect level at the middle or lowest      early in the drug development process, to assure
dose.                                                     drug developers and regulators that no obvious
   Usual experimental procedures include the de-          risk of such adverse effects exists, albeit know-
termination of body weights and food consump-             ing that more definitive studies to evaluate terato-
tion on at least a weekly basis, evaluation of            genic and carcinogenic effects will not come until
hematology and blood chemistry parameters                 later.
during the treatment period, ophthalmoscopic                 An exhaustive review of the various components
examinations, the recording of macroscopic exam-          of a mutagenicity evaluation will not be attempted
inations at necropsy, and the determination of            here. Multiple guidelines are available. Those
organ weights. A complete histopathological exam-         issued by the ICH include general guidelines
ination of tissues from animals is required. In           (Federal Register April 24 1996) and specifics re-
rodent studies, this can take the form of examin-         lated to the core battery of studies required (Fed-
ation of all high-dose and control animals and the        eral Register April 3 1997). Tennant et al (1986)
examination of target organs at the two lower             have summarized the correlation between the
doses. In non-rodent studies, it is typical to exam-      results of a battery of mutagenicity assays and
ine tissues from all animals in the study.                the probability of the material producing a positive
   It is crucial that plasma concentrations of drug       carcinogenic response in long-term rodent studies.
are measured in these studies to allow for determin-      Obviously, mutagenicity studies cannot address
ation of effects on the basis of exposure. Frequently     issues of non-genetic carcinogenicity or teratogeni-
this is a more appropriate measure of comparing           city.
effects in animals and man, since rates of absorp-           Positive results in one or more mutagenicity
tion, distribution, and excretion can vary exten-         assays do not necessarily translate into human
sively between these species. This aspect, now            risks. Mechanistic studies may show that such
commonly referred to as `toxicokinetics', has             responses would not occur in the human cell popu-
been outlined in an ICH guideline (Federal Register       lation, or the concentrations at which positive
March 1 1995). This guideline specifies minimum           responses occurred may far exceed any concentra-
requirements in terms of number of time points            tion of drug that may occur in the clinical setting.
examined, number of animals per time point,               Many drugs are on the market today that have
and the requirements for calculation of various           produced some type of positive response in these
pharmacokinetic parameters such as Cmax ,                 studies and yet it has been concluded that no
AUC, etc. These will become important for com-            human risk is present or the potential risk is

not known (e.g. aspirin causes chromosomal              week, of a 90 day toxicity study. In rodent studies,
breaks). A Fairly standard worked example is pro-       satellite groups of animals are required due to
vided by Fox et al, 1996.                               the blood volumes needed for assay. For larger
                                                        non-rodents the main study animals can usually
                                                        provide the samples. Guidelines have been made
           Pharmacokinetic Studies                      available covering most aspects of the collection
                                                        and analysis of these data (Federal Register
In the early stages of drug development, it is im-      March 1 1995)
portant to identify important parameters that               Lastly, pharmacokinetic assessment requires
relate to the absorption and excretion pathways         tissue distribution studies in non-clinical models
for the drug. In the later stages of development,       to determine the extent of localization of the drug
studies on the extent of tissue distribution and the    in tissues. In some situations, where single-dose
identification of metabolites become important.         tissue distribution studies suggest drug localiza-
Another reason why this is important is that it         tion, a tissue distribution study following repeated
assists the investigator in knowing that the appro-     dosing may be indicated. The conditions under
priate species has been selected for the non-clinical   which such studies may be necessary have been
toxicology program. It is important to human            delineated in an ICH guideline (Federal Register
safety evaluation that the non-clinical models          March 1 1997)
chosen are representative of the metabolism of the
drug in man. Therefore, it is necessary to have
pharmacokinetic information early in the program,                     Safety Pharmacology
so that it can be compared to the data generated in
the early clinical studies.                             Studies related to safety pharmacology (sometimes
   Drug metabolism is a highly specialized field,       confusingly termed `general pharmacology' studies)
and is increasing in sophistication all the time. A     tend now also to be performed earlier in the drug
relatively new technique that is available to the       development process than was previously the case.
preclinical investigator is the use of in vitro         While in some respects considered an aspect of the
methods to establish and confirm similar mechan-        discipline of pharmacology, the purpose of safety
isms in drug metabolism between animals and man         pharmacology is to evaluate the potential pharma-
(see Chapter 10). These procedures involve the use      cological properties that may be unrelated to the
of liver slices and/or liver hepatocyte homogenates     intended indication for the drug. An example of this
and can be done in human and animal cultures at         would be significant effects of a drug on the cardio-
the earliest stages of drug development.                vascular system that may actually be under devel-
   Toxicokinetic data is generally obtained from        opment for the treatment of gastric ulcers.
repeated-dose toxicity studies, and generally deter-       Most major developed countries have stated
mines whether: (a) the plasma concentrations of         guidelines indicating that safety pharmacology
the drug increase in a linear fashion over the          studies are required. Table 6.3 lists the guidelines
range of the increasing doses used in the studies;      from major countries. As can be seen from these
and (b) plasma concentrations increase over time,       guidelines, it is not always clear when such studies
suggesting an accumulation of the drug in plasma        are required. All of the major organ systems need
or tissues; (c) there is a relationship between the     to be evaluated, and therefore studies need to be
plasma concentrations of the drug (or metabolites)      performed that would identify potential effects on
and the toxicity associated with higher levels of the   the central nervous, cardiovascular, and gastro-
drug; and (d) the effects are more closely related to   intestinal systems, as well as an evaluation of
peak concentrations or to overall exposure (meas-       renal function and possibly immunogenicity.
ured by the area under the concentration time              Like many other disciplines, there are a multi-
curve, AUC).                                            tude of protocols and procedures that can be
   Toxicokinetic data are generally collected on        followed for each pharmacology study. A decoiled
the first day of dosing in a repeated-dose study,       review of each available procedure is outside the
and near the last day of dosing, i.e. during the last   purview of this discussion.
                                                  NON-CLINICAL TOXICOLOGY                                                                61

Table 6.3 International regulatory guidelines for safety pharmacology studies. Excerpts from international regulatory documents

USAÐ`Studies that otherwise define the pharmacological properties of the drug or are pertinent to possible adverse effects'
  (21CFR314.50, para 2)
EUÐ`A general pharmacological characterization of the substance, with special reference to collateral effects' (EC Directive 91/507/
UKÐ`A general pharmacological profile of the substance is required, with special reference to collateral effects . . . the aim should be to
  establish a pattern of pharmacological activity within major physiological systems using a variety of experimental models' (MAL2,
  p. A3F-1)
CanadaÐ`Secondary actionsÐstudies related to secondary pharmacological actions of the new drug which may be relevant to
  expected use or to adverse effects of the new drug' (Canada RA5, exhibit 2, p. 21).
AustraliaÐ`Studies should reveal potentially useful and harmful properties of the drug in a quantitative manner, which will permit an
  assessment of the therapeutic risk . . . Investigations of the general pharmacological profile should be carried out' (Guidelines under
  the Clinical Trial Exemption Scheme, pp. 12, 15)
Nordic countriesÐ`New drugs should be studied in a biological screening program so as to define any action over and above that which
  is desirable for the therapeutic use of the product'
JapanÐ`The objective of general pharmacological studies is to examine extensively the kind and potency of actions other than the
  primary pharmacological actions, predict potential adverse effects likely to manifest in clinical practice . . . ' (Japanese Guidelines, 29
  January 1991).

       Non-clinical Summary Documents                                       Second, the drug development strategy estab-
                                                                         lished by the company may call for the availabi-
Prior to the initiation of initial studies in man, it                    lity of proof of absorption and perhaps even
is important that all of the non-clinical infor-                         preliminary proof of efficacy (sometimes called
mation available is made into an integrated sum-                         `proof of principle') before expending resources
mary. This information must be included in the                           for the longer and more expensive studies. On
clinical investigators' brochure and so that the clin-                   the other hand, the company may have deter-
ical protocol can be modified to include relevant                        mined that the drug in question is on a `fast
biochemical or other markers to minimize human                           track', and is willing to expend resources early
risk. The regulatory authority and ethics commit-                        in the hopes of getting an earlier approval, and
tees are further target audiences, and the company                       thus to the market faster. The following sec-
may wish to use this for formal, internal proceed-                       tions will summarize the areas that need to be
ings to justify the decision to proceed with initial                     addressed.
human exposure.

                                                                                        Chronic Toxicity Studies
           PHASE II±III STUDIES                                          As discussed above, the extent of additional
                                                                         repeated-dose studies are generally outlined in
Non-clinical toxicology studies required to support                      Tables 6.2 and 6.3. The maximum duration of
Phase II and Phase III stages of the program depend                      chronic studies is generally 6 months. The ICH
upon a variety of factors. First, as shown in Tables                     guidelines describe situations where studies of 9±
6.1 and 6.2, the ultimate clinical regimen, i.e. dur-                    12 months duration in a non-rodent species may be
ation of therapy or treatment, determines the ultim-                     necessary, particularly for the US FDA. The ICH
ate duration of the animal studies. For example, a                       has issued a guideline relative to the duration of
diagnostic agent or a drug with a 3± 4 day regimen                       chronic toxicity studies.
[as might be the cases for disease or trauma situation                      Protocols for these studies are similar to those for
that are handled in the intensive care unit may                          studies of shorter duration, except that a minimum
require little in the way of additional repeated-                        of 10 ±15 rodents/group and four non-rodents/sex/
dose toxicity studies. In comparison, a new antihy-                      group are required. Toxicokinetic measurements
pertensive agent may require all of the longer-term                      are still required. The usual in-life and post-mortem
studies.                                                                 observations are performed.

     Reproduction and Teratology Studies                  Several different ICH guidelines have been
                                                       issued that address the various aspects of the car-
The Thalidomide tragedy demonstrated the need to       cinogenicity testing of drugs, including when stud-
evaluate new drugs in reproductive toxicology stud-    ies are needed (duration of clinical therapy; Federal
ies. Some of the earliest guidelines were issued by    Register March 1 1996). Other features of the new
the US FDA (the `Goldenthal guidelines'). An ICH       drug may mandate carcinogenicity testing, such as
guideline now covers the performance of these stud-    structure±activity similarities to known carcino-
ies (Federal Register September 22 1994), as           gens, evidence of preneoplastic lesions in repeated-
amended in 1995 to address possible effects on         dose non-clinical studies, or long-term tissue
male reproduction.                                     sequestration of the drug. Another guideline (Fed-
   In general, there are three phases of the repro-    eral Register March 1 1995) addresses the complex
ductive process that are evaluated. The first phase    issue of the selection of doses for these studies;
(historically referred to as Segment I study, and      this responds to much criticism of the prior recom-
now under ICH as Stage A) evaluates the effect of      mendation to use the maximum tolerated dose
the new drug on fertility and the early implantation   (which had been suggested by the National Toxi-
stages of embryogenesis. In these studies, breeding    cology Program; Haseman and Lockhart 1994).
animals of one species (usually rats or rabbits) of    The current ICH guideline recommends a high
both sexes will be treated for 2 or more weeks prior   dose causing up to a 25-fold greater plasma AUC
to mating, and then the females will be further        in rodents compared to the AUC in man at steady
dosed until day 6 of gestation. The second stage       state. A subsequent amendment to this guideline
(historically Segment II, now ICH Stage B) is the      (Federal Register December 4 1997) adds a further
teratology study (sometimes termed `the develop-       proviso that the highest dose in a carcinogenicity
mental toxicity study') and is done in both of the     study need not exceed 1500 mg/kg/day when (a)
same two species. The third stage (Segment III or      there is no evidence of genotoxicity and (b) the
ICH Stage C) evaluates treatment during late ges-      maximum recommended human dose is no bigger
tation, parturition, and lactation. Behavioral and     than 500 mg/day. The basis for species selection,
neurodevelopmental assessments in the offspring        circumstances needing mechanistic studies, and ex-
are often made in Segment III studies. In some         ploitation of pharmacokinetic information in car-
cases, two of the studies can be combined and still    cinogenicity testing is described in yet another
satisfy the ICH guideline.                             guideline (Federal Register August 21 1996).
   The period in the drug development process at          Modern protocols for carcinogenicity studies
which results of these studies are required varies     have changed little since first established in the
somewhat from country to country, and is discussed     early 1970s. In recent years, the use of mice (his-
in the ICH guideline. Hoyer (2001) reviews the cur-    torically the second of the two required species in
rent situation, and provides additional perspective.   addition to rats) has come under scrutiny because
                                                       they may be inappropriate models, with unusual
                                                       sensitivity to certain classes of chlorinated hydro-
            Carcinogenicity Studies                    carbons. The most recent ICH guideline (Federal
                                                       Register August 21 1996) allows for the option of
Carcinogenicity studies involve the treatment of       using transgenic mice and study designs of some-
rodents for long periods of time (18 months to 2       what shorter duration.
years) in order to determine whether the material         Of growing importance is the interaction of
possesses the capability to initiate or promote the    factors that are critical to a successful toxicology
development of tumors. The relevance of these          programme. For example, if a transgenic mouse
models to the human situation has been debated         model is selected, then the choice of strain is im-
for many years. Carcinogenicity studies have been      portant and may depend upon whether the drug is
required for all drugs where clinical therapy may      non-genotoxic (TG.AC model) or genotoxic (p53
extend for 6 months or longer. While the scientific    model). Metabolic and pharmacokinetic data are
debate about relevance of these studies continues,     important to ensure that the selected models handle
they remain required by regulation.                    and metabolize the drug in a fashion at least rea-
                                       NON-CLINICAL TOXICOLOGY                                             63

sonably similar to man, and may vary for the same       ticular importance is the integrated summary. This
drug according to the toxic effect of interest. Per-    document contains an overview of all studies that
haps the most important factor is the relevance of      were conducted, how the pharmacology, pharma-
the doses selected to those in man. While this has      cokinetic, and toxicology study information is
been a subject of controversy for years, a recent       interrelated (Peck et al, 1992), and what signifi-
ICH guideline allows for the use of toxicokinetic       cance the data has to human safety. A well-written
measurements, and states that doses that produce        integrated summary can be beneficial not only to
an AUC in the carcinogenicity model that are 25         the agency reviewer, but also to the NDA Sponsor.
times that seen in man at steady state may no           Some of the information in this summary is also
longer have to be used under some circumstances.        needed for the product's package insert. Crucially,
A recent review of the status of carcinogenicity        it should include comparisons between effects seen
testing (Reno 1997) addresses the many factors          in animals and the likelihood that such findings
that should be considered in a carcinogenicity pro-     would be expected in clinical usage. These compari-
gram.                                                   sons are often quantitative, and must be made both
                                                        on a mg/kg and a surface area (mg=m2 ) basis (Voi-
                                                        sin et al 1990).
                 Special Studies
It is not uncommon in drug development programs                      CANDA Requirements
for specific toxicities to be uncovered. In most
cases, additional studies are then carried out that     The Computer Assisted NDA (CANDA) refers to
will attempt to elucidate additional information        the submission of data to the FDA in machine-
with regard to the mechanism of the effect. For         readable form. At the present time, it is required
example, the identification of a non-specific behav-    that, at a minimum, the data from the carcinogen-
ioral effect (e.g. tremors and/or convulsions) may      icity studies be submitted to the agency in a speci-
trigger the performance of a neurotoxicity study,       fied format that is available from the FDA. This
which includes an exhaustive evaluation of the po-      allows the agency to apply its own criteria and
tential effects on central and peripheral nervous       statistics to the data and independently confirm
tissues. The identification of an effect on reproduc-   the sponsor's conclusions. This agency review is
tion may warrant the performance of detailed stud-      then submitted to the FDA's Carcinogenicity As-
ies to identify the specific mechanism or phase of      sessment Committee for final review and conclu-
the reproductive cycle that is affected. In-depth       sions.
metabolic studies may prove that the effect is re-
lated to a metabolite in animals that has no rele-
vance to man, and prevent the abandonment of an                           Expert Reports
otherwise promising drug. It is rare that a drug
development program does not involve some type          The European Community, and other countries,
of special study.                                       require several expert reports in each dossier, one
                                                        of which examines the non-clinical toxicology of
                                                        the new drug. These documents are typically about
             NDA REQUIREMENTS                           20±30 pages long and summarize all the toxicology
                                                        data, as well as the clinical implications.
   Format and Content of the Application                   Much from the integrated summaries described
                                                        above may be reused in this report, with the excep-
The FDA, in what was referred to as the `New            tion of the expert, who must personally sign the
Drug Application (NDA) Rewrite', issued guide-          report. Expert reports contain the expert's curricu-
lines for how a non-clinical section of an NDA          lum vitae, and part of the regulatory review process
should be organized (the guideline contains not         is to evaluate whether the expert is actually quali-
only requirements on what is required but also a        fied for this role. The choice of expert is important,
basic table of contents for this document). Of par-     and his/her independence is crucial because the role

is that of a reviewer, not of a sponsor. Experts may             Tennat RW, Stasiewicz S, Spalding JW. Comparison of Mul-
nonetheless be drawn from within the sponsoring                    tiple parameters of rodent carcinogenicity and in vitro genetic
                                                                   toxicity Environ Mutagen 1986; 8: 205±227.
company with appropriate protections, although                   Peck CC, Barr WH, Benet LZ et al. Opportunities for integra-
those from outside may carry more credibility in                   tion of pharmacokinetics, Pharmacodynamics, and toxicoki-
some jurisdictions.                                                netics in rational drug development. Clin Pharmacol Ther
                                                                   1992; 51: 465±473.
                                                                 Reno FE (1997) Carcinogenicity Studies. In Comprehensive
                                                                   Toxicology, vol 2, Toxicological Testing and Evaluation,
                                                                   Williams PD, Hottendorf GH (eds), Elsevier: London;
                      REFERENCES                                   121±31.
                                                                 Terrell TG, Green JD (1994) Issues with biotechnology products
Hoyer PB. Reproductive toxicology: current and future direc-       in toxicologic pathology. Toxicol Pathol 22: 187±93.
  tions. Biochem Pharmacol 2001. 62: 1557±1564.                  Voisin EM, Ruthsatz M, Collins JM, Hoyle PC (1990) Extrapo-
Fox Aw, Yang X, Murli H, Lawlor TE, Cifone MA, Reno FE,            lation of animal toxicity to humans: interspecies compari-
  Absence of mutagenic effects of sodium dichloroacetate. Fund     sons in drug development. Regulat Toxicol Pharmacol 12:
  Appl Tox 1996; 32: 87±95.                                        107±16.
Kowalski LA. In vitro carcinogenicity testing: present and       Williams PD, Hottendof GH, (Eds.) Comprehensive Toxicology
  future perspectives in pharmaceutical development. Cure          Vol.2: Toxicological testing & Evaluation. Elsevier: London,
  Opin Drug Discov Devel 2001; 4: 29±35.                           1997.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                     Informed Consent
                                                 Anthony W. Fox
                                       EBD Group Inc., Carlsbad, CA, USA

There is a tendency to assume that the principles of            ization because the treatment options are not
informed consent are self-evident. In fact, evidence            known to be unequally hazardous.
that this is not the case comes from many sources,
such as ethics committees, who are frequently dis-
satisfied with proposed informed consent docu-                         WRITTEN INFORMED CONSENT
ments, and sophisticated Western governments
that, from time to time, have conducted clinical                The large majority of clinical trials use a written
trials without it (e.g. the Tuskeegee travesty).                informed consent document. In the absence of any
   Informed consent was first formulated under                  special circumstances, the essential elements of
international law through the Declaration of Hel-               such a document are:
sinki, and in response to the atrocities of the
Second World War. The principles of informed                    1. A clear statement that the study is a research
consent are under continuous review and discus-                    procedure.
sion (e.g. Marsh, 1990). This is to be expected when            2. A clear statement that participation is volun-
reasonable standards of informed consent are de-                   tary and that there will be no repercussions,
pendent not only upon the design of a particular                   either in the patient's relationship with the in-
study, but also on environmental factors, the cur-                 vestigator, or with the patient's other care-
rent state of medicine, and particular local charac-               givers, should the patient decide not to take
teristics of clinical trials populations, all of which             part in the study.
are themselves continuously changing.                           3. A description of the scope and aims of the
                                                                   research, and whether or not there may be
                                                                   benefits to patients exposed to the test medica-
                  ETHICAL BASIS                                    tions. The foreseeable risks and discomforts
                                                                   should also be disclosed. The possibility of pla-
Although enlarged upon elsewhere in this book,                     cebo treatment, and the probability of being
two ethical principles guiding informed consent                    treated with each test therapy, should be stated.
are those of autonomy and equipoise. Autonomy is                4. Clear descriptions of alternative therapies or
the concept that the patient is an individual who is               standard therapies or procedures (if any), in
under no duress, whether subtle or obvious, actual                 order that the patient can judge whether to
or inferred, and is competent to make a choice                     enter the study.
according to his/her free will. Clinical trials con-            5. The methods for compensation that may be
ducted on persons in custody, or on subordinate                    available in the case of injury (these often
soldiers, may both violate the patient's autonomy.                 have marked international variations).
Equipoise is the concept that the investigator, and             6. Names and telephone numbers of persons who
those sponsoring the trial, are truly uncertain as to              the patient may contact in case of any difficulty
the outcome of the study; in practical terms, this is              during the study. Also, the identity of person(s)
a guarantee to the patient that an unreasonable                    of whom the patient may ask questions during
hazard cannot result from unfavourable random-                     the day-to-day conduct of the study, and an

      expression of willingness on the part of the
                                                            UNWRITTEN INFORMED CONSENT
      investigator to provide answers to any ques-
      tions that the patient may have.
                                                        Informed consent, in law, must be informed,
 7.   A confidentiality statement. This should in-
                                                        but need not be written. Ethics committees and
      clude the degree to which the patient's iden-
                                                        institutional review boards may sanction specific
      tity could be revealed to an inspecting
                                                        methods for the documentation of oral informed
      regulatory authority, and whether informa-
                                                        consent. This is a very rare clinical situation.
      tion from the clinical study will automatically
      be communicated to the patient's primary
      care or referring physician. In any case,
      there should be an assurance that no patient          SURROGATE INFORMED CONSENT
      identity information will be made public.
 8.   A statement of the circumstances under            Some patients are incapable of providing informed
      which the patient will be withdrawn from          consent, whether written or not. These patients are
      the study (e.g. non-compliance with test pro-     often in demographic subgroups which are medic-
      cedures).                                         ally underserved. Consequently, these are patients
 9.   A clear statement that the patient may with-      for whom there is encouragement to the pharma-
      draw from the study at any time and for           ceutical industry by governments, activists, and
      any reason, again without repercussions to        others to increase research into experimental ther-
      his/her relationship with any clinical care-      apies. Children, those with various types of neuro-
      giver.                                            logical disease (e.g. Alzheimer's disease), and
10.   A statement that the patient will be required     emergency patients (e.g. unconscious head injury,
      to give a full and accurate clinical and          stroke, multiple trauma, etc.) are good examples.
      treatment history on study entry and period-      Many of these patients have a very poor prognosis,
      ically thereafter (according to the study         and epitomize the concept of unmet medical need.
      design).                                          For these patients, clinical research would be im-
11.   Assurance that any new information that           possible if written informed consent was an essen-
      arises (e.g. in other studies) and which may      tial prerequisite.
      alter the assessment of hazard of study par-         For children, most ethics committees agree that
      ticipation will be communicated to the patient    provision of written informed consent by a parent
      without delay.                                    or guardian is acceptable. If the child is of sufficient
12.   A statement about the number of patients          age, then his/her concurrence may also be sought;
      taking part in the study, and a brief summary     while this is not sufficient evidence of informed
      of how many patients in the past have been        consent, the refusal to provide concurrence by a
      exposed to the test medication.                   child who is likely to be competent to understand
                                                        the clinical trial conditions should be sufficient to
Written informed consent documents should be            exclude the child from a study.
signed by both the patient and the investigator,           In the case of studies in legally incompetent
and ideally the patient should sign before an impar-    adults, again most ethical committees will accept
tial witness. Informed consent documents should be      a legal guardian or custodian in lieu of the patient
written in a language that is understandable to the     him/herself, provided that there is sufficient evi-
patient, and ideally at a level of complexity that      dence that the custodian has a bona fide and inde-
could be understood by a young adolescent of aver-      pendent interest in the patient's welfare. Again,
age intelligence from the same community as the         forms of concurrence can be employed when pos-
patient. There should be adequate time for the pa-      sible. The ordering of a patient's participation in a
tient to review the document. All written informed      clinical trial by a Court Order would usually be a
consent documents should be approved by an ethics       form of duress and thus violate the concept of
committee or an institutional review board.             autonomy described above.
                                             INFORMED CONSENT                                                    67

      WHEN INFORMED CONSENT IS                            practical constraints. The Informed Consent docu-
             IMPOSSIBLE                                   ment is essentially an agreement between Ethics
                                                          Committee, investigator and patient. However,
Emergency patients have as much right to taking           for example, an investigator is also responsible for
part in clinical research as any other type of patient.   the patient's role in the informed consent; if the
For example, patients with acute head injury and a        investigator suspects that the patient is not truly
low Glasgow Coma Score have a dismal prognosis,           informed, even in the absence of any deficiency on
and therapeutic interventions (if ever likely to be       the part of the investigator, then the investigator
successful) must be instituted quickly. Under these       should nonetheless police the patient's part of the
conditions there is often not even the time to find       agreement. This is entirely different from the
relatives to provide surrogate informed consent.          notion of a contract, where each party to the con-
Even if relatives can be found quickly enough,            tract is responsible only for fulfilling its own com-
then their emotional state may not be suited to           mitments.
becoming truly informed before giving consent.               Audit and policing of some of the elements listed
   Experiments are now under way to investigate           above may also form part of the duty of a regula-
whether some substitute for informed consent may          tory authority. For example, in the USA, the Food
be used. One set of guidelines suggests that such         and Drug Administration (FDA) will audit insti-
clinical trials can be conducted when:                    tutional review boards (IRBs) and issue citations if
                                                          the IRB is not ensuring that written informed con-
1. There is clinical and public agreement that the        sent documents are complete and appropriate.
   disease merits clinical investigation with the         FDA will also audit study sites, and discipline in-
   investigational therapy.                               vestigators (including prosecution) who do not
2. There has been advertising and publicity in the        ensure that appropriate informed consent proced-
   likely catchment area of suitable patients that        ures are being followed.
   such a study is being undertaken.                         Although under law it is not the primary respon-
3. The ethics committee or institutional review           sibility of the typical pharmaceutical company, it
   board has approved, in detail, the methods             nonetheless behooves pharmaceutical physicians to
   used in pursuit of local publicity.                    ensure that appropriate informed consent is being
4. An independent, clinically experienced individ-        obtained in all company-sponsored studies. Many
   ual will confirm that the patient is a member of       companies recognize this within their own Stand-
   the well-defined population that is the subject        ard Operating Procedures, and create patient files
   of the clinical research, and that it is not un-       that require a copy of the signed informed consent.
   reasonable to include the patient in the study         Investigators will often be grateful if the company
   for any other reason.                                  will draft an informed consent document that com-
5. No relative (if any is available in a timely fash-     plies with the guidelines, which the investigator can
   ion) objects.                                          submit to the ethics committee or IRB.

It is likely that these guidelines will be refined,
possibly on an international basis, in the near                               REFERENCES
                                                          Marsh BT (1990) Informed consent. J R Soc Med 83: 603±6.


It is the responsibility of all parties to the informed
consent that all parties remain within its ethical and
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                              Good Clinical Practices
                                 Wendy Bohaychuk and Graham Ball
       Good Clinical Research Practices Consultants, Lakehurst, Ontario, Canada, and Oxford UK

The aim of this chapter is to describe the general              ther, data should be reproducible, i.e., if the study
framework for conducting good clinical practices                were to be conducted in a similar population using
(GCP)-compliant clinical research, particularly                 the same procedures, the results should be the same.
pharmaceutical industry clinical research. Since it             To assure the integrity and reproducibility of re-
is difficult to cover this broad topic in such a short          search results, the whole process should be trans-
chapter, the authors will focus on those areas that             parent, i.e., everything must be documented so that
are most discussed, most problematic, and most                  an external reviewer may verify that the research
critical to achieving a GCP-compliant clinical                  was actually conducted as reported by the research-
study. Thus, there is particular emphasis on ethical            ers.
issues, source data verification and data integrity,
monitoring and safety review, and study medica-
tion/device management.                                         The General Regulatory Framework for GCP
                                                                The regulatory framework for compliance with re-
THE CURRENT RULES FOR CONDUCTING                                search procedures has essentially developed on an
        CLINICAL RESEARCH                                       international basis only in the last two decades,
                                                                except for the US where rules were first established
Conducting GCP-compliant clinical research is a                 in the 1930s. Today, most countries in the Euro-
serious undertaking, and this has been recognized               pean Union, other countries in Europe (e.g. Hun-
by numerous authorities internationally. It is diffi-           gary, Poland and Switzerland) and Japan have
cult to achieve a fully GCP-compliant a clinical                regulations on GCP. Other countries have regula-
study, but the expectation today is that the greatest           tions controlling clinical studies, with guidelines on
effort will be made nevertheless and the documen-               GCP, such as Australia and Canada. In the 1990s,
tation to provide evidence of this effort must be               an attempt was made to harmonise GCP require-
available.                                                      ments in the form of the ICH GCP document
                                                                which has since been adapted in regulation by
                                                                many countries. Some countries have no guidelines
            The Basic Tenets of GCP                             or regulations, but guidance for researchers has
                                                                been provided by oragnisations such as the Council
The primary reason for the presence of a GCP code               for the International Oragnizations of Medical Sci-
of practice is to safeguard human rights, as the                ences (CIOMS) and the World Health Organiza-
welfare of current study subjects and future patients           tion (WHO). (A brief list of existing regulations
is at stake. Therefore, systems must be in place                and guidelines is presented at the end of this chap-
(such as ethics committee review and informed con-              ter.) Regulatory authority review and/or approval
sent) to protect study subjects. Collecting honest              is usually necesary in all countries before, during,
and accurate data is also a major objective of GCP              and after clinical studies.
to ensure that data have integrity and that valid                  In the last few years, there has been increasing
conclusions may be drawn from those data. Fur-                  interest in regulatory inspection of GCP compliance

to ensure validity of the data and protection of                       The USA has vast experience (thousands of inspec-
study subjects, and to compare the practices and                       tions) compared to the handful of inspections in
procedures of the investigator and the sponsor/                        other countries.
CRO with the commitment made in the application                           Within a research organization, other independ-
for marketing. Although inspection has been a                          ent review, auditing, is undertaken internally to
regulatory requirement in the USA for many                             check on compliance with standards and basically
years, inspectorates have only just started in coun-                   to pre-empt the inspectors. Auditing may be con-
tries such as Austria, Denmark, France, Finland,                       ducted at any time during a clinical study to ensure
Germany, Japan, The Netherlands, Norway and                            continued compliance with GCP. Almost all
Sweden. There are problems in finding good in-                         aspects of GCP could be audited. Auditing, by
spectors, in deciding on the final standards for                       definition, must be undertaken by personnel who
inspections, and in imposing sanctions for non-                        are independent of the research being audited.
compliance. An interesting recent development
has been the initiation of inspections in Europe by
the central regulatory authority, the European                                 SETTING UP CLINICAL STUDIES
Medicines Evaluation Agency (EMEA). Regula-
tion of compliance with requirements by ethics                         To ensure that the standards for clinical research
committees is also developing in some parts of the                     are established before studies begin and to check on
world (e.g. France and Denmark). To date, the US                       compliance with those standards, many fundamen-
Food and Drug Administration (FDA) is the only                         tal systems and processes must be defined by
authority that is actively checking on the activities                  pharmaceutical companies and contract research
of institutional review boards (IRBs) by inspection                    organizations (CROs). These are outlined in Table
and licensing.                                                         8.1.
   For non-compliance with regulations, only the                          The sponsor/CRO has a duty to place a study
USA has imposed serious sanctions to date. The                         safely. That is, the sponsor (or the delegated CRO)
`blacklist' (list of all investigators who have been                   must assess and choose a site where study subjects
found to be non-compliant and were barred from                         will not be harmed. Some companies report that, in
clinical research for FDA submissions) is publicly                     practice, they have little choice in this process, as
available through freedom of information rules.                        the marketing department has already selected the

Table 8.1 General systems and procedures for implementation of GCP

The following systems and procedures must be established by clinical researchers to ensure compliance with GCP requirements:
Quality assurance: systems for assuring quality and for checking quality must be established and followed at all stages
Planning: studies must be conducted for valid (ethical and scientific) reasons
Standard operating procedures (SOPs): research procedures must be declared in writing so that reviewers can determine the standards
  which are being applied and so that users have a reference point
Well-designed study: all studies must have a valid study design, documented in a protocol, so that it can be fully reviewed by all
  interested parties. The data collection plans, as described in the CRF, are part of the protocol
Qualified personnel: all personnel (sponsor/CRO and study site) must be experienced and qualified to undertake assigned tasks.
  Documentation of qualifications and training must be evident
Ethics committee review and approval: all studies must be independently reviewed by ethics committees/IRBs, to assess the risk for
  study subjects, before clinical studies begin. Review must continue throughout the study
Informed consent: all study subjects must be given the opportunity to personally assess the risk of study participation by being provided
  with certain information. Their assent to participate must be documented
Monitoring: a primary means of quality control of clinical studies involves frequent and thorough monitoring by sponsor/CRO
Data processing for integrity of data: data must be honest. Data must be reviewed by site personnel, monitors and data processing
Control of study medications/devices: the product being studied must be managed so that study subjects ultimately receive a safe
  product and full accountability can be documented
Archives: documentation of research activities must be securely retained to provide evidence of activities
                                                   GOOD CLINICAL PRACTICES                                                             71

investigators (often those most likely to influence                        Any document used to collect research data on
use of the medications/devices). Another rationale                      clinical study subjects may be generically classed as
for apparent lack of choice is that there are too few                   a data collection form. These completed forms pro-
patients or investigators in a particular therapeutic                   vide evidence of the research conducted. The most
area. None of these reasons is as important as                          common type of data collection form is the CRF.
compliance with the basic GCP principle, which                          Other types of data collection forms include diary
requires the sponsor/CRO to assess, select and                          cards, dispensing records, quality of life forms, etc.
choose safe settings for research.                                      The CRF must allow for proper analysis of the
   Setting up clinical studies is a lengthy process, as                 data and proper reporting of the data in the final
there are many documents to prepare [e.g. proto-                        clinical study report and it must reflect the protocol
cols and case report forms (CRFs) ], study facilities                   exactly: no more and no less data must be collected.
to be assessed (e.g. study sites, CROs, clinical la-                    Thus, a CRF must be created for each clinical
boratories, Phase I units), regulatory review to be                     study and must be prepared in parallel with the
considered, and negotiations and agreements with                        protocol. CRFs are usually also prepared by spon-
study sites (e.g. contracts, finances, confidentiality,                 sors/CROs in pharmaceutical industry research be-
indemnity, insurance) to be undertaken. In add-                         cause of the demanding requirements for their
ition, as will be dealt with in subsequent sections,                    design and contents.
ethical aspects of the study must be considered
(e.g., ethics committee and IRB review and
informed consent requirements) and study medica-                          Selection of Investigators and Study Sites
tions/devices must be organized.
                                                                        The sponsor/CRO must go through a formal as-
                                                                        sessment procedure before placement of a study.
   Protocols and Case Report Forms (CRFs)                               Some of the most important areas requiring assess-
                                                                        ment are described in Table 8.2. All studies involv-
The protocol, with the accompanying CRF, is the                         ing research of investigational medications and
key document governing a clinical study. It for-                        devices require qualified investigators, and the
mally describes how a clinical study will be con-                       internationally accepted standard for `qualified'
ducted and how the data will be evaluated, and it                       usually encompasses three main criteria: medically
must include all the information that an investiga-                     qualified, i.e. legally licensed to practice medicine
tor should know in order to properly select sub-                        as a physician; experienced in the relevant thera-
jects, collect safety and efficacy data, and prescribe                  peutic speciality; and experienced in clinical re-
the correct study medication/device. Protocols                          search.
must be prepared in accordance with a specified                            Many contracts or agreements must be prepared,
and standardized format that is described in guide-                     understood and authorized before clinical studies
lines and regulations (the reader is particularly                       begin. The most common contracts include: the
advised to refer to the ICH GCP document). Proto-                       protocol and CRF; agreements for finances, confi-
cols are usually prepared, at least initially, by the                   dentiality, insurance, and indemnity; and contracts
sponsor or the delegated CRO, although investi-                         between the sponsor and the CRO. A separate in-
gator input is obviously necessary.                                     vestigator agreement, specifying all responsibilities,

Table 8.2 Selection of study sites

The following items should be assessed at study sites by sponsor/CRO monitors before studies begin:
Study site personnel, e.g. qualification, experience, training, availability; specific allocation of responsibilities
Facilities, e.g. offices, wards, archives, pharmacy, clinical laboratory; study medication/device storage areas; clinical laboratories;
  access to source documents; ethics committee/IRB requirements
Suitable study subject population, e.g. access to suitable subjects in sufficient numbers; method of subject recruitment; source, e.g. from
  investigator's subject population, or be referred by other physicians and, if referred, means by which investigator will obtain
  adequate evidence of medical history; use of advertisements; potential subject enrolment (recruitment) rate

Table 8.3 Investigator GCP responsibilities

The following investigator responsibilities must be declared in agreements or contracts:
Adhere to the protocol exactly. No changes to the protocol may be undertaken without following a formal protocol amendment
  procedure and without agreement by the sponsor/CRO
Be thoroughly familiar with the properties of the clinical study medications/devices as described in the investigator brochure
Have sufficient time to personally conduct and complete the study. If more than one investigator is involved at a specific study site, the
  specific responsibilities must be described for each investigator. The investigator must ensure that no other studies divert study
  subjects, facilities or personnel from the study under consideration
Maintain the confidentiality of all information received with regard to the study and the investigational study medication/device
Submit the protocol, information sheet and consent form, and other required documentation, to an ethics committee/IRB for review
  and approval before the study begins. During the study, the investigator is also responsible for submitting any new information, e.g.
  protocol amendments, safety information, which might be important for continuing risk assessment by the ethics committee/IRB
Obtain informed consent from each study subject prior to enrolment into the study
Inform the subjects primary care physician, e.g. general practitioner or family physician, of proposed study participation before
  enrolment into the study
Maintain study subject clinical notes, i.e. source documents, separately from the CRFs. The source documents must support the data
  entered into CRFs and must clearly indicate participation in a clinical study. If the study subject is referred by another physician, the
  investigator must ensure that sufficient evidence is available in the clinical notes to support the eligibility of the study subject
Maintain a confidential list identifying the number/code and names of all subjects entered into the study
Allow authorized representatives of the sponsor/CRO and regulatory authorities direct access to study subject clinical notes (source
  documents) in order to verify the data recorded on CRFs
Ensure CRFs are complete and accurate
Allow monitoring visits by the sponsor/CRO at a predetermined frequency. During these monitoring visits, the monitor must be
  allowed to communicate with all site personnel involved in the conduct of the clinical study
Report all AEs and SAEs to the sponsor/CRO and follow the special reporting requirements for SAEs
Maintain the security and accountability of clinical study supplies, ensure that medications/devices are labeled properly, maintain
  records of clinical study medication/device dispensing, including dates, quantity and use by study subjects; and return or disposition
  (as instructed by the sponsor/CRO) after completion or termination of the study
Archive all CRFs and documents associated with the study for a minimum of 15 years. Notify the sponsor/CRO of any problems with
  archiving in potential unusual circumstances, e.g. investigator retires, relocates, dies; study subject dies, relocates, etc.
Provide reports of the study's progress whenever required
Review the final clinical report, and sign and date the signature page after review
Allow an independent audit and/or inspection of all study documents and facilities
Agree to the publication policy
Agree to the sponsor's/CRO's ownership of the data
Agree to the stated time frames for the study, e.g. start and completion of recruitment, submission of completed CRFs
Work to GCP as defined by the ICH, FDA and local regulations

is usually necessary in addition to the protocol, to                               Ethics Committee/IRB Review
emphasize certain aspects of the protocol. Table 8.3
highlights some of the main investigator GCP re-                        All clinical studies require review by an independent
sponsibilities which might be included in contracts.                    ethics committee/IRB before, during and after
                                                                        the study. Before any study subjects are treated,
                                                                        review by the committee must be documented in
           ETHICAL CONSIDERATIONS                                       compliance with international guidelines and the
                                                                        local regulations of the country in which the
Part of the selection process for a study site in-                      research is conducted. Clinical studies begin (for
volves confirming that ethics committee/IRB                             the study subjects) whenever any procedure is
review will be safe and that all study subjects will                    undertaken by study subjects that they would not
be properly informed prior to consent to study                          normally undergo: ethics committee/IRB review
participation. If the sponsor/CRO cannot obtain                         must be sought before these events. Thus, if a
documented evidence of compliance with these two                        study requires screening procedures, washout
fundamental requirements, it is not safe to work                        from normal treatment, and even completion of
with that site.                                                         a questionnaire that poses personal questions, the
                                         GOOD CLINICAL PRACTICES                                          73

study begins when those procedures are under-              Ethics committees/IRBs also have responsibility
taken. It is a common misconception that studies         for review during and after clinical studies (Table
begin only when study subjects are randomized to         8.5). In other words, committee review is an on-
treatment.                                               going responsibility that extends beyond the initial
   Prior to selection of a clinical study site, the      submission and review of documents to proceed
sponsor/CRO must confirm and document, in                with the study.
the prestudy assessment visit report, that the inves-
tigator has access to a local ethics committee/
IRB. Local committees cannot be bypassed: the                           Informed Consent
only official exception to this requirement is
in France, where, by regulation, a central commit-       Potential study subjects may enter a clinical study
tee may rule for all sites in a multicentre study.       conducted by the sponsor/CRO only after being
However, in the USA, it appears to be com-               properly informed and consenting to participate.
mon practice for a central IRB to rule for the           The researchers must consider who does what,
widely geographically separated areas in the             when, what sort of information must be provided,
country, and researchers may not inform the local        and how this will all be documented. The general
IRB.                                                     principles for the conduct of informed consent are
   Normally, the sponsor/CRO will prepare all            noted in Table 8.6. (See also Chapter 7). All infor-
necessary documentation for submission by the            mation sheets and consent forms should include the
investigator to the ethics committee/IRB (it is not      items listed in Table 8.7 and they must be provided
usual procedure for the sponsor/CRO to directly          before study participation. Obtaining informed
submit items to the committee, unless requested to       consent is a complex issue and it is not easy to
do so by the committee). Whatever the local vari-        comply with these requirements.
ations, the sponsor/CRO is usually responsible
for ensuring the submission of the items in Table
8.4. Some committees require other additional
items.                                                   MONITORING AND SAFETY ASSESSMENT
   The membership of an ethics committee/IRB will
vary nationally and regionally. However, the spon-       The conduct of clinical studies is a cooperative
sor/CRO is only permitted to conduct studies that        undertaking between the sponsor/CRO and the
are approved by ethics committees/IRBs that have         investigator; each is responsible for ensuring that
a sufficient number of qualified members to enable       the study is in conformity with the protocol and in
a medical and scientific review of the proposed          accordance with all applicable laws and regula-
study and to enable a review of all other ethical        tions, and, of course, that study subjects are pro-
aspects of the study. Generally, ethics committees       tected at all times. This responsibility involves
also have to be diverse in composition. Details of       regular and conscientious review of the progress
the membership of the ethics committee/IRB               of the study by the sponsor/CRO and by the inves-
should be obtained and reviewed by the sponsor/          tigator and study site personnel.
CRO, prior to initiating the study, to ascertain the
above and to determine that there is no serious
conflict of interest (e.g. investigator voting on her/                      Monitoring
his study).
   The sponsor/CRO should also request a written         One of the most important means of quality control
copy of the working procedures of the ethics com-        of a clinical study is managed by frequent and
mittees/IRBs. These procedures should provide            thorough monitoring. A monitor's aim is first to
sufficient information to assure sponsors/CROs,          protect the agenda of the sponsor/CRO who
investigators, auditors and inspectors of the integ-     employs him/her. Monitors (often referred to as
rity and independence of the ethics committee/IRB.       CRAs or Clinical Research Associates or Assistants
Unfortunately, today, it is still difficult to obtain    in the pharmaceutical industry) must ensure main-
working procedures from many committees.                 tenance of proper standards, compliance with the

Table 8.4 Review by ethics committees/IRBs before clinical studies begin

The following items should be reviewed by ethics committees/IRBs before clinical studies begin:
Protocol (including annexes, such as the CRF)
Consent procedures (described in the protocol and the appended information sheet and consent form), which specify who will provide
  information and who will obtain consent, how consent will be documented, and whether or not a witness will be present
Consent form/information sheet. Most committees will be particularly interested in these documents to ensure that all necessary
  information is provided to study subjects
Suitability of investigator and facilities, including support personnel. Some committees may request a copy of investigator and other
  site personnel CVs. The committee will be particularly interested in allocation of resources, whether the investigator has enough time
  and study subjects to conduct the study, and whether use of resources for clinical studies will detract from normal medical care
Delegation of responsibility by investigators
Source of study subjects and means of recruitment. The committee will wish to know if study subjects are known to investigators and,
  if not (i.e. referred patients), how investigators will confirm eligibility and whether primary care practitioners will be informed. The
  committee will wish to determine that advertisements are not unduly coercive or misleading or too `inviting'
Appropriateness (eligibility) of study subjects (described in the protocol)
Primary care physician to be informed of study participation
Number of subjects to be studied and justification for sample size (this information should be in the protocol). The committee will be
  interested in how many subjects will be exposed to the risk of treatment. In a multicenter study, the local ethics committee/IRB
  should be informed of the number of subjects to be enrolled at each site and the total number of subjects to be enrolled in the study
Investigator brochure or other authorized summary of information (e.g. preclinical and clinical summaries) about the investigational
  products, including comparator products and placebo. If the study medication/device is a marketed product, the ethics committee/
  IRB must review the most current data sheet, product monograph, etc. The brochure is particularly important for confirming the
  formal declared safety profile of the study treatment and therefore is of great assistance to committees in assessing the relevance of
  AEs. Also, the committee can verify, by reviewing the brochure or product labeling, that the information sheet for obtaining consent
  provides sufficient information with regard to safety
Evidence of regulatory submission and review/approval (if applicable). Committees particularly wish to know whether the drug/device
  is on the market in their country or in other countries, and the details of the stage of the submission
Adequacy of confidentiality safeguards, with regard to protection of identification of the study subject (described in the protocol and
  the appended information sheet and consent form)
Insurance provisions, if any, for injury to study subjects (described in the protocol or provided as a separate document). Committees
  must confirm that there is insurance for protection of the study subjects
Indemnity/insurance provisions for the sponsor/CRO, investigator, institution, etc. (as relevant to the study and if required by local
Payments or rewards to be made to study subjects, if any. Committees must determine that the amount, and schedule of payments, is
  not unduly coercive
Benefits, if any, to study subjects
Payments or rewards to be made to investigators. Many committees are beginning to realize that the financial interests of the
  investigator might have a strong influence on some aspects of the study, particularly recruitment patterns
Assurance of quality/stability of medication/device to be administered
Review decision of other ethics committees/IRBs in multicentre studies
Duration of study
Plans to review data collected to ensure safety

protocol, accurate and complete data capture, and                      or other significant events, and the nature of the
standardization across sites in a multicentre study.                   study medication/device. At the beginning of
Basically, monitors will undertake the review noted                    a study, monitoring may be even more frequent.
in Table 8.8.                                                          The most time-consuming task at the study
   In general, study sites should be visited by                        site is the review of source documents to con-
a monitor at least every 4 ±6 weeks. The freque-                       firm entries in CRFs and compliance with the pro-
ncy of monitoring visits will be defined for each                      tocol.
individual study and will depend on details such                          The monitor will be ever-vigilant for protocol
as the study phase, treatment interval and overall                     violations which can occur during a study and
duration, enrolment rate, complexity of the study                      which can have a serious impact on eligibility and
methodology, occurrence of adverse events (AEs)                        evaluability. Many researchers confuse the terms
                                                   GOOD CLINICAL PRACTICES                                                             75

Table 8.5 Review by ethics committees/IRBs during and after clinical studies

The following items should be reviewed by ethics committees/IRBs during and after clinical studies:
Serious and/or unexpected AEs, if any occur during the study, including the follow-up period
Protocol amendments, if any, and reasons for amendments
Protocol violations which impact on subject safety, if any
Discontinuation of study, if applicable, and any reasons for premature discontinuation
Any new significant information, e.g. information arising from other studies, results of interim analyses, marketing approvals, changes
  in local procedures, updated investigator brochure, supply problems, during study, if any
Amendments to consent forms/information sheets, if any
Annual reports of the study. More frequent review may be necessary, depending on the working procedures of each individual ethics
Final clinical report/summary of study. Some ethics committees/IRBs also review publications, if any

Table 8.6 Principles for the conduct of informed consent

The following principles for conducting informed consent should be implemented for all clinical studies:
Informed consent must be obtained from each study subject. The person receiving the information and giving consent must sign the
  consent form. This is usually the study subject, but may be the study subject's legally acceptable representative (depending on
  national regulations) in the event that the study subject is incapable of providing informed consent, e.g. the subject is unable to write
  or understand the consent documents, or the study subject is in a `vulnerable' population, e.g. children, elderly. Informed consent
  must be obtained before the start of the study
The person providing the information and obtaining consent must sign the consent form. This person should be an investigator who
  must be qualified to adequately inform the study subject, and her/his signature also indicates personal involvement in the consent
  process. If other personnel, e.g. study nurses assist in providing information or obtaining consent, they should also sign the consent
  form, clearly describing their role in the consent procedure
A witness or patient advocate should be present during the consent procedure at the times of providing information and giving
  consent, and should sign the consent form. The witness will ensure that there was no coercion in the obtaining of informed consent
  and that the study subject was given adequate time to consider participation in the study. The witness must be able to confirm that
  the consent procedure was adequate and must have no vested interest in the clinical study, i.e. the witness should be impartial,
  independent, or neutral, as far as this can be achieved. The relationship of the witness to the study subject and to the investigator and
  the study should be documented
All participants should personally date their signatures and all dates should precede the start of the study (for each subject)

Table 8.7 Information to be provided to study subjects before obtaining consent to participate in clinical studies

The information sheets and consent forms should contain the following items:
1. Information about the consent procedure:
   Consent to be given by the study subject's free will
   Adequate time (which should be defined in advance in the protocol) must be allowed for the study subject to decide on
     participation in study
   Adequate time must be allowed to ask questions
   Statement that participation is entirely voluntary
   Statement that refusal to participate will involve no penalties or loss of usual benefits
   Description of circumstances under which participation would be terminated
   Right to withdraw at any time without prejudice or consequences
   Study subject is allowed to keep the written explanation (information sheet and consent form) for future reference
2. Information about the study and medications/devices:
   Instructions on use and storage of study medication/device, if relevant
   Name of sponsor/CRO
   Explanation that the study is a research procedure
   Description of study type and research aims
   Description of study medications/devices
   Description of procedures to be followed
Table 8.7 (contd.)

     Description of experimental procedures to be followed, if any. Experimental procedures might include those which are not
       normally used for the presentation under consideration or procedures which are new or have never been used before
     Comparator treatments (including placebo) described. It is important to explain `placebo' in simple terms
     Randomization procedures. Randomization is not easily understood by many subjects and should also be explained in simple
     Expected duration of participation
     Required number of visits
     Reason for selection of suitable subjects
     Approximate number of other study subjects participating in the study
3. Information about the risks/benefits:
   Foreseeable risks, discomforts, side effects and inconveniences
   Known therapeutic benefits, if any. The benefits must not be `oversold'
   Availability of alternative therapies. If there are other treatments, this must be explained so that the subject does not feel the new
     treatment is the only option
   Any new findings, which might affect the safety of the study subject, and that become available during participation in the study,
     will be disclosed to the study subject
   Assurance of compensation for treatment-induced injury with specific reference to local guidelines (it must not be expected that the
     study subject is familiar with the guidelines, and therefore the guidelines must be explained and/or attached)
   Terms of compensation
   Measures to be taken in the event of an AE or therapeutic failure
   Financial remuneration, if any. Patients, whether receiving therapeutic benefit or not, are not usually paid for participation in
     clinical research, except for incidentals such as travel costs. Healthy volunteers are usually paid a fee for participation, but this
     payment should never be offered to induce the prospective subjects to take risks they would not normally consider
   Explanation of additional costs that may result from participation, if any (this normally only occurs in the USA)
4. Other items:
   Ethics committee/IRB approval obtained (some debate about this)
   Name of ethics committee/IRB (if applicable by local and/or national requirements) and details of contact person on the ethics
     committee/IRB (if applicable by local and/or national requirements)
   Explanation that participation is confidential, but records (which divulge study subject names) may be reviewed by authorized
     sponsor/CRO representatives and may be disclosed to a regulatory authority
   Name, address and telephone number (24 h availability) of contact person at study site for information or in the event of an
     emergency (this information may be provided on a separate card)
   Requirement to disclose details of medical history, any medicines (or alcohol) currently being taken, changes in any other
     medication/device use, and details of participation in other clinical studies
   Medical records will clearly identify study participation
   Conditions as they apply to women of child-bearing potential
   Primary care physician (or general practitioner or family doctor) and/or referring physician will be informed of study participation
     and any significant problems arising during the study. Some subjects may not be comfortable with this requirement, e.g. in a
     study of sexually transmitted diseases, they may not wish the doctor, perhaps a family friend, to be aware of their situation. If
     this is the case, the subject is not eligible for the study as it is vital to confirm history with the primary care physician.
   The information sheet must be written in language which is understandable, e.g. technically simple and in the appropriate national
     language, to the study subject.

`protocol violations' and `protocol amendments'. It                         Reporting and Recording Safety Events
is important to appreciate the differences between
these terms and understand how to avoid protocol                       An issue over which site personnel and monitors
violations and how to manage protocol amend-                           will be particularly watchful is the observation and
ments. Perhaps the easiest way to explain the                          recording of safety information. In many studies,
difference is to stress that violations are not                        safety information is under-reported because of the
planned changes (hopefully) to the protocol,                           tendency to make judgments that are often based
whereas protocol amendments are planned changes                        on subjective and biased clinical opinion. It seems
and are enacted through a formal approval process                      difficult to teach clinical researchers to operate as
(if violations are deliberate or planned, a case of                    `scientists': that is, to observe and record all obser-
fraud should be considered!).                                          vations before making judgments. The monitor
                                                   GOOD CLINICAL PRACTICES                                                               77

Table 8.8 Objectives of monitoring visits

The following tasks should be undertaken by the sponsor/CRO monitor at each study site visit:
Verify accuracy and completeness of recorded data in CRFs, including diary cards, quality of life forms, registration forms, consent
  forms, etc., by comparing with the original source documents (clinic or hospital records). Where discrepancies are found,
  arrangements must be made for corrections and resolution. Resolve any outstanding queries, ensuring completion of any issued data
  queries, since the last monitoring visit
Verify compliance with entry criteria and procedures, for all study subjects, as specified in the protocol. If subjects are found to be
  ineligible or unevaluable, these events must be immediately brought to the attention of the investigator. There may also be
  implications for payment to the study site and requirements for reporting to ethics committees/IRBs. Finally, and most seriously,
  there could be implications for subject safety
Review all AEs, including clinically significant laboratory abnormalities, that have occurred since the previous visit. If a serious or
  unexpected AE has occurred, which was not correctly reported by the investigator, the monitor must ensure that the correct
  reporting procedure is followed immediately
Evaluate the subject recruitment and withdrawal/dropout rate. If recruitment is less than optimal, suggest ways in which it can be
  increased. In particular, query the reasons for withdrawals/dropouts, or unscheduled visits, in case these are related to AEs
Confirm that all source documents will be retained in a secure location. Source documents must be legible and properly indexed for
  ease of retrieval. Check the study site file to ensure that all appropriate documents are suitably archived. Check that the investigator
  files are secure and stored in a separate area which is not accessible to individuals not involved in the study
Conduct an inventory and account for study medications/devices and arrange for extra supplies, including other items, such as CRFs,
  blank forms, etc., if necessary. Resolve discrepancies between inventory and accountability records, and medication/device use, as
  recorded in the CRFs. If a pharmacy is involved in the study, the pharmacy and pharmacist must be visited. Check that the
  medication/device is being dispensed in accordance with the protocol. Check that the medication/device is being stored under
  appropriate environmental conditions and that the expiry dates are still valid. Check that the medication/device is securely stored in
  a separate area that is not accessible to individuals not involved in the study. Check that any supplies shipped to the site since the last
  visit were received in good condition and are properly stored. If applicable, ensure that randomization procedures are being
  followed, blind is being maintained, randomization codebreak envelopes are intact (sealed and stored properly) and a chronological
  sequence of allocation to treatment is being followed
Verify correct biological sample collection (especially number, type, and timing), correct procedures for assays (if applicable), and
  labeling, storage and transportation of specimens or samples. All clinical laboratory reports should be checked for identification
  details, validity and continued applicability of reference ranges, accuracy of transcription to CRFs (if any), comments on all out-of-
  range data, and investigator signatures and dates. The dates of sample collection, receipt, analysis and reporting should be checked
  to ensure that samples are analysed promptly, and that investigators are informed of results and review them promptly
Ensure continued acceptability of facilities, staff and equipment. Ensure that the reference range, documentation of certification and
  proficiency testing, licensing, and accreditation, for the clinical laboratory are still current. Document any changes in clinical site
  personnel and, if changes have occurred, collect evidence of suitability of new personnel. Ensure that new staff are fully briefed on
  the requirements of the protocol and study procedures and arrange any training of new personnel, if necessary. Document any
  changes in overall facilities and equipment and if changes have occurred, collect new evidence of suitability, maintenance,
  calibration and reason for change of new equipment
Advise the investigator and other site personnel of any new developments, e.g. protocol amendments, AEs, which may affect the
  conduct of the study

and all clinical research personnel must ensure that                     information in clinical records at the study sites;
all safety information is documented. This means                         information in data collection forms (e.g. CRFs,
that all adverse events (AEs) occurring in clinical                      diary cards, quality of life forms, psychiatric rating
studies must be recorded in CRFs, their signifi-                         scales, etc.), occurrence of missed and/or unsched-
cance must be assessed, and other information                            uled visits, dropouts and withdrawals; use of
must be provided, for reporting AEs externally                           any concomitant medications/devices; and abnor-
(e.g. to regulatory authorities and ethics commit-                       mal laboratory data. AEs may also occur simply
tees/IRBs). This applies to any study treatment                          as a result of study procedures and study participa-
(including comparator agents, placebo and non-                           tion. Information about definitions of AEs
medical therapy) and any stage of the study (e.g.                        and requirements for reporting AEs must be clearly
run-in, washout, active treatment, follow-up).                           stated in the protocol and explained to the site
   All research personnel must search for clues                          staff, who must also be educated in the correct
about safety events from many sources, such as:                          procedure and immediate requirement for

reporting any adverse event suspected to be             onto computers or direct entry of patient informa-
serious or unexpected per the regulatory defin-         tion onto computers by physicians).
itions.                                                    The primary purpose of source documents is for
   All investigators and other study site personnel,    the care of the study subject from a clinical per-
ethics committees/IRBs, and possibly study sub-         spective: the primary purpose of CRFs is to collect
jects, must be informed of all new significant safety   research data. CRFs (and other data collection
information, including all events occurring with        forms) generally cannot substitute as source docu-
any treatment (e.g. washout, investigational prod-      ments. Data entered in CRFs should generally be
uct, comparator, placebo, etc.) in the study, even if   supported by source data in source documents,
these occurred in another study with the same           except as specifically defined at the beginning of
treatment, or in another country. Significant safety    the study. Nevertheless, some data entered in CRFs
information includes all SAEs and any other events      may be source data (e.g. multiple blood pressure
(e.g. significant trends in laboratory data or new      readings, psychiatric rating scales, etc.) and would
preclinical data) that might have an impact on the      not be found elsewhere. This may be acceptable, if
risk assessment of the study. Safety events may         these data would not normally be entered in med-
necessitate an update to the investigator brochure,     ical records, and if knowledge of such data is not
the protocol and CRF, and the information sheet         required by the investigator or other clinicians who
and consent form.                                       concurrently or subsequently treat the study sub-
                                                        ject (the protocol should specify which data will be
                                                        source data in the CRF).
     COLLECTING DATA WITH INTEGRITY                        How much information is expected to be docu-
                                                        mented in source documents? This is a difficult
Collecting data that are accurate, honest, reliable     issue, but one that must be discussed and resolved
and credible is one of the most important objectives    before the CRFs are completed. Some guidelines
of conducting clinical research. It is difficult to     are provided in Table 8.9.
achieve. However, in general, data in CRFs are             Direct access to source documents is required for
not credible to the regulators unless they can          all studiesÐdirect access means monitors, audit-
be supported by the `real' documents (i.e. the          ors, other authorized representatives of the spon-
source documents maintained at the study site for       sor/CRO, and inspectors are permitted to view all
the clinical care of the study subject).                relevant source documents needed to verify the
                                                        CRF data entries. Other restricted methods of
                                                        access to source documents (e.g. `across-the-
           Source Data Verification                     table', `back-to-back', `interview method') are not
                                                        acceptable, as they do not allow proper verification
Source data verification is the process of verifying    of the data in CRFs. To ensure direct access, the
CRF entries against data in the source documents.       study subject consent form must clearly indicate
Source data verification is only carried out at         that permission for access has been granted by the
the study site, usually by the sponsor/CRO monitor      study subject.
(auditors will also conduct source data verification
on a sample of CRFs; inspectors may conduct
source data verification on a sample or all CRFs).         Other Review to Assure Data Integrity
  Source documents (and the data contained
therein) comprise the following types of documents:     After retrieval from the study site, there are further
patient files (medical notes where summaries of         means of assessing CRFs. First, there is the initial
physical examination findings, details of medical       review at the sponsor/CRO premises: this process is
history, concurrent medications/devices, and dis-       sometimes referred to as `secondary monitoring'.
eases are noted), recordings from automated instru-     Thereafter, review by the data management depart-
ments, traces (eg ECGs, EEGs), X-ray films,             ment is another extremely important means of
laboratory notes, and computer databases (e.g. psy-     quality control. It is a lengthy and complex process
chological tests requiring direct entry by patient      and there are few guidelines and regulations for
                                                   GOOD CLINICAL PRACTICES                                                              79

Table 8.9 Source data verification

For all study subjects, source data veri®cation requires a review of the following items:
Existence of medical records/files at the study site. There must be a medical file, separate from the CRF, which forms a normal part of
  the clinical record for the study subjects. The medical file should clearly indicate the full name, birth date, and hospital/clinic/health
  service number of the study subject
Eligibility of study subjects. The medical file must show compliance with the inclusion and exclusion criteria. At a minimum,
  demographic characteristics, e.g. sex, weight and height, diagnoses, e.g. major condition for which subject was being treated, and
  other `hard' data, e.g. laboratory results within a specified range or normal chest X-ray, should be clearly indicated. All required
  baseline assessments must be evident. If the medical file has little or no information concerning medical history, it would not support
  selection of the subject
Indication of participation in the study. The medical file should clearly show that the subject was in a clinical study in case the
  information is necessary for future clinical care
Consent procedures. The original signed consent form should be maintained with the subject's medical files or in the investigator files
  and an indication that consent was obtained (with the date specified) should be noted in the medical files. Signatures and dates must
  be checked carefully to ensure that the correct individuals were involved in the consent procedure and that consent was obtained
  prior to any study intervention
Record of exposure to study medication/device. The medical file should clearly indicate when treatment began, when treatment finished,
  and all intervening treatment dates
Record of concomitant medications/devices. All notations of previous and concomitant medication/device use must be examined. All
  entries in the CRF should be verifiable in the medical file by name, date(s) of administration, dose and reason (or indication). All
  entries in the medical file during the time period specified by the protocol must be noted in the CRF. Concomitant medication/device
  use must be explicable by an appropriate indication and must be consistent from visit to visit. The reasons (indications) for use of
  concomitant medications/devices, newly prescribed during the study period, must be noted as AEs. The medical history should be
  reviewed to determine whether medical conditions arising during the study already existed at baseline. The dispensing records, which
  are normally separate from the medical file, must also be examined to determine consistency.
Visit dates. All visit dates should be recorded in the medical file. Interim visit dates recorded in the medical file, but not in the CRF,
  should be noted by the monitor in case they signify occurrence of AEs or protocol violations. The final visit date should be so
  indicated, e.g. `study finished' or `withdrew from study'
AEs. All AEs noted in the medical file during the time period specified by the protocol must be recorded in the CRF. The monitor must
  also carefully check other documents (e.g. diary cards, quality of life forms) for sources of information about AEs. Occurrence of
  out-of-range laboratory values, which are considered to be clinically significant by the investigator, must be reported and assessed as
Major safety and efficacy variables (to be decided and documented in advance). It is not necessary for all measured variables to be
  recorded in the medical file. Present and future clinical care of the study subject is the most important factor in determining whether
  or not measured variables should be recorded in the medical file. The investigator should record what he/she would normally record
  to care for the study subject, but also take into account any recording needed because of the special circumstances of a clinical study.
  The entire medical file should be reviewed to ensure that no additional information exists in the medical file that should have been
  recorded in the CRF

reference. These processes will inevitably result in                    requests for data clarification and all resolution of
queries about the data. It is critical that all data                    queries must be documented. All data changes
review procedures be prompt. As time goes by, it                        must be authorized by the investigator ultimately.
becomes more and more difficult to correct data.                        Obviously, the sponsor/CRO cannot arbitrarily
Slow processing usually means that data lose cred-                      make changes of data.
   To ensure that the integrity of clinical research
data is maintained and that there is total agreement                                               Archiving
between the data recorded on CRFs, the data
entered on the computer, the data recorded in                           Systems must be in place to ensure that documents
data listings and cross-tabulations, the data entered                   will be securely retained for a long period of time.
into statistical and clinical study reports, and                        The purpose of archiving is to safeguard all docu-
finally the data in the sponsor/CRO and investi-                        mentation that provides evidence that a clinical
gator archives, it is essential that the data must only                 study has been conducted in accordance with the
be changed by following a formal procedure. Thus,                       principles of GCP. Archives at both the sponsor/

CRO and investigator sites must be reasonably           contact name, address and telephone number is
secure with regard to indexing, controlled access,      available for emergencies and enquiries; and the
fire-resistance, flood-resistance, etc.                 study subject (or the person administering the
   The investigator must be held responsible for        medication/device) is knowledgeable about storing
ensuring that all source documents, especially          and administering the study medication/device,
records acquired in the normal practice of care         and that the packing process can be audited against
and treatment of a study subject, are safely            a standard Operating Procedure.
archived and available for inspection by authorized
company personnel or regulatory authorities. Fur-
ther, the investigator must archive all necessary         Shipment of Study Medications/Devices
documents for a minimum of 15 yearsÐthe usual
industry standard. All appropriate clinical study       Clinical study medications/devices should not be
documents should be archived by the sponsor/            dispatched to study sites until all prestudy activities
CRO, essentially for the lifetime of the product.       have been completed and regulatory requirements
The specific documents to be retained are described     satisfied. The receipt of each shipment of study
in the ICH GCP document.                                medication/device should be confirmed in writing
                                                        by the investigator or pharmacist (or other author-
                                                        ized personnel), who will be instructed to return a
     MANAGING STUDY MEDICATIONS/                        completed `acknowledgement of receipt form' im-
               DEVICES                                  mediately. The recipient at the study site will be
                                                        instructed to contact the sponsor/CRO immedi-
Management of clinical study medications and            ately if there are any problems (e.g. missing or
devices is a complicated activity, and many clinical    broken items, defects in labeling, evidence of excur-
researchers report that they are not particularly       sion from temperature ranges) with the shipment.
interested in this aspect of clinical studies: they     The recipient must be particularly instructed to
assume that it is all handled by other personnel in     record the exact date of receipt of the clinical sup-
the manufacturing facility. Meanwhile, personnel        plies at the study site. This information is necessary
in the manufacturing facility usuall report that        so that the monitor can determine that the supplies
once the supplies are released, they assume no fur-     were secure and correctly stored environmentally
ther responsibility!                                    during the entire period of shipment.
                                                           After the clinical study supplies have been sent to
                                                        the study site, the monitor must verify as soon as
 Preparation of Study Medications/Devices               possible that the supplies have arrived satisfactorily.
                                                        Supplies may not be dispensed to study subjects
The preparation of study medications or devices         until the monitor has checked their condition. The
for clinical studies is a time-consuming process        monitor will verify that the amount shipped
and often rate-limiting in initiating the study, par-   matches the amount acknowledged as received. If
ticularly with double-blind designs. Requisition,       there is a lack of reconciliation, or if the shipment is
labeling and packaging are some of the important        not intact, recruitment may be delayed until the
considerations.                                         situation is resolved.
   Requisition of study medication/device (includ-
ing placebo and comparator products, if relevant)
must be initiated at an early stage to allow suffi-       Control of Study Medications/Devices at
cient time to procure the study medications/devices                      Study Sites
and to prepare the final labeling and packaging,
taking into account any special circumstances for       Evidence of careful control at the study site is
blind studies and for import requirements.              imperative and naturally it is difficult to standard-
   The principles of safe labeling and packaging        ize the situation across many study sites and many
require compliance with the following principles:       countries. Security, correct storage, and accurate
the contents of a container can be identified; a        documentation of dispensing and inventory are
                                         GOOD CLINICAL PRACTICES                                           81

necessary. Systems to ensure and assess compliance       to also allow assessment of possible detrimental
with the required use of the product being studied       environmental impact. All unused and returned
must be established. Monitors must be trained to         medications/devices, empty containers, devices,
check on these features and ensure that all site         equipment, etc., which are returned to the investi-
personnel are fully briefed.                             gator by the study subjects, must be stored securely
   The expectations with regard to maintenance of        and under correct environmental conditions at the
study medications/devices at study sites focus on        study site until retrieval by the monitor. The moni-
security and appropriate environmental condi-            tor will check the supplies returned and verify that
tions. Concerns for security require that supplies       they reconcile with the written specifications. All
be maintained under locked conditions. All agree-        discrepancies and the reasons for any non-returns
ments between the sponsor/CRO and the study site         must be documented and explained.
must specify that supplies are only for clinical study      Generally, destruction of returned study medica-
subjectsÐthis information must also be clearly           tions/devices by the sponsor/CRO may not take
stated on the labeling. The main concern for ap-         place until the final report has been prepared and
propriate environmental conditions is usually tem-       until there is no further reason to question the
perature requirements, but other factors (e.g. light,    accountability of the study medication/device. The
humidity) might also be important. Terms such as         actual destruction process must be documented in a
`room temperature' and `ambient temperature',            manner which clearly details the final disposition of
which have different meanings in different coun-         the unused medications/devices and the method of
tries, should always be avoided and specific tem-        destruction. The information is particularly neces-
peratures must be stated. At each monitoring visit,      sary in case of any query regarding environmental
the monitor will ensure that the correct procedures      impact. In exceptional circumstances, unused study
are being followed.                                      medications (e.g. cytotoxics, radio-labeled prod-
   Compliance with medication/device use (by the         ucts) may be destroyed at the study site, with ap-
study subject) should be assessed in all studies. If     propriate documentation.
supplies are dispensed to subjects for self-adminis-
tration, methods to assure compliance (e.g., diary
cards, instructions on labeling, supervised adminis-              Randomization and Blinding
tration) and methods to check compliance (e.g.
tablet counts, plasma/urine assays, diary card           Randomization procedures are employed to ensure
review) must be in place. At each study visit, the       that study subjects entered into a comparative
study subjects should be asked to return all unused      study are treated in an unbiased way. Blinding (or
supplies and empty containers to the investigator,       masking) procedures (e.g. single-blind or double-
who will check the supplies for assessment of com-       blind) further minimize bias by ensuring that out-
pliance and store them for return to the sponsor/        come judgments are not based on knowledge of the
CRO. The monitor will review all relevant docu-          treatment. If the study design is double-blind, it is
ments (e.g. source documents, CRFs, medication/          essential that all personnel who may influence the
device inventory, dispensing forms) to ensure that       subject or the conduct of the study are blinded to
the data in the CRFs reflect the subjects' compli-       the identity of the study medication/device assigned
ance with the study medications/devices.                 to the subject, and therefore do not have access to
                                                         randomization schedules.

       Overall Accountability of Study
           Medications/Devices                                              SUMMARY

Overall accountability must documented and               The code of Good Clinical Practice was established
reviewed. A reconciliation of the initial inventory      to ensure subject safety and arose because of biases
and the final returns must be undertaken and all         inherent in clinical research (e.g. pressures to re-
discrepancies must be explained. Final disposition       cruit subjects for payment, publication, etc.), which
and destruction must be carefully documented             needed some counterbalance. It is hoped the reader

will appreciate that GCP is not `bureaucratic non-                  Commission Directive 91/507/EEC modifying the Annex to Coun-
sense' (as argued by some researchers), but is a                      cil Directive 75/318/EEC on the approximation of the laws of
                                                                      Member States relating to the analytical, pharmacotoxicologi-
logical, ethical, and scientific approach to stand-                   cal and clinical standards and protocols in respect of the testing
ardizing a complex discipline.                                        of medicinal products, Official Journal of the European Com-
                                                                      munities, 1991.
                                                                    Commission Directive 2001/20/EC of the European Parliament
                                                                      and of the Council of 4 April 2001 on the approximation of
       SOURCES OF INTERNATIONAL                                       the laws, regulations and administrative provisions of the
     GUIDELINES/REGULATIONS FOR GCP                                   Member States relating to the implementation of good clinical
                                                                      practice in the conduct of clinical trials on medicinal products
Australia                                                             for human use.
National Statement on Ethical Conduct in Research Involving           en_301L0020.html
  Humans, National Health and Medical Research Council              Manufacture of Investigational Medicinal Products, Annex to the
  Act, 1992.       EC Guide to Good Manufacturing Practice, EEC 111/3004/91-
  ses/e35syn.htm                                                      EN, Brussels, 1992.
Note for Guidance on Good Clinical Practice (CPMP/ICH/135/          Biostatistical Methodology in Clinical Trials in Applications for
  95). Annotated with TGA comments. Therapeutic Goods                 Marketing Authorizations for Medicinal Products, Committee
  Administration [TGA] (Australia), Commonwealth Depart-              for Proprietary Medicinal Products [CPMP] EEC 111/3630/
  ment of Health and Aged Care. The TGA has adopted                   92-EN, 1994.
  CPMP/ICH/135/95 in principle but has recognised that
  some elements are, by necessity, overridden by the National       UK
  Statement (and therefore not adopted) and that others             Clinical Trial Compensation Guidelines, Association of the Brit-
  require explanation in terms of `local regulatory require-          ish Pharmaceutical Industry (ABPI), 1994.
  ments', July 2000.        Conduct of Investigator Site Audits, ABPI, 1993.
  ich13595.htm                                                      Good Clinical (Research) Practice, ABPI, 1996.
Note for Guidance on Clinical Safety Data Management (CPMP/         Good Clinical Trial Practice, ABPI, 1995.
  ICH/377/95). Annotated with TGA comments. Therapeutic             Introduction to the Work of Ethics Committees, ABPI, 1997.
  Goods Administration (Australia), Commonwealth Depart-            Patient Information and Consent for Clinical Trials, ABPI, 1997.
  ment of Health and Aged Care. The TGA has adopted the             Phase IV Clinical Trials, ABPI, 1993.
  Note for Guidance on Clinical Safety Data Management:             Set of Clinical Guidelines, ABPI, 2000.
  Definitions and Standards for Expedited Reporting in              Structure of a Formal Agreement to Conduct Sponsored Clinical
  principle, particularly its definitions and reporting time          Research, ABPI, 1996.
  frames. However, there are some elements of CPMP/ICH/             Fraud and Misconduct in Clinical Research, Royal College of
  377/95 which have not been adopted by the TGA and other             Physicians of London, 1991.
  elements which require explanation in terms of `local regula-     Guidelines for Clinicians Entering Research, Royal College of
  tory requirements', 2000.        Physicians of London, 1997.
  html/ich37795.htm                                                 Guidelines on the Practice of Ethics Committees in Medical Re-
                                                                      search Involving Human Subjects, Royal College of Physicians
Canada                                                                of London, 1997.
Code of Ethical Conduct for Research Involving Humans,              Research Involving Patients, Royal College of Physicians of
  Medical Research Council of Canada, Natural Sciences and            London, 1990.
  Engineering Research Council of Canada, Social Sciences           Research on Healthy Volunteers, Royal College of Physicians of
  and Humanities Research Council of Canada, 1998. http://            London, 1986.                     title.htm
Clinical Trial Review and Approval, Drugs Directorate, Policy       Governance Arrangements for NHS Research Ethics Committees:
  Issues, Health and Welfare Canada, 1995. TPP (Therapeutic           (Section A ± General Standards and Principles, Department of
  Products Program, Canada)          Health [DOH], Central Office for Research Ethics Commit-
  therapeut/htmleng/whatsnew.html                                     tees [OREC], 2001
Clinical Trial Framework, Schedule 1024, Food and Drug Regu-        Guidelines for Good Pharmacy Practice in Support of
  lations. Therapeutic Products Directorate, Health Products          Clinical Trials in Hospitals, Royal Pharmaceutical Society,
  and Food Branch, Health Canada, 2001. http://www.hc-                1994.   Guidance on Good Clinical Practice and Clinical Trials in the
  sch-1024_e.pdf                                                      NHS, National Health Service, 1999.
European Union                                                      Research Ethics Guidance for Nurses Involved in Research or
Good Clinical Practice for Trials on Medicinal Products in the        Any Investigative Project Involving Human Subjects, Royal Col-
  European Community, Committee for Proprietary Medicinal             lege of Nursing Research Society,1998 http://www.doh.
  Products [CPMP] EEC 111/3976/88-EN, Brussels, 1990.                 research/rd3/nhsrandd/researchgovernance/govhome.htm
                                                 GOOD CLINICAL PRACTICES                                                         83
USA                                                                   ucts (Human and Animal), Department of Health & Human
Regulations:                                                          Services, FDA, April 1991.
Code of Federal Regulations [CFR], 21 CFR Ch 1, Food and              old042fn.pdf
  Drug Administration [FDA], Department of Health and
  Human Services [DHHS]:                                            Inspection and Warning Letters:
Part 11 ± Electronic Records; Electronic Signatures http://         Clinical Investigator Inspection List              regulatory/investigators/default.htm
Part 50 ± Protection of Human Subjects http://www.access.gpo.-      Debarment List
  gov/nara/cfr/waisidx_01/21cfr50_01.html                             default.htm
Part 54 ± Financial Disclosure by Clinical Investigators http://    Disqualified/Restricted/Assurances List for Clinical Investiga-              tors
Part 56 ± Institutional Review Boards      sur.htm
  nara/cfr/waisidx_01/21cfr56_01.html                               Notice of Initiation of Disqualification Proceedings and Oppor-
Part 312 ± Investigational New Drug Application;       tunity to Explain (NIDPOE) Letters                   nidpoe/default.html
Part 314 ± Applications for FDA Approval to Market a New            Public Health Service (PHS) Administrative Actions Listings
  314 _01.html                                                      Warning Letters
Compliance Program Guidance Manuals for FDA Staff:
Compliance Program 7151.02. FDA Access to Results of Quality        Forms:
  Assurance Program Audits and Inspections, 1996. [Same as          Form FDA 1571 ± Investigational New Drug Application
  Compliance Policy guide 130.300]            (IND)
  compliance_ref/cpg/cpggenl/cpg130-300.html                        Form FDA 1572 ± Statement of Investigator http://forms.psc.
Compliance Program 7348.001 ± Bioresearch MonitoringÐIn               gov/forms/FDA/FDA-1572.pdf
  Vivo Bioequivalence, 1999.       Form FDA 3454 ± Certification: Financial Interests and Ar-
  ce_ref/bimo/7348_001/Default.htm       rangements of Clinical Investigators
  compliance_ref/bimo/7348_001/foi48001.pdf                           forms/fda3454.pdf
Compliance Program 7348.809 ± Institutional Review Boards,          Form FDA 3455 ± Disclosure: Financial Interests and Arrange-
  1994.          ments of Clinical Investigators
  irb-cp7348-809.pdf                                                  FDA/FDA-3455.pdf
Compliance Program 7348.810 ± Bioresearch Monitoring ± Spon-
  sors, Contract Research Organizations and Monitors, 2001.         International       ICH:
  .htm         Clinical Safety Data Management: Definitions and Standards for
  48-810.pdf                                                          Expedited Reporting, International Conference on Harmon-
Compliance Program 7348.811 ± Bioresearch Monitoring ± Clin-          isation [ICH] of Technical Requirements for the Registration
  ical Investigators, FDA, 1997.          of Pharmaceuticals for Human Use, 1994. http://www.ifp-
  ora/ftparea/compliance/48_811.pdf                                 Clinical Safety Data Management: Periodic Safety Update
                                                                      Reports for Marketed Drugs, International Conference on
Information Sheets:                                                   Harmonisation [ICH] of Technical Requirements for the
Computerised Systems Used in Clinical Trials. FDA, 1999 http://       Registration of Pharmaceuticals for Human Use, 1996.         
Enforcement Policy: Electronic Records; Electronic SignaturesÐ      Note for Guidance on Structure and Content of Clinical Study
  Compliance Policy Guide; Guidance for FDA Personnel, FDA,           Reports, International Conference on Harmonisation [ICH]
  1999           of Technical Requirements for the Registration of Pharma-
  dates/cpg-esig-enf-noa.htm                                          ceuticals for Human Use, 1995.
Guidance. Financial Disclosure by Clinical Investigators. FDA,        pdfifpma/e3.pdf
  2001.            Guideline for Good Clinical Practice. International Conference
Guidance for Institutional Review Boards and Clinical Investiga-      on Harmonisation [ICH] of Technical Requirements for the
  tors, FDA, 1998. .htm       Registration of Pharmaceuticals for Human Use, 1996. http://
Guidance for Institutional Review Boards, Clinical Investigators,
  and Sponsors: Exceptions from Informed Consent Require-           General Considerations for Clinical Trials. International Confer-
  ments for Emergency Research, FDA, 2000. http://www.fda.-           ence on Harmonisation [ICH] of Technical Requirements for
  gov/ora/compliance_ref/bimo/err_guide.htm                           the Registration of Pharmaceuticals for Human Use, 1997
Guideline for the Monitoring of Clinical Investigations, FDA,
  1988.               Statistical Principles for Clinical Trials. International Confer-
Guideline on the Preparation of Investigational New Drug Prod-        ence on Harmonisation [ICH] of Technical Requirements for
  the Registration of Pharmaceuticals for Human Use, 1998.          1996, and the 52nd General Assembly, Edinburgh, Scotland,                              October, 2000.

WHO:                                                              Other Related Publications by the Authors:
Good manufacturing practices for pharmaceutical products          Bohaychuk W, Ball G (1994) Good Clinical Research Prac-
  supplementary guidelines for the manufacture of investiga-        tices. An Indexed Reference to International Guidelines and
  tional pharmaceutical products for studies in humans, 1994.       Regulations, with Practical Interpretation (available from         authors).
  MET+DST+2000/0/49                                               Bohaychuk W, Ball G (1996) GCP. A Report on Compliance.
International Ethical Guidelines for Biomedical Research In-        (available from authors)
  volving Human Subjects, Council for                             Bohaychuk W, and Ball G (1998), GCP Audit Findings ± Case
International Organizations of Medical Sciences [CIOMS] in          Study 1, Quality Assurance Journal, Volume 3, Issue 2.
  collaboration with the World Health Organization [WHO],         Bohaychuk W, and Ball (1998), 101 GCP SOPs for Sponsors and
  1993.     CROs (available from authors, paper and diskette).
  17:28+MET+DST+2000/0/49                                         Bohaychuk W, and Ball (1998), GCP Audit Findings ± Case
Guidelines for Good Clinical Practice (GCP) for Trials on           Study 2, Quality Assurance Journal, Volume 3, Issue 3.
  Pharmaceutical Products, Division of Drug Management &          Bohaychuk W, Ball G, Lawrence G, Sotirov K (1998) A Quan-
  Policies, World Health Organization, 1994. http://saturn.         titative View of International GCP Compliance. Appl Clin               Trials February: 24±29 (first in a series of articles published
  2000/0/49                                                         approximately every 2 months).
Operational Guidelines for Ethics Committees that Review Bio-     Bohaychuk W, Ball G (1999) GCP compliance assessed by
  medical Research, World Health Organization, 2000. http://        independent auditing: international similarities and differ- +           ences. In Hamrell M (ed.), The Clinical Audit in Pharmaceut-
  DST+2000/0/49                                                     ical Development Marcel Dekker: New York.
                                                                  Bohaychuk W, Ball G (1999) GCP compliance: national similar-
World Medical Association:                                          ities and differences. In Eur Pharmaceut Contract September.
Declaration of Helsinki. Recommendations Guiding Physicians in    Bohaychuk W, Ball G (1999) Conducting GCP-compliant Clin-
  Biomedical Research Involving Human Subjects, Adopted by          ical Research (available from John Wiley & Sons Ltd, Baffins
  the 18th World Medical Assembly, Helsinki, Finland, June          Lane, Chichester, West Sussex PO19 1UD, UK, www:inters-
  1964, amended by the 29th World Medical Assembly, Tokyo,
  Japan, October 1975, the 35th World Medical Assembly,           Bohaychuk W, and Ball G (2000), GCP Compliance Assessed by
  Venice, Italy, October 1983, and the 41st World Medical           Independent Auditing. International Similarities and Differ-
  Assembly, Hong Kong, September 1989, the 48th General             ences, In The Clinical Audit in Pharmaceutical Development,
  Assembly, Somerset West, Republic of South Africa, October        edited by M Hamrell. (Marcel Dekker publishers)
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

Quality Assurance, Quality Control and Audit
                                                   Donna Cullen
                                            Auditrial, Fairlawn, NJ, USA

The Food and Drug Administration (FDA) and                      cited for either falsifying or inadequately docu-
regulatory agencies throughout the world require                menting clinical study data. As recently as 1997, a
assurance, from those who seek to market drugs,                 physician investigator in California pled guilty to
biologics, medical diagnostics or medical devices in            fraud after substituting his own blood for some
their countries, that these products are safe and               clinical study tests. Another investigator in Geor-
effective for their intended indications. Companies             gia was convicted of fraudulently diverting ap-
engaged in the research, development, licensing,                proximately $10 million in grants for clinical
and marketing of pharmaceuticals must comply                    studies, and was cited for failing to participate in
with regulations governing the manufacturing and                the care of study subjects enrolled at his location
testing of their products. They must ensure that                and inadequately supervising his research staff.
risks to test animals and human subjects are min-               These are but a few of the historical cases that
imized throughout the research and development                  demonstrate the need for quality assurance in
process and that the data generated in support of               pharmaceutical research and development, despite
their objectives are true and accurate. Monitoring              years of regulation and control.
and inspection by companies and regulatory agen-                   The FDA was established as a law enforcement
cies are required to assess and control the integrity           agency in 1930. At that time, the Federal Food and
of research and development.                                    Drugs Act of 1906, which prohibited false and
                                                                misleading statements about a drug or its ingredi-
                                                                ents, and the Shirley Amendment (1911), which
                  BACKGROUND                                    subsequently prohibited false therapeutic claims
                                                                in drug labeling, were the only two substantive
In 1963 a Brooklyn hospital undertook an investi-               regulations governing the research and develop-
gation to determine whether pre-existing cancer or              ment of pharmaceuticals. In 1938 Congress passed
other debilitations in cancer patients compromised              the Federal Food, Drug and Cosmetic (FD&C)
their ability to defend against cancer cells. Twenty-           Act, both in response to recommendations from
two control patients, without cancer, were given an             the FDA to revise the obsolete 1906 Federal Food
injection of live cancer cells without their written            and Drugs Act, and as a result of an elixir of
consent, without being informed about the injec-                sulfanilamide, which caused 107 deaths, mostly in
tions, and without the approval of the hospital's               children, because of its adulteration with glycol.
research committee. Both the investigator and the               The FD&C Act required that drug safety be estab-
medical director were found guilty of fraud, deceit,            lished before marketing and extended regulations
and unprofessional conduct. In 1974 eight prison-               to cosmetics and therapeutic devices. It did not
ers in a Maryland correctional facility were coerced            address the issue of drug efficacy, however. Fac-
into testing vaccines for typhoid and malaria. In               tory inspections and standards for food quality and
1977 a physician investigator claimed that all                  containers were also authorized.
source records for patients enrolled in a clinical                 Between 1941 and 1945, Congress amended the
study at his location were lost in a rowboat acci-              1938 FD&C Act to reflect the development of
dent. In 1978 three separate investigators were                 insulin, penicillin, and other new antibiotics.

The FDA was given authority to approve these               The Declaration of Helsinki, published in 1964,
products as both safe and effective and to require      established free and informed consent, preferably
certification that each batch or lot conforms to the    written, for research subjects, or consent by a legal
established standards of purity and potency. The        representative if a subject is legally incompetent. It
regulation of biological products (serum, vaccines      also required research to be supervised by compe-
and blood products) began in 1944 with the Public       tent medical persons.
Health Service Act and was supplemented in 1986            Drug abuse and control amendments were
by the Childhood Vaccine Act, which required pa-        enacted in 1965 in response to abuse of depressants,
tient information on vaccines, and gave the FDA         stimulants and hallucinogens. In 1966 the FDA, in
authority to recall biologics and authorized penal-     cooperation with the National Academy of Sci-
ties for violations.                                    ences and the National Research Council, began
   The Nuremberg Code, adopted in 1949, required        an evaluation of the effectiveness of approximately
voluntary consent for all research subjects, guaran-    4000 drugs that had been approved between 1938
teed their right to leave a research study at any       and 1962 on the basis of safety alone. The Fair
time, protected them from remote possibilities of       Packaging and Labeling Act was passed in the
injury, disability, or death, and stipulated that the   same year and required true and informative label-
results of research studies must be intended for the    ing for all consumer products involved in interstate
good of society.                                        commerce. A review of the safety and effectiveness
   The Durham Humphrey Amendment (1951) de-             of over-the-counter drugs sold without prescrip-
fined the kinds of drugs that required medical super-   tion began in 1972, and was followed 10 years
vision for safe use and restricted them to sale by      later by FDA-mandated tamper-resistant pack-
prescription. Two years later, in 1953, the US Public   aging regulations, in reaction to reports of deaths
Health Service issued a policy document entitled        from Tylenol capsules laced with cyanide.
Group Consideration of Clinical Research Proced-           In 1971, in a move to affirm the Declaration of
ures Deviating from Accepted Medical Practice or        Helsinki, and further guarantee the protection of
Involving Unusual Hazard. This policy stated that       research subjects, the US Department of Health,
the potential risks of clinical studies must be care-   Education and Welfare (DHEW) issued The Guide
fully assessed and that informed consent was essen-     to DHEW Policy on Protection of Human Subjects.
tial. Ethical review of research proposals was also     This publication required every institution receiv-
suggested. In this same year, the Factory Inspection    ing DHEW funds for research to establish a com-
Amendment was passed, requiring the FDA to give         mittee to monitor the ethical integrity of that
written inspection reports to manufacturers.            institution's human research, and made both the
   The Color Additives Amendments (1960) re-            institution and the investigator responsible for vio-
quired manufacturers to establish the safety of         lations.
these substances in foods, drugs and cosmetics.            Congress passed the National Research Act in
Following reports of birth defects associated with      1974 and created the National Commission for the
the use of thalidomide in Australia and Western         Protection of Human Subjects of Biomedical and
Europe, public support for tougher drug regulation      Behavioral Research. The commission was given
prompted the passage, in 1962, of the Kefauver±         the task of developing ethical guidelines for con-
Harris Drug Amendment. Many US patients who             ducting research in human subjects, particularly
had received thalidomide in clinical studies during     children, prisoners, and the mentally impaired.
this period were not informed that its status was       The commission recommended to DHEW and
investigational. This legislation required drug         Congress the enactment of federal regulations
manufacturers, for the first time, to prove that        governing ethical review of all human research by
their products were effective prior to marketing,       institutional review boards, including accreditation
and required the affirmative act of approval by         and education. The commission also recommended
the FDA. In the same year, President John F.            that these boards be required to review informed
Kennedy proposed the Consumer Bill of Rights,           consent procedures for research subjects, and that
which include the rights to safety, information,        DHEW establish an office to monitor compliance.
choice, and the right to be heard.                      In response to these recommendations, the FDA
                          QUALITY ASSURANCE, QUALITY CONTROL AND AUDIT                                    87

and DHEW drafted regulations and guidelines for         containers and closures, production and process
clinical investigators and institutional review         controls, laboratory and stability testing, pack-
boards.                                                 aging and labeling, holding and distribution,
                                                        returned and salvaged drug products, and records
                                                        and reporting. In addition, there are guidelines and
            QUALITY ASSURANCE                           regulations issued by the Occupational Safety and
                                                        Health Administration (OSHA), the Environmen-
It is the responsibility of each company to identify    tal Protection Agency (EPA), the International
and meet the regulatory standards and to follow         Standards Organization (ISO) and the United
the regulations and guidelines applicable to its        States Pharmacopoeia (USP) that are also applic-
products. Failure to do so will delay licensing and     able to most phases of the manufacturing process.
marketing approval. The company must give its               The FDA has established regulations and re-
promise or guarantee (assurance) of the character       quirements for extensive animal testing of a prod-
of its product with respect to the grade of excel-      uct before human investigations are permitted.
lence (quality). Compliance at each phase of the        GLP regulations govern every aspect of preclinical
development and manufacturing process must be           development, including the humane and proper
carefully documented. The FDA and most foreign          care of the animals used in research and the quali-
regulatory agencies assess the degree of compliance     fications of testing facilities.
with current accepted standards, and guidelines for         Good clinical practice (GCP) guidelines ensure
manufacturing, preclinical/animal testing, and the      that risks to human subjects participating in re-
conduct of clinical studies in human subjects. In the   search studies are minimized, that they are properly
USA these standards are published as regulations        informed about these risks and that their participa-
under Title 21 of the Code of Federal Regulations       tion is entirely voluntary. Procedures are estab-
21 CFR for Good Manufacturing Practice (GMP)            lished for preparing investigational new drug
and Good Laboratory Practice (GLP), and Guide-          applications (IND), the responsibilities of sponsors
lines for Good Clinical Practice (GCP).                 and investigators, the composition and responsibil-
   Elsewhere in the world, individual countries         ities of institutional review boards, and obtaining
have established their own standards and guide-         and documenting written informed consent. Special
lines or have formed alliances with other countries,    consideration is also given to prisoners used in re-
such as the Commission of the European Commu-           search, drugs intended to treat life-threatening or
nity (CEC), to unify and facilitate approvals. The      debilitating illnesses, foreign studies, and adminis-
EU, Australia, Canada Japan, the Nordic coun-           trative actions for non-compliance.
tries, the USA and the World Health Organization            Compliance to applicable GMP and GLP regu-
(WHO) have already drafted an international eth-        lations and GCP guidelines can be assured by es-
ical and scientific standard for designing, conduct-    tablishing policies and standard operating
ing, recording and reporting clinical trials with       procedures (SOPs) that address them, and by docu-
human subjects. The International Conference on         menting that these procedures are followed. SOPs
Harmonization (ICH) Guideline of Good Clinical          should be reviewed routinely and updated when
Practice, published in May 1997, has its roots in the   required. Deviations from SOPs must be carefully
Declaration of Helsinki and assures the rights,         explained and documented and should be ad-
safety and well-being of human subjects participat-     dressed with a view towards correction and pre-
ing in research.                                        venting recurrence. If a quality assurance unit or
   GMP regulations encompass every phase of             department exists within a company, it should be
product development (manufacturing, testing, ac-        empowered to assure compliance to regulations,
ceptance, etc.), from the purity of the raw materials   guidelines and SOPs. It should also be organized
to the quality of the finished product. Established     in a way that will permit the inspection and
procedures include staffing and training, quality       reporting of processes and procedures without the
control, facilities and maintenance, selection,         possibility of coercion or pressure from the person-
calibration and maintenance of equipment, pro-          nel and departments that it monitors. For example,
curement, testing and storage of raw materials,         it should not be under the control of manufacturing

or clinical research. Frequently, the organizational      Quality control in manufacturing involves: phys-
structure places it within regulatory affairs, but     ical and chemical testing of raw materials, inter-
free-standing QA/QC departments are also               mediates, and the final product; potency testing;
common.                                                microbiological testing; product stability testing;
   To assure that a final product meets recognized     environmental monitoring and analysis; process
standards for safety and efficacy, the quality of      reliability; process validation; equipment qualifica-
the ingredients, the quality of the manufacturing      tion and validation; product disposition; complaint
processes, and the quality of the research and         handling; auditing of suppliers; and periodic prod-
testing used to develop the product must be verifi-    uct review. Statistically valid sampling plans are
able. The basic elements required to accomplish this   needed to support the testing and inspection of
include written quality standards, systems to in-      raw materials, components and final product, as
spect, control, validate, and maintain quality,        well as systems to identify each of these during the
systems to manage changes in quality, and records      production process. The stability and reliability
to document that these systems are in place and        testing of these items before and after distribution
operating properly. The burden of proof rests with     must also be monitored. Regular communication
the holder of the IND and the supporting documen-      with executive management is essential. There
tation.                                                should be a system to identify trends and handle
                                                       customer complaints, and there should be provi-
                                                       sions for special situations that require updates on
             QUALITY CONTROL                           the quality of the system and to manage changes,
                                                       both planned and unplanned. Unplanned changes
The proactive arm of quality assurance is quality      within a system (ingredients, processes, equipment,
control. Compliance to regulations throughout          documentation or personnel) require investigation
each phase of the research and development             and appropriate corrective action, and must be
process can only be assessed by routine monitoring     carefully documented.
with repeated sampling and testing using validated        The basic characteristics of a new product are
methods, by comparing the results to the desired       determined in the preclinical development phase.
or accepted standards, and by making corrections       This includes the development of new product for-
when deviations from the standards are identified.     mulation, identification of its physical, chemical,
To maintain quality, variability must be controlled.   pharmacologic, and toxicologic properties, as well
To improve quality, variability must be reduced.       as pharmacokinetics, and metabolism and teratol-
   Planning, good scientific methods, and valid-       ogy studies in animals. This information must
ation are the first steps in a quality program, and    be gathered and reported to regulatory agencies
quality control (testing and inspection) is the con-   before the first dose of the product can be adminis-
tinuing verification that validated systems remain     tered for research in human subjects. Thus, it is
in control. Planned change and revalidation assure     imperative that quality control procedures are
continuing quality improvement within the regu-        functioning at the beginning of the development
lated environment.                                     process, to ensure that regulatory requirements
   Another important factor to consider in any         are met and that the product safety information is
quality control program is personnel. Procedures       accurate before proceeding to the next series of
should be established to verify qualifications, pro-   tests. Animals must be adequately fed and housed
vide and evaluate continuous training, and assess      to ensure that they are free of diseases that might
and document job performance. Education and            alter or invalidate the test results. The new product
training are essential to maintaining competency.      and controls must be administered according to
Those individuals who perform the processes within     established protocols, variables must be controlled
research and development and those responsible for     during testing, data must be carefully recorded,
monitoring and control should be qualified, prop-      and specimens must be collected and properly ana-
erly trained, and understand the accepted standards    lyzed and stored. Data recorded in bound note-
and guidelines. There must be a commitment on the      books or recorded on electronic media must be
part of employees to abide by them as well.            archived.
                          QUALITY ASSURANCE, QUALITY CONTROL AND AUDIT                                    89

   In the clinical development phase, many of the       current GCP guidelines and regulatory require-
development processes move `off-site' and away          ments. They should also be capable of suggesting
from the immediate control of the sponsor com-          and directing corrective action when necessary.
pany. The number of variables increases at a time          Orientation of the investigator and clinical staff
when the responsibility for the quality of the          to the study design and test procedures and the
operation must be shared with others at remote          investigational product characteristics is impera-
locations. Clinical investigators must be carefully     tive, to ensure that everyone has a thorough under-
selected after an indication for a product has been     standing of the research goals. A meeting should be
identified and a study plan or protocol has             scheduled with the investigator and clinical staff
been prepared. A visit should be made to each           prior to the enrollment of patients, to review and
investigator's location to evaluate the capabilities    discuss the clinical study in detail. When more than
of the staff and inspect the facilities. The investi-   one investigator is participating in the same study,
gator and any subinvestigators or clinical study        as in a multicenter study, a formal investigator's
support staff who will assist with the evaluation       meeting can be organized at a convenient time and
and treatment of the research subjects should be        location. In advance of the meeting the investigator
qualified by training and experience to undertake       should be provided an opportunity to review a
the research. The investigator must be willing and      copy of the study protocol and current investiga-
able to personally oversee all of the staff who will    tor's brochure. The meeting agenda should include
administer treatments, conduct study-related tests,     a thorough review of the protocol, case report
dispense the investigational product, and record        forms for data collection, investigational product
the data. There should be secure, locked storage        packaging and dispensing, the timing of clinical
space available, away from patient traffic areas, for   laboratory tests and other tests required by the
the investigational product. A laboratory with cur-     study protocol (e.g. physical examination, electro-
rent certification and licensing is needed to perform   cardiogram, etc.). The procedures for reporting
the clinical tests required by the study protocol.      adverse experiences, particularly those that are
The investigator must also have access to a quali-      deemed serious or unexpected, should be thor-
fied institutional review board (IRB) to review and     oughly discussed. It is also helpful to provide an
approve the clinical trial, ensure that the rights of   overview of the investigational product's develop-
human subjects are protected and that they are          ment history, especially toxicology and prior clin-
exposed to no more than minimal risk.                   ical experience, as well as a review of the
   In the early clinical development phases, human      investigator's responsibilities, as outlined in the
subjects, who will usually be healthy, will be con-     GCP guidelines.
fined for varying periods of time while single or          The monitor should arrange a visit to initiate the
multiple doses of the product are administered          clinical study after all of the necessary regulatory
and safety data are recorded. These subjects receive    documents have been filed and the study materials
little in the way of benefit, except monetary com-      have been delivered to the site. This visit will make
pensation. When a reasonably safe and reasonably        it possible to verify their receipt and perform a
effective dose or dose range has been identified, the   complete inventory. If some time has elapsed
safety and efficacy of the product will be deter-       since the prestudy or investigator's meeting, time
mined by studying patients with the indicated           should be set aside to review the protocol and any
symptoms, condition or disease. Some of these pa-       changes that have occurred since the meeting. An
tients will, of necessity, receive no treatment or      inspection of the investigator's regulatory docu-
treatment with inactive substances (placebo).           ments should be made and deficiencies documented
   The sponsor of a clinical study is responsible for   and corrected.
ensuring that it is properly monitored. Monitors           Regulations currently stipulate that an active
should be appropriately trained, and should have        clinical study should be monitored at a minimum
the scientific and clinical knowledge needed to         of once each year, but in practical terms the rate of
monitor a study adequately. They should be famil-       subject enrollment and the volume of data collected
iar with the investigational product, the protocol,     usually dictate a much higher frequency of moni-
written informed consent, applicable SOPs and           toring visits. Each interim or periodic monitoring

visit should include an inspection of the investiga-        An IRB is responsible for the protection of the
tor's regulatory documents, a review of informed         rights and welfare of human subjects. It is author-
consent for subjects enrolled since the previous         ized to conduct the initial review of research studies
monitoring visit, and an inventory of the investiga-     and approve, require modification of, or disap-
tional product. The case report forms used for data      prove any activities covered under existing GCP
collection should be compared to the information         guidelines. A research study includes the study
in the subject's chart, to ensure their accuracy and     plan or protocol, the content of the subject
completeness. Adequate source documentation is           informed consent, and any advertising or other in-
imperative, so that clinical data can be independ-       formation that will be used to direct subjects to the
ently verified, and should include evidence that a       clinical study. An IRB is obligated to continue to
subject is enrolled in the clinical study, that the      review and approve each research study at least
investigational product is being administered            annually, and significant changes to a study plan
according to protocol, and that protocol deviations      or protocol (i.e. changes that may affect or impact
or violations and adverse experiences are properly       the rights and safety of the research subjects) before
documented. Source documents are defined as ori-         the investigator implements them. Investigators
ginal documents, data, and records (e.g. hospital        must provide periodic progress reports to the IRB
records, clinical and office charts, laboratory notes,   as required, and submit a final report at the conclu-
memoranda, subjects' diaries or evaluation check-        sion of the study. Serious adverse experiences must
lists, pharmacy dispensing records, recorded data        be reported promptly.
from automated instruments, copies or transcrip-            Guidelines for IRBs are defined in 21 CFR, Part
tions certified after verification as being accurate     56. They are required to have and to follow written
copies, microfiches, photographic negatives,             procedures. They must consist of at least five
microfilm or magnetic media, X-rays, subject files       members of diverse backgrounds. These individuals
and records kept at the pharmacy, at the laborator-      should be sufficiently qualified through training
ies and at medico-technical departments involved         and experience to consider the research in light of
in the clinical study). After reviewing all of these     the community and institution that they represent,
items, the monitor should be able to confirm that        and be knowledgeable of applicable laws, regula-
the protocol is being followed, that the data are        tions and standards for professional conduct and
true and accurate, that there is adequate supervi-       practice.
sion by the investigator and the institutional review       The investigator must ensure that the subjects
board, and that regulatory obligations are being         selected for the study meet the inclusion criteria,
met. Deficiencies should be addressed at the time        and none of the exclusion criteria, and that they
of the visit and the findings must be documented in      are treated according to the study plan. The IRB
a written report.                                        must ensure that the rights and welfare of the sub-
   The monitor should arrange to close a study site      ject are protected. The monitor must ensure that the
after all subjects have completed the trial, all data    study plan is followed, that all data are true and
have been verified on site, and all investigational      accurate, and that all regulatory requirements are
products have been inventoried and reconciled.           met. Maintaining the quality of clinical research
The investigator's regulatory document files must        depends upon their joint efforts.
be current and complete and archived along with
copies of the case report forms for data collection.
   Provision should be made to communicate any                                 AUDITS
changes to a study plan or protocol, and all serious
adverse experiences, to all investigators involved in    To assure compliance with regulations, govern-
the trial in a timely fashion. These changes must        ment and regulatory agencies throughout the
also be reported to their respective institutional       world must perform official inspections of clinical
review boards. Additions, changes, or corrections        facilities. The FDA employs approximately 1100
to clinical data should be made only by the investi-     investigators and inspectors in 157 cities to inspect
gator, or individuals authorized by the investiga-       almost 95 000 FDA-regulated businesses within the
tor, and must be carefully documented.                   USA. Audits are performed during clinical devel-
                           QUALITY ASSURANCE, QUALITY CONTROL AND AUDIT                                       91

opment to assure the quality of the research data,           During a clinical hold, new subjects may not be
and routinely after product approval to assure ad-        enrolled in clinical studies or treated with the inves-
herence to GMP regulations.                               tigational product. However, in some instances,
   Inspections of clinical study sites can be `routine'   subjects already receiving the investigational prod-
or `for cause'. The former is usually directed to-        uct may be permitted to continue. If significant
wards pivotal trials of a product that is awaiting        deviations from the study program occur or serious
approval, and investigators who enroll large              safety issues are encountered, FDA can request
number of subjects in a study. The latter is usually      termination of a clinical program or withdrawal
reserved for cases where there is reason to suspect       of an IND.
the validity of data generated at a clinical site. Fur-      A routine investigator site audit will focus on two
ther, if significant deviations from regulations are      major areas. First, an auditor will examine the facts
discovered during an audit, a routine inspection          surrounding the conduct of the study to determine
may become a `for cause' inspection and be intensi-       who did what, the degree of delegation of authority,
fied or expanded to include other studies. When           where specific aspects of the study were performed,
violations are discovered, the FDA can request vol-       how and where data were recorded, how investiga-
untary corrective action or recall a product from the     tional product accountability was maintained, and
market if it has already been approved. Failure to        how the monitor communicated with the investi-
voluntarily correct a problem can lead to legal           gator and evaluated the progress of the study.
sanctions, seizure and destruction of a product,          Second, the study case report forms for data collec-
and criminal penalties. Audits by other regulatory        tion will be compared to all available subject
agencies outside the USA follow similar proced-           records and source documents that might support
ures. Given the continuous evolution of oversight         the study data. Subjects' records before and after
activities employed by these nations, their review        the study can also be audited to evaluate medical
may be either more or less stringent than that of the     histories that pre-date enrollment, and to determine
FDA.                                                      what degree of follow-up occurred after treatment
   During its continuing review of product develop-       with the investigational product. Informed consent
ment prior to marketing approval, FDA can re-             for all subjects enrolled at a site are routinely exam-
quest a hold on a clinical development program if         ined, and an auditor will also want to determine
it determines that:                                       whether all adverse experiences have been properly
                                                          reported. Nearly 10 000 inspections at more than
. Human subjects would be exposed to an `un-              4000 investigative sites have been completed since
  reasonable and significant risk of illness or           1964. The most frequently encountered deficiencies
  injury'.                                                continue to be inadequate subject informed con-
. One or more investigators are not qualified by          sent, failure to adhere to the protocol, inadequate
  reasons of scientific training and experience.          and inaccurate records, and inadequate drug ac-
. The investigator's brochure is determined to be         countability. The most frequently encountered de-
  erroneous, incomplete or misleading.                    ficiencies in informed consent are failure to identify
. The information submitted under the IND is              a contact person for questions about research sub-
  insufficient, or the design of a study or clinical      jects' rights, incomplete descriptions of study pro-
  program is deemed to be deficient or lacking in         cedures and compensation/treatment for injury,
  scientific merit.                                       inadequate confidentiality statements, and an in-
. The manufacturing, control, and labeling of the         complete description of available alternative pro-
  investigational product are substandard with            cedures.
  respect to identity, quality, or purity, or insuffi-       FDA can audit the records of the study sponsor,
  cient quantities exist to adequately conduct a          as well as contract research organizations (CROs),
  clinical trial.                                         to review standard operating procedures and all of
. A satisfactory alternative therapy becomes              the documents related to a particular clinical study,
  available or evidence strongly suggests that            including electronic databases. Since 1981, nearly
  the investigational product is unsafe or ineffect-      200 such inspections have occurred. Departures
  ive.                                                    from regulations requiring voluntary corrective

action were reported for 10% of the sponsors             also provide a means to assess both the quality of
and CROs inspected. Two of the inspections               the investigator site and the quality of the monitor-
resulted in official action. The most frequently         ing throughout the study. Audits performed out-
reported deficiencies were records that were un-         side the USA, in conjunction with multinational
available, inadequate or inaccurate, inadequate or       studies, can be problematic due to language and
improper informed consent, inadequate investiga-         cultural differences, as well as a diversity of gov-
tional product accountability, deviations from the       ernment regulations.
protocol and, inappropriate payments to volunteer           When the product development program has
research subjects.                                       been completed and the NDA has been submitted
   Since 1986, FDA has performed nearly 2100             for review, the FDA will likely perform routine
audits of more than 1400 public (e.g. affiliated         inspections of the manufacturing facilities to verify
with a university, hospital or other institution)        GMP compliance and a sample of the clinical study
and private institutional review boards. Fewer           sites to verify GCP compliance.
than 20 of these were `for cause'. Complete compli-         An audit of manufacturing and control processes
ance was observed for only about 13% of these            will include a thorough review of all of the docu-
inspections and 1% uncovered violations that             mentation required to demonstrate that a product
were serious enough to warrant regulatory or ad-         meets the requirements for its intended use, includ-
ministrative sanctions, which included suspension        ing requirements for marketing and consumer use,
of clinical studies. The most frequently encoun-         technical design and performance, regulatory and
tered deficiencies have been inadequate meeting          quality assurance, and product safety. Equipment
minutes, lack of a quorum at meetings; inadequate        systems will also be inspected.
written procedures, and inadequate continuing
   At the conclusion of a FDA audit, the inspector                       FUTURE TRENDS
will conduct an exit interview to discuss deficiencies
observed and, if appropriate, will issue an Estab-       Computer and electronic media, originally
lishment Inspection Report (Form 483). For many          employed for databases and statistical analysis,
sites, voluntary action to correct deficiencies will     have now expanded to include remote data entry
then be requested, some in writing. Failure to           at investigator sites and transfer of documentation
voluntarily correct deficiencies, or the occurrence      of NDA submissions. This has prompted the devel-
of serious deficiencies at some sites, may result in a   opment of regulatory guidelines to ensure that soft-
delay in the approval of a new drug application          ware systems are properly designed, tested,
(NDA), or disapproval of the application, disquali-      validated, and upgraded, that they include ad-
fication of an investigator, restrictions on an          equate security measures and provisions for docu-
investigator, or criminal charges, fines and impris-     menting changes to data (an audit trail), and that
onment. There are currently about 80 physicians          the individuals who use the systems receive ad-
who are ineligible to receive investigational prod-      equate training. Auditors will require training to
ucts and more than two dozen others who have             keep pace with the development of these systems
agreed to some restriction of their use.                 and to effectively audit them in the future.
   The quality assurance department within a com-           The globalization of pharmaceutical research
pany will frequently conduct audits of investigator      and development is expected to shorten the time
sites, either during a clinical study or after it has    required for product approval and registration, by
been completed. Companies that do not have a             increasing the number of subjects available for re-
quality assurance department or unit may contract        search studies and broadening an investigational
this service from a consultant or a third party          product's exposure to ethnic intrinsic factors (gen-
specializing in regulatory audits. The intent and        etic and physiological) and extrinsic factors (cul-
scope of these audits are the same as the FDA,           tural and environmental).
but they afford a company an opportunity to verify          The pharmaceutical industry is already establish-
compliance and to correct deficiencies that might        ing a presence in Central and Eastern Europe and is
otherwise delay product approval. These audits           aggressively moving towards South America and
                              QUALITY ASSURANCE, QUALITY CONTROL AND AUDIT                                                 93

China. Japan remains focused on possible genetic                 Been and Where We are Going. Prentice Hall Law and Busi-
differences, and continues to insist upon studies                ness: New York.
                                                               Code of Federal Regulations, Title 21, Ch. 1 (4-1-97 edn), Part
conducted in its own population as a condition for               58, Good Laboratory Practice for Nonclinical Laboratory
registration and approval. Regional variations,                  Studies.
such as diet, alcohol and tobacco consumption,                 Code of Federal Regulations, Title 21, Ch. 1 (4-1-97 edn), Part
climate, exposure to pollution and other environ-                211, Current Good Manufacturing Practice for Finished
mental factors, socioeconomic status, and differ-
                                                               CPG (1991) Compliance Policy Guide: Fraud, Untrue State-
ences in technology and health care standards,                   ments of Material Facts, Bribery and Illegal Gratuities.
require creative planning for multinational trials.              CPG: 7150.09, September.
   As nations adopt more open trading policies                 DeSain C (1993) Documentation Basics That Support Good
throughout the world and form economic alliances,                Manufacturing Practices. Aster: Eugene, OR.
                                                               DeSain C, Sutton CV (1996) Documentation Practices. A Com-
the trend toward multinational studies will require              plete Guide to Document Development and Management for
increased cooperation between governments and a                  GMP and ISO 9000 Compliant Industries. Advanstar Com-
greater degree of harmonization to level the                     munications: Duluth, MN.
playing field in drug development. In addition to              Department of Health and Human Services, Public Health Ser-
the ICH Guideline for good clinical practice (GCP)               vice, Center for Drug Evaluation and Research, Office of
                                                                 Training and Communication (1998) Freedom of Information
already published, unified GMP and GLP guide-                    Office Staff. HFD-205, January.
lines have been drafted and some portions are al-              Dumitriu H (1998) Impact of the ICH Guideline of Ethnic
ready approved. Efforts such as these will                       Differences. Drug Inf J 32: 141±4.
hopefully pave the way for mutual acceptance of                FDA (1991) Bioresearch monitoringÐhuman drugs; clinical
all research and development data by the regula-                 investigators. In Compliance Program Guidance Manual, Pro-
                                                                 gram 7348.811. September, Chapter 48.
tory authorities of these and other nations.                   FDA (1991) Bioresearch monitoringÐhuman drugs, biologics,
                                                                 medical devices and radiological health products and veterin-
                                                                 ary drug products; sponsor, contract research organizations,
                      SUMMARY                                    and monitors. In Compliance Program Guidance Manual, Pro-
                                                                 gram 7348.810, October, Chapter 48.
                                                               FDA (1992) Bioresearch monitoringÐinstitutional review
Over the years, the pharmaceutical industry has                  boards. In Compliance Program Guidance Manual, Program
endeavored to work in partnership with the regula-               7348.809, January, Chapter 48.
tory agencies to perfect the laws and regulations              FDA (1997) The Food and Drug Administration: An Overview.
that govern research and development. These laws                 Publication No. BG95±13, July 15.
                                                               Guertler-Doyle V (1998) Taking clinical trials into China. Appl
and regulations will continue to evolve as technol-              Clin Res Trials 7(6): 54±60. June.
ogy and health care capabilities change to meet the            ICH (1997) International Conference on Harmonization: Guide-
growing needs of our world. Accompanying this                    line for Good Clinical Practice, May 9.
growth will be the need for the continued assurance            Kingham R (1988) History of FDA Regulation of Clinical
of human safety and well-being. This must remain                 Research. Drug Inf Assoc J 22: 151±5.
                                                               Natoroff BL (1998) Clinical Trials in Central/Eastern Europe:
the foremost consideration in pharmaceutical re-                 Industry Viewpoint. Drug Inf J 32: 129±33.
search and development.                                        Schoichet S, Creasy G, Kasay S (1998) A systems infrastructure
                                                                 for sponsor/CRO collaboration on international clinical
                                                                 trials. Drug Inf J 32: 155±61.
                  BIBLIOGRAPHY                                 FDA (1997) US Department of Health and Human Services,
                                                                 Food and Drug Administration, Center for Drug Evaluation
Barton BL (1990) FDA's inspections of US and non-US clinical     and Research Fact Book, May.
  studies. Drug Inf Assoc J 24: 463±8.                         Warnock-Smith A (1998) Proposed EC harmonization directive
Buc NL, Hutt PB (1988) Recent Changes in the FDA Drug            would increase regulatory burden. Appl Clin Res Trials 7(1):
  Approval Process for Drugs and Biologics: Where We Have        28±31.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

         Phase I: The First Opportunity for
    Extrapolation from Animal Data to Human
 Stephen Curry1, Dennis McCarthy1, Heleen H. DeCory1, Matthew Marler1 and
                            Johan Gabrielsson2
             Astra Arcus USA Inc, Rochester, NY, USA, and 2Astra Arcus AB, Sodertalje, Sweden

There is a need to make reliable and rapid predic-              in itself, as both the end of the research and discov-
tions of human responses from animal data. Al-                  ery process and the beginning of the clinical
though drug discovery is primarily designed to                  process, and none can doubt that the first-in-man
find compounds with desired efficacy, most re-                  study, and achieving an active Clinical Trials Ex-
search-orientated pharmaceutical companies also                 emption (CTX) or Investigational new drug appli-
use data on absorption, metabolism and pharma-                  cation (IND) (or equivalent regulatory permissions
cokinetics in the decision-making process (Welling              in other countries; see Chapters 23, 25 and 26), are
and Tse 1995). Usually, the strategy is to use all              important milestones in the history of a drug. The
available data to choose one or two candidates                  essence of this crucial step of drug development is
from a whole pharmacological class of new drugs                 the making of valid predictions of in vivo drug
for Phase I testing (Welling and Tse 1995). Thus,               effects from in vitro data.
compounds are chosen on the basis of animal data,                  The collection of in vitro data from animal
partly because of suitable bioavailability, half-life,          materials and extrapolation: (a) from physical
and tissue penetration characteristics. The possibil-           properties to in vitro data; (b) from in vitro data to
ity of multiphasic plasma level decay patterns                  non-human in vivo data; and (c) from non-human in
following intravenous doses is an important elem-               vivo data to clinical in vivo responses, can be done
ent in this selection process.                                  more efficiently using on-line analysis and simula-
   Data for chosen compounds will commonly also                 tions. This chapter seeks to show how rapid pro-
have been subjected to simultaneous modeling of                 gression may be achieved for new chemical entities
pharmacokinetic and pharmacodynamic data from                   through this process, using in vitro and in vivo data
animals, again in an effort to optimize the chances             and advanced modeling procedures. This must be
that the drugs chosen will have the properties in               seen in the context of the entire drug discovery
humans specified in a pre-discovery product                     process, which, on a larger scale, is designed to
profile. The pharmacodynamic information avail-                 find potent, safe drugs (in man), based on animal
able typically includes data from receptor-binding              data (Figure 10.1). We anticipate a time when in
studies, in vitro functional assays, and in vivo                vitro pharmacodynamic data will be routinely com-
pharmacological screening experiments. Pharma-                  bined with in vitro drug metabolism data in a ra-
cokinetics, related when possible to observed drug              tional prediction of drug responses in healthy
effects, is a powerful and critical component of the            human volunteers, with consequent acceleration
pivotal step from animal research to human re-                  of the drug discovery effort, and therefore a general
search in the drug development process. For                     trend for more efficient use of resources in early
many drug researchers, Phase I is an end-point                  clinical development.

Physicochemical                                                          Elementary Aspects of Clearance
                                                                  The common, clinical measurement of drug clear-
              In vitro            In vivo
                                                                  ance involves taking serial venous blood samples.
                 rat                rat                           As time passes after dosing, drug concentrations are
                                                                  seen to decline: this is really merely the modeling of
  (i) Microsomes
 (ii) Hepatocytes                       Bioanalysis    Second     drug disappearance, and is essentially a descriptive
                                                      Mammalian   process, requiring actual human exposures. First-
                                                       Species    order elimination, after equilibrium in the circulat-
              In vitro            In vivo                         ing compartment, has a constant (k) with units of
              human               human
                                                                  hÀ1 , and plasma concentration (C ) is then modeled
                                                                  by equations of the general form:
                      Species               Effect
                              (Pharmacokinetic-Pharmacodynamic                         C ˆ AeÀkt
Figure 10.1 General scheme showing the pharmacokinetic pre-       where A is the concentration of drug at time (t) ˆ 0,
diction pathway from physicochemical properties to human          assuming that there was instantaneous and homo-
drug response via in vitro and in vivo studies in laboratory
                                                                  genous equilibration of the dose into the circulating
                                                                  compartment. As the number of compartments
                                                                  increases, then so do the number of terms of the
     THE IN VITRO/IN VIVO PREDICTION                              form shown on the right-hand side of the equation
                                                                  shown above.
The challenge is to predict systemic clearance,                      The elimination rate always has units of (mass/
volume of distribution, and oral bioavailability in               time) for any elimination process. For first-order
man from a combination of in vitro and in vivo                    processes, the elimination rate is represented by a
preclinical data. If this prediction can become reli-             tangent to the elimination curve.
able, then Phase I studies become more confirma-                     In contrast, zero-order elimination processes are
tory. The use of human hepatocytes and isolated                   occasionally encountered. These usually represent
enzymes can form a critical part of the in vitro                  saturation by the drug of the elimination mechan-
database.                                                         ism(s). These `drug disappearance' curves are
   Clearance of almost all drugs is by renal, meta-               straight, and thus described simply, by:
bolic, and/or biliary mechanisms. There are rare
exceptions, such as anaesthetic gases that are ex-                                    C ˆ A À bt
haled unchanged. However, here we shall concen-
trate on the typical situation.                                   where the elimination rate (b) does not change with
   Physicochemical properties, especially lipophili-              time or drug concentration. If followed for long
city, frequently govern the clearance route; lipophi-             enough, most drugs that are subject to zero-order
licity is commonly measured as log D7:4 , where this              elimination eventually fall to such low concentra-
variable equals log10 ([drug in octanol]/[drug in                 tions that the elimination mechanism becomes
buffer]) at pH ˆ 7:4, in a closed system at equilib-              unsaturated, and first-order elimination then super-
rium. Generally compounds with a log D7:4 value                   venes; good examples include ethanol and sodium
below 0 have significant renal clearance values,                  dichloroacetate (Hawkins and Kalant 1972; Curry
whereas compounds with log D7:4 values above 0                    et al 1985; Fox et al 1996).
will principally undergo metabolism (Smith et al                     The elimination rate for zero-order processes
1996). Molecular size also has some effect on                     may also be treated as a maximal rate of reaction
these clearance routes. For example, compounds                    (Vmax ), and thus this type of data may be subject to
with molecular weights greater than 400 Da are                    ordinary Michaelis±Menten analysis (see further,
often eliminated through the bile unchanged,                      below). Note that first-order elimination curves
whilst smaller lipophilic compounds will generally                are so common that `drug disappearance' curves
be metabolized.                                                   are routinely analyzed as semi-logarithmic plots
                                             ANIMAL±HUMAN EXTRAPOLATION                                                             97

(which linearizes the curve). The literature is some-                  dicted using simple enzyme kinetic data (Houston
times ambiguous in its use of the term `linear data',                  1994; lwatsubo etal, 1996 Obach 1996a; Ashforth
authors may or may not assume that the semi-                           et al., 1995). These predictions may be streng-
logarithmic transformation is to be taken as read.                     thened by comparing preclinical in vivo data with
  When the elimination rate is known, then clear-                      the predictions made from in vitro data using
ance (Cl ) is defined simply as:                                       tissues from the same preclinical species. Rane et
                                                                       al 1977 As an illustration, consider a novel com-
                 Cl ˆ elimination rate=C                               pound currently in development, identified as
                                                                       Compound X. This compound has a molecular
where C is again drug concentration. Note that in                      weight less than 400, and a log D7:4 value of ap-
first-order elimination processes, the elimination                     proximately 0.5, suggesting that it could undergo
rate of the drug (with units of mass/time) changes                     both renal and hepatic clearance. Preclinical in vivo
with time (and drug concentration), and thus only                      studies indicate that Compound X is eliminated
instantaneous clearances, specifying time or drug                      largely unchanged in the urine in the rat ($ 90%).
concentration, can be stated.                                          Several oxidative biotransformation pathways
   Urinary clearance, obviously, may only partly                       have nonetheless been identified. In common with
explain the rate of drug disappearance from                            studies of Compound X clearance in humans,
plasma. In any case, the urinary clearance of an                       simple in vitro enzyme kinetic studies were used in
agent may be found from the familiar equation:                         conjunction with knowledge from rat in vivo data.
                                                                       The general strategy for prediction of kinetic stud-
                       Cl ˆ (U  V )=P                                 ies, is shown in Figure 10.2.
                                                                          Using liver microsomes from different species,
where U is the urinary concentration, V is the volu-                   the intrinsic clearance (Clint ) for each species can
me of urine excreted during a specified time period,                   be determined, and then scaled to hepatic clear-
and P is the average plasma concentration during                       ance. This is typically done by first determining in
that time period. Pharmaceutical physicians will                       vitro Km (the Michaelis±Menten constant) and
remember that for inulin and sodium iothalamate,                       Vmax (the maximal rate of metabolism) for each
but not for creatinine or urea, the urinary cleara-                    metabolic reaction, using substrate saturation
nce is a good measure of glomerular filtration rate.                   plots (using the familiar algebra and, because of
   These elementary aspects of clearance may be                        enzyme saturation, finding that Clint ˆ Vmax =Km ).
revised in any textbook (e.g. Curry 1980; Benet                        However, for Compound X, the situation is more
et al 1996). The purpose of the remainder of this                      complicated because we know that the Clint (drug
section is to show how much more informative the                       disappearance) actually is due to several combined
concept of clerance may be, and to provide an                          biotransformation pathways (i.e. Clint (total)
                                                                             H       H         H
illustration of its use.                                               ˆ Clint 1 ‡ Clint 2 ‡ Clint 3 ‡ . . .), thus complicating
                                                                       any Km and Vmax determinations from a simple
                                                                       substrate saturation plot.
     Prediction of Human Drug Clearance                                   To determine the Clint of compound X, we are
                                                                       able to use the in vitro half-life method, which is
For those compounds predominantly cleared by                           simpler than finding all the component Clint values.
metabolism, human blood clearance can be pre-                          When the substrate concentration is much smaller

  Km, Vmax for metabolic pathway
                                                             Scaling                            Blood flow
                  or                        In vitro Clint               In vivo Clint                              Hepatic clearance
                                                                                           Serum protein binding
                                                                                         Microsomal protein binding
         T1/2 for drug loss

Figure 10.2 Strategy for the in vitro±in vivo scaling of hepatic clearance (see for example Iwatsubo et al., 1996)

than the Km , the Michaelis±Menten equation sim-                     The amount of microsomal protein/g liver is con-
plifies from velocity (V ) ˆ Vmax ([S])=(Km ‡ [S ]) to               stant across species (45 mg/g liver). Thus, the only
V ˆ Vmax =Km , because ([S ] is substrate concentra-                 species-dependent variable is the weight of liver
tion) becomes negligible. Furthermore, under these                   tissue/kg body weight.
conditions the in vitro half-life (T1=2 ˆ 0:693=Kel )                   In vivo, hepatic clearance is determined
can be measured, and this, in turn, is related to the                by factoring in the hepatic blood flow (Q), the
Michaelis±Menten equation through the relation-                      fraction of drug unbound in the blood (fu), and
ship velocity (V ) ˆ volume  Kel (where volume is                   the fraction of drug unbound in the microsomal
standardized for the volume containing 1 mg of                       incubations ( fu(inc) ), against the intrinsic clearance
microsomal protein). When both V and Vmax are                        of the drug by the whole liver (the in vivo Clint ). The
known, then the Km is also found. Although simpler                   fu and fu(inc) are included when the drug shows
than finding a complicated Cint , one caveat of the in               considerable plasma or microsomal protein bind-
vitro half-life method is that one assumes that the                  ing (Obach 1996b). Several models are available
substrate concentration is much smaller than the                     for scaling in vivo intrinsic clearance to hepatic
Km . It may be necessary to repeat the half-life deter-              clearance, including the parallel tube model or si-
minations at several substrate concentrations, and                   nusoidal perfusion model, the well-stirred
even model the asymptote of this relationship, be-                   model or venous equilibration model, and the dis-
cause very low substrate concentrations that are                     tributed sinusoidal perfusion model (Wilkinson
beneath biochemical detection may be needed to                       1987).
fulfill the assumptions needed to simplify the                          Thus far, for Compound X, we have obtained
Michaelis±Menten equation.                                           good results in this context with the simplest of
   Note that in this in vitro application, intrinsic                 these, the well-stirred model (see Table 10.1 for the
clearance, like all conventional mathematical evalu-                 equations, with and without significant plasma and/
ation of clearances, has units of volume Á timeÀ1 .                  or microsomal protein binding). Using this well-
It is obtained from Vmax and Km measurements,                        stirred model, it has proved possible to predict the
where Vmax has units of mass Á timeÀ1 . The defin-                   hepatic clearance from in vitro intrinsic clearance
ition of intrinsic clearance as Vmax Á Km should not                 rates in the rat, dog and human (Table 10.2). The
be confused with the historically prevalent calcula-                 hepatic clearance value for the rat (0:972 mlÁ
tion of kel (the first-order rate constant of decay of               minÀ1 Á mgÀ1 protein) was approximately one-
concentration in plasma), calculated from                            tenth the actual clearance found in vivo; well in
kel ˆ Vmax Á Km , where Vmax is the zero order rate                  agreement with the observation that in vivo Com-
of plasma concentration decay seen at high concen-                   pound X was eliminated by the rat, largely un-
trations, and Kmax is the concentration is plasma at                 changed, by the kidneys ($ 90%).
half-maximal rate of plasma level decay.                                To predict hepatic clearance of Compound X in
   Once the in vitro intrinsic clearance has been de-                man, human in vitro intrinsic clearance could then
termined, the next step, scaling in vitro intrinsic                  be scaled to hepatic clearance, using a technique
clearance to the whole liver, proceeds as follows:                   that had been validated in the rat. Ashfortt et al
                                                                     1995Renal clearance is subject to an allometric
          H                H
In vivo Clint ˆ in vitro Clint  weight microsomal                   relationship and can generally be scaled across
                                                                     species (see below). The predicted in vivo renal Cl
                                protein=g liver                      for the rat (estimated by multiplying the predicted
                  Â weight liver=kg body weight                      hepatic Cl by 9) may be scaled allometrically to

            Table 10.1 Equations for predicting hepatic clearance using the well-stirred model

            In the absence of serum or        In the presence of significant       In the presence of both serum and
            microsomal protein binding        serum protein binding                microsomal protein binding
                                H                                      H                                    H
                          Q  Clint                         Q  fu  Clint                       Q  fu  Clint  fu(inc)
            Clhepatic ˆ         H             Clhepatic ˆ              H           Clhepatic ˆ              H
                          Q ‡ Clint                         Q ‡ fu  Clint                       Q ‡ fu  Clint  fu(inc)
                                                ANIMAL±HUMAN EXTRAPOLATION                                                                        99

     Table 10.2 Comparison of the predicted in vivo hepatic clearance and the actual clearance values for Compund X

                 Predicted in vivo hepatic         Predicted in vivo renal          Predicted in vivo total           Actual in vivo
                                   À1                               À1
                 CL (ml minÀ1 kg )                 CL (ml minÀ1 kg )                CL (ml minÀ1 kgÀ1 )               CL (ml minÀ1 kgÀ1 )

     Rat         0.972                             8.75                             9.72                               8.17±10.7
     Man         0.223                             1.93                             2.15                               1.87±2.45
     Dog         0.463                             3.74                             4.20                              21.2±22.5

     Predicted values were scaled from in vitro half-life data using liver microsomes and the well-stirred model of hepatic extraction. Hepatic
     CL predictions were corrected for plasma and microsomal protein binding. Predicted total CL was obtained by adding in renal CL
     estimates which were, in turn, scaled allometrically (Y ˆ aW 0:75 ).

obtain a prediction for human in vivo renal clear-                           encountered in such scaling exercises, which war-
ance. Total or systemic Cl in man can then be                                rant restating.
estimated by adding the two clearance parameters                                The first caveat is that all clearance pathways
(hepatic and renal) together; in practice, for Com-                          (hepatic, renal, biliary, or other) must be taken
pound X, later first-in-man data revealed an actual                          into consideration. If a compound undergoes a
in vivo Cl nearly identical to the predicted total Cl                        high level of hepatic clearance, then in vitro±in
(2.15 vs. 1.87±2.45 ml minÀ1 mgÀ1 , respectively;                            vivo scaling may be used to predict the fraction
Table 10.2). Here, then, is a real-world example                             of systemic clearance expected from this pathway.
of, first, how rat in vitro and in vivo preclinical                          If a compound undergoes a high level of renal
data were used to develop and validate a scaling                             elimination, allometric scaling may be also used
method for Compound X in the rat; and second,                                to predict the clearance attributed to this
how the scaling method successfully predicted in                             pathway.
vivo overall drug clearance in man.                                             The second caveat is that, in order to accurately
   However, if the same methods are used for Com-                            predict hepatic clearance, the correct in vitro system
pound X in the dog, things initially appear to be                            must be chosen. If the candidate drug is primarily
different. Scaling the in vitro intrinsic clearance                          oxidatively metabolized, then liver microsomes will
to hepatic Cl using the rat-validated method, in                             be sufficient. However, if the potential for non-
conjunction with allometric scaling of renal Cl,                             microsomal biotransformation exists, then a differ-
resulted in a five-fold underprediction of total or                          ent in vitro system, such as hepatocyte suspensions,
systemic clearance in vivo. However, further metab-                          should be used. In the illustration above, it turned
olism studies in the dog in vivo revealed that                               out, as far as clearance of Compound X is con-
Compound X undergoes significant additional bio-                             cerned, man is specifically like a rat, and unlike
transformation, particularly N-methylation, which                            a dog.
is unique (as far as we are aware) to this species,                             The third caveat is that one must consider the
and invalidates some of our in vitro assumptions.                            variability in the expression of metabolizing
This canine biotransformation pathway was not                                enzymes between individuals. Oxidative metabol-
detected by our initial microsomal studies because                           ism (seen in vivo and in microsomal enzymes), and
there are no N-methyl transferases in microsomes.                            especially cytochrome P450s, vary tremendously be-
Thus, although we did not successfully predict                               tween human individuals (Meyer 1994; Shimada et
dog systemic clearance for Compound X, our                                   al 1994). Had we used a single donor microsomal
scaling tactics did eventually teach us about a new                          sample, rather than pooled liver microsomes (a pool
clearance mechanism, and how important this                                  consisting of at least eight individual donors), to
was for the systemic clearance of Compound X in                              scale in vitro data to in vivo hepatic clearance, we
the dog.                                                                     might have made greatly misleading predictions
   This is an example of how in vitro studies                                (note that oxidative, initial drug metabolism is
can be combined with in vivo preclinical data,                               sometimes called `Phase I metabolism' in the litera-
leading to useful prediction of human systemic                               ture, causing ambiguity with the stage of drug de-
drug clearance. Nonetheless, several caveats are                             velopment or type of clinical trial).

              Volumes of Distribution                              preclinical species, and in vivo volume of distribu-
                                                                   tion in that same preclinical species:
          Review of Elementary Concepts
                                                                                          VDpreclinical species  fuhuman
Volume of distribution is a theoretical concept                            VDhuman ˆ
that may or may not correspond to the anatomical                                               fupreclinical species
compartment(s) which drugs or metabolites may
access after dosing. When size of the dose (D) is                     Table 10.3 shows the predicted volume of distri-
known, and when drug concentration (C ) may be                     bution of a single intravenous bolus dose of Com-
found by sampling biological fluids, then, in the                  pound X in man; this is found by using the above
simplest case, the volume of distribution (VD) is:                 equation, an in vitro estimate of protein binding
                                                                   data for rat and dog plasma, and the observed
                       VD ˆ D=C                                    volumes of distribution for these two species in
                                                                   vivo. For man, VDhuman was predicted to be 3:48À
Clinical protocols can usually only prescribe the                  4:59 l kgÀ1 using the rat data and 3:01À5:06 l kgÀ1
sampling of a subset of compartments when a                        using the dog data.
drug is known to distribute widely in the body.
For example, a lipophilic drug may penetrate lipo-
philic organs such as brain, and, obviously, brain                   Elementary Aspects of Oral Bioavailability
sampling simply for pharmacokinetic purposes is
usually possible only in animals. In such cases,                   The oral bioavailability (F ) of a drug is dependent
blood concentrations fall far lower than if the dose               on (a) the absorption of the drug from the gastro-
had distributed solely into the circulating compart-               intestinal (GI) tract, and (b) capability of the liver
ment; C becomes very small, and VD becomes cor-                    to clear the drug during its first pass through the
respondingly very large. The opposite effect would                 portal venous system. Oral bioavailability may be
require the drug to be restricted to a fraction of the             described as the fraction of the total oral dose for
compartment that is sampled, essentially suggest-                  which systemic exposure is achieved. It is a meas-
ing that too few compartments have been postu-                     urement of extent of exposure, and contrasts with
lated, and is almost never encountered. Again, see                 the rates of absorption or elimination discussed
Curry (1980) or Benet et al (1996) for expansion                   above.
of these elementary aspects of volume of distribu-                    Clinically, F is found by comparing the systemic
tion.                                                              exposures that result after intravenous and oral
                                                                   doses of the same drug. Note that this comparison
                                                                   need not be for doses of the same size (an import-
 Prediction of Human Volumes of Distribution
                                                                   ant consideration when the pharmaceutical phys-
The free (not plasma protein-bound) volume of                      ician assesses the tolerability aspects of a proposed
distribution of experimental drugs is generally con-               normal volunteer study). It is, in fact, preferable to
sidered to be constant for all species. Thus, the                  achieve concentrations in the same range from the
volume of distribution in man can easily be pre-                   two doses. Typically, Cmax for a standard dose is
dicted through a simple proportionality between in                 going to be higher after bolus intravenous dosing
vitro plasma protein binding data in man and in a                  (IV) than after oral administration (PO), and ad-

      Table 10.3 In vitro plasma protein binding, In vivo volume of distribution and predicted volume of distribution in man

                    Fraction of Compound X            In vivo volume of distribution      Predicted volume of distribution
                    unbound in the plasma ( fu )      (l . kg)                            in man (l . kg)

      Rat           0.45                              3.02±3.97                           3.48±4.59
      Human         0.52                              ±                                   ±
      Dog           0.66                              3.82±6.43                           3.01±5.06
                                      ANIMAL±HUMAN EXTRAPOLATION                                           101

verse effects of new agents are likely to be concen-    or geometric average is a controversy, with re-
tration-dependent. The relevant equation is:            spected protagonists on both sides.

        F (%) ˆ [(AUCPO  DoseIV )=
                 (AUCIV Â DosePO )] Â 100                      Prediction of Oral Bioavailability

where AUC is the area under the time±plasma con-        Oral bioavailability can be predicted using the
centration curve after each of the respective admin-    following equation:
istrations (the dose terms cancel when equally sized
doses are administered by both routes of adminis-                       F ˆ Fa Á (1 À Cl=Q)
tration). A residual of less than 15% (sometimes
10%) of the total AUC is a commonly-used stand-         where Fa represents the fraction of drug absorbed
ard for timing the last plasma sample. These studies    through the intestinal lining, Cl is the hepatic clear-
are usually conducted under standard conditions,        ance (predicted from in vitro studies, see earlier
and using crossover protocols, although, occasion-      section) and Q is the hepatic blood flow in man
ally, a double-label study may be used to measure F     (see, for example, Rane etal., 1977). Octanol/water
instantaneously (see Chapter 12). Comparison of         partitioning has traditionally been used to predict
generic with innovator's formulations, and slow-        the fraction absorbed through the intestinal lining.
release with rapidly absorbed formulations, may         Recently, Caco-2 cell permeability studies have re-
be compared using equations of the same form.           placed the use of octanol/buffer partitioning stud-
Similarly, subcutaneous and intravenous injections      ies. Yee (1997) established a relationship between
can be compared. With very rare exceptions, the         Fa and Caco-2 cell permeability, expressed as the
intravenous administration of a dose is assumed to      apparent permeability constant (Papp ), as follows:
be 100% bioavailable. For example, very short-
acting drugs, e.g. some arachidonate derivatives,       Papp < 10À6 cm=s, then Fa ˆ 0 À 20%
remifentanil, esmolol and adenosine, may be me-         1    Papp    10  10À6 cm=s, then Fa ˆ 20 À 70%
tabolized during their first return circulation after
intravenous administration, and still not achieve       Papp > 10À5 cm=s, then Fa ˆ > 70%
100% `bioavailability'. Also, the concept is not ap-
plicable to topically-acting drugs. However, assess-    The use of Caco-2 cell permeability studies has
ing the bioavailability of these drugs by any other     resulted in more accurate oral bioavailability pre-
route of administration is usually pointless, unless    dictions. Using the predicted hepatic clearance for
there is some highly specialized issue, e.g. absorp-    Compound X in man (see above), estimating Fa by
tion after intrathecal administration or potential      extrapolation from the Caco-2 cell Papp , and an
for drug abuse.                                         assumed hepatic blood flow for man (see, for
   Fluctuation of plasma drug concentration is an       example, Rane et al., 1977) of 20 ml/min/kg, the
important aspect of the bioavailability of slow re-     human oral bioavailability of 69±98% for Com-
lease formulations, which almost always have lower      pound X is predicted. This compares well with the
Cmax values for a standard dose size than, albeit       known oral bioavailability of this compound in
similar AUC to, a more rapidly absorbed tablet.         rats and dogs (83% and 72%, respectively).
Assuming that the assay can handle the inevitably
lower plasma concentrations, then a useful measure
of fluctuation, after the initial absorption phase of        PREDICTION FROM ANIMALS TO
the curve, and during the next four half-lives of                  HUMANS IN VIVO
elimination, is:
                                                                       Elementary Aspects
                (Cmax À Cmin )=Cavg
                                                        Allometric scaling is an empirical method for pre-
where Cavg is the average concentration during the      dicting physiological, anatomical, and pharmaco-
specified time period; whether to use the arithmetic    kinetic measures across species in relation to time

and size (Boxenbaum 1982; Boxenbaum and DiLea                                This allometric relationship between body surface
1995; Ings, 1990). Allometric scaling is based on                            area and species body weight then allows for a
similarities among species in their physiology,                              simple conversion of drug doses across species
anatomy, and biochemistry, coupled with the ob-                              (Figure 10.3), and allometrically eqivalent doses
servation that smaller animals perform physio-                               of drugs (mg/kg) can be calculated for any species
logical functions that are similar to larger                                 (Table 10.4). The conversion factor (km is simply
animals, but at a faster rate. The allometric equa-                          the body weight divided by the body surface area.
tion is Y ˆ aW b , and a log transformation of this                          Thus, by using the km factors, the dose in Species
formula yields the straight line:                                            1 (in mg/kg) is equivalent to (kmspecies2 =kmspecies1 )
                                                                             times the dose in Species 2 (in mg/kg). For example,
                log Y ˆ b log W ‡ log a,                                     a 50 mg/kg dose of drug in mouse would be equiva-
                                                                             lent to a 4.1 mg/kg dose in human, i.e. approxi-
                                                                             mately one-twelfth of the dose (Table 10.4).
                                                                             Likewise, the conversion factor can be used to cal-
                                                                             culate equivalent doses between any species. An
Y ˆ the pharmacokinetic or physiological variable
                                                                             equivalent dose in mg/kg in rat would be twice
    of interest
                                                                             that for the mouse.
a ˆ the allometric coefficient (and log a is the
    intercept of the line)
W ˆ body weight
                                                                                  Allometric Approaches to Drug Discovery
b ˆ allometric exponent (slope of the line)
                                                                             Using limited data, allometric scaling may be used
One of the first applications of allometric scaling                          as a part of drug discovery. To do this we assume
was the use of the toxicity of anticancer agents in
animals to predict toxicity in humans. It was ob-
served that the toxic dose of a drug is similar among                                              0.40
species when the dose is compared on the basis of                                                  0.00
body surface area (Freireich et al 1966). For most                                                                                 Dog       Child
vertebrate species, the body weight/volume ratio
                                                                              Surface area (m2)

varies very little, but the surface area/volume ratio                                             −0.80                            Monkey
increases as species become smaller. Allometric cor-
rection of dose multiples in toxicology (compared
with proposed human doses) is thus important,                                                     −1.60
especially when small rodents provide the principal                                               −2.00
toxicology coverage.                                                                                          Mouse
  Body surface area (Y ) is related to body weight                                                −2.40
                                                                                                      −2.00 −1.50 −1.00 −0.50 0.00 0.50 1.00 1.50 2.00
(W, in kg) by the formula:
                                                                                                                      Log10 Body weight (kg)

                                                                             Figure 10.3 Allometric relationship between body surface area
                       Y ˆ 0:1W 0:67                                         and species body weight on a log vs. log plot

      Table 10.4 Equivalent surface area dosage conversion factors

      Species      Body weight (kg)          Body surface area (kg/m2 )                      Factor (km )    Approximate human dose equivalent

      Mouse         0.02                     0.0067                                           3.0            1/12
      Rat           0.100                    0.0192                                           5.2            1/7
      Dog           8.0                      0.400                                           20              1/2
      Monkey        2.5                      0.217                                           11.5            1/3
      Human        60                        1.62                                            37              N/A

      Dose in species 1 (mg/kg) ˆ dose in species 2 (mg/kg)  (km2 =km1 ).
                                     ANIMAL±HUMAN EXTRAPOLATION                                             103

that, for the formula Y ˆ aW b , the value of the        neuroprotective [these rats were subjected to 2 h of
power function `b' (or slope of the line from a log      focal ischemia by occlusion of the right middle
vs. log plot) is drug-independent, unlike the inter-     cerebral artery (MCA), followed by 22 h of reper-
cept `a', which is drug-dependent. By doing this we      fusion]. With the assumption of 100% systemic
can use data from a single species (the rat) to          absorption, the expected plasma Cmax at this dose
successfully predict the pharmacokinetics of Com-        was 2000 ng/ml. In this model, there was a signifi-
pound X in humans and cats. This method could be         cant reduction (greater than 30%) in cortical infarct
expected to save time and money in the drug dis-         volume, compared with saline controls, when the
covery process by enabling us to:                        drug was given at the time of occlusion and at 0,
                                                         0.5, 1, and 1.5 h post-MCA occlusion.
1. Select the correct dose in an animal model of            Using the data from the neuroprotection models
   disease. These studies are expensive and time-        from rats, we then scaled a dose to the cat that was
   consuming. The selection of the wrong dose in         expected to achieve a neuroprotective plasma con-
   an animal model, especially in a model in a           centration of 1500 ng/ml. To do this, we predicted
   larger species such as cat, could lead to invalid     the volume of distribution (V1cat ) using data col-
   results, either through toxicity (if the dose is      lected from the volume of distribution in rat (V1rat ).
   too high) or inactivity (if the dose is too low).     For our calculations we used a value of 0.938 for
2. Provide confidence that the pharmacological           the power function b (see Ings 1990, Table 2). In
   model will predict efficacy in humans. If a           doing this we made the standard assumption that
   drug is effective in therapeutic models using         in the formula Y ˆ aW b the value of the power
   different species and these animals receive           function b was compound-independent and that
   equivalent exposures (as measured by the max-         the function a was compound-dependent (Ings ob-
   imum plasma concentration, Cmax , or area             served that the power function b is reasonably
   under the plasma concentration curve, AUC),           constant for each pharmacokinetic parameter).
   then the clinician can choose a dose for trials       Substituting into the allometric formula,
   with confidence.                                      log (V1cat ) ˆ b log W ‡ log a, we found:
3. Eliminate unnecessary doses and plasma
   samples in the first trials in humans.                     log 0:426 liters ˆ 0:938 log 0:3 kg ‡ log a

   The discovery process for Compound X, which           Thus:
is efficacious in a number of in vivo models, is again
an illustration of how allometric considerations                             log a ˆ 0:120:
can enhance the development process. The whole
brain concentrations of this compound are in equi-       By substituting back into the formula and using a
librium with plasma concentrations within 5 min          cat weight of 4 kg, we found:
after dosing, and it is also eliminated from the
brain in equilibrium with the declining plasma con-                   V1cat ˆ 4:8 l or 1:21 l=kg:
centration. We also know that Compound X is
$80% orally bioavailable in rats and dogs (see           Our formula for calculating the dose to be adminis-
above), and has linear (first-order elimination)         tered was:
and predictable pharmacokinetics in animals.
   Next, this compound was tested in a model of                     Dosecat ˆ Doserat (V1cat =V1rat )
excitotoxicity, in which the neurotoxin malonate
was injected into the striatum of rats. A subcutane-     The formula for predicting the plasma half-life was:
ous injection of compound X at 9 mg/kg caused an
80% reduction in the lesion activity produced by                   T1=2 cat ˆ T1=2 rat …W cat =W rat †yÀx
malonate. The Cmax plasma levels of Compound X
at this dose would be about 1500 ng/ml.                  in which y is as defined earlier and x is a clearance
   In a study using spontaneously hypertensive           parameter (Boxenbaum and Ronfeld 1983). The
rats, a dose of 12 mg/kg of compound X was also          measured plasma half-life in the rat was 4.53 h.

Filling in the formula (Boxenbaum and Ronfeld          Table 10.5 Predicted and actual pharmacokinetic parameters
1983), we predicted a plasma half-life in the cat      for humans
of 7.3 h (ˆ 4:53  (4=0:3)0:938À0:75 ). The measured
                                                       Pharmacokinetic parameter                                      Predicted              Actual
plasma half-life in the cat was 6 h. We knew from
data collected in the rat that a dose of 3.06 mg/kg    Clearance                                                       0.138 l/h/kg           0.123
administered over 15 min would give a plasma Cmax      Half-lifea                                                     14.5 h                 13.6 h
of 1500 ng/ml of plasma. This equated to a dose in     V1                                                              1.01 l/kg              1.02 l/kg
the cat of 2.6 mg/kg over 15 min or 175 mg=kg=min      Vdss                                                            2.4 l/kg               2.1 l/kg
for 15 min.                                            a
                                                         Plasma half-life is the average from three values by three different
   When we performed studies to determine the          methods: (a) T1=2 human ˆ (0:693  Vd )=Clp ; (b) T1=2 human ˆ T1=2 rat
Cmax in cats following a dose of 2.6 mg/kg adminis-    (Whuman =Wrat ) yÀx ; and (c) log T1=2 human ˆ log a ‡ b log W human .
tered over 15 min, our predicted values were very
close to the actual values, with a measured Cmax of                                   104
1240 Æ 100 ng=ml.                                                                                                                     Rat
   Data from the rat can also be used to predict the                                                                                  Human

                                                           Normalized concentration
pharmacokinetics of Compound X in humans. As                                          103
with the cat we made our predictions prospectively,
by assuming, as stated earlier, that for the formula
Y ˆ aW b , the value of the power function b (or                                      102
slope of the line from a log vs. log plot) was drug-
independent, and that the intercept function a was                                              3            6     9       12           15
drug-dependent. We assigned values of 0.75, 0.938,                                    101                     Human time (h)
and 0.25 for clearance, volume of distribution,                                                 1          2       3        4
and plasma half-life, respectively, using data                                                            Rat time (h)
taken from the literature and discussed above.                                              0       100         200        300         400         500
The intercept function a was then determined for                                                          Apolysichrons (time/W0.19)
each parameter by substituting the pharmacoki-
                                                       Figure 10.4 Complex Dedrick plot of rat and human data for
netic data from rats, i.e. clearance ˆ 0.54 l/h/kg,
                                                       Compound X
V1 ˆ 1:42 l=kg, Vdss ˆ 3:33 l=kg. We estimated the
pharmacokinetic parameters for humans by substi-
tuting the calculated intercept function back into     again showing very good scaling between rat and
the formula and solving for Y for a 70 kg human.       human (Figure 10.4).
The prediction of the plasma half-life in humans          This illustrates how allometric scaling is a useful
was determined by three separate methods. For our      part of the drug discovery process: we avoided
predictions, we also assumed that the protein bind-    studying irrelevant doses and saved time. Ideally,
ing was the same in rats and in humans and that the    allometric scaling should be done using pharmaco-
metabolism of Compound X was similar in both           kinetic data from at least four species, even though
species. Clearly, approaches like this could be a      accurate predictions can be made using data from a
routine part of drug discovery.                        single species. If possible, information about differ-
   The values estimated by allometric scaling were     ences in metabolism among species should be con-
compared with those observed in the single-dose        sidered when making predictions.
human volunteer study (Table 10.5). We predicted
that for Compound X in humans the plasma half-
life would be 14.5 h, the plasma clearance would be                                             PHARMACOKINETIC/
0.138 l/h/kg, and the V1 , Vdss , and Vdb would be                                          PHARMACODYNAMIC MODELS
1.01, 2.37, and 2.56 L/kg, respectively. The predic-
tions using rat data were within 15% of the actual                                                  Elementary Aspects
mean values in human volunteers. A complex
Dedrick plot of the rat and human data showed          The possibility that time since lose changes the
nearly superimposable concentration±time curves,       relationship between pharmacological effect size
                                     ANIMAL±HUMAN EXTRAPOLATION                                            105

and drug concentrations in plasma has been known         size for a standard degree of receptor occupation
for a long time (Levy 1964, 1966; Levy and Nelson        (Jenkinson et al 1995). When an agonist must
1965; Wagner 1968; Curry 1980). The pioneering           occupy 100% of available receptors to cause Emax ,
work was done by Levy and his colleagues in the          its efficacy may be said to be unity. If occupation of
1960s on single dose±plasma level±effect relation-       all receptors achieves a response that is less than
ships, and on the duration of action of drugs as a       Emax , then the agonist's efficacy is less than one,
function of dose. Brodie and colleagues had shown        and equal to the ratio of observed maximal effect/
even earlier how complicated the relationships are       maximal effect for an agonist with efficacy ˆ 1 (we
when drugs with multicompartment distribution            call these partial agonists or agonist±antagonists).
are studied in this context (e.g. Brodie 1967).          Some agonists need occupy only a subset of the
Lasagna and colleagues, using diuretics, found           available receptors, in order to achieve Emax , and
that depending on whether a cumulative effect            these have efficacy greater than unity. In the latter
(24 h urine production) or an `instant' effect (rate     case, the concentration±response curve lies to the
of urine flow at a particular time) were measured,       left of the concentration±receptor occupancy curve
different relationships of response were possible        (e.g. Minneman et al 1983). Drugs with efficacy ! 1
(Murphy et al 1961). Nagashima et al (1969) dem-         are also called full agonists.
onstrated the relative time courses of anticoagulant        Below, we present some model relationships be-
concentration and effect. Thus, the relationship         tween observed concentration and effect size, as
between effect size and concentration of drug in         examples from a considerable volume of literature.
plasma should not be expected to be constant or          The reader is referred to key texts for comprehen-
simple, and can vary with time.                          sive coverage of this topic (e.g. Smolen 1971;
   The objectives of modern analysis of drug action      Gibaldi and Perrier 1982, Dayneka et al 1993;
are to delineate the chemical or physical interactions   Levy 1993; Lesko and Williams 1994; Colburn
between drug and target cell and to characterize the     1995; Derendorf and Hochhaus 1995; Gabrielsson
full sequence and scope of actions of each drug (Ross    and Weiner 1997; Sharma and Jusko 1997).
1996). Preclinical models describing the relationship
between the concentration of drug in blood or
plasma, and drug receptor occupancy or functional             Pharmacokinetic±Pharmacodynamic
response, provide clinically useful tools regarding                  (PK/PD) Modeling
potency, efficacy, and the time course of effect.
   Potency is an expression of the activity of a          Single-compartment, Time-independent PK/PD
compound, in terms of either the concentration or                           Models
amount needed to produce a defined effect. Emax is
                                                         The simplest model is where: (a) the drug distributes
the maximal drug-induced effect. EC50 is the con-
                                                         into a single compartment, represented by plasma;
centration of an agonist that produces 50% of the
                                                         and (b) the effect is an instantaneous, direct func-
maximal possible response. An EC50 can be de-
                                                         tion of the concentration in that compartment. In
scribed for drug concentrations using in vitro
                                                         this situation, the relationship between drug con-
assays, or as a plasma concentration in vivo. IC50
                                                         centration (C) and a pharmacological effect (E ) can
is the concentration of an antagonist that reduces a
                                                         be simply described by the linear function:
specified response to 50% of its former value.
   A measure of the tendency of a ligand and its
                                                                              E ˆSÁC
receptor to bind to each other is expressed as Kd in
receptor occupancy studies. Kd is the equitibrune
                                                         where S is a slope parameter. If the measured effect
contant for the two processes of drug-receptor
                                                         has some baseline value (E0 ), when drug is absent
combination and dissociation. Kd may be found
                                                         (e.g. physiological, diastolic blood pressure, or
for both agonists and antagonists, although some-
                                                         resting tension on the tissue in an organ bath),
times the former poses more technical challenge,
                                                         then the model may be expressed as:
due to alterations to the conformation of the bind-
ing site. In contrast, efficacy is a relative measure,
amongst different agonists, describing response                            E ˆ E0 ‡ S Á C

The parameters of this model, S and E0 , may be                                      Imax C
                                                                         E ˆ E0 À
estimated by linear regression. This model does not                                 IC50 ‡ C
contain any information about efficacy and po-
tency, cannot identify the maximum effect, and           In the Emax model above, plasma concentration
thus cannot be used to find EC50 .                       and EC50 are raised to the power of n (Hill factor)
   When effect can be measured for a wide concen-        equal to 1. A more general form of the equation is
tration range, the relationship between effect and       the sigmoid curve:
concentration is often observed to be curvilinear. A
semi-logarithmic plot of effect vs. log concentra-                              Emax C n
tion commonly linearizes these data within the ap-                             EC n ‡ C n
proximate range 20±80% of maximal effect. This
log-transformation of the concentration axis facili-     where, by addition of a single parameter (n) to the
tates a graphical estimation of the slope of the         Emax model, it is possible to account for curves
apparently linear segment of the curve:                  which are both shallower and steeper than when
                                                         n ˆ 1 (i.e. unlike the ordinary Emax models). Note
                E ˆ m Á ln(C ‡ C0 )                      that the sigmoidicity parameter (n) does not neces-
                                                         sarily have a direct biological interpretation and
where m and C0 are the slope and the hypothetical        should be viewed as an extension of the original
baseline concentration (usually zero, but not for        Emax model to account for curvature.
experiments of add-on therapy or when adminis-              The larger the value of the exponent, the more
tering molecules that are also present endogen-          curved (steeper, concave downwards) is the line. A
ously), respectively. In this equation, the              very high exponent can be viewed as indicating an
pharmacological effect may be expressed, when            all-or-none effect (e.g. the development of an action
the drug concentration is zero, as:                      potential in a nerve). Within a narrow concentra-
                                                         tion range the observed effect goes from all to noth-
                  E0 ˆ m Á ln(C0 )                       ing, or vice versa. An exponent less than unity (< 1)
                                                         sometimes indicates active metabolites and/or mul-
  As mentioned earlier, for functional data based        tiple receptor sites.
on biophase, plasma or tissue measurements, we              The corresponding inhibitory sigmoid Emax
often represent potency as EC50 , and when two           model is functionally described as follows:
compounds are compared with respect to potency,
the one with the lowest EC50 value has the highest                                   Imax C n
potency. A general expression for observed effect,                      E ˆ E0 À
                                                                                   IC n ‡ C n
by analogy with the Michaelis±Menten equation                                         50
(above) is:
                                                         In vivo, these models, analogous to the classical
                         Emax C                          dose or log dose±response curves of in vitro
                   Eˆ                                    pharmacology, are limited to direct effects in single
                        EC50 ‡ C
                                                         compartment systems. These models make no al-
There are various forms of this function for agonist     lowance for time-dependent events in drug re-
(stimulatory) and antagonist (inhibitory) effects.       sponse.
For example, if there is a baseline effect (E0 ), then
this may be added to the right-hand side of the           Complex PK/PD and Time-dependent Models
                                                         The most common approach to in vivo pharmaco-
                            Emax C                       kinetic and pharmacodynamic modeling involves
                E ˆ E0 ‡                                 sequential analysis of the concentration vs. time
                           EC50 ‡ C
                                                         and effect vs. time data, such that the kinetic
Alternatively, the relationship between concentra-       model provides an independent variable, such as
tion and effect for an antagonist, including a base-     concentration, driving the dynamics. Only in limited
line value, is:                                          situations could it be anticipated that the effect
                                           ANIMAL±HUMAN EXTRAPOLATION                                      107

influences the kinetics, for example, effects on         chosen to illustrate a single dose of drug causing
blood flow or drug clearance itself.                     the reversal of a symptom (pain). Many other types
   Levy (1964), Jusko (1971), and Smolen (1971;          of examples exist.
1976) described the analysis of dose±response              The plasma kinetics of the analgesic were describ-
time data. They developed a theoretical basis for        able by the following expression after the intraven-
the performance of this analysis from data               ous bolus dose, with C0 ˆ 45:0 and K ˆ 0:50 hÀ1 :
obtained from the observation of the time course
of pharmacological response, after a single dose of                         C ˆ 45:0 eÀ0:50t
drug, by any route of administration. Smolen
(1976) extended the analysis to application of
dose±response time data for bioequivalence testing.      In the same study, effect measurements were
   In dose±response time models the underlying as-       recorded during 80 min, as shown in Figure 10.5.
sumption is that pharmacodynamic data gives us              Often, drug effects do not parallel changes in
information on the kinetics of drug in the biophase      plasma concentration. This can result from distri-
(i.e. the tissue or compartment precisely where the      bution phenomena, such as when the effect occurs
drug exhibits its effect). In other words, apparent      outside the plasma compartment (e.g. the sedative
half-life, bioavailability and potency can be            effect of a dose of benzodiazepine, which occurs in
obtained simultaneously from dose±response±time          the brain), or when the effect recorded reflects, for
data. Considering such a model, assuming (a) first-      example, a chain of biochemical events triggered by
order input/output processes and (b) extravascular       the presence of drug (e.g. the aborting of a migraine
dosing, the kinetic model then drives the inhibition     attack by a serotoninergic drug). In relation to the
function of the dynamic model. It is the dynamic         first of these possibilities, a model sometimes called
behaviour which is described by the response model.      a `link model' (also called the `effect-compartment'
A zero-order input and first-order output governs        or the `effect-distribution' model) allows estima-
the turnover of the response. This permits us to         tion of the in vivo pharmacodynamic effect from
consider situations where the plasma concentration       non-steady-state effect (E ) vs. time and concentra-
represents delivery of the drug to an effect compart-    tion (C ) vs. time data, within which potential exists
ment; the time course of drug concentration and of       for observed E and C to display temporal displace-
effect (both in the biophase), is different from that    ment with respect to each other (Segre 1968;
simply observed in plasma concentrations.                Wagner 1968; Dahlstrom et al 1978; Sheiner et al
   The amount of drug in a single hypothetical           1979). The rate of change of drug amount (Ae ) in
compartment after an intravenous (IV) dose is usu-       a hypothetical effect compartment can be ex-
ally modeled with mono-exponential decline, and          pressed as:
analogous to the `plasma disappearance' curve
(above):                                                                dAe
                                                                            ˆ kle A1 À ke0 Ae
                     XIV ˆ DIV eÀKt
                                                         where A is the amount of drug in the central com-
The amount of drug in a single hypothetical com-         partment of a pharmacokinetic model, linked to
partment after an extravascular dose is then mod-        the effect compartment, with first-order rate con-
eled with first-order input/output kinetics:             stants kle and ke0 . The corresponding expression
                                                         for the amount of drug in the effect compartment,
               Ka FDpo  ÀK(tÀtlag )                 Ã
                                                         for a one-compartment model with bolus input of
       Xpo ˆ            e            À eKa (tÀtlag )
               Ka À K                                    dose (D) is:

Concentration±time effect modelling is illustrated                            kle D Â ÀKt           Ã
                                                                     Ae ˆ            e    À eÀke0 t
by the example which follows. This example was                              ke0 À K










                       0             10            20          30      40        50            60           70          80
                                                                    Time (h)
Figure 10.5 Observed effect-time data for an analgesic

where K is the elimination rate constant. The con-                    At equilibrium, C will be equal to Ce =Kp by defin-
centration of drug in the effect compartment, Ce , is                 ition, and thus:
obtained by dividing Ae by the effect compartment
volume, Ve :                                                                              ke0 D     Â ÀKt           Ã
                                                                               Ce ˆ                  e    À eÀke0 t
                                                                                      V1 (ke0 À K )
                                  kle D     Â ÀKt          Ã
                       Ce ˆ                  e    À eke0 t
                              Ve (ke0 À K )                           This is how the link-model relates the kinetics in
                                                                      plasma to the kinetics of drug in the effect com-
At equilibrium, the rates of drug transfer between                    partment. When used together with the Emax model
the central and effect compartments are equal:                        for estimation of the maximal drug-induced effect,
                                                                      the concentration at half-maximal effect (apparent
                                  kle A ˆ ke0 Ae                      EC50 ), and the rate constant of the disappearance
                              kle Vc C ˆ ke0 Ve Ce                    of the effect (ke0 ):

If the partition coefficient, Kp , equals Ce =C at
equilibrium (steady-state), then we can rearrange                                              Emax C e
                                                                                        Eˆ       n ‡ Cn
the above equation:                                                                           EC 50     e

                                          kle V1                        Computer fitting of the equations to the effect
                                  Ve ˆ
                                          Kp ke0                      data and estimation of the rate constant for the
                                                                      disappearance of the effect, ke0 , EC50 , and Emax
Substituting for Ve in the above equation (i.e.                       follows, assuming the sigmoidicity factor (n) to be
kle ˆ ke0 ) yields:                                                   equal to unity.
                                                                        At steady-state, Ce is directly proportional to the
                             ke0 DKp  ÀKt           à                plasma concentration (C ), since Ce ˆ Kp C. Conse-
                   Ce ˆ                  e À eÀke0 t
                           V1 (ke0 À K )                              quently, the potency (EC50 ) obtained by regressing
                                         ANIMAL±HUMAN EXTRAPOLATION                                               109

the last two equations, represents the steady-state           kinetics of (s)-warfarin were described by the
plasma concentration producing 50% of Emax .                  following mono-exponential expression:
   Note that the effect equilibration rate constant
(ke0 ) may be viewed as a first-order distribution                             Cw(s) ˆ 1:05eÀ0:0228t
rate constant. It can also be thought of in terms
of the rate of presentation of a drug to a specific           and the equation for the turnover of clotting factor
tissue, determined by, for example, tissue perfusion          [P] was:
rate, apparent volume of the tissue, and eventual
                                                                                   P                   Q
diffusion into the tissue. The results of the data
fitting in this exercise with the analgesic are: Emax                     dP      T   P0          U
4.5; EC50 0.61 ngÁmlÀ1 ; and ke0 0:07 hÀ1 .                                  ˆ kd T
                                                                                  R         !n À PU
                                                                          dt          Cw(s)
   Effect compartment or link models are limited                                   1‡
by their applicability to situations in which the                                     IC50s
equilibrium between plasma and response is due                In this equation, kd is the apparent first-order deg-
to distributional phenomena. In reality, there is             radation rate constant (also called kout ). This con-
often a delay between occurrence of maximum                   stant can be obtained experimentally from the slope
drug concentration in the effect compartment and              of a ln (P) vs. time plot, after administration of a
maximum intensity of effect caused by slow devel-             synthesis-blocking dose of coumarin anticoagulant
opment of the effect, rather than by slow distribu-           (Nagashima et al 1969; Pitsui et al 1993). P0 is the
tion to the site of action. In this situation, indirect       baseline value of the prothrombin time, Cw(s) the
or `physiological substance' models are more ap-              concentration of (s)-warfarin, and IC50S the con-
propriate (Dayneka et al 1993; Levy 1994; Sharma              centration of warfarin at 50% of maximal blocking
and Jusko 1997). Warfarin is a good example,                  effect. It was also possible to estimate the half-life
where this drug inhibits the prothrombin complex              of the apparent first-order degradation.
activity (PCA) (inhibition of production of effect).             An alternative model, including a lag-time to
This is illustrated by the following example, which           allow for distributional effects embedded in the
relates changes in (s)-warfarin concentration to ob-          observed time delay of the onset of the effect after
served PCA. The dose was intravenous. The                     warfarin administration, was published by Pitsui
change in PCA is shown in Figure 10.6. The plasma             et al (1993). Setting the baseline value of clotting
                                                              factor activity in the absence of warfarin (P0 ) to a
                                                              fixed mean of three predose measurements, the pro-
                                                              gram can estimate that parameter.
                                                                 The model equations are as follows:
                                                                           dPCA     Kin
                                                                                ˆ           À kd  P
      100                                                                    dt   I(Cw(s) )

                                                              where I(Cw(s) ) is the inhibition function of warfarin

                                                              (see next equation). It is appropriate to substitute
                                                              Kin with kd  P0 . Inhibition of synthesis (rate in) has
                                                              an impact upon the peak (trough) level rather than
                                                              the time to the peak. This is similar to a constant-
       40                                                     rate of drug infusion into a one-compartment
                                                              system. The time to steady state is only governed
       20                                                     by the elimination-rate constant and not the rate of
            0   24     48       72      98     120     144    infusion. At steady state:
                            Time (days)

Figure 10.6 Observed PCA time course following the adminis-                dR     Kin
                                                                              ˆ           À kout P ˆ 0
tration of an intravenous bolus dose of warfarin                           dt   I(Cw(s) )

If the baseline condition for PCA with no inhib-          be commonly measured. The reader is again referred
ition of drug is:                                         to standard texts for more thorough treatment of
                                                          models of this kind (Sharma and Jusko 1997).
                     PCA ˆ P0

then the steady-state condition for the pharmaco-                         COMMENTARY
logical response (PCAss ) with drug present be-
comes:                                                    We have not sought here to describe Phase I studies
                                                          as such. This is a postgraduate textbook, and we
                     P0            1                      wish to convey how in vitro and in vivo data of
         PCAss ˆ         ˆ P0            !
                    I(C)           Cw(s) n                various kinds may be used to help extrapolate ob-
                                1‡                        served drug effects from simple experimental
                                                          systems to more complex situation. The ultimate
and where I(Cw(s) ) is a function of Cw(s) , n, and       need is to obtain useful predictions of response in
IC50s , then:                                             healthy human subjects (Phase I studies) from ob-
                                    !                     served drug effects in animals or in the test tube.
                              Cw(s) n                        What are the strengths and weaknesses of these
              I(Cw(s) ) ˆ 1 ‡
                              IC50s                       approaches? The use of intrinsic clearance in vitro
                                                          permits predictions between species for the particu-
   As stated before, the intensity of a pharmaco-         lar enzyme/route of metabolism concerned. If
logical response may not be due to a direct effect of     humans have qualitatively different routes of me-
the drug on the receptor. Rather, it may be the net       tabolism for any particular compound, then this
result of several processes only one of which is          will weaken the predictive value of the in vitro
influenced by the drug. The process that is influ-        observation. Similarly, allometric scaling works
enced by the drug must be identified and an at-           best for compounds with a high component of
tempt made to relate plasma drug concentration to         non-enzymatic elimination, such as our model
changes in that process. Warfarin provides a good         compound with approximately 90% excretion as
example of this, as the anticoagulant (hypothrom-         unchanged drug. This prediction weakens as vari-
binemic) effect is an inhibition of the synthesis of      ations in rates of enzymatic reactions become more
certain vitamin K-dependent clotting factors.             important. The pharmacokinetic±pharmacody-
   Initial parameter estimates were obtained from         namic modelling approaches use existing in vivo
the PCA vs. time data. The baseline value (120 s)         data to calculate constants which can be applied
was obtained from the intercept on the effect axis.       to other in vivo data, but does not, in its present
This value is the ratio Kin =kd . From the intercept      form, link in vitro and in vivo data.
and slope, Kin was calculated to be 3:5 s hÀ1 . The          Significantly, none of these approaches uses
plasma concentration at the time of the trough of the     drug-receptor binding data. Although Kd values
effect corresponded approximately with the EC50           are generated during initial screening of the scores
value. Thus, IC50 ˆ 0:35 mg Á lÀ1 , kd ˆ 0:3 hÀ1 ,        of compounds emerging from medicinal chemistry
n ˆ 3:5, P0 ˆ 130 s, and tlag ˆ 0 h. The computer         laboratories, it has been a traditional problem
fitting gave 0.262 Æ 9.46 for the IC50 , 0.033 Æ 17.9     that relative efficacy remains unknown (this does
for kd , 2.68 Æ 39.6 for n, and 121 Æ 58 for P0 (limits   not detract from their value in chemical, structure±
are CV%) with no lag time. Precision increased            activity analyses). Neither do any of these ap-
when a finite lag time was included in the fitting.       proaches use results of in vitro functional assays
   As stated earlier, these are two of many examples      which emerge from screening of the compounds in
that can be chosen to illustrate principles. These two    biochemistry laboratories. It should be added that
cases, however, are especially relevant to the rela-      there are exceptions, however: drug±receptor bind-
tionship between animal work, and Phase I studies         ing constants and EC50 values from in vivo studies
in which only the simplest effects, such as counter-      in animals were used by Danhof and Mandema
action of a painful stimulus, or raising /lowering of a   (1995) to model drugs effects at benzodiazepine
physiological parameter such as PCA, are likely to        receptors and effects on EEG (Figure 10.7). Row
                                                        ANIMAL±HUMAN EXTRAPOLATION                                         111


                       EC50u (ng/ml)                                                   O


                                              0.5   1             10                 100              1000
                                                                       Ki (ng/ml)

Figure 10.7 Correlation (r ˆ 0:993, p < 0:001) between benzodiazepine free drug concentrations EC50 u) producing 50% of the
maximal EEG effect (change in amplitudes in the b frequency band, as determined by aperiodic EEG analysis) and affinity to the
GABA±benzodiazepine receptor complex (Ki ). Binding to the benzodiazepine receptor was determined on basis of displacement of
[3 H] flumazenil in washed brain homogenate at 378C. (Reproduced with permission from Danhof and Mandema, 1995)

ley et al (1997) have taken a similar approach with                       be simultaneous or not simultaneous but do not
NMDA antagonists.                                                         require much interaction between the investigators
                                                                          involved. Unlike the flow chart of a computer pro-
                                                                          gram, after which the diagram is modeled, most of
                       Prospectus                                         the decisions are made in discussions among com-
                                                                          mittee members, and may not necessarily be based
In the future, models will exist which will link                          on hard and fast criteria. Also, unlike a computer
constants for in vitro binding to cloned human                            flow chart, the decision concerning a particular
receptors (Kd ), data from in vitro functional assays                     drug will usually be based in part on the results of
(IC50 ), and animal and human in vivo EC50 values.                        work with other compounds that have the same
A composite prediction matrix will be applied rap-                        indication.
idly and accurately to the process of synthesis of                           In the boxes representing tasks to complete in the
new compounds for Phase I testing.                                        Phase I study in humans, we have used the symbol
   In the shorter term, what can we now do to                             1 to represent work that can be expedited by good
expedite the drug selection process? Figure 10.8                          validated preclinical data. The symbol 2 represents
represents a flow chart illustrating one form of                          the tasks that can be expedited by on-line pharma-
metabolism/pharmacokinetics input into the drug                           cokinetic modeling. Among the pharmacokinetic
discovery process. Arrows (indicating the flow of                         questions that will be asked on-line in the Phase
work and communication) pointing to the right                             I trial are the following:
represent perceived progress, whereas arrows
pointing to the left represent `disappointments'                          1. As the doses are escalated, do the kinetics of
(and other feedback) leading to corrections and                              the drug appear to be linear or non-linear over
revisions. The numbered asterisks indicate continu-                          the dose range?
ations. The `flow of time' is from left to right, and                     2. With repeated dosing, is there any evidence of
from the top panel to the bottom panel. The rect-                            a change in kinetics, e.g. a higher elimination
angles indicate tasks that are to be completed, and                          rate that might be indicative of autoinduction?
rectangles in a column within a panel represent                           3. Does the drug accumulate in tissues more than
work done by different departments which may                                 predicted with repeated dosing?

                                         Binding                                                                        No
                                         IC50                                                                  PK              Yes
                                         Functional Assay     Criteria                          In vivo
      Medicinal         New                                                                                  adequate
                                         EC50                                  First              PK
      chemistry         chemical                                              in vivo
                        entity(NCE)                                              rat
                                           In vitro                            study                                                 Project
                                                                                              In vivo                    Yes         group      *1
                                           intrinsic CL           Criteria                                   Criteria
                                                                   met                       efficacy                                evaluation


                                                  General                                                        No
           Project                                pharmacology
      *1   group                                                             Project
                             Candidate                                                                    Meets               Yes        First
           evaluation                                                        group                                                               *2
                             drug                                                                         safety                         human
                                                              2                                           criteria                       dose

                                  Restart at
                                      ?                                                                                             No

                         1   2
                                Allometric scaling                                                doses              Dose tolerated
                        Choose first dose and                                                                        PK adequate
      *2 Phase I         administration route
                                                                     pharmacology                                                           Yes *3
         trial                                                        laboratory            1                        PK model
                        Choose plasma target concentration                                                           adequate
                        Choose dose escalation                                                 Measure
                         schedule                                                              plasma

                                                     No                 Dose

               Human PK model
      *3   Review/revise sample                                      Yes                            Yes              Phase
                                                     Phase I
            times                                                                  Decision to                         II
           Review/revise escala-                                                   proceed                            trial
            tion schedule


Figure 10.8 Flow diagram for involvement of pharmacokinetic and pharmacodynamic mode/computer-generated feedback into the
iterative process of drug discovery from medicinal chemistry to the decision to enter Phase II trials. This is not a comprehensive flow
diagram for all aspects of drug discoveryÐit is restricted to the components of the process discussed in this chapter. This flow diagram
emphasizes efficient involvement of in vitro and in vivo experimental science and computer modeling, in review of data obtained in
Phase I studies, in the decisions related to selection of the best compound for patient studies
                                       ANIMAL±HUMAN EXTRAPOLATION                                                      113

4. If preclinical work identified metabolite(s) to         the first human exposure to the drug. In later expos-
   measure in humans, are the pharmacokinetics             ures, the dose is escalated according to some pre-
   of metabolite(s) linear and as predicted?               arranged criteria until the drug concentrations in
5. Does the relationship between concentration             plasma associated with undesirable properties in
   and effect change with dose, time and duration          animals are reached, and/or until some other
   of treatment?                                           limiting response is threatened or observed in the
                                                           human volunteers. Doses may be single or short
   We expect that the task lists represented by some       multiple-dose series. Simple physiological and bio-
of the boxes will increase. For example, within the        chemical measurements are routinely made in order
box including `in vitro intrinsic clearance', there may    to monitor for safety. If possible, responses to the
be in vitro predictors of oral availability, and meas-     drug are also measured when relevant to the
ures of potentially toxic metabolites. The `in vivo        intended therapeutic use. A drug successfully passes
pharmacokinetics' in rats may include an increasing        to Phase II if, with appropriate plasma levels, re-
number of compartments whose concentrations are            sponses are predictable, reversible, related to the
measured by microdialysis, and may include meas-           known pharmacological mechanisms of the drug,
ures of a few selected metabolite concentrations.          and there is a viewpoint among the investigators
   This diagram is not a comprehensive guide to            concerned that the drug could safely be given in
drug discovery. However, it does show that the             initial studies to patients from its target population.
chemists discover new chemical entities with desir-        Hopefully, all or most of what is observed in Phase I
able properties. In vitro biochemistry is followed by      is in line with predictions based on the pharmacoki-
initial in vivo work in the rat, which is conducted        netic and pharmacodynamic properties of the drug
with pharmacokinetic support and in vitro drug             in animals.
metabolism in parallel. Compounds meeting pre-                 Once Phase I is complete, the human becomes
arranged criteria proceed through pharmacological          the first-choice test species, under all but the most
screening to general pharmacology and toxicology,          specialized of circumstances (e.g. effects on repro-
all with pharmacokinetic support, which involves           duction). In this context, Phase I serves as the
the development of pharmacokinetic and pharma-             interface between preclinical research and clinical
codynamic models. As a chemical series develops,           development, and the validity of the predictions
correlations such as that in Figure 10.6 are de-           from animals to man involved is of paramount
veloped. Eventually, a compound or compounds               importance.
is/are chosen for Phase I studies.                             We believe that with enhanced integrated study of
   In this scheme, Phase I is influenced by pharma-        animals and humans, and with data feedback based
cokinetic and pharmacodynamic modeling. This               on computer models, the process of drug discovery
modeling is used to refine the Phase I protocol, pro-      from synthesis to proof of safety in humans could be
viding advice on sampling times, doses, and warning        dramatically improved in its efficiency. This is
signs of difficulty if they occur, as well as permitting   beyond what has traditionally been expected from
comparison of, for example, EC50 data from humans          departments of drug metabolism and pharmacoki-
with EC50 data from animals, and in vitro/in vivo          netics (Welling and Tse 1995). The time saved could
comparisons. The objective is expeditious choice of        be used to permit a larger number of compounds
the best compound, with the ever-present limitations       with better prospects, from a single research pro-
on information available. Note that this scheme can        gram, to be compared in Phase I studies. Conse-
involve feedback from Phase I to renewed chemical          quently, the extremely costly testing programs in
synthesis, as well as choice of a second or third          patients which follow Phase I could be started
compound for human testing.                                sooner and conducted better.
   Currently, Phase I studies themselves tend to be
quite straightforward and focus on single com-
pounds. Typically, after adequate preclinical char-                               REFERENCES
acterization of a candidate drug and 14 day and/or 3
month multiple-dose toxicology studies in two              Ashforth EIL, Carlile DJ, Chenery R, Houston JB (1995) Pre-
mammalian species, a very low dose is chosen for             diction of in vivo disposition from in vitro systems: clearance
  of phenytoin and tolbutamide using rat hepatic microsomal          Jenkinson DH, Barnard EA, Hoyer D et al (1995) International
  and hepatocyte data. Journal of Pharmacology and Experi-             union of pharmacology committee on receptor nomenclature
  mental Therapeutics 274: 761±6. (1995).                              and drug classification. IX. Recommendations on terms
Benet LZ, Kroetz DL, Sheiner LB (1996) Pharmacokinetics: the           and symbols in quantitative pharmacology. Pharmacol Rev
  dynamics of drug absorption, distribution and elimination. In        47: 225.
  Hardman JG et al (eds), Goodman and Gilman's Pharmaco-             Jusko WJ (1971) Pharmacodynamics of chemotherapeutic
  logical Basis of Therapeutics, 9th edn. McGraw-Hill; New             effects: dose±time response relationships for phase-non-
  York; 3±28.                                                          specific agents. J Pharm Sci 60: 892.
Boxenbaum H (1982) Interspecies scaling, allometry, physio-          Lesko LJ, Williams RL (1994) Regulatory perspectives: the role
  logical time and the ground plan for pharmacokinetics.               of pharmacokinetics and pharmacodynamics. In Cutler NR,
  J Pharmacokin Biopharm 10: 201±27.                                   Sramek JJ, Narang PK (eds), Pharmacodynamics and Drug
Boxenbaum H, DiLea C (1995) First-time-in-human dose selec-            Development: Perspectives in Clinical Pharmacology, 1st edn.
  tion: allometric thoughts and perspectives. J Clin Pharmacol         Wiley: Chichester.
  35: 957.                                                           Levy G (1964) Relationship between elimination rate of drugs
Boxenbaum H, Ronfeld R (1983) Interspecies pharmacokinetic             and rate of decline of their pharmacologic effects. J Pharm Sci
  scaling and the Dedrick Plots. Am J Physiol 245: R768±74.            53: 342.
Brodie BB (1967) Physical and biochemical aspects of pharma-         Levy G (1993) The case for preclinical pharmacodynamics.
  cology. J Am Med Assoc 202: 600±609.                                 In Yacobi A, Shah VP, Skelley JP, Benet LZ (eds), Integration
Colburn WA (1995) Clinical markers and endpoints in bioequi-           of Pharmacokinetics, Pharmacodynamics, and Toxicokinetics
  valence assessment. Drug Inf J 29: 917.                              in Rational Drug Development. Plenum: New York.
Curry SH (1980) Drug Disposition and Pharmacokinetics, 3rd           Levy G (1994) Mechanism-based pharmacodynamic modeling.
  edn. Blackwell Scientific: Oxford.                                   Clin Pharmacol Ther 56: 356.
Curry SH, Chu P, Baumgartner TG, Stacpoole PW (1985)                 Levy G (1966) Kinetics of pharmacological effects. Clin Phar-
  Plasma concentrations and metabolic effects of intravenous           macol Ther 7: 362.
  sodium dichloroacetate. Clin Pharmacol Ther 37: 89±93.             Levy G, Nelson E (1965) Theoretical relationship between dose,
Dahlstrom B, Paalzow LK, Segre G et al (1978) Relation be-             elimination rate and duration of pharmacological effect of
  tween morphine pharmacokinetics and analgesia. J Pharma-             drugs. J Pharm Sci 54: 872.
  cokin Biopharm 6: 41.                                              Meyer UA (1994) The molecular basis of genetic polymorphisms
Danhof M, Mandema JW (1995) Modeling of relationships                  of drug metabolism. J Pharm Pharmacol 46(suppl 1): 409±15.
  between pharmacokinetics and pharmacodynamics. In                  Minneman KP, Fox AW, Abel PA (1983) Occupancy of a-
  Welling PG, Tse FLS (eds), Pharmacokinetics: RegulatoryÐ             adrenergic receptors and contraction of rat vas deferens. Mol
  IndustrialÐAcademic Perspectives, 2nd edn. Marcel Dekker:            Pharmacol 23: 359±68.
  New York; 139±94.                                                  Murphy J, Casey W, Lasagna L (1961) The effect of dosing
Dayneka NL, Garg V, Jusko W (1993) Comparison of four                  regimen on the diuretic efficacy of chlorothiazide in human
  basic models of indirect pharmacodynamic responses. J Phar-          subjects. J Pharmacol Exp Ther 134: 286.
  macokin Biopharm 21: 457.                                          Nagashima R, O'Reilly RA, Levy G (1969) Kinetics of pharma-
Derendorf H, Hochhaus G (eds) (1995) Pharmacokinetic/Phar-             cologic effects in man: the anticoagulant action of warfarin.
  macodynamic Correlation. CRC Press: Boca Raton, FL.                  Clin Pharmacol Ther 10: 22.
Fox AW, Sullivan BW, Buffini JD et al (1996) Reduction of            Obach RS (1996a) Prediction of human pharmacokinetics using
  serum lactate by sodium dichloroacetate, and human phar-             in vitro±in vivo correlations. In Schlegel J (ed.), Pharmacoki-
  macok9inetic-pharmacodynamic relationships. J Pharmacol              netic/Pharmacodynamic Analysis: Accelerating Drug Discovery
  Exp Ther 279: 686±93.                                                and Development. Biomedical Library Series. International
Freireich EJ, Gehan EA, Rall DP et al (1966) Quantitative com-         Business Communications: Southborough, MA.
  parison of toxicity of anticancer agents in mouse, rat, hamster,   Obach RS (1996b) The importance of nonspecific binding in
  dog, monkey and man. Cancer Chemother Rep 50: 219±40.                vitro matrices, its impact on kinetic studies of drug metabol-
Gabrielsson J, Weiner D (1997) Pharmacokinetic and Pharma-             ism reactions, and implications for in vitro±in vivo correl-
  codynamic Data Analysis: Concepts and Applications, 2nd edn.         ations. Drug Metab Disposit 24: 1047±9.
  Apotekarsocieteten: Stockholm.                                     Pitsui M, Parker E, Aarons L, Rowland M (1993) Population
Gibaldi M, Perrier D (1982) Pharmacokinetics, 2nd edn. Marcel          pharmacokinetics and pharmacodynamics of warfarin in
  Dekker: New York; 231±2.                                             healthy young adults. Eur J Pharm Sci 1: 151.
Hawkins RD, Kalant H (1972) The metabolism of ethanol and            Rane A, Wilkinson G, Shand D (1977) Prediction of hepatic
  its metabolic effects. Pharmacol Rev 24: 242±9.                      extraction from in vitro measurement of intrinsic clearance.
Houston JB (1994) Utility of in vitro drug metabolism data in          J Pharmacol Exp Ther 200: 420±24.
  predicting in vivo metabolic clearance. Biochem Pharmacol 47:      Ross EM (1996) Pharmacodynamics: mechanisms of drug
  1469±79.                                                             action and the relationship between drug concentration and
Ings RMJ (1990) Interspecies scaling and comparisons in drug           effect. In Goodman and Gilman's Pharmacological Basis of
  development and toxicokinetics. Xenobiotica 20: 1201±31.             Therapeutics, 10th edn. Pergamon: New York.
Iwatsubo T, Hirota N, Ooie T et al (1996) Prediction of in vivo      Rowley M, Kulagowski JJ, Walt AP et al (1997) Effect of plasma
  drug disposition from in vitro data based on physiological           protein binding on in vivo activity and brain penetration of
  pharmacokinetics. Biopharmaceut Drug Disposit 17: 273±310.           glycine/NMDA receptor antagonists. J Med Chem 40: 4053±68.
                                           ANIMAL±HUMAN EXTRAPOLATION                                                       115
Segre G (1968) Kinetics of interaction between drugs and bio-      general considerations and procedures. J Pharmacokin Bio-
  logical systems. Il Farmaco 23: 907.                             pharm 4: 337.
Sharma A, Jusko WJ (1997) Characterization of four basic         Smolen VF (1971) Quantitative determination of drug bioavail-
  models of indirect pharmacological responses. J Pharmacokin      ability and biokinetic behavior from pharmacological data for
  Biopharmaceut 24: 611±35.                                        ophthalmic and oral administration of a mydriatic drug.
Sheiner LB, Stanski DR, Vozeh S et al (1979) Simultaneous          J Pharm Sci 60: 354.
  modelling of pharmacokinetics and pharmacodynamics:            Wagner JG (1968) Kinetics of pharmacological response:
  application to d -tubocurarine. Clin Pharmacol Ther 25: 358.     I. Proposed relationships between response and drug concen-
Shimada T, Yamazaki H, Minura M et al (1994) Inter-individual      tration in the intact animal and man. J Theoret Biol
  variations in human liver cytochrome P450 enzymes involved       20: 173.
  in the oxidation of drugs, carcinogens, and toxic chemicals:   Welling PG, Tse FLS (eds) (1995) Pharmacokinetics: Regula-
  studies with liver microsomes of 30 Japanese and 30 Cauca-       tory±Industrial±Academic Perspectives, 2nd edn. Marcel Dek-
  sians. J Pharmacol Exp Ther 270: 414±23.                         ker: New York.
Smith DA, Jones BC, Walker DK (1996) Design of drugs in-         Wilkinson GR (1987) Clearance approaches in pharmacology.
  volving the concepts and theories of drug metabolism and         Pharmacol Rev 39: 1±47.
  pharmacokinetics. Med Res Rev 16: 243±66.                      Yee S (1997) In vitro permeability across Caco-2 cells (colonic)
Smolen VF (1976) Theoretical and computational basis for drug      can predict in vivo (small intestinal) absorption in manÐfact
  bioavailability determinations using pharmacological data I:     or myth. Pharm Res 14: 763±6.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

          Phase II and Phase III Clinical Studies
                                                 Anthony W. Fox
                                       EBD Group Inc., Carlsbad, CA, USA

THE PHASES OF DRUG DEVELOPMENT: AN                              asked of companies, and the earlier stages of drug
         OBSOLETE MODEL?                                        development when these questions are asked, have
                                                                driven change in clinical study design. Increasingly
In former times, it was assumed that development                sophisticated data are now developed at earlier and
of drugs proceeded in step-wise fashion from Phase              earlier stages of drug development.
I, through Phase II, to Phase III, prior to filing a               In the later stages of the development of success-
Product Licence Application (PLA) or new drug                   ful drugs, the interval between PLA or NDA filing
application (NDA). Phase I was conducted in                     and product launch is not wasted. The term `Phase
`normal volunteers' (although some medical stu-                 IIIb' has been invented for the conduct of Phase
dents might hardly characterize this term!). Phase              IV-type studies during the pre-approval period.
II trials were initial studies in selected patients, and        Furthermore, in some companies, the old `Phase
Phase III trials were seen as wide-scale studies in             IV', is now divided into Phases IV and V, without
broader patient populations. After approval, cer-               any generally agreed definitions except, perhaps,
tain studies to find new indications, address special           that the studies are run by different teams.
patient subpopulations for marketing purposes, or                  Quite apart from these general trends blurring the
otherwise broaden product labeling might or might               distinctions between Phases I, II, and III, there are
not be conducted. All post-approval studies were                (and always have been) sound medical or pharma-
termed Phase IV.                                                cological reasons for doing so. Good examples
   In modern practice, the distinctions between                 might be:
Phases I, II, III, and IV are very often blurred.
Three principal, and interlocking, pressures have               . It would be unreasonable to study the pharma-
caused this blurring: time, finance, and an evolving              cokinetics of relatively toxic agents, at poten-
regulatory environment.                                           tially therapeutic doses, in normal volunteers
   Of these three pressures, the most important is                due to the near-certainty of the adverse events.
time. Strategies such as the overlapping of develop-              Typically, this information can be gained in
ment `phases', as well as the use of early dose-                  patients with diseases potentially responsive to
ranging studies as pivotal, and choosing doses                    these agents. Thus, the first-in-man studies in
based on surrogate end-points, are technical re-                  this case are `Phase II', using the classic nomen-
sponses to this challenge. Financial pressures,                   clature. Cytotoxic and antiviral drugs are two
even for the largest pharmaceutical companies,                    important classes of agent where this is com-
are generally much greater than in the past. The                  monly the case.
technical response is to maximize resources,                    . There is little point in testing the tolerability of
avoiding any and all redundant clinical studies.                  drugs in normal volunteers, when only patients
   The regulatory pressures come both from the                    with the disease of interest are able to demon-
regulatory authorities and from within the pharma-                strate a relevant pharmacodynamic effect. The
ceutical companies themselves. Regulatory author-                 doses at which tolerability must be confirmed
ities have increased in their scientific sophistication           are unknown until the exposure of patients can
during the last 30 years. The questions that are now              indicate the doses that may be effective. The

  development of potent opioids, such as alfenta-                                    CONCEPTS OF BIAS AND STATISTICAL
  nil, sufentanil and remifentanil, as anaesthetic                                             NECESSITIES
  agents, are a good example.
. There are some diseases which have no animal                                   Bias is a general consideration in clinical trial
  model or relevant pharmacodynamic or surro-                                    design, regardless of the type of trial being con-
  gate end-points in normal volunteers. Such dis-                                ducted. It is considered here as an overarching
  eases may also alter the pharmacokinetics of                                   issue, to be applied to the systematic description
  the drug, thus invalidating anything that                                      of the types of study design, as considered below.
  might be learned from normal volunteers. A                                        The word `bias' has many definitions, but in this
  good example is the migraine syndrome. No                                      context it is best described as a distortion of, or
  animal species has migraine, and normal volun-                                 prejudice towards, observed effects that may or
  teers cannot report an antimigraine effect.                                    may not truly be due to the action of the test
  Nausea, vomiting, and gastric stasis are common                                drug(s). Many things can distort the true measure-
  during migraine attacks and may be expected                                    ment of drug action, and bias is the trialist's most
  to alter the pharmacokinetics and effectiveness                                unremitting enemy, which comes from many quar-
  of oral therapies.                                                             ters (Table 11.1). The clinical trialist must be suffi-
                                                                                 ciently humble to realize that he/she, him/herself,
There is nonetheless little hope that the Phase I, II                            may be a source of bias.
III aphorism will die. Nevertheless, it is quite                                    The pharmaceutical physician may not be
wrong to assume that these `classical' terms and                                 expected to be a specialist statistician, and statistics
definitions still apply to how drugs are developed                               are not the subject of this chapter. However, the
according to modern practice. The classical four-                                ability to talk to and understand statisticians is
phase strategy of drug development is far too                                    absolutely essential (sine qua non: involve a good
stereotyped, simplistic and pedestrian to have sur-                              statistician from the moment a clinical trial is
vived into the modern era of drug development.                                   contemplated). Furthermore, the pharmaceutical
None of today's successful companies actually use                                physician should be confident of a sound under-
such a strategy. We are simply shackled with an                                  standing of the concepts of type I and type II error,
outmoded terminology.                                                            and the probabilities a and b (e.g. Freiman et al.

Table 11.1 Some example sources of bias in clinical trials

   Poorly matched placebos
   Subtle or obvious non-randomization of patients
   Failure of double-blinding, e.g. when pharmacodynamic effects cannot be controlled
   Prompting of prejudiced subjective responses
   Non-uniform medical monitoring
   Protocol amendments with unequal effects on treatment groups
   Peculiarities of the study site itself (e.g. psychotropic drug effects in psychiatric institutions which fail to predict effects in
   Differing medical definitions across languages, dialects or countries (e.g. `mania')
   CRF with leading questions, either toward or away from adverse event reporting
   Informal, `break the blind' games played at study sites
   Selective rigour in collection and storage of biological samples
   Selectively incomplete data sets for each patient
   Inappropriate use of parametric or non-parametric statistical techniques
   Failure to adequately define end-points prospectively, and retrospective `data dredging'
   Acceptance of correlation as evidence of causation
   Averaging of proportionate responses from non-homogenous treatment groups, also known as Simpson's paradox; see Spilker
   Unsceptically accepting anecdotal reports
   Tendency to publish only positive results

CRF case report form; the term `controlled' is used in its technical sense (see section on Bias and Statistical Necessities, this chapter).
                                      PHASE II AND III CLINICAL STUDIES                                      119

1978). This is one of your best defences against          (Normal), Chi-squared, f, binomial, etc., probabil-
bias.                                                     ity density functions. An intrinsic property of bio-
                                                          logical variables is that when measured 100 times,
                                                          then, on the average and if Normally distributed,
 PROSPECTIVE DEFINITIONS: THE ONLY                        5% of those measurements will be more than Æ 2
WAY TO INTERPRET WHAT YOU MEASURE                         standard deviations from the mean (there are
                                                          corollaries for the other probability density func-
It does not require a training in advanced statistics     tions). This meets a typical, prospective, `p < 0:05,
to hold a commonsense and accurate approach to            and therefore it's significant' mantra. It is also true
creating clinical hypotheses, translate them into the     that if you measure 100 different variables, on two
precise quantities of a measured end-point, and           occasions only, before and after administration of
then interpret the results. Whilst the finer points       the test material, then, on the average, 5% of those
of statistics are presented in Chapter 19, it is com-     variables are going to be significantly different
monsense that the only way to interpret what you          after treatment (this masquerades sometimes in
measure is to define this whole process, before the       findings among `selected secondary end-points').
experiment starts.                                        A sound interpretation, of course, is based upon
   Thinking carefully about what might actually           only those end-points that were selected before
constitute an observed response before you meas-          the experiment began, and comparing these with
ure it removes at least one important source of bias.     those for which no such statistical differences were
That bias is the clinical trialist him/herself. There     found.
has been too little emphasis in recent years on the
fundamentals of end-points, their variability and
how they are measured. Furthermore, the relation-                HISTORICAL CLINICAL TRIALS
ship between what is measured and its clinical rele-
vance is always debatable: the tendency is to             Any general work must include these classic bits of
measure something that can be measured, rather            history. Perhaps unusually, clinical trials appear to
than something that needs validation as clinically        be a European scientific invention. There is no
relevant. Good examples include rheumatological           evidence that either the ancient world or the medi-
studies: counts of inflamed joints before and after       aeval Arabs carried out prospective studies (al-
therapy may be reported, but do not reveal whether        though there are some anachronisms in recent
the experimental treatment or the corresponding           fiction). It is generally accepted that the earliest
placebo caused some of the patients to recover the        clinical trial was held by James Lind.
ability to write or others the ability to walk (Cha-         Thomas (1997) has pointed out that sailing men-
put de Saintonge and Vere 1982).                          of-war frequently went many months without
   Most clinical trialists experience the urge, espe-     docking, e.g. Nelson spent 24 unbroken months
cially in early studies, to collect every piece of data   on HMS Victory while blockading French ports,
that they possibly can, before and after every drug       and it is said that Collingwood once went 22
exposure. This urge comes from natural scientific         months without even dropping anchor. Scurvy
curiosity, as well as a proper ethical concern,           was rampant in the Royal Navy, often literally
because the hazard associated with clinical trials is     decimating ships' crews. Sailors survived on the
never zero. It behooves us to maximize the amount         poor diets carried aboard for long months, with
of information gained in return for the risk that         water-weevils and biscuit-maggots constituting im-
the patient takes for us, and for medicine in gen-        portant dietary protein! Before Lind's time, the
eral.                                                     Dutch had already learned to treat scurvy by re-
   Consequently, large numbers of variables are           plenishing their ships at sea with fresh fruit and
typically measured before and after drug (or pla-         vegetables. This was also known by Cook: when
cebo) administration. These variables all exhibit         in command of HM Barque Endeavour, men were
biological variation. Many of these variations            flogged for not eating their vegetables.
have familiar, unimodal, symmetrical distribu-               Lind had been pressed into the Royal Navy, as a
tions, which are supposed to resemble Gaussian            surgeon's mate, in 1739, and with some experience

as an apprentice surgeon in Edinburgh. It is a nice        subsequently developed a large private practice,
irony that the first prospective clinical study was        but little fame amongst his peers, and was buried
actually conducted by a surgeon!                           at Gosport in 1794. The Royal Navy was even
   The clinical trial was held at a single site, HMS       slower to act on his findings, and did not introduce
Salisbury, a frigate in the English Channel, during        citrus juice in sailors' diets until the year after
the early summer of 1747 (Lind 1753; Frey 1969;            Lind's death, following much administrative resist-
Thomas 1997). The experimental controls included           ance but no scientific controversy (Bardolph and
that all 12 patients met the same inclusion criteria       Taylor 1997). The British, especially those in the
(putrid gums, spots on the skin, lassitude, and            Royal Navy, are still known as `limeys', which is
weakness of the knees). All patients received the          the unique example of a national nickname based
same diet, except for the test materials. All treat-       on a therapy proven by clinical trial.
ments were administered simultaneously (parallel              Thus, Lind illustrates some other aspects of clin-
group). Compliance with therapy was confirmed              ical trials: first, he had little academic kudos, al-
by direct observation in all cases. The trial had six      though he was clearly qualified by experience and
groups, with n ˆ 2 patients per group. The test            training (a requirement of trialists by law in the
medications were (daily doses): (a) cider, 1 quart;        USA). Second, he did not publish his results rap-
(b) elixir of vitriol, 25 drops; (c) vinegar, two          idly. Third, his results were not implemented
spoonfuls plus vinegar added to the diet and used          promptly in the interests of the public health. It is
as a gargle; (d) seawater, `a course'; (e) citrus fruit,   important to realize that these undesirable aspects
two oranges, plus one lemon when it could be               of clinical trials persist to this day.
spared; and (f ) nutmeg, a `bigness'. Lind noted,
with some disdain, that this last treatment was
tested only because it was recommended by a sur-                 LIMITATIONS OF CONTROLLED
geon on land. The famous result was that within 6                      CLINICAL TRIALS
days only 2 of the 12 patients had improved, both
in the citrus fruit group, one of whom became fit          Progress in therapeutics has not always arisen from
for duty, and the other at least fit enough to nurse       controlled clinical trials. Chance observations have
the remaining 10 patients.                                 historically led to huge advances. Today's three
   We should note the absence of dose-standardiza-         most commonly used cardiovascular drugs are
tion and probably of randomization, because                good examples: digoxin is a component of digitalis
Lind's two seawater patients were noted to have            (famously reported by Withering after observing
`tendons in the ham rigid', unlike the others. How-        the treatment of a dropsical lady by a gypsy); as-
ever, the result had been crudely replicated by using      pirin is derived from the willow tree bark, first
n ˆ 2 in each group. If we accept that the hypoth-         reported by the Reverend Edmund Brown to treat
esis was that the citrus-treated patients alone            his own malarious fevers; and warfarin is the result
would improve (Lind was certainly skeptical of             of a University of Wisconsin investigation into a
the anecdotal support for the other five alternative       hemorrhagic disease of cattle. Lest we forget, Jen-
treatments), then, using a binomial probability dis-       ner's experiments would be ethically impossible
tribution, the result has p ˆ 0:0075. But statistics       today: they included deliberate exposure to small-
had hardly been invented, and Lind had no need of          pox, and aspirin is a drug that would probably fail
them to interpret the clinical significance of this        in a modern preclinical toxicology program due to
brilliant clinical trial.                                  chromosomal breaks and gastrointestinal adverse
   Lind was not quick to publish his most famous           effects due to systemic exposures in rodents.
treatise reporting this clinical trial (Lind 1753).        Modern clinical trials are therefore not necessarily
Indeed, in 1748, his Edinburgh MD thesis was on            the Holy Grail of therapeutic progress.
an entirely unrelated subject. Subsequently, Lind             Statistical theory must also be held not only with
was Treasurer of the Royal College of Surgeons of          respect but also with healthy skepticism (although
Edinburgh, and then appointed physician to the             this is really the subject of Chapter 19). It should be
Royal Naval Hospital, Haslar (one-fifth of his             remembered that the development of statistics, as
first 6000-odd admissions were for scurvy). He             they have come to be applied to clinical trials, has
                                     PHASE II AND III CLINICAL STUDIES                                      121

arisen from a variety of non-mammalian biological        with the same formulation and manufacturing pro-
sources. Experimental agriculture stimulated the         cess that is proposed to be taken to market. While
early giants (Drs Fisher, Yates) to explore prob-        the nuances of pharmaceutical constructs are de-
ability density functions. While epidemiological         scribed in Chapter 5, it is important to understand
studies have confirmed much that is similar in           the sometimes grave consequences when this rule of
human populations, it is unknown whether these           thumb is not observed.
probability density functions apply uniformly to all        Most regulatory authorities will want reassur-
disease states. Any statistical test that we employ      ance that the pharmacokinetic properties of the
makes assumptions that are usually not stated.           marketed product closely resemble those in which
                                                         the pivotal studies are carried out. This is not un-
                                                         reasonable: if the pharmacokinetic (PK) properties
   THE CLINICAL DEVELOPMENT PLAN                         differ, then so may dose size and frequency. Occa-
                                                         sionally, a Phase III study will be `bridged' to the
It is impossible to consider clinical trial protocol     marketed formulation by the demonstration, for
design in isolation. All clinical protocols should be    example, that two different tablets have the same
written after a clinical development plan has been       PK profile. However, the risk is that different for-
agreed by the diverse membership of the clinical         mulations will not turn out to possess the same PK
development team. The clinical development plan          profile: either new pivotal studies will have to be
should itself follow the construction of an hypo-        conducted with the new formulation, or registration
thetical drug label (see Chapter 29). The goals of       will be delayed until the new formulation is adapted
such a plan might be as limited as to provide for the    so that it does match the Phase III test material. For
start of Phase II, or as complex as mapping an           inhaled drugs, this is especially difficult. Time and
entire route from first-in-man studies to product        money is often lost in both cases. It is a risky gamble
registration. The path from the present status to        to leave development of the final formulation until
the overall goal can then be understood. It may be       the end of a clinical development plan.
added that, within a large company, this is also a
good way for clinical and marketing departments
to communicate.                                                          Informed Consent
                                                         This is considered in detail in Chapter 7. The clin-
PROTOCOLS, CASE-REPORT FORMS, AND                        ical trialist should remember, however, that he/she
    INVESTIGATORS' BROCHURES                             ultimately carries the ethical responsibility for this
                                                         document, regardless of what corporate lawyers
Chapters 23±26 describe the regulatory governance        and others may wish to do with it. Typically, insti-
of clinical trials, and little needs to be added here.   tutional review boards in the USA are more likely
These clinical trial documents are central to these      to be tolerant of long forms than ethics committees
processes. Equally the regulatory requirements           in Europe.
(which still vary from country to country), and
the documents needed to support them, must be
taken into account when constructing the clinical                     Toxicological Coverage
development plan.
                                                         This is covered in more detail in Chapter 6. How-
                                                         ever, the clinical trialist is encouraged to consider
   OBJECTIVES AND PREREQUISITES OF                       this for every protocol. A useful method is to start
           PHASE II STUDIES                              with the general case: what is the relationship be-
                                                         tween duration and dose sizes of animal studies and
                Gallenical Forms                         the clinical protocol-specified dose size and dur-
                                                         ation? This exercise ought to be conducted using
A good rule of thumb is that pivotal clinical trials     methods that standardize for both body weight and
for registration purposes ought to be conducted          body surface area across species. Next, review

closely all the prior human exposure to the test          erably) a relevant pharmacodynamic effect has
drug (if any) to see whether any unexpected signals       been observed in normal volunteers (see Chapter
for investigation may be found. Last, consider            10), then the first task is to reassess all of these in a
from the known pharmacology of the drug whether           relevant disease state. This is slower and uses more
there are likely to be any particular tolerability        patients than the professor's uncontrolled observa-
issues for which special monitoring methods are           tions. But at the end of a small number of such
needed, and think laterally.                              small studies, there ought to be good information
   For example, what is likely to be the adverse          about the feasibility of a pivotal clinical trials pro-
effects of a potassium channel-blocking drug being        gram, and, if not, then the feasible course correc-
investigated for a central nervous system indica-         tions (e.g. alternative indications). Note that one
tion? The answer may lie in all the excitable tissues     such course correction may be ceasing to develop
that contain potassium channels. Is there is any          the drug, and switching to another member in the
preclinical evidence that the drug discriminates          series. Arguably, the appropriate `killing' of drugs
between potassium channels in different tissues?          is the most valuable thing that a Phase II program
Are there changes in the EEG or ECG that may be           can accomplish, before too much time and money
found in the non-human database or among prior            has been wasted.
human exposures to the test agent, that escaped              When choosing a clinical trial design (Table
being reported because `not thought to be clinically      11.2), economic factors include numbers of pa-
significant'?                                             tients, time that will elapse, drug supply and total
                                                          cost. While these economies are important and
                                                          relevant in all design choices, they should also be
 COMMON PHASE II/III STUDY DESIGNS                        factored against the end-points that may or will be
                                                          measured. The relevance of an end-point, and its
Many initial studies are conducted in an uncon-           sensitivity to detect a drug-related effect may be
trolled fashion. Eminent professors will treat a few      primarily dependent upon the duration of patient
of their patients with a test medication (perhaps         exposure, e.g. a short period of observation is un-
under an investigators' IND in the USA) and               likely to detect a difference in time to next seizure in
form opinions about the worth (or otherwise) of a         a study of an anti-epileptic drug with an add-on
new therapy. While this may be grist for the mill of      design in patients who are only moderately dis-
press releases and fund-raising for small companies,      abled by epilepsy. On the other hand, the identifi-
these uncontrolled observations often mistakenly          cation of a PK interaction between a new and an
become a cast-iron credo for the sponsoring com-          established therapy in the same population may
pany. An observed effect, any effect, is viewed as        only require very short observation periods.
better than none, and the relative lack of scientific        There are several common classes of study
controls permits large biases to arise.                   design. These classes apply to almost all phases of
   The first risk from this haphazard start to clinical   drug development. No list of trial designs can be
development is that potentially good options for a        exhaustive, because almost all clinical trials are
test compound may be needlessly rejected. The             different. What follows is an attempt to briefly
professor's patient population may not include a          review the classes of clinical trial design that will
disease state or disease subtype for which the new        encompass a large majority of studies, and to com-
drug is actually well-suited. Equally, efficacy and       ment on their economy and end-point possibilities.
tolerability may be dose-dependent, and this can
only be assessed when studied in a systematic fash-
ion. Lastly, most drugs are just one of a series of                     Parallel-group Studies
compounds which share closely related properties
in preclinical testing. It is impossible to know which    These are typically thought of as the most straight-
of these is the most promising, when only one has         forward design case. In fact, a bewildering array of
been tested.                                              variations exist within this class.
   Assuming that reasonable tolerability, reason-           In the simplest case of parallel-group study, a
able understanding of pharmacokinetics and (pref-         group of patients presenting sequentially are
                                            PHASE II AND III CLINICAL STUDIES                                                       123

Table 11.2 Basic trial designs and the factors that are suited and unsuited to each

Trial type                                 Factors suited                              Factors unsuited

Parallel-group, single treatment           Episodic disease                            Rare disease
                                           Imperfect placebo matching
                                           Blinding difficult (e.g. surgical
                                             procedures, psychtropic drugs)
Parallel-group, chronic treatment          Stable disease state                        Unethical to use active comparator or
Cross-over with washout                    Stable disease state                        Untreated washout not ethical
                                           Ethical to use placebo after active
Sequential                                 Rare disease                                Complicated tolerability profile
                                           Homogenous disease state                    Many concomitant disease factors
                                           Urgent need to save life
`n of 1'                                   Stable disease state                        Few or no feasible alternative therapies
`Large simple'                             Very common disease                         Tolerability issues not closely related to
                                           Easily measured end-points                    efficacy variable
                                           Well-understood drug
Open label                                 Tolerability issues only                    Spontaneous adverse event frequency high
Within-patient dose ranging                Stable disease state                        Drug tolerance
                                           Intolerable high initial dose
Combination therapy                        A priori reason to expect favourable drug   Unethical to use single therapy

randomized to one of two equally-sized treatment                      Patients are randomized in cohorts to either active
groups, until a prospectively determined total                        or placebo treatment; frequently there are fewer
number of patients has been recruited. All these                      placebo-treated patients in each cohort. The ob-
patients are followed for a predetermined period of                   jective is to accumulate tolerability experience as
time, or until some end-point is achieved. The data-                  dose size gradually increases. If the treatments in
base is quality assured and locked before the ran-                    the first cohort prove to be well-tolerated, then the
domization code is broken. The patients are then                      next cohort is randomized in the same way except
sorted according to their treatment, the end-point                    that the active-treated patients receive a larger dose
measurements are subjected to a statistical test, and                 size. Note that this judgment can be made without
an interpretation of the effect (or absence thereof )                 breaking the blind. A comparable number of pla-
of the drug is made. What could possibly go wrong?                    cebo-treated patients to any single active-treatment
   The answer is that little can go wrong when there                  group can be accumulated across several cohorts,
are ample patients, plenty of drug available, the                     each cohort having fewer placebo treated than
choice of dose size has been perfect, the end-points                  active-treated patients. This economizes on patient
are incontrovertible, the measurements are possible                   numbers in comparison to randomizing each
using a ratio or absolute scale, there is ample toxi-                 cohort in a 1:1 fashion, and may also economize
cological coverage for all the dose sizes employed,                   on both drug and patients if two doses are found to
and the trialist has an unlimited budget! This com-                   be similarly effective and well-tolerated, albeit not
bination of utopian conditions never exists.                          the highest dose that was projected.
                                                                         Sequential cohort designs do not usually econo-
                                                                      mize on time. Treatment codes can be broken at the
The Ascending Dose-ranging Cohort Design                              end of each cohort (and not introduce bias into
                                                                      observations of succeeding cohorts). Sometimes
This is one variant within the parallel-group class.                  this can lead to early closure of the study when
It is best suited when there is no cast-iron assurance                the desired pharmacodynamic effect is observed
of tolerability for all the dose sizes of interest.                   at a lower dose than the maximum projected by

the study. However, the deliberations of safety          response curves to the right (i.e. tend to overesti-
committees at the end of each cohort can often be        mate the ED50 ) in comparison to a parallel-group
time-consuming.                                          study in the same patients with the same end-

      Within-patient Dose Titration Designs
                                                                        Cross-over Studies
These may be conceptualized as the application of
an ascending dose cohort design within a single          Generally, cross-over studies are more complicated
patient. The advantages of such designs are when         than parallel-group designs. Patients are exposed
immediate high-dose therapy is contraindicated for       to more than one test medication, in sequential
tolerability reasons, and when there are likely to be    treatment periods, perhaps with periods of no ther-
large variations between patients in the tolerability    apy intervening between those of active therapy.
and efficacy of the test drug.                           Active therapies may be different drugs, or differ-
   Patients are reviewed during and after comple-        ent doses of the same drug, or, in complicated
tion of a course of therapy, which may include pro-      studies, both.
grammed changes in dose size. If the drug was well-         The most famous problem is eliminating carry-
tolerated, they may progress to a course of therapy      over effects (`washout'). Ideally, end-points should
at higher dose. A prospective limit on dosing and the    be measured and unambiguously attributable to
number of courses of treatment is made (e.g.             one of the test regimens. This requires no residual
according to toxicology coverage). Dosing may be         effects of the previous regimen(s) (see Laska et al
curtailed at any time when either there is unreason-     1983). If this involves intervening placebo-treat-
able intolerance of the drug, or when acceptable         ment periods in between test medications, then
efficacy and simultaneous tolerability has been ob-      clearly this approach is not possible when placebos
served. This is not unlike the approach to therapy       are ethically unjustifiable.
under ordinary clinical circumstances. For example,         Usually, patients are randomized to a particular
patients with epilepsy are often treated by dose al-     treatment order, and all patients are eventually
terations. Another advantage of this design is that at   exposed to the same variety of treatments. Large
the end of the study, the range of tolerated and         numbers of treatment periods, assigned using a
efficacious doses can be examined among all treated      Latin square, have been reported; however, the
patients in comparison to demographic factors, dis-      logistics and patient retention in such studies is
ease subtypes, etc.                                      usually difficult, and these ideal designs are likely
   The greatest difficulty with ascending-dose,          to be successful only when treatment periods are
within-patient designs is usually in treatment-          short; ideal designs are commonest for normal vol-
masking. Double-blind requirements have to take          unteer studies (e.g. Amin et al 1995).
into account a wide variety of dose sizes, and that         In later-phase studies, if there are still numerous
contemporaneous placebo formulations will be             treatments or dose sizes that need to be tested, then
needed. Some studies of this type are hybridized         partial cross-over designs can be used. These
with a cross-over strategy (see below). Dose-tailing     expose patients to a random subset of all the
at the end of the study may be viewed as the same        study treatments, again in a random order. Partial
procedure in reverse, although may be conducted          cross-over designs necessarily require the availabil-
open-label and more rapidly (guided by suitable          ity of large numbers of patients. However, there
PK information) than when therapy is being intro-        can be economies of the amounts of test drug
duced.                                                   needed, and in the time needed to conduct the
   Sources of bias in this study design arise from the   study in comparison to an equivalent, complete,
exposure of patients to lower doses first. Patients      cross-over design. Shorter durations of patient par-
obligatorily must tolerate, and fail to respond to,      ticipation are also usually associated with fewer
lower doses before being exposed to higher doses.        missing data and fewer patients lost for adminis-
Any degree of treatment familiarization, tachy-          trative reasons. Overall patient recruitment is more
phylaxis, or patient withdrawal rate biases dose±        efficient.
                                              PHASE II AND III CLINICAL STUDIES                                                     125

   Pharmaceutical physicians should be wary of                         they have the same objective, which is to detect
using randomized, cross-over designs when there                        a treatment effect at the earliest moment possible,
are likely to be appreciable numbers of patients                       using the fewest possible patients, while retaining
who are withdrawn before completing the study.                         statistical robustness. Both types are suited for
This can cause serious imbalance among treatment                       exploratory clinical research, and both types are
groups and seriously jeopardize the likelihood of                      suited for diseases which are rare.
achieving a statistically robust result. Cross-over
studies with three or more periods have a substan-
tial advantage over two-period designs, when the                               The Dixon `Up±Down' Technique
amount of missing data is likely to be large, and
statistical salvage is necessary (Ebbutt 1984).                        This was first described in the statistical literature
                                                                       in 1947. It is designed to estimate an ED50 in
                                                                       clinical trials or toxicological tests, when a quantal
              MINIMIZATION TRIALS                                      response is measured (see Figure 11.1). However, it
                                                                       should be remembered that continuous responses
Less common are trial designs that specifically and                    can be converted into quantal responses with
adaptively minimize the number of patients                             appropriate, prospective efficacy criteria, e.g.
needed, while preserving design integrity for appro-                   blood pressure is a continuous variable, but a
priate statistical analysis. Early `evolutionary'                      drug may be deemed effective or ineffective by
designs are now being succeeded by independent                         stating prospectively that a desired response
treatment allocation, in pursuit of this goal. All                     is quantal-positive after a 15 mmHg fall in diastolic
minimization designs involve arduous statistical                       blood pressure within 60 days of commencing
planning, and the pharmaceutical physician should                      therapy. Theoretically, this strategy can be imple-
seek expert help from the outset.                                      mented with groups of patients treated in
                                                                       the same way instead of individuals. Sometimes
                 Evolutionary Designs                                  this technique is termed an `adaptive' trial design,
                                                                       because dose size is adapted according to the
These were devised by Dixon and Armitage.                              response of the previous patient or group of pa-
Although the statistical analysis is rather different,                 tients.

Figure 11.1 Illustration of the Dixon `up±down' method. Patients are treated in order of randomization, and the size of each test dose
is determined by the response of the previous patient. In this example, patients who do not respond ( ) are followed by a patient
treated at the next higher dose size; vice versa, patients who do respond ( ) are followed by a patient treated at the next lower dose.
When the line has changed direction at least six times, the mean effective dose is about the EC80 (in this case, 5 u, broken line). There
are many, often more sophisticated, variants of this basic technique

    The Armitage Technique or `Sequential                                Contemporaneous Independent Treatment
                  Analysis'                                                           Allocation
This technique was originally employed in the                          Taves (1974) has described a study design that
testing of explosive ordnance. Patients or groups                      requires an independent coordinator who allocates
of patients are paired, and then treated with alter-                   each patient, as he/she is recruited, to one or other
native therapies. A control chart is developed that                    treatment group. The independent coordinator al-
records the result of each comparison with time,                       locates each patient so as to minimize the difference
and crossing a boundary on the chart, after an                         between the two treatment groups, according to
unpredictable number of paired comparisons,                            prospectively defined patient characteristics, e.g.
gives the trial result. For a trial of a new therapy                   age, sex, genotype, disease state or stage, or con-
that can both benefit and harm the patient, a typical                  comitant therapy. This allocation is therefore also
probability control chart forms a `double-triangle'                    based upon the accumulating characteristics of the
pattern, as shown in Figure 11.2.                                      treatment groups, as has developed during the
   The original methods have been extended in                          study to date. Patients are therefore not allocated
many ways. The design of control charts is always                      to a treatment group by the chance of a random-
prospective, and their shape depends upon the a                        ization schedule.
priori expectations of the development team. For                          Bias in minimization trials can be avoided when
example, when it is important to test only the tol-                    three conditions are met. First, those performing
erability of a compound, the chart can have an                         the clinical trial itself, i.e. administering test medi-
`open top': this is when it is important to the devel-                 cations and measuring end-points, should be
opment team to detect drug toxicity early, but not                     double-blind and unaware of which treatment the
efficacy. Similarly, depending upon the hypotheses                     patient has received. Second, the independent co-
under test, control charts can be rhomboidal, par-                     ordinator need only allocate patients to anonym-
allelogram-shaped, or many other shapes. White-                        ous groups A or B, and the study pharmacist need
head (1999) is the best entry to the literature on this                be the only person who knows which treatments
specialized topic.                                                     these codes represent. Third, the criteria for which

Figure 11.2 Illustration of the Armitage `sequential analysis' study design. Patients are paired, and one of each pair receives each
alternative treatment. If the patient receiving treatment A does better than the one receiving treatment B (A > B), then the line moves
upwards; vice versa, if the patient receiving B does better than the one receiving A (B > A), then the line moves downward. If the
treatments cannot be distinguished within a pair of patients, then the line moves horizontally. The critical boundaries (broken lines) are
computed from prospective measures of a and b (e.g. p ˆ 0:05 and 80% power, respectively). The technique derives from an
engineering control chart and, once again, can be adapted to more sophisticated forms, including limits on the study size for
indeterminate results
                                      PHASE II AND III CLINICAL STUDIES                                      127

the treatment groups should be balanced must be                THE `LARGE, SIMPLE STUDY' AND
prospectively identified and rigidly adhered to,                  STRATIFICATION DESIGNS
using a recorded, quantitative system of scoring
the factors.                                              These similar classes of study require large
   In its simplest form, this class of minimization       numbers of patients. The choice between them lies
designs usually results in treatment groups of            in being able to `hedge one's bets' with a partial
nearly equal size. By equitably assigning patients        indication approval, vs. `all or nothing' with huge
to three or more treatment groups, and yet having         logistical costs and potentially huge rewards.
identical treatments for two or more of these, un-
balanced sample sizes can be created. This is of use
when, for example, it may be desirable to expose                        Stratification Studies
fewer patients to placebo than to active therapy,
especially when conducting a trial of compounds           In pivotal studies, large numbers of patients are
whose properties are fairly well known or may be          studied so that their diverse clinical characteristics
predicted with some confidence.                           can imitate better the ordinary patient population
   Note that minimization trials can only alter           than in earlier, more selective, trials. When a var-
power calculations when assumptions of the size           iety of concomitant factors (e.g. other diagnoses,
of worthwhile differences in effect are also pro-         wider degree of disease severity, concomitant medi-
spectively defined. For example, from a clinical          cations, etc.) are suspected, and may interact with
point of view, a small-sized improvement in out-          drug tolerability or efficacy, then patients may be
come (perhaps a few per cent of patients more than        stratified into randomization groups according to
that observed for placebo treatment) may be viewed        the presence or absence of such factors. For
as very worthwhile in an extremely heterogenous           example, patients with Crohn's disease might be
patient population when subjected to multivariate         stratified according to whether or not they also
analysis (this is common in large, simple studies; see    have cutaneous manifestations, and each stratum
below). On the other hand, when designing a mini-         then randomized to active or placebo for a total of
mization study, the assumption is that the treat-         four treatment groups, although only two test
ment groups will be devoid of relevant differences        treatments. Separate statistical analyses for the
in baseline characteristics, and therefore clinical       strata can then be planned, and the study size
significance might only be assumed to follow from         adjusted accordingly. The efficacy of the new
a large-sized difference in patient response. The size    drug may be found to be restricted to one or more
of the difference that is assumed to be of interest, as   particular patient subset(s). Regulatory authorities
it increases, may compensate for the reduction in         will often approve indications with caveats based
variability amongst study group samples, and thus         on such subsets. For example, in the USA, one
have less than expected impact on the sample sizes        indication for aprotonin is `. . . to reduce periopera-
needed to conduct the clinical trial.                     tive blood loss . . . in selected cases of primary cor-
   Minimization designs are probably under-used           onary artery bypass graft surgery where the risk of
by the pharmaceutical industry. This approach is          bleeding is especially high, e.g. impaired hemo-
not well-designed for pivotal clinical trials, or for     stasis, presence of aspirin, or coagulopathy of
diseases with large numbers of prognostic factors,        other origin'. The risk of stratification studies is
where, in any case, large numbers of patients are         that conservative regulatory authorities will
needed for a tolerability database. If the controlled     want to see statistical significance in all patient
clinical trial is a gold standard, then it would be       subsets before allowing a short, broad indication
wrong to assert that the independent treatment al-        in labeling.
location design is the `platinum standard' ( pace
Treasure and MacRae 1998). The interested reader
is referred to a good published example (Kallis et al                The `Large, Simple Study'
1994), and to more detailed statistical treatments
(Pocock and Simon 1975; Freedman and White                This is a recently recognized alternative to strati-
1976).                                                    fication, pioneered by Peto. Large numbers of

unselected patients are subjected to a single ran-        TREATMENT WITHDRAWAL AND OTHER
domization. If enough patients are recruited, and if             SPECIALIZED DESIGNS
the randomization is truly unbiased, then the large
sample sizes will allow all the potentially interacting   There are rare cases where established treatments
variables (concomitant drugs, concomitant dis-            are without strong evidence-based support. Two
eases, demographic variables, etc.) to balance out        good examples exist for digoxin: the treatment of
between the treatment groups.                             mild heart failure, and the treatment of cardiac
   The `simple' part of this approach is that, in         asthenia, a diagnosis that is especially common in
fundamental terms, the case report form can               Europe, and for which relatively small doses are
be very short. There is no need to collect lots of        prescribed. When the effect of such treatments on
information about the patient's clinical condition        the natural progression of disease is unknown, it
because there is no use for these data. Trials of         can then be ethical to recruit patients into a study
cardiovascular drugs, on an almost epidemi-               with inclusion criteria that include that they are
ological scale, have been the most significant            already being treated with the drug of interest.
examples of this alternative approach. Literally          Almost any of the designs discussed above may
tens of thousands of patients have been recruited         then be used, where patients are randomized either
under these protocols with case report forms              to remain on the treatment of interest or to be
having fewer than 10 pages for each patient.              withdrawn from that treatment. All the usual
Dr Robert Temple (1997; Director of the Office            needs for precisely defined prospective end-points
of Drug Evaluation I, at FDA) has commented               and sound statistical advice before starting the
that it may even be possible to conduct large sim-        study apply.
ple studies in treatment investigative new drug ap-          Early-phase clinical trials in patients with cancer
plication (IND) situations, thus permitting the           often use a two-stage design that has been pro-
generation of efficacy data outside of orthodox           moted by Gehan and others (Gehan 1979; Ellen-
`Phase III' clinical trial programs. However, in          berg 1989). With progressive, fatal diseases, the
this case the end-point would have to be just             problem of preventing an untoward number of
as simple, e.g. survival or death of the patient,         patients from being treated with a useless therapy
during a documented period of observation;                increases. These two-stage designs usually include a
Kaplan±Meier analysis and other epidemiological           small number of open label-treated patients (usu-
approaches may also be applied to such data-              ally n 14) in the first stage. The proportion and
bases.                                                    degree of tumor responsiveness is then used to fix
   Whilst the conditions under which large simple         the number of patients in the second stage of the
trials can provide efficacy data are fairly well          design, which may use an active comparator or no
worked out, it is important to consider whether           therapy as the alternative treatment, depending
(or which) tolerability issues can be precisely           upon whether an active comparator therapy can
addressed in this way. If a tolerability factor           be identified. Such studies cannot produce funda-
(adverse event) relates to the efficacy variable of       mental evidence of efficacy but, in the hands of
interest (e.g. a fatal adverse event in a patient         experienced statisticians and development teams,
survival study), then a simple case report form           can predict whether wider trials are justified.
may provide relevant information. However, if
the adverse event type is rare or unanticipated
(e.g. the test drug causes unanticipated, significant             STOPPING CLINICAL TRIALS
anemia in 0.1% of patients, and the protocol
and case report form do not collect hemoglobin                               Safety Issues
values before and after treatment), then it is
very likely that the adverse event will be missed.        Stopping a clinical trial because of an emergent
Large simple studies can thus create undue confi-         safety problem, either by a medical monitor or by
dence in product tolerability (`thousands of pa-          a safety committee, is always a unique situation.
tients were exposed to the agent during clinical          Little useful, generalizable guidance can be pro-
trials').                                                 vided here. These are decisions that are always
                                       PHASE II AND III CLINICAL STUDIES                                      129

taken in consultation, and the safety of potential         should ask whether a minimization design would
future trial recruits must be the paramount concern        have achieved the same thing with even fewer pa-
(including the abrupt cessation of therapy). Trial         tients, and thus actually feel chastened.
suspension is usually the best immediate option,              It is not the purpose of this chapter to delve into
allowing time for collective thought, notification         the mechanics of statistics. However, a few com-
of regulatory authorities and wider consultations          ments about the relationships between values for a
as appropriate.                                            at the stage of an interim and complete statistical
                                                           analysis of a clinical trial may be in order. There are
                                                           several statistical points of view on this subject, and
                   Efficacy Issues                         regulatory authorities have a habit of believing
                                                           only the most conservative.
Pocock (1992) has succinctly summarized most of               At the time of writing, the O'Brien and Fleming
the situations that obtain when it is considered           rule is becoming an acceptable standard. As a rule
whether to stop a clinical trial. Efficacy, like safety,   of thumb, pharmaceutical physicians should expect
can cause ethical concerns to the pharmaceutical           statisticians to provide alternatives that obey a
physician when he/she suspects that patients will be       simple subtraction rule. For example, clinicians
exposed to alternative therapies that are sub-             might agree that the study should stop due to
optimal.                                                   great efficacy when p ˆ 0:01 at an interim analysis,
                                                           when sufficient patients (power of 0.8) to detect
                                                           such a difference have been recruited. In that
            Interim Efficacy Analyses                      case, if the study continues after the interim analy-
                                                           sis fails to achieve p < 0:01, then it will be required
These usually make a mess! They require either that        to achieve approximately p < 0:04 for the whole
the overall size of the trial has to be greater than if    patient population in the final statistical analysis
no interim analysis was performed, or that a               in order to demonstrate the efficacy of the test
smaller a must be accepted as indicating statistical       drug. Even so, Pocock and Geller (1986) have
significance at the end of the whole study.                shown that trials stopped by reason of efficacy at
   Pharmaceutical physicians will hear loud com-           an interim stage are likely to have exaggerated the
plaints about these drawbacks of interim analyses,         size of the difference between treatment groups.
especially from senior management with purely              Marketing departments should be aware of this
commercial backgrounds. Everyone will want to              error in their extrapolations to the commercial
know as soon as possible whether `the drug                 worth of the product.
is working', but lax scientific thinking is behind
these complaints. Common statements are: `We
don't want to stop the study at the halfway stage,                      Bayesian Trial Designs
we just want to see how it is going'. When asked
why, the answer is usually something like, `There          A typical Bayesian design might be where, for
would be no point in spending more money on the            example, there are several drugs with preclinical
study if there is no chance of achieving a statistic-      rationale for the treatment of cancer; since none
ally significant result'. This is a popular misratio-      are clinically proven, one of the test treatments is
nalization: the decision not to stop a study is a          placebo. Patients are then recruited sequentially
decision to allow it to continue. Any interim deci-        into the study, and the results (e.g. tumor size
sion introduces a bias on the dataset that is eventu-      reduction) are recorded. After a while, the propor-
ally analysed.                                             tions of patients responding to each treatment
   Spectacularly effective drugs may achieve a very        are compared using a sophisticated probabalistic
small a at the time of the interim analysis. Stopping      method which takes into account the uncertainties
the trial by reason of the unethical basis for treating    associated with small and unequal treatment group
the patients with anything else is a rare and pleas-       sizes. The randomization code is then adjusted to
ant event for the clinical trialist. However, in that      favor more patients being allocated to the treat-
spectacular success, the pharmaceutical physician          ments that have started out looking better than

the others, while very poor, placebo-equivalent,         eral pharmaceutical physician should be advised
treatments might be dropped altogether. Eventu-          that, when considering a new trial, he/she should
ally, the several test therapies are reduced to two      at least consider whether a Bayesian approach
and a definitive demonstration of superiority or         might help. If this option is rejected, then that is
non-superiority for that pair of treatments can be       fine, but the brief consideration, as a matter of
reported.                                                routine, might occasionally lead to a superior trial
   The difficulties with interim analyses do not arise   design.
when a Bayesian approach to the original design
has been taken (Berry 1985). The Bayesian meth-
odology essentially revises the proportionate                    SERIES OF PUBLISHED CASES
patient allocation among the test therapies
according to the latest and best information avail-      Some diseases are so rare that the prospects of
able (e.g. Berry 1995): essentially, after some min-     conducting a clinical trial are remote. It is unlikely
imum number of patients have entered the trial, an       that enough patients could ever be collected at any
interim analysis is done every time another patient      reasonably small number of study sites for any
completes the trial. The important distinction be-       useful randomization. These diseases may be
tween Bayesian and sequential designs (above) is         found in the literature as case reports. In these
that although the number of patients required to         cases, probably the best that can be accomplished
complete a sequential design study are undefined at      is to collect and retrospectively analyze as many
the beginning, the treatment allocations are none-       such cases as possible. If the drug of interest has
theless according to a fixed randomization               been used in a sufficient number of patients, then
schedule. Thus, the sequential designs are still,        retrospective risk ratios for benefit and harm can
essentially, a frequentist methodology, and not          be calculated. This may be the strongest evidence
Bayesian.                                                that can ever be collected about a particular drug
   Bayesian approaches currently find little under-      under these rare conditions, albeit never as strong
standing on the part of regulatory authorities, and      as a controlled clinical trial. One example is the
thus are, probably unduly, little utilized by gener-     effectiveness of dantrolene in malignant hyperther-
alist pharmaceutical physicians. However, Baye-          mia (Strazis and Fox, 1993).
sian methods are finding increased uses in
specialized areas, e.g. trials of cancer chemother-
apy, and studies in rare diseases. The potential            OBJECTIVES AND PREREQUISITES OF
benefits of Bayesian methods include the use of                 PIVOTAL CLINICAL TRIALS
fewer patients to demonstrate efficacy, as well as
potential seamlessness of Phase II and Phase III         Licensing requirements typically are greater than
development when the number of drugs or dose             reporting data from by multicenter Phase III stud-
sizes of interest has been reduced during the trial      ies. Special populations may require small-scale
from several to one or two; patients recruited after     studies to supplement a traditional two-study,
this transition may be regarded as patients in a         large-scale registration development scheme. Simi-
pivotal trial by an enlightened regulatory authority.    larly, if (in the USA) the proposed indication has
   The general pharmaceutical physician cannot be        an approved orphan drug designation, then small-
expected to be able to generate Bayesian statistical     scale Phase II-type studies may be all that is pos-
plans for him/herself. These require an experienced      sible due to disease rarity (see Chapter 16). Fur-
statistician and, it may be added, a statistician who    thermore, even for conventional indications, the
is not, him/herself, philosophically opposed to          resource implications of pivotal studies are usually
Bayesian rather than frequentist thinking. The de-       much greater than any earlier phase of develop-
cision to employ a Bayesian design for a clinical        ment, and efficient resource utilization becomes
trial will be viewed as courageous in most com-          exponentially more important than before. The
panies, and there will be many clinical trials for       incorporation of pharmacoeconomic and human-
which an orthodox, frequentist approach will be          istic outcomes alongside the primary registration
selected, for several good reasons. Overall, the gen-    end-points is becoming essential, and preparatory
                                     PHASE II AND III CLINICAL STUDIES                                      131

work is best done in conjunction with the smaller,          There are highly mathematical approaches to
earlier studies, and must also factor treatment com-     benefit±risk assessment. When a single (binary)
pliance.                                                 end-point of interest can be balanced against a
                                                         single adverse event of concern, then the number
                                                         of patients required, and the number of required
           BENEFIT±RISK ANALYSIS                         therapeutic events, can be defined, and the confi-
                                                         dence intervals can be calculated, in order to exam-
The accumulation of all the data from the clinical       ine what the true benefit±risk ratio might be, e.g.
trials of a new drug product, assuming a fairly          for the global utilization of streptokinase and t-PA
orthodox regulatory strategy for a typical dossier       for occluded coronary arteries (GUSTO; Willan et
or NDA, will form the largest fraction of the appli-     al 1997). The number needed to treat, number
cation. However, these data are also needed for          needed to harm (and corresponding reciprocals)
derivative documents within the application, one         can be used to compare drugs for this purpose.
of which is a benefit±risk analysis, which forms the     However, this is an highly unusual and artificial
last part of an integrated safety summary (Section 9     situation, and the sophisticated statistical answers
of the NDA), and is a central objective of the           that result are unlikely to have more that a partial
Expert Report in European applications. These            impact on the more non-numerical approach taken
benefit±risk assessments must be derived from the        by clinicians.
clinical study reports and summaries elsewhere in           Usually, however, the clinical trialist has to stick
the applications.                                        out his/her neck, based upon an highly personal,
   All clinicians constantly weigh benefit±risk          non-numerical assessment of benefit±risk. The
in their daily practice. Their assessment of this        highly mathematical approaches usually work
`ratio' in everyday practice, using approved drugs,      best in retrospect, and this is not the situation of
is usually not as numerical as it sounds. In practice,   the clinician who must decide whether to prescribe,
clinicians make prescribing decisions based upon:        or the clinical trialist who must decide whether to
(a) a subset of the published information that           commit patients to a particular study design, both
might be available about the drug (labeling, drug        being prospective decisions. Furthermore, both in
representatives, comments from colleagues, etc.);        clinical trials and general medical practice, it is a
(b) their current and prior experience with this         rare situation where the benefit to the patient arises
particular patient; and (c) prior experience with        from a single binary variable, and there are no
other patients. This prior experience, even if per-      drugs which possess a single type of adverse
sonal, may or may not be consciously recalled.           event, whose probability may be confidently, pro-
Furthermore, we all operate algorithms taught us         spectively estimated for any given patient. Even the
by others whom we respect, and thus we use others'       simplest case, a drug with substantial history and
experience with drugs and patients, quite apart          experience, cannot fit the contrived mathematical
from the often hard-learned lessons from our own         approach described above. Penicillin has three ad-
therapeutic adventures (pace `evidence-based medi-       verse events of primary interest (anaphylaxis, bac-
cine').                                                  terial drug resistance, and sodium load at high
   Clinical trialists also weigh benefit±risk, every     doses). The mechanism by which infection recedes,
time a protocol is written. Often, unlike for ap-        if it is to recede, is only partly due to the action of
proved drugs, there is much less information to          the drug, because the extreme variability intro-
go on. In early clinical development, extrapolations     duced by the concomitant condition of the patient.
are obligatory. However, unlike in general medical       Whether to prescribe penicillin is a common deci-
practice, these extrapolations are often not from        sion for doctors and dentists: the mathematical
clinical experience, but rather from pharmacoki-         analysis of the benefit±risk `ratio' is unlikely to
netic models or animal data, or at best from             affect most prescribing decisions.
patients who are clearly dissimilar from those pro-         The informed consent document is where we ask
posed in the new trial. This is obligatory: if the       patients to make their own benefit±risk assess-
answers to the clinical trial questions were known,      ments, albeit with some guidance (Marsh 1990).
then there would be little point in doing the trial.     Certainly, the mathematical approach cannot be

expected on the part of the patient, neither will it be              Berry DA (1995) Decision analysis and Bayesian methods in
useful in a balanced and fair communication with                       clinical trials. Cancer Treat Res 75: 125±54.
                                                                     Chaput de Saintonge DM, Vere DW (1982) Measurement in
the patient about the nature of the clinical trial.                    clinical trials. Br J Clin Pharmacol 13: 775±83.
   Benefit±risk, then, is a central part of the practice             Ebbutt AF (1984) Three-period crossover designs for two treat-
of pharmaceutical medicine and its regulation. It                      ments. Biometrics 40: 219±24.
can almost never be reduced to a numerical exer-                     Ellenberg SS (1989) Determining sample sizes for clinical trials.
                                                                       Oncology 3: 39±42.
cise. Benefit±risk assessment of clinical trials data is
                                                                     Freedman LS, White SJ (1976) On the use of Pocock and
an important part of all new drug applications.                        Simon's method for balancing treatment numbers over prog-
Good people will differ in their benefit±risk assess-                  nostic factors in the controlled clinical trial. Biometrics 32:
ment, even when using the same body of clinical                        691±4.
trials data.                                                         Freiman JA, Chalmers TC, Smith H, Kuebler RR (1978) The
                                                                       importance of beta, the type II error and sample size in the
                                                                       design and interpretation of the randomized control trial. N
                                                                       Eng J Med 299: 690±4.
                        SUMMARY                                      Frey WG (1969) British naval intelligence and scurvy. N Eng J
                                                                       Med 281: 1430±33.
This chapter has attempted to provide a philoso-                     Gehan EA (1979) Clinical trials in cancer research. Environ
                                                                       Health Perspect 32: 31±48.
phy of clinical trials. The place of clinical trials in              Kallis P, Tooze JA, Talbot S, Cowans D et al (1994) Pre-
the overall development plan and what the                              operative aspirin decreases platelet aggregation and increases
pharmaceutical physician must know about rather                        post-operative blood loss: a prospective, randomized,
than be able to actually implement him/herself, has                    placebo-controlled double-blind, clinical trial in 100 patients
been emphasized. Almost all clinical trials are                        with chronic stable angina. Eur J Cardiothor Surg 8: 404±9.
                                                                     Laska E, Meisner M, Kushner HB (1983) Optimal crossover
unique because of the infinite combinations of hy-                     designs in the presence of carryover effects. Biometrics 39:
pothesis to be addressed, pharmacological proper-                      1087±91.
ties of the drug under investigation, the types of                   Lind J (1753) A Treatise of the Scurvy. 460 Edinburgh.
patients that are likely to be available, and likely                 Marsh BT (1990) Informed Consent. J R Soc Med 83: 603±6.
                                                                     Pocock SJ (1992) When to stop a clinical trial. Br Med J 305:
users of the resulting data. The major categories of
trial designs have been surveyed in some detail; it is               Pocock SJ, Geller NL (1986) Interim analyses in randomized
hoped that, when challenged with testing any clin-                     clinical trials. Drug Information J 20: 263±9.
ical hypothesis, a good pharmaceutical physician                     Pocock SJ, Simon R (1975) Sequential treatment assignment
would consider all these broad categories, select                      with balancing for prognostic factors in the controlled clinical
                                                                       trial. Biometrics 31: 103±15.
that most relevant to the clinical situation, and                    Spilker B (1991) Guide to Clinical Trials. Raven: New York;
then refine the proposed trial design from that                        1156 pp.
point. Some of the subtle interactions between stat-                 Strazis KP, Fox AW (1993) Malignant hyperthermia: review of
istical, financial and psychological aspects of trial                  published cases. Anesth Analg 77: 297±304.
design have been hinted at. The pharmaceutical                       Taves DR (1974) Minimization: a new method of assigning
                                                                       patients to treatment and control groups. Clin Pharmacol
physician will only really grow in this discipline                     Ther 15: 443±53.
through experience and good mentorship.                              Temple R (1997) Public hearings on Myotrophin: Peripheral
                                                                       and Central Nervous System Drugs Advisory Committee,
                                                                       May 8, 1997.
                                                                     Thomas DP (1997) Sailors, scurvy and science. J R Soc Med 90:
                       REFERENCES                                      50± 4.
                                                                     Treasure T, MacRae KD (1998) Minimisation: the platinum
Amin HM, Sopchak AM, Esposito BF, Henson LG et al (1995)               standard for trials? Br Med J 317: 362±3.
  Naloxone-induced and spontaneous reversal of depressed             Whitehead J (1999) A unified theory for sequential clinical trials.
  ventilatory responses to hypoxia during and after continuous         Statist Med 18: 2271±86.
  infusion of remifentanil or alfentanil. J Pharmacol Exp Ther       Willan AR, O'Brien BJ, Cook DJ (1997) Benefit±risk ratios in
  274: 34±9.                                                           the assessment of the clinical evidence of a new therapy.
Bardolph EM, Taylor RH (1997) Sailors, scurvy, science and             Control Clin Trial 18: 121±30.
  authority. J R Soc Med 90: 238.
Berry DA (1985) Interim analyses in clinical trials: classical vs.
  Bayesian approaches. Stat Med 4: 521±6.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

 Phase IV Drug Development: Post-marketing
                                            Lisa R. Johnson-Pratt
                                    Merck & Co. Inc., North Wales, PA, USA

 OBJECTIVES OF THE PHASE IV CLINICAL                            Table 12.1 Typical goals and tactics of Phase IV clinical trials
                                                                Extension of tolerability   Broader range of patients than in
                                                                  information                 NDA/PLA
While the earlier phases of drug development (I,                                            Larger numbers of patients
II, III) are geared toward demonstrating safety                                               (to detect rarer adverse events)
and efficacy and obtaining regulatory approval,                 Competitive efficacy        Active comparator studies
Phase IV clinical trials are usually conducted                    claims
                                                                New indications             Supplemental efficacy studies
after regulatory approval. Often, however, these                Market expansion            All of the above
trials are started prior to approval (despite risks                                         `Formulary tests'
that the drug or one of its indications may not                 Ethnopharmacology           Ethnic minority or overseas studies
be approved) to try to gain a competitive advan-                Outcomes assessment         Pharmacoepidemiology
tage.                                                                                       Pharmacoeconomics
   While Phases I, II and III have regulatorily-de-
termined functions and goals Phase IV trials are
much more flexible and cover a variety of purposes
(e.g. support label changes or demonstrate cost±
effectiveness). Depending on the organization,                             TYPES OF PHASE IV TRIALS
Phase IV trials may be conducted by the clinical
research team that conducted the original NDA or                The scope and goals of Phase IV clinical trials are
by a separate research team. If the latter, communi-            broader than Phases I±III. In general, the trials
cation links between the two research teams must                are larger (often having up to several thousand
be established early in the clinical program, to                patients), the inclusion/exclusion conditions are
provide information necessary to the conduct of                 less restrictive, and the end-points may be less ob-
later trials and to advise the marketing department             jective (e.g. `quality of life').
and corporate affairs during the launch and post-                  As the market becomes more congested and
approval life of the drug.                                      products compete for position on formularies and
   The objectives of the Phase IV program (see                  for reimbursement, additional data of use to con-
Tables 12.1 and 12.2) include further safety evalu-             sumers and health organizations may become crit-
ation, comparative safety and efficacy vs. other                ical. Unless the product is novel, sponsors can no
products, discovery of new indications, and market              longer be satisfied only with data from Phase III
expansion. Phase IV trials are not restricted to                trials demonstrating efficacy superior to placebo.
randomized, double-blind, placebo-controlled                    Sponsors must now demonstrate superiority (or at
trials, but may include other study formats, such               least equivalence) to competing products, and
as open-label trials and case control and cohort                meaningful and sustained efficacy, together with
studies.                                                        acceptable long-term safety and tolerability.

             Table 12.2 Practical aspects and problems of Phase IV clinical trials

             Active comparators             Obtaining competitor's drug
                                            Blinding, reformulation and bioequivalence
                                            Giving new ideas to competitor's marketing department
                                            Placebo groups must be ethically justifiable
                                            Dose ranges must be appropriate for both drugs
             Equivalence trials             Useful when superiority is unlikely to be demonstrated
                                            Scientific problem of proving a negative
                                            Large sample sizes
                                            Standards of care not previously studied
             Megatrials                     Unusual statistical strategy
                                            Few inclusion/exclusion criteria
                                            Representativeness to marketing target only known at end of study
             Open-label studies             Patient-introduced bias
                                            Non-random designs possible
                                            Scientifically limited
             Safety surveillance studies    Few options for control groups
             New indication                 Similar to NDA efficacy studies
             Dosage form comparisons        Patient satisfaction end-points
             Special patient populations    See Chapter 14±16
             Drug interaction studies       Almost infinite set of alternatives

        Comparative Superiority Trials                            request supplies from the competing sponsor or
                                                                  purchase the product on the open market and
In comparative superiority trials, the test drug is               modify it (e.g. encapsulation; reformulation) or,
typically compared with a drug that is the standard               via new packaging technologies, mask the identity
of care or is the class market leader. The goal is to             of the product. These options increase the time and
demonstrate superiority in efficacy, cost-effective-              costs associated with the trial. Modifying the prod-
ness, patient satisfaction, quality of life, or another           uct requires the development of a pharmaceutical
parameter that may be helpful in increasing market                process and evidence of bioequivalence. This may
share. The aim is to publish the trial (preferably in a           be particularly difficult for a product with a unique
peer-reviewed journal) and use the data in a pro-                 dosage form (e.g. wafer or patch). There is also a
motional campaign. Although historically such                     risk that the competing sponsor will challenge the
studies were often of poor design (e.g. inadequate                study results, citing formulation differences that
sample size; inappropriate end-points) to obtain a                may affect product efficacy.
superiority claim for use in promotional pieces,                     Although obtaining clinical supplies from the
the FDA now requires two adequate and well-                       competing sponsor is scientifically the best option,
controlled clinical trials or one study large enough              requests for such supplies may not be handled
to detect clinically useful differences. These studies            quickly or may be denied altogether. A study syn-
must have well-defined end-points that are relevant               opsis is usually required by the competing sponsor,
for both products and include a range of doses                    so that the design and hypotheses can be evaluated
for each product to ensure an optimal response.                   for scientific rigor. The synopsis also alerts the
Double-blind and, if necessary, double-dummy                      competing sponsor to the ultimate aims of the
studies are generally favored over open-label trials.             study, giving their marketing group time to respond
A placebo group is often added to minimize an                     to potentially damaging data. As the marketplace
active control bias (Spilker 1991).                               becomes increasingly competitive, companies are
   One of the difficulties in conducting double-                  less willing to provide supplies for comparative
blind comparative trials is acquiring suitable sup-               studies, so that alternatives should always be ex-
plies of the comparison drug. The sponsor can                     plored.
                                         PHASE IV DRUG DEVELOPMENT                                            135

                Equivalence Trials                          tered. In the earlier phases of drug development,
                                                            open-label trials are rarely appropriate, except for
Sometimes, it is not possible to demonstrate super-         very early safety and efficacy (e.g. dose-ranging)
iority but may be sufficient to demonstrate efficacy        studies. Open-label studies are often useful in the
and safety equivalence, particularly if the                 later phases of drug development and when con-
marketing strategy is based on price.                       ducting `real-world' safety/efficacy trials, particu-
   Equivalence trials usually require very large            larly when treatment will be prolonged and when
sample sizes due to small efficacy differences be-          cost or utilization data are required by purchasers
tween the two treatments. Equivalence trials typic-         of health care.
ally do not include a placebo group; therefore, both           Open-label studies may be conducted using a
treatments should have previously demonstrated              longitudinal trial design, in which a subject's treat-
superior efficacy to placebo, which may be difficult        ment data may be compared to his own baseline
to achieve when the standard-of-care drug has not           data or those of a treatment group may be com-
been stringently evaluated for efficacy and is used to      pared to those of a control group. Longitudinal
treat a condition for which it is not approved              designs are often used in epidemiologic studies,
(Makuch 1989). In addition, equivalence trials re-          such as the famous Framingham study, but can
quire the trialist to choose a clinically relevant treat-   also be used in drug studies where the effects of
ment difference to determine the appropriate sample         treatment on various variables (e.g. pharmacoeco-
size, which is often difficult to do (Makuch 1986).         nomics, quality of life) can be assessed over a finite
                                                            time period. Since subjects are enrolled in a non-
                                                            randomized manner, some of the ethical issues in-
                     Megatrials                             herent in randomized trials, in which subjects have
                                                            no choice as to which treatment group they are
Megatrials are commonly used to establish small
                                                            allocated, can be avoided. Criticisms of this ap-
differences of new drugs over existing drugs, e.g.
                                                            proach include the following (Friedman 1985):
the global utilization of streptokinase and t-PA for
occluded coronary arteries (GUSTO) study. Thus,
                                                            . The cohort being followed may not be repre-
they can be important marketing tools and can
                                                              sentative of the larger population for whom the
justify the use of the new and usually more expen-
                                                              drug is intended.
sive agents instead of older therapies of proven
                                                            . It is difficult to obtain comparable treat-
value (Hampton 1996).
                                                              ment groups, since patients are typically
   Unlike conventional randomized clinical trials,
                                                              only matched on one of many possible vari-
megatrials do not usually attempt to control for
confounding variablesÐrather, their goal is to
                                                            . Subjects may be inadvertently assigned to treat-
distribute bias evenly between groups of subjects.
                                                              ment groups based on characteristics that can
The group, rather than the individual, is the subject
                                                              influence the outcome.
of analysis. Megatrials appear to have had their
biggest impact when used to compare different allo-
                                                            Further problems are that open-label studies may
cated protocols with respect to a selected outcome.
                                                            result in increased patient dropouts, from patients
These trials consist of large numbers of patients (up
                                                            being randomized to receive an agent they view as
to tens of thousands) and typically have very loose
                                                            less beneficial and that is not viewed as favorably
entry criteria, although the ideal megatrial study
                                                            by the medical community or by regulatory author-
population should be representative of the target
population and have rigorous diagnostic criteria
                                                               On the positive side, open-label studies are
for entry (Charlton 1996).
                                                            usually administratively easier to conduct, offer
                                                            investigators more comfortable decision-making
                 Open-label Trials                          if they know which treatment the subject has
                                                            received, and may be the only option when blin-
In an open-label trial, both the subject and the            ded study supplies are not available (Friedman
investigator know the treatment being adminis-              1985).

            SAFETY SURVEILLANCE                          supplemental NDA (sNDA). To assist independent
                                                         investigators with this process, many large pharma-
Once a product is released onto the market, the          ceutical companies have set up investigator grant
number of patients exposed to it increases signifi-      programs.
cantly. As a result, the sponsor has an opportunity
to identify adverse drug reactions that are rare or
idiosyncratic. The sponsor does this in two ways:                    NEW DOSAGE FORMS
by monitoring spontaneously reported adverse
events and by conducting large clinical trials (post-    In a clinical development program, initially one
marketing safety surveillance studies). These trials,    lead dosage form is typically taken through full
which may be mandated by the regulatory agency           development. The chosen form usually represents
in order to gain approval, are usually conducted         the most easily developed, stable and marketable
under the Phase IV development program in an             form that is suitable for adults. For some products,
unblinded manner and under normal-use condi-             the initial form is found to be suboptimal and
tions.                                                   alternative forms are introduced early in the clin-
                                                         ical development process. For other products, the
                                                         need for multiple forms is not recognized until
   SEARCHING FOR NEW INDICATIONS                         future competitors are found to have a more ac-
                                                         ceptable or a unique formulation that may lead to a
Once a drug is approved for an indication, many          greater market share.
years of its patent life have usually expired. Infor-       In other cases, a new dosage formulation is re-
mation gathered from Phase III clinical trials or        quired to satisfy a new market segment. Oral solu-
from use of the drug under normal conditions             tions and chewable tablets are beneficial for a
(postmarketing) may suggest alternative indica-          pediatric population. A slow-release preparation
tions that may extend the patent life. For an inven-     (such as percutaneous patches) may be preferred
tion to be patented in the USA, it must be a new         in elderly patients taking multiple drugs if it allows
and useful improvement and not obvious (see also         once-daily (or less frequent) dosing. Trials demon-
Chapter 35). Patentable categories in drug develop-      strating safety and efficacy of new formulations are
ment include pharmaceutically active compounds,          often conducted in Phase IV and typically act as the
processes for making compounds, and compos-              supporting information for an sNDA.
itions comprising an active compound and a
pharmaceutical carrier (Hammer 1990). New for-
mulations may satisfy these criteria and extend the          Developing Dosage Forms to Improve
patent life, e.g. minoxidil (originally used orally as         Patient Compliance/Satisfaction
an antihypertensive agent) as a topical preparation
for the treatment of hair loss. New indications also     Although there are numerous reasons for lack of
help to keep up the `noise level' for the product,       compliance (e.g. lack of perceived benefit of treat-
enabling the marketing team to capitalize on new         ment, few disease symptoms), the sponsor may be
information that could potentially expand the            able to improve compliance by developing formu-
market for existing indications.                         lations that are more acceptable to the patient (e.g.
   The independent investigator can play an im-          once-daily rather than q.i.d. regimen, oral rather
portant role in identifying new indications by iden-     than injectable form). Dosage forms can also influ-
tifying and studying a particular drug effect. When      ence overall patient satisfaction, which can be
published, such findings may lead to widespread          useful in achieving market expansion or patient
use of the product for a new indication and, ultim-      product switches.
ately, adoption by the medical establishment. This          If the sponsor has developed a new, more accept-
process can occur without the intervention of the        able dosage form and can demonstrate improved
product sponsor or any formal clinical develop-          patient satisfaction, patients may switch products
ment program. However, generally new indications         and demand that the new drug be placed on a
are evaluated in a formal trial and submitted as a       formulary. Dosage forms may be evaluated in
                                      PHASE IV DRUG DEVELOPMENT                                          137

a Phase IV comparative trial and may require               EVALUATING DRUG INTERACTIONS
the use of a double-blind, double-dummy tech-
nique.                                                  Rarely is a drug devoid of potential for drug inter-
                                                        actions. A sound early clinical development pro-
                                                        gram is used to assess such interaction potential,
 Role of New Dosage Forms in Maintaining                typically with drugs that are expected to be used
            Formulary Status                            concomitantly with the investigative drug, or with
                                                        those that may have a significant effect on the
Multiple dosage formulations of a drug give             pharmacokinetics of the drug (e.g. the effect of
a sponsor a significant advantage over other prod-      oral contraceptives on the pharmacokinetics of a
ucts in its class. Not only can the sponsor reach       drug that will be used by women of childbearing
groups of patients that may have not accepted a         potential).
particular formulation (e.g. bulky tablet for a pedi-      After the drug is marketed, and more patients
atric population), but in the formulary wars, a drug    use the product under a multitude of conditions
with multiple formulations is generally considered      with a higher degree of concomitant medication
a better choice because different subgroups of          use, the potential for drug interactions can be
patients can be treated with the same product.          more fully assessed. Potential interactions are usu-
Additionally, patients may be able to switch            ally reported via the spontaneous reporting system,
formulations, depending on their medical require-       and can result in further evaluation in large Phase
ments, e.g. as they move from an inpatient (i.v.)       IV post-marketing safety surveillance trials. If a
to an outpatient (oral) setting without changing        beneficial effect of concomitant medication use is
products. This obviates the need for the pharmacy       anecdotally reported, such concomitant usage can
to make multiple product purchases in a particu-        be formally explored in a Phase IV clinical trial.
lar drug class, simply to accommodate patient
                                                        CLINICAL±LEGAL INTERFACE: IMPROVING
                                                              AND CLARIFYING THE LABEL
      USE IN SPECIAL POPULATIONS                                 (TACTICAL ASPECTS)

The entry and exclusion criteria for Phase I±III        In the USA, the concern regarding product liability
trials are usually very strict, eliminating as many     has led arguably to an ultra-conservative approach
opportunities for confounding of data as possible.      in labeling, advertising and promotion, and com-
As a result, patients with multiple medical condi-      pensation for investigators and patients in clinical
tions are routinely excluded from these trials. In      trials. This conservatism is expressed not only in
addition, entire classes of patients are excluded       the development of the original label that is sub-
because of ethical or liability concerns and recruit-   mitted for approval by the regulatory agency, but
ment difficulties. These patients include women of      also in requests by the legal department to improve
childbearing potential, children and adolescents,       the label after marketing to minimize risk of liabil-
the elderly, and ethnic minorities (see also Chapters   ity. On occasion, clinical trials are conducted to
below). Since these patient subgroups will un-          evaluate specific risks. This may be in response to
doubtedly use the product, trials are conducted in      pharmacodynamic effects seen in Phase III or
Phase IV to evaluate the safety and efficacy in these   to serious adverse events reported spontaneously.
populations. The goal is to gather information that     Results from these trials will then be used to add to
will supplement information in the label and in-        or clarify information in the label, to encourage
struct physicians on any differences in pharmaco-       safer use of the drug. Phase IV trials that are used
kinetics/dynamics in the special population,            for this purpose include: large safety surveillance
compared to the Phase III population. This infor-       trials (e.g. to assess rare serious adverse events
mation may also be used to expand the potential         or pregnancy outcomes); additional dose-ranging
market by broadening the population for whom            studies (e.g. to assess a reduction in dose on
the drug is indicated.                                  the risk±benefit profile); double-blind placebo-

controlled trials, that assess effects on a target      published in a key journal, they represent an
organ (e.g. using angiography or cognitive function     extremely small percentage of physicians who will
tests); or pharmacokinetic trials (e.g. assessing the   actually be prescribing the product. Physicians
effect of alcohol or food on drug concentration).       chosen for participation in Phase IV studies are
   The impact of anticipated study results on label-    those who can potentially increase market share
ing should be evaluated carefully prior to initiating   by serving as regional and local advocates for
a trial. Although the ultimate goal of such labeling    the products, and are commonly referred by the
changes is to convey important, clinically useful       business group. These are the physicians who
information to the physician and patient (via the       may sit on formulary committees of hospitals, de-
Patient Package Insert) to allow better clinical        velop local treatment guidelines, see high volumes
decisions, haphazard label additions purely to de-      of patients, are active in local medical societies, and
crease legal liability may confuse rather than assist   serve as consultants to the sponsor. In the
the prescriber.                                         USA, with increased use by managed-care organ-
                                                        izations of a `gatekeeper' generalist physician, use
                                                        of family practitioners, general internists and
      CLINICAL/MARKETING INTERFACE:                     gynecologists as clinical trial physicians may
          EXPANDING THE MARKET                          prove to be extremely beneficial for ultimate prod-
                                                        uct acceptance. Generalists are often the physicians
  Conduction of Studies to Promote Local                who will most prescribe a product and therefore
       Operating Company Needs                          should be given the opportunity to test it in phase
                                                        IV trials.
To obtain reimbursement and formulary accept-
ance as health-care dollars become scarcer, the
sponsor at launch or soon after should be able to                          CONCLUSION
demonstrate cost±effectiveness, superiority, con-
venience, and patient satisfaction, usually through     Phase IV clinical trials are the next step in drug
data from Phase IV pharmacoeconomic trials (see         development after completion of pivotal safety
also Chapter 19).                                       and efficacy trials. No longer viewed as, Phase
                                                        IV clinical trials are becoming an increasingly inte-
                                                        gral part of the overall development program.
       Choosing Investigators: Advocate                 The questions they are designed to answer are
                Development                             those asked by the purchasers of pharmaceuticals,
                                                        with goals of demonstrating the overall benefit
When selecting investigators for the pivotal Phase      of using a product and distinguishing it from its
III clinical trials, the sponsor typically searches     competitors in terms of pharmacoeconomics,
for physicians who are well-known specialists in        safety, and efficacy. These trials can provide timely
a particular therapeutic area. These physicians         data to respond to a rapidly changing market
are typically referred to as `thought leaders' or       place and fill information gaps that might other-
`opinion leaders' and include editors and authors       wise preclude successful market penetration.
of major textbooks or other publications and
leaders in important medical societies. They may
also be representatives on advisory committees to
the FDA or other regulatory agency who will ul-                             REFERENCES
timately determine whether the drug is approved
                                                        Charlton BG (1996) Megatrials are based on a methodological
for marketing.
                                                          mistake. Br J Gen Pract 46: 429±31.
   One goal of Phase IV studies is to expand the        Friedman LM, Furberg CD, DeMets DL (1985) Fundamentals
numbers and types of physicians who are using the         of Clinical Trials. Mosby-Year Book: St. Louis, MO.
product. While a few `thought-leaders' may lend         Hammer CE (1990) Drug Development, CRC Press: Boca
credibility when attempting to get a manuscript           Raton, FL.
                                              PHASE IV DRUG DEVELOPMENT                                                      139
Hampton JR (1996) Alternatives to mega-trials in Cardiovascu-     Makuch, RW, Johnson Mary F (1989). Issues in planning and
 lar Disease. Cardiovasc Drugs Ther 10: 759±65.                     interpreting active control equivalence studies. J Clin Epide-
Makuch RW, Johnson MF (1986) Some issues in the design and          miol 42(6): 503±11.
 interpretation of `negative' clinical studies. Arch Intern Med   Spilker B (1991) Guide to Clinical Trials, Raven: New York,
 146: 986±9.                                                        NY.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

          The Unique Role of Over-the-counter
                                                   Paul Starkey
            Formerly Vice-President, Smithkline Beecham Healthcare, Parsippany, New Jersey.

         THE EXPANDING PLACE OF                                    in the handling of many common types of
             SELF-MEDICATION                                       illness.
                                                                5. There is a need to increase treatment effective-
In recent years, the role of over-the-counter (OTC)                ness, which is not ordinarily considered an
medication in the overall health system has in-                    advantage of self-medication. The increase in
creased dramatically. The increased interest in                    effectiveness depends on the generally more
and availability of OTC medications is being                       rapid availability of OTC medications than
driven by several factors:                                         prescription medications, so that treatment
                                                                   may begin sooner. This can significantly
1. There is a growing recognition of the capability                shorten the total length of suffering, especially
   of patients to treat themselves in a rational and               when the natural course of a disease is brief or
   safe manner. The older authoritarian model of                   when severe discomfort makes prompt therapy
   medicine is being gradually replaced by a more                  especially helpful.
   participative model.
2. There is an increasing desire by patients to                    An example of this last phenomenon is in the
   participate in their own medical care. This is               treatment of vaginal candidiasis. Prior to the
   not just a result of changes in philosophy but               OTC availability of topical antifungals, it was
   also of the dramatic increase in average educa-              often necessary for a woman who had already rec-
   tional level over the past half-century. The                 ognized the symptoms of the disease to call and
   world increasingly possesses a well-informed                 arrange a physician's appointment. This often
   and intellectually capable population that                   took several days. Delaying treatment caused
   demands an active and inclusive role in its                  much unnecessary suffering and encouraged dis-
   own health care.                                             ease progression. Many physicians, recognizing
3. The quantity of information now available to                 these difficulties, would prescribe over the phone,
   the average person, both through formal edu-                 based solely on the woman's description of symp-
   cation and through the media, has increased                  toms. Research has shown that the accuracy of the
   enormously, giving increasing awareness of                   physician's diagnosis in this setting is no better
   treatment options.                                           than that of the woman herself. This constituted
4. There is a growing need to contain medical                   an ideal situation for the switching of an important
   costs. OTC drugs are not only cheaper than                   class of drugs from prescription to OTC status. The
   prescription drugs, due to their simpler and                 patient obtained equally accurate diagnosis and far
   more efficient distribution channels, but they               more rapid treatment for a disease that is very
   also eliminate the need for an expensive visit               uncomfortable. Severe cases of vaginal candidiasis
   with the doctor for each episode of illness. The             with heavy discharge are now much less common.
   professional intervention required to prescribe                 A second example is in the treatment of the
   pharmaceuticals represents the dominant cost                 common cold. Anti-cold medications have been

available OTC for many years, because of the com-                          Self-diagnosis
pelling need for rapid treatment. A cold evolves
quickly, the entire illness lasting only a few days.     Surprisingly, the first of these considerations has
A delay of only a day or two in seeing the physician     nothing to do with the drug itself but rather with
for a prescription may eliminate any possibility of      the condition it is to treat. Self-treatment implies
obtaining effective treatment for at least half of the   self-diagnosis. Only diseases that can be made self-
duration of the illness. The prompt availability of      diagnosable through appropriate labeling can be
self-medication improves treatment efficacy, while       considered for OTC treatment.
reducing costs and enhancing patient satisfaction           Fortunately, there are many common conditions
with the medical system.                                 that are self-diagnosable. Indeed, it should not be
   The above factors have combined to greatly in-        assumed that a diagnosis made by the patient is
crease public awareness of the importance of self-       necessarily inferior to that made by the physician.
medication in the total health care scheme. The          The patient can actually feel the symptoms as well
pharmaceutical physician should recognize the op-        as observe the signs of a diseaseÐa real advantage
portunities for OTC use of medications and the           in the diagnosis of conditions where symptoms
advantages and pitfalls attendant upon such use.         predominate.
As self-medication becomes a central part of                Of course, disorders vary in the extent to which
the health care system, the skillful and appropriate     laboratory tests and other sophisticated techniques
movement of pharmaceuticals from prescription to         are important in their diagnosis. If these are a
OTC availability will increasingly become a vital        major factor, a satisfactory patient diagnosis is
role of the pharmaceutical physician in optimizing       unlikely, no matter how skillfully written labeling
the nation's health.                                     may be. At the other extreme are many conditions
                                                         whose diagnosis rests primarily on history and
                                                         symptoms, where a patient may be educated to
 CRITERIA FOR OTC USE OF MEDICINES                       give an equal or better diagnosis than that of the
The criteria by which a drug may be judged as               An example of this is headache, where the diag-
suitable for self-medication are never absolute. A       nosis rests largely on history and symptoms. The
drug's `OTC-ness' is always a matter of careful          patient has lived the history and experienced the
judgment. The Food and Drug Administration               symptoms. The physician has at best a description
(FDA) has lessened the requirements for OTC use          of these symptoms, which a patient may be able to
substantially in recent years. In time, it may dra-      communicate well or poorly. Even with the most
matically shift the standards for conditions and         skillful physician eliciting the history, there is a
drugs considered suitable for self-medication. The       degradation of information as it moves from pa-
old tendency to restrict OTC treatment to condi-         tient to physician. If patients can be educated about
tions of short duration and primarily to symptom-        the criteria for diagnosis, they may be as capable of
atic therapy is rapidly disappearing. Neither is the     rendering the diagnosis as accurately as the phys-
suitability of a medication for OTC use solely the       ician.
result of its pharmacologic characteristics. Appro-         Even when a fully adequate description of symp-
priate labeling and advertising of the medication        toms and signs is not practicable for patient label-
can have a major impact on the extent to which           ing, this barrier may be surmounted by limiting use
patients understand its proper use. An OTC prod-         to patients who have previously had the condition
uct should be envisioned not just as the drug but as     and been diagnosed by a professional. Once some
the complete package of drug, labeling, and adver-       diseases have been experienced, they are unmistak-
tising, designed to encourage safe and effective         able. This approach emphasizes the need for the
self-medication. With this is mind, several vital        pharmaceutical physician to think creatively in
considerations concern suitability of a drug for         evaluating whether or not a disease can be made
OTC marketing.                                           self-diagnosable.
                          THE UNIQUE ROLE OF OVER-THE-COUNTER MEDICINE                                  143

             Differential Diagnosis                       for the treatment of ulcers has extremely low
                                                          toxicity, whereas OTC systemic decongestants
Once a condition is established as self-diagnosable,      typically have a lower therapeutic index than
a related consideration is the differential diagno-       have most prescription drugs.
sisÐthe potential consequence of confusing the          . The effects and consequences of toxicity and
disease with other similarly presenting ones, pos-        overdosage.
sibly resulting in a major delay in treatment. This     . The ease of recognition of early signs of toxicity
consideration can often be a dominant factor              to allow reduction in dosage or professional
in determining whether a condition is safely self-        assistance.
treatable. In conditions where minimal conse-
quences are likely from a misdiagnosis, a modest        Safety (negative propensity to cause harm) must be
level of diagnostic inaccuracy is tolerable to obtain   distinguished from tolerability (negative propen-
the benefits of self-medication. If the major down-     sity to cause annoying side effects). Tolerability
side of misdiagnosis is simply the persistence or       can limit OTC, use even when safety is good.
modest worsening of symptoms without serious            This is particularly true for topical agents such as
health consequences, even more difficult self-diag-     anti-acne preparations, most of which are of little
noses may reasonable. However, it is usually wise       safety concern but can produce very substantial
to place a time limit on the length of self-treatment   irritation.
without a satisfactory response.                           However, the effect of a drug on the general
                                                        population is only part of the story. The accept-
                                                        ability of a drug for market, particularly an OTC
                   Drug Safety                          drug without a physician intermediary, is often
                                                        determined by its effect on special populations,
When evaluating the safety and tolerability of a        including those patients who are particularly sensi-
drug for possible OTC use, one must first consider      tive to its effects. Care should be taken to examine
the quality of available information. Many drugs,       atypical patients in a study population, as well as
particularly those used for a long time as prescrip-    individual adverse reaction reports. Precautions
tion medications, have extensive safety databases.      may be required in the labeling for populations at
However, some do not, particularly older drugs          particular risk.
that predate modern research standards and newer           The conclusion that a drug is not acceptable for
drugs with insufficient usage. Also, with some          OTC use based on safety should be reached only
drugs, the tolerability of one formulation may differ   after determining that satisfactory labeling cannot
greatly from that of another. One example is benzyl     be developed. The pharmaceutical physician must
peroxide, in which formulations may very greatly,       weigh safety and tolerability against efficacy, both
even at the same strength but with different            in the general population and in special popula-
excipients. Where such problems mean that there         tions. Here the responsibility rests directly on the
is an inadequate database for an intended OTC           pharmaceutical physician, since there will be no
formulation, clinical testing will be needed before     other medical professional between the drug and
launch.                                                 the patient using it.
   Safety is usually the controlling factor in deter-
mining suitability for OTC use, and involves sev-
eral factors:                                                                Efficacy
. The therapeutic index (the ratio of therapeutic       Efficacy is a central issue with all pharmaceutical
  dose to toxic dose), which varies widely for both     products, and especially with OTC drugs. In the
  prescription and OTC drugs and is often less of       context of self-medication drugs, it is traditional
  a safety determinant than might be supposed.          to accept a somewhat lesser degree of efficacy in
  For example, the prescription drug sucralfate         order to improve the safety profile. Also, a lesser

standard of efficacy is normally expected by the         the physician generally focuses on a single thera-
patient, since OTC medication tends to be a first        peutic area, enjoys a large support staff that pro-
step in therapy. Failure to obtain satisfactory effi-    vide him/her with in-depth assistance, and uses a
cacy typically results in the patient seeking profes-    limited number of research protocols and tech-
sional advice, at which point more potent                niques that can be thoroughly mastered. In con-
treatments can be utilized. This does not mean,          trast, the OTC physician must be an expert on
however, that OTC drugs should not be effective          smoking cessation one day, gastroenterology the
for the conditions they treat.                           next, and dermatology the next. The OTC phys-
                                                         ician must also be concerned with detailed issues of
                                                         formulation and manufacturing.
                Dosage Selection                            The regulations governing OTC medications are
                                                         substantially different from those in the prescrip-
The extent of efficacy will depend considerably on       tion field, and the OTC physician is typically more
dosage. In the past, there was an automatic ten-         involved in regulatory matters than his/her non-
dency to reduce the dosage to half or less of pre-       OTC colleagues.
scription strength. Today, it is widely realized that       Because staff are fewer and the hierarchy sim-
dosage should not be reduced simply as a matter of       pler, the OTC physician has much more general
course; rather, a considered judgment on optimum         authority, with broad responsibility for in-line,
dosage should be made. It is being progressively         new, and forthcoming products. On the prescrip-
appreciated by both the pharmaceutical industry          tion side, this would not be true of any job short of
and the regulatory agencies that inappropriate re-       the Vice President of Research.
duction of dosage can result in reduced efficacy            Another difference concerns marketing. Typic-
with little or no safety and tolerability benefits,      ally in the prescription area, interaction with the
thus leading to needlessly ineffective treatment.        marketing department is infrequent, although
The goal is to provide the lowest effective dose. It     sometimes intense. In the OTC area, it is constant.
is vital to retain medically meaningful efficacy that    The physician educates the marketing department
will provide patients with satisfying results if self-   on medical issues surrounding a particular drug and
treatment is to fulfill its proper role in the medical   on the opportunities and limitations that these pre-
care system.                                             sent. In particular, the physician must understand
                                                         the needs of the brand managers and be able to offer
                                                         guidance. For instance, when difficulties occur in
THE UNIQUE CHARACTERISTICS OF THE                        the implementation of marketing plans, the phys-
OTC FIELD FROM THE PHARMACEUTICAL                        ician must be able to assist in developing alternative
       PHYSICIAN'S VIEWPOINT                             strategies. An OTC business is subject to intense
                                                         market pressures. The physician must help the mar-
The role of the pharmaceutical physician in the          keters deal with them effectively by frequently
OTC division of a major pharmaceutical company           playing the roles of educator and creative thinker
is substantially different from that played in the       as well as medical expert.
research or medical affairs departments dealing             One of the most surprising aspects of the
with prescription drugs. One might assume that           pharmaceutical physician's role in OTC medica-
OTC work is simpler and less involved than that          tion is the very high degree of creativity required
related to prescription medications. In many ways,       of the physician. With prescription medication, one
the opposite is true.                                    must work with whatever compounds have been
   The OTC pharmaceutical physician must be a            previously developed by chemistry and toxicology.
generalist, requiring a broad expertise in medicine,     These are brought to the physician for clinical
toxicology, and regulatory affairs. The OTC phys-        testing. There is seldom any input by the physician
ician deals with a vast variety of drugs from many       into drugs he/she will select to work on. Sometimes
different areas of medicine, including some that are     the project on which the physician will be spending
little taught in formal medical education. This con-     years of his/her life is of considerable medical inter-
trasts with research on new chemical entities, where     est, in other cases it is not. No matter what the case,
                          THE UNIQUE ROLE OF OVER-THE-COUNTER MEDICINE                                    145

the physician will be able to exercise only minimal     each time a new formulation of one of these com-
control over what compounds he/she is working on        pounds is to be brought to market. The FDA has
at any given time. While it is possible for the clin-   provided a series of numerous monographs, each
ical development of a new chemical entity to be         one of which deals with a particular narrow thera-
poorly handled, it is not possible for the clinical     peutic area, ranging from acne and anthelminthics
researcher to add any characteristic that the par-      to hormones and weight control. The therapeutic
ticular chemical entity did not possess when it was     area is discussed in some detail and specific require-
synthesized.                                            ments for well-established drugs in that area are set
   In contrast to this, in the OTC area, the phys-      forth. As long as a new formulation remains within
ician is actually in a position to greatly influence    the exact requirements set forth in the monograph
the choice of compounds on which he/she and the         for type of drug, dosage, indication, and labeling
company will do research. He/she can even cre-          statements, a compound may be formulated and
atively discover new indications suitable for OTC       marketed on the judgment of the company alone.
therapy. The OTC physician typically enjoys major       No further pre-approval or examination of any
input into all decisions involved in the company's      application to the FDA is necessary. However, if
commitment to particular compounds and formu-           the requirements set forth in the monograph for a
lations. This is true for OTC switch and for new        particular compound are to be changed in any way
formulations of older products. The formulators in      by a different dosage, a new indication, or by
an OTC operation seek extensive input from their        changes in labeling, the formulation no longer is
medical colleagues and the corporation looks to the     covered by the monograph and it is necessary to
physician for more than just straightforward opin-      submit a full NDA. As long as the monograph
ions. Creativity is required and he/she has an op-      requirements are strictly met, the physician in
portunity to devise concepts that are actually          charge will make the final judgment on whether a
developed by the company.                               new formulation is satisfactory for market. This
   Since the development cycle of OTC drugs is          system exists only in the USA and it provides for
much shorter than that of prescription compounds,       a striking amount of speed and flexibility in the
the physician is often able to see an idea of his/her   OTC marketing of products.
own brought to fruition in the form of an actual           However, it also places a very substantial
product. Typically, it requires only 3 years or less    amount of responsibility on the OTC physician.
for the development of an OTC drug, as opposed to       You can never appreciate the value of having a
7±10 years for a new chemical entity. The skillful      regulatory agency review your work and make the
use of medical knowledge and its creative applica-      final decision to allow marketing until you don't
tion to new products can make all the difference in     have them and must take the responsibility your-
the medical and business success of an OTC com-         self. This is particularly true with regard to the
pany.                                                   tolerability of new formulations. It is unlikely that
   The extent to which the OTC physician is a key       major safety problems will arise with well-known
decision-maker is especially clear in dealing with      drugs dosed at well-known levels for indications
the release to market of new formulations of drugs      that are thoroughly understood. However, with
that have monograph status. Here the pharmaceut-        topical drugs, where irritation and allergenicity
ical physician makes direct judgments on the safety     are a problem, the judgment of suitability for
and marketability of products without the inter-        market can be difficult. These drugs tend to be
vention of a regulatory agency. The US Food and         very dependent on the details of individual formu-
Drug Administration has provided for the direct         lations and you want to be sure that you have
marketing of a wide variety of OTC drugs which it       enough information before you release them to
has pre-approved in the so-called `monograph'           market.
system. The underlying concept of this system is           The need for specific clinical testing must be
that there are many drugs that have long been on        determined by the physician in each individual in-
the OTC market and for which abundant informa-          stance. A wide variety of situations may arise,
tion already exists. Therefore, it would be redun-      varying from those in which no particular testing
dant and wasteful for a new NDA to be submitted         is required to those in which an extensive series of

tests is needed before full confidence can be felt in a    bility for the OTC physician. The physician must
formulation. In short, the American monograph              ensure that the research undertaken will be com-
system provides unparalleled speed and flexibility         plete and adequate for both safety and efficacy
of drug development for those compounds which              determination purposes and must make a solo
come under it, but it requires particular vigilance        judgment as to the safety of the research subjects
on the part of the OTC physician. For all the delay        involved, with no FDA oversight.
and difficulty involved in obtaining approvals from           The details of the clinical research process are
FDA, it does have the major advantage that it              little different for OTC and prescription work.
provides a second source of learned judgment               What changes most is the role of the pharmaceut-
prior to the marketing of products. Even in the            ical physician. This role is greater in scope and
limited scope of monograph drugs, the physician            responsibility in the OTC area and everything
can often find it necessary to use all his/her abilities   must be done with greater speed.
to ensure that adequate testing is done and that
careful judgments are made before individual for-
mulations are allowed to reach the marketplace.                  PRESCRIPTION-TO-OTC SWITCH
   Because of the monograph system, one of the
more striking features of OTC drug development             One of the most dynamic areas in the pharmaceut-
is the speed with which new formulations may be            ical industry today is the prescription-to-OTC
moved from the conceptual stage to actual product          switch, commonly called the Rx-to-OTC switch.
realization. This contributes in a major way to the        This is the process by which a drug that has previ-
job satisfaction of the pharmaceutical physician,          ously been used only by prescription is converted to
but it also creates a need to act with much more           self-medication status. We have already considered
speed in advancing one's own portion of the devel-         the criteria for OTC use of medications and these
opment efforts. There is a need for the physician to       criteria represent a sound guide to determining
participate in every phase of early planning of a          what drugs are suitable for switching. There are
development program. This is the only way to               no hard and fast guidelines for determining which
ensure that it is properly handled and can be              drugs may become suitable for OTC switch, but a
quickly executed. Frequently, several companies            consideration of self-diagnosability of the disease
will be moving forward with similar projects.              state to be treated, the general safety and tolerabil-
Both commercial and personal success rely on               ity of the drug, its ability to show efficacy in the
being the first to market. Thus, the program must          hands of non-professionals, and a relative absence
be planned for success on the first try. If major          of problems with masking of symptoms all contrib-
delays in research occur, the product will usually         ute to making a drug more OTC-able.
be so far behind competition in reaching the                  The first question that arises when considering
market that it will have little commercial value.          the possibility of an OTC switch is, why has the
   Several factors accelerate the entire process of        drug not been available OTC before and what can
research in the OTC area. Since it is much quicker         be done to remove the obstruction? It is possible
and simpler for a product to remain within the             that a drug may simply not have had adequate
monograph requirements, every effort is made to            prescription experience in the past. It takes time
do so if it is possible. For research with monograph       to accumulate a substantial use database of real-
drugs, it is perhaps surprising to learn that an           world experience. This is essential to make it pos-
investigational new drug (IND) exemption is not            sible to form a judgment about safety in prescrip-
always required prior to undertaking research. This        tion use and, therefore, projected safety in OTC
is only logical, however, since for a monograph            use. What constitutes substantial use is always a
drug there is pre-approval from the FDA to actu-           relative matter. Typically, at least 3 years of data
ally launch the product into the market. It would          accumulation with a widely marketed drug is re-
not be sensible to require special pre-approval to         quired to be able to feel some security in making
perform human research via the IND system. This            judgments from the adverse reaction database ac-
considerably speeds and simplifies the course of the       cumulated. For small selling drugs, this can easily
research effort but again results in greater responsi-     take 10 years or more. The fewer problems this
                          THE UNIQUE ROLE OF OVER-THE-COUNTER MEDICINE                                     147

database reveals, the better the drug will be as a      We will discuss the peculiar aspects of the design of
switch candidate.                                       clinical studies suitable for such purposes later, but
   It is sometimes possible to accelerate the accu-     for now it is sufficient to note that they may usually
mulation of data for a promising OTC candidate by       proceed with the objectives of establishing efficacy
specialized Phase IV studies. These studies acceler-    and common side effects only, and that very rare
ate the process of data collection by conducting        side effects have already been evaluated in the pre-
what amounts to a survey amongst physicians             scription use setting.
using the drug on a prescription basis. Since the          Regardless of the reasons underlying a drug's
sole interest is the gathering of adverse reaction      prescription-only status, once a decision is taken
data, with special emphasis on rare and serious         that a drug should be moved OTC, it is extremely
events, record forms are kept very minimal, often       important to get early interaction with the regula-
to a single page. The study design consists simply of   tory agencies. This can establish at an early stage
a survey, done without control groups. Hundreds         whether or not they have concerns that the
of physicians, or even thousands, must be con-          company has not yet considered. The obvious con-
tacted to participate in the survey by submitting       cerns of safety and efficacy are not always the
brief record forms on patients they treat in their      principal issues obstructing a switch. It is possible
usual manner with the prescription drug. Such a         in some cases that FDA has no concerns in these
survey can rapidly provide a much more reliable         two areas, but that self-medication use is prevented
database than spontaneous reporting. With a             by some other peripheral but still highly important
survey, you get both a frequency of the various         consideration. Examples of such problems are in-
side effects and a reasonable estimate of the           dications which the FDA does not regard as self-
number of patients treated, which permits the cal-      diagnosable, and such problems as spread of
culation of accurate rates for the adverse effects      antibiotic resistance by OTC use of an antibiotic
observed. This is in marked contrast to the data        that might adversely affect the overall public health
obtained from an entirely spontaneous adverse re-       picture.
action database, where it is impossible to determine       It should be remembered that FDA's principal
what the efficiency of reporting is. Therefore, it is   concern in considering an Rx-to-OTC switch is
extremely difficult to estimate correct rates of oc-    from a public health perspective. This is in contrast
currence of individual adverse effects. The spontan-    to the usual viewpoint of the pharmaceutical com-
eous databases are more useful for the qualitative      panies, which tends to be focused on the treatment
evaluation of what can happen with a drug than for      of the individual patient. The FDA's positions may
the quantitative evaluation of its true frequency.      be better understood by remembering that they
This type of adverse reaction survey study can          tend to take the big picture, whereas the pharma-
pave the way for a switch effort in much less time      ceutical company tends to concentrate on narrower
than needed if reliance is placed solely on spontan-    issues. There is nothing that will facilitate the Rx-
eous reports for collection of data.                    to-OTC switch of a drug more powerfully than
   If the principal barrier to switch has been a lack   convincing the Agency that its OTC availability
of clinical experience with a drug, this can be rem-    will benefit public health.
edied by the collection of a large adverse reaction        Other issues that may concern the FDA far more
database. Once this is done, it is usually straight-    than the industry, are precedent-setting issues. The
forward to establish that the drug is safe in pre-      FDA functions in an environment in which they
scription use. This is a major advance on the road      have ongoing responsibility for the approval and
to OTC approval, but it certainly does not yet          switch of many different types of drugs. They may
prove that the drug will be safe and effective in       feel that the precedent set by a particular switch
the hands of consumers without the benefit of a         could be damaging in terms of their overall policy
learned intermediary. In order to establish this        posture, even though they have relatively little con-
additional point, it is almost always necessary to      cern about the switch itself. If this is the situation,
supplement the analysis of adverse reaction data-       careful negotiation with the FDA and care in how
bases with clinical studies in realistic conditions,    the switch is presented to minimize setting an awk-
using the labeling composed for the OTC product.        ward precedent can often resolve the issue.

   Another broad-scale public health concern which      will be too late to complete the necessary studies and
may worry the FDA is the implied message given to       regulatory applications in time for a smooth transi-
the consumer by the OTC availability of a particu-      tion. In order to make a seamless transition at the
lar compound. This concern is illustrated by the        time of greatest opportunity for a company, it is
situation with soluble fiber cholesterol-lowering       necessary to plan for the OTC switch many years
agents of the psyllium type. These agents have          in advance. Nothing can substitute for an awareness
been shown to lower cholesterol, but only to a          of the OTC potential of the company's portfolio of
very small degree. It was felt by the FDA that, if      drugs so that suitable programs can be designed and
they become established with claims of cholesterol      put into action on a timely basis.
reduction, the population may be misled into feel-         There are two fundamentally different types of
ing that they have made major inroads into their        Rx-to-OTC switches from the standpoint of the
cholesterol problems when, in fact, they have not.      scope of the research program required. Switch
The message communicated to the consumer by             programs can vary from large NDA programs as
making these compounds available constituted a          extensive and expensive as anything found in the
barrier to this use.                                    new chemical entity development, to programs
   Early negotiation with the FDA will assist in        consisting of little more than a single study. What
identifying their concerns and will help in obtaining   influences the basic size and expense for an Rx-to-
their insight into the types of studies required to     OTC switch is whether or not you propose to
resolve the difficulties presented by any given case.   change either the indication or the dose of
The skill of the regulatory department in structur-     the drug when you bring it into the OTC market.
ing and presenting a switch proposal is vital to its    If the indication or the dose is to be changed, you
success.                                                will be involved with an entirely new NDA, which
   The timing of an Rx-to-OTC switch project is         is needed to show the fundamental efficacy and
always a major contributor to its success. The          safety of the drug, either at its new dose or in its
timing is influenced by both regulatory and com-        new indication. Such a program obviously will re-
mercial considerations. The completeness of the         quire several years and involve extensive expend-
available database is critical and can set the timing   iture. In startling contrast to this are the programs
of a switch. Often, however, the key factor in de-      of modest size often required for the switch of
ciding when a switch effort should begin will be        drugs that will be taken into the OTC market at
commercial. This involves the major benefits to a       their existing prescription dosage and for their
company in moving a drug to OTC status at ap-           existing prescription indications. Here, the regula-
proximately the same time as its patent expires. By     tory agencies will generally accept the concept that
doing this, the company may greatly mitigate the        there is no need to prove again the basic safety and
loss of patent protection. A switch can offset          efficacy of the drug, because this has already been
the precipitous decline in the innovator's share of     done in the primary new chemical entity NDA.
the Rx market. At this difficult time for a company,    Such a repetition would not provide useful new
nothing is welcomed more than the rapid growth of       data. What will be required is an actual use study,
an OTC market that, in some cases, can be even          to show that the proposed labeling for OTC use is
larger than the original prescription sales. Typic-     effective in enabling patients to use the drug prop-
ally, once a drug has gone OTC, it is sold at a lower   erly. Also, it may be necessary to address whatever
price with smaller profit margins, but the total        specific factor it is that has been previously
volume tends to increase several-fold. The result       obstructing the drug from OTC use. For example,
can be a highly favorable economic situation for        if there is a question as to whether the prescription
the company.                                            indication that will now be taken OTC is self-
   Unfortunately, in many cases, a switch at the time   diagnosable, then a study of self-diagnosis will be
of patent expiration does not occur and a long delay    required. This occurred with the vaginal anti-
intervenes before OTC status is reached. This is        fungal compounds, which were long kept on pre-
generally due to the tendency within companies to       scription status because of questions as to whether
seriously examine the need for an OTC switch only a     women could effectively diagnose vaginal candidia-
year or two before patent expiration looms. This        sis themselves. Only a single study was required to
                           THE UNIQUE ROLE OF OVER-THE-COUNTER MEDICINE                                     149

resolve this issue. It was extremely unusual for            Realism is the key to OTC research design. The
the pharmaceutical industry, in that it involved         best OTC studies are often called `slice-of-life' stud-
no drugs of any kind. It was simply a study of           ies. This communicates effectively the idea that you
women's ability to self-diagnose but it resolved         want to get a realistic real world impression of what
the one outstanding issue that had blocked OTC           the drug will do when subjected to all the variables
approval.                                                of an actual use setting. In order to achieve this,
   The time required to carry out studies on such        selection criteria should be minimized rather than
special questions can vary, considerably depending       maximized. You want to study the same kind of
on the complexity of the question. However, it is        population that will come into drug stores and even
typically a brief program and its budget is com-         into supermarkets and actually buy this drug.
monly small by the standards of the pharmaceut-          Eliminating large segments of this population by
ical industry. It is obvious that in the planning and    strict admission criteria will only produce a result
preparation of a switch program, it is essential not     that is redundant of what has been previously dem-
to assume that a full safety and efficacy program        onstrated and which is not relevant to the actual
will be required. Rather, early communication with       conditions of use the drug will encounter. Every
the regulatory agencies is needed in order to estab-     effort should be made to simulate the way in which
lish what barriers actually exist.                       patients will actually use the drug. In some cases, it
                                                         is even known to go so far as to actually have
                                                         patients pay for the drug in order to obtain the
    SPECIAL STUDY DESIGNS FOR THE                        motivational factors associated with a purchase.
               OTC AREA                                  This illustrates the degree to which some studies
                                                         go to ensure that they're actually studying the con-
The philosophy for OTC study design is signifi-          ditions of use and not some artificial situation
cantly different from that of prescription medica-       which is set up for the study and may give seriously
tion studies. With prescription medications, you         misleading results. In the same philosophical vein,
are typically striving to answer the basic scientific    it is important to design the study for minimum
questions of `can this drug work effectively?' and `is   interference with the patient. He/She must be left
it safe to administer to people?'. Therefore, it is      free to act, guided only by the labeling. Excessive
appropriate to study these new chemical entities         intervention by the investigator will distort the
primarily in controlled settings with extensive in-      results.
clusion and exclusion criteria. This provides in-           Some may react to studies of this type by feeling
creased safety for the study participants, who will      that they are unscientific. Actually, they are just
be using a drug of relatively unknown toxicity, and      addressing a different question. What is important
allows a reduction in the inherent variability of the    to remember is what hypothesis is being tested. At
study population, so as to obtain a clearer scientific   the stage where a drug is being considered for a
answer to the questions of basic safety and efficacy.    switch, the question is, `What impact will this drug
Every effort is made in studies of this type to con-     have on the public health as it will really be used
trol for all possible variables and to reduce random     by the lay public?'. This is the question that the
real-world circumstances to a minimum.                   FDA needs answered. Excessively controlled study
   In the case of drugs being prepared for the self-     designs will only detract from the studies' rele-
medication market, just the opposite sort of phil-       vance.
osophy applies. In this situation, a great deal of          These studies are tests of the labeling as much as
evidence is already available about the abstract         they are tests of the drug itself. It is essential that
safety and efficacy of the drug. Its basic ability to    the combination of the drug and its OTC labeling
function has already been well established or it         work closely together to enable patients to self-
would not have been approved in the first place.         medicate effectively. One should never study an
The key issue in the switch of drugs is to establish     OTC drug as the pharmaceutical substance alone,
whether the drug can work in the real-world con-         but always as the combination of the labeling and
text, with all the inherent happenstance and ran-        the pharmaceutical. It follows, then, that the devel-
domness that implies.                                    opment of this labeling is very important for OTC

use. It may, in fact, make the entire difference            marketing department and focus group sessions can
between a drug being a switch candidate and                 be invaluable in discovering the possible existence
remaining on prescription status indefinitely. Not          of advantages for a particular formulation over its
only is a great deal of creativity necessary in de-         competitors. Careful review of the pharmacologic
veloping effective labeling but appropriate label           literature is another way in which differences can be
comprehension studies are also important in ensur-          detected. Even small differences may be quite mean-
ing that the best labeling is obtained.                     ingful to patients, even though they may appear
   Research has shown that patients by and large            minor to the pharmacologist, who is not actually
do read labeling and they do heed it, particularly          using the drug him/herself. In the case of an anti-
with new drugs that they have not used before.              nausea drug, for example, a difference in onset of
Prior to any program being advanced to the stage            action of 10±15 min can be highly significant if
of the definitive clinical studies, it is wise to develop   you're the one suffering from the nausea. There-
a variety of different versions of the proposed la-         fore, a careful review of literature can often find
beling, so that these versions can be tested in label       pharmacokinetic and other differences that can be
comprehension studies. These studies are some-              more meaningful than they at first appear. Any
times organized by the medical department and               difference that is not meaningful to the patients
sometimes they are carried out as market research,          will not produce much difference in sales. In
since they need not involve actual ingestion of             selecting market support opportunities for study,
drug. They consist of comparative studies in                it is important to manage expectations so that the
which patients in a realistic setting read the pro-         marketing group can understand that only real dif-
posed labeling and then are quizzed on their com-           ferences that are meaningful to the patient and can
prehension of it. In this way, it is possible to see        be demonstrated with reasonable certainty in clin-
whether they understand how the drug ought to be            ical trials are worth pursuing. If expectations are
used and whether they have understood key pre-              allowed to grow out-of-hand, one will find oneself
cautions. It is best to check both short-term and           designing studies for the most bizarre purposes and
long-term comprehension to see how well the pa-             encountering a very high rate of failure, because
tients are able to remember what they have learned.         you're chasing after advantages that never existed
This sort of pre-screening of labeling can be abso-         in the first place.
lutely essential to success and it has saved many              Once a probable new claim has been identified
careers by avoiding disasters in large-scale defini-        and the chances of its being scientifically valid have
tive studies.                                               been assessed, you must carry out actual studies to
                                                            support the claim. In most situations, two-good
                                                            quality studies are necessary to support an adver-
         MARKET SUPPORT STUDIES                             tising claim, even though these are not necessarily
                                                            regulated by the FDA. In some special situations, a
In addition to the Rx-to-OTC switch, a second type          single study may be enough. Guidance on this issue
of study is very prominent in the OTC area. This is         may be taken from the company's legal depart-
the market support study. These are clinical studies        ment.
intended to demonstrate particular advantages of               To understand what will be needed to support a
one company's drug over another's, or of one par-           claim, it is important to understand how the adver-
ticular formulation of the same drug over other             tising and marketing of OTC drugs is regulated.
competitors. Opportunities for demonstrating                Typically, after a brief initial period, the FDA does
such advantages depend on locating genuine differ-          not take a primary role in such regulation. Rather,
ences that can be successfully studied. While any-          this duty passes to the Federal Trade Commission
thing can be tried in a clinical study, only authentic      and, much more importantly, to the federal courts.
differences will emerge as successful claims at the         The OTC industry's advertising claims tend to be
end of the study process. Locating such possible            self-enforcing. The companies maintain a close
advantages can be done through careful attention            vigilance on each other's advertising and tend
to feedback from actual users of the drug. Usage            to be eager to sue their competitors in the federal
and attitudes (U and A studies) studies done by the         courts if any unsupportable advertising claims
                           THE UNIQUE ROLE OF OVER-THE-COUNTER MEDICINE                                  151

are suspected. The possession of scientifically             In summary, the role of the OTC physician is
sound studies is of great value in preventing law-       significantly different from that of the pharmaceut-
suits and ensuring that if they do occur, they will be   ical physician working in any other area of the
won.                                                     industry. It requires a greater degree of creativity
  Although the pharmaceutical physician rarely           and independence. It features a different kind of
pictures him/herself as a person testifying in           responsibility, since there is no other physician
lawsuits, this is not an uncommon experience for         between you and the patient to share that responsi-
the OTC medical director. This need not be viewed        bility with. One is more completely on one's own
with any particular trepidation if you have              and called upon to utilize an expertise which is
been careful to prepare a satisfactory scientific        broader than that required of the physician in vir-
basis for the advertising claims that you have ap-       tually any other area of pharmaceutical develop-
proved.                                                  ment.
Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                       Copyright # 2002 John Wiley & Sons, Ltd
                                                                    ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                             Section III
                               Special Populations
      Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                             Copyright # 2002 John Wiley & Sons, Ltd
                                                                          ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                          Introduction to section III
                                               Lionel D. Edwards
                             Pharma proplus INC and Novartis, New Jersey, USA

In 1993, the US Food and Drug Administration                       In general, globally and especially in the USA,
(FDA), Europe's Committee for Proprietary Me-                   legislation controlling food and drugs (including
dicinal Products (CPMP) and Japan's Ministries                  devices and biologics) has been stimulated by thera-
of Health and Welfare (MOHW) issued regulatory                  peutic disasters. This, in the USA, caused the im-
requirements for testing and labeling in a `special             plementation of the Food, Drug and Cosmetic Act
population', namely the elderly. These were not                 of 1906, which outlawed the practice of embalming
promulgated in isolation but after consultation                 meat for consumption. Further disasters triggered
with academia and industry. In the USA, initially               subsequent multiple amendments to the Act.
this was done under the auspices of the American                   In special populations, perceived omissions of
Society of Clinical Pharmacology and Therapeut-                 research and development have also resulted in
ics. Industry was allowed to participate and was                specific amendments to this Act. On occasion,
largely credited with aiding the process. The First             these amendments have been due to political pres-
International Conference on Harmonization (ICH)                 sure from special advocate groups, rather than due
held in Europe (November 5±7 1991), again in-                   to a specific therapeutic disaster.
volved the regulators and the regulated and, for                   Why did industry ignore these special popula-
the first time, involved Japan as a major contribu-             tions, which represent major markets? First, the
tor. As a result of pre-, during-, and postconference           costs of additional research would add to the al-
discussions, success was achieved. The `elderly'                ready enormous cost of drug and devices research.
drug guidance was the forerunner of many future                 Second, the ever-present fear of litigation resulting
tripartite agreements in the clinical area.                     from perceived exploitation, coercion, and vulner-
   The special populations covered in the following             ability of these special populations discouraged
chapters include the four major demographic seg-                industry and the FDA from policies of inclusion.
ments: the elderly, women, children and major                      Other influences determining research directions
ethnic groups. While any smaller grouping of                    in drugs and devices were paternalism (protection-
people or diseases may be labeled `special', such               ism) and the money available for grant projects,
may be better described as `orphan' populations,                guided by the numerical male dominance in the
which are the subject of discussions elsewhere in this          reviewing process of research priorities.
book (see Chapter 16). The four major demo-                        For the pharmaceutical industry, it is ironic that
graphic segments were designated `special popula-               attention to these special populations is now
tions' because, despite the large size of each segment          proving `good business', either because of an exten-
(globally, women constitute 51% of the popula-                  sion of protected patent life, or because of the
tion), pharmaceutical research has been sparse in               development of special business units. These units
these groups. The basis for this is multifactorial.             have increased market penetration and retention of
Different responses to needs and medicinal inter-               drugs for third-party reimbursement and allowed
ventions, compared with that in the white male                  niche dominance. The latest of the four major
population, have been only sporadically addressed               special populations rulings by ICH, the final rule
by the research, academic, and industry pharma-                 on Acceptability of Foreign Data, was imple-
ceutical development communities.                               mented in July 1998. While it is the latest, it will
                                    INTRODUCTION TO SECTION III                                         155

not be the lastÐthe future impact of the genome      tabolism in detail, because of the societal sensitivity
project on each of these major demographic seg-      and because of a relative paucity of data in the
ments, and its influence on genomic pharmacology     literature.
and gene therapy with regard to these `special          The chapters on geriatrics and pediatrics (chap-
populations', has yet to be felt.                    ters 14 and 16) focus mainly on the evolution and
  Each chapter will give a limited historical con-   requirements of the drug development process, be-
text. The chapters dealing with drug development     cause data on the physiology and metabolism of
in women (chapter 15) and ethnic populations         these groups are both widely known and available
(chapter 29) explore issues of physiology and me-    in the literature.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                 Drug Research in Older Patients
                                               Lionel D. Edwards
                                      Basking Ridge, NJ, Carlsbad, CA, USA

                 DEMOGRAPHICS                                      The social and healthcare impact of these demo-
                                                                graphics in the USA and globally will lead to an
The elderly (over 64 years old) comprise 12% of the             increased demand for better medicines directed at
US population and 17% of Sweden and Japan. This                 a healthy old age. This elderly population have more
sector continues to grow. In the USA it, is estimated           income than average per capita income. In addition,
that the elderly population will grow to 14% by the             with more time on their hands to lobby, they are
year 2010 and reach 17% by 2030 (US Bureau of the               more likely to vote, and can be expected to use their
Census 1996). This, together with their known sen-              political muscle to make demands on their govern-
sitivity to medications (Everitt and Avorn 1986),               ments. The governments will respond in the usual
contributed to acceptance by industry of additional             knee-jerk reactionÐ`more regulations and con-
requirements for testing in the elderly.                        trols' on industryÐwhile increasing funding for
   The US Bureau of the Census, International                   academic research aimed at improving the quality
Database (1996) (National Center for Health Stat-               of life and the prolongation of active old age. It will
istics 1996) projected that, for the year 2020, the             be interesting to see whether a more extended life
less developed countries would contain only 16.4%               expectancy, over and above the current projections,
of the world population compared to 27.1% in 1996,              will reverse the depopulation trend.
and that by 2020 the mean age of the population in
more developed countries would be 42 years, up
from 36 years in 1996. In developed regions, the                    IMPACT ON SOCIETY OF AN AGING
elderly would outnumber young children by 8:1,                               POPULATION
e.g. in Italy, based on current fertility and survival
rates, only 2% of the population would be 5 years or            In developed countries, by 2020, the working
younger, but 40% would be 65 years and older.                   population aged 15±65 years will fall from 22% in
   There were even more startling projections by the            1996 to 16%. Those aged 65 years and over will
United Nations International Population Division                increase to 20% from 16% (US Bureau of Census
(1996). They projected life expectancy in the `de-              1996). In the USA, 60 years ago, the retirement
veloped' countries to reach 81 years by 2050. For               age for Social Security `pension' was designed
less developed countries, this would still reach 76             for an expected average lifespan of 65 years. Al-
years. However, this increase in the global elderly             ready this has been pushed back to 67 years by year
population would be proportionally offset by a                  2004, and additional legislation will probably push
decrease in fertility rate, now under way, from 1.7             the age requirements back to 70 in 10 years' time,
births/woman down to 1.4 in the Western world.                  when the `baby boomers' swell the retired popula-
This is below the replacement rate. For Second                  tion.
World regions, the rate of about 3.3 births/woman                  To encourage the healthy older person to con-
would decline to 1.6. Even in the least developed               tinue working beyond 65 years, legislation was
(Third World) countries, 5 births/woman would                   passed to remove the penalty (in workers 65±70
fall to 2 by 2050. Thus, the whole world would                  years) of the loss of $1 for every $3 earned from
actually start to `depopulate' in 40 years.                     Social Security benefits in the US. In 1999 it was

proposed that, because of the high cost of medica-      number of drugs increased with age. The elderly
tion and because the older people were the greatest     comprised 18% of the population but received 45%
users, they be eligible for drug cost reimbursement     of all prescription items (Lassila, Stoehrt Gangula
under Medicaid. This would give the US Govern-          1996).
ment reimbursement control on over 58% of drugs            One in 10 admissions to acute geriatric units were
prescribed and the power to `set prices', as in other   caused or partly caused by adverse drug reactions.
countries (e.g. Canada, the UK, France, Italy, Ger-     The drugs involved most commonly were benzodi-
many). This has sent a chill through the US             azepines, warfarin, digoxin, and non-steroid anti-
pharmaceutical industry.                                inflammatories (Deaham and Barnett 1998). Tam-
   Of great concern is the social and financial         blyn (1996), in his review article, cited reports of
impact of Alzheimer's disease, whose incidence          adverse events causing 5±23% of hospitalizations,
per capita increases to 32% of the surviving popu-      nearly 2% of ambulatory visits and 1 in 1000 deaths
lation at ages 80±85 (and declines rapidly after age    in the general population. These rates increase in the
85). Many live with this disease for 5±8 years before   elderly. Errors in prescribing accounted for 19±36%
succumbing. This causes enormous detriment to           of hospital admissions due to drug-related adverse
the surviving spouse and family and to family           events.
finances, and must eventually impact Medicaid              To compound this worrying situation, there is
and Medicare Federal and State budgets. The dur-        the concomitant use of over-the-counter (OTC)
ation of financial burden of terminal care is 1±4       non-prescription drugs. Only 50% of physicians
months in general (1±18 months for Alzheimer's          or health workers ask about OTC drug use, yet
patients) and, even with what would normally be         40% of all drugs used by the elderly are non-
an adequate pension, this burden can financially        prescription drugs. In all, 69% of the elderly use
ruin the surviving spouse.                              OTC drugs and 70% take at least one prescription,
   Immigration from the Third World to the de-          as described earlier. In addition, 31% take alcohol
veloped countries will increase as countries of         frequently (Conn 1992).
aging populations try to replace the loss of their         This new potential for adverse drug interaction is
labor pool. This is already happening in Europe         enormous. NSAIDs and aspirin interact with anti-
and in the USA. This again will put further pres-       coagulants such as warfarin or coumadin, can in-
sure on Medicare and Medicaid, as many of these         crease the bleeding tendency, and not just from the
immigrants will suffer from tuberculosis, hepatitis,    stomach. Antacids can decrease the excretion of
and intestinal disease, endemic to many of their        antidepressant tricyclics, quinidine, pseudoephi-
home countries. In 1997, 39% of tuberculosis            drine and indomethacin. They can also reduce the
cases in the USA were in foreign-born parents; in       absorption of digoxin and b-blocker hypertensive
California this rose to 67% (Satcher, 1999) and the     medication. These are only a few of the multitude
annual cost of diagnosis and treatment of the 1 mil-    of interactive drug effects. This is imposed upon the
lion immigrants was $40 million (Muenning et al         reduced efficacy of hepatic metabolism and elimin-
1999). This will cause further competition for avail-   ation, and renal excretion in the elderly; thus, drug
able health dollars.                                    OTC use can add to the recipe for toxic drug accu-
                                                        mulation and, in the latter case of antacids, cause
                                                        further damage to the kidney by loss of blood
   PRESCRIBING AND ADVERSE EVENTS                       pressure control and worsening cardiac failure.

Studies of drug utilization in the elderly showed
that older people receive disproportionate amounts         PRACTICAL AND ETHICAL ISSUES OF
of medication (Rochon and Gurwitz 1995). A                    DRUG RESEARCH IN OLDER
study in rural persons 65 years or older showed                     POPULATIONS
that, of 967 interviewed, 71% took at least one
prescription drug and 10% took five or more pre-        Traditionally, elderly subjects were frequently ex-
scription medications. Again, women took more           cluded from clinical drug development (unless the
medications than men, and in both groups the            disease being treated was more prevalent in that
                                     DRUG RESEARCH IN OLDER PATIENTS                                         159

age group). The reasons given were that the elderly        contained disclaimers. As a result of this, and
suffer from too many other diseases, require con-          the fact that 30% of prescription drugs by then
comitant medicines, are more frail, and are more           were consumed by just 12% of the population
vulnerable to adverse events. All these can cause          (those over 65 years), a new guideline was issued.
`static' in the interpretation of the data, and give       Thus, the FDA Guideline on Drug Development in
undue weighting to adverse events in the labeling          the Elderly (1990) recommended that, if a drug
and product package insert.                                was likely to have significant use in the elderly,
   In addition, the elderly can exhibit differences,       then studies should be done in an elderly popula-
both physiologically and pathologically compared           tion. These studies should look at effectiveness
with the younger population; the contrast in speed         and adverse events by age. In addition, other stud-
of disease progression of prostate cancer in the           ies should determine whether older people handle
`younger elderly' compared to the slow rate in the         the new drug differently (a 30% decrease in renal
`older old', is an example.                                excretion and liver metabolism is normal in a
   The elderly are often confused or demented,             healthy elderly person). This guideline also re-
making informed consent and their continuation             quired studies of the pharmacokinetics and, where
in a study questionable. Lastly, because the elderly       possible, pharmacodynamic studies of the new
indication may represent only a small use of a drug,       drug in the elderly. The Guideline also urged the
it is uneconomic to include the elderly in a drug's        study of possible drug interactions with drugs com-
development program. These are often the per-              monly used concurrently in this age group. Digoxin
ceived concerns of both investigators and pharma-          was given as an example. Looking even further
ceutical firms.                                            forward to the future, the Guideline encouraged
   What is `geriatric'? Strictly defined, it describes a   the inclusion of patients over 75 years.
person aged 65 years or over, but aging is neither a          Medicines in the elderly had become a world
homogeneous nor a linear process. There are very           issue and, in 1994, the FDA implemented the
fit 80 year-olds who climb mountains, and young            ICH tripartite guidance, Studies in Support of
children dying from genetic advanced aging                 Special Populations: Geriatrics (Fed Reg August
(progeria). The elderly therefore cover a spectrum         1994). The agency followed up with specific re-
of fitness. So many of the above concerns can be           quirements on content and format of labeling for
reduced by selecting `uncomplicated, healthy' older        human prescription drugs; addition of a `Geriatric
patients in Phase I studies, who are increasingly          Use' subsection in labeling (Fed Reg August 1997).
available due to the success of medicines and pre-         This set out priority implementation lists of drug
ventative medicine.                                        categories for information in geriatric population
   However, there is a need to know how medicines          and gave the industry 1 year to comply. It also set
behave in the real worldÐnot just their inter-             out the specific content and format of wording to
actions with other medicines, but in other disease         be used.
states suffered concurrently, which is often the case
in a geriatric population and less so in younger age
   For the elderly, of equal importance to life ex-             OVERVIEW OF INTERNATIONAL
tension and cure is improvement or preservation of              HARMONIZATION CONFERENCE
their activities. Thus, the results of quality of life,                GUIDELINES
disease outcomes and pharmacoeconomic studies
are of even greater relevance to this special popula-      This guideline was very similar to the 1990 FDA
tion and to third-party payers.                            guideline in intent. It requested that:

                                                           1. Studies should be done in new molecular en-
           REGULATORY RESPONSE                                tities (NMEs) or new chemical entities (NCEs)
                                                              likely to be used in the elderly, either to treat a
By the 1980s, most of the new medicines still                 disease of aging or because the disease is also
had little or no information on elderly dosing or             common in the elderly.

2. Studies should include patients 65 years and         dine are given as examples. Finally, other common
   older, and preferably patients aged 75 or            drugs most likely to be used with the test drug are
   older, and advised against arbitrary age cutoff      recommended to be explored for possible synergis-
   (patients aged 60±65 are not considered eld-         tic or antagonistic drug interactions.
3. Meaningful numbers, especially in Phase III: a
   minimum of 100 patients was suggested for a
   non-geriatric specific disease (e.g. hyperten-                   INDUSTRY RESPONSE
4. Analysis of the database for age-related differ-     A survey conducted by the FDA in 1983 (Abrams,
   ences of efficacy, adverse events, dose, and         1993) showed that, for 11 drugs recently approved
   (gender) relationships. A geriatric database         or awaiting approval of New Drug Applications, in
   may contain data from the main Phase II, III         seven applications 30±36% of patients were aged
   studies or from a geriatric-specific study.          over 60. In one application, a study on a drug for
5. Pharmacokinetic studies (PK), either formal          prostate cancer, 76% of patients were, not surpris-
   pharmacokinetic studies or on a population           ingly, over 60 years old (Everitt and Avorn 1986).
   basis. For the latter, a blood sample is taken       An additional survey by the FDA in 1988 of 20
   from many patients on up to four occasions.          NDAs showed similar results but, in addition, an-
   The time of dosing is recorded, and the time of      alysis by age and pharmacokinetic studies in the
   samples. The patients must be at `steady state'.     elderly were frequently included. A survey by the
   This way, an adequate population PK plot can         Pharmaceutical Research and Manufacturers of
   be built.                                            America (PhRMA) (Mossinghoff, 1995) showed
6. Pharmacokinetic studies in renal impaired pa-        that 132 medicines were being studied for potential
   tients if the drug or metabolites are renally        use in the elderly in 157 indications.
   excreted. If the NME is excreted and/or metab-          A private survey of 19 pharmaceutical companies
   olized by the liver, a hepatic-impaired study        operating in the USA (Chaponis, 1998) ranked car-
   should be undertaken. These studies do not           diovascular, depression, Alzheimer hypertension,
   have to be done in elderly patients (they are        rheumatoid arthritis osteoarthritis, and oncology
   usually done on a new NME anyway).                   as the most important therapeutic areas in their
7. Usually, differences in the therapeutic response     company. All of these are commonly found in the
   or adverse events are too small to detect at an      elderly. Why did companies target these thera-
   equivalent plasma level between ordinary adult       peutic areas in the geriatric population? This drew
   and elderly patients to make this a require-         the response: `It's a growing population,' from 77%
   ment. However, separate studies are requested        of respondents, and `increasing market size' from
   of sedative hypnotic psychoactive drugs or           58% of the 27 company respondents. Companies
   drugs having a significant CNS effect, and,          were asked which types of geriatric-based clinical
   similarly, if Phase II, III studies are suggestive   trials they conducted. Safety, efficacy, pharmaco-
   of an age-related difference.                        kinetic- and drug interaction studies were quoted in
8. Drug interaction studies should be done on           that order of frequency, which, because of the
   digoxin and oral anticoagulants, for these           introduction of the guidelines, is to be expected.
   drugs have a narrow therapeutic range and            However, the next most frequent studies were qual-
   are commonly prescribed in the elderly. These        ity-of-life, pharmacoeconomic, drug disease (out-
   drugs frequently have their serum levels altered     comes) and patient satisfaction studies. The later
   by other drugs. Where drugs are heavily me-          studies reflect the elderly and third-party payers'
   tabolized by the liver, the effect of drug enzyme    influences (Chaponis 1998). In its 1999 survey,
   inducers and inhibitors should be explored.          PhRMA reported that over 600 medicines were
   Similarly, drugs which will share the same           then being developed for diseases of aging. This
   cytochrome P450 enzyme pathways should be            reflects the increasing importance of medicines for
   tested. Ketoconazole, macrolides, and quini-         the graying population of America.
                                    DRUG RESEARCH IN OLDER PATIENTS                                        161

   ISSUES OF DISEASES IN THE ELDERLY                     because of limitations on non-clinical toxicity pre-
                                                         dictors and the application of biologic measure-
Heart failure is a leading cause of hospitalization of   ments on a traditional drug appraisal system.
the elderly. About 5 million Americans suffer from          The new non-steroidal anti-inflammatory drugs,
this disease, which has a high mortality rate. Con-      including the Cox II inhibitors, because of the vast
trol of blood pressure, use of b-blockers, ACE           range of arthritic diseases, require that careful se-
inhibitors, and now spironolactone (Pitt et al           lection of indications for initial product approval
1999) will result in further improvement of mortal-      must be undertaken. Rarely do companies have
ity which have started to fall from 117 per 110 000      the time or money to develop all the pain indica-
in 1988 to 108 in 1995, according to the Center for      tions (acute, chronic use) or to study arthritic
Disease Control and Prevention (CDC).                    diseases prior to product launch. As with hyperten-
   Because of its severity, patients are on many         sion, the numbers of patients required in the data-
concomitant medications apart from the aforemen-         base will be large for product approval, especially
tioned drugs, such as diuretics, digoxin, potassium      for safety.
supplements, medicines to improve pulmonary                 Depression is a frequently missed diagnosis in the
function, and antibiotics to control frequent infec-     elderly. The Alliance for Aging Research says that
tion in edematous and often emphysematous lungs.         15% of Americans aged 65 years and older experi-
Measurements of heart function, and the long dur-        ence clinically relevant depression. It can amplify
ation of these studies and large patient numbers         the underlying disabilities in stroke, arthritis, Par-
required for mild to moderate heart failure (end-        kinson's disease, slow or prevent recovery from
point death), make these very challenging and ex-        hip fracture and surgery, and be mimicked or
pensive studies.                                         masked by an underactive thyroid. The latest re-
   Hypertension affects about 50% of the elderly         ceptor-specific medicines have a very much reduced
population. There is also a unique form called           potential for adverse events and drug interactions.
isolated systolic hypertension, which affects 9% of      Difficulties can arise from confusion, memory im-
the geriatric population and is growing as the           pairment and disorientation, which are common in
population ages. The challenges of doing studies         the depressed elderly. This brings challenges of
in this area increase with the age of patients admit-    ensuring both drug compliance and follow-up at-
ted, which correlates with increased concomitant         tendance in clinical studies. It also may require
medications and illness and compliance, but other-       guardian co-signature for informed witnessed con-
wise relates well to study designs in the younger age    sent.
group.                                                      Parkinson's disease affects more than 1 million
   Stroke thrombotic or hemorhagic is the third          Americans and about four in every 100 by 75 years
leading cause of death, killing 160 000 persons in       of age. Ten new drugs are under development. The
the USA each year, Seven out of 10 victims are           patients may become very physically disabled but
aged 65 or older. Of those that survive, one-third       still retain a clear sensorium until the very end
will be permanently disabled. Some improvements          stages of the disease. Thus, drug compliance and
in these figures are hoped for, with earlier use of      follow-up visits are easier to achieve than with
thrombolytics in case of cerebral thrombosis. As of      Alzheimer or depressed patients.
1999, over 20 new drugs were in development to              Alzheimer's disease is the eighth leading cause of
treat this condition.                                    death in the elderly and already affects some 4
   Arthritis causing inflammatory and degenerative       million Americans. The incidence rises from 2% at
changes around joints affects 43 million in the          65 years to 32% at age 85. The National Institute of
USA, and CDC projects that this will rise to 60          Health estimates that at least half of the people in
million by 2020. It can be caused by over 100            nursing homes have this disease. A small study of
different diseases, but the commonest is osteoarth-      donezil showed that this treatment avoided the
ritis and rheumatoid arthritis. New medications,         need for home nursing care by half compared to
such as the antitumor necrotic factor a-blockers,        those who did not receive the medicine (Small,
raise fresh challenges to clinical study methodology     1998).

   Clinical studies in this disease are very expensive,    fluctuating mental ability. Ironically, regulations
often requiring several collaborating disciplines at       governing research in these patients were proposed
each investigative site. A gerontologist, a neurolo-       but never voted upon. The NIH established a
gist, a psychologist and a psychiatrist may be             policy which allowed a patient, when he/she was
required, in addition to the usual support staff.          still in good cognitive condition, to appoint a
Multiple cognitive tests and behavioral ratings of         `Health Care Agent'.
the patient often involving primary caregiver                 For industry, prior written agreement of a family
ratings, will be requiredÐall this in addition to the      member with the potential subject to act as `guard-
basic Alzheimer's Disease Assessment Scale                 ian' is preferred but not always attainable. It is best
(ADAS±COG). These studies at present require               for the researcher him/herself to meet with rela-
large numbers of patients to show the often small          tives, nursing staff, and residents and fully explain
improvement, as well as months of observation to           to them the study purpose, benefit and risks, as well
detect a slowing of progression.                           as to the patient. Not infrequently, any of these
                                                           persons may feel protective of the patient and
                                                           undermine the research objective. It is wise that
 ISSUES IN THE CONDUCT OF CLINICAL                         all family members who are not involved be sent a
        STUDIES IN THE ELDERLY                             letter explaining the research, including a form to
                                                           be completed if they wish to prevent the patient
                Informed Consent                           being involved in research.

In general, the principles are no different with the
elderly than with other adult persons; the elderly                             Compliance
are just as subject to the relationship to the re-
searcher if the clinician and researcher are one           Compliance in the elderly in general is similar to
and the same. Not wishing to offend (by refusal)           that of the general population. If more than six
is very strong in the elderly, and also they are also      drugs are prescribed long-term, or more than
subject to `therapeutic fallacy', i.e. they find it hard   three doses/day are required, then compliance will
to accept that, despite repeated descriptions of risks     suffer (Blackwell, 1979; Gately, 1968). These
and possible benefits, the treating physician could        factors are more common in the elderly. Recom-
be really offering them treatment of uncertain             mendations for improving compliance in older pa-
benefit or risk. The elderly are more likely to have       tients are similar to any other studies, except for
cognitive impairment or mild dementia, and to be           oneÐthat the physician should set priorities for
living alone, in poverty, or under institutional care.     which medications are critical to patients' health
They are also vulnerable to caregiver abuse, often         in a polypharmacy setting. The medication regi-
because of indifference, anger, or physical abuse          mens should be as simple as possible; the caregiver
triggered by the patients' behavior and difficulties       and patient should be educated about the name,
derived from their disease.                                dose, and reason for all medications. Patients
   Hearing or vision problems must be expected;            should be given simple instructions on cards, to-
bright light and large print, together with honest         gether with suggestions on how to remind them-
and simple language, much used for eliciting the           selvesÐ`tick-off cards on fridge', `diary notes' on
informed consent. Research subjects, whether eld-          bathroom mirror for morning dose, or on pantry
erly or not, should be able to understand the              door `with food', etc. Patients and their caregivers
informed consent process, feel free to refuse or to        should be given educational pamphlets about their
withdraw from the study without reprisal, and              diseases. They should be encouraged to ask ques-
understand the uncertain outcomes of the new               tions or report possible adverse events or strange
drug, the use of placebo and the random allocation         feelings. Patients should be asked to repeat back
of treatment.                                              instructions. Lastly, there are telephone call ser-
   The most vulnerable elderly population is found         vices which will call and remind patients to take
in nursing homes or mental institutions and                the medicine, or help organize cabs or transport for
frequently comprises persons of diminished or              follow-up visits, either to the laboratories for
                                   DRUG RESEARCH IN OLDER PATIENTS                                                  163

blood work, etc., or to the investigator appoint-       (ICD) code and charts are available to the investi-
ments.                                                  gators.
                                                           For large studies, mass-mailing to registered
                                                        voters, members of organized groups such as
          Screening and Recruitment                     AARP, or members of a disease association can
                                                        be helpful, with 7±12% response rate (McDerman
The Chaponis (1998) survey of 19 USA-based com-         and Bradford 1982). Use of media campaigns can
panies reported also that 32% reported difficulty       result in up to 11% of first protocol visits (Levenk-
in finding suitable investigative sites for geriatric   row and Farquhar 1982). These need at least 3±6
patients. In addition, those respondents involved in    months of planning for resources to respond to the
Phase IV outcomes, quality-of-life and pharmacoe-       initial wave of inquiries. The approach can be a
conomics studies, etc., said that the lack of `in-      newspaper article and advertisements in regional
company' geriatric expertise and resources was a        papers, TV and radio. Appeals to community phys-
barrier. Locating suitable investigative centers for    icians for referrals are usually disappointing, pos-
geriatric studies is only part of the solution and      sibly caused by the physician believing that he/she
works well for the smaller elderly experience stud-     will lose a paying patient to a research clinic.
ies. Nonetheless, in clinical studies undertaken for
specific diseases in aging, much larger numbers of
patients must be enrolled.                                                   CONCLUSION
   Even the large resources of the NIH can be
strained. The Systolic Hypertension in the Elderly      The growth of the aging population, regulatory
Person (SHEP) investigation recruited 4736 pa-          overview and increased business opportunities will
tients aged 60±96 years (average 72). The patient       ensure the growth of clinical research in the elderly.
screening and selection was organized from 16 sites     Recent reports of the high level of seniors' adverse
but took 31 months to complete, which had ini-          events, many leading to deaths, both in and outside
tially been projected to be 24 months. Nearly           hospitals, will force more monitoring systems for
450 000 patients were screened (SHEP Cooperative        medications. Soon, plastic medicine card chips with
Research Group 1991).                                   imprinted medication recorded by the pharmacist
   Hall (1993) reported on 15 cardiovascular stud-      will be required by third-party insurers. This would
ies funded by the National Heart and Lung Blood         ensure that all current concurrent medications are
Institute (NHLBI) over 10 years. All overran their      captured.
projected recruitment times by an average of 27%.          There is a shortage of geriatric specialists, which
Over-optimistic projections are the norm, and this      will take time to be corrected if the 600 drugs under
norm has been called `Lasagna's Law' (Spilker and       development are to be adequately researched. The
Cramer, 1972). For pharmaceutical clinical phys-        rapid growth of sheltered self-care communal hous-
icians and their staff, similar overruns are not ex-    ing for active seniors, which guarantee health care
cused by management, and raise the temptation           up to terminal status, illustrates that seniors wish to
to `move the target' by closing recruitment at a        stay out of nursing homes. Their expectation of the
lower level. This solution compromises the statis-      pharmaceutical industry is that it should provide
tical robustness of the study; both the problem         them with medications which allow for an active
and this solution are career busters. Better to         old age. The industry has heard.
project realistically and plan recruitment and fall-
back strategies. Hall (1993) also varied the recruit-
ment strategies used; the most successful was
community screening. This can be done through                                  REFERENCES
appeals to senior centers, churches, shopping
                                                        Abrams WB (1993) Food and Drug Administration Guideline
centers and major industrial sites (Melish 1982).        for the study of drugs in elderly patients: an industry perspec-
Medical chart review is also productive if the con-      tive. In Wenger NK (ed.), Inclusion of Elderly Individuals in
dition has a International Classification of Disease     Clinical trials. Marian Mennel Dow: 213±17.
Blackwell B (1979) The drug regimen and treatment complica-        Melish JS (1982) Recruitment by community screenings. 66
  tions. In: Haynes RB, Taylor DN, Sackett DL (eds), Compli-         Circulation IV: (suppl IV) 20±3.
  ance in Health Care John Hopkins University Press; 144±56.       Mossinghoff GJ (1995) Survey of new drug development in the
Chaponis R (1998) Geriatic-based research in the pharmaceut-         elderly. (under Publications).
  ical industry. Private survey (personal correspondence).         Muenning P Pallin D, Sel RC, Chan MS (1999) The cost of
Conn VS (1992) Self-management of over-the-counter medica-           effectiveness of strategies for the treatment of intestinal para-
  tions by older adults Publ Health Nursing Mar 9(1): 22±8.          sites in immigrants. N Eng J Med 340 (10): 773±9.
Denham MJ Barnet NC (1998) Drug therapy and the older              National Center for Health Statistics (1996) US Department of
  person; the role of the pharmacist. Drug Safety 19(4): 243±50.     Health and Human Services (data from 1996)
Everitt DE, Avorn J (1986) Drug prescribing for the elderly.       PhRMA (1999) Survey of new drug developments in the elderly.
  Arch Int Med 146: 2393±6.                                          Web (under publications).
Fed Reg (1994) Studies in support of special populatons: geriat-   Pitt B, Zannad F, Remme WJ et al (1999). The effects of spir-
  ric Fed Reg August 2, 1994, 59 FR: 390±398.                        onolactone on morbidity and mortality in patients with severe
Fed Reg (1997) Specific requirements of content and format of        heart failure. Engl J Med 2341(10): 709±17.
  labeling for huma prescription drugs: addition of `Geriatric     Satcher D (1999) Global health at the cross-roads: Surgeon
  Use' subsection (1997) Fed Reg 62 (166), August 45313±26.          General's report to the 50th World Assembly. J Am Med
Fed Reg (1999) Guidelines for the study of drugs likely to be        Assoc 281: 942±3.
  used in the elderly Fed Reg March 55 FR: 7777.                   SHEP Co-operative Research Group (1991) Prevention of
Gately MS (1968) To be taken as directed. J R Coll Geriat Pract      stroke by hypertensive drugs treatment in old persons with
  16: 39±44.                                                         isolated systolic hypertension: final results of the Systolic
Hall WD (1993) Screening and recruitment of elderly partici-         Hypertensio in the Elderly Program (SHEP). J Am Med
  pants into large-scale cardiovascular studies. In: Wenger NK       Assoc 265; 3255±64.
  (ed.), Inclusion of Elderly Individuals into Clinical Trials.    Small J (1998) An economic evolution of donepezil in the treat-
  Marian Mennel Dow; 67±71.                                          ment of Alzheimer's Disease. Clin Ther 20(4): 838±50.
Lassila HC, Stoehr GP, Ganguli M, et al (1996) Use of pre-         Spilker B, Cramer JA (eds) (1972) Patient Recruitment in Clin-
  scription medications in an elderly rural population. Ann          ical Trials. Rowen: New York.
  Pharmaco-other 30 (6): 589±95.                                   Tamblyn R (1996) Medical use in seniors: challenges and solu-
Levenkrow JC, Farquhar JW (1982) Recruitment using mass              tions. Therapy 51(3): 269±82.
  media strategies. Circulation IV: (suppl IV) 32±6.               United Nations International Population Division (1996).
McDermon M, Bradford RH (1982) Recruitment by use of mass          US Bureau of the Census (1996) current population report
  mailings. Circulation 66(6 pt 2): 27±31.                           series.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

       Drug Development Research in Women
                                               Lionel D. Edwards
                                      Basking Ridge, NJ, Carlsbad, CA, USA

                  BACKGROUND                                       Drug costs have risen slowly compared to other
                                                                health costs, when adjusted for inflation. When
The pharmaceutical industry is in the business of               compared to other health costs, in 1965 the drug/
developing, manufacturing and selling drugs, vac-               device cost was less than a dime per health dollar
cines, and devices. Although basic research has                 and in 1999 is less than a nickel (Health Care
become more important in recent years, it is not                Financing Administration   ). Drug cost is,
the primary aim of industry. However, increasingly              and must remain, one of the most affordable
and usually dictated by opportunity, industry is                aspects of treatment. A large component of drug
investing in a highly targeted fashion in some                  development cost is caused by regulatory needs to
aspects of basic research, but the development of               test for drug safety and efficacy, both for the USA
a product is always to the fore.                                and foreign agencies. Clearly, the cost of any add-
   This thrust, however, need not exclude the                   itional regulation imposed on top of the current
gathering of basic data, which may prove invaluable             burden will also be directly reflected in the eventual
to the research process. Regrettably, these data were           cost to the consumer.
frequently inaccessible, in some instances owing to                Women comprise 51% of the population of most
the needs of confidentiality, product protection, or            nations; in Western countries, 54% of women are of
even legal concerns, but by far the greatest reason is          child-bearing potential (15±49 years). Women ac-
that such data are regarded as a by-product, almost             count for 57% of physician visits (National Disease
`waste data', for they are not part of the mainstream           and Therapeutic Index   ) (FDA, 1986). In the
of product development. Such data are recorded                  age group 20±39 years, women were found to be the
but rarely utilized, frequently residing in notebooks,          biggest users of anti-infectives, especially ampicillin
case records, mainframe databanks, statistical                  and amoxicillin; antidepressants are prescribed
reports, or data tabulations in the back of appendi-            twice as often to women as to men (Stewart,
ces of regulatory submissions.                                  1998); and of some concern was that tetracycline,
   So it is with gender data: it is collected, analyzed         a known teratogen, was the eighth most prescribed
and tabulated by each study and by each drug, but               drug in this group most likely to bear children
data on drugs of the same class and between each                (FDA, 1986).
government agency handling multiple applications                   As major users, it might be postulated that
are virtually inaccessible. Mining this data requires           women, including those of child-bearing age,
more creative solutions than `regulations'. This is             should be the group on which Phase I and Phase
now happening.                                                  II dosing (early efficacy and safety) should be
   It has been estimated that the cost of developing            based. Why is this not so? Critics of the industry,
a new medicine is now $350±550 million (Arling-                 and indeed of the wider research process, claim that
ton, 2000). This estimate mostly comprises costs in             it is entrenched discrimination by males, which is
development, but includes the loss of other revenue             disguised as `concern and gallantry'. Critics also
if the development money had instead been                       point out that both medicine and research are dom-
invested cumulatively. These costs are passed dir-              inated by males, who place research into women's
ectly on to the consumer.                                       diseases on the back burner of their male priorities

and only see data, even on women, from a man's           (Edwards, 1991). This is unlikely to be due to lack
point of view. They point to a report by Coale           of compliance, as women are generally more reli-
(1991) on the `missing 100 million women' in Asia        able than men, although compliance does fall off to
and the Indian subcontinent, whom are speculated         67% over a few weeks for both genders (Cramer et
not to exist becaues of abortion and medical and         al, 1990). This does not exclude self-adjustment of
nutritional neglect. They also point to the misuse of    dose by female patients, a phenomenon seen in
science (ultrasound or amniocentesis) for sex deter-     both sexes and probably much more common
mination.                                                than reported.
   While these are extreme examples of societal atti-       It has also been claimed (because gender data are
tudes, it is true that women have been excluded from     rarely mentioned in clinical studies, papers or
many large, well-published studies, such as the          reports) that gender differences are not sought.
Physicians' Health Study of aspirin in cardiovascu-      This presupposes that data are neither collected
lar disease (Hennekens, 1989). It is also true that      nor examined. In fact, the opposite is much more
many early studies of drugs in Phases I and II were      likely: 94% of surveyed pharmaceutical firms were
conducted in healthy white males 18±40 years old         found to collect gender data in their studies
and the results then extrapolated to women in Phase      (Edwards, 1991). The reality is that findings of no
III studies, primarily aimed at expanded efficacy        differences are rarely reported, but sometimes this
and safety. Only recently, Paul Williams (1996) con-     finding may just be a function of small sample size
firmed that exercise raised HDL cholesterol in           for each individual study or the small degree of
women, many years later than reported in men. It         difference to be found. It must also be recognized
is, however, in most cases, grossly naive to attribute   that many drugs were introduced into medicine
this to deliberate `male discrimination' to exclude      prior to the current modern-day comprehensive
research on women.                                       testing programs. Nonetheless, after many years
   It is also frequently mentioned that fear of em-      and millions of prescriptions, it is of reassurance
bryonic malformation, whether or not drug-               that few have shown significant clinically import-
related, and subsequent litigation is the major de-      ant gender-related differences.
termining factor for exclusion of females from
therapeutic and basic research projects. This overly
simple explanation covers up other difficulties,             Differences in Disease Presentations
such as methodology, lack of relevant baseline in-
formation, and biochemical variables, both hormo-        A report from the National, Heart, Lung and Blood
nal and gender-related. It also ignores the use of       Institute (NHLBI, 1996), showed that the age and
information derived from other groups of women,          incidence (1988±1993) of onset of heart disease
those of no child-bearing potential, sterile or post-    between genders were different; 24% of the 65±74
menopausal, the elderly, or children just entering       year-old males compared to about 18% of females
puberty, where the risk of fetal exposure is non-        in the same age group. This incidence rose in both
existent or minimal.                                     genders at 75±84 years to about 28% males and
                                                         30% females.
                                                            Not only do women develop heart disease later,
                 THE DILEMMAS                            but they also present differently. The signature
                                                         symptom of a heart attack, severe chest pain, is
Do women respond to medications differently to           often absent in women, and pain in the upper
men? If so, in what ways and how frequently are          back or neck, or breathlessness and nausea, may
these changes clinically meaningful? Review of the       present either as a single symptom or as multiple
literature shows some examples of differences be-        symptoms. The American Heart Association states
tween the sexes in drug handling, particularly with      that 44% of women are likely to die in the first year
certain classes of drugs. These will be dealt with       of their heart attack, compared to 27% of men.
later, but it is important to bear in mind that,            It is not surprising that heart attack and angina
despite some detectable differences, usually no          are misdiagnosed more commonly in women than
therapeutically significant differences are seen         men during emergency room visits. The range be-
                                DRUG DEVELOPMENT RESEARCH IN WOMEN                                       167

tween hospitals of misdiagnosis was 0±11%, with                      THE PHANTOM FETUS
an average 2.3% for angina and 2.1% for heart
attacks. The diagnosis was missed in 7% of                              Teratogenic Issues
women under 55 years (Pope, 2000).
                                                         The term `phantom fetus' has been used to describe
                                                         the current apprehension regarding the use of
            What's Representative?                       drugs in women of child-bearing potential. This
                                                         apprehension has dominated industrial, and insti-
An additional dilemma is, what population is `rep-       tutional, and private research. The thalidomide
resentative' for female dose and efficacy determin-      tragedy of the 1960sÐthe 10 000 or so deformed
ation? Women of child-bearing potential (54%)?           children now grown to adultsÐcontinue to haunt
These will have possible hormonal cycling changes        us. It must be recognized that, despite careful
and those on contraceptive hormones will have            animal testing, the full potential for teratogenic
even greater changes, added to a possible basic          activity of any drugs in humans will only come to
gender difference, either amplifying or even sup-        light once the drug is in the marketplace, and then
pressing effects.                                        only when sufficient multiple exposures have oc-
   The needs of women aged 66 years or over are          curred in pregnant patients and their fetuses. It is
already represented in regulatory drug testing           extremely unlikely that deliberate drug testing in
guidelines in the elderly Federal Register (1990),       pregnant women will ever become routine. How-
but women 50±65 years old also can lay claim to          ever, in special circumstances, such as HIV-infected
special consideration, given the special problems        pregnant women, it is justified to include them in
associated with combined hormonal loss and age           appropriate clinical studies. Current predictive
changes (e.g. osteoporosis, loss of possible cardiac     animal screening cannot give complete assurance
estrogen protection and changes in body fat com-         that the potential for teratogenicity will be un-
position and its distribution). Pregnant women,          covered in all cases. It must be remembered that
already isolated from drug development by fear of        the then-current 1956 screens did not discover the
legal tort laws and, indeed, by their physicians'        teratogenicity of thalidomide, nor the 16-year
reluctance to even prescribe in early pregnancy,         delayed hyperplasia and neoplasia effects on the
can also stake a claim to require additional studies.    cervix and uterus of female adolescents exposed
Finally, when studying females of child-bearing          to stilbestrol (given to prevent miscarriages during
potential, should we include patients on oral            their mothers' pregnancies).
contraceptives, with their large levels of regulated        Both historically and currently, the major deter-
fluctuating but synthetic hormones, or rely on           mination of teratogenicity is made from findings
females not taking oral contraceptives? The latter       from animal screening; many agents have been
option will increase the risk of potential fetal         eliminated from further development, and only
exposure.                                                rarely does teratogenicity become uncovered in
   It must now be apparent that the female popula-       the marketplace. Nonetheless, it requires large
tion (51%) contains many potential subgroups,            numbers of exposures before the more subtle
none truly `representative', for all have major          embryotoxic or teratogenic effects are found, as
physiological differences from each other. For in-       was demonstrated most recently by the ACE in-
dustry to study all groups would be impractical,         hibitors, which had passed all the screens. Indeed,
uneconomical and would gravely slow the drug             these events may never be exposed. How could this
development process and compromise the number            be? One must take into account the `background
of agents placed into development. To include all        noise' level, the so-called `natural' incidence of
groups within one all-encompassing study, unless         cogenital abnormalities. By far the commonest is
extremely large, offends a basic research nos-           Down's syndrome, whose incidence is known to
trumÐi.e. `stabilize, reduce or remove all the vari-     increase with the age of the mother, although
ables except the one to be measured', or the signals     nearly all other abnormalities appear not to in-
many be lost in the static. This is especially true in   crease with maternal age, according to a recent
Phase II studies.                                        report (Wilson, 1973). Thus, a higher incidence of

`typical' drug-induced teratogenic effects serve as            The Potential for Pregnancy While
an early alert. The commonest abnormalities most                        on a Trial Drug
frequently associated with drug exposure in the
first trimester are neural tube defects, cardiac and     What is the risk of pregnancy occurring in a study
renal anomalies, shortening of limbs and digits,         participant while a new drug is being developed?
and failure of closure of the palate and upper lip.      The author is not aware of any published figures,
More subtle changes associated with exposure to          but from the author's experience in industry and
drugs occur in the third trimester, with hearing and     from questions to colleagues, pregnancy does occur
eye abnormalities predominating (Wilson, 1973).          during drug development, even in those patients
Any such determinations require many, many               apparently taking adequate contraceptive precau-
thousands of exposures before they become ap-            tions. A typical NDA database for most drugs will
parent.                                                  involve between 2000±4000 patients, of which per-
   However, many millions of women become preg-          haps one-third are female and exposed to study
nant before being aware of their pregnancy and           medication. It is not surprising, therefore, that
have been exposed to environmental chemicals             given an average failure rate of the contraceptive
(most of which have never been tested), as well as       pill of 2%, or even with the most stringent compli-
OTC drugs and prescription drugs. Also, a number         ance, a failure fate of 0.5/100 women years will
of embryos are spontaneously aborted and a delay         result in occasional pregnancy (Trussell et al,
to the menstrual period of perhaps two or three          1990). Other methods, such as the diaphragm,
weeks passes unremarked or sometimes unnoticed           condoms and IUDs, can carry even higher failure
in a background of a national miscarriage rate of        rates, depending on whether `usual' or `perfect
1 in 3 pregnancies (Yoder, 1984). Teratologists          compliance' calculation of 18±6%, 12±2% and 3±
have concluded that there is a threshold dose for        0.5%, respectively, are used (Trussell et al, 1990). If
any drug before it shows potential teratogenicity        we assume an average NDA database of 4000 pa-
(in other words, enough must be given), and the          tients, one-third or more female, it is likely that half
effect tends to increase with the duration of expos-     of these will be females of child-bearing potential
ure, with higher concentrations in the plasma or         (the other half being post-menopausal or elderly).
tissues, and with the timing of the developing fetal     Thus, approximately 660 females of child-bearing
tissues and organs (Wilson 1973). In the first 7±8       potential may be exposed to the drug, the compara-
days, the embryo is refractory to any teratogenic        tor, or a placebo. In the best circumstances of
effect but is most susceptible 20±55 days after con-     perfect contraceptive compliance, in a 1-year ex-
ception. Of some reassurance is that most drugs          posure and at a 0.5% failure fate, 3.3 fetuses are
prescribed to women of child-bearing age are anti-       likely to be exposed. With a `typical compliance' of
biotics and tend to be for relatively short durations.   the contraceptive pill, a 3% failure rate would leave
But the tetracyclines and anti-epileptic drugs are       about 19 fetuses exposed to experimental entities,
known to have effects on the developing fetus and        one-third of which would be lost due to spontan-
are frequently prescribed to women (Stewart,             eous miscarriage.
1998).                                                      Few patients would be exposed for a full year,
   It is an irony that the normal tenet of US and        but more typically probably only to between 2
UK law that an individual is `innocent until proven      weeks to 3 months of study medication. Given all
guilty' does not apply to prescribed pharmaceutical      the above assumptions, between 0.8 and 5 early
products or devices. They must be proven safe and        embryos will be exposed in a full drug development
efficacious before they are approved; in other           program. From the author's personal experience of
words, they must be proven to be innocent. Thus,         over 27 years in industry, an average of two chil-
it comes as no surprise that industry and other          dren are born exposed to a new chemical entity.
research groups tend to avoid the potential expos-       This is most likely to occur in Phase 3 studies,
ure of women of child-bearing age in the early           which have many more patients and are often of
clinical development of pharmaceuticals or devices,      longer duration. Currently, pharmaceutical firms,
for many experimental drugs (perhaps 9 out of 10         with the agreement of the FDA, follow up all pos-
tested in man) will never achieve the marketplace.       sible exposures until any resultant child is 12±14
                               DRUG DEVELOPMENT RESEARCH IN WOMEN                                         169

years of age, and a full medical examination (in-      stances; they may be the only drugs suitable for
cluding a full neurological work-up) is done at        some patients and, indeed, frequently can be life-
yearly intervals.                                      saving. Certainly status epilepticus is very injurious
                                                       to the fetus, often resulting in miscarriage or pre-
                                                       mature birth.
                                                          The incidence of neonatal abnormalities in
The Potential for Teratogenic Damage during            mothers taking anticonvulsant treatment is 70/
           Drug Study Programs                         1000 live births (Frederick, 1973). This is 2.4
                                                       times the `spontaneous rate' in the general popula-
As previously mentioned, the best sources for the      tion (29 abnormalities/1000 live births). Thus, even
actual figures for the above calculations resides      using a known `low-incidence' teratogen could
within the FDA but may, as I alluded, be inaccess-     cause 40 additional cases/1000 live births, but to
ible. In recent years, figures given by the Agency,    determine that accurately would require many
e.g. in elderly drug-testing studies, appear to have   thousands of female patient exposures to be detect-
been hand-tallied rather than garnered from            able against the `spontaneous' background inci-
composite computer access. However, the agency         dence.
is now involved in a large effort to `mine' data          So, back to the opening question. What is the
across therapeutic classes, some of which, with        likelihood of detecting low-incidence, drug-induced
meta-analysis, will provide data which individual      congenital effects in a drug development program?
drug programs never could, nor were designed to        With our presumed database of 4000 patients, only
show. In time, the ability to access data across       0.8±5 fetuses would be exposed to a background
drugs and across drug classes will grow as more        `spontaneous' risk of 2.9%. Each program could
firms put in computer-assisted NDAs (CANDAs)           carry a 1 in 33 to 1 in 6 chance of a single `spontan-
in appropriate and compatible programs and             eous' abnormality occurring. If the drug or proced-
formats.                                               ure should have low teratogenic activity (at the level
   What is the risk of a fetus being damaged during    of an anticonvulsant), this risk rises to 1 in 14 to
an `average' NDA drug development program?             1 in 2.5 that a child will be born with a congenital
Obviously small. Clearly, toxic but `life-saving'      abnormality in any drug development program.
treatment will carry a heavy embryotoxic risk;         Both `spontaneous' or drug-induced abnormalities
anticancer, anti-AIDS drugs, and fetal intrauterine    may occur, e.g. a neural tube defect. Thus, on a
surgical procedures are obvious examples, but          single-case basis, the abnormalities will be indistin-
the clear-cut risks involved are usually deemed        guishable for drug causality. This in turn can lead to
acceptable. A more subtle judgment call involves       litigation, and certainly to a reference in the package
the development of anti-epileptic drugs. Let's         label insert.
look at two examples. It has been estimated that          Wilson has estimated that both drugs and envir-
exposure of pregnant women to normal thera-            onmental chemical exposures only account for 2±
peutic doses of valproic acid may give rise to 1%      3% of developmental defects in man (Wilson,
fetal abnormality rate involving the neural tube       1972).Thus, a product-label reference of such an
(Lindhaut and Schmidt 1986)Ð10 times the               occurrence will be undeserved at least 97% of the
natural incidence. Many of these defects are           time, but also may be the first signal of a terato-
correctable with modern surgical techniques. Ex-       genic risk. It may now be appreciated why this
posure to phenobarbitone also has a reported           2±3% risk is termed the `phantom fetus' and also
higher incidence of cleft lip and palate defects       why the difficulty in disproving liability dominates
(Frederick, 1973): most are surgically correctible.    the mainstream concerns of research, regulatory
If used in combination, the incidence of anticon-      authorities, and industry alike. This `ghost risk'
vulsant teratogenic effects are increased (Lindhaut    creates `discrimination' against female patients in
et al 1984). Would either of these drugs be de-        drug research. This `ghost' must be exorcised and
veloped in today's litigious atmosphere? I doubt       contained; possible solutions will be discussed
it. But both drugs are valuable in many circum-        later.

INDUSTRY PRACTICE: FACTORS IN PHASE                     the potential to act as an active transport mechan-
    I AND EARLY PHASE II TESTING                        ism for drugs, pesticides and even environmental
                                                        chemicals to the unfertilized ovum. They may also
Medical journalist Paul Cotton (1990) asked, in a       alter the genetic make-up of either spermatozoa or
thought-provoking article, is there still too much      ovum. In addition, spermatozoa can be made slug-
extrapolation from data on middle-aged white            gish by calcium channel blockers, leading to male
men? Inspection of the demographics of recent           infertility while on medication. Hence, the Euro-
NDAs will give us numbers to debate; however,           pean guidelines call for male animal testing prior to
these data are not readily accessible. Most Phase I     start of Phase 2.
testing is still undertaken in healthy young males,        The blastocyst (early embryo) is relatively resist-
and even for Phase I testing of new contraceptives      ant to damage in the first 7 days, for up to 75%
hormonal for women. Why this occurs is multifa-         of cells can be destroyed before tissue differenti-
ceted.                                                  ation and the embryo can still survive. What might
                                                        happen if garden pesticides, or house builders' for-
                                                        maldehyde containing glue and chemicals, are
                                                        combined into the genetic material? If it is ever
      Timing of Mutagenicity Fertility and              confirmed, then we may have the inkling of what
            Teratogenicity Testing                      makes up the 65% of the `unknown' causes
                                                        of developmental defects mentioned by Wilson
The complete battery of tests with full histology       (1972). If it could be shown that the synthetic
and the development of a final report can take as       chemicals are incorporated into the blastocyst, the
long as 2 years. In general, only some of the muta-     field of male Phase I testing would be transformed,
genicity studies are completed, and perhaps 1±3         as would that of genetic counseling.
month reports of animal testing are available
when male Phase I dosing volunteer studies com-
mence. All animal studies do not commence at the                         Testing Facilities
same time but are usually sequential. Some, such as
postexposure weaning and subsequent second-gen-         Largely because early testing of drugs occurred in
eration drug effect studies will be time-consuming      males rather than females, for reasons discussed
and expensive. Often, if mutagenicity tests, e.g.       above, most commercial and hospital units devoted
Ames' test or mouse lymphoma test, are positive         to human pharmacology testing were set up to deal
(Ames test has 30% false-positive rate), then           with a unisex population. They ran one gender
females will be excluded until more data is col-        study at a time, usually male, in 1993. Sleeping
lected. Thus, only limited data are available prior     and bathroom facilities in the units' dormitory ac-
to the first human exposure. (For further reference     commodations did not provide for mixed gender
Fed Reg 1994, 1996).                                    groups. These were minor but not inexpensive at-
   Volunteer dose-ranging studies will, by design,      tentions but were quickly adopted following the
include high enough doses to provoke unpleasant         publication of the FDA Guidelines for the Study
adverse effects; also, information on `target organs'   and Evaluation of Gender Differences in the Clinical
(organs likely to be most affected or harmed) is        Evaluation of Drugs (Fed Reg, 1993).
usually predictable but unconfirmed at this point.
Generally, as a result of animal studies, it is
thought that the effect of drugs on reproductive           Standardizing for the Menstrual Cycle
function in males is less than that in females and              (Phase I and Early Phase II)
only affects the sperm viability or, rarely, the size
and function of the testicles, which is usually re-     Of much greater concern is the issue of standardiz-
versible. This is unduly optimistic, as one report by   ing the drug administration to the menstrual cycle.
Yazigi et al (1991) suggests that spermatozoa may       Women of child-bearing age do not all have cycles
not be immobilized or destroyed by cocaine, but         for the same length of days; variations of 24±36 day
may interact, and the spermatozoa themselves have       cycles are not unusual between and within the same
                               DRUG DEVELOPMENT RESEARCH IN WOMEN                                         171

women. Thus, unless controlled by oral contracep-       result, it is often hard to recruit older, more mature
tives (OC), women volunteers could not start and        women for these basic types of essential drug de-
finish in a study all together. Indeed, if OC were      velopment programs.
used to standardize cycles, the issue of how really
representative of all women of child-bearing age
this artificial hormone-boosted group might be
would be debatable. Evidence suggests that even                What Is a Representative Female
low-dose contraceptives can affect metabolism                      Population in Phase I?
(Abernathy and Greenblatt, 1981). The logistics
of running Phase I single-dose and multiple-dose        It has been stated that large numbers of mature
ranging studies while controlling for a natural men-    women are volunteering for the new lipid, heart
strual cycle are truly horrendous, both for the         risk, osteoporosis and arthritis studies, due to
Phase I testing units and for the volunteer. The        their concern that women have been represented
duration of any study would be extended by at           so poorly as subjects in the past. Phase I studies are
least 1 month (the time required for the last           of short duration (1±2 weeks), but usually require
patient's cycle to start), and each patient volunteer   confinement of the volunteers to the clinical unit
would have to be measured separately because of         for that time. Because of this time commitment, far
the different days of her cycle. A small but fre-       fewer mature women volunteer, due to career con-
quently argued point is timing. Which is the pre-       flicts or because they are often burdened unequally
ferred day in the cycle for single-dose studies? And    with family management. Those that do volunteer
for a multiple-dose study (usually only 10±14 days      are generally unattached young female students.
long), which segments of the cycle should be            Thus, most female volunteers may not be typical
covered? This may seem academic, but in those           of a `representative', mature, child-bearing popula-
clinically significant drug classes where womens'       tion (if this can ever be defined).
responses to drug handling are different to those          One alternative, a study design of stratification
of men because of biochemical hormone effects           by age and sex, would lead to inordinately long
(not just gender), then the timing of drug dosing       study recruitment times, because the last `cell'
and measurement would be critical.                      (group) always takes a disproportionately long
                                                        time to fill. The most obvious way out of the quan-
                                                        dary for Phase I testing would be to maintain a
     Too Many Young Volunteer Studies                   special cadre of `safe, standard' volunteers. How
                                                        `representative' these much-used `new-drug volun-
Many volunteer studies, especially at commercial,       teers' would become is debatable. For example,
academic and university clinical units, are fre-        studies in arthritic patients show that these `re-
quently young people of college age. Both males         tread' volunteer patients will differ in their toler-
and females will volunteer since financial remuner-     ance to pain and in their judgment of efficacy and
ation, and a free medical check-up and medical          severity of adverse events, when compared to drug-
care play their part in motivation. The young           study `naive' patients (Coles et al 1988). This
also have less career and family commitments inter-     `training effect' increases with multiple drug expos-
fering with their motivation. Time for studying,        ure.
reading and relaxation within an atmosphere of             By far the biggest issue of undertaking additional
camaraderie also contributes to the availability of     dosing Phase I studies on women is expense. Most
younger volunteers, who, because of their age, also     of these studies cost $100±250 thousands each. Al-
tend to be very healthy. It will readily be appreci-    together, single, multiple, and multiple-dose
ated that most drugs or devices are not unique or       ranging studies, with food effect studies and extra
life saving but hopefully an improvement on             staff costs, could add $1 million to development
existing agents, and indeed this applies to most        costs and very rarely show a difference which
basic research experiments. Nearly all drug studies     would prove clinically relevant. Indeed, the differ-
in Phase I are aimed at gathering data on a poten-      ence may not show up at all in Phase I or II gender-
tially safe and possibly efficacious dose range. As a   to-gender studies, due to other variables, e.g. small

numbers, estrogen-cycle levels and oral contracep-       in women after puberty. The composition of `good
tive levels and drug polymorphism.                       fat and bad fat' changes with age, both in increased
                                                         fat, increased bad fat and its relocation to the fat
                                                         around the heart. The quantity and distribution
      DRUG HANDLING DIFFERENCES                          differs between genders. This may have an effect
      BETWEEN MALES AND FEMALES                          on lipid-soluble drugs, regarding the level, the time
                                                         to achieve steady-state, and the time to eliminate
Due to space limitations this subchapter cannot          the drug and its metabolites from such fat storage
discuss the many reports of apparent gender differ-      depots.
ences of psychology, different anatomic brain loca-
tion of functions, skeletal build and muscle-to-fat
mass ratios which might have marginal impact                    Different Gastric Emptying Time
upon drug activity.
                                                         Some studies have shown that women demonstrate
                                                         greater duration in the gastric residence time of
          The Weight/Dose Problem                        medications, which is reflected in an increased lag
                                                         time of absorption, compared to men. This effect is
A casual appraisal of ideal weight-for-height tables     increased when medication is taken with food, even
for males and females (Metropolitan Life Insur-          when adjusted for the timing of the menstrual cycle
ance, 1999) shows clear differences between males        (Majaverian et al, 1987). This was consistent with
and females. The mythical `average' 70 kg (154 l)        other reports that men had faster emptying times
male would be 50 1000 in height and his female coun-     for both liquid and digestible solids than women
terpart 50 400 and weigh 130 lb. This is a 28% differ-   (Majaverian et al, 1988; Wright et al, 1983). The
ence in weight. This mythical male is often used to      length of time and variability of gastric emptying in
calculate dose ranges for `optimal' dose determin-       women was also reported by Notivol et al (1984) to
ations, around which Phase II and Phase III effi-        be altered in relation to the menstrual cycle and
cacy and safety studies evolve. Even more striking       was shortest at mid-cycle (MacDonald 1965; Booth
is the range of normal heights and weights, remem-       et al 1957).
bering that the same dose is usually prescribed to          These changes can affect the amount of drug in
individuals across the range. In males, this varies      the blood. Miaskiewicz et al (1982) showed that,
from 50 at 106 pounds to 60 800 at 226 pounds; in        after a single dose of sodium salicylate, absorption
females, it varies from 85 pounds at 40 900 to 185       was slower and achieved a lower level in women.
pounds at 60 500 ; yet all are ideal weights for the     This has also been shown for ibuprofen. The Tmax
respective heights. For both sexes this represents a     was observed to be more than 54 min in females,
46% differential in healthy weight while taking the      compared to a Tmax of 31.5 min in males. Majaver-
same dose of medication. Why should these great          ian even showed a delay of 9.5 h before absorption
disparities be tolerated by the research community,      occurred in one woman (Majaverian et al, 1987).
industry and agencies? Because most drugs workÐ          Sex differences in plasma salicylate albumin
even over these ranges. First, the majority of the       binding capacity have been reported (Miaskiewicz
population falls towards the middle of the height±       et al, 1982) and, for other agents (Allen and Green-
weight levels, rather than the extremes. Second,         blatt, 1981), g-globulin transport systems have
most drugs have a wide range where which they            been reported to be altered with the menstrual
exert therapeutic effect before efficacy levels off.     cycle.
Third, the level of unacceptable adverse events             Some effects on absorption can be subtle, such as
generally occurs at much higher doses than the           the greater absorption of alcohol in women due to
therapeutic level for most drugs (there are some         their reduced gastric mucosal and liver alcohol
notable exceptions, e.g. lithium, digitalis, warfarin    dehydrogenase activity compared to men. This
etc.).                                                   results in higher circulating levels of alcohol, in
   For lipophilic drugs, the composition of mass to      spite of body weight corrections (Frezza et al,
fat/total body water is a further variable, increasing   1990), with obvious implications. Odansetron, on
                                DRUG DEVELOPMENT RESEARCH IN WOMEN                                        173

the other hand, is more slowly metabolized by            first part of the cycle (M'Buyamba-Kabunga et al,
women and thus may be more effective.                    1985). Androgens transported on the b-globulin
                                                         and albumen fraction are influenced by estrogen,
                                                         which increases their binding. This effect is en-
        Metabolic Gender Differences                     hanced by the use of oral contraceptives (Clark et
                                                         al, 1971).
Propranolol is still one of the most frequently used        In animals, estrogen has been shown to influence
b blockers (National Prescription Audit, 1989), but      the effect of antidepressants on the brain. Wilson
Walle et al (1985) reported that women had higher        showed that estradiol increased the binding of imi-
plasma levels of propranolol than men following          pramine to the uptake of serotonin at membrane
single oral dosing and, in an additional study,          sites. Estrone had no effect, but the addition of
showed that on multiple dosing, propranolol              progesterone to low doses of estrogen increased
steady-state (trough) plasma levels were 80%             this effect. In all, the greatest effect seen was
higher than in men (Walle et al, 1985). This is          about a 20% enhancement of imipramine binding
probably because propranolol is metabolized              (Wilson, 1986).
through three pathways, but in women, the P450              For low-therapeutic/toxic drugs such as lithium,
cytochrome oxidation pathways are less effective         this might prove to be an explanation of the reduc-
than in men (Walle et al 1985).                          tion in efficacy seen at the end of the menstrual
   Methaqualone metabolism has been shown to be          cycle, when these hormone levels fall (Conrad and
significantly increased at the time of ovulation (day    Hamilton 1986). It might also explain the reduction
15), almost double that of day 1, and this was           in efficacy of other central nervous system drugs,
reflected in an area under the curve (AUC) reduced       such as antiepileptics (Shavit et al, 1984; Roscieze-
by half on day 15. It is of interest that men, used as   westa et al, 1986) and antimigraine medications,
a control, only sustained levels at the level of day     seen with the fluctuation of the menstrual cycle
1 in women (Wilson et al 1982).                          (Gengo et al, 1984).
   Differences between males and females in the             Young women appear to be the group most at risk
amount of free drug found in plasma, and of pro-         of developing extrapyramidal reactions when taking
tein binding, have been reported for diazepam            the antinausea drug metoclopramide. This appears
(Greenblatt et al, 1979; Abel et al, 1979) and for       to be strongly age- and gender-related (Simpson et
imipramine (Kristensen 1983). In the latter in-          al, 1987). Another age/gender-related effect is seen
stance, a direct correlation was found with differ-      in older women who have become newly postme-
ences in lipoprotein and orosomucoid protein             nopausal and who are still taking antipsychotic
(1-a-acid glycoprotein) fractions (Greenblatt et al,     medications, because the symptoms of tardative
1980). In women, oxazepam has been found to be           dyskinesia may appear or even worsen (Smith
eliminated at a slower rate, about (10%), and for        and Baldossarini, 1973). This is perhaps another
temazepam about (25%) (Diroll et al, 1981). Chlor-       example of the loss of estrogen protection.
diazepoxide was also found to be less bound to              Many of the examples quoted involve central
protein and this was even further reduced if             nervous system drugs. This is very important,
women were also on estrogen oral contraceptives          for gender-related prescription usage is heavily
(Roberts et al, 1979). Free lignocaine levels in         weighted in this area towards women. The FDA
women were 11% higher in estrogen oral contra-           1985 drug utilization report showed that for benzo-
ceptive users and 85% of this effect was due to the      diazepines, the increased usage in women outnum-
reduction of the orosomucoid protein fraction            bers men by 2:1 (339 vs. 171 prescriptions/1000
(Routledge et al, 1981).                                 women and men, respectively). Twice as many
   Circulating hormones, such as aldosterone and         women are treated for depression and anxiety neur-
renin have long been known to fluctuate with the         osis than men, first described by Raskin (1974), and
menstrual luteal phase. If an amenorrheic cycle          confirmed by Weissman and Klerman (1977). It is
occurs, these changes are not seen (Michelakis           by no means certain that this is solely due to bio-
et al, 1975). If oral contraceptives are given, then     chemical differences, for women are more likely to
an increase of these hormones is also seen in the        seek help than men. Of importance from the prior

discussion is that, if women are the greatest users of    Health (NIH) Guide (1989) both recommended
these medications, should not study recruitment           that biomedical and behavioral research should
members be biased in their favor? However, some           be expanded to ensure emphasis on conditions
of the psychotropic CNS drugs also have animal            unique to, or most prevalent in, women of all age
dataÐand a few, even some human dataÐsuggest-             groups: `in addition, studies are needed to study the
ing an increased teratogenic potential (Physician's       metabolism and disposition of drugs and alcohol
Derk Reference, 1991; Jefferson et al, 1987). There       by age and gender'. The National Institute for
is no consistent evidence of class teratogenicity (Elia   Drug and Alcohol Abuse (NIDAA) (1990) policy
et al, 1987), but there is a high association of frac-    provides detailed, almost affirmative-action in-
tured hips with the use of psychotropic medicines,        structions for the inclusion of women and minor-
even when corrected for women's greater age-re-           ities into study designs, according to their
lated hip fracture rate (Ray et al, 1987). One of the     prevalence in the diseases being studied.
commonest causes of the elderly being admitted to            Since 1988, the FDA has requested tabulations
institutional care is urinary incontinence. Women         of gender, age, and racial distributions in NDA
have been found to be more susceptible than men to        submissions. Many of their senior officials, e.g.
medications that can cause incontinence to occur          Drs. Peck and Temple, had forcefully stated that
(Diokuo et al, 1986).                                     women should be included in drug development
                                                          studies. Indeed, the 1977 guideline, General Consid-
                                                          eration for the Clinical Evaluation of Drugs, in-
           Adverse Event Differences                      cluded a policy for the inclusion of women
                                                          of child-bearing potential in clinical trials but
One of the most striking differences between male         excluded them, in general, from Phase I and early
and female responses to drugs is the finding              Phase II studies, with exceptions for life-saving or
reported by Martin et al (1998) in 513 608 patients       life-prolonging treatments. Child-bearing potential
with serious adverse events, which occurred in            was strictly defined as `any woman capable of be-
43.2% males and 55.7% females when adjusted for           coming pregnant', including women using revers-
age. In women of all ages, Tran et al (1998) also         ible contraceptive precautions and those with
reported that, in findings from records of 2367           vasectomized partners.
patients, female patients were at twice greater risk         The FDA issued new guidelines in 1993 (Fed Reg
of adverse reactions than males. More than one            1993), perhaps spurred by its own findings in 1989,
agent was reported to be responsible in 50% of            and confirmed by the General Accounting Office
female patients vs. 33.1% of all male patients.           (GAO), that in only 50% of submissions were
Drug doses in both genders most likely to cause           gender analysis discussed in NDA submissions.
an adverse event were anti-infectives (60.4%) and         Temple (1992) reported that two FDA surveys
nervous system agents (21.5%) (Martin et al, 1998).       demonstrated that women were included routinely
The commonest events were skin-related reactions          and in proportion to the presence in the treatment
(49%). It is possible that bare arms and exposed          population, and young women in large numbers
legs in women may cause more phototoxic reac-             (Bush et al, 1993). Not recorded were his conclud-
tions than in men; nonetheless, this cannot be said       ing remarks, in which he said many NDAs did not
of nervous system agents. Clearly, these two classes      adequately discuss gender difference, which would
of agents need special gender exploration in clinical     be addressed in the new amended guideline. The
development.                                              FDA, in its discourse in the 1993 guidelines, Re-
                                                          vised Policy on Inclusion of Women of Childbearing
                                                          Potential in Clinical Trials, mentions that it was
GOVERNMENT AGENCY AND INDUSTRY                            swayed by a legal precedent. In 1991 the US Su-
   ACTIONS ON GENDER-RELATED                              preme Court found on behalf of the plaintiff
            RESEARCH                                      workers union that their pregnant members had
                                                          been unfairly excluded from jobs by the Johnson
The Public Health Service Task Force on Women's           Control Company, because the working conditions
Health Issues (1985) and the National Institutes of       exposed their fetuses to potential risk. The court
                                 DRUG DEVELOPMENT RESEARCH IN WOMEN                                         175

wrote: `Welfare of future children should be left to      tial data indicating a potential risk with a drug
the parents . . . rather than to employers who hire       class-related compound; and, finally, an FDA
them'. While not quite the same circumstances, the        reviewer's individual comfort level with `high-risk
FDA were of the mind that this opinion would also         population exposure'. Such an event has now
apply to pregnant (informed) women, giving them           become rarer.
the right to enter drug trials irrespective of phase of
   The FDA revised guidelines on this and ethnic                Pharmaceutical Industry Practice
differences which appeared in July 1993 in the
Federal Register, in essence abolished the prior          In July 1991, a survey was completed by this author
ban on women of childbearing age from Phase I             for the Pharmaceutical Manufacturers Association
and Phase II studies, and stipulated additional           (PMA), Special Populations Committee on the
topics, including the embryotoxic and teratogenic         current practice of the industry in handling gender
risk potential, to be covered in the patients'            and minority data (Edwards, 1991). Vice-Presi-
informed consent.                                         dents of headquarters, clinical and regulatory
   Earlier, the NIH had issued its own guidelines to      affairs were contacted at 46 companies; 33 com-
its staff, grant applicants, and academic centers it      panies responded (nearly all the major companies).
supported. It called for all research on human sub-       All 33 responding companies collect gender-related
jects concerning drugs, devices, epidemiology, non-       data on the participant patients in clinical studies.
drug device studies and treatment outcomes, to            Over three-quarters of the companies reported
include both genders and minority representatives         that they deliberately recruit `representative'
whenever possible. In Phase III studies, `women           numbers of women. It should be noted that the
and minorities and their subpopulations in suffi-         term `representative' has not been defined by the
cient numbers should be included, such that valid         FDA or by industry. However, only 10 companies
analyses of differences can be accomplished'. It          (30%) frequently or usually collected data on men-
stipulated that `cost was not an acceptable reason        strual cycle; 56% replied that the FDA at some time
for exclusion, and that programs and support for          or other had requested the inclusion of women in
outreach efforts to recruit these groups be under-        trials. When women of child-bearing potential were
taken'. (NIH, 1986). Failure to ensure adequate           included in protocol proposals, 21% of the re-
effort to implement could be reason for grant re-         spondents said that the FDA never disagreed, but
jection or loss of financial support.                     79% had experience of some FDA reviewers at one
   To amplify the female view both the FDA and            time or another excluding women of child-bearing
NIH during the last decade have appointed women           potential. When excluded, this was usually in the
to significant roles. Dr Bernadette Healy headed          Phase I and Phase II trials, 58% and 45%, respect-
the NIH and created the Office of Research in             ively, correspondents reported.
Women's Health; Dr Henny led the FDA until                   While this survey was qualitative rather than
2001 and within the FDA, Dr Janet Woodcock                quantitative, the results should not be dismissed
and Dr Kathy Zoon were appointed to head                  lightly; because the survey was confidential, no
CDER (drugs), and CBER (biologics), respect-              respondents or their firms were exposed to open
ively, two of the largest centers perhaps partly in       criticism. Because of their experience and senior
response to an article by LaRosa and Pinn (1993),         positions, respondents had reviewed many differ-
both women bemoaning exclusion of women in                ent drugs and NDA applications. The survey re-
decisions of research.                                    plies were therefore likely to be reliable and provide
   The industry is now encouraged by the FDA to           a good approximation of the then-current industry
include women earlier in the clinical development         gender practices and the frequency of clinically
program, but there are also still good reasons why        meaningful differences.
the FDA might deny inclusion of women of child-              When gender differences in safety or efficacy
bearing potentialÐinsufficient toxicology data; a         were found to be clinically significant, most re-
disagreement over the interpretation of such data;        spondent companies (94%) opted to put the
agency knowledge of another company's confiden-           data in the product label, the Physicians' Desk

Reference and the product literature (72%), and to        devices. These could be similar to those now under-
publish in the medical journals (69%). Presumably,        taken in the elderly. First, a single-dose study
the two companies that did not amend their labels         should be undertaken. If important differences
acted thus because the products were only intended        are found compared to men, a multiple-dose
for one-gender use. By December 1999, there were          study ought to be undertaken, and then a shorter
348 medicines in development for diseases only in         duration efficacy and safety study in women. Such
women or where women are disproportionately               studies can be conducted later, perhaps concur-
affected (Holdin 2000). Not only has industry             rently with Phase III of the development program.
stepped up its research efforts, but many large              What do we mean by `a reasonable number'?
firms have units devoted to women's health care.          `Reasonable' is that number which would be
   Finally, correspondents were asked how fre-            expected to show a significant gender clinical dif-
quently gender differences were found; 73% said           ference if a real difference is present, and probably
`occasionally', 3% said `frequently' and the rest         will only apply to efficacy and adverse events 5% or
said `never'. Of those who saw differences, only          larger, because a difference in low-incidence ad-
one-third found these differences to be clinically        verse events will not show up until the drug is on
significant 5% of the time, while 17% of respond-         the market. This would mean at least 300 women
ents said that significant differences occurred 10%       exposed to the new drug. The number of patients
of the time. This was more than expected, and             should be based on what is judged to be a clinically
provides further justification for gender testing.        significant percentage loss or enhancement of effi-
                                                          cacy, e.g. 30%, dependent on the disease or symp-

This author must stress that the opinions and the             Representative Population of Women
suggestions that follow are personal, based on 27
years in industry, over Phase I±IV study experi-          This can be based on the incidence of disease pro-
ence, with five large international pharmaceutical        portional to gender distribution and can be studied
firms.                                                    when drug development and toxicity are well-
                                                          enough advanced, usually by Phase III. Women
                                                          of child-bearing age must be represented if the
      Women's Inclusion as Drug Research                  disease is prevalent in the age group 15±50 years.
                   Subjects                               Indeed, diseases such as endometriosis can only be
                                                          studied in such a population, whereas drugs to treat
Women should be and, indeed, are included into            urinary incontinence would be better undertaken in
new drug and device development programs when             older patients.
not specifically excluded due to male-only disease           In some diseases, such as hypertension, where
or existing pregnancy. If it is predictable that a        both sexes are similarly affected, balanced numbers
drug or device will be used in women (though              of male and female patients in Phase III would not
they may not be the majority users), then a `reason-      seem out of place, although many investigators are
able number' should be included into Phase II and         finding recruitment of sufficient numbers of female
Phase III studies. If the disease occurs more fre-        patients increasingly difficult.
quently in women, e.g. rheumatoid arthritis, then            In diseases such as osteoarthritis, where women
women should be involved in Phase I studies. The          patients outnumber males (80%), a legitimate case
reality is that of the many hundreds of drugs and         can be made for a `female-weighted database', and
devices approved for use today, very few show             also when women are the majority users for medi-
major gender-related differences in either side           cines, such as psychotropic agents (although they
effects or efficacy. Clearly, in the drug classes that    are not necessarily the majority of sufferers). Pro-
have been shown to demonstrate significant gender         vision and timing of adequate animal toxicology
clinical differences, `specific' gender-related studies   and fertility data is critical to avoid expensive
should be included for investigation drugs and            delays and to allow adequate female recruitment,
                                  DRUG DEVELOPMENT RESEARCH IN WOMEN                                          177

so this animal data may be advanced on an `at-risk           the practice of clinical medicines using drugs not
basis', depending on the drug's clinical significance        thoroughly tested and understood, and withhold-
                                                             ing drugs that may be of benefit.
and its market potential. A list of diseases more
prevalent in women is provided in New Medicines
in Women (Holden, 2000).                                     It has been suggested that members of female
                                                          religious orders, women who have had tubal liga-
                                                          tion or lesbians could provide a `no-risk pregnancy'
                                                          pool of volunteers. While possible, this is not gen-
   The Potential Child-bearing Population                 erally a widely applicable solution, because geo-
                                                          graphic, environmental and volunteer numbers
The probability of potential early embryonic ex-          now become added variables.
posure occurring in a drug development program               Should women on oral contraceptives (OC) enter
must be expected and confronted because, despite          studies, could the high level of artificial hormones
careful pregnancy testing and adequate contracep-         could confound the results? Female oral contracep-
tive precautions being undertaken, it happens.            tive users make up 28% of the potential child-bearing
Levine (1975) in his book suggested that, in the          population (Ortho, 1991), and these hormone con-
consent form, there should be `a statement that           centrations (10±20 times higher than the natural hor-
the particular treatment or procedure may involve         mone levels) may cause drug interactions which
risks to the subjectÐ(or to the embryo or fetus if        cannot occur during ordinary menstrual cycling.
the subject is or may become pregnant) which are          Intra-uterine devices are currently regaining popu-
currently unforeseeable'.                                 larity,subdermalimplantshavehadlittleinfluenceon
                                                          contraceptive practice at the epidemiological level.
  When a woman of childbearing age participates in
  a research procedure in which there is a risk to the
  fetus, the nature of the risk being either known or
  unknown, she should be advised that, if she wishes                Liabilities for Fetal Damage
  to be a subject, she should avoid becoming preg-
  nant. Her plans for avoiding conception should be       Given all of the above reasons for including
  reviewed during consent negotiations. At times, if      women of child-bearing potential, the issue of the
  her plans seem inadequate and she does not con-         chilling effect of legal liability for fetal damage on
  sent to the investigator's suggestions, it will be
                                                          firms and institutions is still present, and the neces-
  necessary to exclude her from the research. She
  should be further advised that if she deviates          sary addition to the patient's informed consent
  from the plans discussed at the outset, she should      does not help. The Supreme Court in 1992 rejected
  advise the investigators immediately.                   an attempt to cap the amount juries could award in
                                                          damages as `unconstitutional', i.e., would require a
  Halbreich and Carson (1989) made the point              constitutional amendment. This is highly unlikely
that not to include women of child-bearing age            to occur. The consequence of litigation, particu-
could even increase liability:                            larly in obstetrics, was a dramatic increase in cae-
                                                          sarean section (18±20% of live births; this level was
  The general policy of an academic institution           even higher in 1999), resignation from this spe-
  should be to favor the conduct of research involv-      cialty, and a broader rejection of `high-risk' or
  ing women and children in testing of new drugs          Medicaid patients (O'Reilly et al, 1986; Bello,
  with potential for major therapeutic value to those
                                                          1989). A possible solution might be to follow the
  populations. Such research may expose the insti-
  tution to risk of liability for damage to subjects;     example of the National Vaccine Injury Act of
  however, that is inherent in research involving         October 1988, where a trust fund was set up derived
  human subjects anyway, and there are many               from an excise tax imposed on each vaccine. The
  ways of minimizing such risks. Not to do such           funds, through an arbitration panel, are used to
  research, while it may serve to protect the interests   compensate persons injured by vaccination. It
  of the institution as narrowly conceived, would         should be noted that a Drugs in Pregnancy Regis-
  involve a failure to serve the public interest in a
                                                          try has been set up to follow up early embryonic
  much more serious manner by exposing classes of
  persons to knowable but unknown risks, through          exposure to the anticonvulsants and antiviral drugs

acyclovir and retrovir. This is administered by the         but that greater thought be given to obtaining
American Social Health Association (ASHA),                  `representative' numbers in the early program plan-
Center for Disease Control (CDC), and Glaxo-                ning stage. For drugs intended mainly or entirely
Smithkline. One wonders if it could be expanded             for women, even Phase I testing in women should
(with suitable support) to cover additional agents.         be usually considered. Single-dose testing, even
                                                            in women of child-bearing potential, poses
                                                            minimal risk if done early in the cycle, with ad-
                                                            equate precautions and `consort' consent to
                  Data Gathering                            short sexual abstinence. Alternatively, women
                                                            with tubal ligation could be enrolled for these
Gender data are collected by major pharmaceutical           small studies.
companies; few, however, record the menstrual                  `Representative' could be twofold: a reflection of
dates. Frequently, no drug-handling differences be-         the percentage of women suffering from the dis-
tween the sexes is detected; much less commonly is          ease, or a `reasonable or sufficient' number to show
the absence commented upon in reports or publi-             clinically significant differences in efficacy or safety
cations. It is suggested that LMP dates could be            in the main efficacy and safety studies; alterna-
included in case report forms, and that publications        tively, conducting at least one study just in
and reports should contain statements on the                women in Phase III. What is a `clinically significant
presence or absence of gender differences, also             effect' would depend on the drug and disease, but
giving the patient gender numbers and p values.             effects with a less than 15% difference get harder to
This would allow for later meta-analysis. Both of           detect and generally will be less meaningful. Again,
these suggestions would be inexpensive to imple-            women of child-bearing potential could be in-
ment.                                                       cluded, depending on the age/prevalence of the
   Gender-related data from the FDA is more read-           disease. Women using oral contraceptives may be
ily available as the FDA continues to increase its          compared not only with males but also with non-
computer ability and pharmaceutical firms utilize           OC users. OC and drug interaction studies are
computer-assisted NDAs and increase their efforts           currently required for most drugs.
to adequately power the studies to find differences.           Early embryo drug exposure and the potential
Unified systems and formats would enhance this.             liability for any damage continues to influence in-
The information is included in the Summary Basis            dustry, agencies, and some research workers. It
for Approval or in the Medical Reviewer's Report.           must be recognized that, if an agent has human
Either should be available through the Internet at          teratogenic potential, it is better to detect this before under `New Approvals'.                    it achieves the marketplace. Unfortunately, this is
                                                            unlikely to be detected because the small numbers of
                                                            women becoming pregnant in any NDA program
                                                            make it impossible to detect drug-induced effects
                  CONCLUSION                                from spontaneous birth defects. Data in women
                                                            are needed and the possibility is suggested of an
Gender-related differences do exist in drug hand-           expanded National Register along the lines of the
ling, but in general are relatively clinically insignifi-   International Clearing House for Birth Defects
cant. Theoretically, because of weight differences,         Monitoring to follow up the expected small number
women may receive more medication than men                  of embryos exposed and a Compensation Panel in
for a standard dose when converted to mg/kg.                the event of proven damage, funded by an excise tax,
Greater effects might be expected from the range            as with vaccines.
of normal weights rather than from the effects of              Finally, with all the great strides being made to
gender.                                                     unravel the human genome and determine the gene
   Clinically significant gender effects have been          structures and their influence, we are much nearer
reported with CNS, anti-inflammatory and cardio-            to tailoring drugs to match male and female differ-
vascular drugs. It is suggested that women continue         ences, and with enhanced computer power, this
to be enrolled into most drug study programs,               chapter may become moot.
                                       DRUG DEVELOPMENT RESEARCH IN WOMEN                                                       179

               ACKNOWLEDGMENTS                                       Divoll M, Greenblatt DJ, Harmatz JS, Shader RI (1981) Effect
                                                                       of age and gender on disposition of temazepam. Pharm Sci 70:
                                                                       1104 ±7
The author wishes to acknowledge that much of                        Edwards LD (1991) Summary of survey results on including
this chapter was supported by a grant from the                         women in drug development. PMA. In Development Series,
NIH branch, Office of Protection from Research                         New Medicines for Women, Dec. 1991; 22±8.
Risks.                                                               Elia J, Katz IR, Simpson GM (1987) Teratogenicity of psycho-
                                                                       therapeutic medications. Psychopharmacol Bull 23: 531±86.
                                                                     FDA (1986) Drug Utilization in the US 1986ÐEighth Annual
                                                                       Review. FDA: Washington, DC.
                                                                     Fed Reg (1990) Guideline for study of drugs likely to be used in
           RECOMMENDED READING                                         the elderly. Fed Reg 55: FR 7777; (1997) Labeling: subsection,
                                                                       geriatric use. 21 CFR Pt 201.
Mastroianni AC, Faden R, Federman D (eds).                           Fed Reg (1994). Detection of toxicity to reproduction for medi-
(1994) Women and Health Research: Ethical and                          cinal products. ICH Guideline SSA (59 Fed Reg: 48746)
                                                                     Fed Reg (1996). Detection of toxicity to reproduction for medi-
Legal Issues of Including Women in Clinical Studies                    cinal products. Addendum an toxicity to male fertility. ICH
Academy Press.                                                         Guideline 553 (61 Fed Reg: 15360)
                                                                     Fed Reg. (1993). Guideline for the study and evaluation of
                                                                       gender difference in the clinical evaluation of drugs. (58 Fed
                                                                       Reg: 39406±16)
                       REFERENCES                                    Frederick J (1973) Epilepsy and pregnancy: a report from the
                                                                       Oxford Record Linkage Study. Br Med J ii: 442±8.
Abel JG, Sellers EM, Naranjo CA, et al (1979). Inter and             Frezza M, DiPadova C, Pozzato G et al (1990) High blood
  intrasubject variation in diazepam free faction. Clin Pharma-        alcohol levels in women. The rate of decreased gastric alcohol
  col Ther 26: 247±55                                                  dehydrogenase activity and first-pass metabolism. N Eng J
Abernathy DR, Greenblatt DJ (1981) Impairment of antipyrine            Med 322: 95±9.
  metabolism by low dose oral contraceptive steroids. Clin           Gengo FM, Fagin SC, Kinkel WR, McHugh WB (1984) Serum
  Pharm Ther 29: 106±110.                                              concentrations of propranolol and migraine prophylaxis.
Allen MD, Greenblatt DJ (1981) Comparative protein binding             Arch Neurol 41: 1306±8.
  of diazepam and desmethyldiazepam. J Clin Pharmacol 21:            Greenblatt DJ, Divoll M, Harmatz JS, Shader RI (1980) Oxa-
  219±23.                                                              zepam kinetics: effects of age and sex. J Pharmacol Exp Ther
Arlington S (2000) Pharma 2005 (Industry Trends) Pharmaceut            215: 86±91
  Exec 20(1): 74±80.                                                 Greenblatt DJ, Harmatz JS, Shader RI (1979). Sex differences in
Bello M (1989) Liability crisis disrupts, distorts maternity care.     diazepam protein binding in patients with renal insufficiency.
  National Research Council New Report 39: 6±9.                        Pharmacology 16: 26±9
Booth M, Hunt JN, Miles JM, Murray FA (1957) Comparison              Halbreich U, Carson SW (1989) Drug studies in women of
  of gastric emptying and secretion in men and women with              childbearing age: ethical and methodological consideration.
  reference to prevalence of duodenal ulcer in each sex. Lancet        J Clin Psychopharmacol 9: 328±33.
  1: 657±9.                                                          Health Care Financing Administration US Health and Human
Bush JK, Cook SF, Seigel E (1993) Issues of special populations        Services. www.hcfa gov/stats/stats.htm
  in drug developments. Drug Inf J 27: 1185±1193.                    Henrekens CH et al (1989) Steering Committee of the Physicians
Clark AF, Calandra RS, Bird CE (1971). Binding of Testoster-           Health Study Research Group Final report on the aspirin
  one and 5-dihydrotestosterone to plasma protein in humans.           component of the ongoing Physicians Health Study. N Engl
  Clin Biochem 4189±96                                                 J Med 321: 129±35.
Coale AJ (1991) Population and development review. J Popula-         Holden A (2000) New medicines in development for women: a
  tion Council NY :   .                                          1999 survey. Pharmaceutical Research and Manufacturers'
Coles SL, Fries JF, Kraines RG, Roth SH (1988) Side effects            Association., 29±36.
  of non-steroidal antiinflammatory drugs. Am J Med 74:              Jefferson JW, Greist JH, Ackerman DL (eds) (1987) Lithium
  820±828.                                                             Encyclopedia for Clinical Practice, 2nd edn. American Psychi-
Conrad CD, Hamilton JA (1986) Recurrent premenstrual                   atric Press: Washington, DC. 640±45.
  decline in lithium concentration: clinical correlates and treat-   Kristensen CB (1983) Imipramine serum protein binding in
  ment implications. J Am Acad Child Psychiat 26(6): 852±3.            healthy subjects, Clin Pharmacol Ther 34: 689±94
Cotton P (1990) Medical news and perspective. J Am Med Assoc         LaRosa GH, Pinn VW (1993) Gender bias in biomedical re-
  263 (8): 1049±50.                                                    search. J Am Med Womens Assoc 48(5): 145±51.
Cramer JA, Scheyer RD, Mattson RH (1990) Compliance de-              Levine RJ (1975) Ethics and Regulations of Clinical Research,
  clines between clinic visits. Arch Intern Med 150: 1509±10.          2nd edn.
Diokno SC, Brock BM, Brown MB, Hertzog AR (1986) Preva-              Lindhaut D, Happener RJ, Meinardi H (1984) Teratogenicity of
  lence of urinary incontinence and other urological symptoms          antiepileptic drug combinations with the special emphasis on
  in non-institutionalized elderly. J Urol B6: 1022±5.                 epoxidation of carbamazepine. Epilepsia 25: 77±83.
Lindhaut D, Schmidt D (1986) In utero exposure to valproate         Roscizeweska D, Buniner B, Guz I, Sawisza H (1986) Ovarian
  and neural tube defects. Lancet i: 1392±3.                          hormones anticonvulsant group and seizures during the men-
MacDonald I (1965) Gastric activity during the menstrual cycle.       strual cycle in women with epilepsy. J Neurol Neurosurg Psy-
  Gastroenterology 30: 602±7.                                         chiat 49: 47±51.
Majaverian P, Rocci MC, Connor DP et al (1987) Effect of food       Routledge PA, Stargel NW, Kitchell BB, Barchowski A, Shand
  on the absorption of enteric coated aspirin correlation with        DG (1981) Sex-related differences in plasma protein binding
  gastric residence time. Clin Pharm Ther 41 (1): 11±17.              of lignocaine and diazepam. Br J Clin Pharmacol 11:
Majaverian P, Vlasses PH, Kellner PE, Rocci ML (1988) Effects         245±50
  of gender posture and age on gastric residence time of an         Shavit G, Lerman P, Konczyn AD et al (1984) Phenytoin
  indigestible solid: pharmaceutical considerations. Pharmaceut       pharmacokinetics in catamenial epilepsy. Neurology 34:
  Res 5(10): 639± 44.                                                 959±61.
Martin RM, Biswas PN, Freemantle SN, Pearce GL, Minow               Simpson JM, Bateman DN, Rawlins MD (1987) Using the
  RD (1998) Age and sex distribution of suspected adverse drug        adverse reactions register to study the effects of age and sex
  reactions to newly marketed drugs in general practice in Eng-       on adverse drug reactions. Statist Med 6: 863±7.
  land. Br J Clin Pharmacol 46(5): 505±11.                          Smith JM, Baldessarini RJ (1973) Changes in prevalence, sever-
M'Buyamba-Kabunga JR, Lijen P, Fagard R, et al (1985).                ity and recovery in tardise dyskinesia with age. Arch Gen
  Erythrocyte concentrations and transmembrane fluxes of              Psychiat 29: 177±89.
  sodium and potassium and biochemical measurements during          Stewart DE (1998) Are there special considerations in the pre-
  the menstrual cycle in normal women. Am J Obstet Gynecol            scription of serotonin reuptake inhibitors for women. Can J
  151: 687±93                                                         Psychiat 43(9): 900±904.
Metropolitan Life Insurance (1999). Height ± weight trades.         Tran C, Knowles SR, Liu BA, Shear NH (1998) Gender differ-                                                ences in adverse drug reactions. J Clin Pharmacol 38(1): 1003±9.
Miaskiewicz SL, Shively CA, Vesell ES (1982) Sex differences in     Trussell J, Hatcher RA, Cates W, Stewart FH, Kost K (1990)
  absorption kinetics of sodium salicylates. Clin Pharmacol           Contraceptive failures in the United StatesÐan update. Stud
  Ther 31: 30±37.                                                     Family Planning 21(1).
Michelakis AM, Yoshida H, Dormois JC (1975) Plasma renin            US Public Health Services Office of Womens Health (1996).
  activity and plasma aldosterone during the normal menstrual         Heart disease and stroke in women. Report on data collected
  cycle. Am J Obstet Gynecol 123: 724±6.                              by the National Heart, Lung, and Blood Institute. Chart
National Disease and Therapeutic Index 1991 Plymouth Meet-            Book: Bathesda, Maryland.
  ing, PA IMS 1991 America 1978±1989                                Walle T, Byington RP, Furberg CT et al (1985) Biologic deter-
National Institutes of Health Guide, vol 15. (1986) National          minants of propranolol disposition. Results from 1308 pa-
  Institutes of Health: Bethesda, MD; 14.                             tients in the b-blocker heart attack trial. Clin Pharmacol
National Prescription Audit 1991 Plymouth Meeting, P. A.,             Ther 38: 509±18.
  IMS America, Ltd. 1964 ±1989.                                     Walle T, Walle U, Cowart TD, Conradi EC (1989) Pathway
NIH/ADAMHA (1990) Policy concerning inclusion of women                selective sex differences in metabolic clearance of propranolol
  in study populations PT 343, 11: KW 1014002, 1014006. NIH           in human subjects. Clin Pharm Ther 46(3): 257±63.
  Guide Grants Contracts 19: 18±19.                                 Weissman MM, Klerman GL (1977) Sex differences and the
Notivol R, Carrio I, Cano LE, Storch M, Vilardell F (1984)            epidemiology of depression. Arch Gen Psychiat 34: 98±11.
  Scand J Gastroenterol 18: 1107±14.                                Williams PT (1996) High density lipoprotein cholesterol and
O'Reilly WB, Eakins PS, Gilfix MG, Richwald GA (1986)                 other risk factors for coronary heart disease in female runners.
  Childbirth and the Malpractice Insurance Industry. In Eakins        N Engl J Med 334 (20): 1298±1303.
  JNP (ed.), The American Way of Birth. Temple University           Wilson JF (1973) Environment and Birth Defects. Academic
  Press: Philadelphia, PA; 195±212.                                   Press: New York.
Ortho (1991) 23rd Ortho Annual Birth Control Study. Survey          Wilson JG (1972) Environmental effects on developmentÐtera-
  by Ortho. Rariton, NJ.                                              tology. In Assali NS (ed.), Pathophysiology of Gestation vol 2.
Physician's Desk Reference Montvale, New Jersey: 56th Edition         Academic Press: New York, 269±320.
  Medical Economics.                                                Wilson K, Oram M, Horth CE, Burnett D (1982). The influence
Pope J, Aufderheide TP, Ruthazer R, Woodard RH et al (2000)           of the menstrual cycle on the metabolism and clearance of
  Missed diagnoses of acute cardiac ischemia in the emergency         methaqualone. Br J Clin Pharmac other 14: 333±9.
  department. N Engl J Med 342 (16): 1163±71.                       Wilson MA, Dwyer KD, Roy EJ (1986). Direct effects of
Raskin A (1974) Age±sex differences in response to antidepres-        ovarian hormones on antidepressant binding sites. Brain Res
  sant drugs. J Nerv Ment Dis 159: 120±30.                            Bull 22: 181±5
Ray WA, Griffin MR, Schaffner W et al (1987) Psychotropic           Wilson MA, Roy EJ (1986). Pharmacokinetics of imipramine
  drug use and the risk of hip fractures. N Engl J Med 316:           are affected by age and sex in rates. Life Sci, 38: 711±8.
  361±9.                                                            Wright RA, Krinsky S, Fleeman C et al (1983) Gastric emptying
Report of the Public Health Service Task Force on Women's             and obesity. Gastroenterology 84: 747±51.
  Health Issues (1985) PHS Rep 100: 73±105.                         Yacobi A, Stoll RG, Di'Sanio AR, Levy G (1976) Intersub-
Roberts RK, Desmond PV, Wilkinson GR, Schenker S (1979)               ject variation of warfarin binding to protein in serum of
  Disposition of chlordiazepoxide: Sex differences and effects of     normal subjects. Res Commun Chem Pathol Pharmacol 14:
  oral contraceptives. Clin Pharmacol Ther 25: 826±50                 743± 46.
                                  DRUG DEVELOPMENT RESEARCH IN WOMEN                                              181
Yazigi Odem RR, Polakoski KL (1991) Demonstration of spe-   Yoder MC, Belik J, Lannon R, Pereita GR (1984) Infants
 cific binding of cocain to human spermatozoa, J Am Med      of mothers treated with lithium (Li) during pregnancy
 Assoc 266 (14): 1950±60.                                    have an increased incidence of prematurity. Pediat Res 18:
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                      Clinical Research in Children

                  BACKGROUND                                    were approved and made available for children.
                                                                Together, they included 44 new agents targeting
The world population reached 6 billion in June                  cancer, 14 for cystic fibrosis, 13 for asthma, nine
2000, and half of the world's population (3 billion)            for epilepsy, and 23 potential vaccines.
are less than 15 years old. Sadly, the mortality rate              This chapter will focus on the current regulatory
of children in Third-World countries is 10 times                requirements, their background, the clinical study,
higher than in the developed world. Before 1850,                challenges and the clinical issues of drug research in
half of the children born in the USA died from                  the pediatric population.
infections before 5 years of age. The introduction                 Gordon Still has estimated that over 2500 studies
of sanitation, antiseptics and, in the last century,            in the FDA's pediatric subgroups may need to be
vaccines, and lately medicines, have made such                  conducted over the next 3 years (Still 2000). This
early deaths in the USA now very uncommon.                      includes completion of pediatric studies on the FDA
   Today, accident is the largest killer of children,           priority list of marketed products. Estimates of the
accounting for 2500 deaths in children under 5                  annual cost to the industry of these studies vary. The
years old; this compares to 700 deaths from con-                FDA estimated in 1994 that $13.5±20.9 million/year
genital abnormalities, 518 from cancer and 473                  would be spent by industry (Fed Reg 1998). At a
from murder. AIDS is the leading infectious cause               press conference, Christopher Jennings, President
of death in the under±5 year-olds (200). The major              Clinton's principal health care advisor, said that
causes of death in the 5±14 year-olds are accidents             pediatric label studies would only be about 1% of
(3500), cancer (1053) and murder (570). Again, in               the development cost of a drug. Dr Henry Miller
this age group, AIDS is the leading infectious cause            (1997), a former FDA Director of the Office of
of death (National Center for Health Statistics                 Biotechnology, said that applying Jennings' figure
1996).                                                          will mean an industry cost of $200 million (1% of the
   Many of the childhood cancers are hematologic,               $20 billion spent on R&D). Dr Gordon Still, pre-
and great improvements in survival have been                    senting at the 36th Drug Information Association
achieved. For example, the acute leukemia survival              (DIA) Annual Meeting (1999), estimated the cost at
rate in children has risen from 53% in 1970 to 80%              $892 million if all five pediatric subgroups were to be
by 1989 (American Cancer Society); and new sur-                 studied (based on 1999 study costs).
gical techniques and new devices are improving and                 These additional costs for pediatric studies
sometimes correcting (even by intrauterine surgery)             may be justified if these studies satisfy all global
many previously fatal congenital abnormalities,                 markets. Macleod (1991) estimated that `develop-
e.g. hypoplastic heart.                                         ing countries' by the year 2000 will comprise 36%
   In June 1998 (Holmer 1998), the Pharmaceutical               of the total pharmaceutical market and that half of
Research Manufacturers' Association (PhRMA)                     their populations are children (accounting for 18%
reported, from their survey of pharmaceutical com-              of the market). In the developed countries, children
panies, that 187 medicines or vaccines were in devel-           under 18 years account for 20% of the market. It
opment for children. This was an increase from 146              would seem that the 38% pediatric share of the
the previous year. In addition, 20 new medicines                global market is worth an extra effort.

CHILDREN, THE THERAPEUTIC ORPHANS                           The Pediatric Use Working Group, chaired by
                                                         Miriam Pina (1995) (FDA Division of Pulmonary
The Food, Drug and Cosmetic Act, first passed            Drugs), examined data that the FDA had acquired
in 1906, was dramatically altered by the 1962            on 1994 pediatric prescriptions from IMS. From
Kefauver±Harris amendments as a direct result of         these they identified the top ten drugs used `off
the thalidomide tragedy. This amendment required         label' in children: Albuterol, Phenergan, Ampicillin
that drugs must be both safe and effective before        i.m. or i.v., Auralgan otic solution, Lotison, Prozac,
marketing approval could be given. In addition,          Intal, Zoloft, Ritalin (under 6 years old) and alu-
adequate animal, toxicology and fertility testing        pent syrup (under 6). A combined total of over 5
had to be concluded prior to the first dose in           million of these 10 products were prescribed in 1994.
humans. Substantial additional testing in animals           Clearly, firms needed further encouragement
and in humans was required prior to marketing            to submit additional pediatric data, so in 1997
approval. This led to the era of the Science of          Congress passed the FDA Modernization Act
Clinical Trial Design. Regrettably, the testing of       (FDAMA). This called for firms to submit data
drugs in children did not advance at a similar           on children to support labeling for a new pediatric
pace, and most drugs (unless specifically intended       subsection before the drug could be approved. This
for children) were never tested in children by the       applied to drugs that could be projected to provide
sponsors of new medicines.                               therapeutic benefit to substantial numbers of
   Physicians were thus forced to use most drugs `off-   children. In exchange, Congress felt that an induce-
label' and extrapolate the child dose on a compara-      ment was required and wrote into the Act provision
tive weight basis from that in adults. This often        for an extension of a drug's patent life by 6 months
involved parents splitting or crushing tablets, hiding   if pediatric studies were done. For a $2 billion drug
medication in spoonfuls of honey, or sprinkling          such as Claritin (Loratidine) 6 months' extra exclu-
a crushed tablet onto a meal. Each time this             sivity is not `small change'. The FDA was re-
happened, a little more confidence in and knowledge      quested to provide guidance and, in December,
of the drug was gained, but each child was a `one-off    1998, it issued the Final Rule Amendments to the
experiment' and only provided a learning curve for       Pediatric Subsections to be implemented April
the individual physician. Eventually, academia           1999, governing the need for pediatric studies,
would publish a series of cases, so giving guidance      and extending the requirements to biological
on dosing and likely toxic effects. Even so, the aver-   drugs and already-marketed drugs. The FDA iden-
age pediatrician and family practitioner felt uneasy     tified drugs for which supplemental data was still
and legally vulnerable about off-label use.              needed for pediatric labeling. The FDA has issued
   A few drugs were developed for children in            an annual list of `priority drugs' for which add-
such categories as antibiotics, antihistamines, and      itional pediatric information may be `beneficial'.
antiepileptics. But otherwise, few firms undertook          FDA chose to interpret the patent life extension
studies to develop full pediatric label instructions     as applying to all indications, not just to pediatric
or even pediatric formulations. Liquid formula-          use. As might be expected, the generic companies
tions did exist for some drugs, but mainly for           are appealing this interpretation of the pediatric
use in the elderly. In 1975, Wilson surveyed the         rule.
1973 Physician's Desk Reference for labeling in-
structions for pediatric patients and pregnant or
breast-feeding women. He found that 78% of listed             1994 AND 1998 FINAL RULES ON
drugs either had no information for pediatric                       PEDIATRIC STUDIES
dosing or contained a disclaimer. A subsequent                     (FED REG 1994, 1998)
survey by Gilman (1992) showed that this situation
had not improved qualitatively and had also risen                 Products Subject to the Rule
to 81%. Eventually, the FDA issued the 1994 rule,
which sought to strengthen the 1979 guideline on         For drugs that are new molecular entities (NMEs),
pediatric labeling requirements (Fed Reg 1994).          a determination should be made by the sponsor of
                                       CLINICAL RESEARCH IN CHILDREN                                          185

the potential usefulness of the new drug in a pediat-      when the product is life saving and the only treat-
ric population. If it is likely to generate over 100 000   ment available.
prescriptions/year, this would indicate the need
to develop a pediatric formulation and suitable
pediatric studies. If it is likely to generate less            MAJOR PHYSIOLOGIC VARIATIONS
than 50 000 prescriptions/year, the sponsor may                        IN PEDIATRICS
be granted a waiver by the FDA for pediatric
data, and a disclaimer statement allowed. Either           In the past, the statement that `children are not
way, in a children's disease, if less than 200 000         little people' dominated research thinking. In gen-
patients/year may benefit, then orphan-drug                eral, both in children and in the elderly, drugs and
status with seven/year exclusivity may be applic-          biological products behave similarly to that in the
able. This would then apply only to that pediatric         18±65 year-old population, although this expect-
indication.                                                ation must be adjusted for age-related differences
   The requirements of the Pediatric Final Rule            in pharmacokinetic variables, such as immature or
now includes marketed drugs and biologics. The             aging enzyme metabolism systems as well as elim-
FDA has already listed products affected and sent          ination rates affected by immature or aging organs
pediatric data requests to firms. The firms had until      of excretion.
April 2001 to provide the extra data.                         In neonates, the gastric pH is biphasic, being high
                                                           in the first few days after birth and decreasing by day
                                                           30, but it takes 5±12 years for the adult pattern and
               Data to Be Provided                         value to emerge (Signer and Fridrich 1975). On
                                                           the other hand, the methylation pathway, unim-
If considerable data exists, or is planned, for same       portant in adults, is well developed in children. Fur-
indications in adults, it may be appropriate to ex-        thermore, acetaminophen is less toxic to children
trapolate safety and efficacy from adults to chil-         than to adults, probably because it utilizes the sul-
dren. But pharmacokinetic studies to determine             phate metabolic pathway (Rane 1992).
dosing and, if possible, pharmacodynamics data,               Most infants are slow acetylators and may accu-
will usually be required for children. Discussion          mulate toxic levels of those drugs that are metabol-
with the FDA is recommended early on, to estab-            ized by this second phase of metabolism route.
lish whether pediatric data will be required and           Renal perfusion and glomerular filtration rates
which of the five groups should be covered (pre-           (GFR) vary: for the premature, 2±4 ml/min; for
term, neonate 0±1 month, infants 1±2 years, chil-          neonates, 25 ml/min; and by 1±1.5 years old,
dren 2±12 years and adolescents 12±16 years). One          125 ml/min, which is equivalent to adult clearance
or more adequately sized efficacy and safety studies       rates (Arant 1978). The potential toxic implication
may be required, especially if the drug or disease         of renal metabolites and elimination of unchanged
behaves differently in children, or the drug uses          drug in the very young are obvious (Stewart and
different metabolic pathways. This may occur               Hampton 1987).
if the particular adult enzyme is not present in
children, or is only present in low quantities. If
a different indication to that in adults is being                                 Dosing
sought, then one or two sizable safety and efficacy
studies, in one or more age groups, are likely to be       Without pediatric pharmacokinetic data, dosing
required. This is in addition to pediatric pharmaco-       in children has depended on extrapolations from
kinetic data. Sponsors should also plan for the            the adult data, either by weight or by body surface
major ethnic groups to be represented in these stud-       area. Using weight may result in overdosing neo-
ies.                                                       nates but underdosing infants and children. Using
   Frequently, the FDA may allow approval of a             body surface area may be better because of its
drug with incomplete pediatric data and defer the          linear increase with age and its good correlation
completion to a Phase IV commitment, especially            with cardiac output, renal flow and glomerular

filtration rateÐmore so than weight. Neither            animals may be required (if the product is intended
method compensates for the varying metabolism           for oral delivery), and given by dropper gauage.
aforementioned, nor for differences in drug dispos-
ition between children and adults.
                                                                        CLINICAL STUDIES

  Concerns in Formulations for Pediatrics                                 Pharmacokinetic
If a drug is to be given by injection, i.m. or i.v.,    The traditional pharmacokinetic (PK) study volun-
this may require only volume variations. But most       teer study in healthy children has proved very
drugs developed for adults are given by the oral        hard to set up, because of the attitude of many
route, as tablets, capsules or caplets. The adult       parents and overviewing independent review
formulation is usually determined by marketing          boards (IRBs). Even in pediatric patients, the fre-
considerations. Invariably, for children, especially    quency and total volume requirements for samples
under age 7 years, liquid or syrup must be formu-       for conventional PK studies can cause the same
lated. Most drugs taste bitter or unpleasant (which     refusals. However, there are pediatric research
is why most tablets are sugar-coated). Sometimes,       units that specialize in these studies, with minimum
it may be impossible to completely mask the taste.      needle sticks, minute blood volumes, and IRBs
A commitment to a pediatric formulation requires        sympathetic to the needs of the pediatric commu-
a whole gamut of testing and the development of         nity. The National Institute of Child Health and
specific product specifications. If the liquid formu-   Human Development has set up a `network of
lation changes the bioavailability (faster or slower    pediatric pharmacology units', usually in academic
absorption), then further efficacy and safety studies   regional centers, now numbering 13 units. There
may be required. A further concern is that liquid       are other non-governmental specialized units also
formulations often have a shorter shelf life than       available for pediatric PK work.
tablets. Finally, stability characteristics or other       An alternative method of getting pharmacoki-
factors may make it impossible to make a liquid         netic data is to take a small extra sample of blood
or syrup or glycolated elixir, sprinkle beads or        (and urine) at a child's regular scheduled visit when
powder sachets, and split or crushed tablets may        blood is drawn for routine blood work. The time of
be a last resort. In the later two cases, an even       day of this sample is predetermined by the time of
distribution of active compound and other inactive      the administration of the medicine. If samples are
excipients must be demonstrated. In addition, a         obtained from many children, a weight±age-cor-
lack of effect an bioavailability must be proved if     rected, scatter-plot graph can be constructed and
such advice is to appear in the dosing instructions.    a pharmacokinetic profile be calculated. This is the
                                                        `pharmacokinetic screen' method. A version of this
                                                        method is also utilized to gather ethnic data for
                    Toxicology                          labeling in adults as well as children, and is called
                                                        `population pharmacokinetics'.
The plastic nature of immature organs such as
kidney, liver, brain and lung may indicate the
need for more animal toxicology. Frequently, neo-                            Recruitment
natal acute and subacute toxicology studies are
undertaken in two animal species. Because of the        One of the major problems in running pediatric
small size of both mouse and rat pups, this may         clinical trials is the availability of pediatric patients,
prove a challenge to administer the active drug.        who tend to be scattered, because they are numer-
The common `mixing with chow' is inappropriate          ically less likely to have diseases (other than asthma
in neonates. Dog pups usually provide one of the        and the usual childhood illnesses). This affects the
two species, so a special liquid formulation for        logistics of screening and subsequent clinic visits.
                                      CLINICAL RESEARCH IN CHILDREN                                         187

Another hurdle is finding trained pediatric investi-     as Wards of Court or by a social welfare agency.
gators or pediatric pharmacologists, and overseas        These patients should be rarely used, unless
they are even harder to find. In Europe, there is        the treatment is for serious disease specific to insti-
collaboration between the USA-based Pediatric            tutional settings and no other treatment is avail-
Pharmacology Research Units (PPRU), the Euro-            able.
pean Society of Developmental Pharmacology, and             Emergency situations can arise where it may
the European Network for Drug Investigation in           be impossible to obtain written informed consent
Children (ENDIC). For diseases of children, there        from a parent or guardian. Medications for this
are often self-help organizations who can prove          type of problem will require intense IRB review
invaluable in recruiting children and in reassuring      and overview; only in special circumstances will
their parents.                                           informed consent be waived, and then it must `not
   A large package of data, and two well controlled      adversely affect the rights and welfare of the sub-
pivotal studies of safety and efficacy are rarely        ject'. (Abramson et al 1986). The last category is
required, with the exception of diseases specific to     the use of a research medicine in a child close to
childhood, such as surfactant studies in respiratory     dying who has either no response to standard ther-
Distress Syndrome. This is especially the case if the    apy or where no alternative therapy exists. The
drug has similar effects in both adult and pediatric     agent to be considered must have some evidence
populations e.g., antihistamines.                        of efficacy (animal proof of concept or clinical data
   However, if a disease or drug behaves in differ-      and a good chance of a beneficial result). The risk
ent ways in children compared with adults but a          of unintended coercion of desperate parents is es-
large body of safety data exists in adults, then         pecially to be guarded against.
usually only a single efficacy and safety study is
                                                                       IRBs' Special Emphasis
                                                         IRBs have a duty to make sure the study is of value
                Ethical Concerns                         to children in general and in most cases to the
                                                         patient him/herself; is robust enough to give
The American Academy of Pediatrics formed a              answers; and attempts to minimize risk and maxi-
Committee on Drugs to examine ethical issues of          mize benefit. In reviewing the protocol, the IRBs
pediatric studies for its members and for the            should involve health care specialists who are
guidance of IRBs dealing with pediatric studies.         aware of the special medical, psychological, and
The Committee released its report in 1995. This          social needs of the child, and the disease as special
report (Committee Drugs, American Academy of             medical, psychological and social needs of the
Pediatrics) is very comprehensive, but amongst its       child, and the disease as might be impacted by the
many recommendations the following areas are             study.
highlighted, as follows.                                    In studies conducted on diseases mainly affecting
                                                         pediatric patients, the development will be entirely
                                                         in pediatric patients. However, in addition to the
                                                         appropriate usual toxicology and neonate animal
In this special population, there is a special duty to   toxicology, the first-in-humans studies for toxicity
avoid (unintended) coercion of the patient, parent       and safety are usually done in healthy adult volun-
or guardian. This coercion may arise because             teers. Clearly, however, drugs such as the surfac-
the investigator is usually also the treating phys-      tants would yield no useful data from adult testing.
ician. It would be better to have a colleague explain    For these unique pediatric situations, new meas-
and obtain the informed consent. There are varying       urements and endpoints may need to be developed
degrees of vulnerability. Patients handicapped           and validated. Frequently the FDA will involve
either mentally, emotionally or physically are           an advisory panel to help determine what these
frequently institutionalized and may be supervised       might be.

           The Use of a Placebo Control                     Arant BS (1978) Developmental patterns of renal function mat-
                                                              uration compared to the neonate. J Pediat 92: 705±12.
Placebo control is desired whenever possible if             American Cancer Society.
using a placebo does not place the pediatric pa-            Committee on Drugs, American Academy of Pediatrics. Guide-
                                                              lines for the Ethical Conduct to evaluate drugs in Pediatric
tients at increased risk. The AAP Committee on
                                                              populations. Pediatrics 95: 286±94.
Drugs (1995) outlined other circumstances:                  FDA (1997) Food and Drug Administration Modernization
                                                              Act. Pub. Law 105±115 (1997) Nov. 21), USDC 355a, 111
. No other therapy exists or is of questionable               Stat. 2296.
  efficacy, and the new agent might modify the              Fed Reg (1994) Specific requirements on content and format of
                                                              labeling for human prescription drugs revision of `pediatric
  disease.                                                    use' subsection in the labeling (1994) Fed. Reg. 59(238):
. If the commonly used therapy has a high profile             64240±50.
  of adverse events and risk greater than benefits.         Fed Reg (1998) Final rule regulations requiring manufacturers
. When the disease fluctuates frequently from                 to assess the safety and effectiveness of new drugs and bio-
  exacerbations to remissions thus the efficacy               logical products in pediatric patients (1998) Fed Reg 63:
  of the (new) treatment cannot be evaluated.               Fed Reg (1983), 16d.
                                                            Fed Reg (2000) International Conference on Harmonization:
                                                              clinical investigation of medicinal products in the pediatric
                  CONCLUSION                                  population (2000) Fed Reg 65 (71): (draft).
                                                            Gilman JT, Gal P (1992) Pharmacokinetic and pharmacody-
                                                              namic data collection in children and neonoates. Clin Phar-
The ICH draft guidance on pediatric issues has                macokinet 23: 1±9.
been published in the Federal Register (2000), but          Holmer AF. (1998) Survey or new medicines in children (1995)
at this point of writing has not been issued as a final       Pharmaceutical Research and manufactures of America.
rule. This guidance covers pediatric formulation,   
                                                            MacLeod SM (1991) Clinical pharmacology and optimal thera-
development, ethics, regional and cultural issues,
                                                              peutics in developing countries: aspirations and hopes of the
regulatory expectations, duration and type of stud-           pediatric clinical pharmacology subcomittee. J Clin. Epide-
ies, and age ranges to be studied. The guidance is            miol 44(suppl II): 89±93.
similar to the FDA Final Rules (Fed Reg 1994,               Miller telephone interview (1997) Script No. 2260, Aug. 22,
1998) with the addition of a fifth group, preterm             1997.
                                                            National Center for Health Statistics (1996) Based on data from
newborn infants. It also seems better organized and           US Dept. Health and Human Services.
informative, but then, hindsight is always helpful.         Pina LM (1995) Drugs widely used off-label in pediatrics.
   The face of pediatric pharmacologic medicine has           Report of the Pediatric Use Survey Working Group of the
been changed. In future, for pediatricians there will         Pediatric Subcommittee.
be less uncertainty and better predictive informa-          Rane A (1992) Drug disposition and action in infants and
                                                              children. In Yaffe SJ, Arand AJV (eds), Pediatric Pharmacol-
tion available; for children, safer and more effective        ogy, Therapeutic Principles in Practice. Sanders: New York;
dosages will result. For the industry, the added cost         10±12.
of research will be more than recouped in a new             Signer E, Fridrich R (1975) Gastric emptying in newborns and
global market to which previously they could not              young infants. Acta Paediat Scand 64: 525±30.
promote their products. This is supported by the            Stewart GF, Hampton EM (1987) Effect of maturation on drug
                                                              deposition in pediatric patients. Clin Pharm b: 548±64.
1998 survey by PhRMA, which showed that medi-               Still JG (2000) The pediatric research initiative in the United
cines and vaccines in development for children were           States: implications for global pediatric research. Drug Inf J
up 28% from the previous year.                                35: 207±12.
                                                            Wilson JT (1975) Pragmatic assessment of medicines available
                                                              for your children and pregnant or breast-feeding women. In
                                                              Morsell PL, Garattini S, Serini F (eds), Basic Therapeutic
                    REFERENCES                                Aspects of Perinatal Pharmacology. Raven: New York.

Abramson VS, Meisal A, Sufar P (1986) Deferred consentÐa
 new approach for resuscitation research on comatized pa-
 tients. J Am Med Assoc 225: 2466±71.
Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                       Copyright # 2002 John Wiley & Sons, Ltd
                                                                    ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                                             Section IV
     Applied Aspects of Drug Development
      Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                             Copyright # 2002 John Wiley & Sons, Ltd
                                                                          ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

                         Introduction to Section IV:
                                                Anthony W. Fox
                                  President, EBD Group, Carlsbad, California.

Having covered in Sections II and III the strictly                 These modern aspects of drug development are
clinical, more orthodox aspects of drug develop-                rarely optional. All are crucial for the success of a
ment, we now turn to some applied aspects. In                   product in the marketplace. Several of the next
general, these reflect relatively modern sophistica-            chapters describe methodologies that teach us, on
tions in the development process, compared with,                new dimensions, about properties that are intrinsic
for example, many types of clinical trial design,               to drugs. Some (e.g., pharmaceoeconomics) are
which have been available for decades.                          also becoming of increasing importance in the
                                                                regulatory approval process.
    Principles and Practice of Pharmaceutical Medicine Edited by Andrew J Fletcher, Lionel D Edwards, Anthony W Fox, Peter Stonier
                                                                                           Copyright # 2002 John Wiley & Sons, Ltd
                                                                        ISBNs: 0-471-98655-0 (Hardback); 0-470-84629-1 (Electronic)

               Biotechnology Products and Their
                                  David Shapiro and Anthony W. Fox
                   Scripps Clinic, La Jolla, CA, and EBD Group Inc., Carlsbad, CA, USA

The objectives of this chapter are to describe what             1796, while the last case of the disease occurred in
biotechnology products are, and how their regula-               1979 following a laboratory accident. The Vario-
tion is similar to, or differs from, small molecule             cella vaccine may yet not be redundant in combat-
drugs. It is a common assumption that biotechnol-               ting bioterrorism.
ogy has sprung from nothing, de novo, within a                     It is beyond the scope of this chapter to discuss all
small number of recent years. This is not the case,             potential applications and all present technologies
and we shall show how the recent growth of this                 associated with biological drugs. We merely provide
field actually has a basis which is, in many ways,              an overview. We shall not cover orthodox vaccines,
common to and interconnected with the develop-                  fermented products (e.g. for antibiotics), blood
ment of all other types of drugs.                               products, tissue-extracted hormones, diagnostic
                                                                products (e.g. antibody-based assay systems), and
Definition: Biotechnology products are those that               devices utilizing biotechnology products, which are
are prepared using biological, rather than chemical             not themselves used therapeutically. These are, tech-
synthetic, processes. Biological processes may be               nically, biotechnology products, but here we wish to
manipulated in vivo, ex vivo or in vitro.                       concentrate on the newer technologies.
                                                                   There are approximately 1250 biotechnology
   The list of types of biotechnology products is               companies in the USA and Canada, about half
quite long: peptides, organisms (living, dead or                this number in Europe, and a few more throughout
attenuated), gene `constructs', any type of fer-                the rest of the world. Many of these companies
mented product (even when these may also be syn-                are small organizations. Small companies' research
thesized chemically), and antisense compounds. To               activities may be restricted to the preclinical dis-
date, peptides have been the largest of these prod-             covery and investigation of compounds; therefore
uct groups, and themselves comprise such diverse                their business environment and practices differ
biological agents as hormones, antibodies, cyto-                from large, fully integrated `pharmaceutical'
kines (including interferons), and immune adju-                 companies. On the other hand, the less numerous
vants (including non-mammalian examples, such                   but much bigger pharmaceutical companies are all
as keyhole limpet hemocyanin).                                  engaged in biotechnology. Somewhat arbitrarily,
   Biological products have a longer and more il-               we shall use the term `biotechnology company'
lustrious history that is generally assumed: at one             rather loosely in this chapter to mean the typical
time smallpox accounted for 10% of deaths in some               small organization. The term `biotechnology
countries, but today this is the only infectious dis-           products' refers to the compounds themselves, re-
ease ever to have been eradicated from our planet;              gardless of the size of the organization developing
the development of cow pox vaccination was in                   them.

       REGULATORY CONSIDERATIONS                         example, that hormones are always `orthodox'
                                                         drugs, while HIV diagnostics are `biologicals'.
In most countries regulation of drug and biological      `Blood-devices' and `blood-drugs' are handled by
compound development and marketing has usually           both the drug and biological divisions within FDA.
derived from governmental response to crisis.              However, regulators' approaches and practices
                                                         for drugs and biologics are slowly converging. In
                                                         the USA and elsewhere, evidence of this conver-
                 USA Perspective                         gence may be found in:

The Kefauver±Harris Act in the USA, after the            . Common IND regulations (there are no unique
thalidomide catastrophe, is a good example of re-          regulations for biologics undergoing experi-
sponse to crisis. Such post hoc legislation results in     mental study).
differential oversight of different types of agents.     . Good clinical practices (GCP).
   The initial legislation affecting biologics was the   . International harmonization.
Safe Vaccines and Sera Act of 1904, the focus of         . Good manufacturing practice (GMP).
which was the development of safe, pure, and             . `Fast-track' designation for accelerating review
potent vaccination preparations. This was some-            and approval.
what superseded by the Public Health Service Act         . Treatment INDs.
(1944), written with blood products and prevention       . Compassionate (emergency) INDs.
of the transmission of disease by infusion in mind.
All the previous drug legislation had been super-        However, illogicalities persist, particularly with
vised by the Department of Agriculture. It was not       new Congressional Acts. For example, the Wax-
until 1972 that orthodox drugs and biologics were        man±Hatch legislation gave authority for generics
brought under the same regulatory umbrella. The          and provided patent term exclusivity for drugs,
Food and Drug Administration (FDA), a single             but not for biologics licensed prior to 1972. Simi-
branch of the Public Health Service, then accepted       larly, the pediatric `exclusivity' that the Act initi-
the responsibility for biologics, transferred from       ated also only pertained to drugs with unexpired
the Department of Agriculture.                           patent or Orphan Drug exclusivity. Another
   Similar historical events stimulated other models     example: the Centers for Disease Control remains
in other countries. In spite of the modern uniform-      involved with some compensation issues for pedi-
ity of regulatory authority that exists in the USA,      atric vaccines, a unique administrative arrange-
much of the legislation has not been repealed or         ment.
replaced, and inconsistencies between the regula-
tions relating to biologics, drugs, and devices il-
logically persist. In the USA, regulatory oversight       BIOTECHNOLOGY VS. CONVENTIONAL
of therapeutic biological compounds (i.e. applic-                 DRUG PRODUCTS
able to the prevention, treatment, or cure of disease
or injuries of man) includes:                            There is a widespread, but largely unreasonable,
                                                         perception that biotechnology products differ in
.     Viruses.                                           their properties to conventional small molecules
.     Therapeutic sera, toxin or antitoxins.             drugs. For example, it is widely believed that
.     Vaccines.                                          simple pharmacokinetic models cannot adequately
.     Blood and its components and derivatives.          describe the behavior of biological agents in vivo.
.     Allergenic products.                               However, Table 17.1 lists the factors that need to
.     Arsphenamine and derivatives.                      be considered when modeling pharmacokinetic±
                                                         pharmacodynamic (PK±PD) relationships for
However, `jurisdictional challenges' (i.e. regula-       gene therapy products, and most of these have
tors' turf battles) continue for various compounds       correlates with the behaviour of orthodox drugs.
and combination of drugs, biologics, and devices.        While quantitative data relating to the intracellular
For no biological reason, it has been agreed, for        distribution of these agents may not be known or
                                BIOTECHNOLOGY PRODUCTS AND THEIR DEVELOPMENT                                          193

Table 17.1 Pharmacokinetic considerations for gene therapy         dimensional folding, post-translational modifica-
agents                                                             tion, or glycosylation of proteins can significantly
                                                                   alter biological activity. The potential for the repli-
Pharmacokinetic Property Gene Therapy Property
                                                                   cation of viruses or bacteria in fermenters, and
Absorption                    DNA vector distribution
                                                                   their persistence in finished drug product raise add-
Distribution                  Vector fraction target cell uptake   itional safety concerns about the manufacturing
Distribution                  Genetic material traffic in          process for such compounds. For the clinical trial-
                                organelles                         ist, this leads to a generality: when studying biolo-
Metabolism                    DNA degradation                      gicals, there is usually a greater need for early-stage
Metabolism                    mRNA production
Metabolism                    Protein productionÐquantity          test medications to be as similar as possible to the
Excretion                     Protein productionÐstability         marketed product than for `orthodox' small chem-
Metabolism, excretion         Protein productionÐ                  icals.
Excretion                     Protein productionÐsecretory fate

Adapted from Ledley and Ledley, (1994).
                                                                      PRODUCT CLASSES AND RESULTANT
                                                                           CLINICAL TRIAL ISSUES

easy to measure, the underlying principles of ab-                  Many of the principles outlined in other chapters
sorption, distribution, metabolism and excretion                   for Phase I and II studies of these compounds are
(ADME) are quite constant, even though new                         equally applicable to the testing of most biotech-
paradigms and models are required to describe                      nology products. The same basic principal of dem-
the properties of biological compounds.                            onstrating clinical tolerability as a priority over
   Most biotechnology products have complicated                    proving efficacy must apply. Other chapters also
PK±PD relationships. However, this can also                        discuss some specific toxicology and drug discov-
be the case for orthodox drugs. Consider, for                      ery aspects of biotechnology products.
example, antidepressant therapies, where 3 weeks                      The design of a development program for a bio-
or more is usually needed before any therapeutic                   logical product has the same principles as for ordin-
response may be seen. Again, angiotensin-convert-                  ary drugs, and this should be familiar to the
ing enzyme (ACE) inhibitors are characterized by                   competent clinical trialist. The development pro-
prolonged activity in chronic therapy, yet these                   gram should be determined by the nature and
drugs are still antihypertensive, even after ACE                   needs of the disease; often the pharmacological ac-
activity has compensated and been restored.                        tivity of a biological product is likely to be very
Corticosteroid therapy requires access to the cell                 precise, e.g. an antibody will bind to a previously
nucleus, as do many gene therapies. Among bio-                     identified, narrow range of antigens, and the patho-
technology products, antibody complement fix-                      genesis or source of antigen presentation will have a
ation and cellular attack may be an all-or-none                    fixed relationship to a well-described disease or set
phenomenon, DNA lysis in sputum may not actu-                      of diseases. Equally useful in the case of biologics is
ally require drug absorption at all, and clot lysis                to begin development with an agreed, desirable
may depend on a wide range of endogenous plasma                    package insert or product information leaflet.
proteins, each with its own concentration±response                 That document can then be used to define the de-
relationships. The problems and analysis of tachy-                 velopment strategy. Only those tactics (i.e. clinical
phylaxis are therefore common to both orthodox                     trial designs, milestones, and product-killing find-
and biotechnology products.                                        ings) that are justified or validated by that strategy
                                                                   should be implemented. The nature and seriousness
                                                                   of the disease being treated is just as important as in
            MANUFACTURING ISSUES                                   more orthodox clinical trialsÐthe degree of lethal-
                                                                   ity or morbidity associated with the disease treated
Manufacturing changes are more like to affect the                  with existing therapies is correlated positively with
clinical profile of biological compounds than small                the degree of intolerability of the test agent that may
chemical entities. Small changes in the three-                     be accepted.

   Likewise, the trialist who switches to a biological   specific to the known function of the hormone or
product will find it very familiar to prove efficacy     enzyme. Immunologic adverse events can, in turn,
by comparing study medication to active or placebo       be viewed as either active or passive: what the drug
comparative agents. With `breakthrough' agents           does to the patient (histamine release, B lympho-
that offer the potential for a new type of therapy,      cyte proliferation, etc.) and what the patient does
clinical trials can be conducted, comparing the trial    to the drug (enhanced clearance, peptide cleavage,
agent to a placebo. As usual, the central ethical        etc.). The clinical correlates of these cellular pro-
consideration is whether standard treatment is           cesses then range from a need for ever-increasing
being withheld from patient, and the debate should       doses to maintain biological effect, to the acute
not be about the use of placebo in an absolute sense.    emergencies of anaphylaxis. For example, around
Dose±response relationships need to be evaluated         13% of patients given aglucerase used for Type
for biological agents prior to approval, even when       1 Gaucher's disease develop IgG antibodies to the
the biological response is a `all-or-nothing' type of    enzyme; approximately 25% of these have symp-
response (e.g. serological conversion). It should be     toms of hypersensitivity.
remembered that dose-response relationships must            Clinical trialists should be careful not to ignore
be understood for populations, as well as within         the toxicological potential of the often large
individuals. Therapies like vaccines need to be          amounts of vehicle (ionic strengths, buffering ma-
evaluated in different racial populations.               terials, extreme pH) that may be required to main-
                                                         tain complex peptides in a stable form. These can
                                                         include unanticipated allergens, nephrotoxins and
                     Peptides                            hepatotoxins.
                                                            Historically, insulin is an early and classic
For a long time, interest in biotechnology centered      example of a biotechnology product, illustrating
on the production and properties of administered         the general problems that are associated with
hormones ranging from tripeptides, e.g. cortico-         peptide drugs and how modern technology leads
trophin releasing factor (CRF); through nonapep-         to improved therapy. Prior to the production of
tides, such as vasopressin analogues; to longer          human insulin by cell-based fermentation pro-
polypeptide chains, e.g. insulins and growth hor-        cesses, treatment was with pancreatic extracts of
mones. As the length of the molecule increases, the      porcine or bovine origin. Insulin resistance correl-
three-dimensional structure becomes an important         ated with effective, specific antibody responses in
determinant of in vivo activity and properties.          many diabetics, who had a `career' of increasing
   However, there are also important freedoms that       insulin dose, punctuated by hypoglycemia when
an increase in protein size can bring. Single-peptide    changing from one animal source to another with-
mutations become less important as protein size          out changing dose size. Some patients became so
increases. The scope for post-translational modifi-      competent at clearing bovine or porcine insulin
cation is also greater in large polypeptides than in     that they needed extracts from exotic species such
small ones. Good examples of this are the large          as whales. The modification of recombinant chi-
qualitative changes in pharmacology of the single        meric or pure cell lines to secrete human insulin, the
amino acid that distinguishes arginine-vasopressin       development of large-scale fermenters to multiply
(the human antidiuretic hormone) from human              such cultures, and the ability to purify cell-free
oxytocin, or the marked differences in pharmaco-         insulin from other materials in the broth, has led
dynamics between calcitonin and calcitonin gene-         to a sufficient supply of exogenous, but human,
related peptide (CGRP), which are both encoded           insulin. Now in use by almost all diabetics in the
by the same gene.                                        Western world, immune responses to this homo-
   In comparison with non-peptide, non-nucleic           biotic molecule are very rare, and dose sizes remain
acid drugs, the principal specific adverse events        more stable than before.
associated with peptides may be viewed as the               We shall now consider the principal subsets of
sum of two processes: immunological and physio-          peptide therapeutic classes within the context of
logical changes. The former is non-specific and          similarities to, and differences from, orthodox
familiar to most physicians, the latter is usually       small molecules of purely chemical synthesis.
                         BIOTECHNOLOGY PRODUCTS AND THEIR DEVELOPMENT                                     195

                    Hormones                            diseases, e.g. the prevention of rhesus immuniza-
                                                        tion in the perinatal period.
In addition to insulin, various other hormones             Therapeutically, antibodies may be targeted
made by recombinant methods have been ap-               against a variety of antigens or a single specific
proved or are under development. The most com-          antigen. As an example of a broad-spectrum anti-
monly prescribed examples at present include            body therapy, WinRho1 may be used to prevent
growth hormone (somatrem, Protropiny, Genen-            rhesus immunization in an Rh-negative mother
tech, Inc.) or erythropoietin (epoeitin a, Epogeny;     post partum. There are at least 60 known epitopes
Amgen Inc.).                                            of the rhesus D antigen, and the product, which
                                                        is made from pooled plasma of rhesus-immunized,
                                                        Rh-negative volunteers or patients, binds to
                     Enzymes                            red cells from 99.7% of all Rh-positive blood
There are a few peptide drugs that act as enzymes.         In contrast, rituximab (Rituxan1; IDEC
Dornase alpha (Pulmozyme1; Genentech Inc.) is           Pharmaceuticals) is a highly targeted antibody
an example; in both adult and pediatric patients        that binds principally with CD-20-positive B
with cystic fibrosis, the rate of DNA release from      lymphocytes implicated in one form of non-Hodg-
dead and dying lymphocytes is sufficient to signifi-    kin's lymphoma. Muromonab-CD3 (Orthoclone
cantly increase sputum viscosity. The enzyme (in-       OKT3; Ortho Pharmaceuticals), used to reverse
haled through a nebulizer) digests this released        acute rejection of transplanted kidneys, is another
DNA, liquifies the sputum, enhancing its expector-      example of the specific type of therapeutic anti-
ation, and thus improves the management of chest        body. Specific antigens are far more common as
infections and pulmonary function in these pa-          therapeutic targets.
tients.                                                    Specific targeting of single, infrequently ex-
   The clinical trials of this product were of gener-   pressed antigens forms the basis of the large
ally orthodox design. The fact that the product is      number of monoclonal antibody therapies that
made by fermentation of genetically engineered          are either in development or on the market. These
Chinese hamster ovary cells, containing DNA en-         are generally manufactured by mammalian cell fer-
coding for the native human protein, deoxyribonu-       mentation process, as described above (see Chapter
clease I (DNAase), was essentially irrelevant to the    6 for issues relating to the manufacturing process
design of the clinical trials program. Placebo-con-     and viral contamination of these products). The
trolled trials were conducted, and dose±response        presence of single-antigen targets in or on tumor
relationships were evaluated.                           cells can be further exploited by conjugating the
   Significant other enzymes in common clinical         antibody to a radioactive or cytotoxic molecule.
use, made using similar processes, including tissue     For example, human milk fat globule I monoclonal
plasminogen activator, other thrombolytic agents,       antibody, complexed with 90 Yttrium (Thera-
and aglucerase (Ceredasey, Genzyme Corp.) for           gyny), is under development for the treatment of
Type 1 Gaucher's disease, illustrating the diverse      ovarian carcinoma by Antisoma, PLC.
clinical applications that enzymes may find.

                                                        Fundamentally, cytokine responses to infection or
Since antibodies, by their nature, bind to antigens,    tumors are biologically the most primitive form of
their function is often to augment intrinsic clear-     immune response. Cytokines are generated in re-
ance mechanisms, as well as potentially exerting        sponse to antigen challenge. However, unlike anti-
definitive therapeutic effects themselves. It is not    bodies, the cytokine response is non-specific, and
surprising, therefore, that the therapeutic targets     its principal biological effect is to enhance general,
for antibody therapy are extremely broad, ranging       lymphocyte-mediated attack on the antigen-bear-
from anti-tumor therapy to specific immunological       ing cell.

   Cytokines include the large and ever-increasing      although the latter may (controversially) also be
set of interleukins, various interferons, and numer-    regarded as `autoimmune' or `congenital' when it
ous other factors, e.g. tumor necrosis factor,          is due to the derepression of prion genomes, which
angiogenesis factor II, and various growth factors.     lurk dormant in many normal mammals. The nu-
Granulocyte macrophage colony stimulating factor        merous diseases that plague tropical countries and
(GMCSF; sargramostim, Leukiney; Immunex                 the developing world provide much scope but little
Corp.) is used for myeloid reconstitution after         financial incentive to the traditional pharmaceut-
bone marrow ablation, exploiting its eponymous          ical industry. Drug resistance, occurring in numer-
property initially identified in vitro. Interleukin 2   ous microorganisms ranging from staphylococcus
(IL-2; aldesleukin, Proleukiny, Chiron Corp.) is        to malaria, is another field that could conceivably
approved in the USA for the treatment for renal         be conquered by taking the adjuvant approach.
carcinoma and metastatic melanoma. Platelet-               Not surprisingly, there is considerable interest in
derived growth factor can be used to heal diabetic      using adjuvant tactics for the prevention or treat-
foot ulcers, presumably by imitating normal physi-      ment of non-infectious disease. Spontaneous tumor
ology that is blunted in patients with diabetes (gel    regression (although rarely observed clinically) and
becaplermin, Regranex1; Ortho-McNeil/Chiron             the development of rare tumors in immunocom-
Inc.).                                                  promised patients (such as Kaposi's sarcoma in
   Since cytokines have non-specific effects,           patients with AIDS) are both consistent with
existing drugs are likely to find additional indica-    the usefulness of endogenous host mechanisms to
tions. Similarly, their adverse effects also reflect    either prevent or retard cancer. Tumor-specific
their non-specificity, with symptoms such as fever,     antigens may be used as therapeutic targets for
myalgias, flu-like symptoms and rhabdomyolysis.         exogenous therapy.
                                                           There is a broad range of mechanisms against
                                                        these targets, such as specific cytokines, tumor cell
               Immune Adjuvants                         expression of rejection antigens and inducing
                                                        lymphocyte co-stimulatory molecules on tumor
Immune adjuvants can be classed as:                     cells (vide infra). Non-specific approaches also
                                                        exist, e.g. Bacille Calmet±Guerrin vaccine (BCG),
. Non-specific, e.g. BCG vaccine for bladder            which elicits a T lymphocyte cell-mediated immune
  cancer.                                               response, can be used not only to prevent tubercu-
. Specific, e.g. Salk vaccine for polio prevention.     losis but also to (non-specifically) prevent the recur-
. Genetic, to elicit cytokine responses (vide infra).   rence of bladder cancer.

   Vaccines, used widely in medicine since Jenner's
pioneering work in preventing small pox by inocu-                        Antisense Drugs
lating intradermal cow pox virus, are probably the
best examples of the broad application of biotech-      Antisense drugs are exogenous oligonucleotides
nology products prior to the twentieth century.         that bind to specific endogenous nucleic acid se-
Traditionally, vaccines have been directed against      quences. Binding to mRNA prevents the construc-
the prevention of specific infectious diseases. At-     tion of proteins by ribosomes and similarly,
tenuated live and killed microrganisms are used as      binding to specific gene sequences on DNA pre-
antigens to elicit cellular and humoral responses.      vents both RNA coding and protein production.
They may be viewed as adjuvants, because it is the      The application of antisense technology is broad,
enhanced endogenous physiology that protects            since this approach can be used to inhibit the pro-
against the pathogen, and not the material in the       duction of a wide range of proteins, including
vaccine itself.                                         stimulatory and inhibitory molecules.
   The potential for preventing infectious disease is      While the synthesis of antisense molecules using
far from complete, and there is a continuing need       modern combinatorial chemical approaches is
for worthwhile research programs. Current chal-         easily automated, the delivery of these molecules
lenges include HIV and prion-mediated disease,          to the appropriate intracellular and intranuclear
                          BIOTECHNOLOGY PRODUCTS AND THEIR DEVELOPMENT                                      197

sites is more difficult. The first antisense drug to be     in the nucleus of cells that are successfully
approved (ISIS Pharmaceutical Inc.) was for the             infected, which need not themselves be divid-
treatment of cytomegalovirus retinitis in patients          ing.
with AIDS. The route of administration of this            . Non-pathogenic adeno-associated viruses. These
drug is unique: direct intraocular injection. While         parvoviruses carry single-stranded DNA and
this does deliver sufficient drug to the appropriate        are able to integrate into a broad range of
site, it illustrates well the ADME problems associ-         non-dividing cells.
ated with antisense drugs.                                . Retroviruses. These RNA-containing viruses
   To date, most regulatory authorities have                exist in an envelope derived from host cell mem-
treated antisense drugs in the much same way as             brane, and thus do not usually elicit vigorous
any other biological product. The additional con-           immune responses. Retroviruses also tend to
straints that apply to gene therapies (for example)         replicate only in dividing cells.
have not been imposed. Since these oligonucleo-
tides have specific binding activities, safety consid-    It is perhaps surprising that naked DNA can cause
erations are usually dependent on the potential for       gene expression. Current examples where this con-
non-specific effects of protein synthesis inhibition.     cept has been proven include genes injected into
At present, with the current limited experience,          skeletal and smooth muscle. Gold-coated DNA
there would appear to be sound in vitro methods           containing gold particles may also be inserted into
for the testing of the specificity of antisense drugs     cells by a `gene gun', where electrostatic or gas
to be predictive for their tolerability in man. Fur-      pressure-powered displacement from a plastic
thermore, when the properties of the protein that is      matrix occurs.
inhibited are discrete and consistent across individ-        Liposomal envelopes can also transport sub-
uals, then it is likely that the potential adverse        stances across cell membranes which would other-
effects will be predictable.                              wise be repelled by the hydrophilicity of the gene
                                                          construct. Liposomes may be constructed that are
                                                          either anionic or cationic, and can also be coated
                  Gene Therapy