Ophthalmic Drug

Delivery Systems

Second Edition, Revised and Expanded

edited by

Ashim K. Mitra
University of Missouri-Kansas City
Kansas City, Missouri, U.S.A.

M ARCEl

MARCEL DEKKER, INC.

Copyright 2003 Marcel Dekker, Inc.

NEW YORK BASEL

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress.
ISBN: 0-8247-4124-2
This book is printed on acid-free paper.
Headquarters
Marcel Dekker, Inc.
270 Madison Avenue, New York, NY 10016
tel: 212-696-9000; fax: 212-685-4540
Eastern Hemisphere Distribution
Marcel Dekker AG
Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
tel: 41-61-260-6300; fax: 41-61-260-6333
World Wide Web
http://www.dekker.com
The publisher oers discounts on this book when ordered in bulk quantities. For
more information, write to Special Sales/Professional Marketing at the headquarters
address above.
Copyright Cg2003 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying, microlming, and
recording, or by any information storage and retrieval system, without permission
in writing from the publisher.
Current printing (last digit):
10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

Copyright 2003 Marcel Dekker, Inc.

DRUGS AND THE PHARMACEUTICAL SCIENCES

Executive Editor

James Swarbrick
PharmaceuTech, Inc
Pmehurst, North Carolina

Advisory Board

Larry L Augsburger
University of Maryland
Baltimore, Maryland
Douwe D Breimer
Gorlaeus Laboratories
Leiden, The Netherlands

David E. Nichols
Purdue University
West Lafayette, Indiana
Stephen G. Schulman
University of Florida
Gamesville, Florida

Trevor M. Jones
The Association of the
British Pharmaceutical Industry
London, United Kingdom

Jerome P Skelly
Alexandria, Virginia

Hans E. Junginger
Leiden/Amsterdam Center
for Drug Research
Leiden, The Netherlands

Felix Theeuwes
Alza Corporation
Palo Alto, California

Vincent H L Lee
University of Southern California
Los Angeles, California

Geoffrey T. Tucker
University of Sheffield
Royal Hallamshire Hospital
Sheffield, United Kingdom

Peter G Welling
Institut de Recherche Jouveinal
Fresnes, France

Copyright 2003 Marcel Dekker, Inc.

DRUGS AND THE PHARMACEUTICAL SCIENCES

A Series of Textbooks and Monographs

1. Pharmacokmetics, Milo Gibaldi and Donald Perrier

2. Good Manufacturing Practices for Pharmaceuticals: A Plan for Total

Quality Control, Sidney H. Willig, Murray M. Tuckerman, and William
S. Hitchings IV
3. Microencapsulation, edited by J. R. Nixon
4. Drug Metabolism: Chemical and Biochemical Aspects, Bernard Testa
and Peter Jenner
5. New Drugs: Discovery and Development, edited by Alan A. Rubm

6. Sustained and Controlled Release Drug Delivery Systems, edited by

Joseph R. Robinson
7. Modern Pharmaceutics, edited by Gilbert S. Banker and Christopher
T. Rhodes

8. Prescription Drugs in Short Supply: Case Histories, Michael A.

Schwartz
9. Activated Charcoal" Antidotal and Other Medical Uses, David O.

Cooney
10. Concepts in Drug Metabolism (in two parts), edited by Peter Jenner
and Bernard Testa

11. Pharmaceutical Analysis: Modern Methods (in two parts), edited by

James W. Munson
12. Techniques of Solubilization of Drugs, edited by Samuel H. Yalkowsky

13 Orphan Drugs, edited by Fred E. Karch

14. Novel Drug Delivery Systems: Fundamentals, Developmental Concepts, Biomedical Assessments, Yie W. Chien
15. Pharmacokinetics' Second Edition, Revised and Expanded, Milo
Gibaldi and Donald Perrier
16. Good Manufacturing Practices for Pharmaceuticals: A Plan for Total

Quality Control, Second Edition, Revised and Expanded, Sidney H.

Willig, Murray M. Tuckerman, and William S. Hitchings IV
17. Formulation of Veterinary Dosage Forms, edited by Jack Blodinger

18. Dermatological Formulations. Percutaneous Absorption, Brian W.

Barry

19 The Clinical Research Process in the Pharmaceutical Industry, edited

by Gary M. Matoren

20. Microencapsulation and Related Drug Processes, Patrick B. Deasy

21. Drugs and Nutrients: The Interactive Effects, edited by Daphne A.
Roe and T. Colin Campbell
22. Biotechnology of Industrial Antibiotics, Erick J Vandamme

Copyright 2003 Marcel Dekker, Inc.

23
24
25
26

27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48

Pharmaceutical Process Validation, edited by Bernard T Loftus and

Robert A Nash
Anticancer and Interferon Agents Synthesis and Properties, edited by
Raphael M Ottenbrite and George B Butler
Pharmaceutical Statistics Practical and Clinical Applications, Sanford
Bolton
Drug Dynamics for Analytical, Clinical, and Biological Chemists,
Benjamin J Gudzmowicz, Burrows T Younkin, Jr, and Michael J
Gudzmowicz
Modern Analysis of Antibiotics, edited by Adjoran Aszalos
Solubility and Related Properties, Kenneth C James
Controlled Drug Delivery Fundamentals and Applications, Second
Edition, Revised and Expanded, edited by Joseph R Robinson and
Vincent H Lee
New Drug Approval Process Clinical and Regulatory Management,
edited by Richard A Guarmo
Transdermal Controlled Systemic Medications, edited by Yie W Chien
Drug Delivery Devices Fundamentals and Applications, edited by
Praveen Tyle
Pharmacokinetics Regulatory Industrial Academic Perspectives,
edited by Peter G Welling and Francis L S Tse
Clinical Drug Trials and Tribulations, edited by Alien E Cato
Transdermal Drug Delivery Developmental Issues and Research Initiatives, edited by Jonathan Hadgraft and Richard H Guy
Aqueous Polymeric Coatings for Pharmaceutical Dosage Forms,
edited by James W McGmity
Pharmaceutical Pelletization Technology, edited by Isaac GhebreSellassie
Good Laboratory Practice Regulations, edited by Alien F Hirsch
Nasal Systemic Drug Delivery, Yie W Chien, Kenneth S E Su, and
Shyi-Feu Chang
Modern Pharmaceutics Second Edition, Revised and Expanded,
edited by Gilbert S Banker and Christopher T Rhodes
Specialized Drug Delivery Systems Manufacturing and Production
Technology, edited by Praveen Tyle
Topical Drug Delivery Formulations, edited by David W Osborne and
Anton H Amann
Drug Stability Principles and Practices, Jens T Carstensen
Pharmaceutical Statistics Practical and Clinical Applications, Second
Edition, Revised and Expanded, Sanford Bolton
Biodegradable Polymers as Drug Delivery Systems, edited by Mark
Chasm and Robert Langer
Preclmical Drug Disposition A Laboratory Handbook, Francis L S
Tse and James J Jaffe
HPLC in the Pharmaceutical Industry, edited by Godwin W Fong and
Stanley K Lam
Pharmaceutical Bioequivalence, edited by Peter G Welling, Francis L
S Tse, and Shnkant V Dinghe

Copyright 2003 Marcel Dekker, Inc.

49 Pharmaceutical Dissolution Testing, Umesh V Banakar

50 Novel Drug Delivery Systems Second Edition, Revised and
Expanded, Y/e W Chien
51 Managing the Clinical Drug Development Process, David M Cocchetto and Ronald V Nardi
52 Good Manufacturing Practices for Pharmaceuticals A Plan for Total
Quality Control, Third Edition edited by Sidney H Willig and James
R Stoker
53 Prodrugs Topical and Ocular Drug Delivery, edited by Kenneth B
Sloan
54 Pharmaceutical Inhalation Aerosol Technology, edited by Anthony J
Hickey
55 Radiopharmaceuticals Chemistry and Pharmacology, edited by
Adrian D Nunn
56 New Drug Approval Process Second Edition, Revised and Expanded,
edited by Richard A Guarmo
57 Pharmaceutical Process Validation Second Edition, Revised and Expanded, edited by Ira R Berry and Robert A Nash
58 Ophthalmic Drug Delivery Systems, edited byAshim K Mttra
59 Pharmaceutical Skin Penetration Enhancement, edited by Kenneth A
Walters and Jonathan Hadgraft
60 Colonic Drug Absorption and Metabolism, edited by Peter R Bieck
61 Pharmaceutical Particulate Carriers Therapeutic Applications, edited
by Alam Rolland
62 Drug Permeation Enhancement Theory and Applications, edited by
Dean S Hsieh
63 Glycopeptide Antibiotics, edited by Ramaknshnan Nagarajan
64 Achieving Sterility in Medical and Pharmaceutical Products, Nigel A
Halls
65 Multiparticulate Oral Drug Delivery, edited by Isaac Ghebre-Sellassie
66 Colloidal Drug Delivery Systems, edited by Jorg Kreuter
67 Pharmacokmetics Regulatory Industrial Academic Perspectives,
Second Edition, edited by Peter G Welling and Francis L S Tse
68 Drug Stability Principles and Practices, Second Edition, Revised and
Expanded, Jens T Carstensen
69 Good Laboratory Practice Regulations Second Edition, Revised and
Expanded, edited by Sandy Wemberg
70 Physical Characterization of Pharmaceutical Solids, edited by Harry
G Bnttain
71 Pharmaceutical Powder Compaction Technology, edited by Goran Alderborn and Chnster Nystrom
72 Modern Pharmaceutics Third Edition, Revised and Expanded, edited
by Gilbert S Banker and Christopher T Rhodes
73 Microencapsulation Methods and Industrial Applications, edited by
Simon Benita
74 Oral Mucosal Drug Delivery, edited by Michael J Rathbone
75 Clinical Research in Pharmaceutical Development, edited by Barry
Bleidt and Michael Montagne

Copyright 2003 Marcel Dekker, Inc.

76. The Drug Development Process: Increasing Efficiency and Cost Effectiveness, edited by Peter G Welling, Louis Lasagna, and Umesh
V. Banakar

77 Microparticulate Systems for the Delivery of Proteins and Vaccines,

edited by Smadar Cohen and Howard Bernstein
78. Good Manufacturing Practices for Pharmaceuticals: A Plan for Total
Quality Control, Fourth Edition, Revised and Expanded, Sidney H.
Willig and James R Stoker
79. Aqueous Polymeric Coatings for Pharmaceutical Dosage Forms:
Second Edition, Revised and Expanded, edited by James W
McGimty
80. Pharmaceutical Statistics: Practical and Clinical Applications, Third
Edition, Sanford Bolton
81 Handbook of Pharmaceutical Granulation Technology, edited by Dilip
M. Parikh
82. Biotechnology of Antibiotics' Second Edition, Revised and Expanded,

edited by William R. Strohl

83 Mechanisms of Transdermal Drug Delivery, edited by Russell O Potts
and Richard H. Guy
84. Pharmaceutical Enzymes, edited by Albert Lauwers and Simon
Scharpe
85. Development of Biopharmaceutical Parenteral Dosage Forms, edited
by John A Bontempo
86. Pharmaceutical Project Management, edited by Tony Kennedy
87 Drug Products for Clinical Trials. An International Guide to Formulation . Production Quality Control, edited by Donald C. Monkhouse
and Christopher T. Rhodes

88 Development and Formulation of Veterinary Dosage Forms: Second

Edition, Revised and Expanded, edited by Gregory E. Hardee and J.

Desmond Baggot
89. Receptor-Based Drug Design, edited by Paul Left
90. Automation and Validation of Information in Pharmaceutical Processing, edited by Joseph F deSpautz
91. Dermal Absorption and Toxicity Assessment, edited by Michael S
Roberts and Kenneth A Walters
92. Pharmaceutical Experimental Design, Gareth A. Lewis, Didier
Mathieu, and Roger Phan-Tan-Luu
93 Preparing for FDA Pre-Approval Inspections, edited by Martin D.

Hynes III
94. Pharmaceutical Excipients. Characterization by IR, Raman, and NMR
Spectroscopy, David E Bugay and W. Paul Findlay
95 Polymorphism in Pharmaceutical Solids, edited by Harry G. Brittam
96 Freeze-Drymg/Lyophilization of Pharmaceutical and Biological Prod-

ucts, edited by Louis Rey and Joan C May

97. Percutaneous Absorption- Drugs-Cosmetics-Mechanisms-Methodology, Third Edition, Revised and Expanded, edited by Robert L.
Bronaugh and Howard I. Maibach

Copyright 2003 Marcel Dekker, Inc.

98. Bioadhesive Drug Delivery Systems: Fundamentals, Novel Approaches, and Development, edited by Edith Mathiowitz, Donald E.
Chickering III, and Claus-Michael Lehr
99 Protein Formulation and Delivery, edited by Eugene J McNally

100. New Drug Approval Process: Third Edition, The Global Challenge,
edited by Richard A. Guanno
101. Peptide and Protein Drug Analysis, edited by Ronald E. Reid
102. Transport Processes in Pharmaceutical Systems, edited by Gordon L.
Amidon, Ping I. Lee, and Elizabeth M. Topp
103. Excipient Toxicity and Safety, edited by Myra L Wemer and Lois A.
Kotkoskie
104. The Clinical Audit in Pharmaceutical Development, edited by Michael
R. Hamrell
105 Pharmaceutical Emulsions and Suspensions, edited by Francoise
Nielloud and Gilberte Marti-Mestres
106. Oral Drug Absorption: Prediction and Assessment, edited by Jennifer B.
Dressman and Hans Lennernas
107. Drug Stability: Principles and Practices, Third Edition, Revised and

Expanded, edited by Jens T. Carstensen and C T. Rhodes

108. Containment in the Pharmaceutical Industry, edited by James P.
Wood
109. Good Manufacturing Practices for Pharmaceuticals. A Plan for Total
Quality Control from Manufacturer to Consumer, Fifth Edition, Revised
and Expanded, Sidney H. Willig
110. Advanced Pharmaceutical Solids, Jens T. Carstensen
111. Endotoxms: Pyrogens, LAL Testing, and Depyrogenation, Second
Edition, Revised and Expanded, Kevin L. Williams
112. Pharmaceutical Process Engineering, Anthony J. Hickey and David

Ganderton
113. Pharmacogenomics, edited by Werner Kalow, Urs A. Meyer, and Rachel F. Tyndale
114. Handbook of Drug Screening, edited by Ramakrishna Seethala and
Prabhavathi B. Fernandes
115. Drug Targeting Technology Physical Chemical Biological Methods,
edited by Hans Schreier
116. Drug-Drug Interactions, edited by A. David Rodngues
117. Handbook of Pharmaceutical Analysis, edited by Lena Ohannesian
and Anthony J. Streeter

118. Pharmaceutical Process Scale-Up, edited by Michael Levin

119. Dermatological and Transdermal Formulations, edited by Kenneth A.
Walters
120. Clinical Drug Trials and Tribulations. Second Edition, Revised and

Expanded, edited by Alien Cato, Lynda Sutton, and Alien Cato III
121. Modern Pharmaceutics: Fourth Edition, Revised and Expanded, edited by Gilbert S. Banker and Christopher T. Rhodes

122. Surfactants and Polymers in Drug Delivery, Martin Malmsten

123 Transdermal Drug Delivery: Second Edition, Revised and Expanded,
edited by Richard H Guy and Jonathan Hadgraft

Copyright 2003 Marcel Dekker, Inc.

124 Good Laboratory Practice Regulations Second Edition, Revised and

Expanded, edited by Sandy Wemberg
125. Parenteral Quality Control: Sterility, Pyrogen, Particulate, and Package Integrity Testing- Third Edition, Revised and Expanded, Michael
J Akers, Daniel S Larnmore, and Dana Morion Guazzo
126 Modified-Release Drug Delivery Technology, edited by Michael J
Rathbone, Jonathan Hadgraft, and Michael S. Roberts
127 Simulation for Designing Clinical Trials A Pharmacokmetic-Pharmacodynamic Modeling Perspective, edited by Hui C Kimko and Stephen B. Duffull
128 Affinity Capillary Electrophoresis in Pharmaceutics and Biopharmaceutics, edited by Remhard H H. Neubert and Hans-Hermann Ruttinger
129. Pharmaceutical Process Validation: An International Third Edition,
Revised and Expanded, edited by Robert A. Nash and Alfred H.

Wachter
130 Ophthalmic Drug Delivery Systems Second Edition, Revised and
Expanded, edited by Ashim K Mitra
131 Pharmaceutical Gene Delivery Systems, edited by Alain Rolland and
Sean M. Sullivan

ADDITIONAL VOLUMES IN PREPARATION

Biomarkers in Clinical Drug Development, edited by John Bloom

Pharmaceutical Inhalation Aerosol Technology Second Edition, Revised and Expanded, edited by Anthony J Mickey
Pharmaceutical Extrusion Technology, edited by Isaac Ghebre-Sellassie and Charles Martin
Pharmaceutical Compliance, edited by Carmen Medina

Copyright 2003 Marcel Dekker, Inc.

Foreword

For new medications to be used eectively, and for those now available to
provide maximal benet, improvements in ocular drug delivery are essential.
Drug delivery is no less vital than drug discovery.
Although many drugs can be safely delivered by eye drops, eective
treatment depends on patient compliance. Non-compliance is a major
problem, especially in poorly educated patients and patients who are required to apply drops frequently. Lack of compliance frequently results in
suboptimal therapeutics, which may lead to blindness. People with chronic
conditions or debilitating disease nd complicated eye drop regimens to
be a serious handicap.
Even when drugs can be delivered through the cornea and conjunctiva,
concentrations may be suboptimal and the therapeutic eect minimal. In the
past, a variety of approaches to topical drug delivery have been tested,
including gelatin wafers or soft contact lenses soaked in drugs and placed
on the cornea or in the cul-de-sac, corneal collagen shields, and iontophoresis. The diversity of these approaches is an indication of the need for a
superior method of topical drug delivery and a testament to the fact that no
uniformly acceptable method has been developed to date. Currently, vehicles and carriers such as liposomes and substances that gel, as well as nanoparticles, are being evaluated. Also, prodrugs, such as medicines that
hydrolyze within the eye, are being developed to achieve higher concentrations, prolonged activity, and reduced toxicity of topically applied medications. These important techniques and others are considered in this book.
iii

Copyright 2003 Marcel Dekker, Inc.

iv

Foreword

Perhaps even more important than surface delivery is the need to

apply medications to the posterior segment of the eye. Treatment of blinding
posterior segment diseases, including uveitis, proliferative retinopathy, and
macular degeneration, requires drug delivery to the retina, the choroid, or
the ciliary body in a safe and convenient way. Systemic delivery that can
localize to the retina may be possible. Improving scleral permeability may be
important for periocular delivery, and devices inserted into the vitreous have
certainly been valuable. Both nonbiodegradable controlled-release devices
and biodegradable implants inserted into both aqueous and vitreous show
great promise.
Posterior segment drug delivery is also becoming important for gene
therapy. The need to deliver polypeptide medications and DNA inhibitors
has become clear. The challenge of understanding the pharmacokinetics of
the drug is matched by the challenge of providing a delivery system that can
provide optimal duration of drug delivery in therapeutically sucient concentrations and still be safe and convenient for the patient.
Our approaches to these goals are imperfect at present, but this critically important book describes in vital detail and with great clarity the
progress that has been made so far and the course that needs to be pursued
in the future. In my pharmacological memory, it does not seem so long ago
that we had no treatment for viral diseases, pilocarpine was the only treatment for glaucoma, and antibiotics were crude and relatively ineective.
Similarly, our present achievements in the eld of ocular drug delivery
may seem equally primitive as we follow the paths to future progress
detailed in this book.
Herbert E. Kaufman, M.D.
Boyd Professor of Ophthalmology, Pharmacology, and Microbiology
Louisiana State University Health Sciences Center
New Orleans, Louisiana, U.S.A.

Copyright 2003 Marcel Dekker, Inc.

Preface

A major goal of pharmacotherapeutics is the attainment of an eective drug

concentration at the intended site of action for a desired length of time.
Ecient delivery of a drug while minimizing its systemic and/or local side
eects is the key to the treatment of ocular diseases. The unique anatomy
and physiology of the eye oer many challenges to developing eective
ophthalmic drug delivery systems, but the knowledge in this eld is rapidly
expanding. Systems range from simple solutions to novel delivery systems
such as biodegradable polymeric systems, corneal collagen shields, iontophoresis, and viral and nonviral gene delivery systems, to name a few. An
increase in our understanding of ocular drug absorption and disposition
mechanisms has led to the development of many of these new systems.
The rst edition of this book laid the foundation necessary for understanding barriers to ophthalmic drug delivery and to review the conventional systems available and/or in various stages of research and
development. Since then, signicant advances have been made in understanding the molecular mechanisms involved in ocular drug transport.
The book begins with a brief discussion on the anatomy and physiology
of the eye relevant to ocular drug delivery. The latest techniques, such as
microdialysis, and models developed to study ocular drug disposition are
discussed. A review of both the conventional and novel delivery systems
follows. The book stresses the fact that simple instillation of drug solution
in the cul-de-sac is not always acceptable and emphasizes the need for the
development of newer and more ecient systems. The book concludes with
v

Copyright 2003 Marcel Dekker, Inc.

vi

Preface

the basic information required for pharmaceutical scientists to protect their

inventions.
Part I investigates the fundamental considerations in ocular drug
delivery. The three chapters in this part review the relevant ocular anatomy
and physiology, the constraints imposed by the eye upon successful delivery,
and the associated ion and solute transport processes in the eye. They provide information on the various transport processes as well as recently
identied drug eux pumps, which regulate the transport of endogenous
and exogenous substances.
Part II opens with a discussion of pharmacokinetics relevant to ocular
drug delivery. The next chapter discusses the pharmacokinetic processes
guiding the ocular disposition and expands on the pharmacokinetic/pharmacodynamic modeling processes to determine the appropriate dosage regimen. This chapter is followed by a detailed discussion of the various
mathematical models developed to describe the distribution and elimination
of drugs from the vitreous. This part also includes chapters dealing with the
application of microdialysis technique to study ocular drug delivery and
disposition, and the applicability of the microdialysis sampling approach
for the examination of ocular pharmacokinetics and dynamics of ophthalmics.
Part III is divided into conventional and advanced drug delivery sysstems. The rst section deals with such conventional systems as collagen
shields, iontophoresis, microparticulates, and dendrimers. These chapters
have been updated to include advances in ocular drug delivery achieved in
the past decade. The second section examines the delivery of macromolecules to treat various ocular pathologies. The reader will nd more information on the recent developments in animal models of retino-choroidal
diseases. The viral and nonviral gene delivery systems introduced in this
section are still in their infancy but have the potential to provide enormous
therapeutic benets. This section also focuses on the advances in treating
retinal degenerative diseases. The last chapter in this section discusses
the principles and delivery aspects of gene, oligonucleotide, and ribozyme
therapy.
Part IV provides information on regulatory and patent considerations.
Pharmaceutical scientists will gain knowledge of the regulations governing
animal and human testing and ultimately the release of the product commercially for public use. The nal chapter conveys the legal issues involved
in protecting inventions and the basic legal requirements for obtaining
patents.
Ashim K. Mitra

Copyright 2003 Marcel Dekker, Inc.

Contents

Foreword
Herbert E. Kaufman, M.D.
Preface
Contributors

iii
v
xi

I. Fundamental Considerations
1.

Overview of Ocular Drug Delivery

Sreeraj Macha, Patrick M. Hughes, and Ashim K. Mitra

2.

Membrane Transport Processes in the Eye

Gangadhar Sunkara and Uday B. Kompella

13

3.

General Considerations in Ocular Drug Delivery

James E. Chastain

59

II. Transport Models in Ocular Drug Delivery

4.

Ocular Drug Transfer Following Systemic Drug Administration 109

Nelson L. Jumbe and Michael H. Miller

vii

Copyright 2003 Marcel Dekker, Inc.

viii

Contents

5. Ocular Pharmacokinetics and Pharmacodynamics

Ronald D. Schoenwald
6.

Mathematical Modeling of Drug Distribution in the Vitreous

Humor
Stuart Friedrich, Bradley Saville, and Yu-Ling Cheng

135

181

7.

Anterior Segment Microdialysis

Kay D. Rittenhouse

223

8.

Posterior Segment Microdialysis

Sreeraj Macha and Ashim K. Mitra

251

III. Drug Delivery Systems

A:
9.

Conventional Systems
Ocular Penetration Enhancers
Thomas Wai-Yip Lee and Joseph R. Robinson

281

10.

Corneal Collagen Shields for Ocular Drug Delivery

Shiro Higaki, Marvin E. Myles, Jeannette M. Loutsch,
and James M. Hill

309

11.

The Noncorneal Route in Ocular Drug Delivery

Imran Ahmed

335

12.

Ocular Iontophoresis
Marvin E. Myles, Jeannette M. Loutsch, Shiro Higaki,
and James M. Hill

365

13.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

Thomas P. Johnston, Clapton S. Dias, Hemant Alur,
and Ashim K. Mitra

409

14.

Microparticles and Nanoparticles in Ocular Drug Delivery

Murali K. Kothuri, Swathi Pinnamaneni, Nandita G. Das,
and Sudip K. Das

437

Copyright 2003 Marcel Dekker, Inc.

Contents

15.

Dendrimers: An Innovative and Enhanced Ocular Drug

Delivery System
Jeannette M. Loutsch, Desiree Ong, and James M. Hill

ix

467

B:

Delivery of Macromolecular Therapeutic Agents

16.

Ocular Delivery and Therapeutics of Proteins and Peptides

Surajit Dey, Ramesh Krishnamoorthy, and Ashim K. Mitra

17.

Retinal Disease Models for Development of Drug and Gene

Therapies
Leena Pitkanen, Lotta Salminen, and Arto Urtti

515

New Experimental Therapeutic Approaches for Degenerative

Diseases of the Retina
Joyce Tombran-Tink

535

18.

19.

Gene, Oligonucleotide, and Ribozyme Therapy in the Eye

Sudip K. Das and Keith J. Miller

493

609

IV. Regulatory Aspects

20.

Regulatory Considerations
Robert E. Roehrs and D. Scott Krueger

663

21.

Patent Considerations
Robert E. Roehrs

695

Index

719

Copyright 2003 Marcel Dekker, Inc.

Contributors

Imran Ahmed, Ph.D.

Connecticut, U.S.A.

Pzer Global R&DGroton Laboratories, Groton,

Hemant Alur, Ph.D.

U.S.A.

Murty Pharmaceuticals, Inc., Lexington, Kentucky,

James E. Chastain, Ph.D.

U.S.A.

Alcon Research, Ltd., Forth Worth, Texas,

Yu-Ling Cheng, Ph.D. Department of Chemical Engineering and Applied

Chemistry, University of Toronto, Toronto, Canada
Nandita G. Das Department of Pharmaceutical Sciences, Idaho State
University, Pocatello, Idaho, U.S.A.
Sudip K. Das, Ph.D. Department of Pharmaceutical Sciences, Idaho State
University, Pocatello, Idaho, U.S.A
Surajit Dey Department of Pharmaceutical Sciences, University of
MissouriKansas City, Kansas City, Missouri, U.S.A
xi

Copyright 2003 Marcel Dekker, Inc.

xii

Contributors

Clapton S. Dias Department of Pharmaceutical Sciences, University of

MissouriKansas City, Kansas City, Missouri, U.S.A
Stuart Friedrich, Ph.D. Department of Chemical Engineering and Applied
Chemistry, University of Toronto, Toronto, Canada
Shiro Higaki, M.D. Department of Ophthalmology, LSU Eye and Vision
Center of Excellence, Louisiana State University Health Science Center,
New Orleans, Louisiana, U.S.A.
James M. Hill, Ph.D. Department of Ophthalmology, LSU Eye and
Vision Center of Excellence, Louisiana State University Health Science
Center, New Orleans, Louisiana, U.S.A.
Patrick M. Hughes, Ph.D.
U.S.A.

Allergan Pharmaceuticals, Irvine, California,

Thomas P. Johnston, Ph.D. Department of Pharmaceutical Sciences,

University of MissouriKansas City, Kansas City, Missouri, U.S.A
Nelson L. Jumbe Albany Medical College, Albany, New York, and Amgen
Inc., Thousands Oaks, California, U.S.A.
Uday B. Kompella, Ph.D. Department of Pharmaceutical Sciences and
Ophthalmology, University of Nebraska Medical Center, Omaha,
Nebraska, U.S.A
Murali K. Kothuri Department of Pharmaceutical Sciences, Idaho State
University, Pocatello, Idaho, U.S.A.
Ramesh Krishnamoorthy, Ph.D. Formulation Development,
Pharmaceuticals, Durham, North Carolina, U.S.A.
D. Scott Krueger, Ph.D.
Worth, Texas, U.S.A.

Inspire

Regulatory Affairs, Alcon Research, Ltd., Fort

Thomas Wai-Yip Lee, B. Pharm. School of Pharmacy, University of

WisconsinMadison, Madison, Wisconsin, U.S.A.
Jeannette M. Loutsch, Ph.D. Department of Ophthalmology, LSU Eye
and Vision Center of Excellence, Louisiana State University Health
Science Center, New Orleans, Louisiana, U.S.A.

Copyright 2003 Marcel Dekker, Inc.

Contributors

xiii

Sreeraj Macha, Ph.D.
Department of Pharmaceutical Sciences,

University of MissouriKansas City, Kansas City, Missouri, U.S.A.
Michael H. Miller
Keith J. Miller
U.S.A.

Albany Medical College, Albany, New York, U.S.A.

BristolMyers Squibb Company, Pennington, New Jersey,

Ashim K. Mitra, Ph.D. Department of Pharmaceutical Sciences,

University of MissouriKansas City, Kansas City, Missouri, U.S.A.
Marvin E. Myles, Ph.D. Department of Ophthalmology, LSU Eye and
Vision Center of Excellence, Louisiana State University Health Science
Center, New Orleans, Louisiana, U.S.A.
Desiree Ong Department of Ophthalmology, LSU Eye and Vision Center
of Excellence, Louisiana State University Health Science Center, New
Orleans, Louisiana, U.S.A.
Swathi Pinnamaneniy Department of Pharmaceutical Sciences, Idaho
State University, Pocatello, Idaho, U.S.A.
Leena Pitkanen, Lic. Med. Department of Pharmaceutics, University of
Kuopio, and Department of Ophthalmology, Kuopio University Hospital,
Kuopio, Finland
Kay D. Rittenhouse, Ph.D. Nonclinical Drug Safety, Global Research and
Development, La Jolla Laboratories, Pzer Inc., San Diego, California,
U.S.A.
Joseph R. Robinson, Ph.D. School of Pharmacy, University of Wisconsin
Madison, Madison, Wisconsin, U.S.A.
Robert E. Roehrs, Ph.D.z
U.S.A.

Alcon Research, Ltd., Fort Worth, Texas,

* Current aliation: Boehringer Ingelheim Inc., Ridgeeld, Connecticut, U.S.A.

y Current aliation: Exploratory Biopharmaceutics and Stability, Bristol-Myers Squibb
Company, New Brunswick, New Jersey, U.S.A.
z Retired.

Copyright 2003 Marcel Dekker, Inc.

xiv

Contributors

Lotta Salminen, M.D. Department of Ophthalmology, University of

Tampere, and Department of Ophthalmology, Tampere University
Hospital, Tampere, Finland
Bradley Saville, Ph.D. Department of Chemical Engineering and Applied
Chemistry, University of Toronto, Toronto, Canada
Ronald D. Schoenwald, Ph.D. College of Pharmacy, The University of
Iowa, Iowa City, Iowa, U.S.A.
Gangadhar Sunkara, Ph.D
Department of Pharmaceutical Sciences,
University of Nebraska Medical Center, Omaha, Nebraska, U.S.A.
Joyce Tombran-Tink, Ph.D. Department of Pharmaceutical Sciences,
University of MissouriKansas City, Kansas City, Missouri, U.S.A.
Arto Urtti, Ph.D.
Kuopio, Finland

Department of Pharmaceutics, University of Kuopio,

Current aliation: Clinical Pharmacology Division, Novartis Pharmaceuticals, East Hanover,

New Jersey, U.S.A.

Copyright 2003 Marcel Dekker, Inc.

1
Overview of Ocular Drug Delivery
Sreeraj Macha
and Ashim K. Mitra
University of MissouriKansas City, Kansas City, Missouri, U.S.A.

Patrick M. Hughes
Allergan Pharmaceuticals, Irvine, California, U.S.A.

I.

INTRODUCTION

Opthalmic drug delivery is one of the most interesting and challenging

endeavors facing the pharmaceutical scientist. The anatomy, physiology,
and biochemistry of the eye render this organ highly impervious to foreign
substances. A signicant challenge to the formulator is to circumvent the
protective barriers of the eye without causing permanent tissue damage.
Development of newer, more sensitive diagnostic techniques and novel therapeutic agents continue to provide ocular delivery systems with high therapeutic ecacy. Potent immunosuppressant therapy in transplant patients
and the developing epidemic of acquired immunodeciency syndrome have
generated an entirely new population of patients suering virulent uveitis
and retinopathies. Conventional ophthalmic solution, suspension, and ointment dosage forms no longer constitute optimal therapy for these indications. Research and development eorts to design better therapeutic systems
particularly targeted to posterior segment are the primary focus of this text.
The goal of pharmacotherapeutics is to treat a disease in a consistent
and predictable fashion. An assumption is made that a correlation exists
between the concentration of a drug at its intended site of action and the
resulting pharmacological eect. The specic aim of designing a therapeutic
system is to achieve an optimal concentration of a drug at the active site for

Current aliation: Boehringer Ingelheim Inc., Ridgeeld, Connecticut, U.S.A.

Copyright 2003 Marcel Dekker, Inc.

Macha et al.

the appropriate duration. Ocular disposition and elimination of a therapeutic agent is dependent upon its physicochemical properties as well as the
relevant ocular anatomy and physiology (1). A successful design of a drug
delivery system, therefore, requires an integrated knowledge of the drug
molecule and the constraints oered by the ocular route of administration.
The active sites for the antibiotics, antivirals, and steroids are the
infected or inamed areas within the anterior as well as the posterior segments of the eye. Receptors for the mydriatics and miotics are in the iris
ciliary body. A host of dierent tissues are involved, each of which may pose
its own challenge to the formulator of ophthalmic delivery systems. Hence,
the drug entities need to be targeted to many sites within the globe.
Historically, the bulk of the research has been aimed at delivery to the
anterior segment tissues. Only recently has research been directed at delivery
to the tissues of the posterior globe (the uveal tract, vitreous, choroid, and
retina).
The aim of this chapter is merely to present the challenges of designing
successful ophthalmic delivery systems by way of introduction. The reader is
referred to specic chapters within this book for a thorough discussion of
the topic introduced in this section.

II.

MECHANISMS OF OCULAR DRUG ABSORPTION

Topical delivery into the cul-de-sac is, by far, the most common route of
ocular drug delivery. Adsorption from this site may be corneal or noncorneal. A schematic diagram of the human eye is depicted in Figure 1. The socalled noncorneal route of absorption involves penetration across the sclera
and conjunctiva into the intraocular tissues. This mechanism of absorption
is usually nonproductive, as drug penetrating the surface of the eye beyond
the corneal-scleral limbus is taken up by the local capillary beds and
removed to the general circulation (2). This noncorneal absorption in general precludes entry into the aqueous humor.
Recent studies, however, suggest that noncorneal route of absorption
may be signicant for drug molecules with poor corneal permeability.
Studies with inulin (3), timolol maleate (3), gentamicin (4), and prostaglandin PGF2
(5) suggest that these drugs gain intraocular access by diusion
across the conjunctiva and sclera. Ahmed and Patton (3) studied the noncorneal absorption of inulin and timolol maleate. Penetration of these
agents into the intraocular tissues appears to occur via diusion across
the conjunctiva and sclera and not through reentry from the systemic circulation or via absorption into the local vasculature. Both compounds
gained access to the irisciliary body without entry into the anterior cham-

Copyright 2003 Marcel Dekker, Inc.

Overview of Ocular Drug Delivery

Figure 1 Anatomical structure of the human eye. (From Ref. 12.)

ber. As much as 40% of inulin absorbed into the eye was determined to be
the result of noncorneal absorption.
The noncorneal route of absorption may be signicant for poorly
cornea-permeable drugs; however, corneal absorption represents the major
mechanism of absorption for most therapeutic entities. Topical absorption of
these agents, then, is considered to be rate limited by the cornea. The anatomical structures of the cornea exert unique dierential solubility requirements for drug candidates. Figure 2 illustrates a cross-sectional view of the
cornea. In terms of transcorneal ux of drugs, the cornea can be viewed as a
trilaminate structure consisting of three major diusional barriers: epithelium, stroma, and endothelium. The epithelium and endothelium contain
on the order of 100-fold the amount of lipid material per unit mass of the
stroma (6). Depending on the physiochemical properties of the drug entity,
the diusional resistance oered by these tissues varies greatly (7,8).

Copyright 2003 Marcel Dekker, Inc.

Macha et al.

Figure 2 Cross-sectional view of the corneal membrane depicting various barriers

to drug absorption. (From Ref. 12.)

The outermost layer, the epithelium, represents the rate-limiting barrier for transcorneal diusion of most hydrophilic drugs. The epithelium is
composed of ve to seven cell layers. The basement cells are columnar in
nature, allowing for minimal paracellular transport. The epithelial cells,
however, narrow distal to Bowmans membrane, forming attened epithelial
cells with zonulae occludentes interjunctional complexes. This cellular
arrangement precludes paracellular transport of most ophthalmic drugs
and limits lateral movement within the anterior epithelium (9). Corneal
surface epithelial intracellular pore size has been estimated to be about
60 A (10). Small ionic and hydrophilic molecules appear to gain access to
the anterior chamber through these pores (11); however, for most drugs,
paracellular transport is precluded by the interjectional complexes. In a
recent review, Lee (10) discusses an attempt to transiently alter the epithelial
integrity at these junctional complexes to improve ocular bioavailability.
This approach has, however, only met with moderate success and has the
potential to severely compromise the corneal integrity.
Sandwiched between the corneal epithelium and endothelium is the
stroma (substantia propia). The stroma constitutes 8590% of the total
corneal mass and is composed of mainly of hydrated collagen (12). The
stroma exerts a diusional barrier to highly lipophilic drugs owing to its
hydrophilic nature. There are no tight junction complexes in the stroma, and
paracellular transport through this tissue is possible.

Copyright 2003 Marcel Dekker, Inc.

Overview of Ocular Drug Delivery

The innermost layer of the cornea, separated from the stroma by

Descermets membrane, is the endothelium. The endothelium is lipoidal in
nature; however, it does not oer a signicant barrier to the transcorneal
diusion of most drugs. Endothelial permeability depends solely on molecular weight and not on the charge of hydrophilic nature of the compound
(13,14).
Transcellular transport across the corneal epithelium and stroma is the
major mechanism of ocular absorption of topically applied ophthalmic
pharmaceuticals. This type of Fickian diusion is dependent upon many
factors, i.e., surface area, diusivity, the concentration gradient established,
and the period over which concentration gradient can be maintained. A
parabolic relationship between octanol/water partition coecient and corneal permeability has been described for many drugs (1519). The optimal
log partition coecient appears to be in the range of 13. The permeability
coecients of 11 steroids were determined by Schoenwald and Ward (15).
The permeability versus log partition coecient t the typical parabolic
relationship, with the optimum log partition coecient being 2.9.
Narurkar and Mitra studied a homologous series of 50 aliphatic esters of
5-iodo-20 -deoxyuridine (IDU) (16,17). In vitro corneal permeabilities were
optimized at a log partition coecient of 0.88, as can be seen graphically in
Figure 3 and in Table 1, where CMP represents the corneal permeability
values as measured by in vitro perfusion experiments on rabbit corneas (I =
IDU, II = IDU-propionate, III = IDU-butyrate, IV = IDU-isobutyrate,

Figure 3 A plot depicting the parabolic relationship between in vitro CMP and
ester chain length. (From Ref. 16.)

Copyright 2003 Marcel Dekker, Inc.

Macha et al.

Table 1

Physicochemical Properties of IDU and Its 50 -Ester Prodrugs

Compound
I
II
III
IV
V
VI
a

m.p. (8C)

Solubilitya in pH 7.4
phosphate buer, 258C
(M/L  SD [103 ])

168171 (dec)
167168
145146
144145
142143
106107

5.65
3.48
1.45
1.75
0.40
0.44

(0.5)
(0.3)
(0.1)
(0.3)
(0.2)
(0.1)

Ka  SD
(octanol/water)
0.11
4.77
7.50
6.92
27.54
22.10

(0.02)
(0.1)
(0.3)
(0.8)
(2.0)
(1.5)

N = 3.
See text for compound identication.

V = IDU-valerate, VI = IDU-pivalate). A homologous series of n-alkyl-paminobenzoate esters in a study of Mosher and Mikkelson t the parabolic
relationship displaying optimal permeability at a log partition coecient of
2.5 (18). Maximizing bioavailability of ophthalmic mediations, then,
requires that the active compound be neither extremely hydrophilic nor
lipophilic. To this end, the pH of the postinstillation precorneal uid
becomes an important factor. The postinstillation pH time course will be
dictated by the buer concentration of the formulation. Most ophthalmic
formulations are formulated in the pH range of 56; hence, depending on
the pKa of the drug to be administered, the postinstillation buering capacity of the formulation may greatly aect the drugs bioavailability. Mitra
and Mikkelson studied the eect of varying the concentration of citrate
buer in a pH 4.5 formulation on the miosis versus time prole of a 1%
pilocarpine solution (20). The area under the miosis-time prole, maximum
pupillary response, and duration of mitotic activity were all decreased with
increasing buer concentrations. Figure 4 displays the eect of increasing
buer concentration on the mitosis-time proles for dierent total molar
citrate values (0.0, 0.055, 0.075, 0.110). The ratio of pilocarpinium ion to
pilocarpine increases with the postinstillation buering capacity, thus reducing the net transcorneal ux of pilocarpine.

III.

CONSTRAINTS TO OCULAR DRUG DELIVERY

Ocular tissues are protected from exogenous toxic substances in the environment or bloodstream by a variety of mechanisms, notably, tear secretion
continuously ushing its surface, an impermeable surface epithelium, and a

Copyright 2003 Marcel Dekker, Inc.

Overview of Ocular Drug Delivery

Figure 4 Miosis-time proles: Plots of the average observed changes in pupillary

diameter (
PD) as a function of time following the instillation of 25.0 mL of the
isotonic 1% pilocarpine nitrate solutions, which contained the dierent concentrations of citrate buer. The vertical lines through the data points are SD (data
points with standard deviation lines omitted is for clarity of the gure). (From
Ref. 20.)

transport system actively clearing the retina of agents potentially able to

disturb the visual process. However, the same protective mechanisms may
cause subtherapeutic drug levels at the intended site. The diculties can be
compounded by the structure of the globe itself, where many of its internal
structures are isolated from the blood and the outside surface of the eye.
A major goal in ocular therapeutics is to circumvent these structural
obstacles and protective mechanisms to elicit desired pharmacological
response.
Physiological barriers to the diusion and productive absorption of
topically applied ophthalmic drugs exist in the precorneal and corneal
spaces. Anterior chamber factor also greatly inuence the disposition of
topically applied drugs. Precorneal constraints include solution drainage,
lacrimation and tear dilution, tear turnover, and conjunctival absorption.
For acceptable bioavailability, a proper duration of contact with the cornea
must be maintained. Instilled solution drainage away from the precorneal
area has been shown to be the most signicant factor reducing this contact

Copyright 2003 Marcel Dekker, Inc.

Macha et al.

time and ocular bioavailability of topical solution dosage forms (21,22).

Instilled dose leaves the precorneal area within 5 minutes of instillation in
humans (21,23). The natural tendency of the cul-de-sac is to reduce its uid
volume to 710 mL (2426). A typical ophthalmic dropper delivers 30 mL,
most of which is rapidly lost through nasolacrimal drainage immediately
following dosage. This drainage mechanism may then cause the drug to be
systemically absorbed across the nasal mucosa or the gastrointestinal tract
(27). Systemic loss from topically applied drugs also occurs from conjunctival absorption into the local circulation. The conjunctiva possesses a relatively large surface area, making this loss signicant.
Simple dilution of instilled drug solution in the tears acts to reduce the
transcorneal ux of drug remaining in the cul-de-sac. Lacrimation can be
induced by many factors, including the drug entity, the pH, and the tonicity
of the dosage from (2830). Formulation adjuvants can also stimulate tear
production (20).
Tear turnover acts to remove drug solution from the conjunctival culde-sac. Normal human tear turnover is approximately 16% per minute,
which can also be stimulated by various factors, as described elsewhere
(21,25). These factors render topical application of ophthalmic solutions
to the cul-de-sac extremely inecient. Typically, less than 1% of the instilled
dose reaches the aqueous humor (27,31). The low fraction of applied dose
(1%) of drug solution reaching the anterior chamber further undergoes
rapid elimination from the intraocular tissues and uids. Absorbed drug
may exit the eye through the canal of Schlemm or via absorption through
the ciliary body of suprachoroid into the episcleral space (27). Enzymatic
metabolism may account for further loss, which can occur in the precorneal
space and/or in the cornea (32,33). Age and genetics have been determined
to be two important factors in ocular metabolism (34,35).
Clearly, the physiological barriers to topical corneal absorption are
formidable. The result is that the clinician is forced to recommend frequent
high doses of drugs to achieve therapeutic eect. This pulsatile dosing not
only results in extreme uctuations in ocular drug concentrations but may
cause many local and/or systemic side eects. Approaches taken to circumvent this pulsatile dosing and their ramications on ocular therapies are the
subject matter of this text.
For the eective treatment of diseases involving the retina, drugs must
cross the blood-ocular barrier in signicant amounts to demonstrate therapeutic eect. The blood-ocular barrier is a combination of microscopic
structures within the eye, which physiologically separate it from the rest
of the body. It is comprised of two systems: (a) blood-aqueous barrier,
which regulates solute exchange between blood and the intraocular uid,
and (b) blood-retinal barrier, which separates the blood from the neural

Copyright 2003 Marcel Dekker, Inc.

Overview of Ocular Drug Delivery

retina. Both barriers contain epithelial and endothelial components whose

tight junctions limit transport.
A transient increase in the blood-retinal barrier permeability can be
achieved by modication of the barrier properties. For instance, opening of
the blood-retinal barrier can be achieved by intracarotid infusion of a hyperosmotic solution, such as mannitol or arabinose. Perfusion with such a
solution for about 30 seconds is shown to open the blood-retinal barrier
reversibly. Osmotically induced shrinkage of the retinal and brain capillary
endothelial cells causes opening of the tight junctions. Other methods
include perfusion with oleic acid or protamine. These methods, however,
produce a nonspecic opening of the blood-retinal barrier, possibly with
associated retinal and central nervous system toxicity.
Chemical modication is more commonly employed to enhance drug
transport across biological barriers. Lipophilic analogs of the parent drug
increase lipid solubility and thereby their blood-retinal barrier permeability. Another approach to enhance transport across the blood-retinal
barrier could involve utilizing specic carrier systems on the epithelial
membrane. Drugs may be modied in such a way that their structures
resemble endogenous ligands for a specic carrier system on the bloodretinal barrier.
Drug delivery through nutrient transport systems has been reported
previously with intestinal absorption (3638). -Lactam antibiotics and
other compounds that share the structural features of the endogenous peptides are recognized by the peptide transporters. Recently valacyclovir (valyl
ester of acyclovir) (39,40) and valganciclovir (valyl ester of ganciclovir) (41)
were shown to be the substrates for peptide transporters. These prodrugs
increased the oral bioavailability of acylclovir and ganciclovir signicantly
(42,43), thus reducing the daily oral dose requirement.
Various transporters/receptors are reported to be present on the retina
and/or the blood-ocular barriers. The reader is referred to specic chapters
in this volume for detailed description. However, very few studies have been
carried out to explore the transporters present on the retina or the bloodocular barrier. The transporters/receptors present on the retina or the
blood-ocular barrier may be exploited to increase ocular bioavailability of
drugs with poor intrinsic permeability.

ACKNOWLEDGMENTS
Supported by NIH grants R01 EY09171 and R01 EY10659.

Copyright 2003 Marcel Dekker, Inc.

10

Macha et al.

REFERENCES
1. M. Gibaldi and D. Perrier. Pharmacokinetics. New York: Marcel Dekker,
1982.
2. D. M. Maurice. The eye. In: H. Davsoned, ed. Vegetative Physiology and
Biochemistry. New York: Academic Press, 1969.
3. I. Ahmed and T. F. Patton. Importance of the noncorneal adsorption route in
topical ophthalmic drug delivery. Invest. Ophthalmol. Vis. Sci. 26:584587,
1985.
4. S. E. Bloomeld, T. Miyata, M. W. Dunn, N. Bueser, K. H. Stenzel, and A. L.
Rubin. Soluble gentamicin ophthalmic inserts as a drug delivery system. Arch.
Ophthalmol. 96:885887, 1978.
5. L. Z. Bito and R. A. Baroody. The penetration of exogenous prostaglandin
and arachidonic acid into, and their distribution within, the mammalian eye.
Curr. Eye Res. 1:659669, 1981.
6. D. G. Cogan and E. D. Hirch. Cornea: Permeability to weak electrolytes.
Arch. Ophthalmol. 32:276, 1944.
7. V. E. Kinsey. Physiology of the eye. In; F. Adler, H., eds. St. Louis: Mosby,
1965.
8. H. S. Huang, R. D. Schoenwald, and J. L. Lach. Corneal penetration behavior of beta-blocking agents II: Assessment of barrier contributions. J. Pharm.
Sci. 72:12721279, 1983.
9. G. M. Grass and J. R. Robinson. Mechanisms of corneal drug penetration. II:
Ultrastructural analysis of potential pathways for drug movement. J. Pharm.
Sci. 77:1523, 1988.
10. V. H. L. Lee. Mechanisms and facilitation of corneal drug penetration, J
Controlled Rel, 11:79 (1990).
11. S. D. Klyce and C. E. Crosson. Transport processes across the rabbit corneal
epithelium: a review. Curr. Eye Res. 4:323331, 1985.
12. A. K. Mitra. Ophthalmic drug delivery. In: P. Tyle, ed. Drug Delivery
Devices. New York: Marcel Dekker, 1988.
13. A. M. Tonjum. Permeability of rabbit corneal epithelium to horseradish peroxidase after the inuence of benzalkonium chloride. Acta Ophthalmol.
(Copenh.) 53:335473, 1975.
14. S. Mishima. Clinical pharmacokinetics of the eye. Proctor lecture. Invest.
Ophthalmol. Vis. Sci. 21:504541, 1981.
15. R. D. Schoenwald and R. L. Ward. Relationship between steroid permeability
across excised rabbit cornea and octanol-water partition coecients. J.
Pharm. Sci. 67:786788, 1978.
16. M. M. Narurkar and A. K. Mitra. Prodrugs of 5-iodo-20 -deoxyuridine for
enhanced ocular transport. Pharm. Res. 6:887891, 1989.
17. M. M. Narurkar and A. K. Mitra. Synthesis, physicochemical properties, and
cytotoxicity of a series of 50 -ester prodrugs of 5-iodo-20 -deoxyuridine. Pharm.
Res. 5:734737, 1988.

Copyright 2003 Marcel Dekker, Inc.

Overview of Ocular Drug Delivery

11

18. G. L. Mosher and T. J. Mikkelson. Permeability of the n-alkyl-p-aminobenzoate esters across the isolated corneal membrane of the rabbit. Int. J. Pharm
2:239, 1979.
19. R. D. Schoenwald and H. S. Huang. Corneal penetration behavior of betablocking agents I: Physiochemical factors. J. Pharm. Sci. 72:126672, 1983.
20. A. K. Mitra and T. J. Mikkelson. Ophthalmic solution buer systems I. The
eect of buer concentration on the ocular absorption of pilocarpine. Int. J.
Pharm. 10:219, 1982.
21. S. S. Chrai, T. F. Patton, A. Mehta, and J. R. Robinson. Lacrimal and
instilled uid dynamics in rabbit eyes. J. Pharm. Sci. 62:11121121, 1973.
22. S. S. Chrai, M. C. Makoid, S. P. Eriksen, and J. R. Robinson. Drop size and
initial dosing frequency problems of topically applied ophthalmic drugs. J.
Pharm. Sci. 63:333338, 1974.
23. J. W. Sieg and J. R. Robinson. Mechanistic studies on transcorneal permeation of pilocarpine. J. Pharm. Sci. 65:18161822, 1976.
24. D. M. Maurice. The dynamics and drainage of tears. Int. Ophthalmol. Clin.
13:103116, 1973.
25. S Mishima, A. Gasset, S. D. Klyce, Jr., and J. L. Baum. Determination of tear
volume and tear ow. Invest. Ophthalmol. 5:264276, 1966.
26. N. Ehlers. The precorneal lm, biomicroscopical, histological, and chemical
investigations. Acta Ophthalmol. 81(Suppl.):1, 1965.
27. V. H. Lee and J. R. Robinson. Topical ocular drug delivery: Recent developments and future challenges. J. Ocul. Pharmacol. 2:67108, 1986.
28. A. Kupferman, M. V. Pratt, K. Suckewer, and H. M. Leibowitz. Topically
applied steroids in corneal disease. 3. The role of drug derivative in stromal
absorption of dexamethasone. Arch. Ophthalmol. 91:373376, 1974.
29. J. W. Sieg and J. R. Robinson. Vehicle eects on ocular drug bioavailability
II: Evaluation of pilocarpine. J. Pharm. Sci. 66:122212228, 1977.
30. J. M. Conrad, W. A. Reay, R. E. Polcyn, and J. R. Robinson. Inuence of
tonicity and pH on lacrimation and ocular drug bioavailability. J. Parenter.
Drug Assoc. 32:149161, 1978.
31. T. J. Mikkelson, S. S. Chrai, and J. R. Robinson. Competitive inhibition of
drug-protein interaction in eye uids and tissues. J. Pharm. Sci. 62:19421945,
1973.
32. V. H. Lee, K. W. Morimoto, and R. E. Stratford, Jr. Esterase distribution in
the rabbit cornea and its implications in ocular drug bioavailability.
Biopharm. Drug dispos. 3:291300, 1982.
33. V. H. Lee, H. W. Hui, and J. R. Robinson. Corneal metabolism of pilocarpine
in pigmented rabbits. Invest. Ophthalmol. Vis. Sci. 19:210213, 1980.
34. V. H. L. Lee, J. Stratford, and K. W. Morimoto. Age related changes in
esterase activity in rabbit eyes. Int. J. Pharm. 13:183, 1983.
35. H. Shichi and D. W. Nebert. Genetic dierences in drug metabolism associated with ocular toxicity. Environ. Health Perspect. 44:107117, 1982.
36. J. P. Bai, M. Hu, P. Subramanian, H. I. M osberg, and G. L. Amidon.
Utilization of peptide carrier system to improve intestinal absorption: target-

Copyright 2003 Marcel Dekker, Inc.

12

37.

38.

39.

40.

41.

42.

43.

Macha et al.
ing prolidase as a prodrug-converting enzyme. J. Pharm. Sci. 81:113116,
1992.
M. Hu, P. Subramanian, H. I. Mosberg, and G. L. Amidon. Use of the
peptide carrier system to improve the intestinal absorption of L-alpha-methyldopa: carrier kinetics, intestinal permeabilities, and in vitro hydrolysis of
dipeptidyl derivatives of L-alpha-methyldopa. Pharm. Res. 6:6670, 1989.
I. Tamai, T. Nakanishi, H. Nakahara, et al. Improvement of L-dopa absorption by dipeptidyl derivation, utilizing peptide transporter PepT1. J. Pharm.
Sci. 87:15421546, 1998.
R. L. de Vrueh, P. L. Smith, and C. P. Lee. Transport of L-valine-acyclovir
via the oligopeptide transporter in the human intestinal cell line, Caco-2. J.
Pharmacol. Exp. Ther. 286:11661170, 1998.
M. E. Ganapathy, W. Huang, H. Wang, V. Ganapathy, and F. H. Leibach.
Valacyclovir: a substrate for the intestinal and renal peptide transporters
PEPTT1 and PEPT2. Biochem. Biphys. Res. Commun. 246:4705, 475, 1998.
M. Sugawara, W. Huang, Y. J. Fei, F. H. Leibach, V. Ganapathy, and M. E.
Ganapathy. Transport of valganciclovir, a ganciclovir prodrug, via peptide
transporters PEPT1 and PEPT2. J. Pharm. Sci. 89:781789, 2000.
M. D. Pescovitz,J. Rabkin, R. M. Merion, et al. Valganciclovir results in
improved oral absorption of ganciclovir in liver transplant recipients.
Antimicrob. Agents Chemother. 44:28112815, 2000.
P. Reusser. Oral valganciclovir: A new option for treatment of cytomegalovirus infection and disease in immunocompromised hosts. Expert. Opin.
Invest. Drugs 10:17451753, 2001.

Copyright 2003 Marcel Dekker, Inc.

2
Membrane Transport Processes in
the Eye
Gangadhar Sunkara
and Uday B. Kompella
University of Nebraska Medical Center, Omaha, Nebraska, U.S.A.

I.

INTRODUCTION

Epithelial and/or endothelial cells sealed by tight junctions serve as the

major membrane barriers for the transport of nutrients, ions, and drugs
into intraocular tissues. The principal membrane barriers of the eye are
located in the cornea, conjunctiva, iris-ciliary body, lens epithelium, and
retina (Fig. 1). Various specialized transport processes in these barriers
control the movement of solutes into and out of intraocular chambers.
These processes maintain the visual function, control intraocular pressure,
provide nutrients to avascular cornea and lens, and protect ocular tissues
from xenobiotics. An alteration in the function of these transporters is often
the underlying cause of various ocular diseases. In addition, these transport
processes can play a role in drug transport.
Drug therapy is useful in the treatment of corneal epitheliopathy,
keratitis, conjunctivitis, extracellular infections, glaucoma, iritis, and cataractdiseases that aict the anterior segment of the eyeand vitreoproliferative disorders, endophthalmitis of bacterial and fungal origins, uveitis,
viral retinitis, diabetic retinopathy, and macular degenerationdiseases that
aict the posterior segment of the eye. Ocular drug therapy often involves
topical or systemic administration of drugs. Therefore, for eective delivery
to intraocular tissues, drugs must penetrate across cornea, conjunctiva, and/
or sclera following topical administration or across endothelial barriers

Current aliation: Novartis Pharmaceuticals, East Hanover, New Jersey, U.S.A.

13

Copyright 2003 Marcel Dekker, Inc.

14

Sunkara and Kompella

Figure 1

Various epithelial and endothelial barriers of the eye.

along with the epithelial barriers of the iris-ciliary body or retina following
systemic administration. However, as these barriers restrict the transport of
therapeutic agents, the ocular bioavailability of drugs following topical or
systemic administration is low. These barriers also play a role in the drug
clearance following periocular or intraocular administration.
This chapter describes the various ocular epithelial and endothelial
barriers and the associated ion and solute transport processes in the eye.
The discussion includes ocular drug transport processes as well as recently
identied drug eux pumps.

II.

GENERAL MECHANISMS AND ROUTES OF

TRANSPORT

A.

Mechanisms of Transport

The membrane, whether it is plasma membrane or a membrane encompassing cellular organelles, imposes a barrier to the free movement of molecules.
Simple diusion, facilitated diusion, primary active transport, secondary
active transport, transcytosis, and group translocation are six primary
mechanisms of solute transport across a membrane (Saier, 2000a). Simple

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

15

or passive diusion is a process that does not require cellular energy, but
requires a chemical gradient for the solute to be transported. Facilitated
diusion is similar to simple diusion in not requiring energy and following
chemical gradient, but it requires a membrane transporter or carrier to facilitate the transport. Two primary modes of facilitated transport have been
recognized in the biological systems: channel type and carrier type. In
channel-type facilitated diusion, the solute passes in a diusion-limiting
process from one side of the membrane to the other via a channel or pore
that is lined by appropriately hydrophilic, hydrophobic, or amphipathic
amino acyl residues of the constituent proteins. In carrier-type facilitated
diusion, some part of the transporter is classically presumed to pass through
the membrane together with the substrate. Carriers usually exhibit rates of
transport, stereospecicity, and saturation kinetics in higher magnitude compared to channels. Solute transport processes where energy is required to
transport the solutes against a concentration gradient are referred to as active
transport processes. In primary active transport, energy is directly expended
by the transporter in the form of ATP hydrolysis or electron ow. In
secondary active transport, the transporter does not directly expend cellular
energy but relies on a primary active transport process for its driving force.
Some macromolecules cross the cell baers through endocytosis (receptormediated, adsorptive, or uid-phase) followed by exocytosis, a process
known as transcytosis. Finally, with group translocation, the transported
substance is chemically modied by the membrane transporter during the
transport event, which may require energy either directly or indirectly.
B.

Routes of Transport

Across any continuous epithelium, a solute can be transported either through

the cells or between the cells. The cellular pathway is known as the transcellular route, and the intercellular pathway is known as the paracellular route
(Fig. 2). The transport of most of the solutes is contributed by both
pathways. Lipophilic molecules and molecules with specialized transport
processes prefer the transcellular route, whereas hydrophilic molecules lacking membrane transport processes prefer the paracellular route. However,
dissecting the fraction contributed by each pathway is dicult (Ho et al.,
1999). One measure of the overall barrier permeability is its electrical
resistance, a measure of the resistance of the tissue to ion transport. The
electrical resistance of various ocular barriers is summarized in Table 1.
The mechanisms of transcellular transport include simple diusion,
facilitated diusion, active transport, and endocytosis. The plasma membrane of a cell consists of a lipid bilayer with an array of peripherally and
integrally associated proteins, some of which serve as specic transporters or

Copyright 2003 Marcel Dekker, Inc.

16

Sunkara and Kompella

Figure 2 Transcellular and paracellular pathways for drug transport across epithelial and endothelial barriers.

receptors, which allow solute uptake into the cell via active transport,
facilitated diusion, or receptor-mediated endocytosis. However, for a
majority of the currently available drug molecules, no such transporters
or receptors exist, and these drugs are transported by passive diusion
through the apical membrane, through the cell proper, and across the basolateral membrane to move across the cell. During transcellular movement,
the solute must interact with some components of the cell membrane. The
interaction of the solute with the cell is inuenced by both the structural
characteristics of the solute and the cell itself.
Transport along the paracellular route is passive and is only limited by
the size and charge of the intercellular spaces. The paracellular pathway is
an aqueous route involving diusion of the solute between adjacent epithelial cells/endothelial cells restricted by the presence of a series of junctional
strands known as tight junctions or zonula occludens (ZO), which are joined
at the apical pole (Madara and Trier, 1982; Yu, 2000). Freeze-fracture
electron microscopy studies demonstrated that ZO forms fusion sites that
appear as a complex network of protein strands that seal the intercellular

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

Table 1

17

Transmembrane Electrical Resistances of Various Ocular Barriers

Tissue
Cornea
Intact rabbit cornea
Primary rabbit corneal epithelial
cultures (air interface)
Ciliary body
Intact rabbit ciliary body
Rabbit nonpigmented epithelial
cell cultures
Human nonpigmented epithelial
cell culture
Conjunctiva
Excised rabbit conjunctiva
Primary rabbit conjunctival
cultures (liquid interface)
Primary rabbit conjunectival
cultures (air interface)

Rt
(ohm-cm2 )
7500
5000

77
2030
20

Ref.
Marshall and Klyce, 1983
Chang et al., 2000

Noske et al., 1994

Cilluo et al., 1993
Noske et al., 1995

1400
1900

Kompella et al., 1993

Saha et al., 1996

1100

Yang et al., 2000

Retina
RPE-Choroid
Adult human
Fetal human
Bovine
Rabbit
Cat

79
206
138
350
133259

Cultured RPE
Fetal human
Fetal bovine
Adult human

70
329
225

Quinn and Miller, 1991

Quinn and Miller, 1991
Steinberg et al., 1978
Quinn and Miller, 1992
Joseph and Miller, 1991
Hernandez et al., 1995
Hernandez et al., 1995
Hu et al., 1994

space all around the cell perimeter (Lapierre, 2000). The tightness of the
barrier increases with an increase in the number of junctional strands. In
general, the tight junctions are essentially impermeable to molecules with
radii greater than 15 A. Various proteins that form the tight junctional
complex include ZO-l, ZO-2, cingulin. and occludin (Lapierre, 2000). In
addition to tight junctions, intercellular spaces possess intermediate junctions or zonula adherens, desmosomes, and gap junctions. Zonula adherens
and desmosomes are sites of rm adhesion between cells that ensure

Copyright 2003 Marcel Dekker, Inc.

18

Sunkara and Kompella

mechanical stability. Gap junctions are responsible for electrical and metabolic coupling between cells (Simon and Goodenough, 1998).

C.

Transporters

Transport systems serve the cell by allowing the cellular entry and exit of
essential ions and nutrients, and by expelling the toxic metabolites out of the
cell. All transmembrane transport processes are mediated by integral membrane proteins, sometimes functioning in conjunction with extracytoplasmic
receptors or receptor domains and/or cytoplasmic energy coupling and regulatory proteins or protein domains. These integral membrane proteins and
any associated protein or protein domains are referred to as a transport
process, transport system, transporter, porter, permease system, or permease (Saier, 2000a). Many transporters are localized in a polarized manner
in an epithelium or endothelium to perform vectorial transport of solutes.
All transporters are proteins with specic conformations, which makes them
unique in their function. Broadly, various types of transporters on the cell
membrane can be classied as channels, pumps, uniporters, symporters or
cotransporters, and exchangers or antiporters (Table 2) (Saier, 2000a). Some
transport proteins having multiple substrate domains are structured so that
these domains are arranged in the plane of the membrane in a circle, thereby
forming a barrel to transport small solutes such as H+, Na+, and K+ into
and out of the cell. When ions or solutes pass through such a transporter by
passive diusion, the transporter is called a channel. Alternatively, the
transporter protein may directly utilize energy, usually derived from ATP,
Table 2

Commonly Used Symbols for Various Transporters

Transporter

Symbol

Example

Channel

Na+, K+, and Cl channels

Cotransporter
(symporter)

Na+-glucose and Na+amino acid transporters

Exchanger
(antiporter)

Na+/H+ and Cl /HCO

3
exchangers

Uniporter

Glucose transporter
(GLUT)

Pump

Na+/K+-ATPase
P-glycoprotein/MRP

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

19

to actively drive solutes from one side of the plasma membrane to the other,
in which case it is called as a pump. A uniporter transports one solute at a
time. A symporter transports the solute and one or more cotransported
solutes at the same time in the same direction. An antiporter transports
the solute in (or out) and another solute in the opposite direction.
Transporter proteins may exist in dierent isoforms with the same function
but with variable anities, uptake kinetics, and tissue expression proles.
For example, facilitated glucose transport is mediated by a family of glucose
transporter (GLUT) proteins, seven isoforms of which are known to date
(Thorens, 1996). GLUT1, GLUT2, GLUT3, and GLUT4 are involved in
cellular glucose uptake; GLUT1, GLUT3, and GLUT4 are high-anity
glucose transporters, and GLUT2 has a signicantly lower anity.
GLUT5 is a high-anity fructose transporter with poor glucose transport
capacity. GLUT7 is closely related to GLUT2, but it is retained in the
endoplasmic reticulum and not transported to the plasma membrane. The
role of GLUT6 is still unknown. Similarly, various families of transmembrane transporters exist in the mammalian plasma membranes for amino
acids and their derivatives (Palacin et al., 1998; Saier, 2000b). Transfer of
amino acids across the hydrophobic domain of the plasma membrane is
mediated by proteins that recognize, bind, and transport these amino
acids, which exhibit broad substrate specicity and stereospecicity. There
are three best-known amino acid transport systems present in the plasma
membrane of mammalian cells based on the type of amino acid the protein
moves and the thermodynamic properties of the transport: (a) zwitterionic
amino acid systems (A, ASC, N, BETA, GLY, IMINO, PHE, B0, L), (b)
cationic amino acid systems (B0,+, b+, y+, y+L, b0,+), and (c) anionic
amino acid systems (X AG, X
C ). A detailed description of these transporters
is provided elsewhere (Palacin et al., 1998).

III.

TRANSPORT BARRIERS IN THE EYE

A.

Cornea

While primarily being a refractive element of the eye, the cornea acts as a
mechanical and chemical barrier to intraocular tissues. The mammalian
cornea, with relatively few exceptions, consists of ve to six distinct layers,
with a total thickness of 300500 mm (Pepose and Ubels, 1992). These layers
include (Fig. 3): corneal epithelium, underlying basement membrane, acellular Bowmans layer, corneal stroma, Descemets membrane, and corneal
endothelium. Bowmans layer is absent in the cornea of rabbit, a routinely
used animal model in eye research (Fig. 3).

Copyright 2003 Marcel Dekker, Inc.

20

Figure 3

Sunkara and Kompella

The cornea. Cellular organization of various transport limiting layers.

1. Corneal Epithelium
The corneal epithelium plays critical roles in the maintenance of a barrier
against tearborne agents and in maintaining a balanced stromal hydration.
The corneal epithelium, with a total thickness of 3550 mm, is made of ve
cell layers constituted by supercial, wing, and basal cells and is maintained
by several cell-cell and cell-substrate interactions (Pepose and Ubels, 1992).
The supercial cells are joined by numerous desmosomes, and the barrier
function of these cells is due to the presence of tight junctional complexes
that are exclusive to supercial cells in the epithelium. Wing cells are
attached to supercial cells and to one another by desmosomes. Large
gap junctions are also present between the wing cells, allowing a high degree
of intercellular communication in this layer. The basal cells also have desmosomes and gap junctions that are fewer in number. Corneal epithelium
overlies on an array of collagen brils and glycosaminoglycans (GAGs)
known as acellular Bowmans membrane followed by stroma, a hydrated
(7578% water) matrix of collagen brils and GAGs, which constitutes 90%
of the corneal thickness. This arrangement is followed by Descemets membrane and corneal endothelium. Descemets membrane forms the basement
membrane for the corneal endothelium (Fig. 3).

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

21

The corneal epithelium is considered a tight epithelium due to its

high shunt resistance, which is in the range of 1216 kohm-cm2 (Marshall
and Klyce, 1983). The tightness of the corneal epithelium is due to the
existence of zonulae occludens or tight junctions between the supercial
cells. These intercellular junctions maintain the asymmetrical distribution
of several receptors for endogenous molecules such as hormones or growth
factors and transporters such as Na+/K+-ATPase, Cl channel, and Na+
amino acid cotransporter in order to maintain the vectorial transport (Klyce
and Crosson, 1985).
The resistance to transport across the cornea is the sum of the resistances to transport across each of the individual corneal layers, i.e., Rcornea
= Repithelium + Rstroma + Rendothelium. Although the corneal epithelium
represents less than 10% of the entire corneal thickness, it contributes
99% of the resistance to solute diusion across cornea. Of this, the supercial epithelial cells contribute as much as 60%. The stroma and endothelium contribute 3% to the total corneal electrical resistance. The
transepithelial or transmembrane resistance of various corneal preparations
is summarized in Table 1.
2.

Corneal Endothelium

Corneal endothelium is a monolayer of polygonal cells, most of which are

hexagonal in shape with about 20 mm diameter and 46 mm thickness. These
cells play a key role in maintaining corneal transparency through their
transport, synthetic, and secretory functions. Contrary to the corneal epithelial cells, macular occludens, not entirely occlusive junctions, rather than
zonula occludens exist between the endothelial cells (McLaughlin et al.,
1985). This results in a leaky barrier between aqueous humor and stroma.
The endothelium has a passive permeability that allows the passage of large
molecules up to 70 kDa in size (Pepose and Ubels, 1992). Although the
endothelium is leaky, it is the site of the major active ion and uid transport
mechanisms that maintain constant corneal thickness. Gap junctions present on the lateral membranes of the endothelial cells contribute to intercellular communication.
3.

Ion and Organic Solute Transport

Corneal surface and stromal hydration is maintained by the movement of

Na+, K+, and Cl across the apical and basolateral membranes of the
epithelium and the corneal endothelium (Fig. 4). The active transport of
these ions is responsible for the tear-side negative potential dierence of the
corneal epithelial cells. The ouabain-sensitive Na+/K+-ATPase on the
basolateral side of the epithelium actively transports Na+ out of the cells

Copyright 2003 Marcel Dekker, Inc.

22

Sunkara and Kompella

Figure 4 Putative ion and solute transport processes in the mammalian corneal
epithelium and endothelium.

in exchange for K+ (Green, 1970). Cl entry across the basolateral membrane of the epithelium is mediated by a loop diureticsensitive Na+/K+/
Cl cotransporter (Klyce, 1972; Marshall and Klyce, 1983). This Cl entry
occurs against an electrochemical potential and is ouabain sensitive, suggesting that Cl entry is coupled to the Na+/K+-ATPase activity. The Na+/
K+/Cl cotransporter elevates the intracellular Cl concentration to three
to four times greater than that expected from simple equilibrium, enabling
the passive transport of Cl through Cl channels located on the apical side.
K+ has low paracellular permeability, and it enters the cells through Na+/
K+-ATPase (Marshall and Klyce, 1983) and exits through K+ channels
located on the basolateral side.
In endothelial cells, the Na+/K+-ATPase is located on the basolateral
side (Whikehart and Soppet, 1981; Whikehart et al., 1987). Inhibition of this
pump with ouabain stops sodium transport, causes corneal swelling, and
eliminates the transendothelial potential dierence. Stromal Cl concentration is in part determined by Cl entry through a C1 /HCO
3 exchanger
located on the apical membrane of the endothelial cells and exit through Cl
channel on the basolateral membrane of the endothelium (Bonanno et al.,
1998). Also, Na+/K+/Cl cotransporter present on the lateral membrane
of the corneal endothelium contributes to the transendothelial Cl ux
(Jelamskii et al., 2000).

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

23

In corneal epithelial and endothelial cells, Na+/H+ exchanger, Na+symport, and Cl /HCO
3 exchanger are involved in the regulation of
intracellular pH (Jentsch et al., 1985; Bonanno et al., 1999). Na+/H+
exchanger is present in the basolateral membranes of both epithelial and
endothelial cells. Na+/HCO
3 transporter is predominantly localized on the
basolateral side of the corneal endothelium and is weakly expressed in the
corneal epithelium (Sun et al., 2000).
H+-lactate cotransport is present on the baslolateral side of the rabbit
corneal epithelium (Bonanno, 1990) and on both sides of the rabbit corneal
endothelium (Giasson and Bonanno, 1994). In addition, the corneal
endothelium has a Na+-lactate cotransport process on the basolateral
side (Giasson and Bonanno, 1994). These transport processes eciently
remove lactate from the highly glycolytic cornea, thereby preventing lactate-mediated corneal swelling.
The corneal epithelium is relatively impermeable to water-soluble
compounds such as amino acids derived from tears (Thoft and Friend,
1972). The limbal blood supplies less than 20% of the corneal nutrients,
with the aqueous humor being the primary source of amino acids (Thoft and
Friend, 1972). Indeed, corneal endothelium, the principal barrier between
the aqueous humor and the corneal epithelium, transports amino acids from
the aqueous humor to the extracellular uid of the stroma against a concentration gradient in the rabbit cornea (Riley, 1977). Consistent with this,
the steady-state concentrations of most free amino acids in the stroma of the
rabbit cornea are higher than those in the aqueous humor (Thoft and
Friend, 1972; Riley et al., 1973; Riley and Yates, 1977). Similarly, the
high glucose requirements of the corneal epithelium are met in part by the
facilitated glucose transporter, GLUT 1, present on the basolateral side of
the epithelium (Takahashi et al., 1996).
HCO
3

B.

Conjunctiva

The conjunctiva is a thin, transparent mucous membrane lining the inside of

the eyelids and is continuous with cornea. In rabbits, the conjunctiva occupies 9 times larger surface area, and in humans, it occupies 17 times larger
surface area than the cornea (Watsky et al., 1988). The conjunctiva can be
divided into three layers: (a) an outer epithelium, a permeability barrier, (b)
substantia propria, containing nerves, lymphatics, and blood vessels, and (c)
submucosa, which provides a loose attachment to the underlying sclera. The
conjunctiva diers from cornea in its rich vasculature, the presence of many
goblet cells, and its ability to transdierentiate.

Copyright 2003 Marcel Dekker, Inc.

24

Sunkara and Kompella

1. Conjunctival Epithelium
The conjunctival epithelium is similar to the cornea with respect to the
organization of epithelial cellsit contains supercial, wing, and basal
cells, the three principal corneal epithelial cell types. The epithelium is two
to three cell layers thick, nonkeratinized, stratied and sqamous at the
eyelids, and columnar towards the cornea. The tight junctions in the apical
pole of the conjunctival epithelium render it a relatively impermeable barrier. The epithelial cells of the conjunctiva are attached to one another by
means of desmosomes and to the basal lamina through hemidesmosomes.
2. Ion and Organic Solute Transport
The conjunctiva is a tight epithelium with an electrical resistance of 1.4
kohm-cm2 (Kompella et al., 1993) (Fig. 5). The potential dierence of the
conjunctiva is lumen-negative like the cornea, suggesting a net secretion of
anions and/or a net absorption of cations.

Figure 5 Putative ion and solute transport processes in the mammalian conjunctival epithelium.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

25

The conjunctiva actively secretes Cl (Kompella et al., 1993), and about
80% of the conjunctival active ion transport is accounted for by Cl secretion
through apically localized channels. Cl enters the basolateral side via Na+/
K+/Cl cotransport energized by the basolateral Na+/K+-ATPase.
Conjunctiva secretes uid secondary to active Cl secretion (Shiue et al.,
2000). Active Na+ absorption can counter this uid secretion. Net uid secretion rate across the conjunctiva is altered by pharmacological agents known
to aect active Cl secretion or Na+ absorption (Shiue et al., 2000).
Evidence exists for the presence of ion-dependent solute transport
processes such as Na+-glucose, Na+-amino acid, and Na+-nucleoside
transporters in the conjunctival epithelium (Kompella et al., 1995; Hosoya
et al., 1996; Hosoya et al., 1998b). After bathing conjunctiva in a glucosefree glutathione bicarbonate Ringer (GBR) solution, mucosal addition of
D-glucose elevated short-circuit current (Isc) by a maximum of 20% in a
dose-dependent manner. Phlorizin, a specic inhibitor of Na+-glucose
cotransporter, reduced the Isc in a dose-dependent manner, suggesting the
possible apical localization of Na+-glucose cotransporter in the pigmented
rabbit conjunctiva (Hosoya et al., 1996). Similarly, mucosal addition of
glycine, L-arginine, D-arginine, and L-glutamic acid have increased Isc in
the presence but not in the absence of sodium in the medium, suggesting
the possible existence of Na+-amino acid cotransporters on the apical side
of the conjunctiva (Kompella et al., 1995). Furthermore, B0,+, a sodiumdependent transporter of neutral (L-Glu) and basic amino acids (L-Arg,
L-Lys, and NG-nitro-L-arginine) is present in the conjunctiva (Hosoya
et al., 1998a).
Uridine, a nucleoside, is transported preferentially from the mucosal
to serosal direction in a temperature- and phlorizin-sensitive manner across
the conjunctiva (Hosoya et al., 1998b). At constant Na+ concentration (141
mM) on mucosal side, uridine increased Isc in a dose-dependent fashion. At
constant uridine concentration (10 mM), increasing Na+ concentration
increased Isc with a Hill coecient of 1.1, suggesting that there is a 1:1
coupling in the transport of Na+ and uridine. Na+-dependent uridine transport was inhibited by 10 mM adenosine, guanosine, and inosine, but not by
thymidine, suggesting that the transport process may be mainly selective for
purine nucleosides.
C.

Iris-Ciliary Body

The iris, ciliary body, and choroid comprise the vascular uveal coat of
eye. The anterior iris is immersed in the aqueous humor, which enters
iris stroma through openings or crypts along its anterior surface. The
receives its blood supply from the major arterial circle, which lies in

Copyright 2003 Marcel Dekker, Inc.

the
the
iris
the

26

Sunkara and Kompella

stroma of the ciliary body near the iris root. The ciliary body can be divided
into the following regions: nonpigmented ciliary epithelium, pigmented ciliary epithelium, stroma, and ciliary muscle. The main arterial blood supply
to the ciliary body is through the long posterior and the anterior ciliary
arteries. These capillaries are fenestrated (Cunha-Vaz, 1979) and leaky
(Stewart and Tuor, 1994).
The ciliary body secretes aqueous humor into the posterior chamber,
from where it ows through the pupil into the anterior chamber and leaves
the eye in a bulk ow through trabecular and uveo-scleral routes at the angle
of the anterior chamber. The aqueous humor is a transparent, aqueous
so1ution, and its composition is dierent from that of plasma in its low
concentration of plasma proteins and 20 to 60-fold higher concentration of
ascorbic acid (Caprioli, 1992). The aqueous humor production and the
intraocular pressure are maintained by membrane transport processes.
For instance, the aqueous humor production can be controlled by agents
that directly or indirectly alter the release of Cl through basolateral channels of nonpigmented ciliary epithelium, which is the rate-limiting step in the
aqueous humor secretion in the eye (Bowler et al., 1996).
1. Blood-Aqueous Barrier
The blood-aqueous barrier restricts the penetration of solutes such as acriavine (MW 540) into the posterior chamber as well as the anterior chamber
(Rodriguez-Peralta, 1975). With the blood-aqueous barrier, inward movement of solutes from the blood to the eye is more restrictive compared to the
outward movement. The blood-aqueous barrier is principally constituted by
two discrete layers of cells: the endothelium of the iris and ciliary blood
vessels and the nonpigmented ciliary epithelium. To pass from the blood
vessels of iris into the posterior chamber, a substance has to cross the iris
vessels, the stroma, a layer of iris muscle, and the iris epithelium. Transport
from the stroma into the anterior chamber is easier because the cellular layer
on the anterior surface of the iris is incomplete. In the anterior chamber
angle, there is continuous drainage of aqueous humor, which limits the
movement of solutes from the anterior chamber to posterior chamber. To
pass from the blood vessels of the ciliary body into the posterior chamber, a
solute has to cross the ciliary microvessels, the loose connective tissue of the
stroma, and the two-layered ciliary epithelium. The capillaries of the iris
have a relatively thick wall layered by the continuous-type endothelial cells
that are attached to each other by tight junctional complexes (Raviola, 1977;
Freddo and Raviola, 1982). The number of strands in the zonulae occludens
of these cells in the monkey varies from one to eight. As in the blood-brain
barrier (BBB), GLUT-l and P-glycoprotein are present in iris and ciliary

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

27

vessel endothelial cells, while endothelial antigen PAL-E (Pathologische

Anatomie Leiden-Endothelium) and the transferrin receptor are absent
(Schlingemann et al., 1998). This suggests a phenotypic similarity of iris
and ciliary vessels with brain microvessels. The iris vessels restrict the transport of circulating horseradish peroxidase (HRP) and uorescein due to the
presence of tight junctions between the endothelial cells. In this regard, the
vessels of the iris, retina, and brain share a common behavior (Hank et al.,
1990; Holash and Stewart, 1993; Stewart and Tuor, 1994). However, ciliary
vessels are permeable to circulating HRP, and it is speculated that it reaches
ciliary muscle by diusion from the permeable vessels of the ciliary processes.
The ciliary body epithelium is made of an outer pigmented epithelium
(PE) and an inner nonpigmented epithelium (NPE) juxtaposed at their apical surfaces (Fig. 6). The tight junctions between NPE cells impose a diusion barrier between blood and aqueous humor (Schlingemann et al., 1998).
Intravenously injected HRP passes through the fenestrated ciliary endothelial cells and reaches the intercellular spaces around PE cells and those
between PE and NPE cells. However, the tight junctions at the apico-lateral

Figure 6 Putative ion and solute transport processes in the mammalian pigmented
and nonpigmented epithelial cells of the ciliary body.

Copyright 2003 Marcel Dekker, Inc.

28

Sunkara and Kompella

fusion points of the NPE cells block further movement of HRP into the
posterior chamber (Schlingemann et al., 1998). Gap junctions are ubiquitous
in the ciliary body epithelium, connecting PE-to-PE, NPE-to-NPF, and PEto-NPE cells (Freddo, 1987) (Fig. 6). The gap junctions between PE and
NPE cells are permeable to molecules at least as large as 9001000 daltons
(Spray and Bennett, 1985).
In general, the breakdown of the blood-aqueous barrier takes place
when the mammalian eye is subjected to painful or irritant stimuli in
response to mechanical trauma or by carotid infusion of hyperosmotic
agents, leading to the leakage of plasma proteins into the aqueous humor
(Butler et al., 1988). This breakdown may be due to shrinkage of the pigmented epithelial cells, degeneration of the junctional complexes, and
separation of the two epithelial cell layers. In addition, stimulation of ocular
motor nerve, local administration of prostaglandin (PGE 1), and systemic
injection of
-MSH causes dilatation of iris and ciliary capillaries and
relaxation of aerent vessels, leading to the disruption of the blood-aqueous
barrier.

4. Ion and Organic Solute Transport

Aqueous humor formation was once considered a simple process of diusion or ultraltration of uid from the plasma across the ciliary epithelia. It
is now well established that active transport of certain solutes by the ciliary
epithelia is the most important process in the formation of aqueous humor
(Caprioli, 1992). Na+/K+-ATPase (Usukura et al., 1988; Okami et al.,
1989; Martin-Vasallo et at., 1989), H+/K+-ATPase (Fain et al, 1988) and
active ascorbate transport system (Socci and Delamere, 1988) are present in
the ciliary epithelia. Immunoblot analysis revealed the presence of Na+/
K+-ATPase mainly on the basolateral surfaces of PE and NPE cells, with
the labeling density being twofold higher in PE cells relative to NPE cells
(Okami et al., 1989). H+/K -ATPase is present extensively on the apical
membrane of both PE and NPE cells (Fain et al., 1988). In addition, the
basolateral membrane of pigmented layer contains the entry step for
Cl /Na+/K+ cotransporter, a Cl /HCO3 -exchanger, a Na+/HCO
3
cotransporter, a Na+/H+ exchanger, and a Na+/ascorbate cotransporter
(Zadunaisky, 1992). These systems produce low levels of sodium in the two
cells, a high level of potassium, and a high ascorbic content and control the
intracellular pH. The basolateral membrane of the nonpigmented epithelium contains the channels for the exit of chloride, sodium, bicarbonate, and
ascorbate into the posterior chamber and induce the movement of water for
the formation of the aqueous humor (Zadunaisky, 1992).

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

29

In spite of the tight barrier properties of ciliary epithelium, the concentration of glucose in aqueous humor is kept at a level similar to that in
plasma. This is facilitated by GLUT1, a glucose transporter localized in the
ciliary epithelial cells (Hank et al., 1990; Takata et al., 1997). The rate of
glucose transport into posterior chamber is very high. GLUT 1 at the basal
infoldings of NPE cells, which face the posterior chamber, may play a role in
the exit of glucose from the NPE cells to the aqueous humor (Fig. 6).
Immunocytochemical studies indicated a two fold higher expression of
GLUT1 in basal plasma membrane of PE cells compared to NPE cells.
From blood, glucose crosses fenestrated ciliary endothelial cells and then
enters PE cells by GLUT1 located in the basal infoldings of their plasma
membrane. From PE, glucose enters NPE cells by passing through the gap
junctions that connect PE and NPE cells. Finally, glucose leaves the NPE
cells via GLUT1 at the infolded basal plasma membrane and enters the
aqueous humor.
The concentration of most amino acids is higher in the aqueous than
in the plasma, possibly due to active transport of amino acids across the
ciliary epithelium in several mammalian species (Zlokovic et al., 1992, 1994).
In sheep, concentrations of aspartic acid and glutamic acid in the aqueous
were 913 times higher compared to plasma, and cystine and lysine levels
were approximately 1.7 times those in the plasma. For other amino acids,
such as alanine, arginine, glycine, isoleucine, leucine, methionine, phenylalanine, serine, threonine, tyrosine, and valine, the aqueous/plasma ratio was
greater than 2. A descriptive statistical study of the covariation of the concentration of amino acids and related compounds in human aqueous suggested the existence of six transport systems in the ciliary epithelia: three for
neutral amino acids and one each for basic amino acids, acidic amino acids,
and urea (Ehlers et at., 1978; Zlokovic et at., 1992). The protein concentration in the aqueous humor is always less than 1% of its plasma concentration, and it consists primarily of lower molecular weight proteins such as
albumin and -globulin. Heavy molecular proteins such as -lipoproteins
and heavy immunoglobulins are present only in trace quantities in the aqueous humor. Diusion of protein from the stroma of the ciliary body
through the iris stroma and into the aqueous humor has been proposed to
account for the major fraction of protein in the rabbit aqueous humor
(Freddo et al., 1990).
Ciliary epithelium expresses transporters to allow the transmembrane
ux of organic acids such as prostaglandins and eicosanoids (Bito, 1986).
Organic acids such as iodopyracet are actively removed from the eye when
introduced into the vitreous body. The ciliary epithelium contains at least
two systems for the transport of organic acids, the liver-like or L-system,
and the classical hippurate or H-system. These systems limit the accumula-

Copyright 2003 Marcel Dekker, Inc.

30

Sunkara and Kompella

tion of eicosanoids in the eye, which are produced in greater amounts under
pathological conditions (Bito and Wallenstein, 1977). The concentrative
accumulation of prostaglandins by the isolated anterior uvea is energy
and sodium dependent (Dibenedetto and Bito, 1980). Prostacycline, 6keto-PGF1
, and the classical E and F prostaglandins appear to be substrates for the same transport process. Tissues surrounding the anterior
chamber and those covering the surface of the eye, such as the cornea,
conjunctiva, sclera, and the anterior surface of the iris, do not appear to
possess transporters for prostaglandins (Bito, 1986).
In bovine pigment ciliary epithelial cells, Na+-dependent [14C] ascorbic acid accumulation was seen to proceed against a 40-fold intracellular
gradient (Helbig et at., 1989). This uptake was inhibited by phloretin, ouabain, amphotcriein, and D-isoascorbate, but not by glucose. The Na+
dependence of the uptake indicated that two or more Na+ ions translocate
with each molecule of ascorbate, resulting in an electrogenic transport. In
general transport terms, ocular ascorbic acid ts a type of pump-leak
model: the pump consists of ascorbic acid transport into aqueous humor;
the leak is combined loss of ascorbic acid by uid drainage through the
canal of Schlemm and chemical loss through oxidation. The ascorbic acid
uptake in excised iris-ciliary body is saturable and can be inhibited by Disoascorbate, ouabain, and metabolic inhibitors including dinitrophenol,
iodoacetate, and cyanide. The iris-ciliary body resembles retina, ileum,
and kidney in having an active transport process for ascorbic acid (Socci
and Delamere, 1988). Unlike the uptake process, the eux of [14C]ascorbic
acid was not inhibited by ouabain or other above-mentioned metabolic
inhibitors (Chu and Candia, 1988).
Endogenous nucleosides enter nonpigmented epithelial cells through a
sodium-nucleoside cotransporter localized on the apical side of the nonpigmented layer of the ciliary body facing the pigmented epithelium and exit
into the aqueous humor across the basolateral membrane via an equilibrative transport system (Redzic et at., 1998). Knowledge of the transport of
nucleosides through the blood-aqueous barrier is now made more important
with the increasing use of nucleoside analogs in the treatment of acquired
immunodeciency malignancies and syndrome (AIDS). As synthetic nucleosides are now known to be transported from blood into the aqueous humor
(Redzic et at., 1995), such molecules may compete for the carrier systems
used by endogenous nueleosides.
D.

Lens

The lens is a transparent tissue, with 65% of its weight consisting of water
and the remainder principally of proteins. Anteriorly, the lens is in contact

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

31

with the pupillary portion of the iris, and posteriorly it ts into a hollow
depression of the anterior vitreous surface (Paterson and Delamere, 1992).
The major components of the lens are capsule, epithelium, and lens ber
cells (Fig. 7). The lens capsule is acellular, transparent, elastic, and acts as an
unusually thick basement membrane that encloses the epithelium and lens
ber cells. It is mainly composed of type IV collagen together with 10%
glycosaminoglycans (GAGs). The inner portion of the anterior capsule is in
immediate contact with lens epithelium, while the posterior capsule is in
contact with the most supercial lens ber cells. A single layer of epithelial
cells forms a cap on the inner anterior surface of the lens, and hundreds of
thousands of dierentiated ber cells form its bulk. Lens cortex or periphery
is composed of ber cells that are formed from the dierentiation of epithelial cells in the equatorial zone. Young and supercial ber cells contain
cytoplasmic inclusions similar to those of epithelial cells. Since the lens does
not shed any of its cellular components from embryonic development
onward, the older cells of the lens will be displaced towards the center or
nucleus of the lens. Most of the cells in the lens nucleus lack nuclei and
particulate cytoplasmic contents. Membrane transport proteins in the lens

Figure 7 The lens. Representation of pump-leak system and cellular barriers in the
movement of ions and nutrients. The ber cells are distributed through out the lens
body.

Copyright 2003 Marcel Dekker, Inc.

32

Sunkara and Kompella

play an important role in cell volume regulation, nutrient supply, and lens
transparency (Goodenough, 1992, Rae, 1994). All cells of the lens are interconnected by gap junctions that form low-resistance pathways between the
cytoplasm of adjacent cells.
1. Lens Epithelium
The lens epithelium lies beneath the anterior and equatorial regions of lens
capsule, but not under the posterior region of the capsule. The apices of the
epithelial cells face the interior of the lens, and their bases face the lens
capsule. There are several junctional relationships between the cells of the
lens epithelium. The lateral and apical aspects of the plasma membranes
have desomosomes for cell adhesion. Zonula occludens is also present
between epithelial cells in both frog and human lens, and these junctions
could restrict the movement of high molecular weight solutes (Lo and
Harding, 1986). There is a small number of gap junctions between the
lens epithelial cells and the underlying lens bers, which allow cell-to-cell
communication (Brown et al., 1990). The cells of the lens epithelium are
anatomically and physiologically polarized, thereby bringing net transport
of many essential nutrients. While the lens epithelial cells are selectively
permeable to K+, lens ber cells have a low and nonselective overall conductance (Robinson and Patterson, 1982).
2. Ion and Organic Solute Transport
In the mammalian lens, the concentration of ions inside the cells plays a
major role in the development of cataract. Ions such as Ca2+ have been
shown to form aggregates with lens
-crystallins, thereby causing opacity
(Benedek, 1971; Benedek et al., 1999). Low intracellular calcium is maintained by Na+/Ca2+ exchanger and Ca2+-ATPase (Hightower et al, 1980).
Evidence exists for the presence of Na+/Ca2+ exchanger in the apical membrane of lens epithelial cells (Ye and Zadunaisky, 1992a,b). In rabbit,
bovine, dog, and rat lenses, while the lack of sodium did not aect the eux
of Ca2+ from the cells, inhibitors of Ca2+-ATPase, lanthanum and propranolol inhibited Ca2+ eux (Hightower et al, 1980). ATPase is present in
both the lens epithelium and cortex but absent in the nucleus, with the
expression being the highest in the epithelium (Ye and Zadunaisky,
1992b) (Fig. 8). Na+/H+ exchanger regulates intracellular pH in the lens
epithelial cells as well as the lens ber cells (Wolosin et al., 1988).
Because of lens avascularity, there is a need for the continuous supply
of nutrients such as glucose, amino acids, and ascorbic acid for metabolic
and synthetic reactions. Lens derives most of its nutrients from aqueous
humor. Entry of solutes into the lens occurs by both saturable and nonsa-

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

33

Figure 8 Putative ion and solute transport processes in the mammalian lens epithelial and ber cells.

turable paths (Merrimann-Smith et al., 1999). While water, chloride, urea,

and glycerol are likely to traverse the plasma membrane of the lens epithelial
cells by simple diusion, amino acids are transported actively. The uptake of
D-glucose and nonmetabolizable 3-O-methyl-D-glucose into the rat lens was
saturable and the capacity of the system for D-glucose was similar at both
anterior and posterior surfaces. Transport studies with a variety of sugars
indicated that a C-1 conformation and the presence of H or OH on carbon
number 1 are critical to transport. Furthermore, the GLUT family of proteins, GLUT 1 and GLUT3, is predominantly responsible for the uptake of
glucose in the anterior epithelium (Kern and Ho, 1973; Merrimann-Smith et
al., 1999). These transporters are abundant in the lens nucleus and present in
the cortex to a lower extent in both human and rat eyes.
Lens contains a higher concentration of amino acids compared to the
surrounding aqueous and vitreous humors. In vitro lens uptake studies with
labeled amino acids indicated saturation kinetics, energy and temperature
dependency, consistent with their active transport (Zlokovic et al., 1992).
Furthermore, separate systems exist for the transport of neutral, basic, and
acidic amino acids in the lens. Lens expresses A-, L-, Gly-, -, and Ly+systems (Kern et al., 1977). The ASC system, selective for three- or fourcarbon neutral amino acids, is present in the lens epithelial cell, and it is the

Copyright 2003 Marcel Dekker, Inc.

34

Sunkara and Kompella

preferred pathway at physiological levels of L-alanine, L-serine, and L-cystein

(Kern et al., 1977). The lens behaves like a pump-leak system in which the
substances actively transported by the epithelium are concentrated in the
lens and then diuse toward the posterior pole and eventually leave the lens
across the posterior capsule.
The rate of dierentiation of lens epithelial cells into ber cells is
dependent on the synthesis of phosphoinositides from myoinositol
(Zelenka and Vu, 1984). Since myoinositol synthesis in the lens is negligible,
membrane transport is the major source of cellular myoinositol. The lens
allows both sodium-dependent active transport and passive diusional
transport of myo-inositol (Diecke et al., 1995).
The oxidative balance in the lens is maintained by ascorbic acid, whose
primary source is aqueous humor (Kern and Zolot, 1987). The active uptake
of ascorbic acid by lens epithelium is 20 times greater than L-glucose. Within
7 minutes following systemic administration, [14C]ascorbic acid is concentrated more in the lens epithelium than the aqueous humor.
Nucleotides enter the lens via Na+-dependent transport processes,
with the uptake being comparable at the two surfaces of the lens, similar
to sugars (Redzic et al., 1998). Saturable uptake was observed for purines
such as guanosine, inosine, and adenosine, but not for pyrimidines and
adenine.
E. Blood-Retinal Barrier
Fully developed retina is organized into the following layers (Fig. 9): retinal
pigment epithelium (RPE), photoreceptor layer, outer limiting membrane,
outer nuclear layer, outer plexiform layer, inner nuclear layer, inner plexiform layer, ganglion cell layer, nerve ber layer, and inner limiting membrane (Wu, 1995). Large retinal vessels are present in the optic nerve ber
layer, and retinal capillaries are present between the inner nuclear layer and
the outer plexiform layer. The outer and inner limiting membranes of the
retina are quite permeable. The outer limiting membrane has traditionally
been described as a layer of zonulae adherens that connect Muller cells to
photoreceptors, and it is permeable to macromolecules. The inner limiting
membrane is a continuous glycoprotein coating identical to the basal lamina
of epithelia or endothelia. The inner limiting membrane contains the cell
bodies of most neurons and forms a basement membrane for the Muller and
glial cells and separates the base of the Muller cells from the vitreous body.
The intercellular clefts between the Muller cell processes that abut the inner
limiting membrane are open and lack specialized intercellular junctions.
Freeze-fracture studies conrmed the absence of zonulae occludens between
the basal feet of the Muller cells in both humans and monkey. The perme-

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

35

Figure 9 Cellular organization of the retina.

ability properties of inner limiting membrane are likely similar to the basal
laminae, which retain particulate matter but allow the transport of molecules up to the size of ferritin (400 kDa) (Wu, 1995). Among all the retinal
layers, the principal barrier to solute transport is the blood-retinal barrier
(BRB), which includes retinal pigment epithelium and retinal vessels.
The BRB is located at two levels: the outer BRB, consisting of the
retinal pigment epithelium (RPE), and the inner BRB, consisting of the
endothelial membranes of the blood vessels of the retina. BRB plays a
critical role in the homeostatis of the neural retina by limiting the entry of
xenobiotics into the extravascular spaces of the retina and by preventing the
loss of essential solutes. Blood-retinal barrier breakdown, a key factor
underlying pathological conditions such as diabetic retinopathy, is a leading
cause of vision loss. In neovascular disorders such as proliferative diabetic
retinopathy or age-related choroidal neovascularization, elevation of angiogenic factors such as vascular endothelial growth factor (VEGF165) and
transforming growth factor (TGF-), can disrupt the blood-retinal barrier
(Behzadian et al., 2001). Quantication of the degree of damage to bloodretinal barrier can be performed by locating 125I- and 135I-albumin tracers
(Miyamoto et al., 1999) or Evans blue in the retinal tissue and/or vitreous
following their systemic injection (Xu et al., 2001).
1.

Retinal Pigment Epithelium

RPE is a monolayer of hexagonal cells separating the outer surface of the

neural retina from the choriocapillaries. Many of these cells are multinu-

Copyright 2003 Marcel Dekker, Inc.

36

Sunkara and Kompella

cleated, and the nuclei are located in the basal portion of the cell. RPE plays
a central role in regulating the microenvironment surrounding the photoreceptors in the distal retina, where the phototransduction takes place. The
outer segments of rods and cones are closely associated with the RPE via
villous and pesudopodial attachments. The RPE phagocytoses distal
potions of rod and cone outer segments.
Following intravitreal injection, horseradish peroxidase crosses the
inner limiting membrane, ganglion cell layer, inner plexiform layer (synaptic
layer), inner nuclear membrane, outer limiting membrane, and the layer of
the photoreceptors but blocked by the tight junctions of the retinal pigment
epithelial cells (Tornquist et al., 1990). Thus, the retinal pigment epithelium
is a principal barrier to solute transport. Macromolecules such as horseradish peroxidase that escape from the permeable vessels of the choriocapillaries, cross Bruchs membrane and penetrate the intercellular clefts of the
retinal pigment epithelium. Further progression into the retina is blocked by
the junctional complexes of the RPE (Tornquist et al., 1990). In a number of
vertebrate species, these junctional complexes consist of zonula occludens,
zonula adherens, and gap junctions, and their surface specializations are
dierent from those observed in other epithelia (Hudspeth and Yee,
1973). The gap junctions lie apically (vitread), the zonula adherens lie basally
(sclerad), and the zonula occludens overlap the other two junctions. With the
presence of tight junctions, RPE forms a polarized monolayer of cells with
morphologically and functionally distinct apical and basolateral membranes. The apical membrane of RPE faces the photoreceptor outer segment
across the subretinal space, and the basolateral membrane is juxtaposed to
the choriocapillaries across Bruchs membrane. Various transporter proteins
are distributed in a polarized manner in RPE (Rodriguez-Boulan and
Nelson, 1989; Mays et al., 1994).
RPE is extremely restrictive for paracellular transport of solutes due to
the presence of tight junctions. However, it is capable of a variety of specialized transport processes (Betz and Goldstein, 1980). To understand the
barrier properties of RPE, experimental models such as isolated RPE-choroid preparations (Crosson and Pautler, 1982; Tsuboi and Pederson, 1988;
Joseph and Miller, 1991; Quinn and Miller, 1992; la Cour et al., 1994),
neural retina-RPE-choroid preparations (Shirao and Steinberg, 1987), and
conuent monolayers of cultured RPE (Defoe et al., 1994; Hernandez et al.,
1995; Gallemore et al., 1995) can be used. The electrical resistance of various
RPE and choroid preparations is summarized in Table 1. The electrical
resistance of RPE preparations ranges from 70 to 350 ohm-cm2 in various
species and preparations. The low in vitro resistance of these preparations
does not appear to be representative of the formidable in vivo blood-retinal
barrier.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

2.

37

Retinal Vessels

The studies led by Cunha-Vaz et al. (1979) proposed that the endothelial cells
along with their junctional complexes are the main sites of the blood-retinal
barrier for substances like thorium dioxide, trypan blue, and uorescein.
Many investigations demonstrated the barrier properties of the retinal
endothelium (Tomquist et al., 1990; Xu et al., 2001). Following intravenous
injection of thorium dioxide or horseradish peroxidase, tight junctions
blocked tracer progression along the clefts between the endothelial cells of
the retinal capillaries. Also, the tracer transport into the endothelial cytoplasm was negligible. Compared to several other blood vessels in the body,
the retinal vessels have more extensive zonula occludens, which render them
impermeable to both horseradish peroxidase (MW 40 kDa; hydrodynamic
radius 5 nm) and microperoxidase (MW 1.9 kDa; hydrodynamic radius 2
nm) (Smith and Rudt, 1975). These junctional complexes restrict solute
movement in either direction, as indicated by the inability of vitreal horseradish peroxidase in reaching the lumen of retinal vessels (Peyman and Bok,
1972). These junctions may restrict small solutes dierently because uorescein does not cross the blood-retinal barrier when injected intraperitoneally
into rabbits (Cunha-Vaz and Maurice, 1967), but it is easily absorbed by the
retinal vessels following vitreal administration, suggesting that the retinal
vessels and the pigment epithelium are capable of removing organic anions
from the vitreous. Poor permeability into the vitreous was seen for solutes
such as urea, sodium, potassium, chloride, phosphate, inulin, sucrose, antibiotics, and proteins such as albumin, myoglobin, and horseradish peroxidase following intravenous injection (Tornquist et al., 1990). Higher
quantities were observed in the anterior vitreous following entry from the
posterior chamber or ciliary circulation, suggesting that blood-aqueous barrier is more leaky compared to BRB. The BRB permeability of uorescein
with a molecular radius of 5.5 A was estimated to be about 0:14  105 cm/s,
which is similar to uorescein permeability across the blood-brain barrier,
suggesting that BRB and BBB are functionally similar (Vinores, 1995). The
passive transport of solutes across the BRB is 50 times lower than in most
other vessels of the body and 10 times lower than in the iris vessels (Tornquist
et al., 1990). Lipophilic substances such as rhodamine penetrate freely across
the BRB into the vitreous after systemic administration, with the blood-tovitreous ratio being 1 (Maurice and Mishima, 1984). Indeed, with increasing
drug partition coecients, the vitreal concentrations as well as the rate of
vitreal penetration of dierent antibiotics increased (Bleeker et al. 1968;
Lesar and Fiscella, 1985). The low permeability surface area products (PS)
for sucrose (0:44  105 mL/g/s) and mannitol (1:25  105 mL/g/s) across
BRB are similar to that of BBB (Lightman et al., 1987).

Copyright 2003 Marcel Dekker, Inc.

38

Sunkara and Kompella

3. Ion and Organic Solute Transport

There is substantial evidence for the vectorial transport of solutes across the
BRB. Various solute transport processes present in RPE are shown in
Figure 10. Unlike other epithelial tissues and similar to the choroid plexus,
RPE expresses Na+/K+-ATPase primarily in the apical membrane (Quinn
and Miller, 1992). RPE cells secrete Na+ actively. Indeed, the active 22Na
secretion was inhibited by apical ouabain (Miller and Edelman, 1990). RPE
cells of human as well as rat origins express tetrodotoxin-sensitive Na+
channels (Wen et al., 1994; Botchkin and Matthews, 1994). Besides Na+
secretion, Cl absorption is the principal contributor to the active ion transport across the RPE. Cl absorption is primarily determined by furosemideand bumetanide-sensitive Na+/K+/Cl cotransporter, which allows apical
Cl entry (La Cour, 1992). Ca2+- cAMP-activated Cl channels are present
on the basolateral side to allow Cl exit (Ueda and Steinberg, 1994; Strauss
et al., 1996). Also, patch-clamp studies on single cells demonstrated a swelling-activated Cl channel in RPE. Bovine and human RPE express CFTR,
a cAMP-regulated Cl channel (Miller et al., 1992). Na+/Ca2+ exchanger is
also present in the apical membrane of bovine and dogsh RPE cells
(Fijisawa et al., 1993).

Figure 10 Putative ion and solute transport processes in the mammalian retinal
pigmented epithelium.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

39

RPE has a reverse polarization of Na+/K+-ATPase, because this

pump is localized on the apical membrane as opposed to the basolateral
membrane. To determine whether such reverse polarization is also the case
with protein tracking and sorting in RPE cells, Bok et al. (1992) determined the budding of viruses whose progeny bud from specic membrane
domains in epithelial cells as directed by the sorting of their envelope glycoprotein. Upon infection of human and bovine RPE with these envoloped
viruses, cultured human and bovine RPE exhibited the same pattern of viral
budding as has been observed in other polarized epithelia, with the inuenza
hemagglutinin sorted to the apical membrane and the vesicular stomatitis
glycoprotein sorted to the basolateral membrane.
In addition to the above-mentioned ionic transport processes, there
are specic transporters that regulate the transport of nutrients and metabolites such as glucose, amino acids, nucleosides, folic acid, lactic acid,
ascorbic acid, and retinoids in the retina. A facilitated glucose transporter,
GLUT1, a 50 kDa protein, is expressed in various cells, including retinal
pigmented epithelial cells, choroidal cells, retinal Muller cells, and the outer
segments of the photoreceptor cells in the adult eye. Immunouorescence
and immunogold studies revealed that GLUT1 is present on both apical and
basolateral sites of RPE cells (Mantych et al., 1993). Immunoreactivity for
GLUT3, a 5055 kDa protein, was observed in the adult inner synaptic
layer of the retina (Mantych et al., 1993).
In the blood-retinal barrier of rats, carrier systems exist for the transport of neutral and basic amino acids (Tornquist and Alm, 1986; Tornquist
et al., 1990). Also, taurine and myo- inositol enter RPE cells via carriermediated transport mechanisms (Miyamoto et al., 1991). A purine nucleoside transporter is present in RPE cells and retinal neurons, indicated by an
increase in the accumulation of [3H]phenylisopropyl adenosine and
[3H]adenosine in the presence of nitrobenzylthioinosine, an inhibitor of
purine nucleoside transporter (Blazynski, 1991). The localization of these
transporters is still not clear. The apical reduced-folate transporter (RFT- 1)
and the basolateral folate receptor alpha (FR
) mediate vectorial transfer of
reduced folate from choroidal blood to the neural retina in mouse and
human RPE cells (Chancy et al., 2000).
To regulate lactate levels in the neural retina, RPE transports lactate
between two tissue compartments, the interphotoreceptor matrix and the
choriocapillaries. In isolated bovine RPE, Na+-lactate cotransporter located
on the basolateral side moves lactate out of cells and the apically localized
H+-lactate cotransporter moves lactate into the cells (Kenyon et al., 1994).
The transport of lactate and other monocarboxylates in mammalian cells is
mediated by a family of monocarboxylate transporters (MCTs), a group of
highly homologous proteins that reside in the plasma membrane of almost all

Copyright 2003 Marcel Dekker, Inc.

40

Sunkara and Kompella

cells and mediate the 1:1 electroneutral transport of a proton and a lactate
ion. MCT3 has been identied in RPE cells with basolateral distribution
(Yoon et al., 1997). Unlike GLUT1, MCT1 is highly expressed in the apical
processes of RPE and absent on the basal membrane of pigment epithelium
(Philip et al., 1998; Gerhart et al., 1999; Bergersen et al., 1999). MCT1 is also
associated with Muller cell microvilli, the plasma membranes of the rod inner
segments, and all retinal layers between the inner and external limiting membranes. MCT1 functions to transport lactate between the retina and the
blood at the level of retinal endothelium as well as the pigment epithelium.
MCT2, on the other hand, is abundantly expressed on the inner (basal)
plasma membrane of Muller cells and glial cells surrounding retinal microvessels. MCT4 is weekly expressed in RPE cells (Philip et al., 1998).
In primary or subcultured bovine and cat retinal pigment epithelium,
ascorbate transport was observed to be coupled to the movement of sodium
down its electrochemical gradient (Khatami et al., 1986). In cultured bovine
capillary pericytes, ascorbate was transported via facilitated diusion
(Khatami, 1987). The uptake was specic for ascorbate, and this process
was not sensitive to metabolic inhibition, the presence of ouabain, or the
removal of Na+ from the bathing medium, consistent with ascorbate entry
into the cells by facilitated diusion.
RPE cells are critical in the maintenance of the visual or retinoid cycle,
which involves the back-and-forth movement of vitamin A (retinol) and
some of its derivatives (retinoids) between the rods and cones (photoreceptors) and the RPE (Bok et al., 1984). Binding of retinol to retinol-binding
proteins such as cellular retinol-binding protein and an interphotoreceptor
retinoid-binding protein increases its solubility and delivers it to the RPE by
a receptor-mediated process. The presence of cellular retinol binding protein
and an interphotoreceptor retinoid-binding protein was shown in human,
monkey, bovine, rat, and mouse retinas (Bunt-Milam and Saari, 1983; Bok
et al., 1984). Cellular retinol-binding protein is predominantly localized in
the space that surrounds the photoreceptor outer segments and the apical
surface of RPE cells. It is also present in Muller glial cells.
Interphotoreceptor retinoid-binding protein is localized in the space bordered by three cell typesRPE, photoreceptor, and Mullerwhich is consistent with its proposed role in the transport of retinoids among cells.

IV.

DRUG EFFLUX PUMP

Drug eux pumps belong to a large family of ATP binding cassette (ABC)
transporter proteins. These pumps bind their substrate and export it
through the membrane using energy derived from ATP hydrolysis. The

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

41

original concept of multidrug resistance was introduced in 1970s to designate cells resistant to one drug, which develop cross-resistance to unrelated
drugs that bear no resemblance in structure or cellular target (Biedler and
Riehm, 1970). Multidrug resistance was rst associated with the presence of
one or more drug eux transporters such as P-glycoprotein (P-gp) and
multidrug resistanceassociated protein (MRP) in tumor cells. Current evidence suggests that these transporters are present in the normal tissues, and,
therefore, they may play a role in the drug disposition (Lum and Gosland,
1995). In this section, the expression of P-gp and MRP in various ocular
epithelia is summarized.

A.

P-Glycoprotein

P-gp, a 170 kDa membrane glycoprotein, is expressed in tumor cells as well

as various normal tissues associated with the eye, small intestine, liver,
kidney, lung, and the blood-brain barrier (Thiebaut et al., 1989). In the
eye, P-gp is expressed in retinal capillary endothelial cells, iris, and ciliary
muscle cells (Holash and Stewart, 1993), and retinal pigmented
(Schlingemann et al., 1998), ciliary nonpigmented (Wu et al., 1996), corneal
(Kawazu et al., 1999), and conjunctival epithelial (Saha et al., 1998) cells. Pgp exports a variety of structurally and pharmacologically unrelated hydrophobic compounds such as vinblastine, vincristine, cyclosporin (CsA), glucocorticoids, and lipophilic peptides such as N-acetyl-leucyl-leucylnorleucinal (Sharma et al., 1991; Hunter et al., 1991, Tsuji et al., 1992).
Rhodamine-123 (Rho-123), a ourescent P-gp substrate, is often used
to examine the function of P-gp in vitro and in vivo. Kajikawa et al. (1999)
determined the contribution of P-gp to rabbit blood-aqueous barrier by
analyzing the distribution of Rho-123 in aqueous humor after intravenous
injection in the presence or absence of topically administered P-gp inhibitors. Following topical quinidine (12.5 mM eye drops, ve applications at 5minute intervals) and cyclosporin A (12.5 mM eye drops, ve applications at
5-minute intervals) treatments, the aqueous humor distribution of intravenously injected Rho-123 was increased 11.2- and 11.3-fold, respectively,
compared to controls. This suggests that P-gp is functionally active in the
blood-aqueous barrier. Functional studies with other P-gp substrates, such
as cyclosporin A, verapamil, and dexamethasone, were also investigated to
determine the presence of P-gp activity in the conjunctival epithelium (Saha
et al., 1998). The basolateral-apical apparent permeability coecients of
cyclosporin A, verapmil, and dexamethasone were 9.3, 3.4, and 1.6 times
that of apical-basolateral permeability coecients, respectively. Cyclosporin
A eux was reduced by 5070% in the presence of verapamil and anti-Pgp

Copyright 2003 Marcel Dekker, Inc.

42

Sunkara and Kompella

antibody (4E3 mAb), consistent with P-gp activity in the conjunctival

epithelium.
B.

Multidrug ResistanceAssociated Protein

MRP belongs to a group of mammalian ABC transporters that can be

clearly dierentiated from others such as P-gp or cystic brosis transmembrane regulator (CFTR). The most important dierence in MRP and P-gp
is that MRP is an organic anion transporter with high activity towards
compounds conjugated to glutathione (GSH), glucuronide, or sulfate
(Muller et al., 1994; Jedlitschky et al., 1994). Currently, MRP is known to
exist in seven isoforms, including MRP1, MRP2, MRP3, MRP4, MRP5,
MRP6, and MRP7 (Cole et al., 1992; Flens et al., 1996; Kool et al., 1997,
1999a,b). In polarized epithelial cells, MRP1, MRP3, and MRP5 are usually
present on the basolateral side, whereas MRP2 is present on the apical
surface.
The expression of MRP 1 has been demonstrated in the human retinal
pigment epithelial cell line (ARPE- 19) and in the primary cultures of human
retinal pigment epithelium (HRPE) using Western blot analysis and RTPCR studies (Aukunuru and Kompella, 1999a; Aukunuru et al., 2001).
Accumulation of uorescein, an MRP substrate, was increased in both
ARPE-19 and HRPE cells in the presence of MRP inhibitors, including
probenecid, verapamil, and indomethacin. Similar observations were made
in ARPE-19 cells with benzoylaminophenylsulfonylglycine (BAPSG), an
anionic aldose reductase inhibitor intended for diabetic complications
(Fig. 11). Thus, MRP inhibition may enhance the drug uptake into RPE
cells. MRP1 is present in rabbit conjunctival epithelial cells as indicated by a
190 kDa protein corresponding to MRP1 in Western blots (Yang and Lee,
2000). The apical to basolateral transport of leukotriene (LTC4), an MRP
substrate, was abolished by basolateral probenecid, suggesting that MRP is
likely localized on the basolateral side.

V. DRUG TRANSPORT ACROSS OCULAR BARRIERS

The general goal of ophthalmic drug delivery is to maximize drug levels in
the target eye tissues while minimizing the levels in the remainder of the
body. Ocular delivery can be achieved by topical administration, systemic
administration, periocular injections, and intraocular injections. Topical
administration of drugs results in higher drug concentrations in the extraocular barriers (conjunctiva, cornea), followed by the anterior chamber and
its structures (aqueous humor, lens), with minimal drug entering the poster-

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

43

Figure 11 Inuence of MRP inhibitors on N-4(benzoylamino)phenylsulfonyl glycine (BAPSG) uptake in ARPE-19 cells (
P < 0:05. (From Aukunuru et al., 2001.)

ior ocular structures including vitreous, retina, and choroid. Following

systemic administration, the concentration of drug obtained within the
eye depends on the blood concentration-time prole of the free drug and
the rate of drug clearance from the eye compartments. The blood-ocular
barriers, which include the iris blood vessel endothelium, ciliary body
epithelium, retinal blood vessel endothelium, and retinal-pigmented epithelium, inuence the overall amount of drug entering the anterior and posterior compartments of the eye from blood. As the blood-retinal barrier
presents a greater hindrance to drug penetration than does the blood-aqueous barrier, the aqueous humor concentrations are typically higher than
vitreous humor concentrations following systemic administration of drugs
(Lesar and Fiscella, 1985). Periocular injections such as subconjunctival
injections and intravitreal injections place drug more proximal to the target
tissues, thereby overcoming some of the ocular barriers. In this section,
physiochemical factors that aect the transport across blood-ocular barriers
and case reports of drug transport are presented.
Ocular tissue permeability following topical administration can be
inuenced by the physicochemical properties of the drug. Molecular size
restricts paracellular transport in cornea, conjunctiva, and sclera. The paracellular and aqueous penetration routes in cornea, conjunctiva, and sclera
were characterized in rabbits using a mixture of polyethyleneglycols (PEGs)
with mean molecular weights of 200, 400, 600, and 1000, which display some
basic features of peptides and oligonucleotides (hydrophilicity, hydrogen-

Copyright 2003 Marcel Dekker, Inc.

44

Sunkara and Kompella

bonding capacity, and size) (Hamalainen et al., 1997). The conjunctival and
scleral tissues were more permeable to PEGs than the cornea. The conjunctival permeability was less inuenced by molecular size compared to that of
cornea, which is expected because the conjunctiva has 2 times larger pores
and 16 times higher pore density than the cornea. The total paracellular
space in the conjunctiva was estimated to be 230 times greater than that in
the cornea. Conjunctiva is commensurately permeable to hydrophilic molecules up to 40 kDa. Prausnitiz and Noonan (1998) summarized the corneal, conjunctival, and scleral penetration of various drugs as a function of
lipophilicity, molecular size, molecular radius, partition coecient, and distribution coecient. They observed an increase in corneal as well as corneal
endothelial permeability with an increase in the drug distribution coecient.
Cornea as well as corneal endothelium exhibited molecular size-dependent
drug permeability. Conjunctiva did not show clear dependence on distribution coecient, but it did show a possible dependence on molecular size.
Scleral transport was not dependent on either molecular radius or distribution coecient.
Size is one determinant that inuences the transport of molecules
across blood-ocular barriers (Bellhorn, 1981). In a systematic study, the
permeability of the ocular blood vessels and neuroepithelial layers in neonatal and adult cats was assessed using FITC-dextrans of various sizes. The
iris and ciliary vessels were permeable to molecules with eective diusion
radius as large as 85 A. The choriocapillaries were permeable to molecules
with an eective diusion radius of 3258 A. Iris vessels in humans, monkey, rabbit, and rat were not permeable to free and proteinbound sodium
uorescein, whereas marked permeability was observed in cats (Sherman
et al, 1978).
Conjunctiva expresses organic cation transport processes (Ueda et al.,
2000). The permeability of guanidine and tetramethylammonium in the
mucosal-to-serosal direction was temperature and concentration dependent, and it was much greater than that in the serosal-to-mucosal direction.
Guanidine transport was also inhibited by dipivefrine (72%), brimonidine
(70%), and carbachol (78%). Also, acidication of mucosal uid, apical
exposure of a K+ ionophore, as well as high K+ levels reduced the transport of guanidine. However, it was not aected by the serosal presence of
0.5 mM ouabain. These observations suggest that transport of certain
amine-type ophthalmic drugs may be driven by an inside-negative apical
membrane potential dierence.
Propranolol transport was assessed in conjunctival epithelial cells in
the presence and absence of P-gpcompeting substrates and anti-P-gp
monoclonal antibody or a metabolic inhibitor, 2,4-DNP (Yang et al.,
2000). Propranolol was transported preferentially in the basolateral-to-api-

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

45

cal direction. When exposed apically, inhibitors of P-gp, cyclosporin A,

progesterone, rhodamine 123, verapamil, and 2,4-DNP increased propranolol accumulation by 4366%. These results suggest that P-gp is likely localized on the apical plasma membrane to restrict the conjunctival absorption
of some lipophilic drugs.
Cornea, conjunctiva, RPE, and iris pigment epithelial cells exhibit particulate uptake processes (Mayerson and Hall, 1986; Zimmer et al., 1991;
Rezai et al., 1997). Zimmer et al. (1991) determined the uptake of 120 nm
particles in excised rabbit cornea and conjunctiva. Following 30 minutes of
incubation of rhodamine 6G nanoparticle suspension, particle uptake was
observed in both tissues. No particles were observed in intercellular junctions
probably because the openings for the intercellular space are too small for the
120 nm particles. Also, penetration of uorescein was not seen in these cells
or across the whole cornea when uorescein solution instead of particles were
used, suggesting the superiority of nanoparticles as a drug delivery system.
Penetration through whole corneal tissue did not occur either. RPE cells
possess nonspecic phagocytic activity and are capable of binding and ingesting latex particles (Mayerson and Hall, 1986; Aukunuru and Kompella,
1999b, 2002). Also, rod outer segments enter RPE via phagocytosis, which
involves recognition, attachment, internalization, and degradation of the rod
outer segments. With respect to particle uptake, iris pigment epithelial cells
are functionally similar to RPE cells (Rezai et al., 1997).
Another important factor that may limit the ocular concentrations of
some classes of drugs is the presence of an active transport system that
removes drugs from ocular compartments and drains into blood. Barza et
al. (1982) determined the kinetics of intravitreally injected carbenicillin, an
organic anion antimicrobial, in rabbits following concomitant intraperitoneal administration of probenecid, an organic anion transport inhibitor.
Probenecid increased the vitreous half-life of carbenicillin from 5 to 13
hours. In addition, it increased the drug concentrations in cornea, aqueous,
and iris. Similar observations were made in rhesus monkeys, wherein probenecid increased the vitreous half-life of carbenicillin and cefazolin from 10
to 20 hours and 7 to 30 hours, respectively (Barza et al., 1983). These
ndings suggest that a probenecid-sensitive active transport system present
in the retinal pigmented epithelium may actively remove organic acids such
as penicillins and cephalosporins. Lipid-soluble compounds are also lost
through the retina due to their ability to cross the blood-retinal barrier
(Barza et al., 1982, 1983). Confounding ocular inammation results in elimination of nontransported drugs such as aminoglycosides (Barza et al.,
1983). However, the eect of inammation on the loss of actively transported carbenicillin from the vitreous is less due to a decrease in the function
of the active transport systems (Lesar and Fiscalle, 1985).

Copyright 2003 Marcel Dekker, Inc.

46

Sunkara and Kompella

From the drug delivery point of view, monocarboxylate transport

processes such as proton or Na+-coupled lactate systems in epithelial cells
may serve as conduits for anionic drugs such as cromolyn, which is used in
the treatment of vernal conjunctivitis and urbiprofen and dicolfenac, which
are used in the treatment of herpes conjunctivitis. Evidence exists for a Na+dependent carrier-mediated monocarboxylate transport process on the
mucosal side of the conjunctival epithelium (Horibe et al., 1998). This process may be used by ophthalmic nonsteroidal anti-inammatory drugs
(NSAIDs) and uoroquinolone antibacterial drugs. While NSAIDs and
uoroquinolones reduced L-lactate transport across conjunctiva, cromolyn
and prostaglandins (PGE2 and PGF2) did not aect L-lactate transport,
probably because cromolyn is a dicarboxylic acid and the hydrophobicity
of PGE2 and PGF2 may hamper their recognition by the monocarboxylate
transporter (Nord et al., 1983).
Besides Na+-lactate transporter, various ion transport processes discussed in Sec. III are likely to inuence drug transport. The various ion
transport processes can inuence cell surface pH, uid transport, tight junctional permeability, and vesicular tracking, thereby altering drug transport
(Kompella and Lee, 1999). Active transport of ions such as Na+, K+, and
Cl contributes to net uid transport across various epithelia. For instance,
cornea and conjunctiva secrete Cl towards tears. In association with this
Cl secretion, net uid secretion occurs towards tears. A change in this uid
secretion is likely to aect the transport of hydrophilic solutes. Exposure of
nutrients such as amino acids to the apical side of conjunctiva can induce
Na+ absorption in association with uid absorption. This uid absorption
in conjunction with possible opening of tight junctions by these amino acids
can allow increased absorption of hydrophilic solutes across conjunctiva.
Elevation of intracellular Ca2+ through various transporters such as Ca2+
channels and Na+/Ca2+ exchanger is another likely approach to increase
paracellular permeability. Function of Cl channels such as CFTR correlate
with the extent of endocytosis in various cells. Activation of Cl channels
with 8Br-cAMP and terbutaline have been shown to increase the transport
of horseradish peroxidase (Kompella and Lee, 1999).
pH in the microclimate of the cell surface can be dierent from that in
the surrounding uid bulk. This is due to unstirred water layers and the
presence of several surface transporters such as Na+/H+ exchanger, Na+/


HCO
3 cotransporter, and Cl /HCO3 exchangers. Because transcellular diffusion of drugs is dependent on partitioning, which is dependent on pH, the
function of these transporters is likely to inuence drug transport. Indeed,
inhibition of Na+/H+ exchanger with hexamethylene amiloride has been
shown to elevate drug transport across conjunctiva (Kompella and Lee,
1999).

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

VI.

47

CONCLUSIONS

The principal membrane barriers located in the cornea, conjunctiva, irisciliary body, lens, and retina express highly specialized transport processes
that control the movement of endogenous as well as exogenous solutes into
and out of intraocular chambers. Ion transport processes such as Na+/K+ATPase, Na+/K+/Cl and Na+/Cl cotransporter, K+, Cl , and Na+
channels are primarily responsible for the maintanence of potential dierences and uid transport across various cellular barriers. Ion transport
+
processes such as Na+/H+ exchanger, Cl /HCO
3 exchanger, and Na 
HCO3 cotransporter regulate cellular pH. Transporters such as Na+/
Ca2+ exchanger and Ca2+-ATPase regulate the cellular levels of Ca2+, a
messenger involved in multiple membrane transport events. By regulating
uid transport, cellular pH, and intracellular Ca2+, ion transport processes
are likely to inuence drug transport.
The various ocular barriers express transport processes for hydrophilic
solutes such as organic anions, glucose, amino acids, and nucleosides in one
or both sides of the cell. These transporters may play a role in the transport
of structurally related drug molecules. In addition, the ocular barriers
express eux pumps such as MRP and P-gp, which can export several
structurally diverse anionic drugs and lipophilic drugs, respectively.

ACKNOWLEDGMENTS
We are thankful to Mr. Jithuan V. Aukunuru for his editorial assistance in
the preparation of the manuscript.

REFERENCES
Aukunuru, J. V., Sunkara, G., Bandi, N., Thoreson, W. B., and Kompella, U. B.
(2001). Expression of multidrug resistance protein (MRP1) in retinal pigment
epithelial cells, and its interaction with BAPSG, a novel aldose reductase
inhibitor. Pharm. Res 18(5):565572.
Aukunuru, J. V., and Kompella, U.B. (1999a). Functional and biochemical evidence
for the expression of multidrug resistance associated protein (MRP) in a
human ertinal pigment epithelial cell line. PharmSci., 1(4):2609.
Aukunuru, J. V., and Kompella, U. B. (1999b). Size- and concentration-dependent
uptake of microspheres by human retinal pigment epithelial cells. PharmSci.,
1(4):3294.

Copyright 2003 Marcel Dekker, Inc.

48

Sunkara and Kompella

Aukunuru, J.V., and Kompella, U.B. (2002). In vitro delivery of nano- and microparticles to retinal pigment epithelial cells. Drug Del. Technol., 2:5057.
Barza, M., Kane, A., and Baum, J. (1982). The eects of infection and probenecid on
the transport of carbenicillin from the rabbit vitreous humor. Invest.
Ophthalmol. Vis. Sci., 22:720726.
Barza, M., Kane, A., and Baum, J. (1983). Pharmacokinetics of intravitreal carbenicillin, cefazolin and gentamicin in rhesus monkeys. Invest. Ophthalmol. Vis.
Sci., 24:16021606.
Behzadian, M. A., Wang, X., Windsor, J., Ghaly, N., and Caldwell, R. B. (2001).
TGF- increases retinal endothelial cell permeability by increasing MMP-9:
possible role of glial cells in endothelial barrier function. Invest. Ophthalmol.
Vis. Sci., 42(3):853859.
Bellhorn, R. W. (1981). Permeability of blood-ocular barriers of neonatal and adult
cats to uorescein-labeled dextrans of selected molecular sizes. Invest.
Ophthalmol. Vis. Sci., 21(2):282290.
Benedek, G. B. (1971). Theory of transparency of the eye. Appl. Optics, 10:459473.
Benedek, G. B., Pande, J., Thurston, G. M., and Clark, J. I. (1999). Theoretical and
experimental basis for the inhibition of cataract. Prog. Retin. Eye Res.,
18(3):391402.
Bergersen, L., Johannsson, E., Veruki, M. L., Nagellius, E. A., Halestrap, A.,
Sejersted, O., M., and Ottersen, O. P. (1999). Cellular and subcellular expression of monocarboxylate transporters in the pigment epithelium and retina of
the rat. Neuroscience, 90(1):319331.
Betz, A. L., and Goldstein, G. W. (1980). Transport of hexoses, potassium and
neutral amino acids into capillaries isolated from bovine retina. Exp. Eye
Res., 30(5):593605.
Biedler, J. L., and Riehm, H. (1970). Cellular resistance to actinomycin D in Chinese
hamster cells in vitro: cross-resistance, radioautographic, and cytogenetic studies. Cancer Res., 30(4):11741184.
Bito, L. Z. (1986). Prostaglandins and other eicosanoids: their ocular transport,
pharmacokinetics and therapeutic eects. Trans. Ophthalmol. Soc. UK,
105(Pt2):162170.
Bito, L. Z., and Wallenstein, M. C. (1977). Transport of prostaglandins across the
blood-brain and blood-aqueous barriers and the physiological signicance of
these absorptive processes. Exp. Eye Res., 25(suppl):229243.
Blazynski, C. (1991). The accumulation of [3 H]phenylisopropyl adenosine ([3 H]PIA)
and [3 H]adenosine into rabbit retinal neurons is inhibited by nitrobenzylthioinosine (NBI). Neurosci. Lett., 121(12):14.
Bleeker, G. M., van Haeringen, N. J., Mass, E. R., and Glasius, E. (1968). Selective
properties of the vitreous barrier. Exp. Eye. Res., 7(1):3746.
Bok, D., Ong, D. E., and Chytil, F. (1984). Immunocytochemical localization of
cellular retinol binding protein in the rat retina. Invest. Ophthalmol. Vis. Sci.,
25(8):877883.
Bok, D., ODay, W., and Rodriguez-Boulan, E. (1992). Polarized budding of vesicular stomatitis and inuenza virus from cultured human and bovine retinal
pigment epithelium. Exp. Eye Res., 55(6):853860.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

49

Bonanno, J. A. (1990). Lactate-proton cotransport in rabbit corneal epithelium.

Curr. Eye Res., 9(7):707712.
Bonanno, J. A., Yi, G., Kang, X. J., and Srinivas, S. P. (1998). Reevaluation of Cl /
HCO
exchange in cultured bovine corneal endothelial cells. Invest.
3
Ophthalmol. Vis. Sci., 39(13):27132722.
Bonanno, J. A., Guan, Y., Jelamskii, S., and Kang, X. J. (1999). Apical and basolateral CO2 -HCO
3 permeability in cultured bovine corneal endothelial cells.
Am. J. Physiol., 277(3 pt 1):C545C553.
Botchkin, L. M., and Matthews, G. (1994). Voltage-dependent sodium channels
develop in rat retinal pigment epithelium cells in culture. Proc. Natl. Acad.
Sci. USA, 91:45644568.
Bowler, J. M., Peart, D., Purves, R. D., Carre, D. A., Macknight, A. D. and Civan,
M. M. (1996). Electron probe X-ray microanalysis of rabbit ciliary epithelium.
Exp. Eye. Res., 62(2):131139.
Brown, H. G., Pappas, G. D., Ireland, M. E., and Kuszak, J. R. (1990).
Ultrastructural, biochemical, and immunologic evidence of receptor-mediated
endocytosis in the crystalline lens. Invest. Ophthalmol. Vis. Sci., 31(12):2579
2592.
Bunt-Milam, A. H., and Saari, J. C. (1983). Immunocytochemical localization of two
retinoid-binding proteins in vertebrate retina. J. Cell. Biol., 97(3):703712.
Butler, J. M., Unger, W. G., and Grierson, I. (1988). Recent experimental studies on
the blood-aqueous barrier: the anatomical basis of the response to injury, Eye,
2:S213S220.
Caprioli, J. (1992). The ciliary epithelia and aqueous humor. In: Adlers Physiology
of the Eye. Hart, W. M., Jr. (Ed.). Mosby-Year Book, Inc., St. Louis, pp. 228
247.
Chancy, C. D., Kekuda, R., Huang, W., Prasad, P. D., Kuhnel, J. M., Sirotnak, F.,
Roon, P., Ganapathy, V., and Smith, S. B. (2000). Expression and dierential
polarization of the reduced-folate transporter-1 and the folate receptor
in
mammalian retinal pigment epithelium. J. Biol. Chem., 275(27):2067620684.
Chang, J. E., Basu, S. K., and Lee, V. H. (2000). Air-interface condition promotes
the formation of tight corneal epithelial cell layers for drug transport studies.
Pharm. Res., 17(6):670676.
Chu, T. C., and Candia, O. A. (1988). Active transport of ascorbate across the
isolated rabbit ciliary epithelium. Invest. Ophthalmol. Vis. Sci., 29(4):594
599.
Cilluo, M. C., Fain, M. J., and Fain, G. L. (1993). Tissue culture of rabbit ciliary
body epithelial cells on permeable support. Exp. Eye Res., 57(5):513526.
Cole, S. P., Bhardwaj, G., Gerlach, J. H., Mackie, J. E., Grant, C. E., Almquist, K.
C., Stewart, A. J., Kurz, E. U., Duncan, A. M., and Deeley, R. G. (1992).
Overexpression of a transporter gene in a multidrug-resistant human lung
cancer cell line. Science, 258(5088);16501654.
Crosson, C. E., and Pautler, E. L. (1982). Glucose transport across isolated bovine
pigment epithelim. Exp. Eye Res., 35:371377.
Cunha-Vaz, J. G. (1979). The blood-ocular barriers. Surv. Ophthalmol., 23(5):279
296.

Copyright 2003 Marcel Dekker, Inc.

50

Sunkara and Kompella

Cunha-Vaz, J. G., and Maurice, D. M. (1967). The active transport of uorescein by

the retinal vessels and the retina. J. Physiol., 191(3):467486.
Defoe, D. M., Ahmad, A., Chen, W., and Hughes, B. A. (1994). Membrane polarity
of the Na+-K+ pump in primary cultures of Xenopus retinal pigment epithelium. Exp. Eye Res., 59(5):587596.
Dibeneditto, F. E., and Bito, L. Z. (1980). The kinetics and energy dependence of
prostaglandin transport processes. I. In vitro studies on the rate of PGF2

accumulation by the rabbit anterior uvea. Exp. Eye Res., 30(2):175182.
Diecke, F. P., Beyer-Mears, A., and Mistry, K. (1995). Kinetics of myo-inositol
transport in rat ocular lens. J. Cell Physiol., 162(2):290297.
Ehlers, N., Kristensen, K., and Scholnheyder, F. (1978). Amino acid transport in
human ciliary epithelium. Acta Ophthalmol., 56(5):777784.
Fain, G. L., Smolka, A., Cilluo, M. C., Fain, M. J., Lee, D. A., Brecha, N. C., and
Sachs, G. (1988). Monoclonal antibodies to the H+-K+ ATPase of gastric
mucosa selectivity stain the non-pigmented cells of the rabbit ciliary body
epithelium. Invest. Ophthalmol. Vis. Sci., 29(5):785794.
Fijisawa, K., Ye, J., and Zadunaisky, J. A. (1993). A Na+/Ca2 exchange mechanism in apical membrane vesicles of the retinal pigment epithelium. Curr. Eye
Res., 12(3):261270.
Flens, M. J., Zaman, G. J., van der Valk, P., Izquierdo, M. A., Schroeijers, A. B.,
Scheer, G. L., van der Groep, P., de Haas, M., Meijer, C. J., and Scheper, R.
J. (1996). Tissue distribution of the multidrug resistance protein. Am. J.
Pathol., 148(4);12371247.
Freddo, T. F. (1984). Intercellular junctions of the iris epithelia in Macaca mullata.
Invest Ophthalmol. Vis. Sci., 25(9):10941104.
Freddo, T. F. (1987). Intercellular junctions of the ciliary epithelim in anterior
uveitis. Invest. Ophthalmol. Vis. Sci., 28(2):320329.
Freddo, T. F., and Raviola, G. (1982). Freeze-fracture analysis of the interendothelial junctions in the blood vessels of the iris in Macaca mullata. Invest.
Ophthalmol. Vis. Sci., 23(2):154167.
Freddo, T. F., Bartles., S. P., Barsotti, M. F., and Kamm, R. D. (1990). The source
of proteins in the aqueous humor of the normal rabbit. Invest. Ophthalmol.
Vis. Sci., 31(1):125137.
Gallemore, R. P., Hu, J., Fambach, D. A., and Bok, D. (1995). Ion transport in
cultured fetal human retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci.,
36:S316.
Gerhart, D. Z., Leino, R. L., and Drewes, L. R. (1999). Distribution of monocarboxylate transporters MCT1 and MCT2 in rat retina. Neuroscience, 92(1):367
375.
Giasson, C., and Bonanno, J. A. (1994). Facilitated transport of lactate by rabbit
corneal endothelium. Exp. Eye Res., 59(1):7381.
Gliula, N. B., Reeves, O. R., and Steinbach, A. (1972). Metabolic coupling, ionic
coupling and cell contacts. Nature, 235(5336):262265.
Goodenough, D. A. (1992). The crystalline lens. A system networked by gap junctional intercellular communication. Semin. Cell. Biol., 3(1):4958.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

51

Green, K. (1970). Stromal cation binding after inhibition of epithelial transport in

the cornea. Am. J. Physiol., 218(6):16421648.
Hamalainen, K. M., Kananen, K., Auriola, S., Kontturi, K., and Urtti, A. (1997).
Characterization of paracellular and aqueous penetration routes in cornea,
conjunctiva, and sclera. Invest. Ophthalmol. Vis. Sci., 38(3):627634.
Harik, S. I., Kalaria, R. N., Whitney, P. M., Andersson, L., Lundahl, P., Ledbetter,
S. R., and Perry, G. (1990). Glucose transporters are abundant in cells with
occluding junctions at the blood-eye barriers. Proc. Natl. Acad. Sci. USA,
87(11):42614264.
Helbig, H., Korbmacher, C., Wohlfarth, J., Berweck, S., Kuhner, D., and Widerholt,
M. (1989). Electrogenic Na+-ascorbate cotransport in cultured bovine pigmented ciliary epithelial cells. Am. J. Physiol., 256(1 Pt 1):C44C49.
Hernandez, E., Hu, J., Frambach, D., and Gallemore, R. (1995). Potassium conductances in cultured bovine and human retinal pigment epithelium. Invest.
Ophthalmol. Vis. Sci., 36:113122.
Hightower, K. R., Leverenz, V., and Reddy, V. N. (1980). Calcium transport in the
lens. Invest. Ophthalmol. Vis. Sci., 19(9):10591066.
Ho, N. F. H., Raub, T. J., Burtoon, P. S., Barsutin, C. L., Adson, A., Audus, K. L.,
and Borchardt, R. T. (1999). Quantitative approaches to delineate passive
transport mechanisms in cell culture monolayers. In: Transport Processes in
Pharmaceutical Systems. Amidon, G. L., Lee, P. I., and Topp, E. M. (Eds.).
Marcel Dekker, Inc., New York, pp. 219316.
Holash, J. A., and Stewart, P. A. (1993). The relationship of astrocyte-like cells to the
vessels that contribute to the blood-ocular barriers. Brain Res., 629(2):218
124.
Horibe, Y., Hosoya, K., Kim, K. J., and Lee, V. H. L. (1998). Carrier mediated
transport of monocarboxylate drugs in the pigmented rabbit conunctiva.
Invest. Ophthalmol. Vis. Sci., 39(8):14361443.
Hosoya, K., Kompella, U. B. Kim, K. J., and Lee, V. H. (1996). Contribution of
Na(+)-glucose cotransport to the short-circuit current in the pigmented rabbit
conjunctiva. Curr. Eye Res., 15(4):447451.
Hosoya, K. I., Horibe, Y., Kim, K. J., and Lee, V. H. (1998a). Carrier-mediated
transport of NG-nitro-L-arginine, a nitric oxide synthase inhibitor, in the
pigmented rabbit conjunctiva. J. Pharmacol. Exp. Ther., 285(1):223227.
Hosoya, K., Horibe, Y., Kim, K. J., and Lee, V. H. (1998b). Nucleoside transport
mechanisms in the pigmented rabbit conjunctiva. Invest. Ophthalmol. Vis. Sci.,
39(2);372377.
Hudspeth, A. J., and Yee, A. G. (1973). The intercellular junctional complexes of
retinal pigment epithelia. Invest. Ophthalmol., 12(5):354365.
Hunter, J., Hirst, B. H., and Simmons, N. L. (1991). Transepithelial vinblastine
secretion mediated by P-glycoprotein is inhibited by forskolin derivatives.
Biochem. Biophys. Res. Commun., 181(2):671676.
Jedlitschky, G., Leier, I., Buchholz, U., Center, M., and Keppler, D. (1994). ATPdependent transport of glutathione S-conjugates by the multidrug resistanceassociated protein. Cancer Res., 54(18):48334836.

Copyright 2003 Marcel Dekker, Inc.

52

Sunkara and Kompella

Jelamskii, S., Sun, X. C., Herse, P., and Bonanno, J. A. (2000). Basolateral Na+K+-2Cl cotransport in cultured and fresh bovine corneal endothelium.
Invest. Ophthalmol. Vis. Sci., 41(2):488495.
Jentsch, T. J., Keller, S. K., and Wiederholt, M. (1985). Ion transport mechanisms in
cultured bovine corneal endothelia cells. Curr. Eye Res., 4(4);361369.
Joseph, D., and Miller, S. (1991). Apical and basolateral membrane ion
transport mechanisms in bovine retinal pigment epithelium. J. Physiol.,
435:439463.
Kajikawa, T., Mishima, H. K., Murakami, T., and Takano, M. (1999). Role of Pglycoprotein in distribution of rhodamine 123 into aqueous humor in rabbits.
Curr. Eye Res., 18(3):240246.
Kawazu, K., Yamada, K., Nakamur, M., and Ota, A. (1999). Characterization of
cyclosporin A transport in cultured rabbit corneal epithelial cells: P-glycoprotein transport activity and binding to cyclophilin. Invest. Ophthalmol. Vis. Sci.,
40(8):17381744.
Kenyon, E., Yu, K., La Cour, M., and Miller, S. S. (1994). Lactate transport
mechanisms at apical and basolateral membranes of bovine retinal pigment
epithelium. Am. J. Physiol., 267(6Pt1):C15611573.
Kern, H. L., and Ho, C. K. (1973). Localization and specicity of the transport
system for sugars in the calf lens. Ex. Eye Res., 15(6):751765.
Kern, H. L., and Zolot, S. L. (1987). Transport of vitamin C in the lens. Curr. Eye
Res., 6(7):885896.
Kern, H. L., Ho, C. K., and Ostrove, S. A. (1977). Comparison of transport at
the anterior and posterior surfaces of the calf lens. Exp. Eye Res., 24(6):
559570.
Khatami, M. (1987). Na+-linked active transport of ascorbate into cultured bovine
retinal pigment epithelial cells: heterologous inhibition by glucose. Membr.
Biochem., 7:115130.
Khatami, M., Stramm, L. E., and Rockey, J. H. (1986). Ascorbate transport in
cultured cat retinal pigment epithelial cells. Exp. Eye Res., 43(4):607615.
Klyce, S. D. (1972). Electrical proles in the corneal epithelium. J. Physiol.,
226(2):407429.
Klyce, S. D., and Crosson, C. E. (1985). Transport processes across the rabbit
corneal epithelium: a review. Curr. Eye Res., 4(4);323331.
Kompella, U. B., Kim, K. J., and Lee, V. H. L. (1993). Active chloride transport in
the pigmented rabbit conjunctiva. Curr. Eye Res., 12:10411048.
Kompella, U. B., Kim, K. J., Shiue, M. H., and Lee, V. H. (1995). Possible existence
of Na)+)-coupled amino acid transport in the pigmented rabbit conjunctiva.
Life Sci., 57(15):14271431.
Kompella, U. B., and Lee, V. H. L. (1999). Barriers to drug transport in ocular
epithelia, In: Transport Processes in Pharmaceutical Systems. Amidon, G. L.,
Lee, P. I., and Topp, E. M. (Eds.). Marcel Dekker, Inc., New York, pp. 317
375.
Kool, M., de Haas, M., Scheer, G. L., Scheper, R. J., van Eijk, M. J., Jujin, J. A.,
Baas, F., and Borst, P. (1997). Analysis of expression of cMOAT (MRP2),

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

53

MRP3, MRP4, and MRP5, homologues of the multidrug resistance-associated

protein gene (MRP1), in human cancer cell lines. Cancer Res., 57(16):3537
3547.
Kool, M., van der Linden, M., de Haas, M., Scheer, G. L., de Vree, J. M., Smith,
A. J., Jansen, G., Peters, G. J., Ponne, N., Scheper, R. J., Elferink, R. P., Baas,
F., and Borst, P. (1999a). MRP3, an organic anion transporter able to transport anti-cancer drugs. Proc. Natl. Acad. Sci. USA, 96(12);69146919.
Kool, M., van der Linden, M., de Haas, M., Baas, F., and Borst, P. (1999b).
Expression of human MRP6, a homologue of the multidrug resistance protein
gene MRP1, in tissues and cancer cells. Cancer Res., 59(1):175182.
la Cour, M. (1992). Cl transport in frog retinal pigment epithelium. Exp. Eye Res.,
54(6):921-931.
la Cour, M., Lin, H., Kenyon, E., and Miller, S. S. (1994). Lactate transport in
freshly isolated human fetal retinal pigment and epithelium. Invest.
Ophthalmol. Vis. Sci., 35:434442.
Lapierre, L. A. (2000). The molecular structure of the tight junctions. Adv. Drug
Deliv. Rev., 41(3):255264.
Lesar, S. T., and Fiscella, R. G. (1985). Antimicrobial drug delivery to the eye. Drug
Intel. Clin. Pharm., 19:642653.
Lightman, S., Rechthand, E., Latker, C., Plaestine, A., and Rapoport, S. (1987).
Assessment of the permeability of the blood-retinal barrier in hypertensive
rats. Hypertension, 10(4):390395.
Lo, W. K., and Harding, C. V. (1986). Structure and distributionof gap junctions in
the lens epithelium and ber cells. Cell Tissue Res., 244(2):253263.
Lum, B. L., and Gosland, M. P. (1995). MDR expression in normal tissues.
Pharmacologic implications for the clinical use of P-glycoprotein inhibitors.
Hematol. Oncol. Clin. North Am., 9(2);319326.
Madara, J. L., and Trier, J. S. (1982). Structure and permeability of goblet cell tight
junctions in rat small intestine. J. Membr. Biol., 66(2):145147.
Mantych, G. J., Hageman, G. S., and Devaskar, S. U. (1993). Characterization of
glucose transporter isoforms in the adult and developing human eye.
Endocrinology, 133(2):600607.
Marshall, W. S., and Klyce, S. D. (1983). Cellular and paracellular pathway resistances in the tight Cl -secreting epithelium of rabbit cornea. J. Memb. Biol.,
73:275282.
Martin-Vasallo, P., Ghosh, S., and Coca-Prados, M. (1989). Expression of Na+/
K+-ATPase alpha subunit isoforms in the human ciliary body and cultured
ciliary epithelial cells. J. Cell. Physiol., 141(2):243252.
Maurice, D. M., and Mishima, S. (1984). Ocular pharmacokinetics. In:
Pharmacology of the Eye, Sears, M. L. (Ed.). Springer-Verlag, New York.
Mayerson, P. L., and Hall, M. O. (1986). Rat retinal pigment epithelial cells show
specicity of phagocytosis in vitro. J. Cell Biol., 103:299308.
Mays, R. W., Beck, K. A., and Nelson, J. W. (1994). Organizaton and function of
the cytoskeleton in polarized epithelial cells: a component of the protein sorting machinery. Curr. Opin. Cell Biol., 6:1624.

Copyright 2003 Marcel Dekker, Inc.

54

Sunkara and Kompella

McLaughlin, B. J., Caldwell, R. B., Sasaki, Y., and Wood, T. O. (1985). Freezefracture quantitative comparison of rabbit corneal epithelial and endothelia
membranes. Curr. Eye Res., 4(9):951961.
Merriman-Smith, R., Donaldson, P., and Kistler, J. (1999). Dierential expression of
facilitative glucose transporters GLUT1 and GLUT3 in the lens. Invest.
Ophthalmol. Vis. Sci., 40(13):32243230.
Miller, S., and Edelman, J. (1990). Active ion transport pathways in the bovine
retinal pigment epithelium. J. Physiol.,4 24:283300.
Miller, S. S., Rabin, J., Strong, T., Iannuzzi, M., Adams, A. J., Collins, F., Reenstra,
W., and McCray, P., Jr. (1992). Cystic brosis (CF) gene product is expressed
in retina and retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci., 33:1009
(abstr).
Miyamoto, K., Khosrof, S., Bursell, S. E., et al. (1999). Prevention of leukostatis and
vascular leakage in streptozotocin-induced diabetic retinopathy via intercellular adhesion molecule-1 inhibition. Proc. Natl. Acad. Sci. USA, 96:10836
10841.
Miyamoto, Y., and Del Monte, M. A. (1994). Na+-dependent glutamate transporter
in human retinal pigment epithelial cells. Invest. Ophthalmol. Vis. Sci.,
35:35893598.
Miyamoto, Y., Kulanthaivel, P., Leibach, F. H., and Ganapathy, V. (1991). Taurine
uptake in apical membrane vesicles from the bovine retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci., 32:25422551.
Muller, M., Meijer, C., Zaman, G. J., Borst, P., Scheper, R. J., Mulder, N. H., de
Vries, E. G., and Jansen, P. L. (1994). Overexpression of the gene encoding the
multidrug resistance-associated protein results in increased ATP-dependent
glutathione S-conjugate transport. Proc. Natl. Acad. Sci. USA, 91(26):13033
13037.
Nord, E. P., Wright, S. H., Kippen, I., and Wright, E. M. (1983). Specicity of the
Na+-dependent monocarboxylic acid transport pathway in rabbit renal brush
border membranes. J. Membr. Biol., 72(3):213221.
Okami, T., Yamamoto, A., Omori, K., Akayama, M., Uyama, M., and Tashiro, Y.
(1989). Quantitative immunocytochemical localization of Na+,K+=ATPase
in rat ciliary epithelial cells. J. Histochem. Cytochem., 37:13531361.
Palacin, C., Tarrago, C., and Ortiz, J. A. (1998). Molecular biology of mammalian
plasma membrane amino acid transporters. Physiol. Rev., 78(4):9691054.
Paterson, C. A., and Delamere, N. A. (1992). The ciliary epithelia and aqueous
humor. In: Adlers Physiology of the Eye. Hart, W. M. Jr. (Ed.). MosbyYear-Book, Inc., St. Louis, pp. 412441.
Pepose J. S., and Ubels, J. L. (1992). The cornea. In: Adlers Physiology of the Eye.
Hart, W. M., Jr. (Ed.). Mosby-Year Book, Inc., St. Louis, pp. 2970.
Peyman, G. A., and Bok, D. (1972). Peroxidase diusion in the normal and lasercoagulated primate retina. Invest. Ophthalmol., 11(1):3545.
Philip, N. J., Yoon, H., and Grollman, E. F. (1998). Monocarboxylate transporter
MCT1 is located in the apical membrane and MCT3 in the basal membrane of
rat RPE. Am. J. Physiol., 274 (6 Pt 2):R1824R1828.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

55

Prausnitz, M. R., and Noonan, J. S. (1998). Permeability of cornea, sclera, and

conjunctiva: A literature analysis for drug delivery to the eye. J. Pharm.
Sci., 87(12):14791488.
Quinn, R. H., and Miller, S. S. (1992). Ion transport mechanisms in native human
retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci., 33:35133527.
Rae J. L. (1994). Outwardly rectifying potassium currents in lens epithelial cell
membranes. Curr. Eye Res., 13(9):679686.
Raviola, G. (1977). The structural basis of the blood-ocular barriers. Exp. Eye Res.,
25:2763.
Redzic, Z. B., Markovic, I. D., Jovanovic, S. S., Zlokovic, B. V., and Rakic, L. M.
(1995). Penetration of [3 H]tiazofurin into guinea pig eye by a saturable
mechanism. Eur. J. Ophthalmol., 5(2):131135.
Redzic, Z. B., Markovic, I. D., Vidovic, V. P., Vranic, V. P., Gasic, J. M., Duricic, B.
M., Pokrajac, M., Dordevic, J. B., Segal, M. B., and Rakic, L. M. (1998).
Endogenous nucleosides in the guinea-pig eye: analysis of transport and metabolites. Exp. Eye. Res., 66:315325.
Rezai, K. A., Lappas, A., Farrokh-Siar, L., Kohen, L., Widemann, P., and
Heimann, K. (1997). Iris pigment epithelial cells of long evans rats demonstrate phagocytic activity. Exp. Eye Res., 65:2329.
Riley, M. V. (1977). A study of the transfer of amino acids across the endothelium of
the rabbit cornea. Exp. Eye Res., 24(1):3544.
Riley, M. V., and Yates, E. M. (1977). Glutathione in the epithelium and endothelium of bovine and rabbit cornea. Exp. Eye Res., 25(4):385389.
Riley, M. V., Campbell, D., and Linz, D. H. (1973). Entry of amino acids into the
rabbit cornea. Exp. Eye Res., 15(6):677681.
Robinson, K. R., and Patterson, j. W. (1982). Localizastion of steady currents in the
lens. Curr. Eye Res., 2(12):843847.
Rodriguez-Boulan, E., and Nelson, W. J. (1989). Morphogenesis of the polarized
epithelial cell phenotype. Science, 245:718725.
Rodriguez-Peralta, L. (1975). The blood-aqueous barrier in ve species. Am. J.
Ophthalmol., 80(4):713725.
Saha, P., Kim, J., and Lee, V. H. (1996). A primary culture model of rabbit conjunctival epithelial cells exhibiting tight barrier properties. Curr. Eye Res.,
15(12):11631169.
Saha, P., Yang, J. J., and Lee, V. H. (1998). Existence of a P-glycoprotein drug eux
pump in cultured rabbit conjunctival epithelial cells. Invest. Ophthalmol. Vis.
Sci., 39(7):12211226.
Saier, M. H. (200a). A functional phylogenetic classication system for transmembrane solute transporters. Microbiol. Mol. Biol. Rev., 64(2):354411.
Saier, M. H., Jr. (2000b). Families of transmembrane transporters elective for amino
acids and their derivatives. Microbiology, 146:17751795.
Schlingemann, R. O., Hofman, P., Klooster, J., Blaauwgeers, H. G., Van der Gaag,
R., and Vrensen, G. F. (1998). Ciliary muscle capillaries have blood-tissue
barrier characteristics. Exp. Eye Res, 66(6):747754.

Copyright 2003 Marcel Dekker, Inc.

56

Sunkara and Kompella

Sharma, R. C., Assaraf, Y. G., and Schimke, R. T. (1991). A phenotype conferring

selective resistance to lipophilic antifolates in Chinese hamster ovary cells.
Cancer Res., 51(11);29492959.
Sherman, S. H., Green, K., and Laties, A. M. (1978). The fate of anterior chamber
uorescein in the monkey eye. 1. The anterior chamber outow pathways.
Exp. Eye Res., 27(2);159173.
Shirao, Y., and Steinberg, R. (1987). MEchanisms of eects of small hyper-osmotic
gradients on the chick RPE. Invest. Ophthalmol. Vis. Sci., 280:20152025.
Shiue, M. H., Kulkarni, A. A., Gukasyan, H. J., Swisher, J. B., Kim, K. J., and Lee,
V. H. (2000). Pharmacological modulation of uid secretion in the pigmented
rabbit conjunctiva. Life Sci., 66(7):PL105111.
Simon, A. M., and Goodenough, D. A. (1998). Diverse functions of vertebrate gap
junctions. Trends Cell Biol., 8(12);477483.
Smith, R. S., and Rudt, L. A. (1975). Ocular vascular and epithelial barriers to
microperoxidase. Invest. Ophthalomol., 14(7):556560.
Socci, R. R., and Delamere, N. A. (1988). Characteristics of ascorbate transport in
the rabbit iris-ciliary body. Exp. Eye Res., 46(6):853861.
Spray, D. C., and Bennett, M. V. (1985). Physiology and pharmacology of gap
junctions. Ann. Rev. Physiol., 47:281303.
Stewart, P. A., and Tuor, U. I. (1994). Blood-eye barriers in the rat: correlation of
ultrastructure with function. J. Comp. Neurol., 340(4):566576.
Strauss, O., Widerholt, M., Wienrich, M. (1996). Activation of Cl currents in
cultured rat retinal pigment epithelial cells by intracellular applications of
ionositol-1,4,5,-triphosphate: dierences between rats with retinal dystrophy
(RCS) and normal rats. J. Membr. Biol., 151:189200.
Sun, X. C., Bonanno, J. A., Jelamskii, S., and Xie, Q. (2000). Expression and localization of Na+-HCO
3 cotransporter in bovine corneal endothelium. Am. J.
Physiol. Cell Physiol., 279(5):C1648C1655.
Takahashi, H., Kaminski, A. E., and Zieske, J. D. (1996). Glucose transporter 1
expression is enhanced during corneal epithelial wound repair. Exp. Eye Res.,
63(6):649659.
Takata, K., Hirano, H., and Kasahara, M. (1997). Transport of glucose across the
blood-tissue barriers. Int. Rev. Cytol., 172:153.
Theibaut, F., Tsuruo, T., Hamada, H., Gottesman, M. M., Pastan, I., and
Willingham, M. C. (1989). Immunohistochemical localization in normal tissues of dierent epitopes in the multidrug transport protein P170: evidence for
localization in brain capillaries and crossreactivity of one antibody with a
muscle protein. J. Histochem. Cytochem., 37(2):159164.
Thoft, R. A., and Friend, J. (1972). Corneal amino acid supply and distribution.
Invest. Ophthalmol. Vis. Sci., 11(9);723727.
Thorens, B. (1996). Glucose transporters in the regulation of intestinal, renal, and
liver glucose uxes. Am. J. Physiol., 270(4 Pt 1): G541G553.
Tornquist, P., and Alm, A. (1986). Carrier-mediated transport of amino acids
through the blood-retinal and the blood-brain barriers. Grafes. Arc. Clin.
Exp. Ophthalmol., 224(1):2125.

Copyright 2003 Marcel Dekker, Inc.

Membrane Transport Processes in the Eye

57

Tornquist, P., Alm, A., and Bill, A. (1990). Permeability of ocular vessels and transport across the blood-retinal-barrier. Eye, 4 (Pt 2):303309.
Tsuboi, S., and Pederson, J. E. (1988). Eect of plasma osmolality and intraocular
pressure on uid movement across the blood-retinal barrier. Invest.
Ophthalmol. Vis. Sci., 29(11):17471749.
Tusji, A., Terasaki, T., Takabatake, Y., Tenda, Y., Tamai, I., Yamashima, T.,
Moritani, S., Tsuruo, T., and Yamashita, J. (1992). P-glycoprotein as the
drug eux pump in primary cultured bovine brain capillary endothelial
cells. Life Sci., 51(18);14271437.
Ueda, Y., and Steinberg, R. (1994). Chloride currents in freshly isolated rat retinal
pigment epithelial cells. Exp. Eye Res., 58:31342.
Ueda, H., Horibe, Y., Kim, K. J., and Lee, V. H. (2000). Functional characterization
of organic cation drug transport in the pigment rabbit conjunctiva. Invest.
Ophthalmol. Vis. Sci., 41(3):870876.
Usukura, J., Fain, G. L., and Bok, D. (1988). [3 H]Ouabain localization of Na+-K+
ATPase in the epithelium of rabbit ciliary body pars plicata. Invest.
Ophthalmol. Vis. Sci., 29(4):606614.
Vinores, S. A. (1995). Assessment of blood-retinal barrier integrity. Histol.
Histopathol., 10(1):141154.
Wang, L., Chen, L., Walker, V., and Jacob, T. J. (1998). Antisense to MDR1 mRNA
reduces P-glycoprotein expression, swelling-activated CL current and volume
regulation in bovine ciliary epithelial cells. J. Physiol., 511(Pt 1):3344.
Watsky, M. A., Jablonski, M. M., and Edelhauser, H. F. (1988). Comparison of
conjunctival and corneal surface areas in rabbit and human. Curr. Eye Res.,
7:483486.
Wen, R., Lui, G. M., and Steinberg, R. H. (1994). Expression of a tetrodoxinsensitive Na+ current in cultured human retinal pigment epithelial cells. J.
Physiol (Lond)., 475:187196.
Whikehart, D. R., and Soppet, D. R. (1981). Activities of transport enzymes located
in the plasma membranes of corneal endothelial cells. Invest. Ophthalmol. Vis.
Sci., 21(6):819825.
Whikehart, D. R., Montgomery, B., and Hafer, L. M. (1987). Sodium and potassium
saturation kinetics of Na+-K+ ATPase in plasma membranes from corneal
endothelium: fresh tissues vs. tissue culture. Curr. Eye Res., 6(5):709717.
Widerholt, M., Jentsch, T. J., and Keller, S. K. (1985). Electrogenic sodium-bicarbonate symport in cultured corneal endothelial cells. Pugers Arch., 405(suppl
1):S167171.
Wolosin, J. M., Alvarez, L. J., and candia, O. A. (1988). Cellular pH and Na+-H+
exchange activity in the lens epithelium of Bufo marinus toad. Am. J. Physiol.,
255(5 Pt 1):C595C602.
Wolosin, J. M., Alvarez, L. J., and Candia, O. A. (1990). HCO
3 transport in the
toad lens epithelium is mediated by an electronegative Na(+)-dependent symport. Am. J. Physiol., 258(5 Pt 1):C855C861.
Wu, G. (1995). Anatomy and physiology of the retina/ocular embryology. In:
Retina: The fundamentals. W. B. Saunders Company, Philadelphia.

Copyright 2003 Marcel Dekker, Inc.

58

Sunkara and Kompella

Wu, J., Zhang, J. J., Koppel, H., and Jacob, T. J. (1996). P-glycoprotein regulates a
volume-activated chloride current in bovine non-pigmented ciliary epithelial
cells. J. Physiol., 491(Pt 3):743755.
Xu, Q., Quam, T., and Adamis, A. P. (2001). Sensitive blood-retinal barrier breakdown quantitation using evans blue. Invest. Ophthalmol. Vis. Sci., 42(3):789
794.
Yang, J. J., and Lee, V. H. (2000). Role of multidrug resistance protein (MRP) in
drug transport in rabbit conjunctival epithelial cells. PharmSci., 2(4):S3168.
Yang, J. J., Kim, K. J., and Lee, V. H. (2000). Role of P-glycoprotein in restricting
propranolol transport in cultured rabbit conjunctival epithelial cell layers.
Pharm. Res., 17(5):533538.
Ye, J. J., and Zadunaisky, J. A. (1992a). Ca2+/Na+ exchanger and Na+, K+, 2Cl
cotransporter in lens ber plasma membrane vesicles. Exp. Eye Res.,
55(6):797804.
Ye, J. J., and Zadunaisky, J. A. (1992b). Study of the Ca2+/Na+ exchange mechanism in vesicles isolated from apical membranes of lens epithelium of spiny
dogsh (Squalas acanthias) and bovine eye. Exp. Eye Res., 55(2):243250.
Yoon, H., Fanelli, A., Grollman, E. F., and Philip, N. J. (1997). Identication of a
unique monocarboxylate transporter (MCT3) in retinal pigment epithelium.
Biochem. Biophys. Res. Commun., 234(1):9094.
Yu, A. S. (2000). Paracellular solute transport more than just a leak? Curr. Opin.
Nephrol. Hypertens., 9(5):513515.
Zaunaisky, J. A. (1992). Membrane transport in ocular epithelia. In: Barriers and
Fluids of the Eye and Brain. Segal, M. B. (Ed.). CRC Press, Inc., Ann Arbor.
Zelenka, P. S., and Vu, N. D. (1984). Correlation between phosphatidylinositol
degradation and cell division in embryonic chicken lens epithelia. Dev. Biol.,
105(2):325329.
Zimmer, A., Kreuter, J., and Robinson, J. R. (1991). Studies on the transport pathway of PBCA nanoparticles in ocular tissues. J. Microencapsulation. 8(4):497
504.
Zlokovic, B. V., Mackic, J. B., McComb, J. G., Kannan, R., and Weiss, M. H.
(1992). An in-situ perfused guinea pig eye model for blood-ocular transport
studies: application to amino acids. Exp. Eye Res., 54:471477.
Zlokovic, B. V., Mackic, J. B., McComb, J. G., Kaplowitz, N., Weiss, M. H.,and
Kannan, R. (1994). Blood-to-lens transport of reduced glutathione in an in
situ perfused guinea pig eye. Exp. Eye Res., 59:487496.

Copyright 2003 Marcel Dekker, Inc.

3
General Considerations in Ocular
Drug Delivery
James E. Chastain
Alcon Research, Ltd., Fort Worth, Texas, U.S.A.

I.

INTRODUCTION

The use of pharmacokinetic principles has become widespread in the

pharmaceutical industry, primarily because of their utility in relating
ecacy and toxicity to drug concentrations in plasma or some other
appropriate body compartment. In general, pharmacokinetics is the process of absorption, distribution, and excretion of a drug. Excretion is
usually coupled with metabolism, which typically converts drug to a
more water-soluble form, more amenable to excretion. Most of the
pharmacokinetic literature deals with systemically administered drugs
that reach their pharmacological target by way of the blood following
oral or parenteral administration.
There are various approaches to studying the pharmacokinetics of a
drug. Classical pharmacokinetics empirically derives one or more exponentials to mathematically describe concentration versus time data.
Physiological pharmacokinetics associates compartments with specic anatomical tissues or organs and usually includes blood ow and drug clearance
in individual organs/tissues as part of the model. Noncompartmental pharmacokinetics, as its name implies, makes no assumptions regarding compartments but usually employs statistical moment theory to derive basic
parameters such as volume of distribution and clearance. All of these methods can prove useful in describing a drugs pharmacokinetic behavior, an
essential step toward determining an appropriate dosing regimen for a drug
relative to its ecacy and toxicity proles.
59

Copyright 2003 Marcel Dekker, Inc.

60

Chastain

Ocular pharmacokinetics includes the features of absorption, distribution, and excretion found with systemic administration but applied to the
eye. However, owing to the unique anatomy and physiology of the eye and
surrounding tissue, ocular pharmacokinetics is considerably more dicult
to describe and predict than its systemic counterpart. The task is further
complicated by the various formulations, routes, and dosing regimens typically encountered in ophthalmology.
Pharmacodynamics is the measurement of pharmacological response
relative to dose or concentration. The pharmacological response induced by
a drug can vary greatly from individual to individual due to dierences in
factors such as eye pigmentation, the pathological state of the eye, tearing,
or blink rate. The application of pharmacological endpoints is particularly
useful in the study of drugs in the human eye, where the ability to determine
the ocular pharmacokinetics based on ocular tissue concentrations is
severely limited.
This chapter discusses the ocular pharmacokinetics associated with
topical ocular, intravitreal, periocular, and systemic administration. In addition, the pharmacodynamics related to ophthalmic drugs and the role of
ocular drug metabolism are reviewed.

II.

OCULAR PHARMACOKINETICS

Application of classical pharmacokinetics to ophthalmic drugs is problematic because of the complexities associated with eye anatomy and physiology. As a result, most of the literature is limited to measuring
concentrations in ocular tissues over time following single or multiple
administration. This approach, while informative, does not easily yield
quantitative predictions for changes in formulation or dosage regimen.
Compounding the problem is the fact that most studies have been
conducted in rabbit eyes, which dier signicantly from human eyes in
anatomy and physiology (see Table 1). the most obvious dierences are in
blink rate and the presence or absence of a nictitating membrane. An
overall, detailed discussion of these factors and ocular pharmacokinetics
as a whole has been presented elsewhere (19).
A.

Topical Ocular Administration

1. Absorption
The general process of absorption into the eye from the precorneal area
(dose site) following topical ocular administration is quite complex. The

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

61

Table 1 Comparison of Anatomical and Physiological

Factors in the Rabbit and Human Eye
Factor

Rabbit

Human

Tear volume (mL)

Tear turnover rate (mL/min)
Spontaneous blinking rate
Nictitating membrane
Lacrimal punctum/puncta
pH of lacrimal uids
Milliosmolarity of tears
Corneal thickness (mm)
Corneal diameter (mm)
Corneal surface area (cm2 )
Ratio of conjunctival surface
and corneal surface
Aqueous humor volume (mL)
Aqueous humor turnover
rate (mL/min)

510
0.60.8
45 times/h
Present
1
7.37.7
305
0.40
15
1.52.0
9

730
0.52.2
615 times/min
Absent
2
7.37.7
305
0.52
12
1.04
17

0.250.3
34.7

0.10.25
23

Source: Adapted from Refs. 8, 9.

classical sequence of events involves drug instillation, dilution in tear uid,

diusion through mucin layer, corneal penetration (epithelium, stroma,
endothelium), and transfer from cornea to aqueous humor. Following
absorption, drug distributes to the site of action (e.g., iris-ciliary body).
Parallel absorption via the conjunctiva/sclera provides an additional
pathway to eye tissues but, for most drugs, is minor compared with corneal
absorption. Also, nonproductive, competing, and parallel pathways
(e.g., nasolacrimal drainage or systemic absorption via the conjunctiva)
work to carry drug away from the eye and limit the time allowed for the
absorption process. Moreover, in some species, such as the rabbit, nonproductive absorption into the nictitating membrane can occur. Figure 1
presents a summary of these precorneal events, along with a relatively
simplied view of the kinetics in the cornea, aqueous humor, and anterior
segment.
a. Corneal Penetration. Drug absorption through the cornea into
the eye is dependent to a large degree upon a drugs physicochemical
properties, such as octanol-water partition coefcient, molecular weight,
solubility, and ionization state. In addition, corneal penetration is layer
(corneal epithelium, stroma, and endothelium) selective. Schoenwald
and Ward demonstrated a parabolic relationship between log corneal

Copyright 2003 Marcel Dekker, Inc.

62

Chastain

Figure 1 Model showing precorneal and intraocular events following topical

ocular administration of a drug. (Adapted from Ref. 2.)

permeability coefcients and log octanol-water coefcients of various

steroids (10) (see Fig. 2). The optimum log octanol-water coefcient was
2.9. Schoenwald and Huang showed a correlation between octanol-water
partitioning of beta-blocking agents and their corneal permeabilities using
excised rabbit corneas (11). Over a fourfold logarithmic range, the best t
was also a parabolic curve. In a renement of this parabolic relationship,
Huang et al. demonstrated in vitro a sigmoidal relationship between
permeabiilty coefcient and distribution coefcient (12) (see Fig. 3). In
this study, the endothelium offered little resistance and the stroma posed
even less. Lipophilic drugs penetrated the cornea more rapidly; however,
the hydrophilic stroma was rate limiting for these compounds. Maren et
al. studied 11 sulfonamide carbonic anhydrase inhibitors (CAIs) of varied
physicochemical characteristics with respect to transcorneal permeability
and reduction of intraocular ow (13). In isolated rabbit cornea with a
constantly applied drug concentration, the rst-order rate constants
ranged from 0.140  103 h1 , nearly proportional to lipid solubility,
with water-insoluble drugs tending to have higher rate constants.
For most drugs, the multicell layered corneal epithelium presents the
greatest barrier to penetration, primarily due to its cellular membranes.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

63

Figure 2 Log-log plot of corneal permeability coecient (pH 7.65) versus distribution coecient (pH 7.65). Observed data (*) and predicted curve () are presented.
(Replotted from Ref. 10.)

Stroma and particularly endothelium oer little resistance, except for highly
lipophilic drugs. In fact, these two layers are often lumped together, along
with aqueous humor, as a single compartment for purposes of pharmacokinetic modeling. The inuence of the epithelium is most clearly demonstrated by studying corneal penetration following removal of the epithelium.
Cox et al. showed in rabbits that when the epithelium was intact, no
14
C-dexamethasone was detected in cornea or aqueous humor following
topical ocular administration (14,15), while detectable levels were observed
after removal of the corneal epithelium. Chien et al. studied the relationship between corneal epithelial integrity and prodrug lipophilicity with
corneal penetration of a homologous series of timolol prodrugs (16).
Deepithelization of the corneal in vitro did not aect corneal permeability
of O-acetyl, propionyl, or butryl timolol but reduced penetration of the
other prodrugs. In contrast, deepithelization in vivo only reduced timolol
aqueous humor concentrations derived from O-propionyl and octanoyl
esters. Therefore, factors other than the corneal epithelium may play a
role in penetration.

Copyright 2003 Marcel Dekker, Inc.

64

Chastain

Figure 3 Log-log plot of corneal permeability coecient versus distribution coecient (octanol-Sorensens buer, pH 7.65). Intact corneal data (*) and computergenerated, model-derived curve () are presented. (Replotted from Ref. 12.)

Corneal penetration is also aected by the composition of the drug

formulation. Whether a formulation is a solution, suspension, contains a
buer, viscosilating agent, or penetration enhancer, can inuence absorption. Burstein and Anderson have reviewed corneal penetration and ocular
bioavailability of drugs relative to optimizing formulations for typical ocular use (1). They evaluated the eects of preservatives, vehicles, adjunct
agents, and anatomy and developed model systems for selecting the best
formulations for preclinical and clinical use. Vehicle pH was one important
factor considered. Adjusting the pH so that the drug was mostly in the
unionized form greatly enhanced corneal penetration. Furthermore, it was
concluded that buering capacity, which keeps drug mainly in the ionized
form, should be minimized to allow for adequate neutralization by tear lm.
Other formulation components have been examined for their eect on
absorption. Madhu et al. studied the inuence of benzalkonium chloride
(BAC)/EDTA on ocular bioavailability of ketorolac tromethamine follow-

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

65

ing topical ocular instillation onto normal and deepithelialized rabbit corneas in vitro and in vivo (17). BAC/EDTA caused a statistically signicant
increase in the ocular bioavailability of ketorolac through deepithelialized
cornea but not intact cornea in vitro and in vivo. Jani et al. demonstrated
that inclusion of ion exchange resins in an ophthalmic formulation of betaxolol increased the ocular bioavailabilty of betaxolol twofold (18).
Hyaluronic acid, which can adhere to the corneal surface, is also capable
of prolonging precorneal residence time (19).
b. Noncorneal, Ocular (Productive) Absorption. In addition to the
classical corneal pathway, there is a competing and parallel route of absorption via the conjunctiva and sclera, the so-called conjunctival/scleral
pathway. For most drugs this is a minor absorption pathway compared
to the corneal route, but for a few compounds its contribution is signicant. Ahmed and Patton investigated corneal versus noncorneal penetration of topically applied drugs in the eye (20,21). They demonstrated that
noncorneal absorption can contribute signicantly to intraocular penetration. A productive noncorneal route involving penetration through the
conjunctiva and underlying sclera was described. Drug can therefore bypass the anterior chamber and distribute directly to the uveal tract and
vitreous. This route was shown to be particularly important for drugs
with low corneal permeability, such as inulin. In a separate study, Ahmed
et al. evaluated in vitro the barrier properties of the conjunctiva, sclera,
and cornea (22). Diffusion characteristics of various drugs were studied.
Scleral permeability was signicantly higher than that in cornea, and permeability coefcients of the -blockers ranked as follows: propranolol >
penbutolol > timolol > nadolol for cornea, and penbutolol > propranolol > timolol > nadolol for the sclera. Resistance was higher in cornea
versus conjunctiva for inulin but similar in the case of timolol. Chien et
al. studied the ocular penetration pathways of three
2 -adrenergic agents
in rabbits both in vitro and in vivo (23). The predominant pathway for
absorption was the corneal route, with the exception of p-aminoclonidine,
the least lipophilic, which utilized the conjunctival/scleral pathway. The
results suggest that the pathway of absorption may be inuenced in part
by lipophilicity and that hydrophilic compounds may prefer the conjunctival/scleral route.
Some investigators have employed a dosing cylinder axed to the
cornea to study corneal and noncorneal absorption. Drug is applied within
the cylinder for corneal dosing and outside the cylinder for noncorneal
(conjunctival/scleral) dosing. In a study by Schoenwald et al., the conjunctival/scleral pathway yielded higher iris-ciliary body concentrations for all
compounds evaluated with the exception of lipophilic rhodamine B (24).

Copyright 2003 Marcel Dekker, Inc.

66

Chastain

Romanelli et al. demonstrated the absorption of topical ocular bendazac

into the retina/choroid via the conjunctival/scleral pathway (25).
Absorption by way of this extracorneal route was inuenced by physicochemical features and not by vehicle, while the transcorneal route was
aected by vehicle.
c. Noncorneal, Nonproductive Absorption and Precorneal Drainage.
Routes that lead to the removal of drug from the precorneal area and do
not result in direct ocular uptake are referred to as nonproductive absorption pathways. These noncorneal pathways, which are in parallel with corneal absorption, include conjunctival uptake and drainage via the
nasolacrimal duct. Both lead to systemic absorption by way of conjunctival
blood vessels in the former case or via the nasal mucosa and gastrointestinal tract in the latter case. As discussed previously, the conjunctiva can absorb drug and, via the sclera, deliver drug to the eye; however, blood
vessels within the conjunctiva can also lead to systemic absorption.
Nonproductive, noncorneal absorption and drainage loss greatly
impact the precorneal residence time and the time for ocular absorption.
Drainage, in particular, is very rapid and generally limits ocular contact at
the site of absorption to about 310 minutes (8). However, the lag timethe
time for drug to traverse the cornea and appear in the aqueous humoris
suciently long to extend time to maximal concentration in the aqueous to
between 20 and 60 minutes for most drugs (8). Due to rapid loss of drug
from the precorneal region, less than 10%, and more typically less than 1
2%, of a topical dose is absorbed into the eye. At the same time, systemic
absorption can be as high as 100%, indicating that most of the drug dose is
unavailable for ecacy. For example, Ling and Combs showed that the
ocular bioavailability of topical ocular ketorolac was 4% in anesthetized
rabbits, while systemic absorption was complete (26). Ocular tissue levels
were about 13-fold higher than those in plasma, and peak concentrations
were achieved by 1 hour in both aqueous humor and plasma postdose.
Tang-Liu et al. showed that topical ocular levobunolol was rapidly
absorbed, with an ocular bioavailability of 2.5% and systemic bioavailability of 46% (27). Patton and Robinson have investigated the contribution of
tear turnover, instilled solution drainage, and nonproductive absorption to
precorneal loss of drug (28). Instilled solution drainage was shown to be the
predominant factor in precorneal loss, while the inuence of tear turnover
was minor. It was concluded that noncorneal, nonproductive loss was
potentially signicant due to the large surface area of noncorneal tissue;
however, its role was minimal compared to drainage. Ocular (aqueous
humor) and systemic bioavailabilities for various drugs are presented in
Table 2.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

67

Table 2 Ocular and Systemic Bioavailabilities (Percent of Dose) Following

Topical Ocular Administration
Drug
Ketorolac
Levobunolol
Flurbiprofen
Imirestat
Propranolol
Pilocarpine nitrate

Clonidine
Cortisol
Tetrahydrocannabinol

Ocular
(%)

Systemic
(%)

Test subject

Ref.

3.7
2.5
710
ND
55, 5.6
0.62
2.12
0.80
1.87
1.6
ND
ND

 100
46
74
5075
ND
ND
ND
ND
ND
ND
3035
23

Anes. rabbit
Rabbit
Anes. rabbit
Anes. rabbit and dog
Rabbit, dog
Rabbit, duct open
Rabbit, duct plugged
Anes. rabbit, duct open
Anes. rabbit, duct plugged
Rabbit
Rabbit
Rabbit

26
27
46
120
51
28
28
28
28
67
2
2

ND, Not determined; Anes, anesthetized.

Volume instilled is one factor that can inuence drainage and noncorneal absorption. Chrai et al. evaluated the eect of instilled volume on
drainage loss using miosis data in albino rabbits (29). A radioisotopic
method was used to determine lacrimal volume and tear turnover.
Unanesthetized rabbits had lacrimal volume of 7.5 mL, whereas anesthetized
rabbits had a slightly larger volume of 12.0 mL. Lacrimal turnover was
slower in anesthetized rabbits, in which the rate was negligible. Instilled
volume drainage was found to be rst order, and drainage was dependent
on the volume administered in unanesthetized but not anesthetized rabbits.
In a separate study, using 99mTc (technetium), Chrai et al. demonstrated that
drug loss through drainage increased with increasing drop size (30a).
Concentration in the precorneal tears was higher, while drainage was
increased, following a larger drop volume. An instillation volume of 510
mL containing a larger concentration of drug was recommended. For comparison, most commercial ophthalmic droppers deliver 3070 mL. In addition, it was shown that 5-minute spacing between drops was optimal for
minimizing drainage loss. Also, for two drugs given as two separate drops,
the second drop will negatively inuence the rst, arguing for combination
therapy. Interestingly, Keister et al. showed that for drugs with high corneal
permeability, ocular bioavailability and corneal permeability are relatively
unaected by drug volume (30b). In contrast, for drugs with low corneal

Copyright 2003 Marcel Dekker, Inc.

68

Chastain

permeability, reducing to a small dose volume can increase ocular bioavailability up to fourfold.
It is known that increasing the formulation viscosity has the potential
to decrease drainage rate, thereby increasing precorneal residence time and
prolonging the time for ocular absorption. Zaki et al. studied the precorneal
drainage of radiolabeled polyvinyl alcohol or hydroxymethylcellulose formulations in the rabbit and humans by gamma scintiography (31).
Signicant retardation of drainage in humans was observed at higher polymer concentrations. Patton and Robinson also used polyvinyl alcohol,
along with methylcellulose, to evaluate the relationship between viscosity
and contact time or drainage loss (32). The optimum viscosity range was 12
15 centipoise in rabbits; however, the relationship was not direct, presumably due to shear forces acting on the formulation lm at the eye surface.
Chrai et al. also demonstrated, using methylcellulose vehicle, that increasing
viscosity of an ophthalmic solution results in decreased drainage (33). Over
the range of 115 centipoise, there was a threefold change in drainage rate
constant and another threefold change over the range of 15100 centipoise.
Dierences in the anatomical and physiological characteristics of the
eye, particularly between species, can also aect drainage and noncorneal
absorption. A comparison of attributes between rabbit and human eyes has
been presented in Table 1. In vivo evaluations in humans and rabbits by
Edelhauser and Maren demonstrated lower permeability of a series of sulfonamide CAIs in humans, possibly due to a greater blinking rate, a twofold
greater tear turnover, and a twofold lower corneal-conjunctival area (34).
Maurice has shown that corneal penetration is enhanced in the rabbit due to
low blink rate (35). This can increase area under the aqueous humor concentration-time curve threefold over that in humans. For many drugs,
epithelial permeability is suciently high to mitigate eects of blink rate,
although blinking may be critical to the proper ocular absorption and distribution of certain other drugs (36).
As one might expect, occlusion of the nasolacrimal duct substantially
reduces drainage and prolongs precorneal residence time. As a result, an
increase in ocular bioavailability and a decrease in systemic exposure typically occur. Kaila et al. studied the absorption kinetics of timolol following
topical ocular administration to healthy volunteer subjects with eyelid closure, nasolacrimal occlusion (NLO), or normal blinking (37). NLO reduced
total timolol systemic absorption, although, in some subjects, the initial
absorption was enhanced. In another example, Zimmerman et al. showed
that there were lower uorescein anterior chamber levels and a shorter
duration of uorescein in the absence of NLO or eyelid closure (38).
Systemic drug absorption in normal subjects was reduced more than 60%
with these techniques. Linden and Alm studied the eect of tear drainage on

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

69

intraocular penetration of topically applied uorescein in healthy human

eyes using uorophotometry (39). Upper and lower punctal plugs in one
eye caused a signicant (p < 0:025) increase in aqueous humor uorescein
concentrations 18 hours postdose of 20 mL of 2% solution of sodium
uorescein in the lower conjunctival sac. Compressing the tear sac and/or
closing the eyelids for 1 minute after application had no eect on corneal or
aqueous levels of uorescein. Lee et al. evaluated the eect of NLO on the
extent of systemic absorption following topical ocular administration of
various adrenergic drugs (40). Table 3 summarizes the results of this
study. Hydrophilic atenolol and lipophilic betaxolol, which were not
absorbed into the circulation as well as timolol and levobunolol, were not
aected in their systemic absorption by 5 minutes of NLO. However, systemic bioavailability decreased 80% by prolonging precorneal retention of
the dose to 480 minutes. It was concluded that modest formulation changes
will have little eect on systemic absorption for extremely hydrophilic drugs.
Drugs similar in lipophilicity to timolol will be well absorbed systemically,
while extremely hydrophilic drugs or extremely lipophilic drugs will be
absorbed to a lesser extent.
As alluded to earlier in this chapter, the rate and extent of systemic
absorption via the conjunctiva relative to corneal absorption is dependent
on the physicochemical properties of a drug or its formulation. Ahmed and
Patton showed that the conjunctival pathway is particularly important for
drugs with low corneal permeability and that noncorneal permeation is
limited by nonproductive loss to the systemic circulation (21). Hitoshe et
Table 3 Systemic Bioavailabilitiesa
(Percent of Dose) Following Topical
Ocular Administration with Various
Durations of Nasolacrimal Duct
Occlusion
Ocular
Drug
Atenololb
Timolol
Levobunolol
Betaxolol

0 min

5 min

120 min

41
106
82
66

39
88
54
56

12
36
56
58

Relative to subcutaneous route.

For atenolol, the duration of occlusion was
480 min.
Source: Ref. 40.

Copyright 2003 Marcel Dekker, Inc.

70

Chastain

al. demonstrated that drugs and prodrugs could be designed to selectively

reduce conjunctival absorption and thus suppress systemic exposure (41).
This can be accomplished by taking advantage of the apparently lower
lipophilicity of the conjunctiva versus that of the cornea. Ashton et al.
studied the inuence of pH, tonicity, BAC, and EDTA on conjunctival
and cornea penetration of four beta blockers: atenolol, timolol, levobunolol,
and betaxolol (42). Isolated pigmented rabbit conjunctiva and cornea were
used. The conjunctiva was more permeable than cornea, and formulation
changes had greater inuence on corneal versus conjunctival penetration.
This was particularly true for the hydrophilic compounds; therefore,
changes in formulation can eect both ocular and systemic absorption.
d. Absorption Kinetics. As discussed, lag time and drainage severely
curtail the absorption process in the eye. These events make it somewhat
difcult to estimate absorption kinetics. In particular, an anomaly arises
when attempting to derive the absorption rate constant, Ka , in that there
is a large discrepancy between the theoretical and actual times during
which absorption occurs. For example, phenylephrine has an absorption
half-life of 278 hours (43); therefore, the theoretical time to complete absorption would be an enormous 1200 hours (45 half-lives). In actuality,
the absorption process terminates within 310 minutes. Consistent with
this, Makoid and Robinson showed that extensive parallel absorption
pathways resulted in an apparent Ka one to two orders of magnitude larger than actual for 3H-pilocarpine (44). Aqueous ow accounted for most
of the drug elimination. The predominating effects on absorption and
elimination, independent of drug structure, suggested that similar pharmacokinetics may be found for a variety of drugs. Table 4 shows absorption
half-lives ranging from about 3 hours for pilocarpine to as high as 278
hours for phenylephrine. These values illustrate that corneal absorption is
relatively slow and that the theoretically derived times for absorption are
substantially longer than the actual 310 minutes typically encountered.
Rapid drainage of drug from the precorneal area largely determines
the time to peak concentration in the aqueous humor irrespective of a drugs
physicochemical properties. As a general rule, virtually all drugs will reach
peak concentration in the aqueous humor within 2060 minutes postinstillation (8). Makoid and Robinson have derived an equation for calculating
time to peak (tp ):
 
Kna
Ka
tp
Kna  Ka
ln

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

71

Table 4 Transcorneal First-Order Absorption Rate Constant (Ka ) and

Absorption Half-Life t1=2abs for Corneal Absorption
Drug

Ka (min1 )

t1=2abs (h)

Ref.

Pilocarpine
6-Hydroxyethoxy-2-benzothiazolesulfonamide
Ethoxzolamide
Aminozolamide
Clonidine
2-Benzothiazolesulfonamide
Ibuprofen
Ibufenac
Phenylephrine

4  103
4:2  103
1:5  103
1:4  103
1:4  103
1:3  103
9:64  104
6:03  104
4:15  105

2.88
5.37
7.7
8.25
8.25
8.88
12
18.3
278

64
48
48
111
67
48
121
121
43

Source: Adapted from Ref. 8.

where Kna and Ka are nonabsorptive loss rate constant and the transcorneal
absorption rate constant, respectively, and Kna  Ka (44). Since for most
drugs Kna is about twofold larger than Ka , this formula is widely applicable.
Huang et al. have shown for various -blockers in rabbits that aqueous
humor Tmax is inversely proportional to corneal permeability coecient
(45), and that Cmax and area under the concentration-time curve (AUC)
for aqueous humor generally correlate with permeability. Table 5 shows

Table 5 Time to Maximal Concentration

(Tmax in Aqueous Humor Following Topical
Ocular Administration
Drug
Falintolol
Lincomycin hydrochloride
Bufuralol
Acebutolol
Propranolol
L-662,583a
Timolol
Ketorolac
a

Tmax (min)

Ref.

3060
3045
10
85
30
120
10
25
60

122
123
45
45
124
125
126
45
26

5-(3-Dimethylaminomethyl-4-hydroxyphenylsulfonyl) thiophene-2-sulfonamide.

Copyright 2003 Marcel Dekker, Inc.

72

Chastain

times to maximal aqueous humor concentration (Tmax ) for various drugs.

Tmax falls within a range of 10 minutes to 2 hours.
As mentioned, the typical percentage of drug absorbed into the eye
following a topical ocular dose is in the range of 110%, while systemic
bioavailability can be as high as 100% (Table 2). For calculating bioavailability fraction (F), aqueous humor AUC can be determined by intracameral injection (27,46) or topical instillation with plugged drainage ducts (28).
In the former case, aqueous humor AUCs, normalized for dose, are compared between topical and intracameral doses to derive F. In the latter case,
Patton and Robinson (28) calculated F using the following equation:
k10  V  AUC
2
D
where D is the instilled dose, k10 is the loss of drug from the precorneal area,
and V is the estimated aqueous humor volume of 0.3 mL.
For some drugs, such as topical anti-infectives, tear bioavailability and
not aqueous humor bioavailability is the determining pharmacokinetic factor for ecacy. Unlike aqueous humor, tear lm is readily sampled in
humans. Depending on the method of collection, analytical sensitivity
may be a limiting factor due to small volumes typically collected and the
short precorneal residence time typically encountered. Nevertheless, when
area under the tear concentration-time curve values are available, bioavailabilities can be calculated and compared between formulations. Tear sampling technique may be an important factor in obtaining this type of data.
For example, Ding et al. compared two tear sampling techniques (capillary
tubes and Schirmer strips) and one recovery technique (cotton swabs) for
suitability for determining precorneal drug levels as a function of time for
ophthalmic gels (47). All three methods yielded similar corneal contact times
(about 1 hour). Capillary tubes proved eective for tear sampling, while the
strip method suered from gel carry-over. The cotton swab technique was
gentle, easy, and nondestructive and yielded recovery of total drug in the
cul-de-sac, but failed to provide information for tear lm.
F

2. Ocular Distribution
Compared to absorption and elimination, distribution is generally more
dicult to describe kinetically. In the concentration-time curve shown in
Figure 4, distribution is associated with the log-linear decline in aqueous
humor concentration immediately following peak concentration and before
the terminal elimination phase.
a. Distribution Within the Anterior Segment. The fundamental parameter to describe distribution is volume of distribution Vd , which is de-

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

73

Figure 4 Ideal sampling times for determining the pharmacokinetics of topical

ocular drugs. (Adapted from Ref. 8.)

ned as a proportionality constant relating concentration to amount. Volume of distribution at steady-state (Vss ) most closely reects the distribution capacity and, as such, is the most useful measure of Vd . Unlike
whole body Vss , determining Vss in the eye is particularly difcult because
the amount of drug in the eye at any time is not known following a single
topical ocular administration. Two approaches have been developed to
overcome this obstacle. The rst method uses a well afxed over the cornea of an anesthetized rabbit so that drug solution is in constant contact
with the cornea and not the surrounding sclera (48,49). Drug solution remains on the cornea for 90160 minutes until concentration of drug in the
aqueous humor reaches steady state. This so-called topical infusion study,
with constant rate input, can yield absorption rate constant, ocular clearance, and Vss .
The second method for determining Vss and distribution involves
intracameral injection, which delivers a bolus dose directly into the anterior
chamber. Conrad and Robinson used this approach to determined the
volume of distribution in albino rabbits (50). Inulin yielded an aqueous

Copyright 2003 Marcel Dekker, Inc.

74

Chastain

humor volume of 287 mL in close agreement with that previously determined

by other methods. The apparent volume of distribution of pilocarpine was
twofold larger than the assumed apparent volume of 250300 mL, indicating
distribution into surrounding tissues. Rittenhouse et al. examined the ocular
uptake and disposition of topical ocular -adrenergic antagonists in individual dogs and rabbits (51). Radiolabeled (3H) propranolol was administered
intracamerally and topically, and microdialysis was performed to monitor
concentrations of radioactivity in the anterior chamber. Ocular bioavailability was 5.6% and 55% in anesthetized dog and rabbit, respectively. The
value in rabbit was outside the 110% range typically observed with topical
ocular instillation. The high bioavailability may be related to the use of
anesthesia, which can cause a reduction in blinking and precorneal drainage.
Table 6 presents ocular Vd values determined using either technique for
various drugs.
Reliable estimates of Vss are useful for establishing multiple dose regimens; however, because of the somewhat specialized methods needed to
determine Vss in the eye following topical ocular administration, it is not
surprising that Vss has been estimated for only a few drugs. An obvious
alternative for evaluating distribution is to simply measure directly the concentration of drug in ocular tissues. Researchers have taken this approach

Table 6

Ocular Volumes of Distribution Vd for Various Drugs

Drug
2-Benzothiazolesulfonamide
Ethoxzolamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide
Phenylephrine
Clonidine
Aminozolamide
Pilocarpine
Flurbiprofen
Levobunolol
Dihydrolevobunolol
Ketorolac tromethamine
a

Vd (mL)

Method

Ref.

0.24
0.28
0.33
0.42
0.53
0.53
0.58
0.62
1.65
1.68
1.93

SSa
SS
SS
SS
SS
SS
EXTb
EXT
EXT
EXT
EXT

48
48
48
43
67
111
50
46
27
27
26

SS = steady-state method. A constant concentration of drug is applied to the

cornea of anesthetized rabbit. VSS is calculated from aqueous humor concentration
versus time data.
b
EXT = Extrapolated method. C at t0 is determined by log-linear extrapolation
and divided into the intracameral dose.
Source: Adapted from Ref. 8.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

75

for many years, generating a fairly large database of tissue concentration

information. Many of these studies have employed radioactive methods to
achieve the detection limits needed to measure the small amounts of drug
often encountered with small-sized ocular tissues. Unfortunately, unless
amounts are high enough for chromatographic separation or metabolism
is known not to occur, the identity of the radioactivity, i.e., parent drug
versus metabolite, remains in question. This problem is overcome by the use
of highly sensitive and selective bioanalytical methods, such as high-performance liquid chromatography coupled with mass spectrometry (HPLC/MS
or HPLC/MS/MS).
Most topical ocular drugs demonstrate a tissue distribution pattern
within the eye consistent with corneal penetration, that is, a concentration
gradient of cornea > conjunctiva > aqueous humor > iris-ciliary body >
lens, vitreous, and/or choroid-retina (8). However, with certain topical ocular drugs, such as clonidine, timolol, dapiprazole, oxymetazoline, and ketorolac tromethamine, iris-ciliary body concentrations are higher than those in
aqueous humor (8). A number of explanations have been oered for this
phenomenon, but the most plausible appears to involve conjunctival/scleral
absorption with direct distribution to the iris-ciliary body. Several studies
support this idea as discussed previously in reference to noncorneal, productive absorption (2025).
Distribution kinetics can be signicantly inuenced by binding of drug
to tissue. In addition, if binding anity is high, elimination from the tissue
will be delayed. A common observation with ophthalmic drugs is binding to
pigmented tissues, such as the iris-ciliary body. Binding to melanin is usually
evidenced by greater concentrations in pigmented rabbit eyes than nonpigmented, albino eyes. Lindquist has presented a comprehensive review of
binding of drugs to melanin (52). Patil and Jacobowitz investigated the
accumulation of adrenergic drugs by pigmented and nonpigmented irides
(53). At the highest concentration of 10 mM, accumulation of norepinephrine, epinephrine, and phenylephrine by the pigmented iris was twofold
higher than that in the nonpigmented iris. Araie et al. demonstrated that
-adrenergic blockers can also bind to melanin-containing tissues and may
be slowly eliminated from these tissues (54). Their data suggest that ecacy
may be reduced short-term and binding may occur after long-term use in
heavily pigmented subjects. Lyons and Krohn demonstrated that pigmented
irides and ciliary bodies accumulated two- to threefold more pilocarpine
than nonpigmented tissue (55), and Chien et al. showed a pigment binding
eect with 14C-brimonidine (56). In a study by Achempoing et al., iris-ciliary
body levels of brimonidine peaked at 40 minutes and 1.5 hours and declined
with a half-life of 1 hour and 160 hours in albino and pigmented rabbits,
respectively (57). These results indicate that elimination from pigmented

Copyright 2003 Marcel Dekker, Inc.

76

Chastain

tissue is prolonged compared to that in nonpigmented tissue. Pigment binding may also reduce ecacy as demonstrated by Nagata et al (58). In vivo,
topically applied timolol and pilocarpine lowered intraocular pressure (IOP)
in albino but not in pigmented rabbits.
b. Distribution to the Posterior Segment. The aging of the general
population, along with the higher incidences of eye diseases, such as
age-related macular degeneration or retinal edema, has created a need to
deliver drugs to the posterior segment (i.e., retina and choroid).
Although treatment of posterior diseases usually involves intraocular or
periocular injections, the advantages of topical ocular administration are
obvious.
It has generally been observed that drugs applied topically to the eye
do not reach therapeutic levels in the posterior segment tissues, except
perhaps by way of absorption from the percorneal area into the systemic
circulation and redistribution into the retina/choroid (59). However, there
are a few studies suggestive of topical ocular drugs reaching the posterior
segment by direct distribution. In a study in monkeys, Dahlin et al.
estimated the contributions of local ocular versus systemic delivery to
posterior-segment concentrations of betaxolol at steady state following
multiple topical dosing of Betoptic S (60). Signicant levels of betaxolol
were found in the retina and optic nerve head. A comparison of dosed
versus nondosed eye tissue concentrations revealed that most of the drug
in the posterior segment was from local delivery (absorption) with some
contribution from the systemic plasma. High concentrations in the irisciliary body, choroid, and sclera suggested the presence of a depot, which
possibly facilitated transfer to the retina and optic nerve head. In another
example, Chien et al. evaluated the ocular distribution of brimonidine in
albino and pigmented rabbits following a single topical ocular dose of 14Clabeled drug (56). The results indicated that drug was retained in choroid/
retina and optic nerve head. Levels in the nondosed contralateral eyes were
much lower than those in the treated eyes for both albino and pigmented
rabbits, suggesting that the majority (>99%) of the intraocularly
absorbed drug was due to local topical application and not to redistribution from plasma.
The mechanism by which drugs may be locally delivered to the posterior segment from the precorneal area is unknown, but the evidence seems to
indicate a noncorneal route, possibly involving conjunctival/scleral absorption followed by distribution to choroid, vitreous, and retina. Romanelli et
al. conrmed the existence of a noncorneal alternative route in their investigations of the posterior distribution of drugs (61). Concentrations in retina
were lower than those in aqueous for drugs that easily penetrate into the

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

77

aqueous, while levels in retina were equal or higher for drugs with poor
penetration into aqueous. The authors concluded that topically applied
drugs reach the retina not by passing through the vitreous or anterior
chamber, but possibly by an alternative route of drug penetration into
the eye.
3.

Elimination

The elimination phase typically appears as the terminal log-linear portion of

the concentration versus time plot (see Fig. 4). Elimination rate from the
aqueous humor within the anterior chamber is of particular interest.
Elimination from the anterior chamber can involve aqueous humor outow,
elimination by distribution into the lens, and/or metabolism.
Elimination rate from the aqueous humor varies little from drug to
drug, as shown in Table 7 in terms of half-life. The half-lives fall within a
relatively narrow range of about 0.36 hours for a wide variety of drugs.
Determining the aqueous humor half-life is largely dependent on the sensitivity of the analytical method employed. If the method is insuciently
sensitive, terminal concentrations will be missed and the half-life may be
underestimated. On the other hand, if the method reports false concentrations at the low end of analytical sensitivity, the half-life will be overestimated.
Clearance is the most common and useful parameter for expressing
elimination and is dened as a proportionality constant relating concentration to rate of drug loss. Ocular clearance Qe ) can be calculated using any
one of the following equations (8):
Qe Ke  Vd

Qe

K0  T
AUCinf

Qe

Dic
AUCinf

where Ke is the rst-order rate constant for elimination from the aqueous
humor, Vd is the apparent ocular volume of distribution, K0 is the constant
rate input into the anterior chamber, T is the duration of constant rate
input, Dic is the intracameral dose, and AUCinf is the area under the aqueous humor concentration-time curve extrapolated to innity. These equations require the accurate determination of Ke and AUCinf . Table 8 shows
aqueous humor clearances for a number of drugs.

Copyright 2003 Marcel Dekker, Inc.

Table 7 Aqueous Humor Half-Livesa of Ophthalmic Drugs

Administered to the Rabbit Eye Topically, Subconjunctivally or
Intracamerally
Drug
Dapiprazole
Timolol
Imirestat
6-Mercaptopurine
Ciprooxacin
Fusidic acid
Suprofen
Histamine
Cimetidine
Noroxacin
Ketorolac tromethamine
Benzolamide
I-643,799b
Flurbiprofen
Lomeoxacin
Phenylephrine
Timolol
6-Amino-2-benzothiazolesulfonamide
L-662,583c
Acetbutolol
Cefsulodin
5-Fluorouracil
Brimonidine
Dihydrolevobunolol
6-Hydroxyethoxy-2-benzothiazolesulfonamide
6-Acetamido-2-benzothiazolesulfonamide
L-650,719d
Triuoromethazolamide
Tobramycin
Levobunolol
Ethoxzolamide
Pyrilamine
Methazolamide
Cefotazime
Bufuralol
Clonidine
BCNUe
2-Benzothiazolesulfonamide
a

Half-life (h)

Ref.

5.8
4.8
4.75
4.6
3.5
2.8
2.52.75
2.2
2.2
2.2
2.1
2.0
1.8
1.7
1.6
1.4
1.2
0.84
1.15
1.11
1.1
1.0
1.0
1.3
0.98
0.97
0.93
0.81
0.78
0.75
0.67
0.63
0.23
0.61
0.58
0.57
0.50
0.49
0.38
0.29

127
126
120
85
68
128
129
130
130
131
26
132
133
134
131
43
135
45
111
125
45
136
86
57
27
48
111
133
132
62
27
48
132
130
132
137
45
67
94
48

Half-lives based on terminal phase of log concentration-time curve.

6-Hydroxybenzothiazide-2-sulfonamide.
c
5-(3-Dimethylaminomethyl-4-hydroxy-phenylsulfonyl) thiophene-2-sulfonamide.
d
6-Hydroxy-2-benzothiazidesulfonamide.
e
1,3-Bis(2-chloroethyl)-1,1-ntirosurea.
Source: Adapted from Ref. 8
b

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

79

Table 8 Ocular Clearances (Qe from Aqueous Humor for

Various Drugs
Drug
Levobunolol
Dihydrolevobunolol
Clonidine
Phenylephrine
Fluriboprofen
Pilocarpine
Ketorolac tromethamine
Ethoxzolamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide
2-Benzothiazolesulfonamide

Qe (mL/min)

Ref.

28.7
19.7
14.9
14.6
14.4
13.0
11.0
9.0
3.0
1.15

27
27
67
43
46
138
26
49
49
49

Source: Adapted from Ref. 8.

4.

Ocular Pharmacokinetic Models

Various approaches to modeling pharmacokinetic data from ocular studies

have been developed. These primarily involve (a) classical, empirical compartmental modeling, (b) physiological modeling, and (c) noncompartmental modeling employing statistical moment theory. Critical to pharmacokinetic modeling is an adequate description of the concentration versus
time curve. Unfortunately, the rationale for selecting the number and
timing of sample intervals is not always clear. In many cases, a sampling
scheme may be dictated by bioanalytical sensitivity, practicality, or even
economics. However, Schoenwald (8) has provided general sampling
guidelines that may prove useful in properly designing ocular studies
(Fig. 4).
Not only are an appropriate number of time intervals necessary,
but there should also be a sucient number of replicates per time
point. Rabbits studies have used as few as 4 and as many as about
20 eyes per time interval, depending on the study objectives. Higher
numbers are preferred for bioequivalence studies. Moreover, because
the animal is typically euthanized for ocular tissue/uid sampling,
each animal subject (e.g., rabbit or monkey) yields at most two data
points (one/eye) per interval. Mean concentrations, derived from individual animal data, are then used to calculate the pharmacokinetic parameters. A statistical approach to this sparse sampling scheme has been

Copyright 2003 Marcel Dekker, Inc.

80

Chastain

developed for comparison of area under the concentration-time curves

(62,63).
a. Classical Empirical Pharmacokinetics. Classical modeling uses
compartments, representing kinetically homogeneous groups of tissues/organs, linked together by various rate constants. For ocular pharmacokinetics, the simplest model is one employing a single compartment as
shown in Figure 5a. However, this model fails to account for precorneal
loss. The model in Figure 5b corrects for this but is still very simplied
and treats the cornea as a homogeneous tissue, lumping all precorneal

Figure 5 Schematics of various models of topical ocular drug pharmacokinetics

FD = Bioavailability times dose; C = cornea; PC = precorneal area; AH =
aqueous humor; AC = anterior chamber; R = reservoir; Epi = corneal epithelium;
Str = corneal stroma; Endo = corneal endothelium. (Adapted from Refs. 4, 44, 64.)

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

81

rate constants together into one. The model in Figure 5c divides the anterior segment of the eye into cornea and aqueous humor, although this still
excludes precorneal loss and does not adequately describe disposition beyond entry into the aqueous. Makoid and Robinson proposed a fourcompartment caternary model for pilocarpine, as shown in Figure 5d,
which combines precorneal loss with differentiation of cornea and anterior
chamber (44). However, cornea is still treated as a homogeneous tissue,
when the epithelium is known to be the major barrier to ocular uptake.
The model in Figure 5e portrays epithelium as one compartment and stroma/endothelium/aqueous humor as a separate lumped compartment and
incorporates elimination from each of the compartments (64). An example
of one of the more sophisticated models is shown in Figure 6, which includes a conjunctival compartment, along with sclera, intraocular, and
systemic circulation compartments, as well as redistribution to the contralateral eye (21).
b. Physiological Model. Beyond the classical compartmental
modeling approach is one that incorporates more realistic physiological components. Physiological pharmacokinetic models are intuitively
more predictive by their use of actual anatomical and physiological
parameters, such as tissue blood ow and volume (65). Ocular pharmacokinetics appears to be an ideal candidate for physiological modeling,
since it is relatively simple to remove the tissue components of the eye
for measurement of drug levels. For example, Himmelstein et al. developed

Figure 6 Schematic of an ocular pharmacokinetic model showing precorneal

events, absorption into the eye via the cornea or conjunctive/sclera, distribution to
the systemic circulation, and redistribution to the contralateral (undosed) eye. PC =
Precorneal area; Conj = conjunctiva; C = cornea; S = sclera; AC = anterior
chamber; IT = intraocular tissues; OC = ocular circulation; SC = systemic circulation; CE = contralateral eye. (Adapted from Ref. 21.)

Copyright 2003 Marcel Dekker, Inc.

82

Chastain

a simple physiological pharmacokinetic model, shown in Figure 7, for

predicting aqueous humor pilocarpine concentration following topical
application to rabbit eyes (66). The model takes into account instilled
volume and drug concentration and can predict the effect of precorneal
drainage.
Physiological modeling may not be the best approach in all cases. For
example, Chiang and Schoenwald determined the concentrations of clonidine in seven dierent ocular tissues and plasma after a single topical ocular
dose of clonidine was administered to rabbits (67). The data were t to a
physiological model and a classical diusion model with seven ocular tissue
compartments and a plasma reservoir. The complex classical model was
subdivided into fragmental models. While predicted and observed concentrations proles closely agreed with the physiological model, the classical
model t the data better than the physiological model.
c. Other Models. A few other modeling approaches have been proposed, including noncompartmental modeling and population pharmacokinetics. Eller et al. applied noncompartmental statistical moment theory
to topical infusion data to describe the disposition of various compounds
with a range of transcorneal permeabilities within the rabbit eye (48).
Morlet et al. used population pharmacokinetics to evaluate pharmacokinetic data in plasma and vitreous of the human eye (68). Gillespie et al.
applied principles and methods of linear system analysis to the analysis of
ocular pharmacokinetics (69). Using convolution integral mathematics, a

Figure 7 Schematic of a two-compartment model consisting of the precorneal area

and the aqueous humor. PC = precorneal; QT = normal tear production rate; VT
= total volume in the precorneal area any given time; K = proportionality constant
that is a function of instilled drop size; V0 = normal tear volume; VAH = normal
aqueous humor volume A = corneal area; L = corneal thickness; Kel = lumped
rst-order clearance parameter from aqueous humor. (Adapted from Ref. 66.)

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

83

mechanistic model of precorneal disposition was used to predict concentration. The authors adequately predicted betaxolol levels resulting from a
multiple dose regimen and from single doses of prototype controlled-release ocular inserts. This approach appears to require fewer, less restrictive assumptions than compartmental or physiological model methods.
B.

Intravitreal Administration

Intravitreal injection is the most direct approach for delivering drug to the
vitreous humor and retina; however, this method of administration has been
associated with serious side efects, such as endophthalmitis, cataract,
hemorrhage, and retinal detachment (59). In addition, multiple injections
are usually required, further increasing the risk. Nevertheless, intravitreal
injection continues to be the mode of choice for treatment of acute intraocular therapy.
The kinetic behavior of intravitreally delivered drugs is complicated by
the stagnent, nonstirred nature of the normal vitreous. Mechanisms that
may inuence movement of molecules within the vitreous include diusion,
hydrostatic pressure, osmotic pressure, convective ow, and active transport
(70). For small to moderately sized molecules, such as uorescein or dextran, diusion is the predominant mechanism of transvitreal movement
(4,70). Although low-level convective ow has been observed within the
vitreous (71), this ow has only a negligible eect on transvitreal movement
in comparison to diusion. For small to moderately sized molecules, diusion within the vitreous is generally unimpeded and similar to that observed
in water or saline (4,70).
1.

Distribution and Elimination

As shown in Figure 8, drug distribution and elimination can occur in two

main patterns: diusion from the lens region toward the retina with elimination via the retina-choroid-sclera or anterior diusion with elimination
via the hyloid membrane and posterior chamber (4). A molecules path of
distribution and elimination in the vitreous largely depends on its physiochemical properties and substrate anity for active transport mechanisms in
the retina. Lipophilic compounds, such as uorescein (72) or dexamethasone
(73), and transported compounds tend to exit mainly via the retina. On the
other hand, hydrophilic substances, such as uorescein glucuronide, and
compounds with poor retinal permeability, such as uorescein dextran, diffuse primarily through the hyloid membrane into the posterior chamber and
eventually into the anterior chamber (72). Table 9 shows vitreal half-lives
for a variety of drugs and eye conditions. Generally, shorter half-lives are

Copyright 2003 Marcel Dekker, Inc.

84

Chastain

Figure 8 Schematic of exit pathways from the vitreous humor: (a) transretinal, (b)
by way of drainage out of the aqueous humor; (c) compartmental model showing
kinetic relationships between a, anterior chamber, p, plasma, and v, vitreous.
(Reprinted with permission from Maurice, D. M. and Mishima, S. (1984). Ocular
pharmacokinetics. In: Pharmacology of the Eye (M. L. Sears, ed.). Springer-Verlag,
Berlin, p. 73.) (Ref. 4).

associated with elimination through the retina, with its high surface area,
while longer half-lives are reective of elimination through the hyloid
membrane and the anterior segment.
Injection volume and position within the vitreous body can also inuence the distribution and elimination pattern of a drug. Friedrich et al.
demonstrated that these factors had a substantial eect on vitreal distribution and elimination of uorescein and uorescein glucuronide (74). Four
extreme positions and injection volumes of 15 or 100 mL were considered.
The mean concentration of drug remaining in the vitreous 24 hours postdose
varied by up to 3.8-fold depending on injection position, and increasing
injection volume dampened this eect.
Retinal inammation (which can cause a breakdown of the bloodretinal barrier) and aphakia are common pathophysiological states that
can alter a drugs vitreal kinetics. Friedrich et al. showed that drug diu-

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

85

Table 9 Vitreous Humor Half-Lives of Drugs Administered

Intravitreally
Drug

Species

ISIS 2922a
Foscarnet
Octreotide acetate
Vancomycin
Amphotericin B

Rabbit
Rabbit
Cat
Rabbit
Rabbit

Rifampin
Carbenicillin

Rabbit (i.v.)
Rabbit
Monkey
Monkey
Cat

Gentamicin

Dexamethasone
Ganciclovirb
Cefazolin
Ceftizoxime
Ceftriazone
Ceftazidime
Cefepime
a
b

Monkey
Rabbit
Rabbit
Rabbit
Human
Monkey
Monkey
Rabbit
Rabbit
Rabbit
Rabbit
Rabbit
Rabbit
Rabbit
Rabbit

Half-Life (h)

Ref.

62
34
16
12.514.5
6.915.1
1.8 days (aphakic)
5.59
5.0
10
20 (+ probenecid)
12.6
6.5 (infected)
33
3.0
3.48
0.017 cm2 hr1
0.015 cm2 hr1
7
30 (+ probenecid)
5.7
9.4 (infected)
9.1
13.1 (infected)
20.0
21.5 (infected)
14.3
15.1 (infected)

139
140
141
142
76
76
95
143
81
81
77
77
81
73
144
145
145
81
81
78
78
78
78
78
78
78
78

Phosphorothioate oligonucleotide.
Half-lives for ganciclovir normalized for volume and retinal surface area.

sivity and retinal permeability are important factors that determine elimination from the vitreous particularly in the case of blood-retinal barrier compromise (75). Furthermore, drug is eliminated faster in aphakic eyes,
especially for drugs with low retinal permeability and injected proximal to
the lens capsule. Injection close to the primary elimination barrier appears
to be a key factor. Wingard et al. showed that intravitreally injected amphotericin B progressively accumulated in the sclera-choroid-retina in control
phakic eyes, a phenomenon not observed in aphakic eyes (76). Whole phakic

Copyright 2003 Marcel Dekker, Inc.

86

Chastain

eye half-life was 6.915.1 days,while aphakic half-life was only 1.8 days. In
the case of infected eyes, Ben-Nun et al. showed, with intravitreal injection
of gentamicin, that the elimination rate of drug was greater in infected than
normal eyes, presumably due to an alternation in blood-retinal barrier (77).
In another evaluation of the vitreal kinetics of ceftizoxime, ceftriazone,
ceftazindime, and cefepime in rabbits, T1=2 values ranged from 5.7 to 20
hours in rabbits with uninamed eyes and from 9.4 to 21.5 hours in rabbits
with infected eyes (78). The longer T1=2 suggested a predominant anterior
route of elimination, while the shorter T1=2 and low aqueous/vitreous concentration ratios suggested retinal elimination.
As mentioned, some compounds are actively transported out of the
vitreous leading to a faster elimination than expected based on physicochemical properties; for example, Mochizuki investigated the transport of
indomethacin in the anterior uvea of the albino rabbit in vitro and in vivo
(intravitreal injection) (79). An energy-dependent carrier-mediated transport mechanism with low anity was observed in the anterior uvea of the
rabbit that could have accounted for the drugs rapid clearance (30% per
hour) from the eye. Yoshida et al. characterized the active transport
mechanism of the blood-retina barrier by estimating inward and outward
permeability of the blood-retinal barrier in monkey eyes using vitreous
uorophotrometry and intravitreally injected uorescein and uorescein
glucuronide (80). Outward permeability (Pout ) was 7.7 and 1.7  104
cm/min, respectively. Pout =Pin was 160 for uroescein and 26 for uorescein glucuronide. Intraperitoneal injection of probenecid caused a signicant decrease in Pout for uorescein but had no eect on uorescein
glucuronide Pout . The data suggest that uorescein is actively transported
out of the retina. In another example, Barza et al. studied the ocular
pharmacokinetics of carbenicillin, cefazolin, and gentamicin following
intravitreal administration to rhesus monkeys (81). Vitreal half-lives
ranged from 7 to 33 hours. Concomitant intraperitoneal injection of
probenecid prolonged the vitreal half-life of the cephalosporins, indicating
a secretory mechanism. The results are consistent with the hypothesis that,
in primates (as in rabbits), -lactam antibiotics are eliminated by the
retinal route and aminoglycosides by the anterior route.
2. Vitreal Pharmacokinetic Models
Several models have been proposed to describe the kinetics of intravitreally injected drugs. The simplest models assume a well-stirred vitreous
body compartment in an eort to reduce the complexity of the mathematics. This may be a closer approximation in studies employing injection
volumes of 100 mL or more, where the normally nonstirred vitreous can

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

87

become agitated; however, injection volumes of greater than 20 mL

generally require removal of an equal volume of vitreous to avoid a precipitous rise in intraocular pressure. This act alone may alter the physiology of the vitreous body. More sophisticated modeling takes into account
diusion through the relatively stagnant vitreous humor by employing
Ficks law of diusion. For example, Ohtori and Tojo determined the
elimination of dexamethasone sodium m-sulfobenzoate (DMSB) following
injection in the rabbit vitreous body under in vivo and in vitro conditions
(82). The rate of elimination was greater in vivo versus in vitro. A
general mathematical model, based on Ficks second law of diusion,
was developed to describe the pharmacokinetics. The model assumed a
cylindrical vitreous body with three major elimination pathways: posterior
aqueous chamber, retina-choroid-scleral membrane, and lens (see Fig. 9).
Concentration in the vitreous decreased rapidly near the posterior aqueous
chamber, indicating that the annular gap between the lens and ciliary body

Figure 9 Cylindrical model of the vitreous body of rabbits. The posterior chamber,
the retina-choroid-sclera (RCS), and the lens constitute elimination pathways out of
the vitreous. (From Ref. 82.)

Copyright 2003 Marcel Dekker, Inc.

88

Chastain

(posterior chamber) was the major route of elimination. The concentration

gradient near the retina was considerable. It was concluded that, because
of its large surface area, the retina can be a signicant route of elimination. In a seprate study, Tojo and Ohtori used the cylindrical model
approach to demonstrate three potential pathways of elimination including
the annular gap, the lens, and the retina-choroid-sclera (83). The concentration in the retina was aected by the site of injection or initial distribution proles, while concentration at the lens was independent of dose site.
Drug injected into the anterior segment of the vitreous rapidly exited
through the annular gap into the posterior chamber. The authors reasoned
that drugs should be injected into the posterior vitreous to prolong therapeutic levels in the retina. Their results also showed that half-life was
proportional to molecular weight and elimination into the lens was neglible due to the barrier function of the lens capsule.
Probably the most precise modeling of vitreal pharmacokinetics uses
nite element analysis, a method commonly employed in engineering. This
approach accounts for the detailed geometry and boundary conditions of
the vitreous and precisely predicts the concentration gradients within the
vitreous. Friedrich and colleagues adapted nite element modeling to the
study of the drug distribution in the stagnent vitreous humor of the rabbit
eye after an intravitreal injection of uorescein and uorescein-glucuronide
(74,75,84). The computer-generated concentration prole in the vitreous
humor is shown in Figure 10. Retinal permeability of uorescein and uorescein glucuronide were estimated by the model at 1:94  105 to 3:5  105
cm/s and from 0 to 7:62  107 cm/s, respectively. Simulations have also
been performed for the human eye (74). In both rabbit and human eyes, the
eect of injection position was found to be an important variable, as indicated in Figure 11 (84).
C.

Periocular Administration

As previously mentioned in this chapter, while there is some evidence for

direct drug delivery via the topical ocular route, drugs usually do not reach
therapeutically relevant levels in the posterior segment following topical
ocular instillation. If signicant concentrations are achieved at the back of
the eye, they are usually the result of redistribution from the systemic circulation, not local delivery. Consequently, to treat diseases of the posterior
segment, drug must typically be administered intravitreally, periocularly, or
systemically. Systemic administration will be discussed later in this chapter.
Intravitreal injection has been discussed and is quite eective but, as has
been mentioned, presents a serious risk to the eye. Periocular drug administration, using subconjunctival, sub-Tenons, or retrobulbar injection, is

Copyright 2003 Marcel Dekker, Inc.

Figure 10 Finite element model of the vitreous body. The injection is hyloid displaced cylindrical. (Reprinted with permission from Friedrich et al. (1997) Finite
element modeling of drug distribution in the vitreous humor of the rabbit eye.
Ann. Biomed. Eng., 25:306.) (Ref. 84).

Figure 11 Relationship betwen normalized concentration of uorescein in the

vitreous humor and distance from the lens. (Reprinted with permission from
Friedrich et al. (1997) Finite element modeling of drug distribution in the vitreous
humor of the rabbit eye. Ann. Biomed. Eng., 25:310.) (Ref. 84).
89

Copyright 2003 Marcel Dekker, Inc.

90

Chastain

another and, in many cases, preferred, route for delivering drugs to the
posterior segment (59).
1. Subconjunctival Administration
Subconjunctival injection oers the advantage of local drug delivery
without the invasiveness of intravitreal injection. This route also allows
for the use of drug depots to prolong the duration of drug therapy
and avoids much of the toxicity encountered with systemic administration. Drug concentrations in the eye are typically substantially higher
following subconjunctival versus systemic administration, while systemic
exposure is greatly reduced with subconjunctival dosing. For example,
following subconjunctival injection of 6-mercatopurine, mean peak concentrations in aqueous and vitreous were 15 and 10 times those following intravenous administration, while serum levels were about half (85).
In another example, rabbits were administered 14 C-5-uorouracil either
subconjunctivally or intravenously (86). Peak levels of parent in the
serum and urine were similar for the two routes; however, subconjunctival injection resulted in peak aqueous concentrations of 125 and 380
times that after intravenous injectiotn. The localized deliver of hydrocortisone by the subconjunctival route has been demonstrated by
McCartney et al. in the rabbit eye (87). Their results showed that
hydrocortisone penetrated directly into the eye with minimal spread
beyond the site of administration.
Various studies have explored the mechanism by which drugs are
absorbed into the eye following subconjunctival administration. Maurice
and Mishima point to direct penetration to deeper tissues as the main
pathway of entry into the anterior chamber (4). A necessary rst condition, however, is the saturation of the underlying sclera with drug.
This is followed by diusion by various possible routes: laterally into
corneal stroma and across the endothelium, across trabecular meshwork, through the iris stroma and across its anterior surface, into
the ciliary body stroma and into newly generated aqueous humor,
and into the vitreous body via the pars plana and across its anterior
hyloid membrane (4). In addition to these pathways, depending on the
injection volume, regurgitation out the dose site with subsequent spillage onto the cornea can lead to direct transcorneal absorption. For
example, Conrad and Robinson investigated the mechanism of subconjunctival drug delivery using pilocarpine nitrate, albino rabbits, and
instillation volumes ranging from 60 to 500 mL (88). At high injection
volumes (>200 mL), the primary mechanism for uptake into the aqueous
was reux of the drug solution from the injection site followed by corneal

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

91

absorption. At lower volumes, the mechanism involved reux and transconjunctival penetration, permeation of the globe, and systemic absorption
followed by redistribution.

2.

Periocular Injection (Sub-Tenons and Retrobulbar)

Sub-Tenons injection involves delivery of drug, usually as a depot, between

the sub-Tenons capsule and sclera or episclera. This route of administration
has the advantage of placing drug in very close proximity to the sclera. Drug
can subsequently diuse through the sclera, which is quite permeable to a
wide range of molecular weight compounds (59,89). Because the diusion of
drug from the dose site can be very localized, the preferred location of dose
is directly over the target. For example, Freeman et al. performed localization of sub-Tenons repository injection of corticosteroid (90). Echography
showed drug within the sub-Tenons space over the macula in 11 of 24 cases.
The lack of therapeutic response to repository steroids was attributed to
placement relative to target.
Except for the observation that bleb retrobulbar injection spreads
forward, unlike subconjunctival injection which spreads backward, these
two injection routes yield very similar results. Bodker et al. compared
ocular tissue levels of dexamethasone 1 and 4 hours after subconjunctival
or retrobulbar injection in rabbits (91). In both dosage route groups,
concentrations in all three tissues (aqueous, vitreous, retina) were similar
1 hour postdose. After 4 hours, levels in the two groups were again
similar except in choroid. Dosed and contralateral undosed eye tissues
contained similar levels after 4 hours with the exception of retina,
which had lower levels in undosed eye versus dosed. Retrobulbar dosing,
however, provided a more sustained drug delivery than with subconjunctival administration.
Retrobulbar (or peribulbar) injection is another option for delivering
drug to the posterior segment and the vitreous. Hyndiuk and Reagan determined the penetration and persistence of retrobulbar depot-corticosteroid in
monkey ocular tissues (92). High concentrations were found in posterior
uvea with persistence of lower concentrations. Steroid tended to concentrate
in the optic nerve after retrobulbar but not systemic administration, and no
drug was detected in other ocular tissues with the exception of lens and
vitreous after 2 and 9 days. Weijten et al. studied the penetration of dexamethasone into the human vitreous and its systemic uptake following peribulbar injection (93). Mean levels in the vitreous peaked at 13 ng/mL at 67
hours postdose, and maximal serum level was 60 ng/mL 2030 minutes
postdose.

Copyright 2003 Marcel Dekker, Inc.

92

D.

Chastain

Systemic Administration

Because local drug delivery to the eye generally provides direct access to the
site of action and, in most cases, substantially reduces systemic exposure and
toxicity, systemic administration of drugs to treat ocular diseases is generally not preferred. Moreover, lower ocular concentrations are usually
achieved compared to those following direct ocular dosing. However, for
drug delivery to the posterior segment or vitreous body, systemic administration could be the best choice depending on the drugs ability to penetrate
the blood-retinal barrier or blood-vitreous barrier and its systemic toxicity
prole. For example, in a study by Ueno et al., concentrations of BCNU
were measured in aqueous and vitreous of rabbits following intravenous,
subconjunctival, and topical ocular administration (94). Distribution was
dependent on dose route in that topical, followed by subconjunctival, was
best for distribution into the iris, while intravenous, was best for distribution
into the choroid-retina. In another example, Liu et al. demonstrated that
rifampin penetrated vitreous humor after an intravenous single dose (95).
Compromising the blood-vitreous or -retinal barrier can enhance
intraocular absorption following systemic administration. Wilson et al. treated the right eyes of rabbits with triple or single freeze-thaw cryotherapy at
one or two locations one day before intravenous carboplatin with or without cyclosporine (96). Cryotherapy increased the intravitreal penetration of
carboplatin. In a study by Elliot et al., following intravenous injection of
ganciclovir with and without RMP-7, a compound known to increase the
permeability of the blood-brain barrier, RMP-7 enhanced retinal uptake
through the blood-retinal barrier (97). Interestingly, Palastine and
Brubaker demonstrated that systemically administered (intravenous or
oral) uorescein can enter the vitreous through other means beyond an
increase in blood-retinal barrier permeability (98).

III.

OCULAR METABOLISM

The metabolic capacity of the eye is low compared to that of a primary

metabolizing organ such as liver. However, sucient enzymatic activity is
present to cause the breakdown of ophthalmic drugs in the eye. While this
biotransformation usually results in loss of ecacy, the development of
prodrugs takes advantage of the increased corneal permeability of the prodrug and the subsequent hydrolysis of the prodrug to active compound (16,
99,100). Prodrugs of pilocarpine (101), phenylephrine (102,103), timolol
(104), and prostaglandin F2
tromethamine have shown improved corneal
penetration.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

93

A variety of esterases responsible for prodrug hydrolysis have been

identied in the eye. Lee et al. reported the presence of acetyl-, butyryl-,
and carboxylcholinesterases in pigmented eyes (99). Over 75% of cholinesterase activy was due to butyrylcholinesterases in all ocular tissues but corneal epithelium. Their results suggested that esters with chain lengths
exceeding four carbons are hydrolyzed principally by butyrylcholinesterase.
In a separate study, Lee demonstrated that esterase activity ranked as
follows in ocular tissues: iris-ciliary body > cornea > aqueous humor
(100). Activities in the pigmented rabbit cornea and iris-ciliary body were
greater than in the nonpigmented eye, indicating a higher metabolic capacity
in pigmented ocular tissue. Narurkar and Mitra conrmed Lees ranking of
enzymatic hydrolysis activity with aliphatic esters of 5-iodo-20 -deoxyuridine
(105). Camber et al. evaluated the permeability and absorption of PGF2
and
its methyl and benzyl esters through isolated pig cornea in vitro. Hydrolysis
was shown to be due to butyrlcholinesterase in the corneal epithelium (106).
Although esterases have been studied to the largest extent, a number of
other enzymes have been identied within the eye. These include catechol-Omethyltransferase, monamine oxidase, steroid 6-betahydroxylase, oxidoreductase, lysosomal enzymes, peptidases, and glucuronide and sulfate transferase (107). In addition, cytochrome P450 enzymatic activity has been
observed in ocular tissue (108,109). Campbell et al. demonstrated, using paminobenzoate, aminozolamide, and sulfamethazine as substrates that there
is signicant arylamine acetyltransferase activity in the eye with the highest
to lowest ranking compared to a nonocular tissue as liver > iris-ciliary body
> corneal epithelium > stroma-endothelum (110). Phenotype (slow versus
fast acetylators) had no eect on activity in the eye. Putnam et al. also studied
the disposition of aminozolamide in the rabbit eye (111). Their results suggested that the topical IOP-lowering activity of aminozolamide was the result
of both metabolite levels in aqueous humor and iris-ciliary body and the
99%+ inhibition of carbonic anhydrase. Metabolite levels were highest in
cornea and iris-ciliary body, with less in aqueous.
Aldehyde reductase and ketone reductase have also been found in
ocular tissues (8). Compounds such as N-oxide, hydroxamic acid, sulfoxide,
and nitro drugs are reduced by aldehyde reductase in bovine ciliary body.
Using levobunolol as substrate, Lee et al. showed that the rank order of
ketone reductase activity was corneal epithelium > iris-ciliary body >
conjunctiva > lens. No activity was detected in tears, corneal stroma, sclera,
or aqueous humor. Models describing metabolism have been developed for
aminozolamide (111) and levobunolol (27).
In general, the iris-ciliary body and corneal epithelium possess the
greatest capacity for metabolic activity in the anterior segment. Moreover,
activity may be higher in pigmented versus nonpigmented ocular tissue.

Copyright 2003 Marcel Dekker, Inc.

94

IV.

Chastain

OCULAR PHAMACODYNAMICS

Fundamentally, pharmacodynamics is the study of the biological eects of

drugs (112). More specically, it is the study of how the concentration of
drug in the biophase (site of action) relates to the magnitude of biological
and therapeutic eects produced (113). However, in the evaluation of systemic pharmacodynamics, drug concentrations are typically determined in
plasma or serum and not the biophase. As a result, there is not always a
direct relationship between measured concentration of drug and biological
response. This shortcoming is somewhat overcome with ophthalmic drugs
since dose delivery and pharmacokinetic sampling are usually in very close
proximity to the biophase. Also, there are several quantiable ocular
responses, which are relatively easy to evaluate in vivo, including miosis,
mydriasis, intraocular pressure, corneal sensation, aqueous humor ow,
blink rate, tear secretion, and electroretinogram (4).
In conducting human ocular studies, measuring ocular pharmacokinetics in intraocular uids and tissues is dicult if not impossible. These
samples can only be removed at death of the patient or in the case of
removal of a diseased eye. As a result, pharmacological response measurements usually replace concentration assays in describing the eect of a drug
in the human eye. Unfortunately, responses can vary widely between individuals, presumably because of dierences in dose-response relationships
and ocular pharmacokinetics (8). Eye pigmentation, wearing of contact
lenses, allergies to drug or formulation components, and physiological factors all play a role in causing high variability in response data (8).
Nevertheless, pharmacodynamic measurements continue to be applied,
and where possible pharmacokinetic data are coupled with pharmacodynamic data.
A mathematical approach developed for miotic or mydriatic response
(3,4) is shown in the following equation:
R

R
qC
Rmax  R

where Rmax is the maximal response achieved by a drug with all receptors
occupied, q is a proportionality constant, and C is the concentration of
drug. Mishima showed that Eq. (6) predicted the miosis of carbachol and
pilocarpine on the sphincter muscle of the cat (see Fig. 12) (3).
Based on maximum and minimum pupil diameters of 8.5 and 1 mm,
respectively, it can be shown that for a miotic response,
Rl

D0  D
D  1

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

95

Figure 12 Dose-response relationship from carbachol (*) and pilocarpine (*) in

isolated human sphincter muscle strip. (Replotted from Ref. 3.)

and for a mydriasis response,

Rl

D  D0
8:5  D

where D0 is the pupil diameter before drug application and D is the pupil
diameter following drug administration (9). Figure 13 shows a log-linear
plot of the response versus time, from which is derived the following equation:
Rl RL eAtt0  eBtt0

where RL is the intercept value of Rl , t0 if the lag time after instillation when
response is observed, and A and B are apparent elimination and absorption
rate constants (9).
Chien and Schoenwald investigated aqueous humor concentration of
phenylephrine and the corresponding mydriatic response in rabbit eyes following topical ocular administration of phenylephrine or a prodrug of phenylephrine (114). Ocular bioavailability of 10% prodrug was eightfold
higher than that of the 10% HCl salt. However, mydriatic activity only
improved fourfold, demonstrating that the mydriatic response did not accurately reect drug distribution to the iris. The observed clockwise hysteresis
in the response versus concentration plot was suggestive of pharmacological
tolerance. An Emax model with a Michaelis-Menten relationship between

Copyright 2003 Marcel Dekker, Inc.

96

Chastain

Figure 13 Log-linear plot of pupillary response (miotic or mydriasis) versus time.

Dotted line indicates absorption phase after stripping o elimination phase.
(Reprinted with permission from Maurice, D. M., and Mishima, S. (1984). Ocular
pharmacokinetics. In: Pharmacology of the Eye. (M. L. Sears, ed.), Springer-Verlag,
Berlin, p. 56.) (Ref. 4).

response and concentration of phenylephrine in the iris was developed and

expressed as follows:

Emax  Ca t
10

Et
Km0 Ca t
where
Et is the dierence between pupillary diameter before and after
drug instillation at time t, Km0 is the drug concentration in the iris at half
maximal mydriatic response (one-half
Emax , and Ca t is the concentration of drug in aqueous humor at time t. In the case of tolerance, Km0 is no
longer constant but changes with time and can be calculated for each time
interval as:




Emax
0
Km t
11
 1  Ca t

Et
A plot of Km0 t versus time was linear up to 90 minutes but increased thereafter through 240 minutes postinstillation verifying the development of tolerance (114)

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

97

Smolen (115) described a methematical model for pharmacological

response from pilocarpine based on its dose-response relationship:
Imax

719:4D
0:5D
I 20:4D

12

Equation (12) requires the assumptions of nonhysteresis, rapid and reversible binding to the receptor, and rst-order kinetics. If these assumptions
are met, the equation can be generalized to other drugs.
Smolen and Schoenwald further developed a theoretical basis for the
performance of drug-absorption analysis from pharmacological response
versus time data following single, multiple, or continuous dosing of drug
by any route of administration (116119). Pharmacological and pharmacokinetic parameters characterizing the mydriatic behavior of tropicamide
were derived, and pharmacological response versus time plots yielded two
and three-compartment models.
An Emax model, similar to that for miosis, has also been proposed for
intraocular pressure (IOP) reduction (9). Mishima reviewed the pharmacodynamics related to IOP lowering (3). By virtue of the complexities of IOP
lowering, this response does not necessarily relate directly to drug concentrations as is the case for pupillary response.
Drugs typically have eects other than the measured response.
These extraneous actions might include alterations in aqueous humor
turnover or blood ow within certain ocular tissues. As a result, nonlinear behavior is likely to be observed, however, as Schoenwald has
discussed (8), the diculty in measuring concentrations in the human
eye will likely perpetuate the use of pharmacological responses to evaluate ophthalmic therapies.

V.

CONCLUSIONS

Understanding the ocular pharmacokinetic characteristics of a drug is critical to the optimization of its single and multiple dose therapeutic regimens.
However, because of the complex nature of ocular absorption, distribution,
and elimination, accurately measuring the ocular pharmacokinetics can be
dicult. The topical ocular route is preferred for reasons of ease of
administration, noninvasiveness, direct drug delivery, and minimization of
systemic eects. Even considering the complexities of ocular pharmacokinetics following topical ocular administration, several valid approaches
to its study have been developed. Many take into account the competing
pathways of absorption, leading to either ocular or systemic exposure.
Models in varying detail have also been developed for intravitreal adminis-

Copyright 2003 Marcel Dekker, Inc.

98

Chastain

tration. Compared to the kinetics associated with the topical ocular route,
vitreal pharmacokinetics is more straightforward in that it avoids convoluted precorneal events. However, due to the nonstirred nature of the
vitreous humor, the mathematics can be quite complex due to the inclusion
of Ficks law of diusion. Periocular administration is also dependent to a
large extent on diusion, which can be very localized. Both periocular and
systemic administration can be used to deliver drug to the posterior segment, however, the latter is usually the last resort due to the greater risk for
systemic drug exposure and toxicity.
Ocular pharmacokinetics, with all its limitations, is an important part
of ophthalmic drug development. Coupled with the knowledge of a drugs
metabolism and its pharmacodynamics, ocular pharmacokinetics provides
invaluable information needed to properly design single and multiple dose
regimens. In the future, with the discovery and renement of animal models
and the advent of more sophisticated modeling techniques and computer
programs, ocular pharmacokinetics/pharmacodynamics will likely receive
even more widespread use.

REFERENCES
1. Burstein, N. L., and Anderson, J. A. (1985). Review: Corneal penetration and
ocular bioavailability of drugs. J. Ocular Pharmacol., 1:309.
2. Lee, V. H. L., and Robinson, J. R. (1986). Review: Topical ocular drug
delivery: Recent developments and future challenges. J. Ocular Pharmacol.,
2:67.
3. Mishima, S. (1981). Clinical pharmacokinetics of the eye. Invest. Ophthalmol.
Vis. Sci., 21:504.
4. Maurice, D. M., and Mishima, S. (1984). Ocular pharmacokinetics. In:
Pharmacology of the Eye (M. L. Sears, ed.). Springer-Verlag, Berlin, pp. 19
116.
5. Schoenwald, R. D. (1990). Ocular drug delivery. Pharmacokinetic considerations. Clin. Pharmacokinet., 18:255.
6. Lang, J. C., and Stiemke, M. M. Biological barriers to ocular delivery. In:
Ocular Therapeutics and Drug Delivery (I. K. Reddy, ed.). Technomics
Publishing, Lancaster, pp. 51132.
7. Davies, N. M. (2000). Biopharmaceutical consideration in topical ocular drug
delivery. Clin. Exp. Pharmacol. Physiol., 27:558.
8. Schoenwald, R. D. (1993). Ocular pharmacokinetics/pharmacodynamics,
Ophthalmic Drug Delivery Systems (A. K. Mitra, ed.), Marcel Dekker, Inc.,
New York, pp. 83110.
9. Worakul, N., and Robinson, J. R. (1997). Ocular pharmacokinetics/pharmacodynamics. Eur. J. Pharm. Biopharm., 44:71.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

99

10. Schoenwald, R. D., and Ward, R. L. (1978). Relationship between steroid

permeability across excised rabbit cornea and octanol-water partition coecients. J. Pharm. Sci., 67:786.
11. Schoenwald, R. D., and Huang, H. S. (1983). Corneal penetration behavior of
-blocking agents I: Physicochemical factors. J. Pharm. Sci., 72:1266.
12. Huang, H. S., Schoenwald, R. D., and Lach, J. L. (1983). Corneal penetration
behavior of -blocking agents II: Assessment of barrier contributions. J.
Pharm. Sci., 72:1272.
13. Maren, T. H., Jankowska, L., Sanyal, G., and Edelhauser, H. F. (1983). The
transcorneal permeability of sulfonamide carbonic anhydrase inhibitors and
their eect on aqueous humor secretion. Exp. Eye Res., 36:457.
14. Cox, W. V., Kupferman, A., and Leibowitz, H. M. (1972). Topically applied
steroids in corneal disease. I. The role of inammation in stromal absorption
of dexamethasone. Arch. Ophthal., 88:308.
15. Cox, W. V., Kupferman, A., and Leibowitz, H. M. (1972). Topically applied
steroids in corneal disease. II. The role of drug vehicle in stromal absorption
of dexamethasone. Arch. Ophthal., 88:549.
16. Chien, D. S., Bundgaard, H., and Lee, V. H. L. (1988). Inuence of corneal
epithelial integrity on the penetration of timolol prodrugs. J. Ocular
Pharmacol., 4:137.
17. Madhu, C., Rix, P. J., Shackleton, M. J., Nguyen, T. G., and Tang-Liu, D. D.
S. (1996). Eect of benzalkonium chloride/EDTA on the ocular bioavailability of ketorolac tromethamine following ocular instillation to normal and deepithelialized corneas of rabbits. J. Pharm. Sci., 85:415.
18. Jani, R., Gan, O., Ali, Y., Rodstrom, R., Hancock, S. (1994). Ion exchange
resins for ophthalmic delivery. J. Ocular Pharmacol., 10:57.
19. Gurny, R., Ibrahim, H., Aebi, A., Buri, P., Wilson, C. G., Washington, N.,
Edman, P., and Camber, O. (1987). Design and evaluation of controlled
release systems for the eye. J. Controlled Release, 6:367.
20. Ahmed, I., and Patton, T. F. (1985). Importance of the noncorneal absorption
route in topical ophthalmic drug delivery. Invest. Ophthalmol. Vis. Sci.,
26:584587.
21. Ahmed, I., and Patton, T. F. (1987). Disposition of timolol and inulin in the
rabbit eye following corneal versus non-corneal absorption. Int. J. Pharmaceutics, 38:9.
22. Ahmed, I., Gokhale, R. D., Shah, M. V., and Patton, T. F. (1987).
Physicochemical determinants of drug diusion across the conjunctiva, sclera,
and cornea. J. Pharm. Sci., 76:583.
23. Chien, D. S., Homsy, J. J., Gluchowske, C., and Tang-Liu, D. D. S. (1990).
Corneal and conjunctival/scleral penetration of p-aminoclonidine, AGN
190342, and clonidine in rabbit eyes. Curr. Eye Res., 9:1051.
24. Schoenwald, R. D., Deshpande, G. S., Rethwisch, D. G., and Barfknecht, C.
F. (1997). Penetration into the anterior chamber via the conjunctival/scleral
pathway. J. Ocular Pharmacol. Ther., 13:41.
25. Romanelli, L., Valeri, P., Morrone, L. A., Pimpinella, G., Graziani, G., and
Tita, B. (1994). Ocular absorption and distribution of bendazac after topical
administration to rabbits with dierent vehicles. Life Sci., 54:877.

Copyright 2003 Marcel Dekker, Inc.

100

Chastain

26. Ling, T. L., and Combs, D. L. (1987). Ocular bioavailability and tissue distribution of [14 C]ketorolac tromethamine in rabbits. J. Pharm. Sci., 76:289.
27. Tang-Liu, D. D. S., Liu, S., Ne, J., and Sandri, R. (1987). Disposition of
levobunolol after an ophthalmic dose to rabbits. J. Pharm. Sci., 76:780.
28. Patton, T. F., and Robinson, J. R. (1976). Quantitative precorneal disposition
of topically applied pilocarpine nitrate in rabbit eyes. J. Pharm. Sci., 65:1295.
29. Chrai, S. S., Patton, T. F., Mehta, A., and Robinson, J. R. (1973). Lacrimal
and instilled uid dynamics in rabbit eyes. J. Pharm. Sci., 62:1112.
30a. Chrai, S. S., Makoid, M. C., Eriksen, S. P., and Robinson, J. R. (1974). Drop
size and initial dosing frequency problems of topically applied ophthalmic
drugs. J. Pharm. Sci., 63:333.
30b. Keister, J. C., Cooper, E. R., Missel, P. J., Lang, J. C., and Hager, D. F.
(1991). Limits on optimizing ocular drug delivery. J. Pharm. Sci., 80:50.
31. Zaki, I., Fitzgerald, P., Hardy, J. G., and Wilson, C. G. (1986). A comparison
of the eect of viscosity on the precorneal residence of solutions in rabbit and
man. J. Pharm. Pharmacol., 38:463.
32. Patton, T. F., and Robinson, J. R. (1975). Ocular evaluation of polyvinyl
alcohol vehicle in rabbits. J. Pharm. Sci., 64:1312.
33. Chrai, S. S., and Robinson, J. R. (1974). Ocular evaluation of methylcellulose
vehicle in albino rabbits. J. Pharm. Sci., 63:1218.
34. Edelhauser, H. F., and Maren, T. H. (1988). Permeability of human cornea
and sclera to sulfonamide carbonic anhydrase inhibitors. Arch. Ophthalmol,
106:1110.
35. Maurice, D. (1995). The eect of the low blink rate in rabbits of topical drug
penetration. J. Ocular Pharmacol. Therap., 11:297.
36. Doane, M. G., Jensen, A. D., and Dohlman, C. H. (1978). Penetration routes
of topically applied eye medications. Am. J. Ophthal., 85:383.
37. Kaila, T., Huupponen, R., and Salminen, L. (1986). Eects of eyelid closure
and nasolacrimal duct occlusion on the systemic absorption of ocular timolol
in human subjects. J. Ocular Pharmacol., 2:365.
38. Zimmerman, T. J., Kooner, K. S., Kandarakis, A. S., and Ziegler, L. P.
(1984). Improving the therapeutic index of topically applied ocular drugs.
Arch. Ophthalmol., 102:551.
39. Linden, C., and Alm, A. (1990). The eect of reduced tear drainage on corneal
and aqueous concentrations of topically applied uorescein. Acta Ophthal.,
68:633.
40. Lee, Y.-H., Kompella, U. B., and Lee, V. H.-L. (1993). Systemic absorption
pathways of topically applied beta adrenergic antagonists in the pigmented
rabbit. Exp. Eye Res., 57:341.
41. Hitoshe, S., Bundgaard, H., and Lee, V. H. L. (1989). Design of prodrugs to
selectively reduce systemic timolol absorption on the basis of the dierential
lipophilic characteristics of the cornea and the conjunctiva. Invest.
Ophthalmol. Vis. Sci., 30 (Suppl.):25.
42. Ashton, P., Podder, S. K., and Lee, V. H. L. (1991), V. H. L. (1991).
Formulation inuence on conjunctival penetration of four beta blockers in

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

43.

44.
45.

46.

47.

48.

49.

50.
51.

52.
53.
54.

55.
56.

57.

101

the pigmented rabbit: A comparison with corneal penetration. Pharm. Res.,

8:1166.
Schoenwald, R. D., and Chien, D. S. (1988). Ocular absorption and disposition of phenylephrine and phenylephrine oxazolidine. Biopharm. Drug
Dispos., 9:527.
Makoid, M. C., and Robinson, J. R. (1979). Pharmacokinetics of topically
applied pilocarpine in the albino rabbit eye. J. Pharm. Sci., 68:435.
Huang, H. S., Schoenwald, R. D., and Lach, J. L. (1983). Corneal penetration
behavior of -blocking agents III: In vitro in vivo correlations. J. Pharm.
Sci., 72:1279.
Tang-Liu, D. D. S., Liu, S. S., and Weinkam, R. J. (1984). Ocular and systemic bioavailability of ophthalmic urbiprofen. J. Pharmacokin. Biopharm.,
12:611.
Ding, S., Chen, C.-C., Salome-Kesslak, R., Tang-Liu, D. D.-S., and
Himmelstein, K. J. (1992). Precorneal sampling techniques for ophthalmic
gels. J. Ocular Pharmacol., 8:151.
Eller, M. G., Schoenwald, R. D., Dixson, J. A., Segarra, T., and Barfknecht,
C. F. (1985). Topical carbonic anhydrase inhibitors IV: Relationship between
excised corneal permeability and pharmacokinetic factors. J. Pharm. Sci.,
74:525.
Eller, M. G., Schoenwald, R. D., Dixson, J. A., Segarra, T., and Barfknecht,
C. F. (1985). Topical carbonic anhydrase inhibitors III: Optimization
model for corneal penetration of ethoxzolamide analogues. J. Pharm. Sci.,
74:155.
Conrad, J. M., and Robinson, J. R. (1977). Aqueous chamber drug distribution volume measurement in rabbits. J. Pharm. Sci., 66:219.
Rittenhouse, K. D., Peier, R. L., and Pollack, G. M. (1998). Evaluation of
microdialysis sampling of aqueous humor for in vivo models of ocular absorption and disposition. J. Pharm. Biomed. Anal., 16:951.
Lindquist, N. G. (1973). Accumulation of drugs on melanin. Acta Radiol.
325(Suppl.): 5.
Patil, P. N., and Jacobowitz, D. (1974). Unequal accumulation of adrenergic
drugs by pigmented and nonpigmented iris. Am. J. Ophthalmol., 78:470.
Araie, M., Takase, M., Sadai, Y., Ishii, Y., Yokoyama, Y., and Kitagawa, M.
(1982). Beta-adrenergic blockers: ocular penetration and binding to the uveal
pigment. Jpn. J. Ophthalmol., 26:248.
Lyons, J. S., and Krohn, D. L. (1973). Pilocarpine uptake by pigmented uveal
tissue. Am. J. Ophthalmol., 75:885.
Chien, D.-S, Richman, J., Zolezio, H., and Tang-Liu, D. (1992). Drug distribution of brimonidine in albino and pigmented rabbit eyes. Pharm Res.,
9:S336.
Achempoing, A. A., Schackleton, M., and Tang-Liu, D. D-S. (1995).
Comparative ocular pharmacokinetics of brimonidine after a single dose
application to the eyes of albino and pigmented rabbits. Drug Metab.
Disposition, 23(7):708.

Copyright 2003 Marcel Dekker, Inc.

102

Chastain

58. Nagata, A., Mishima, H. K., Kiuchi, Y., Hirota, A., Kurokawa, T., and
Ishibashi, S. (1993). Binding of antiglaucomatous drugs to synthetic melanin
and their hypotensive eects on pigmented and nonpigmented rabbit eyes.
Jpn. J. Ophthalmol., 37:32.
59. Geroski, D. H., and Edelhauser, H. F. (2000). Drug delivery for posterior
segment eye disease. Invest. Ophthalmol. Vis. Sci., 41(5):961.
60. Dahlin, D. C., Curtis, M. A., DeSantis, L., Struble, C. B. (2000). Distribution
of betaxolol to posterior ocular tissues of the cynomolgus monkey following a
30-day BID topical ocular regimen of Betoptic S. Invest. Ophthalmol. Vis. Sci.,
41:S509.
61. Romanelli, L., Morrone, L. A., Guglielmotte, A., Piccinelli, D., and Valeri, P.
(1992). Distribution of topically administered drugs to the posterior segment
of rabbit eye. Pharmacol. Res., 25(1):39.
62. Schoenwald, R. D., Harris, R. G., Turner, D., Knowles, W., and Chien, D. S.
(1987). Ophthalmic bioequivalence of steroid/antibiotic combination formulations. Biopharm. Drug Dispos., 8:527.
63. Tang-Liu, D. D. S., and Burke, P. J. (1988). The eect of azone on ocular
levobunolol absorption: Calculating the area under the curve and its standard
error using tissue sampling compartments. Pharm. Res., 5:238.
64. Lee, V. H. L. and Robinson, J. R. (1979). Mechanistic and quantitative evaluation of precorneal pilocarpine disposition in albino rabbits. J. Pharm. Sci.,
68:673.
65. Gibaldi, M., and Perrier, D. (1982). Pharmacokinetics, 2nd ed., Marcel
Dekker, New York, pp. 355384.
66. Himmelstein, K. J., Guvenir, I., and Patton, T. F. (1978). Preliminary pharmacokinetic model of pilocarpine uptake and distribution in the eye. J.
Pharm. Sci., 67:603.
67. Chiang, C. H., and Schoenwald, R. D. (1986). Ocular pharmacokinetic models of clonidine-3 H hydrochloride. J. Pharmacokin. Biopharm., 14:175.
68. Morlet, N., Graham, G. G., Gatus, B., McLachlan, A. J., Salonikas, C.,
Naidoo, D., Goldberg, I., and Lam, C. M. (2000). Pharmacokinetics of ciprooxacin in the human eye: A clinical study and population pharmacokinetic
analysis. Antimicrob. Agents Chemother., 44:1674.
69. Gillespie, W. R., Missel, P. J., and Lang, J. C. (1989). Application of system
analysis to ocular pharmacokinetics: Prediction of betaxolol concentrations in
ocular structures following various modes of administration. Pharm. Res.,
6:S224.
70. Sebag, J. (1992). The vitreous, Adlers Physiology of the Eye: Clinical
Applications (W. M. Hart, Jr., ed.), Mosby Year Book, St. Louis, pp. 305308.
71. Fatt, I. (1975). Flow and diusion in the vitreous body of the ey. Bull. Math.
Biol., 37:85.
72. Araie, M., and Maurice, D. M. (1991). The loss of uorescein, uorescein
glucuronide and uorescein isothiocyanate dextran from the vitreous by the
anterior and retinal pathways. Exp. Eye Res., 52:27.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

103

73. Graham, R. O., and Peyman, G. A. (1974). Intravitreal injection of dexamethasone. Arch. Ophthalmol., 92:149.
74. Friedrich, S., Cheng, Y.-L., and Saville, B. (1997). Drug distribution in the
vitreous humor of the human eye: The eects of intravitreal injection position
and volume. Curr. Eye Res., 16:663.
75. Friedrich, S., Saville, B., and Cheng, Y.-L. (1997). Drug distribution in the
vitreous humor of the human eye: the eects of aphakia and changes in retinal
permeability and vitreous diusivity. J. Ocular Pharmacol. Therap., 13:445.
76. Wingard, L. B., Zuravle, J. J., Doft, B. H., Berk, L., and Rinko, J. (1989).
Intraocular distribution of intravitreally administered amphotericin B in normal and vitrectomized dyes. Invest. Ophthalmol. Vis. Sci., 30(10)):2184.
77. Ben-Nun, J., Joyce, D. A., Cooper, R. L., Cringle, S. J., and Constable, I. J.
(1989). Pharmacokinetics of intravitreal injection. Assessment of a gentamicin
model by ocular dialysis. Invest. Ophthalmol. Vis. Sci., 30:1055.
78. Barza, M., Lynch, E., and Baum, J. L. (1993). Pharmacokinetics of newer
cephalosporins after subconjunctival and intravitreal injection in rabbits.
Arch. Ophthalmol., 111:121.
79. Mochizuki, M. (1980). Transport of indomethacin in the anterior uvea of the
albino rabbit. Jpn J. Ophthalmol, 24:363.
80. Yoshida, A., Ishiko, S., and Kojima, M. (1992). Outward permeability of the
blood-retinal barrier. Graefes Arch. Clin. Exp. Ophthalmol., 230:78.
81. Barza, M., Kane, A., and Baum, J. (1983). Pharmacokinetics of intravitreal
carbenicillin, cefazolin, and gentamicin in rhesus monkeys. Invest.
Ophthalmol. Vis. Sci., 24:1602.
82. Ohtori, A., and Tojo, K. (1994). In vivo/in vitro correlation of intravitreal
delivery of drugs with the help of computer simulations. Biol. Pharm. Bull.,
17:283.
83. Tojo, K. J., and Ohtori, A. (1994). Pharmacokinetic model of intravitreal drug
injection. Math. Biosci., 123:59.
84. Friedrich, S., Cheng, Y.-L., and Saville, B. (1997). Finite element modeling of
drug distribution in the vitreous humor of the rabbit eye. Ann. Biomed. Eng.,
25:303.
85. Gudauskas, G., Kumi, C., Dedhar, C., Bussanich, N., and Rootman, J.
(1985). Ocular pharmacokinetics of subconjunctivally versus intravenously
administered 6-mercaptopurine. Can. J. Ophthalmol., 20:110.
86. Rootman, J., Ostry, A., and Gudauskas, G. (1984). Pharmacokinetics and
metabolism of 5-uorouracil following subconjunctival versus itnravenous
administration. Can. J. Ophthalmol., 19:187.
87. McCartney, H. J., Drysdale, I. O., Gornall, A. G., and Basu, P. K. (1965). An
autoradiographic study of penetration of subconjunctivally injected hydrocortisone into the normal and inamed rabbit eye. Invest.Ophthalmol., 4:297.
88. Conrad, J. M., and Robinson, J. R. (1980). Mechanisms of anterior segment
absorption of pilocarpine following subconjunctival injection in albino rabbits. J. Pharm. Sci., 69:875.

Copyright 2003 Marcel Dekker, Inc.

104

Chastain

89. Maurice, D. M., and Polgar, J. (1977). Diusion across the sclera. Exp. Eye
Res., 25:577.
90. Freeman, W. R., Green, R. L., and Smith, R. E. (1987). Echographic localization of corticosteriods after periocular injection. Am. J. Ophthalmol., 103:281.
91. Bodker, F. S., Ticho, B. H., Feist, R. M., and Lam, T. T. (1993). Intraocular
dexamethasone penetration via subconjunctival or retrobulbar injections in
rabbits. Ophthalmic Surg., 24:453.
92. Hyndiuk, R. A., and Reagan, M. G. (1968). Radioactive depot-corticosteroid
penetration into monkey ocular tissue. I. Retrobulbar and systemic administration. Arch. Ophthalmol., 80:499.
93. Weijten, O., van der Sluijs, F. A., Schoemaker, R. C., Lentjes, E. G. W. M.,
Cohen, A. F., Romijn, F. P. H. T. M., and van Meurs, J. C. (1997). Peribulbar
corticosteroid injection: vitreal and serum concentrations after dexamethasone
disodium phosphate injection. Am. J. Ophthalmol., 123:358.
94. Ueno, N., Refojo, M. F., and Liu, L. H. S. (1982). Pharmacokinetics of the
antineoplastic agent 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) in the aqueous and vitreous of rabbit. Invest. Ophthalmol. Vis. Sci., 23:199.
95. Liu, Q. F., Dharia, N., Mayers, M., and Miller, M. (1995). Rifampin pharmacokinetics in serum and vitreous humor after single dose systemic administration. Invest. Ophthalmol. Vis. Sci., 36:S1018.
96. Wilson, T. W., Chan, H. S. L., Moselhy, G. M., Heydt, D. D., Frey, C. M.,
and Gallie, B. L. (1996). Penetration of chemotherapy into vitreous is
increased by cryotherapy and cyclosporine in rabbits. Arch. Ophthalmol.,
114:1390.
97. Elliot, P. J., Bartus, R. T., Mackic, J. B., and Zlokovic, B. V.(1997).
Intravenous infusion of RMP-7 increases ocular uptake of ganciclovir.
Pharm. Res., 14:80.
98. Palastine, A. G., and Brubaker, R. F. (1981). Pharmacokinetics of uorescein
in the vitreous. Invest.Ophthalmol. Vis. Sci., 21:542.
99. Lee, V. H. L., Chang, S. C., Oshiro, C. M., and Smith, R. E. (1985).Ocular
esterase composition in albino and pigmented rabbits: Possible implications in
ocular prodrug design and evaluation. Curr. Eye Res., 4:1117.
100. Lee, V. H. L. (1983). Esterase activities in adult rabbit eyes. J. Pharm. Sci,
72:239.
101. Bungaard, H., Falch, E., Larsen, C., Mosher, G. L., and Mikkelson, T. J.
(1986). Pilocarpine prodrugs II. Synthesis, stability, bioconversion and physicochemical properties of sequentially labile pilocarpine acid diesters. J. Pharm.
Sci., 75:775.
102. Chien, D. S., and Schoenwald, R. D. (1986). Improving the ocular absorption
of phenylephrine. Biopharm. Drug Dispos., 7:453.
103. Schoenwald, R. D., Folk, V. K., and Piper, J. G. (1987). In vivo comparison of
phenylephrine and phenylephrine oxazolidine instilled in the monkey eye. J.
Ocular Pharmacol., 3:333.
104. Chang, S-C., Bundgaard, H., Buur, A., and Lee, V. H. L. (1987). Improved
corneal penetration of timolol by prodrugs as a means to reduce systemic drug
load. Invest. Ophthalmol. Vis. Sci., 28:487.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

105

105. Narurkar, M. M., and Mitra, A. K. (1989). Prodrugs of 5-indo-20 -deoxyuridine for enhanced ocular transport. Pharm. Res., 6:887.
106. Camber, O., Edman, P., and Olsson, L. I. (1986). Permeability of prostaglandin F2
and prostaglandin F2
esters across cornea in vitro. Int. J. Pharm.,
29:259.
107. Plazonnet, B., Grove, J., Durr, M., Mazuel, C., Quint, M., and Rozier, A.
(1977). Pharmacokinetics and biopharmaceutical aspects of some anti-glaucoma drugs. In: Ophthalmic Drug Delivery Biopharmaceutical Technological
and Clinical Aspects. Vol. 11 (M. F. Saettone, M. Bucci, and P. Speiser,
eds.)., Liviana Press, Springer-Verlag, Berlin, pp. 118139.
108. Conners, M. S., Stolz, R. A., and Schwarzman, M. L. (1996). Chiral analysis of
12-hydroxyeicosatetraenoic acid formed by calf corneal epithelial microsomes.
J. Ocular Pharmacol. Therap., 12:19.
109. Shichi, H., Mahalak, S. M., Sakamoto, S., and Sugiyama. T. (1991).
Immunocytochemical study of phenobarbital- and 3-methylcholanthreneinducible cytochrome P450 isozymes in primary cultures of porcine ciliary
epithelium. Curr. Eye Res, 10:779.
110. Campbell, D. A., Schoenwald, R. D., Duel, M. W., and Barfknecht, C. F.
(1991). Characterization of arylamine acetyltransferase in the rabbit eye.
Invest. Ophthalmol. Vis. Sci., 32:2190.
111. Putman, M. L., Schoenwald, R. D., Duel, M. W., Barfknecht, C. F., Segarra,
T. M., and Campbell, D. A. (1987). Ocular disposition of aminozolamide in
the rabbit eye. Invest. Ophthalmol. Vis. Sci., 28:1373.
112. Wagner, J. G. (1975). Fundamentals of Clinical Pharmacokinetics. Drug
Intelligence Publications, Hamilton, IL, p. 42.
113. Rowland, M., and Tozer, T. N. (1989). Clinical Pharmacokinetics, Lea and
Febiger, Philadelphia, p. 3.
114. Chien, D. S., and Schoenwald, R. D. (1990). Ocular pharmacokinetics and
pharmacodynamics of phenylephrine and phenylephrine oxazolidine in rabbit
eyes. Pharm. Res., 7:476.
115. Smolen, V. F. (1981). Noninvasive pharmacodynamic and bioelectrometric
methods for elucidating the bioavailability mechanisms of ophthalmic drug
preparations. Prog. Drug Res., 25:421.
116. Smolen, V. F., and Schoenwald, R. D. (1971). Drug-absorption analysis from
pharmacological data I: Method and conrmation exemplied for the mydriatic drug tropicamide. J. Pharm. Sci., 60:96.
117. Schoenwald, R. D., and Smolen, V. F. (1971). Drug-absorption analysis from
pharmacological data II: Transcorneal biophasic availability of tropicamide. J.
Pharm. Sci., 60:1039.
118. Smolen, V. F. (1971). Quantitative determination of drug bioavailability and
biokinetic behavior from pharmacological data for ophthalmic and oral
administrations of a mydriatic drug. J. Pharm. Sci., 60:354.
119. Smolen, V. F., and Schoenwald, R. D. (1974). Drug absorption analysis from
pharmacological data III: Inuence of polymers and pH on transcorneal biophasic availabiilty and mydriatic response of tropicamide. J. Pharm. Sci.,
63:1582.

Copyright 2003 Marcel Dekker, Inc.

106

Chastain

120. Brazzel, R. K., Wooldridge, C. B., Hackett, R. B.,and McCue, B. A. (1990).

Pharmacokinetics of the aldose reductase inhibitor imirestat following topical
ocular adminsitration. Pharm. Res., 7:192.
121. Rao, C. S. (1991). Physicochemical and Biopharmaceutical Evaluation of
Ibufenac, Ibuprofen and Their Hydroxyethoxy Analogs in the Rabbit Eye.
Ph.D. thesis, University of Iowa College of Pharmacy, Iowa City.
122. Andermann, G., Guggenbuhl, P., de Burlet, G., and Himber, J. (1989).
Pharmacokinetics of falintolol II. Absorption, distribution and elimination
from tissues and organs following ocular administration and intravenous
injection of falintolol in albino rabbits. Meth. Find. Exp. Clin. Pharmacol.,
11:747.
123. Kleinberg, J., Dea, F. J., Anderson, J. A., and Leopold, I. H. (1979).
Intraocular penetration of topically applied lincomycin hydrochloride in rabbits. Arch. Ophthalmol., 97:933.
124. Hussain, A., Hirai, S., and Sieg, J. (1980). Ocular absorption of propranolol in
rabbits. J. Pharm. Sci., 69:738.
125. Sugrue, M. F., Gautheron, P., Mallorga, P., Nolan, T. E., Graham, S. L.,
Schwam, H., Shepard, K. L., and Smith, R. L. (1990). L-662,583 is a topically
eective ocular hypotensive carbonic anhydrase inhibitor in experimental animals. Br. J. Pharmacol., 99:59.
126. Vuori, M.-L., and Kaila, T. (1995). Plasma kinetics and antagonist activity of
topical ocular timolol in elderly patients. Graefs Arch. Clin. Exp. Ophthalmol.,
233:131.
127. Valeri, P., Palmery, M., Severini, G., and Piccinelli, D. (1986). Ocular pharmacokinetics of dapiprazole. Pharmacol. Res. Com., 18:1093.
128. Taylor, P. B., Burd, E. M., and Tabbara, K. F. (1987). Corneal and intraocular penetration of topical and subconjunctival fusidic acid. Br. J.
Ophthalmol., 71:598.
129. Leibowitz, H. M., Ryan, W. J., Kupferman, A., and DeSantis, L. (1986).
Bioavailability and corneal anti-inammatory eect of topical suprofen.
Invest. Ophthalmol. Vis. Sci., 27:628.
130. Hui, H. W., Zeleznick, L., and Robinson, J. R. (1984). Ocular disposition of
topically applied histamine, cimetidine, and pyrilamine in the albino rabbit.
Curr. Eye Res., 3:321.
131. Sato, H., Fukuda, S., Inatomi, M., Koide, R., Uchida, N., Kanda, Y., Kiuchi,
Y., and Ougchi, K. (1996). Pharmacokinetics of noroxacin and lomeoxacin
in domestic rabbit aqueous humour analyzed by microdialysis. J. Jpn.
Ophthalmol. Soc., 100:513.
132. Maren, T. H., and Jankowska, L. (1985). Ocular pharmacology of sulfonamides: The cornea as barrier and depot. Curr. Eye Res., 4:399.
133. Grove, J., Gautheron, P., Plazonnet, B., and Sugrue, M. F. (1988). Ocular
distribution studies of the topical carbonic anhydrase inhibitors L-643,799
and L-650,719 and related alkyl prodrugs. J. Ocular Pharmacol., 4:279.

Copyright 2003 Marcel Dekker, Inc.

General Considerations in Ocular Drug Delivery

107

134. Anderson, J. A. Chen, C. C., Vita, J. B., and Shackleton, M. (1982).

Disposition of topical urbiprofen in normal and aphakic rabbit eyes. Arch.
Ophthalmol., 100:642.
135. Huupponen, R., Kaila, T., Salminen, L., and Urtti, A. (1987). The pharmacokinetics of ocularly applied timolol in rabbits. Acta Ophthalmol., 65:63.
136. Mester, U., Krasemann, C., and Werner, H. (1982). Cefsulodin concentrations in rabbit eyes after intravenous and subconjunctival administration.
Ophthalmic Res., 14:129.
137. Vigo, J. F., Rafart, J., Concheiro, A., Martinez, R., and Cordido, M. (1988).
Ocular penetration and pharmacokinetics of cefotaxime: An experimental
study. Curr. Eye Res., 7:1149.
138. Miller, S. C., Himmelstein, K. J., and Patton, T. F. (1981). A physiologically
based pharmacokinetic model for the intraocular distribution of pilocarpine in
rabbits. J. Pharmacokin. Biopharm., 9:653.
139. Leeds, J. M., Henry, S.P., Truong, L., Zutshi, A., Levin, A. A., and
Kornbrust, D. (1997). Pharmacokinetics of a potential human cytomegalovirus therapeutic, a phosphorothioate oligonucleotide, after intravitreal injection in the rabbit. Drug Metab. Dispos., 25:921.
140. Berthe, P., Baudouin, C., Garrao, R., Hofmann, P., Taburet, A.-M., and
Lapalus, P. (1994). Toxicologic and pharmacokinetic analysis of itnravitreal
injections of foscarnet, either alone or in combination with ganciclovir. Invest.
Ophthalmol. Vis. Sci., 35:1038.
141. Robertson, J. E., Westra, I., Woltering, E. A., Winthrop, K. L., Barrie, R.,
ODorisio, T. M., and Holmes, D. (1997). Intravitreal injection of octreotide
acetate. J. Ocular Pharmacol. Therap., 13:171.
142. Coco, R. M., Lopez, M.I., Pastor, J. C., Vallelado, A. I., Nozal, M. J., and
Pampliega, A. (1995). Intravitreal pharmacokinetics of 0.5 and 1 mg of vancomycin in endophthalmic rabbit eyes. Invest. Ophthalmol. Vis. Sci., 36:S613.
143. Barza, M., Kane, A., and Baum, J. (1982). The eects of infection and probenecid on the transport of carbenicillin from the rabbit vitreous humor.
Invest. Ophthalmol. Vis. Sci., 22:720.
144. Kwak, H. W. and DAmico, D. J. (1992). Evaluation of the retinal toxicity
and pharmacokinetics of dexamethasone after intravitreal injection. Arch.
Ophthalmol., 110:259.
145. Ashton, P., Brown, J. D., Pearson, P. A., Blandford, D. L., Smith, T. J.,
Anand, R.,Nightingale, S. D., and Sanborn, G. E. (1992). Intravitreal ganciclovir pharmacokinetics in rabbits and man. J. Ocular Pharmacol., 8:343.

Copyright 2003 Marcel Dekker, Inc.

4
Ocular Drug Transfer Following
Systemic Drug Administration
Nelson L. Jumbe
Albany Medical College, Albany, New York, and Amgen Inc., Thousand
Oaks, California, U.S.A.

Michael H. Miller
Albany Medical College, Albany, New York, U.S.A.

I.

INTRODUCTION

A discussion of transport models following systemic, intraocular, as well as

conventional eyedrop drug administration is important since ocular pharmaceuticals are often administered by more than one route. The primary
advantages of direct topical drug administration include targeted drug delivery to the anterior segment of the eye and avoidance of systemic drug
toxicity. Many of the disadvantages of eyedrops, such as limited penetration
into vitreous and retina, compliance, and the advantages of the use of drug
delivery systems, are discussed in other sections of this text.
Intravenous and oral therapy are used in the treatment of generalized
systemic disease with ocular involvement (e.g., diabetes mellitus, inammatory diseases, autoimmune diseases, and sarcoidosis), malignancies such as
lymphomas, or when penetration following topical drug administration does
not attain therapeutic levels at the site of disease (e.g., vitreous, choroid, or
retina). Specic uses of systemic therapy for ocular diseases include photosensitization therapy for the treatment of choroidal neovascularization,
cytotoxic agents for the treatment of autoimmune diseases, nonsteroidals
and steroids for the treatment of uveitis and scleritis, acetazolamide for
macula edema, and antivirals or antibiotics for the prophylaxis or therapy
of keratitis, chorioretinitis, and endophthalmitis. Additional applications of
systemic therapy include thalidomide for the treatment of proliferative dis109

Copyright 2003 Marcel Dekker, Inc.

110

Jumbe and Miller

eases, antioxidants for protection against macular degeneration, and the use
of neuroprotective agents in patients with glaucoma.
This chapter primarily deals with the intercompartmental drug translocation (entry and eux) of antimicrobial agents in the posterior eye.
Antimicrobial agents were chosen as the paradigm for systemically administered drugs because the principles governing the intercompartmental drug
transfer of antimicrobials are similar for most pharmaceutical agents.
Moreover, new methods with enhanced discriminative capabilities used to
characterize inter-compartmental drug transfer have primarily characterized
the ocular pharmacokinetics and pharmacodynamics of antibiotics, antifungals, and antiviral agents (17,18,52,54,72,73,75).
Systemic, topical, and intraocular administration of antimicrobial
agents are all used in the therapy of infectious diseases of the eye. When
compared to topical drug administration, the systemic toxicity of antimicrobial agents may be less of an issue since they generally do not alter
physiological functions of other organ systems or cause dose-related systemic toxicity. Systemic therapy is used to treat cytomegalovirus (CMV)retinitis in patients with immunosuppressive diseases such as acquired
immunodeciency syndrome (AIDS) or those on immunosupressive medications such as organ or bone marrow transplant patients. While AIDS
patients with CMV-chorioretinitis are often treated with antivirals administered via implanted, long-term drug delivery devices, supplemental systemic
therapy is also used to protect the contralateral eye. Recurrent herpes simplex virus (HSV) keratitis can be prevented using oral acyclovir. Finally,
while the preferred treatment of deep-seated eye infections such as bacterial
and fungalendophthalmitis is direct intraocular (IO) drug administration,
the role of systemic therapy remains unclear. In the National Eye Institute
(NEI) Endophthalmitis Vitrectomy Study (EVS), systemically administered
antibiotics did not improve the outcome in patients with postsurgical, bacterial endophthalmitis (5). However, the drugs used in this trial exhibited
poor penetration into noninamed eyes. As a result, the potential role of
adjuvant therapy with systematically administered antimicrobials that show
better penetration into the vitreous humor still needs to be addressed. In
fact, adjuvant therapy with systemically administered quinolones has been
used successfully for the treatment of bacterial endophthalmitis (14,19,58).
While vitrectomy with intravitreal drug administration is the preferred
mode of therapy for endophalmitis (23), signicant morbidity from this
infection persists. Moreover, intraocular drug administration for other ocular infections (e.g., CMV-chorioretinitis) is often associated with serious
untoward eects, including endophthalmitis, vitreous hemorrhage, retina
detachment, retinal toxicity, cataract formation, and, in the absence of systemic therapy, infection of the contralateral eye. Furthermore, the incidence

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

111

of opportunistic ocular infections has increased with use of potent immunosuppressive agents in patients with organ transplants, cancer, and autoimmune diseases. While highly active antiretroviral therapy (HAART) for
AIDS has decreased the incidence of infections of the eye, viral and fungal
infections continue to be seen in patients with HIV diseases. This, in conjunction with the use of more potent immunosuppressive agents in selected
patients such as transplant patients, suggest that the use of potent and less
toxic systemically administered antimicrobial drugs as adjuncts or alternative therapy of ocular disease has merit in the therapy of deep ocular infections.
A considerable amount of work demonstrating relatedness between
drug-specic pharmacodynamic parameters to clinical outcome has been
published over the last several years (10,27,30,37,38,50,87). Normally, optimization of the plasma concentration-versus-time prole translates into a
similar concentration-versus-time prole for an infection site. However,
drugs administered systemically often have poor access to the inside of
the eye because of the blood-aqueous and blood-retinal barriers. Thus,
ophthalmic drug discovery and intervention therapy development must
also deal with the challenge of achieving eective concentrations of these
drugs within the eye. The primary focus of this chapter is to review important principles of ocular pharmacokinetics of antimicrobial agents following
intravitreal and systemic drug administration and to discuss available strategies for developing eective ocular drug therapies using systemically administered drugs.

II.

OVERVIEW OF BLOOD-OCULAR BARRIER

TRANSPORT BIOLOGY

The parenteral and oral routes are the principal routes for the systemic
delivery of drugs. The most commonly used parenteral routes are intravenous (IV), intramuscular (IM), subcutaneous (SC), and intradermal (ID).
The distribution of the drug throughout the body depends on the rates of
absorption, distribution, metabolism and excretion from the blood as well
as the extent of binding to plasma proteins. The major advantages of intravenous drug administration are rapid and complete absorption and avoidance of rst-pass metabolism. However, unlike other parts of the body,
specialized barriers regulate drug distribution from the blood into protected
compartments like the prostage, eye, and brain. The eye, prostage, and brain
are privileged sites in which the concentration-versus-time prole may dier
substantially from that observed for the plasma (5). Consequently, under-

Copyright 2003 Marcel Dekker, Inc.

112

Jumbe and Miller

standing of the concentration-versus-time prole of drugs in these sites is of

crucial importance.
The blood-brain barrier (BBB), bloodcerebrospinal uid barrier
(BCSFB),and blood-ocular barrier (BOB) share similarities in microscopic
structure and function (47,22,62). Structurally, the barriers consist of tight
endothelial junctions. Functionally, the barriers regulate transfer of sugars,
amino acids, organic acids, and ions according to molecular size, proteinbinding anity, lipophilicity, and degree of ionization at the relevant anatomical compartment pH (40,53,59,6668,80,97). Furthermore, active transport systems and enzymatic degradation contribute to this barrier and
regulate the eective penetration of a variety of chemotherapeutic compounds (9,22,53,86). Recent studies have shown that the pharmacokinetics
of several antimicrobial agents are similar in both the eye and cerebrospinal
uid (CSF) following systemic drug administration (48,53,61). This observation may have practical as well as theoretical implications since, in the
absence of data for site-specic pharmacokinetics, one site may serve as a
pharmacokinetic surrogate for the other (48,53,61).
The primary blood supply to the eye is the external ophthalmic artery,
a branch of the internal carotid artery. However, because of selective permeability and the presence of tight cellular junctions between the
endothelial cells lining capillaries, great limitations are placed on which
blood components actually reach the intraocular tissue space. The bloodocular barrier is comprised of three barriers: the corneal epithelial barrier
(tear-corneal stroma barrier), retinal capillaries, and the retinal pigmented
epithelium (RPE, or blood-retinal barrier). The blood-ocular barrier selectively allows some materials to traverse it, while preventing others from
crossing. This anatomical arrangement shields delicate ocular tissue from
biochemical uctuations and toxins in the bloodstream. Although these
barriers function as protective mechanisms, they also greatly aect penetration of therapeutic agents into the eye.
Systemically administered drugs gain access into the eye anterior and
posterior chambers principally by crossing the blood-aqueous barrier via
capillaries perfusing the iris and ciliary body. Once in the posterior chamber, the drugs can diuse through the lens-iris barrier into the anterior
portion of the vitreous body. This barrier is important for two reasons: it
can act as an anterior-posterior barrier against infective bacteria, and it
imposes limited diusion of drugs from the anterior chamber to the vitreous humor. The blood-aqueous barrier is composed of the ciliary body,
which is posterior to this iris and is far more permissive than the posterior
portion of the eye. The iris and ciliary body have fenestrated capillaries,
whereas the pigment epithelium and endothelial cell of the retinal
capillaries have tight intercellular junctions. These barriers separate

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

113

drugs equilibrated within the extravascular space of the choroids from

entry into the retina and vitreous body.
Drug translocation rates are dependent on the functional anatomy and
physiology of the eye as well as pharmacokinetics in the serum. Ocular drug
transfer may occur by passive diusion and active transport. Ultimately, the
concentrations of drug achieved in the posterior eye depend upon competing
rate constants describing uptake and eux. Studies showing that the renal
and ocular eliminations of uorescein (20) quinolones, and -lactam antibiotics in humans and rabbits are blocked by probenecid (52,70,71) and
inammation (4,49,52,85) provide evidence that these barriers have active
transport systems that aect drug concentrations in the eye. In vivo and in
vitro studies characterizing the rates of renal elimination of zwitterionic
quinolones suggest the presence of separate and distinct carrier-mediated
systems (44,52). Studies from our laboratory have shown that the elimination rates of four quinolones examined following direct intravitreal injection
were prolonged by both probenecid and heat-killed bacteria (52).
Unpublished data from our laboratory demonstrated that similarities in
drug penetration into the CSF and eye can be explained by the presence
of eux pumps lining these anatomical sites that are targets for probenecid.
There is a sidedness to these pumps, which are dierentially aected by
systemic and intraocular probenecid administration (4,52,62,84,85,102).
Furthermore, intracerbroventricular and intravitreal but not systemically
administered probenecid block the eux of quinolone antimicrobials.

III.

ALTERATIONS OF THE BLOOD-OCULAR BARRIER IN

DISEASED STATE

The blood-ocular barrier shares similar embryological origin, microanatomy, and many physiological functions with the blood-brain barrier.
There are many natural (e.g., diabetes, hypertension) or iatrogenic (chemotherapy, retinal photocoagulation) conditions that cause blood-ocular
barrier breakdown. Disruption of the tight junctions between the endothelium of the retinal blood vessels (inner blood-retinal barrier) and the tight
junctions between adjacent RPE cells (outer blood-retinal barrier) results in
breakdown of the BOB and subsequent changes in drug ocular penetration.
Infectious and noninfectious (uveitis and surgery) causes of ocular
inammation represent one category of retinal vascular disorders causing
BOB breakdown. Fungi, viruses, and bacteria can be very destructive when
they infect the eye. Candida endophthalmitis occurs in 530% of patients
with disseminated Candida infections (15). Bacterial endophthalmitis is a
severe and often blinding infection of the eye (26,69). Noninfectious inam-

Copyright 2003 Marcel Dekker, Inc.

114

Jumbe and Miller

mation is associated with diseases of immune regulation. Models for noninfectious uveitis include intravitreal antigen injection or heat-killed bacteria
(7,33,34,52). Bacterial lipopolysaccharide induces endotoxin-induced uveitis
in rabbits, mice, and rats and is a useful tool for investigating ocular inammation due to immunopathogenic, rather than autoimmune processes
(39,94,95).
The integrity of the blood-retinal barrier can be demonstrated both
experimentally and clinically by intravenous injection of tracer molecules
normally excluded from the retina by the healthy blood-retinal barrier.
Horseradish peroxidase (99) and carboxyuoroscein (20) are the most commonly used tracers). Disruption of the inner or outer blood-retinal barrier is
demonstrated by passage of these tracers either between pigment epithelial
cells or of retinal blood vessels. Fluorescein isothiocyanate linked to high
molecular weight dextrans of varying size can also be used experimentally to
evaluate the signicance of the molecular weight on dierential passage
through the disrupted blood-retinal barrier in dierent pathological conditions (8).
Chemical, mechanical, inammatory, and infective insults modulate
the penetration of drugs into the eye following systemic administration.
This is particularly important, especially when the blood-ocular barrier is
suciently insulted and reduces the integrity of the barrier against normally
nonpermeable drugs. As a result, drugs that are normally excluded from the
eye may penetrate into the eye due to degenerative eects on junctional
barriers. For example, for hydrophilic antimicrobials such as the ciprooxacin, infection may increase their mean vitreous concentration by more than
vefold (52).
The role of inammation in drug penetration is particularly informative. Inammation reduces the elimination rates of intravitreally injected
drugs that are not actively exported by the posterior route. For systemically
administered quinolones and other antimicrobials, inammation increases
ocular drug accumulation (i.e., penetration) most likely due to increased
entry and decreased eux rates (45,52).

IV.

PHYSICOCHEMICAL PROPERTIES GOVERNING DRUG

PENETRATION INTO THE EYE

Carrier-independent penetration of antibiotics and other drugs into the eye

(11,16,23,63,81), like that at other anatomical sites (16,23,53,63,66), is
related to the physicochemical properties of the compound, which include
lipophilicity and molecular weight/size and protein binding. The eects of
these independent variables on drug penetration into bacteria (96) and into

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

115

dierent anatomical compartments have been investigated by several

laboratories (35,40,52,66,80,97). Multiple linear regression analysis shows
that the logarithm of the penetration of compounds across planar lipid
bilayers and tissue membranes correlates with the oil/water partition
ratio, the inverse of the square root of the molecular weight (8) and the
free fraction of drug (24,53,63). Almost all studies show that the oil/water
partition ratio (lipophilicity) is generally the most predictive variable of drug
transport into and out of privileged anatomical spaces (Fig. 1).
In relation to ocular drug penetration following systemic administration, only the unbound fraction of drug is in diusional exchange across the
blood-ocular barriers. In principle, protein binding can and should be
included in dose determination. Thus, true comparison of drug penetration
should be on the basis of the unbound fraction in the plasma, so that
binding does not enter as a factor into the ocular kinetics. However, as

Figure 1 Relationship between the quinolone partition coecients and levels of

drug penetration into the vitreous humor following systemic drug administration,
and elimination rate half-lives following direct intravitreal injection. The only statistically signicant (when examined univariately; multiple linear regression did not
improve model ts) quinolone physicochemical property related to ocular drug
translocation was lipophilicity. The partition coecients for direct and systemic
drug administration are shown on the top and bottom abscissas, respectively.

Copyright 2003 Marcel Dekker, Inc.

116

Jumbe and Miller

already noted, many factors inuence drug entry into protected compartments, and ceftriaxone, which is >95% protein bound, penetrates into the
CSF better than other, less protein-bound cephalosporins. Since intracameral penetration is inversely related to molecular weight/size, drugs with a
large molecular weight are likely to be excluded from aqueous and vitreous
humors after systemic administration, similar to blood proteins and macromolecules. However, compounds may cross the blood-ocular barriers in
spite of their large molecular size, and the rate of penetration is correlated
with their lipid solubility, as expressed by their oil/water partition ratio.
Unfortunately, the ability to further delineate the relationship between lipophility and molecular size/weight is curtailed due to the limit molecular
weight/size within families of chemotherapeutic compounds.
Independent studies (52,66) have demonstrated the correlation
between lipophilicity and penetration into or elimination from the eye following systemic (17,45,52,73,74) or intravitreal injection (3335,65) in nonpigmented, uninfected rabbits (42,78). A similar relationship between
lipophilicity, penetration, and eux occurs in prokaryotes [96].

V. METHODS FOR IN VIVO ANTRAOCULAR

PHARMACOKINETIC MEASUREMENTS
The study of the pharmacokinetics of drugs that penetrate into eye following systemic administration requires assaying ocular specimens obtained at
dierent time points. In humans these samples are generally obtained by
needle aspiration, often in patients with diseases that alter the BOB. The risk
of iatrogenic complications when perforating the globe with a needle has
restricted human ocular pharmacokinetic studies to a single sample obtained
from patients at the time of intraocular surgery (3). In early studies, the
concern that serial sampling of individual eyes would introduce artifact also
limited animal studies to only one specimen per animal. As a result, single
samples from dierent individuals are generally pooled to generate singlesubject estimates (1,51,59). This technique is referred to as the naive pooled
data (NPD) approach. A validated animal model from our laboratory
involves sequential sampling of the vitreous humor and CSF in individual
animals to obtain complete concentration-versus-time proles (52,55,60
62). In addition, several laboratories have used microdialysis to investigate
drug kinetics in the aqueous and vitreous humors (25,32,41,42,45,46). Both
methods are useful, but it is important to recognize several potential methodological problems that are inherent in both in vivo methods for collecting
ocular drug kinetic data in animals and humans (29,43,60). This section

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

117

evaluates these three models and maps out future directions and developments in the art of in vivo intraocular pharmacokinetics measurements.
A.

Nave Pooled Data

Drug penetration into the eye can be estimated by comparing the means of
serum and ocular uid using the mean values of multisubject data sets
obtained at sequential time points or at a single sampling time (43,89,90).
Determining penetration by either method is imprecise. Two-compartment
models best describe ocular kinetics of most systemically administered
drugs. The ratios of discrete single pairs of ocular-serum drug concentrations change with time until the subject is well into pseudo-distribution
equilibrium. Consequently, obtaining mean penetration ratios from these
discrete data will give erroneous estimates of the ability of the drug to
penetrate into a specic site, because these ratios would depend largely on
the time of collection of the samples. Thus, systemic drug administration is
quite dierent from direct drug administration.
Pharmacokinetics based upon samples obtained sequentially throughout the steady-state dosing interval for a large number of patients should
result in reliable pharmacokinetic estimates. Besides diculties in obtaining
serial samples in humans, associated ocular direct-aspiration sampling techniques also include (a) the eects of sampling on BOB and (b) awareness
that vitreal aspirates are almost never obtained in the absence of ocular
pathology, often with diseases that also alter the BOB. Consequently, comparison of data sets obtained at single time points is also subject to system
hysteresis, intersubject variation, and, most noteworthy, parameter estimations generated from fragmentary data. The naive pooled data method prevents the characterization of dierences within a population. Therefore, (a)
intersubject variation prohibits rigorous determination of pharmacokinetic
parameters from small groups of animals (or patients), and (b) parameter
estimates from this approach discard the general population information,
which reduces the value of the information obtained from these studies
(29,60). One way to improve the quality of the data obtained from these
models is to study serial drug concentrations in dierent subjects. Analysis
of these data can be evaluated using population models for clinical decision
making since exposure-penetration relationships can then be extrapolated to
make treatment decisions based on achievable serum steady-state levels.
B.

Direct Aspiration

Previous studies have shown that even nontraumatic aspiration of the aqueous humor (a) is associated with BOB breakdown, (b) increased penetra-

Copyright 2003 Marcel Dekker, Inc.

118

Jumbe and Miller

tion of drugs, and/or (c) decreased aqueous humor ow-dependent drug

elimination. In contrast, our laboratory has developed and validated a
model that permits repeated vitreous humor paracentesis from the same
animal but does not alter the parameters listed above or ocular drug kinetics
(60,62). Complete concentration-versus-time curves over the observation
period can be obtained using this method. Serial sampling did not alter
the pharmacokinetics or terminal elimination rates following direct injection
into the vitreous humor (52). Furthermore, serial paracentesis does not alter
the BOB or the penetration of the drug into the vitreous from the serum in
the majority of the animals (52). BOB breakdown (likely due to trauma to
retinal vessels) is easily recognized by measuring albumin concentrations in
the vitreous, permitting exclusion of these samples from analysis. The
absence of a measurable eect of serial paracentesis on drug pharmacokinetics observed using this method is likely related to both the absence of
inammation and BOB breakdown and the small volume of vitreous humor
removed.
Analysis of population pharmacokinetic data using complete concentration-versus-time data following serial aspiration in animals that do not
have BOB breakdown is not only theoretically preferable to pooling of data
from dierent subjects (29,56,60,78), but also provides robust kinetic information obtained from a small number of animals. Since pharmacokinetic
data are important in the design of clinical trials, this model has fundamental clinical implications for all classes of ocular pharmaceuticals.
Furthermore, this model permits the evaluation of the eects of inammation or surgery on antibiotic penetration after drug elimination following
intravitreal injection (52). We have successfully used this model to study (a)
similarities between blood-aqueous and blood-cerebrospinal uid barriers
(4,53,61), (b) the relationship between physicochemical properties and ocular drug penetration (52), and (c) determination of mechanisms of active
drug transport from the eye and CSF (45). Recent studies characterizing
ocular pharmacokinetics of systemically administered ciprooxacin following a single dose, multiple doses, and continuous infusion to rabbits have
shown that microbiological outcome is dependent upon the mean rather
than the peak concentrations in the serum (56).
C.

Microdialysis

The method of microdialysis provides an important alternative for obtaining intra-ocular pharmacokinetic data (10,32,42,47,100). Complete concentration-versus-time proles can be obtained in individual animals by
perfusing a ne cannula with normal saline at a xed ow rate. Drug in
the surrounding tissue (solute) diuses down its concentration gradient into

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

119

the perfusate and is collected for analysis. Details of this method are covered
extensively elsewhere in this text.
Microdialysis has proven to be the gold standard of measuring ocular
pharmacokinetics following direct intraocular drug injection when drug
concentrations are high. Experimental conditions following systemic drug
administration are quite dierent. First, high drug concentrations are not
achieved in the eye. Second, and more importantly, the kinetic prole
following systemic drug administration follows a rapid distribution phase
as drug enters the eye, followed by multiexponential elimination. Therefore,
ocular kinetics following systemically administered drug are transient and
initially rapidly changing, but change more slowly when the drug is well into
pseudo-distribution equilibrium during the elimination phase. This has
direct bearing on the methods used for determining probe extraction coecients. Finally, the lower drug concentrations seen in conjunction with
small volumes of dialysate may make the measurement of drug concentrations using conventional HPLC or microbiological assays dicult.
Microdialysis probe calibration is aected by the cannula deadspace,
temporal resolution insensitivity, substrate physicochemical properties, and
potential BOB breakdown from an indwelling probe. For example, early
studies characterizing intravitreal concentrations of aminoglycosides
showed time-dependent dierences in pharmacokinetic parameters suggesting progressive BOB breakdown. BOB breakdown can be minimized by
careful probe placement and/or by allowing the eye to heal completely
after probe placement (42). Since BOB is invariable with serial aspirations
of the aqueous humor, microdialysis via previously implanted probes in
quiet eyes is clearly preferable when measuring drug concentrations in the
aqueous humor (54,82,83). Both the zero-net ux and retrodialysis methods
are used for the determination of probe recovery. The zero-net ux recovery
is a steady-state measurement determined at xed solute concentrations
(32,101). Retrodialysis recovery is determined from loss of a drug analog
in the solute from the perfusate (32,101). The retrodialysis recovery determination method is most appropriate for in vivo pharmacokinetic studies.
However, potential limitations exist in the availability of drug-analog pairs.
Microdialysis temporal resolution issues have been greatly reduced by the
ability to interface this technique directly with high-performance liquid
chromatography mass spectroscopy (HPLC-MS) for sample analysis.
This is possible because collected samples are clean and do not require
any preparation before analysis. Moreover, assays with increased sensitivity
greatly reduce the sampling time and sample volumes, allow for continuous
sampling, and permit faster/slower ow rates, which in turn erase system
hysteresis if cannula deadspace noise is removed by integration. Since microdialysis sampling gives an integrated (mean concentration) response over the

Copyright 2003 Marcel Dekker, Inc.

120

Jumbe and Miller

sampling period, the drug area under the concentration-versus-time curve

(AUC) is obtained by pooling the measured concentrations from each sample interval and summing them together (47).
While direct vitreous humor sampling requires paracentesis, a repeated
invasive procedure not generally possible in humans, microdialysis provides
inroads for ocular pharmacokinetic studies using indwelling probes.
Furthermore, our laboratory has shown that the CSF and eye are pharmacokinetic surrogates for systemically administered drugs. Thus, microdialysis measurements in the CSF may provide a practical way to obtain
ocular pharmacokinetic data where none is available, and vice versa.

VI.

PHARMACOKINETICS AND PHARMACODYNAMICS OF

SYSTEMIC DRUG ADMINISTRATION

A.

Compartmental Model Analysis

The pharmacokinetic processes of drug absorption, distribution, metabolism, and elimination determine the concentration-versus-time prole in
a target organ. Ocular pharmacokinetic models for intraperitoneal (IP),
intravenous (IV), and intraocular drug administration are described by
three-, two-, or one-compartment models, respectively. These models are
not representative of real tissue spaces. Physiologically based models
using mass balances and blood-ow limited elimination are least useful
for ocular pharmacokinetics because of the nonidentiable nature of drug
transfer by ow, blood diusional exchange, and anterior-posterior
drainage. Compartmental open models are depicted with elimination from
the central compartment.
Drugs that distribute rapidly and homogeneously into the plasma and
well-perfused, extracellular uid follow a one-compartment model (and
often show rst-order kinetics). Drug elimination from the eye generally
follows rst-order kinetics. One-compartment models best describe monoexponential (log-linear) elimination kinetics (Fig. 2). Two-compartment
kinetic models demonstrate identiable distribution and elimination phases
where competing elimination rates govern both the distribution and elimination phases (Fig. 3).
Analysis of concentration-versus-time data optimally co-models
plasma samples and ocular samples simultaneously for pharmacokinetic
parameter estimation (16). Pharmacokinetic analyses of these data provide
apparent volumes of distribution and areas under the concentration-versustime curve of the serum and peripheral compartments as part of the pharmacokinetic parameter identication procedure. AUCs are determined by
numerical integration (or trapezoidal approximation). The calculated

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

121

Figure 2 On-compartment open model. The gure depicted shows the pharmacokinetic model for changing drug concentrations in the eye, where drug is injected
directly into the vitreous either as a bolus [dt or rapid prolonged-bolus (infusion)
Ft. Infusion of drug into the eye is of xed rate Ft and known duration. kel is a
rst order removal rate constant from the eye. k0el is the rate of removal of drug due
to active pumping. Amte is the drug amount in the eye compartment. Given most
achievable concentrations, this behaves in a pseudo-linear fashion. Ce =
Concentration in the eye; Ve = volume of distribution of the eye.

Copyright 2003 Marcel Dekker, Inc.

122

Jumbe and Miller

Figure 3 Two-compartment open model. Rates of change in drug concentrations

in the eye can alternatively be described by a two-compartment open model, where
ce;e;p and Vc;e;p represent drug concentrations and apparent volumes of distributions
in the central, eye, and peripheral compartments, respectively. F; k0el are as dened
previously. kcp ; kpc ; kce and kec are rst-order transfer rate constants, and kel is a rstorder elimination rate constant. Ve is as previously dened; Vc = volume of central
compartment. Amtc = Drug amount in the central compartment; Amtp = drug
amount in the peripheral compartment; Amte = drug amount in the eye compartment.

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

123

volumes of distribution do not necessarily correspond to real anatomical

tissue space. They are proportionality constants, which relate observed drug
concentrations to the amount given. Since ocular compartment volumes are
much smaller than the central compartment, the volume of the ocular compartment can be xed at the physiologically determined mean values to
improve identication of the other parameters. The mean drug concentrations after a single dose (or at steady state) are inversely proportional to the
AUC. Since the AUC is related to average observed concentrations, the
ratio of serum (plasma) AUC to aqueous/vitreous humor AUC provides
the most accurate determination of drug penetration into the eye (Fig. 4).
B.

Statistical and Population Approaches to Ocular

Pharmacokinetics

Optimally, one would like to extend the ocular pharmacokinetics knowledge

base to patients. This has been made possible by the recent transfer of
adaptive control theory and optimization methods from control systems
engineering to the eld of pharmaceutical pharmacokinetics/pharmacodynamics (PK/PD). These combined mathematical and statistical tools enable
the robust determination of population pharmacokinetic mean parameter
values and their dispersions from fragmentary data. There is now software
available to support these population-modeling approaches, such as the
nonlinear mixed eects modeling (NONMEM) (89), nonparametric expectation maximization (NPEM2) (88), an iterative two-stage (IT2S) population modeling software package (37,38,93) and WinNonMix (Pharsight
Corporation, Mountain View, CA). These algorithms have been validated
and employed to obtain population pharmacokinetic parameter values from
sparse pharmacokinetic data sets (13,29), and comparisons of these
packages are available in literature (57,76,77,98).
The utility of these mathematical models can be readily extended to
the patient care situation. Most notably, adaptive population models
require only a single pair of data where a prior complete kinetic data set
exists to begin analysis, no matter how many system parameters are available in the population pharmacokinetic model. Thus, each new data set
supplements the population data, which keeps the general population database growing. Most importantly, the more data samples that are obtained
for an individual patient, the more patient-specic the model parameter
estimation becomes, thereby resulting in customized therapy.
We have used a population modeling approach to characterize ocular
pharmacokinetics. We used IT2S population modeling software to characterize and validate this approach following systemic administration of ciprooxacin in a rabbit model (29). Sequential serum samples were obtained in

Copyright 2003 Marcel Dekker, Inc.

124

Jumbe and Miller

Figure 4 Simulation of mean concentrations of sparoxacin in the serum and

vitreous humor of rabbits following a single intravenous bolus (40 mg/kg) using
mean pharmacokinetic parameter estimates from observed data. Penetration was
expressed as a cumulative percentage by dividing the area under the concentration
versus time curve (AUC) in the vitreous by that in the serum.

dierent subjects along with a single, sequential vitreous sample from one
eye in each subject. We simultaneously also obtained sequential vitreous
samples from the contralateral eye of all subjects. Data were analyzed by
IT2S to determine the robustness of pharmacokinetic estimates obtained

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

125

from single data point experiments. Penetration and derived pharmacokinetic constants were equivalent when single or complete ocular data sets
were used. This nding suggests that intensive plasma sampling and optimized single sample experimental design holds promise for the analysis or
penetration of drugs into privileged spaces, such as the eye and CSF, from
which only single samples can be obtained.

C.

Pharmacodynamics

Pharmacodynamics characterizes the relationship between drug concentrations (pharmacokinetics) and a quantiable eect. Modeling this relationship using dierent dose schedules and determining the eect of derived
pharmacodynamic relationships provides a robust approach to optimize
therapy. While this approach has primarily focused on antimicrobials, it
also can provide useful information for pharmaceutical agents used in the
therapy of ocular disease. There has been a long-standing controversy
regarding the relative merits of continuous, intermittent, and single dose
administration on drug penetration into extravascular foci. It has been
argued that the high peak levels achieved by intermittent administration
may be more eective than low sustained levels in reaching protected compartments. These questions are best answered via pharmacodynamic analysis. For bacteria and fungi, there are three pharmacodynamic relationships
that correlate response to exposure (Fig. 5): the ratios of the area under the
concentration-time curve to the MIC (AUC/MIC), the maximal concentration to the MIC (Cmax /MIC), and the time the drug concentration is above
the MIC (time > MIC). The MIC and the 50% inhibitory concentrations
(IC50 ) often closely approximate one another and are often used to signify
the same relative eect. These exposure-therapeutic response relationships
are used for the selection of the best dose and dose-schedule treatment
strategies. To optimize outcome for time dependent drugs, the total daily
dose of these compounds is administered in multiple equally divided doses
(or as a continuous infusion) over 24 hours. Optimal outcome for peak
concentration dependent agents is seen in a single large drug dose per
day, while AUC-driven drugs are schedule independent. Pharmacodynamic
determinants that link microbial response and antimicrobial exposure have
been well characterized in the eld of antimicrobial chemotherapy
(28,30,31,3638,79,87). The characterization of the pharmacodynamically
linked variables requires the determination of a readily measurable outcome. Once this has been established, separate pharmacokinetic and pharmacodynamic experiments may be designed. Data from these experiments
may be analyzed separately, or pharmacokinetic data can be incorporated

Copyright 2003 Marcel Dekker, Inc.

126

Jumbe and Miller

Figure 5 This gures shows the relationship between pharmacokinetic data and
the minimum inhibitory concentration (MIC). The relationships that correlate with
outcome for antimicrobial agents discussed in the chapter are shown.

into the nonlinear dynamic model analysis. This approach takes advantage
of the potential strengths of population modeling outlined above.
Pharmacodynamically based studies comparing the outcome of rabbits
with S. epidermidis endophthalmitis treated with either ciprooxacin or
trovaoxacin [drugs for which ocular outcome correlates with the AUC/
MIC ratio (29) support the utility of PK/PD approaches in optimizing the
design of future studies and characterizing the relationship between antimicrobial drug exposure and microbial response for ocular infections (46).
Pharmacodynamic modeling may also be useful in determining the optimal
mode of systemic drug administration for infectious and noninfectious ocular disease. For example, quinolone antimicrobials show excellent serum
and, often, ocular bioavailability following oral administration. Ocular
AUCs following systemic drug administration are independent of the schedule of drug administration (single dose, intermittent dose, or continuous
infusions) for a given total exposure (56). As a result, for quinolones that
show good ocular penetration [e.g., levooxacin or sparoxacin, but not
ciprooxacin (47,53)], outcome should be similar following intravenous
and oral dosing as long as the AUC/MIC ratios are equivalent. On the
other hand, for drugs for which the pharmacodynamically linked parameter
is the Cmax /MIC ratio, intravenously administered agents would be preferred to those given orally.

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

VII.

127

SUMMARY

Despite the potential benets of parenteral drug administration as adjunctive or alternative therapy in ocular diseases, there are limited data characterizing ocular pharmacokinetics following systemic drug administration.
New population modeling tools for pharmacokinetic data analysis have
made it possible to obtain precise and unbiased estimates of the population
kinetics for chemotherapeutic agents in a privileged site such as the eye
through the use of a complete set of concentration-versus-time proles as
prior estimates plus discreet single serum eye compartment concentration
determinations. This approach can be readily extended to the patient care
situation to provide optimal information for dose and schedule development
and customized patient care management. These methods will prove useful
for determining the role of systemically administered drugs in achieving the
best outcomes for patients with serious eye disease.

REFERENCES
1. Axelrod JL, Kochman RS. Cefamandole levels in primary aqueous humor in
man. Am J Ophthalmol 1978; 85:342348.
2. Badenoch PR, McDonald PJ, Coster DJ. Eect of inammation on antibiotic
penetration into the anterior segment of the rat eye. Invest Ophthalmol Vis Sci
1986; 27:958965.
3. Barza M, Kane A, Baum J. Oxacillin for bacterial endophthalmitis: subconjunctival, intravenous, both, or neither? Invest Ophthalmol Vis Sci 1980;
19:13481354.
4. Barza M, Kane A, Baum J. The eects of infection and probenecid on the
transport of carbenicillin from the rabbit vitreous humor. Invest Ophthalmol
Vis Sci 1982; 22:720726.
5. Barza M, Cuchural G. General principles of antibiotic tissue penetration. J
Antimicrob Chemother 1995; 15 (suppl A):5975.
6. Barza M. Anatomical barriers for antimicrobial agents. Eur J Clin Microbiol
Infect Dis 1993; 12 (suppl 1):S3135.
7. Barza M. Pharmacokinetics of antibiotics in shallow and deep compartments.
J Antimicrob Chemother 1993; 31 (suppl D):1727.
8. Bellhorn MB, Bellhorn RW, Poll DS. Permeability of uorescein-labelled
dextrans in fundus uorescein angiography of rats and birds. Exp Eye Res
1977; 24:595605.
9. Bito LZ. A new approach to the medical management of glaucoma, from the
bench to the clinic, and beyond: the Proctor Lecture. Invest Ophthalmol Vis
Sci 2001; 42:11261133.
10. Blaser J, Stone BB, Groner MC, Zinner SH. Comparative study with enoxacin
and netilmicin in a pharmacodynamic model to determine importance of ratio

Copyright 2003 Marcel Dekker, Inc.

128

11.
12.
13.

14.
15.
16.

17.

18.

19.

20.
21.

22.

23.
24.
25.
26.

27.

Jumbe and Miller

of antibiotic peak concentration to MIC for bactericidal activity and emergence of resistance. Antimicrob Agents Chemother 1987; 31:10541060.
Bleeker GM, van Haeringen NJ, Glasius E. Urea and the vitreous barrier of
the eye. Exp Eyue Res 1968; 7:3036.
Bleeker GM, van Haeringen, NJ, Maas ER, Glasius E. Selective properties of
the vitreous barrier. Exp Eye Res 1968; 7:3746.
Bondareva IB, Sokolov AV, Tischenkova IF, Jellie RW. Population pharmacokinetic modelling of carbamazepine by using the iterative Bayesian
(IT2B) and the nonparametric EM (NPEM) algorithms: implications for
dosage. J Clin Pharm Ther 2001; 26:213223.
Brook I, Ledney GD. Quinolone therapy in themanagement of infection after
irradiation. Crit Rev Microbiol 1992; 18:235246.
Brooks RG. Prospective study of Candida endophthalmitis in hospitalized
patients with candidemia. Arch Intern Med 1989; 149:22262228.
Burns-Bellhorn MS, Bellhorn RW, Benjamin JV. Anterior segment permeability to uorescein-labeled dextrans in the rat. Invest Ophthalmol Vis Sci 1978;
17:857862.
Cekic O, Batman C, Yasar U, Basci NE, Bozkurt A, Kayaalp SO. Human
aqueous and vitreous humour levels of ciprooxacin following oral and topical
administration. Eye 1999; 13:555558.
Cekic O, Batman C, Yasar U, Basci NE, Zilelioglu O, Bozkurt A. Subretinal
uid levels of topical, oral, and combined administered ciprooxacin in
humans. Br J Ophthalmol 2000; 84:10611063.
Cochereau-Massin I, Bauchet J, Marrakchi-Benjaafar S, Saleh-Mghir A,
Faurisson F, Vallos JM, Vallee E, Pocidalo JJ. Ecacy and ocular penetration
of sparoxacin in experimental streptococcal endophthalmitis. Antimicrob
Agents Chemother 1993; 37:633636.
Cunha-Vaz JG, Maurice DM. The active transport of uorescein by the retinal
vessels and the retina. J Physiol 1967; 191:467486.
Dacey RG, Sande MA. Eect of probenecid on cerebrospinal uid concentrations of penicillin and cephalosporin derivatives. Antimicrob Agents
Chemother 1974; 6:437421.
Dacey RG, Welsh JE, Scheld WM, Winn HR, Jane JA, Sande MA.
Alterations in cerebrospinal uid outow resistance in experimental bacterial
meningitis. Trans Am Neurol Assoc 1978; 103:142146.
Davidson SI. Post-operative bacterial endophthalmitis. Trans Ophthalmol Soc
UK 1985; 104:278284.
Davson H, Segal MB. Eect of cerebrospinal uid on volume of distribution
of extracellular markers. Brain 1969; 92:131136.
de la Pena A, Liu P, Derendorf H. Microdialysis in peripheral tissues. Adv
Drug Deliv Re 2000; 15(45):189216.
Driebe WT Jr, Mandelbaum S, Forster RK, Schwartz LK, Culbertson WW.
Pseudophakic endophthalmitis. Diagnosis and management. Ophthalmology
1986; 93:442448.
Drusano GL. Role of pharmacokinetics in the otucome of infections.
Antimicrob Agents Chemother 1988; 32:289297.

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

129

28. Drusano GL. Pharmacodynamics of antiretroviral chemotherapy. Infect

Control Hosp Epidemiol 1993; 14:530536.
29. Drusano GL, Liu W, Perkins R, Madu A, Madu C, Mayers M, Miller MH.
Determination of robust ocular pharmacokinetic parameters in serum and
vitreous humor of albino rabbits following systemic administration of ciprooxacin from sparse data sets by using IT2S, a population pharmacokinetic
modeling program. Antimicrob Agents Chemother 1995; 39:16381687.
30. Drusano GL, Craig WA. Relevance of pharmacokinetics and pharmacodynamics in the selection of antibiotics for respiratory tract infections. J
Chemother 1997; 9 (suppl 3):3844.
31. Drusano GL. Infection in the intensive care unit: beta-lactamase-mediated
resistance among Enterobacteriaceae and optimal antimicrobial dosing. Clin
Infect Dis 1998; 27 (suppl 1):S111116.
32. Elmquist WF, Sawchuk RJ. Application of microdialysis in pharmacokinetic
studies. Pharm Res 1997; 14:267288.
33. Ficker L, Meredith TA, Gardner S, Wilson LA. Cefazolin levels after intravitreal injection. Eects of inammation and surgery. Invest Ophthalmol Vis Sci
1990; 31:502505.
34. Ficker LA, Kirkness CM, Steele AD, Rice NS, Gilvarry AM. Intraocular
surgery following penetrating keratoplasty: the risks and advantages. Eye
1990; 4:693697.
35. Fiscella RG, Nguyen TK, Cwik MJ, Phillpotts BA, Friedlander SM,
Alter DC, Shapiro MJ, Blair NP, Gieser JP. Aqueous and vitreous penetration of levooxacin after oral administration. Ophthalmology 1999; 106:2286
2290.
36. Fluckiger U, Segessenmann C, Gerber AU. Integration of pharmacokinetics
and pharmacodynamics of imipenem in a human-adapted mouse model.
Antimicrob Agents Chemother 1991; 5:19051910.
37. Forrest A, Ballow CH, Nix DE, Birmingham MC, Schentag JJ. Development
of a population pharmacokinetic model and optimal sampling strategies for
intravenous ciprooxacin. Antimicrob Agents Chemother 1993; 37:1065
1072.
38. Forrest A, Nix DE, Ballow CH, Goss TF, Birmingham MC, Schentag JJ.
Pharmacodynamics of intravenous ciprooxacin in seriously ill patients.
Antimicrob Agents Chemother 1993; 37:10731081.
39. Forrester JV, McMenamin PG. Immunopathogenic mechanisms in intraocular inammation. Chem Immunol 1999, 73:159185.
40. Habgood MD, Begley DJ, Abbott NJ. Determinants of passive drug entry
into the central nervous system. Cell Mol Neurobiol 2000; 20:231253.
41. Hillered L, Persson L. Theory and practice of microdialysisprospect for
future clinical use. Acta Neurochir Suppl (Wien) 1999; 75:36.
42. Hughes PM, Krishnamoorthy R, Mitra AK. Vitreous disposition of two acycloguanosine antivirals in the albino and pigmented rabbit models: a novel
ocular microdialysis technique. J Ocul Pharmacol Ther 1996; 12:209224.
43. Jacquez JA. Parameter identiability is required in pooled data methods. J
Pharmacokinet Biopharm 1996; 24(3):301305.

Copyright 2003 Marcel Dekker, Inc.

130

Jumbe and Miller

44. Jaehde U, Sorgel F, Reiter A, Sigl G, Naber KG, Schunack W. Eect of

probenecid on the distribution and elimination of ciprooxacin in humans.
Clin Pharmacol Ther 1995; 58:532541.
45. Jumbe NL, W Liu, Drusano, GL Louie, A Miller MH. Comparison of
Quinolone Active Eux after Systemic, Intra-vitreous or Intra-cerebroventricular Administration. 39th Interscience Conference on Antimicrobial Agents
and Chemotherapy, Abstract 1924, 1999, p. 46.
46. Jumbe NL, Drusano GL, Miller MH. Parametric Comparison of
Midrodialysis and Direct Vitreous Sampling Techniques at Non Steady
State Conditions. 39th Interscience Conference on antimicrobial Agents and
Chemotherapy, Abstract 1923, 1999, p. 46.
47. Jumbe NL, W Liu, Kaw P, Drusano G, Louie A, Miller M. Fluoroquinolonerifampin potentiation in S. epidermidis [SE] endophthalmitis: emergence of
rifampin resistance. Abstract 3641B499, IOVS 40, 1999, p. 5690.
48. Kaw P, Jumbe N, Liu W, Mayers M, Drusano G, Miller M. The penetration
of trovaoxacin into the eye and CSF of rabbits. Abstract 469B429, IOVS
40, 1999, p. 588.
49. Kra MC, Sanders DR, McGuigan L, Raanan MG. Inhibition of bloodaqueous humor barrier breakdown with diclofenac. A uorophotometric
study. Arch Ophthalmol 1990; 108:380383.
50. Leggett JE, Fantin B, Ebert S, Totsuka K, Vogelman B, Calame W, Mattie H,
Craig WA. Comparative antibiotic dose-eect relations at several dosing
intervals in murine pneumonitis and thigh-infection models. J Infect Dis
1989; 159:281292.
51. Lesar TS, Fiscella RG. Antimicrobial drug delivery to the eye. Drug Intell
Clin Pharm 1985; 19:642654.
52. Liu W, Liu QF, Perkins R, Drusano G, Louie A, Madu A, Mian U, Mayers
M, Miller MH. Pharmacokinetics of sparoxacin in the serum and vitreous
humor of rabbits: physicochemical properties that regulate penetration of
quinolone antimicrobials. Antimicrob Agents Chemother 1998; 42:1417
1423.
53. Liu W, Jumbe N, Kaw P, Lomastro B, Louie A, Miller M. Comparison of
quinolone pharmacokinetics in vitreous humor and CSF following IV drug
administration in albino rabbits. Abstract 468B428, IOVS 40, 1999.
54. Macha S, Mitra AK. Ocular pharmacokinetics in rabbits using a novel dual
probe microdialysis technique. Exp Eye Res 2001; 72:289299.
55. Madu A, Cioe C, Mian U, Burroughs M, Tuomanen E, Mayers M, Schwartz
E, Miller M. Pharmacokinetics of uconazole in cerebrospinal uid and
serum of rabbits: validation of an animal model used to measure drug concentrations in cerebrospinal uid. Antimicrob Agents Chemother 1994;
38:21112115.
56. Madu AA, Mayers M. Ocular manifestation of systemic infections. Curr Opin
Ophthalmol 1996; 7:8590.

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

131

57. Mahmood I, Miller R. comparison of the Bayesian approach and a limited

sampling model for the estimation of AUC and Cmax : a computer simulation
analysis. Int J clin Pharmacol Ther 1999; 37:439445.
58. Marrakchi-Benjaafar S, Cochereau I, Pocidalo JJ, Carbon C. Systemic prophylaxis of experimental staphylococcal endophthalmitis: comparative ecacy of sparoxacin, peoxacin, imipenem, vancomycin, and amikacin. J
Infect Dis 1995; 172:13121316.
59. Maurice DM, Mishima S. Ocular pharmacokinetics. In: Sears ML, ed.
Pharmacology of the Eye. Handbook of Experimental Pharmacology. Vol.
69. New York: Springer-Verlag, 1984: 19116.
60. Mayers M, Rush D, Madu A, Motyl M, Miller MH. Pharmacokinetics of
amikacin and chloramphenicol in the aqueous humor of rabbits. Antimicrob
Agents Chemother 1991; 35:17911798.
61. Mian UK, Mayers M, Garg Y, Liu QF, Newcomer G, Madu C, Liu W, Louie
A, Miller MH. Comparison of uconazole pharmacokinetics in serum, aqueous humor, vitreous humor, and cerebrospinal uid following a single dose
and at steady state. J Ocul Pharmacol Ther 1998; 14:459471.
62. Miller MH, Madu A, Samathanam G, Rush D, Madu CN, Mathisson K,
Mayers M. Fleroxacin pharmacokinetics in aqueous and vitreous humors
determined by using complete concentration-versus-time data from individual
rabbits. Antimicrob Agents Chemother 1992; 36:3238.
63. Moog E, Knothe H. [Passage of various tetracyclines into human aqueous
humor in dependence on their protein binding in the blood seruem]. Ber
Zusammenkunft Dtsch Ophthalmol Ges 1969; 69:536539.
64. Moog, F, Etzler ME, Grey RD. The dierentiation of alkaline phosphatase in
the small intestine. Ann NY Acad Sci 1969; 166:447-465.
65. Murray DC, Christopoulou D, Hero M. Intravitreal penetration of teicoplanin. Eye 1999; 13:604605.
66. Nau R, Sorgel F, Prange HW. Lipophilicity at pH 7.4 and molecular
size govern the entry of the free serum fraction of drugs into the cerebrospinal
uid in humans with uninamed meninges. J Neurol Sci 1994; 122:6165.
67. Oldendorf WH. Stereospecicity of blood-brain barrier permeability to amino
acids. Am J Physiol 1973; 224:967969.
68. Oldendorf WH, Crane PD, Braun LD, Wade LA, Diamond JM. Blood-brain
barrier transport of basic amino acids is selectively inhibited at low pH. J
Neurochem. 1983; 40:797800.
69. Olson JC, Flynn HW Jr, Forster RK, Culbertson WW. Results in the treatment of postoperative endophthalmitis. Ophthalmology 1983; 90:692699.
70. Ooie T, Suzuki H, Terasaki T, Sugiyama Y. Characterization of the transport
properties of a quinolone antibiotic, eroxacin, in rat choroid plexus. Pharm
Res 1996; 13:523527.
71. Ooie T, Suzuki H, Terasaki T, Sugiyama Y. Kinetics of quinolone antibiotics
in rats: eux from cerebrospinal uid to the circulation. Pharm Res 1996;
13:10651068.

Copyright 2003 Marcel Dekker, Inc.

132

Jumbe and Miller

72. Ozdamar A, Aras C, Ozturk R, Karacorlu M, Bahcecioglu H, Ozkan S.

Ocular penetration of cefepime following systemic administration in humans.
Ophthalmic Surg Lasers 2001; 32:2529.
73. Ozturk F, Kortunay S, Kurt E, Ilker SS, Basci NE, Bozkurt A. Penetration of
topical and oral ciprooxacin into the aqueous and vitreous humor in
inamed eyes. Retina 1999; 19:218222.
74. Ozturk F, Kortunay S, Kurt E, Inan UU, Ilker SS, Basci N, Bozkurt A. The
eect of long-term use and inammation on the ocular penetration of topical
ooxacin. Curr Eye Res 1999; 19:461464.
75. Ozturk F, Kurt E, Inan UU, Kortunay S, Ilker SS, Basci NE, Bozkurt A. The
eects of prolonged acute use and inammation on the ocular penetration of
topical ciprooxacin. Int J Pharm 2000; 25(204): 97100.
76. Park K, Verotta D, Gupta SK, Sheiner LB. Use of a pharmacokinetic/pharmacodynamic model to design an optimal dose input prole. J Pharmacokinet
Biopharm 1998; 26:471492.
77. Parker TJ, Della Pasqua OE, Loizillon E, Chezaubernard c, Jochemsen R,
Danhof M. Pharmacokinetic-pharmacodynamic modelling in the early development phase of anti-psychotics: a comparison of the eects of clozapine, S
16924 and S 18327 in the EEG model in rats. Br J Pharmacol 2001; 132: 151
158.
78. Perkins RJ, Liu W, Drusano G, Madu A, Mayers M, Madu C, Miller MH.
Pharmacokinetics of ooxacin in serum and vitreous humor of albino and
pigmented rabbits. Antimicrob Agents chemother 1995; 39:14931498.
79. Preston SL, Drusano GL, Berman AL, Fowler CL, Chow AT, Dornseif B,
Reichl V, Natarajan J, Corrado M. Pharmacodynamics of levooxacin: a new
paradigm for early clinical trials. JAMA 1998; 279:125129.
80. Rabkin MD, Bellhorn MB, Bellhorn RW. Selected molecular weight dextrans
for in vivo permeability studies of rat retinal vascular disease. Exp Eye Res
1977; 24:607612.
81. Records RE. The penicillins in ophthalmology. Surv Ophthalmol 1969;
13:207214.
82. Rittenhouse KD, Peier RL Jr, Pollack GM. Evaluation of microdialysis
sampling of aqueous humor for in vivo models of ocular absorption and
disposition. J Pharm Biomed Anal 1998; 16:951959.
83. Rittenhouse KD, Peier RL Jr, Pollack GM. Microdialysis evaluation of the
ocular pharmacokinetics of propranolol in the conscious rabbit. Pharm Res
1999; 16:736742.
84. Salminen L. Ampicillin penetration into the rabbit eye. Acta Ophthalmol
(Copenh) 1978; 56:977983.
85. Salminen L. Cloxacillin distribution in the rabbit eye after intravenous injection. Acta Ophthalmol (Copenh) 1978; 56:1119.
86. Sande MA, Sherertz RJ, Zak O, Dacey RG, Bodine JA, Stausbaugh LJ.
Factors inuencing the penetration of antimicrobial agents into the cerebrospinal uid of experimental animals. Scand J Infect Dis Suppl 1978; 14:160163.

Copyright 2003 Marcel Dekker, Inc.

Ocular Drug Transfer

133

87. Schentag JJ, Strenekoski-Nix LC, Nix DE, Forrest A. Pharmacodynamic

interactions of antibiotics alone and in combination. Clin Infect Dis 1998;
27:4046.
88. Schumitzky, A. Nonparametric EM algorithms for estimating prior distributions. Appl Math Comput 1991; 45: 141157.
89. Sheiner LB, Rosenberg B, Marathe VV. Estimation of population characteristics of pharmacokinetic parameters from routine clinical data. J
Pharmacokinet Biopharm 1977; 5:445479.
90. Sheiner LB, Beal SL. Evaluation of methods for estimating population pharmacokinetics parameters. I. Michaelis-menten model: routine clinical pharmacokinetic data. J Pharmacokinet Biopharm 1980; 8:553571.
91. Sheiner BL, Beal SL. Evaluation of methods for estimating population pharmacokinetic parameters. II. Biexponential model and experimental pharmacokinetic data. J Pharmacokinet Biopharm 1981; 9:635651.
92. Sheiner LB, Beal SL. Evaluation of methods for estimating population pharmacokinetic parameters. III. Monoexponential model: routine clinical pharmacokinetic data. J Pharmacokinet Biopharm 1983; 11:303319.
93. Steimer JL, Mallet A, Golmard JL, Boisvieux JF. Alternative approaches to
estimation of population pharmacokinetic parameters: comparison with the
nonlinear mixed-eect model. Drug Metab Rev 1984; 15:265292.
94. Streilein JW, Ksander BR, Taylor AW. Immune deviation in relation to ocular immune privilege. J Immunol 1997; 15(158):35573560.
95. Streilein JW. Immunoregulatory mechanisms of the eye. Prog Retin Eye Res
1999; 18:357370.
96. Tulkens PM. Intracellular distribution and activity of antibiotics. Eur J Clin
Microbiol Infect Dis 1991; 10:100106.
97. van de Waterbeemd H, Camenisch G, folkers G, Chretien JR, Raevsky OA.
Estimation of blood-brain barrier crossing of drugs using molecular size and
shape, and H-bonding descriptors. J Drug Target 1998; 6:151165.
98. Vermes A, Math t RA, van der Sijs IH, Dankert J, Guchelaar HJ. Population
pharmacokinetics of ucytosine: comparison and validation of three models
using STS,NPEM, and NONMEM. Ther Drug Monit 2000; 22:676687.
99. Vinores SA. Assessment of blood-retinal barrier integrity. Histol Histopathol
1995; 10:141154.
100. Waga J, Nilsson-Ehle I, Ljungberg B, Skarin A, Stahle L, Ehinger B.
Microdialysis for pharmacokinetic studies of ceftazidime in rabbit vitreous.
J Ocul Pharmacol Ther 1999; 15:455463.
101. Wang Y, Wong SL, Sawchuk RJ. Microdialysis calibration using retrodialysis
and zero-net ux: application to a study of the distribution of zidovudine to
rabbit cerebrospinal uid and thalamus. Pharm Res 1993; 10:14111419.
102. Yoshida A, Ishiko S, Kojima M. Outward permeability of the blood-retain
barrier. Graefes Arch Clin Exp Ophthalmol 1992; 230:7883.

Copyright 2003 Marcel Dekker, Inc.

5
Ocular Pharmacokinetics and
Pharmacodynamics
Ronald D. Schoenwald
College of Pharmacy, The University of Iowa, Iowa City, Iowa, U.S.A.

I.

INTRODUCTION

The study of pharmacokinetic processesabsorption, distribution, and

eliminationis fundamental to determining the appropriate dosing regimen
for an individual patient administered a drug by systemic routes.
Pharmacokinetics has also been indispensable in designing an improved
therapeutic agent or the means by which it is delivered. Quite often, pharmacokinetic analyses of systemically useful drugs is approached by dividing
the body into a series of compartments that mathematically represent tissue
levels of drug over time as a summation of exponentials. The number
of compartments that are chosen are equal to and limited by the number
of exponentials that can be assigned to the concentration of drug found in a
particular body tissue and/or uid over time.
The compartments are often interpreted physiologically or anatomically to contain tissues and/or organs which are kinetically homogeneous.
The term kinetically homogeneous compartments refers to tissues and/or
organs that show similar if not identical kinetics for certain drugs even
though the tissues and/or organs within the compartment may be dissimilar
physiologically or anatomically. Although the rates within these groups of
tissues and/or organs are similar for a particular drug, the extent of drug
distribution to each is usually not equal.
The ophthalmic literature contains many reports of drug concentrations measured in eye tissues over time, administered either topically,
subconjunctively, orally, intravenously, or vitreally. The latter route of
135

Copyright 2003 Marcel Dekker, Inc.

136

Schoenwald

administration has been of major interest because of the prevalence of cytomegalovirus retinitis associated with human immunodeciency virus (HIV).
When classical pharmacokinetic approaches (compartmental modeling)
have been applied to ophthalmic drugs, a number of limitations have
been found to restrict the usefulness of pharmacokinetics in the practice
of ophthalmology. Noncompartmental approaches, such as moment analysis, circumvent the diculty in choosing anappropriate compartmental
model, but the assumption of linearity remains, as do various intractable
experimental diculties. Consequently, the majority of the ocular pharmacokinetic reports do not employ a classical approach but focus on
whether or not drug reaches target tissues and whether or not the tissue
concentrations are within the therapeutic range.
Limitations on the use of classical pharmacokinetic approaches, diculties in designing and implementing animal studies for the eye, as well as
interpretation of the results obtained from animal models are discussed in
the following sections.

II.

LIMITATIONS ON THE PRACTICAL USE OF CLASSICAL

MODELING

At the present time it is not possible to predict optimal dosing regimens in

the human eye for dierent drugs or for the same drug in dierent dosage
forms. The most signicant reason for not conducting ocular pharmacokinetic studies in the human eye is the inability to sample tissues or uids from
the intact eye without risking pain and/or injury. Although the rabbit eye is
useful in predicting human ocular toxicities (1), the eyes of each species are
dissimilar in anatomy and physiology (see Table 1) such that predicting
human ocular pharmacokinetics from rabbit data may not be very precise
for certain drugs.
As a further complication in predicting human ocular pharmacokinetics based upon animal data, samples from eye tissues cannot be continuously sampled over time. Although a number of tissues can be removed
quickly and precisely from the rabbit eye, a dierent animal is used to
determine drug concentration at a single time point, increasing intersubject
variability. From this pooling approach a kinetic prole of drug concentration over time is constructed from a number of rabbits, which are sacriced
at each time point. As many as 150250 rabbit eyes may be required to
complete a statistically signicant bioequivalence test between two ophthalmic drugs. It is often assumed that a representative pharmacokinetic prole
can be constructed from noncontinuous sampling if enough time intervals

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

137

Table 1 Anatomical and Physiological Differences in the New

Zealand Rabbit and Human Eye Pertinent to Ophthalmic
Phamacokinetics
Pharmacokinetic factor
Tear volume
Tear turnover rate
Spontaneous blinking rateb
Nictitating membranec
pH of tears
Milliosmolarity of tears
Corneal thickness
Corneal diameter
Aqueous humor volume
Aqueous humor turnover rate

Rabbit eye

Human eye

7.5 mL
0.60.8 mL/min
45 times/min
Present

7.030.0 mLa
0.52.2 mL/min
15 times/min
Absent
7.147.82
305 mOsm/L
0.52 mm
12 mm
310 mL
1.53 mL/min

d
d

0.40 mm
15 mm
d
d

Range depending on blinking rate and conjunctival sac volume.

Occurs during normal waking hours without apparent external stimuli.
c
Signicance of nictitating membrane is small relative to overall loss rate
from precornea area.
d
Approximately same measurement as human.
b

and sucient eyes per time interval are chosen, but this number is not
always based on a sound statistical approach.
A.

Approaches Designed to Overcome the Pooling of

Individual Samples

Because of the use of pooled data obtained from animal eyes, pharmacokinetic proles of various ocular tissues are limited in their interpretation of
the data. Sophisticated models that more adequately explain the data are
often too complex and tend to be overly scattered and not as smooth as data
obtained in a serial fashion. Therefore, the models have to be oversimplied.
As a result of the oversimplication of the modeling and the use of animal
eyes instead of human eyes, the value of ocular pharmacokinetic studies are
limited to:
1. Validating the minimum eective concentrations in various
tissues from dierent routes of administration
2. Determining the best route of administration for entry into the
eye as well as accumulation at the site of action
3. Determining the elimination half-life, which can serve as a guide
to selecting dosing regimens for further study

Copyright 2003 Marcel Dekker, Inc.

138

Schoenwald

4. Estimating the safety of various drugs by determining their

accumulation in tissues of interest, such as the lens or retina
Approaches to extending the application of pharmacokinetics have recently
focused on two approaches: serial sampling and microdialysis.
1. Serial Sampling
Serial sampling has not been advocated as a reliable method for obtaining
samples from animal eyes because of the possibility of a breakdown of the
blood-aqueous barrier, resulting in an alteration of the drug clearance from
the eye. Studies by Tang-Liu and coworkers (2,3) established that for a
single intracameral injectkion of 5 mL solutions containing urbiprofen or
levobunolol, no signicant breakdown of the blood aqueous barrier had
occurred since protein concentration was <1 mg/mL. However, serial sampling is a repetitive process not just a single injection. Miller et al. (4) studied
the eect of paracentesis on the pharmacokinetics of eroxacin following
direct intravitreal or systemic drug administration and serially measuring
drug levels in serum, aqueous, and vitreous humor. Samples were obtained
from aqueous humor with the use of a sterile 30-gauge needle fused to a
calibrated 20 mL capillary tube. A volume of 7 mL was removed at each time
interval. A 28-gauge needle was used to sample vitrous humor and 20 mL
samples were removed. Figure 1 gives the results, for which no statistically
signicant dierences were observed in assigning a correct model for the
data, nor were there signicant dierences in the half-lives for serum, aqueous, and vitreous humor2.34,3.20, and 3.88 hours, respectively. Similar
experiments with similar results were conducted for the pharmacokinetics of
amikacin and chloramphenicol in rabbit aqueous humor following anterior
chamber injection and serial sampling using 30-gauge needle and removing
7 mL of aqueous (5).
2. Microdialysis Approach
The microdialysis approach was widely used in the analysis of drugs in brain
tissue and cerebral spinal uid with the intention of measuring free drug
concentration at the sampling site (6,7). More recently the approach has
been adapted for use in the eye, particularly in the measurement of aqueous
and vitreal humor concentrations of drug (812). Its primary advantage is to
measure complete concentration-time proles in individual animal eyes,
reducing the number of animals required, and to more accurately dene
the pharmacokinetics of ocular drugs.
The procedure consists of surgically placing a probe within the aqueous or vitreous chambers of an anesthetized animal. The microdialysis

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

139

Figure 1 Average concentrations of eroxacin in serum, aqueous humor and

vitreous humor following intravenous administration to 10 rabbits. (Adapted
from Ref. 4.)

probe is designed to allow for perfusate to ow into and out of the unit at a
xed rate, usually 23.5 mL/min. The base of the probe is constructed with a
semipermeable membrane (e.g., a 4 mm polycarbonate membrane with a
20,000 dalton cuto), which allows drug to diuse into a perfusate (11).
Probe inlet and outlet lines allow for collection of dialysate and analysis of
drug over time. It is essential for the percent relative recovery to be determined so that accurate concentrations can be estimated, which may likely
vary depending on the membrane used in the probe, the size of the probe
and the perfusion rate. For example, reported values for propranolol and
two nucleoside antivirals (ganciclovir and acyclovir) were 46 and 15%,
respectively (12,13), when probe and perfusion rate varied. Of concern is
the insertion of the probe, which appears to result in an increase in protein
concentrations in aqueous humor (12), suggesting an alteration of clearance.
The use of systemic anesthesia has also been shown to alter the pharmacokinetics of propranolol by increasing the area under the aqueous humor
time-concentration plot, possibly by decreasing aqueous humor turnover
(12). Nevertheless, results using this technique have reduced the number
of animals needed to estimate ocular pharmacokinetic parameter values

Copyright 2003 Marcel Dekker, Inc.

140

Schoenwald

and have produced curves that are smooth in appearance and, therefore,
more reliable in itnerpreting the tting of models to the results.

III.

ADEQUATE CHOICE OF TIME INTERVALS

Studies using pooled data from dierent rabbits often use as few as 4 or 5
rabbit eyes per time interval or as many as 22, depending on the objective of
the study (1419). A statistical basis for choosing an appropriate number of
rabbit eyes in ophthalmic bioavailability studies has been established using
areas under the concentration-time curve (AUC) (20,21). The AUC is
related to the extent of absorption, which, similar to systemically administered drugs, is most important in determining therapy for chronic medication. The rate of absorption into the eye is less precise a measurement and
also less pertinent to chronic therapy.
How many intervals and the length of time between each interval has
not always been decided upon using a rational approach. Some general
guidelines can be applied based upon the importance of absorption, distribution, and elimination to the kinetic prole. Each kinetic process is practically complete after ve half-lives (96.875%); therefore, if the half-life for
each kinetic process is known or can be easily arrived at, the theoretical
length of time that should be used to generate a concentration  time prole
can be estimated (see Fig. 2).

A.

Elimination Phase

Elimination of drug from the eye is the least complicated process and can be
discussed rst. It occurs over the entire concentration-time prole, and as
long as it is the slowest of pharmacokinetic processes, the latter log-linear
phase of the drug concentration-time prole represents the elimination
phase and its slope allows for the calculation of elimination half-life and
the time necessary for completion of the process. For example, the half-life
for elimination of foscarnet from rabbit vitreous humor is 34 hours (18)
after intravitreal injection; therefore, the process is complete in 5  34 hours,
or 170 hours. Foscarnet can be measured to a sensitivity of 4.5 mg/mL using
an electrochemical detector with high-peformance liquid chromatography
(HPLC) methodology (22).
In general, at least four or ve time intervals should be spaced over the
latter log llnear phase following the completion of distribution within the
eye. For drugs with a much shorter half-life and a less sensitive assay than
foscarnet, a practical limitation to measuring drug in the postdistributive

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

141

Figure 2 Sampling times and critical problems associated with measuring absorption, distribution, and elimination of an ocularly administered drug.

phase is the sensitivity of the assay, which could limit measuring the decline
of the drug over three to ve half-lives.
B.

Absorption Phase

For many years absorption has been assumed to occur via the cornea;
however, recently, scleral routes of administration have been suggested as
routes of entry into the eye. Absorption is complex to estimate in the eye
since lag time and drainage lengthen and shorten, respectively, the length of
time that the absorption process is operative. For example, the rst-order
absorption rate constant ka for phenylephrine is 4:5  105 min1 , which
gives a half-life of 128.3 hours; however, drainage permits drug to remain at
the corneal absorption site for approximately 36 minutes only, depending
on the volume and viscosity of the instilled solution (16,2326).
Consequently, the absorption process is abruptly terminated from theoretical expectations. The short residence time of drug at the absorption
site results in exceptionally poor bioavailability.
Phenylephrine is atypical since it is very hydrophilic and transverses
the cornea at a very slow rate. On the other hand, pilocarpine is moderately
lipophilic and therefore more rapidly absorbed. Nevertheless, the dierence

Copyright 2003 Marcel Dekker, Inc.

142

Schoenwald

between the theoretically expected time and the actual time during which
absorption occurs remains large. For example, Makoid and Robinson (27)
reported an accurate ka of 6:2  103 min for pilocarpine 15, which gives an
absorption half-life of 111.8 minutes and predicts that the process is complete in about 10 hours. However, the drainage rate constant is approximately 100 times larger with an estimated completion of the process in about
6 minues (27). As a result of the very short residence time of drug at the
absorption site, the extent of absorption of ophthalmic drugs is about
110% (2,14,16,19,28).
The initial time for drug to transverse the cornea, referred to as the lag
time, is suciently long that most often the time to peak occurs over a much
longer time period than the time that a drug solution resides on the cornea
(i.e., 36 min). The majority of ophthalmic drugs require between 20 and 60
minutes after instillation to reach the time to peak. Therefore, a reasonable
time period exists so that up to three to four time intervals can be chosen to
characterize the absorption phase. Because of the very low extent of absorption for ophthalmic drugs as well a the large correction needed for the
drainage rate, the classical deconvolution techniques (Wagner-Nelson and
Loo-Riegelman) (29) cannot be used without severely overestimating the
rst-order absorption rate consta,t ka .
The other routes of administration that have been suggested as routes
of entry into the eye are penetration across the conjunctiona/sclera (30) and/
or absorption via vessels imbedded in the conjunctiva and sclera (e.g., anterior ciliary artery). Vessels embedded in the conjunctiva/sclera only partially
empty into sites of action within the eye (i.e., the iris/ciliary body) (31).
These routes of entry are in parallel and also contribute to the total value
for ka , the absorption rate constant.
The competing processes, referred to as nonabsorptive processes, have
been specically dened as drainage, conjunctival absorption with ultimate
emptying into venous circulation, nictitating membrane absorption, tear
turnover, drug-tear protein binding, and drug metabolism (16).
C.

Distribution Phase

In an ideal curve, a distribution phase can be identied visually as the

concave portion of the log concentration-time curve immediately following
the time to peak. The latter log linear phase is the elimination or postdistributive phase (see Fig. 2). The distribution phase, which is expected to be
shorter than the elimination phase, cannot be visually identied as easily as
absorption and elimination. Depending on the numerical relationship
between the microtransfer constants associated with drug disposition, it is
possible that the prole representing drug concentration over time cannot be

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

143

characterized by the proper number of exponentials. Often one less exponential than actually exists, paticularly when k21 > ka and alpha > ka (32).
In addition, the concave portion of the log concentration-time curve may
not be apparent because distribution is rapid. On the other hand, when drug
concentrations from just after the time to peak to the beginning of the
postdistributive phases are concave to the time axis, then the distributive
phase is discernible and the concave portion of the prole should be represented by an additional two to four time intervals.
As indicated in Figures 2, 1012 time intervals will suce if they are
properly chosen to represent absorption, distribution, and elimination.
However, if variabiilty is too great, a larger number of eyes could be chosen
at each interval to characterize adequately the drugs ocular pharmacokinetics. It is assumed that a smooth representative prole will result if a large
enough sample size is chosen and pooling of individual rabbit eyes is used.
For bioequivalence, the time to peak (tp is a critical concentration to obtain
since it represents a maximum concentration. Because of the relatively large
and constant nature of the drainage rate for topically administered solutions
and suspensions, the tp only varies from about 20 to 60 minutes for ophthalmic drugs regardless of a drugs physicochemical behavior. As discussed
previously, in practice the distribution phase is usually not apparent either
because of inherent variabiilty or because the magnitude of the microconstants representing distribution to peripheral tissues prevents observation of
the characteristic concavity shown in Figure 2.
D.

Other Factors Inuencing Choice of Sampling Times

Unfortunately, more subtle factors may be operating to complicate the

choice of sampling times and obscure the expected shape of the prole
shown in Figure 2. For example, for some drugs, most notably cyclosporine (19,33,34), pilocarpine (35), and verapamil (36), rapid equilibration
with aqueous humor is not apparent. As a result, the terminal log-linear
elimination phase of the concentration of drug in aqueous humor vs. time
is dicult to identify unless time intervals are extended well beyond what
is usually considered adequate for ocular studies (i.e., 25 hours). For
pilocarpine, when drug concentration was measured for 11 hours, a biexponential decline was evident and an elimination half-life of 2.75 hours
was observed (27) compared to earlier studies, which estimated the half-life
to be about 4060 minutes. However, the latter log-linear decline of
pilocarpine may be due to residual drug slowly redistributing from the
lens. Since these levels are low and released slowly, the use of 4060
minutes is probably better in practical situations. For cyclosporine, the
distribution phase was not apparent for 810 hours after subconjunctival

Copyright 2003 Marcel Dekker, Inc.

144

Schoenwald

administration (34). Acheampong et al. (34) observed that cyclosporine in

both rabbit and beagle dogs accumulated signicantly in the cornea, lens,
lacrimal gland, and iris/ciliary body. These tissues acted as reservoirs to
prolong retention in the eye likely because of the drugs lipophilic nature
and poor water solubility.
Another limiting factor is the sensitivity of the assay. Signicant
advances have been made in recent years so that it is not necessary to
prepare a radioactive tracer for adequate detection of drug in eye tissues.
HPLC with ultraviolet, radioactive, or uorescence detectors, laser scanning
confocal Raman spectroscopy, or radioimmunoassay can measure drug
concentrations in ocular tissues at levels as low as 1 ng per mL or g of tissue
(3743).

IV.

OCULAR PHARMACOKINETIC MODELING

Although there are intractable diculties in clearly and unambiguously

dening pharmacokinetic parameters exclusively for the eye, such as clearance, volume of distribution, as well as rate and extent of absorption, it is
also true that ophthalmic drugs are a relatively small commercial market
compared to systemically adminstered drugs. Nevertheless, the classical
pharmacokinetic approach of expressing the concentration-time curve
into a sum of exponentials has been applied to the eye (4,13,14,17,19,28,
36,4451), but much less extensively than other routes of administration.
At the present time, limited application has resulted from these studies in
developing new ophthalmic drugs with optimal pharmacokinetic behavior
or more specically in estimating dosing regimens.
Although the classical pharmacokinetic approach has limitations
when applied to the eye, in situ or in vitro experiments have been developed
that directly measure kinetic phenomena and are more accurate when compared to results obtained from classical compartmental approaches. Very
specic techniques have been developed to measure precorneal drug disposition, permeability of various corneal layers, metabolism is various eye
tissues (52,53) and to measure uorescein kinetics, which have led to
the understanding of tear, aqueous, and vitreous dynamics (54,55) and
have been responsible for the development of very useful clinical applications. These latter approaches are discussed in another chapter.
A.

Classical Modeling Approaches

Initially, a classical pharmacokinetic approach is applied to concentrationtime curves derived from topical instillation to the eye in much the same

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

145

manner as data derived from systemically administered drugs. The curve is

rst expressed by a computer-determined sum of exponentials, which closely
t the experimental data and likewise show no systemic deviation. In the
eye, aqueous humor is most often assigned to the central compartment,
which is reversibly connected to one or more peripheral compartments
and/or a reservoir compartment in the various models (or schemes) that
have been derived. This choice is compatible with physiological reality
since aqueous humor, which lls the anterior and posterior chambers, is
the circulating uid bathing the peripheral tissues. Drugs instilled topically
on the eye primarily reach the rst third of the eye, which encompasses
these regions. Topically applied ophthalmic drugs do not reach the retina
in signicant concentrations. Consequently, models can be devised that treat
the viscous region as a central compartment particularly when drug is
administered intravitreally.
In the eye, the cornea, conjunctiva, lens, iris/ciliary body, choroid, and
vitreous are specic tissues that are often lumped together into one or more
peripheral compartments. A peripheral compartment can be reversibly connected to a central compartment, but if redistribution into aqueous humor is
negligible or nonexistent, peripheral tissues can act as a sink or reservoir
compartment. The exit out of the eye is into the blood or circulating uid of
the body.
Figure 3 lists the classical compartmental schemes most commonly
applied to ophthalmic drugs following topical application. For many
years pilocarpine received the most attention (35,4447). More recently,
publications focusing on classical modeling have decreased, but those that
have been published used a varied assortment of drugs for study
(5,19,22,36,40,4851). The most appropriate model seems to depend heavily
on the design of the study (i.e., number and length of sampling periods and
number of tissues measured for drug content over time), the specicity and
sensitivity of the assay, the sophistication of the curve-tting routine used to
analyze the data, and the likelihood that the data can be interpreted by a
complex modeling scheme. When aqueous humor or vitreous concentrations
of drug are measured over time, either mono- or biexponential equations
adequately describe the disposition of drug (5,13,14,22,4951).
Since it is relatively easy to remove cornea, conjunctiva, lens, iris/
ciliary body, and/or choroid tissues along with aqueous humor, the assignment of barriers and peripheral and/or reservoir compartments is not
particularly dicult with the exception of the cornea. Although the
cornea is clearly a physical barier to drug entry into the anterior chamber,
it is actually divided into three signicant kinetic barriers: the multilayered
lipophilic epithelium, the aqueous-like stroma, and the single-celled
lipophilic endothelium. The specic corneal permeability rate for each

Copyright 2003 Marcel Dekker, Inc.

146

Schoenwald

Figure 3 Classic pharmacokinetic schemes used to express compartmentalization

for topically applied ophthalmic drugs. PC = precorneal area; EP = corneal epithelium; AH = aqueous humor; P = peripheral or intraocular tissues accessible from
aqueous humor; ST/AH = stroma, endothelium and aqueous humor (i.e., kinetic
homogeneity), RE = reservoir (i.e., blood and systemic uids).

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

147

drug depends on the drugs partitioning properties and molecular weight

relative to the individual properties of each barrier (5662). The sum of the
resistances of each layer represents the apparent corneal permeability rate.
However, for pilocarpine a closer study of its kinetics in each barrier by Lee
and Robinson (44) indicated that the stroma and endothelium could be
more correctly associated with the aqueous humor compartment, and
only the corneal epithelium was a barrier for entry of drug into the anterior
chamber. Although the endothelium is lipophilic, it is only one cell thick and
is apparently not as signicant a barrier as the much thicker, more tortuous
epithelium with 9 or 10 layers.
This latter division of corneal layers into a single epithelial barrier and
the assignment of stroma and endothelium into a compartment along with
aqueous humor may be correct not only for pilocarpine, but also for other
hydrophilic drugs for which the epithelium is the major signicant barrier.
Water-soluble drugs of relatively small molecular weight (<500900) likely
penetrate the epithelium predominantly by paracellular pathways (30,53).
Consequently, as long as the drug is soluble, a high topical concentration
can be applied tot he cornea to promote a high penetration rate and overcome poor penetrability.

B.

Pharmacometric Measurements

Today computer curve-tting routines are designed for microcomptuers to

easily accommodate mathematical treatment of tissue concentrations of
drug over time (63). Regardless of the statistical approach employed to
arrive at the best representation of the data, it is desirable to characterize
and express absorption, distribution, and elimination of drugs. Over the
years, attempts that have been made to express pharmacokinetic drug behavior are the fraction absorbed (F), the rst-order rate absorption rate constant ka or the time to peak tp , the volume of distribution Vd , and
clearance. Clearance is expressed as either excretory (Cle , metabolic Clm
or their total, which represents elimination out of the body (Clt . In the
eye, Clt becomes ocular clearance, Clo . Noncompartmental approaches can
be used, for which mean residence times can be calculated to dene ocular
pharmacokinetics (21,42,51,52). On occasion, no parameter values are given
other than half-lives; however, tables are presented summarizing various
tissue levels over time (16,34,64). Strictly speaking, classical linear models
as well as noncompartmental approaches require that processes follow rstorder kinetics and the superposition principle. These assumptions are sometimes not conrmed with ocular pharmacokinetic data, mostly because of
intersubject variability, which is often signicant enough that reasonable

Copyright 2003 Marcel Dekker, Inc.

148

Schoenwald

experimentation is not adequate. Diculties in measuring ocular pharmacokinetic behavior are discussed in more detail in the following sections.
1. Absorption
As discussed more fully in another chapter, the rate and extent of ophthalmic
drug absorption is restricted largely by noncorneal absorption in the precorneal area and the drainage rate. The latter process, because of its rapidity,
limits the ocular contact time for drugs residing at the absorption site to
about 36 minutes, whereas the rate and extent of penetration across the
cornea is restricted by the physicochemical properties of the drug and/or its
formulation (1416,6569). Drug may also enter the ey by scleral penetration
(30,31,53,7072) or by uptake of drug into the anterior ciliary artery (31).
a. Fraction Absorbed. For drugs administered systemically, the fraction absorbed is determined by taking the ratio of the area under the plasma concentration-time curve or the total amount excreted in the urine for
a bolus intravenous dose and an oral dose, respectively. When applied in
an analogous manner to the eye, a small volume of drug is injected intracamerally using a small-bore needle and in a separate group of animals is
given topically, each followed by sampling of aqueous humor and measuring drug concentration over time (2,3). When given topically, the fraction
absorbed for 75 mg and 150 mg ocular doses of urbiprofen was 10 and
7%, respectively.
Patton and Robinson (46) estimated the fraction absorbed using a
dierent approach. Topical dosing studies were conducted in both anesthetized and awake rabbits and with the drainage ducts plugged and
unobstructed. Pilocarpine ntirate was measured as the disappearance of
drug from the tears of rabbit eyes whose ducts had been plugged and also
as appearance of drug in aqueous humor in both experiments. After correcting for dilution due to tear production and determining the areas under the
tear concentration-time curve (AUC), the fraction absorbed in anesthetized
rabbits whose ducts had been plugged was calculated as 0.0187. Equation
(1) was used to calculate F:
AUC

FD
k10 V

In Eq. (1), D is the instilled dose (25 mL), k10 is the loss of drug from the
precorneal area (1  102 min1 ), and V is estimated as 0.3 mL.
Extrapolating the results to normal rabbits, assuming a volume of distribution approximately equal to the aqueous humor volume, and assuming that
negligible drug distributes out of the aqueous humor were approximations
required in order to estimate F.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

149

Since drug absorption across the cornea and loss of drug from the
precorneal area are parallel loss processes, it was possible to conrm the
results of Patton and Robinson (46) by the use of Eq. (2):
F

k10
k10 ka

The fraction absorbed for a 25 mL instilled dose of pilocarpine nitrate in the

rabbit eye was estimated to be 12% (46,47). Using Eq. (2), Chiang and
Schoenwald calculated that for a 30 mL dose of 0.4% clonidine, an F of
1.6% was calculated for the rabbit eye (28).
Ling and Combs (73) compared the areas under the concentrationtime curves for ketorolac tromethamine 0.5% administered topically (50 mL)
and intracamerally, respectively. The ratio of the areas indicated a fraction
absorbed of 3.7%. An identical approach was used by Tang-Liu et al. (3) in
determining that 2.5% of the instilled dose of levobunolol was absorbed
after a topical dose (50 mL of 0.5%), similar in range to other drugs
administered topically.
b. Rate of Absorption. The loss of drug from the precorneal area is
a net effect of tear secretion, drainage, and noncorneal and corneal absorption rate processes. When drug concentration in aqueous humor is described biexponentially, the latter, shallower log-linear slope represents
elimination out of the eye, whereas the steeper slope is the net effect of
the precorneal processes. Because the drainage rate is approximately 100
times more rapid than ka (44,46), it is correct to assign the more shallow
log-linear slope from aqueous humor concentration-time curves as elimination from aqueous humor. Consequently, estimation of the steepest
slope and subsequent calculation of ka can be obtained from nonlinear
curve-tting techniques if the other precorneal processes are known.
Table 2 lists drugs for which ka has been estimated. These values, when
interpreted as half-lives, show how slowly drug is actually absorbed across
the cornea. Because of the very rapid loss of drug from the precorneal area,
and in particular the drainage rate, it is not surprising to nd that only a
small fraction of the instilled dose is actually absorbed across the cornea. In
addition, the cornea is relatively thick, variable in hydrophilic/lipophilic
properties for each layer, and small in surface area. These factors, along
with the rapid precorneal loss rate, combine to have a signicant eect on
the time to peak following topical instillation of drugs to the eye regardless
of the drugs physicochemical properties or its elimination rate from internal
eye tissues. In general, the time to peak is 2060 minutes for nearly all
ophthalmic drugs when instilled topically to the eye. This has been shown
by Makoid and Robinson (27), who determined the ophthalmic pharmaco-

Copyright 2003 Marcel Dekker, Inc.

150

Schoenwald

Table 2 Transcorneal First-Order Absorption Rate Constant and

Accompanying Half-Life for Corneal Absorption
Drug

ka (min1 )

t1=2 h

Ref.

Pilocarpine
Clonidine
Ketapron
Phenylephrine
Ibuprofen
Ibufenac
Ethoxyzolamide
Aminozolamide
2-Benzothiazolesulfonamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide
N-Methylacetazolamide
Acetazolamide

0.004
0.0014
0.0000462
0.00001
0.00129
0.00608
0.0015
0.0014
0.0013
0.0042
0.00126
0.00153

2.88
8.25
250
1155
9
19
7.7
8.25
8.88
5.37
9.17
75.5

44
28
52
120
51
51
76
78
76
76
121
121

kinetics of pilocarpine following topical application to the rabbit eye. From

their work an equation was developed for tp :
tp

lnkna =ka
Kna  Ka

In Eq. (3), kna and ka are the nonabsorptive loss rate constant and the
transcorneal absorption rate constant, respectively. Equation (3) assumes
that kna is much larger than uptake into the epithelium of the cornea
from the precorneal area. This assumption can be applied to most, if
not all, ophthalmic drugs, since kna is approximately twofold larger
than ka .
On occasion, a drug cannot be given systemically by a bolus intravenous dose, and therefore equations have been developed so that a slow
intravenous infusion could be used instead (74,75). Eller et al. (76) developed a topical infusion technique for estimating the ophthalmic pharmacokinetics of drugs, which was based upon noncompartmental methods. The
procedure consisted of maintaining a constant concentration of drug on the
cornea through the use of a plastic cylinder secured over the sclera, which
allowed only the cornea to be exposed to drug solution (see Table 3). A
volume of 0.7 mL is maintained over the cornea of an anesthetized rabbit
until steady-state ocular concentrations are reached. Rabbits are sacriced
at various time intervals, ocular tissues aer excised, and then assayed for
drug content. This topical infusion approach permits an estimate of ka the
apparent volume of distribution at steady state Vss , and ocular clearance

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

Table 3

151

Topical Infusion Method Depicting Wella and Equationsb

Ocular well used for topical infusion

Equations used for topical infusion method

ka

VA dCa =dtt
CW VW

3a

Cl0

K0 T
AUC

3b

VSS

K0 T AUMC K0 T 2

2AUC
AUC2

3c

K0 dCa =dtt VA

3d

Well is a xed over cornea of anesthetized rabbit; drug solution is maintained at a constant
concentration for 90160 minutes until steady-state concentrations of drug are reached in the
aqueous humor.
b
Ka = rst-order transcorneal rate constant; Va = volume of the anterior chamber; (dCa =dtI
= initial rate of appearance of drug in aqueous humor minus lag time; Cl0 = ocular clearance
(mL/min); K0 = constant rate of input into anterior chamber; AUC and AUMC = areas (to
innity) under the aqueous humor concentration  time curve and concentration  time-time
curve, respectively; CW = constant concentration maintained on the cornea over time (90160
min); VW = volume of drug solution maintained in well.

(Clo ). Figure 4 shows a semilogarithmic plot of the aqueous humor concentration of ibufenac and ibuprofen following topical infusion of 0.3% maintained on the cornea for 120 minutes (51). Table 3 lists the equations and
pharmacokinetic parameter values for drugs for which this procedure has
been used.
2.

Distribution

The volume of distribution represents a proportionality constant to relate

concentration to amount or dose and also as a relative measure of tissue
accumulation. It is the most dicult pharmacokinetic process to measure
primarily because the amuont of drug in the eye at any time is not known.
Aqueous humor is the circulating uid of the eye, and in order to measure
Vd accurately, an instantaneous input (i.e., intracameral injection) or an
injection at a known rate must be introduced. Estimates of apparent
volumes of distribution are lacking for ocular drugs. Table 4 lists values
estimated either from an intracameral injection (see gure 5) or from topical

Copyright 2003 Marcel Dekker, Inc.

152

Schoenwald

Figure 4 Average aqueous humor concentrations of ibufenac and ibuprofen following topical infusion of a 300 mg/mL solution to anesthetized rabbit eyes (n 6)
for 120 minutes. (Adapted from Ref. 51.)

infusion. Estimating the volume of distribution for amikacin and for chloramphenicol yielded values of 2.67 and 3.33 mL (5), which are about 8.5- and
10.5-fold larger than aqueous humor volume. Although these values are the
largest summarized in Table 4, they are relatively small relative to aqueous
humor volume. For systemically administered drugs, for which serum or
plasma is measured, Vd can easily be 100 times larger than plasma volume,
the latter of which is about 4.3% of body weight.
Nevertheless, tissues in the anterior and posterior chamber are adjacent to circulating aqueous humor and therefore readily available for transfer of drug. Also, protein concentration in aqueous humor is about 10% of
proteins in plasma, so that it is not likely that tissue distribution is prevented
because of binding to components in aqueous humor. Consequently, it is
surprising that Vd is so small. Although accurate determinations of Vd are
necessary for estimates of optimal multiple dosing regimens, few are available, mainly because of the pharmacometric diculties in obtaining such
measurements.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

Table 4

Volumes of Distribution Vd for Drugs of Ophthalmic Interest

Drug
Pilocarpine
Clonidine
Phenylephrine
Flubiprofen
Ibufenac
Ibuprofen
Ketanserin
Amikacin
Chloramphenicol
Levobunolol
Dihydrolevobunolol
Phenylephrine
Ketorolac tromethamine
Ethoxzolamide
Aminozolamide
2-Benzothiazolesulfonamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide
N-Methylacetaxolamide
Acetazolamide
a

153

Vd (mL)

Method

Ref.

0.58
0.53
0.42
0.62
0.21
0.53
0.97
2.67
3.33
1.65
1.68
0.42
1.93
0.28
0.53
0.24
0.33
0.42
0.47

EXTa
SSb
SS
EXT
EXT
EXT
EXT
EXT
EXT
EXT
EXT
SS
EXT
SS
SS
SS
SS
SS
SS

47
28
120
2
51
51
52
5
5
3
3
120
73
57
78
76
76
121
121

EXT = extrapolated method; determined by extrapolating log-linear elimination

phase t0 C at t0 and dividing into the intracameral dose.
SS = steady-state method; determined by maintaining a constant concentration of
drug on the cornea of an anesthetized rabbit and measuring aqueous humor
concentrations of drug over time during infusion and postinfusion. See Table 3 for
equatiaon for VSS .

The pharmacokinetic parameter, Vd , also provides a general assessment of distribution. However, it does not dierentiate between sites that
are active and those that are either devoid of activity of potentially responsible for side eects. In lieu of making these measurements, it is not dicult
to measure eye tissue concentrations over time, which provides a direct
indication of whether or not drug distributes to the active site. In contrast
to the paucity of pharmacokinetic measurements for ophthalmic drugs,
tissue concentrations of drug have been widely reported for many years.
With recent improvements in assay methodology, tissue proles over time
are routinely measured for aqeous humor and cornea, but less frequently for
lens, iris/ciliary body, conjunctiva, choroid, lacrimal gland, and/or retina.
In the interest of therapy, it is of importance to determine the percent
of the dose that reaches a particular tissue and to determine if the concentrations that are reached meet and/or exceed the minimum therapeutic

Copyright 2003 Marcel Dekker, Inc.

154

Schoenwald

Figure 5 Average (n 5) aqueous humor concentrations of amikacin (2 mg/mL)

and chloramphenicol (5 mg/mL) following direct injection of into the anterior chamber of the rabbit eye (10 mL). (Adapted from Ref. 5.)

range. For example, in a recent study by Morlet et al. (77), ciprooxacin,

which is active against a wide variety of bacteria, was given orally in doses of
750 or 1500 mg prior to eye surgery and then assayed for levels in blood,
aqueous, and vitreous humor during surgery. Population pharmacokinetics
was applied to the data, and the mean half-lives for loss of ciprooxacin
were determined for aqueous and vitreous humor, which were found to be
3.5 and 5.3 hours, respectively. At steady state, the concentrations of drug in
aqueous and viterous were 23 and 17% of serum levels, respectively. The
authors concluded that the use of oral ciprooxacin would be below the
minimum inhibitory concentration (MIC) of 0.5 mg/mL in treating bacterial
endophthalmitis or for use for perioperative prophylaxis.
It is reasonable to expect that if drug is instilled topically to the eye
and no unusual tissue anity occurs for a particular drug, the order of
decrasing tissue concentrations are as follows: cornea > conjunctiva >
aqueous humor > iris/ciliary body > lens, vitreous, and/or choroid/retina.
Many studies (37,51,52,67,78,80,81) show that iris/ciliary body concentrations are higher than aqueous humor concentrations, even though the dose
is instilled topically to the cornea. A number of reasons have been suggested

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

155

to account for this phenomenon. For example, the drug may distribute
extensively to iris/ciliary body, but drug may equilibrate rapidly with aqueous humor so that the elimination rate during distribution equilibrium is
the same as elimination from aqueous humor. If this explanation is correct,
the iris/ciliary body would have a large capacity for drug but not exhibit
an unusually high binding anity. It is also conceivable that the binding
anity as well as its capacity is very high so that the elimination rate from
iris/ciliary body and aqueous are not the same but drug remains in the
former tissue much longer.
This latter observation was observed by Putnam et al. (78) for aminozolamide, a topically active carbonic anhydrase inhibitor (CAI), which
showed a relatively slow elimination for drug as well as metabolite, 6-acetamido-2-benzothiazolesulfonamide, from iris/ciliary body compared to aqueous humor following topical administration to the rabbit eye. In another
example, a derivative of methazolamide, 5-acetoxyacetylimino-4-methyl
2 -1,3,4,-thiadiazoline-2-sulfonamide, is an ester prodrug that lowered
intraocular pressure in albino New Zealand rabbits but was found to be
active in pigmented Dutch Belt rabbits (79). From various studies it was
concluded that melanin binding in the iris, and to some extent metabolism
occurring only in the Dutch Belt rabbits, were valid explanations.
Alternatively, support for the possibility of high concentrations of
drug in the iris-ciliary body after topical application come from the likelihood that drugs reach iris-ciliary body by scleral absorption (i.e., paracellular penetration) and/or uptake by vessels imbedded in the conjunctiva that
deposit drug. Certain drugs, when applied topically to the eye, may be
preferentially absorbed by the sclera, as opposed to the cornea, and enter
the iris/ciliary body without rst entering the aqueous humor. In a study by
Ahmed et al. (71), corneal and scleral penetration were determined for
propranolol, timolol, nadalol, penbutolol, sucrose, and inulin. The results
of the study showed that the outer layer of the sclera provides much less
resistance to penetrabiilty for hydrophilic drugs than the corneal epithelium.
Hamalainen et al. (30) estimated that the conjunctival and scleral tissues
were 1525 times more permeable than the cornea and that conjunctival
permeability was less aected by molecular size than the cornea. In addition,
the total paracellular space of the conjunctiva was estimated to be 230 times
greater than that in the cornea. For lipophilic drugs, such as propranolol,
timolol, and penbutolol, the dierence in penetrability between the tissues
has been estimated to be similar. In a study by Schoenwald and Zhu (52),
ketanserin and its metabolite, ketanserinol, were infused topically to
anesthetized rabbits for 120 minutes. Drug was instilled either within the
well and, therefore, in direct contact with the cornea, or outside the well,
excluding the cornea but allowing drug to come in contact with conjunctiva

Copyright 2003 Marcel Dekker, Inc.

156

Schoenwald

(see Fig. 3). The results showed that irisciliary body concentrations were
three and two times higher for ketanserin and ketanserinol when applied
exclusively to the sclera as opposed to the cornea (52).
In a similar study by Chien et al. (79), various anterior chamber tissue
levels were measured over time for clonidine, p-aminoclonidine, and a
6-quinoxalinyl derivative of chlonidine (AGN 190342) using the infusion
method. Whenever drug was maintained on conjunctiva or cornea for a
period of 60 minutes, tissue concentration followed a clear trend. the order
of highest to lowest concentration of drug following conjunctival contact was
conjunctiva > cornea > ciliary body > aqueous humor. When drug solution was in contact with cornea only, the order was cornea > aqueous humor
> ciliary body > conjunctiva. From the results of recent studies (30,31,79
82), various pathways of ocular absorption have been proposed, as summarized in Figure 6.
Although many anterior and, to a lesser extent, posterior tissues have
been measured for drug concentration over time following topical or systemic administration, little denitive information exists regarding the ability
of tissues to accumulate drug through either partitioning or binding of drug.
Of critical importance is the iris/ciliary body, which is the biophasic location
for many pharmacological responses acting on the eye. Also of interest is the
lens, since drug accumulation may be responsible for inducing cataract
formation. Both of these tissues are easily removed and often treated as a
kinetically homogeneous tissue, but drug concentrations within these tissues
are likely to be quite variable in concentration since these tissues are not
anatomically homogenous.
For example, iris tissue in the rabbit eye is porous and highly vascular
with a large surface area in direct contact with aqueous humor; consequently,
distribution equilibrium between iris and aqueous humor should occur
rapidly. Also, as the iris becomes darker, the capacity for pigmented iris to
bind to catecholamines increases. This was also observed for carbonic anhydrase inhibitors (CAIs) (82). Dark-eyed individuals have a delayed onset and
a reduced but prolonged response to catecholamines, presumably due to the
reservoir eect that the pigmented iris has on ocular disposition of catecholamines. Latanoprost, an isopropyl ester of prostglandin F2
, can darken the
iris in susceptible individuals due to increased pigmentation of the anterior
surface of the iris (83). In contrast to iris tissue, the ciliary body epithelial
layer of the pars plicata, containing a high concentration of carbonic anhydrase and presumed to be the active site for CAIs, is not easily reached via
topical instillation (84), requiring a lipophilic drug substance to readily penetrate cellular membranes. Although the separation of these tissues from one
another would provide useful information, it is usually not done because of
the diculty in separating one tissue entirely from the other.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

157

Figure 6 Probable ocular penetration routes for topically applied drugs: (1) transcorneal pathways; (2) noncorneal pathways; (3) systemic pathways.

The entire lens is also easy to remove during kinetic studies; nevertheless, the lens should not be treated as a kinetically homogeneous
tissue. Maurice and Mishima (85) reported that uorescein, a water-soluble dye, spreads laterally and rapidly in the outer layers of the lens but

Copyright 2003 Marcel Dekker, Inc.

158

Schoenwald

does not diuse readily into the lens nucleus. The initial barrier for entry
into the lens is the epithelium, which is a single layer of cells lying just below
the anterior lens capsule. Internal to the epithelial layer are densely packed
lens bers and the nucleus. Because it consists of hard, condensed cellular
material, the nucleus possesses high tortuosity and low porosity. With age,
old bers are not disposed of but become compressed centrally to form a
larger, less elastic nucleus. Consequently, drugs are not expected to readily
penetrate the lens interior. Ahmed et al. (86) came to this conclusion from a
study of the diusion of timolol through the lens bers of rabbit lens. A
pharmacokinetic model for drug entering aqueous humor and distributing
into the lens included the capsule and its epithelium as a compartment adjacent to aqueous, but excluded the lens interior since drug did not reach
signicant levels in this region. It was concluded that unless a steady state
of drug is present and maintained through repetitive dosing, signicant accumulation would not occur. Drug elimination from the eye is too rapid to
allow for signicant concentrations to accumulate into the interior of the lens
from a few doses.
3. Elimination
Drug is eliminated from the anterior chamber, at least in part, by aqueous
humor turnover, which in the rabbit eye is 1.5% of the volume of the anterior
chamber per minute or expressed as a half-life, 46.2 minutes (87). Although
clearance is of greatest interest for drugs used systemically, half-life representing loss from aqueous humor is the most common pharmacokinetic parameter measured following topical administration to the eye. Table 5 lists halflives for drugs of ophthalmic interest. Surprisingly, nearly all the drugs fall
within a range of 0.63 hours, which is less than when these same drugs are
studied systemically. Either there are few tissue-binding sites in the eye to
allow for drugs to resist clearance, or the pathways by which drugs are eliminated are very ecient. The latter may be the most likely explanation, since
aqueous humor volume is relatively large compared to ocular tissue volume.
Elimination can also be expressed as a clearance, and if the anterior
chamber of the rabbit contains about 0.311 mL, an average aqueous humor
clearance due to bulk ow becomes 4.67 mL/min based upon Eq. (4). Ocular
clearances Clo can be calculated from the following equations:
Clo Ke Vd
Ko T
Clo
AUCINF
DIC
Clo
AUCINF

Copyright 2003 Marcel Dekker, Inc.

4
5
6

Ocular Pharmacokinetics and Pharmacodynamics

159

where Kc represents the rst-order elimination rate constant out of aqueous

humor, Vd is the apparent volume of distribution for the eye, Ko is the
constant rate input into the anterior chamber, T is the time for the constant
rate input, DIC is an intracameral dose, and AUCINF is the area (to innity)
under the aqueous humor concentration-time curve.
Each equation above depends on assumption that require experimentation to be carefully planned. In Eq. (4), Ke can be easily obtained. It is
calculated from the latter linear slope of the logarithm of drug concentration
in aqueous humor measured over time or obtained from any other tissue
concentration in distribution equililbrium with aqueous humor. The Vd
term in Eq. (3) has been correctly determined for the eye by either of two
methods. One method, the topical infusion technique (28,37,52,76), has been
discussed previously and is the basis for Eq. (5), also shown in Figure 4. The
other method for determining Vd is based upon measuring drug concentration in aqueous humor over time following an intracameral injection of a
very small volume (5 mL) of drug solution. Whenever an intracameral injection is made, there is concern that drug elimination can be altered because of
a breakdown of the blood-aqueous barrier. However, Mayers, Miller, and
coworkers (4,5,88,89), as well as Tang-Liu and coworkers (2,3), have established that for intracameral injections of 5 mL solutions containing various
drugs, no signicant breakdown of the blood-aqueous barrier had occurred
since protein concentration was <1 mg/mL. Equation (6), sometimes
referred to as a dose-area determination, was used by Tang-Liu et al.
(2,3) to calculate Vd for urbiprofen and levobunolol.
Table 6 contains Clo values for those drugs for which accurate determinations have been made. Values range from 13 to 28.7 mL/min, which are
2.8-6.1 times higher than aqueous humor clearance, suggesting pathways of
elimination other than aqueous turnover. The two most likely alternate
routes of elimination are metabolism and systemic uptake by the vascular
tissues of the anterior uvea. However, accumulation and retention by the
lens (over the time course of the experiment), as well as back diusion into
the cornea and tears followed by subsequent drainage, are all minor routes
that are not likely to signicantly contribute to ocular clearance values.
C.

Metabolic Models

Although the eye is not a primary drug-metabolizing organ, a knowledge of

metabolic pathways in the eye has become increasingly important in order
to optimize drug action and therapeutic eect. Esterases have been studied
most extensively (9093), no doubt because of their importance in the development of prodrugs for use in the eye. Prodrugs that have shown a signicant improvement in corneal penetrability are phenylephrine (94,95),

Copyright 2003 Marcel Dekker, Inc.

160

Schoenwald

Table 5 Aqueous Humor Half-Livesa of Drugs Administered to

the Rabbit Eye Either Topically, Intracamerally, or
Subconjunctivally
Drug
Cyclosporine
Dapiprazole
Imirestat
6-Mercaptopurine
Falintolol
Fusidic acid
Suprofen
Dorzolamide
Histamine
Cimetidine
Ketorolac tromethamine
Benzolamide
Ceftazidime
Gentamicin
L-Alphamethyldopab
1-643,799c
Diclofenac
Diclofenacd
Diclofenace
Diclofenacf
Flurbiprofen
Fluorometholone acetate
Cefamandole
Thiamphenicol
Phenylephrine
D-Alphamethyldopab
Amikacin
Timolol
Timolol
Timolol
Lincomycin
6-Amino-2-benzothiazolesulfonamide
L-662,583g
Acetbutolol
Cefsulodin
Fluorouracil
Chloramphenicol
Dihydrolevobunolol
N-Methylacetazolamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide

Copyright 2003 Marcel Dekker, Inc.

Half-life (h)

Ref.

24.9
5.8
4.75
4.6
3.0
2.8
2.6
2.4
2.2
2.2
2.1
2.0
2.0
1.9
1.8
1.8
1.7
5.0
1.0
0.2
1.7
1.5
1.5
1.5
1.4
1.4
1.4
1.2
0.84
0.2
1.2
1.15
1.11
1.1
1.0
1.0
1.0
0.98
0.98
0.97

19
122
123
124
123
125
126
144
128
127
73
128
129
130
131
132
133
38
38
38
134
20
135
49
120
131
5
136
137
40
138
78
139
137
140
141
5
3
121
76

Ocular Pharmacokinetics and Pharmacodynamics

Table 5

161

Continued

Drug
Propranolol
Propranol
6-Acetamido-2-benzothiazolesulfonamide
L-650,719h
Ketanserrin
Saperconazole
Tobramycin
Triuoromethazolamide
Pilocarpine
Methazolamide
Cefotaxime
Bufuralol
Levobunolol
Ethoxzolamide
Ethoxzolamide
Pyrilamine
Verapamil
Clonidine
Acetazolamide
Tilisolol
Ibuprofen
2-Benzothiazolesulfonamide
Ibufenac

Half-life (h)

Ref.

0.96
1.2
0.93
0.81
0.86
0.86
0.75
0.78
0.72
0.58
0.57
0.50
0.67
0.63
0.23
0.61
0.55
0.49
0.35
0.35
0.31
0.29
0.28

10
12
78
132
64
52
20
128
142
128
143
137
3
76
128
127
36
28
121
40
51
76
51

Concentrations from the last two to four times intervals were used in
calculating t1=2 values.
b
A small dose dependency was observed (statistically NS); IV dose =
10 mg/kg.
c
6-Hydroxybenzol[b]thiophene-2-sulfonamide.
d
Normal rabbit eyes were used.
e
Keratitis model.
f
Uveitis model.
g
6-Hydroxybenzol[b]thiophene-2-sulfonamide.
h
6-Hydroxy-2-benzothiazidesulfonamide.

timolol (96), pilocarpine (97,98), and idoxuridine (99). Two sterioisomers,

both isopropyl ester prodrugs of prostaglandin F2
, latanoprost, and unoprostone, have approved been recently for clinical use (83). Acetyl-,
butyryl-, and carboxylesterases have been identied in the pigmented rabbit
eye (90) and are responsible for rapid conversion of ester prodrugs to the
active drug species.
Bodor and coworkers (100105) have applied knowledge of metabolic pathways in the eye to alter a drugs elimination pathway. New

Copyright 2003 Marcel Dekker, Inc.

162

Schoenwald

Table 6 Ocular Clearances (Cl0 a for Drugs of Ophthalmic

Interest
Drug
Pilocarpine
Clonidine
Ketanserin
Phenylephrine
Flurbiprofen
Levobunolol
Dihydrolevobunolol
Propranolol
Phenylephrine
Ketorolac tromethamine
Ibufenac
Ibuprofen
Ethoxolamide
2-Benzothiazolesulfonamide
6-Hydroxyethoxy-2-benzothiazolesulfonamide
N-Methylacetazolamide
Acetazolamide

Cl0 (mL/min)

Ref.

13.0
14.9
13.6
14.6
14.4
28.7
19.7
0.86
14.6
11.0
18.7
8.4
9.0
1.15
3.0
1.56
5.27

35
28
52
120
2
3
3
12
120
73
51
51
76
76
76
121
121

Cl0 calculated from Eq. 46 in text.

drugs, referred to as soft drugs, are designed for rapid metabolism and
subsequent conversion to an active species before being eliminated from
the eye. For example, an adamantylethyl of metaprolol has shown a prolonged and greater reduction of IOP and also a reduced potential to cause
tachycardia when compared to timolol in the rabbit eye (103). In another
example, timolol maleate was oxidized to a keto analog and coupled with
either hydroxylamine or methoxyamine. The resulting drug candidates,
timolone oxime and timolone methoxime, showed greater reduction of
IOP than timolol in rabbits administered the drugs topically and did not
show cardiovascular eects when administered intravenously to rabbits or
rats (104).
Many enzyme systems have been identied within ocular tissues. As
summarized by Plazonnet et al. (15), these are catechol-O-methyltransferase,
monoamine oxidase, steroid 6-betahydroxylase, oxidoreductase, lysosomal
enzymes, peptidases, glucuronide and sulfate transferase, and glutathioneconjugating enzymes. Arylamine acetyltransaferase activity was demonstrated by Campbell et al. (106) in an anterior chamber tissues using paminobenzoate, aminozolamide, and sulfamethazine as substrates. The
rank order of arylamine acetyltransferase activity regardless of substrate
was liver > iris/ciliary process > corneal epithelium > stroma/endothe-

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

163

lium. Although fast- and slow-acetylating rabbits, classied with respect to

their rate of hepatic acetylation, metabolized aminozolamide at dierent
rates in liver tissue, no dierences between phenotypes were observed for
aminozolamide either in ocular disposition or in decline of IOP following
topical instillation.
The distribution of ketone reductase activity in anterior chamber tissues was determined by Lee et al. (107) in order to explain the metabolism of
levobunolol to dihydrolevobunolol. The rank order of activity was corneal
epithelium > iris/ciliary body > conjunctiva > lens. No activity was
detected in the tears, corneal stroma, sclera, or aqueous humor. Bovine
ciliary body apparently contains aldehyde oxidase, which is responsible
for the reduction of a number of compounds: N-oxide, hydroxamic acid,
and sulfoxide and nitro compounds (108).
Metabolic models have been devised for ketanserin (52) and levobunolol (3) (Fig. 7). It is important to study ocular metabolism so that drug
disposition is better understood and also in anticipation of a new drug
candidate. For example, an endogenous corneal epithelial arachidonic
acid metabolite formed by the cytochrome P450 system and a potent
inhibitor of Na+-K+-ATPase was found to signicantly lower IOP in the
rabbit eye (109).

Figure 7 Metabolic schemes devised for ketanserin and levobunolol following

absorption in the eye. (Adapted from Refs. 3 and 52.)

Copyright 2003 Marcel Dekker, Inc.

164

Schoenwald

In general, the iris/ciliary body and the epithelium of the cornea

appear to be anterior chamber tissues with the greatest capacity for metabolism.

V. PHARMACODYNAMICS
Because ophthalmic drugs are relatively potent and because of the diculty in routinely measuring ophthalmic pharmacokinetics, the measurement of pharmacological responses in the animal or human eye has
become a convenient and necessary aid in anticipating the clinical eect
of an ophthalmic drug. The principal shortcoming to the use of pharmacological measurements either for optimizing therapy or for use as an aid
in the development of new ophthalmic drugs is that the same dose often
produces a dierent intensity of eect in dierent individuals. This variability occurs because of dierences in dose-response relationships as well as
dierences in ocular pharmacokinetic behavior between individuals. Other
factors that contribute to variability are eye pigmentation, whether or not
an individual wears contact lenses, allergies to drugs or preservatives, and
a number of physiological factors, which also determine intrasubject variation, such as eye discomfort leading to induced tearing, changes in blood
pressure, hormonal concentrations, genetic dierences, and/or changes in
autonomic tone.
Although many ophthalmic pharmacological responses, such as miosis, mydriasis, IOP, icker fusion, tear secretion, and aqueous turnover, can
be easily measured over time, variability has been a major reason for the
lack of a mathematical model. Pilocarpine and other muscarinic agonists
that constrict the pupil have been studied extensively with respect to their
miotic response. The response of an isolated strip of human sphincter muscle to carbachol and pilocarpine follows a typical sigmoidal-shaped dose or
concentration-response relationship (110). Figure 8 is a recent representation of a concentration-response curve for S(+)urbiprofen to COX-1 and
COX-2 inhibition activity in homogenate human iris, with the latter treated
with lippolysaccharide (111). Mishima (111) showed that this response,
which represents one molecule binding competitively to a single muscarinic
receptor, can be described as follows:
R
qC
Rmax  R

In Eq. (7), Rmax is the largest response that can be achieved by drug when all
receptors are occupied, C is the drug concentration present in the incubation
media, and q is a proportionality constant, which is equal to 1 if a single

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

165

Figure 8 Concentration-response curves for inhibitin of the COX-1 (untreated)

and COX-2 (treated with lipopolysaccharide) activity in human eye homogenate
(n 46). (Adapted from Ref. 112.)

molecule binds to a single receptor. Maurice and Mishima (85) further

reported that Eq. (7) also correctly characterized the miosis of carbachol
on the sphincter muscle of the cat, mydriasis of isoproterenol on bovine and
rabbit sphincter muscle, mydriasis of 1-epinephrine and atropine on guinea
pig iris, and mydriasis of epinephrine and acetylcholine on cat iris.
Gabrielsson and Weiner (113) analyzed the miotic response in the cat
eye over time after application of three dierent doses of latanoprost. A
simple one-compartment model was assumed with a rst-order input into
the eye with the biophase (site of the miotic response) associated with the
central compartment. The concentration in the biophase was assumed to act
directly on the response according to the sigmoid Emax model equation:
Et E0

Emax C n
n Cn
EC50

Copyright 2003 Marcel Dekker, Inc.

166

Schoenwald

In Eq. (8), E0 is the baseline value for the pupil size, Emax is the maximum
eect, C is the concentration of drug in the biophase estimated from a
kinetic model equation, EC50 is the concentration in the biophase at halfmaximal eect, and n is the sigmoidicity factor.
Smolen and coworkers (114118) developed a mathematical model for
which pharmacological response intensities were transformed into biophasic
drug levels for tropicamide, tridihexethylchloride, carbachol, and pilocarpine. The transformation was accomplished with the use of a precisely
determined dose-response curve. The method requires a graded response
that can be measured over time and expressed according to:
I

K1 Q0B
PQ0B
1 K2 Q0B

where I is the intensity of response at any time and Q0 B is the concentration

of drug in the biophase at the same time of measurement (normalized for
dose: Q0B QB =D) and K1 , K2 , and P are constants derived from the tting
procedure.
Figure 9 shows the kinetics of the expected biophasic concentration of
a drug after tting miosis vs. time data to a kinetic model. After calculating
the expected biophasic drug concentrations to Eq. (8) or (9), the response

Figure 9 Observed (lled circles) and predicted (solid line) kinetics of a drug in the
eect compartment following topical administration of three dierent doses.
(Adapted from Ref. 113.)

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

167

intensities (I) or eect (E) can be t to the concentration of drug estimated

to be in the biophase. Figure 9 represents a t to Eq. (8) for S(+)urbipofen
inhibition of prostaglandin E2 to the COX-1 and COX-2 isoenzymes (113).
Use of Eq. (9) requires that the following assumptions are valid: (1) the
same biophasic concentration of drug produces the same intensity of
response (i.e., nonhysteresis), (2) binding to the receptor site is rapid and
reversible, and (3) the pharmacokinetic processes are rst order and therefore do not dier with dose. Application of this approach permits the optimization of a delivery system for an ophthalmic drug, insight into a drugs
kinetics of response, and determination of a products biophasic bioavailability. The later term refers to the drugs absorption and disposition to the
ocular site of action (114).
A mydriatic tolerance of the pupil response has been interpreted for
phenylephrine from the application of the classic pharmacodynamic Emax
model (119):



Emax
 1 Ca t
KM
10

Et
where Km0 is the drug concentration in aqueous humor required to produce
one half of the maximal mydriatic response of phenylephrine (1/2 Emax ), Et
is the change in the mydriatic response from baseline at any time, and Cat
is the concentration of drug in aqueous humor at the same time of measurement of E. In Eq. (10), Km0 is linearly related to Cat for a drug that does
not develop tolerance, but if mydriatic tolerance is developed following
topical instillation, the value of Km0 will vary with time. Chien and
Schoenwald (119) measured the aqueous humor concentration of phenylephrine and its corresponding mydriatic response over time in the rabbit
eye following a 10 uL topical instillation of phenylephrine HCl viscous
solution (10%). A clockwise hysteresis eect of mydriasis vs. Ca t is
shown in Figure 10, which illustrates the development of tolerance over
time. At the 240-minute time interval, the phenylephrine was twofold higher
than that measured at the 20-minute time interval. However, at these two
time intervals, the mydriatic response was similar. Km0 was linear up to 90
minutes, but steadily increased beyond 90 minutes through 240 minutes
postinstillation.
Nonlinear pharmacokinetic behavior may be more common than has
been observed since many ophthalmic drugs are known to alter physiological processes. Drugs that aect aqueous humor turnover or blood ow
within the iris/ciliary body would be expected to show some degree of nonlinear pharmacokinetic behavior. Nevertheless, the inability to routinely
measure concentrations of drug in the human eye provides a strong incentive to continue exploring the use of pharmacological response intensities to

Copyright 2003 Marcel Dekker, Inc.

168

Schoenwald

Figure 10 Average mydriasis measurement vs. aqueous humor concentrations of

phenylephrine (clockwise hysteresis loop) following a 10 mL topical instillation of
10% phenylephrine HCl viscous solution. (Adapted from Ref. 119.)

either optimize therapy or povide a screning tool for use in developing new
ophthalmic agents.

VI.

CONCLUSION

It is important that the ocular pharmacokinetic behavior for drugs of

ophthalmic interest be determined. Without a detailed knowledge of these
processes, we must rely on the screening of new ophthalmic drugs intended
for systemic use. For many years, drugs originally intended for systemic use
and screened for use in the eye have provided good leads but were not
optimized for ophthalmic potency or structurally altered to be devoid of
systemic side eects when instilled topically to the eye. More recently, dipivefrin, latanoprost, unoprostone, dorzolamide, and brinzolamide were
developed exclusively for the eye. A knowledge of ocular pharmacokinetic
behavior is critical for optimizing therapeutic regimens following single and,
particularly, multiple doses. It is not for lack of interest that prevents study

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

169

of the ocular pharmacokinetic behavior of drugs, but because of the technical problems that have yet to be solved, the application of pharmacokinetics to the design of dosing regimens remains a goal. Specically, a reliable
method for serially measuring drug levels in aqueous humor without injury,
pain, or risk of infection is yet to be developed for widespread use in the
human eye.
At the present time, pharmacokinetic determinations in the human eye
can only be made in subjects undergoing eye surgery. The patient is given
the ophthalmic drug prior to entering surgery, and a sample of aqueous
humor (or other tissue) is removed when it does not interfere with the
surgery. In these experiments the time interval may not be measured accurately, and each patients sample represents only one time interval.
Consequently, a number of patient determinations at dierent time intervals
must be averaged.
In lieu of human measuerments, an animal model is needed that is
readily available and can accurately predict human ocular pharmacokinetics. The ideal animal model, whether a rabbit or monkey eye, should
allow for accurate predictions in the human eye so that dosing regimens
can be predicted. The pharmacokinetic measurements should be capable of
assessing absorption (rate and extent), distribution, and elimination in
volunteers and, in particular, patients. The model may be mathematically
based but useful without extensive training and/or requiring specialized
computer programs for its use. It should also allow for accurate predictions
even though the drug itself may alter its own pharmacokinetic or pharmacodynamic processes over time, which may also complicate pharmacokinetic
modeling since nonlinearity requires a unique set of assumptions for each
drug. No doubt as these limitations are solved, the application of pharmacokinetics will become as extensively studied and applied to the human eye
as it is now to drugs administered systemically. In the interim, it is useful to
study various routes of administration and to measure tissue levels of drug
for the purpose of determining if the minimum eective concentration is
reached.

REFERENCES
1. McDonald, T. O., and Shadduck, J. A. (1977). Eye Irritation.
Determatoxicology and Pahrmacology (H. Maibach and F. N. Marzulli,
eds.). J. Wiley & Sons, New York, pp. 139191.
2. Tang-Liu, D. D. S., Liu, S. S., and Weinkam R. J. (1984). Ocular and systemic
bioavailability of ophthalmic urbiprofen. J. Pharmacokin. Biopharm., 12:611.

Copyright 2003 Marcel Dekker, Inc.

170

Schoenwald

3. Tang-Liu, D. D. S., Liu, S. S., Ne, J., and Sandri, R. (1987). Disposition of
levobunolol after an ophthalmic dose to rabbits. J. Pharm. Sci., 76:780.
4. Miller, M. M., Madu, A., Samathanam, G., Rush, D., Madu, C. N.,
Mathisson, K., and Mayers, M. (1992). Flexoxacin pharmacokinetics in aqueous and vitreous humors determined by using complete concentration-time
data from individual rabbits. Antimicrob. Agents Chemother., 36:32.
5. Mayers, M., Rush, D., Madu, A., Motyl, M., and Miller, M. H. (1991).
Pharmacokinetics of amikacin and chloramphenicol in the aqueous humor
of rabbits. Antimicrob. Agents Chemother., 35:1791.
6. Bito, L., Davson, H., Murray, L. M., and Snider, N. (1966). The concentration of free amino acids and other electrolytes in cerebrospinal uid, in vivo
dialysate of brain, and blood plasma of dog. J. Neurochemistry 13:1056.
7. Wang, Y., Wong, S. L., and Sawchuk, R. J. (1993). Microdialysis calibration
using retrodialysis and zero-net ux: application to a study of the distribution
of zidovudine to rabbit cerebrospinal uid and thalamus. Pharm. Res.,
10:1411.
8. Sato, H., Fukuda, S., Inatomi, M., Koide, R., Uchida, N., Kanda, Y., Kiuchi,
Y., and Oguchi, K. (1995). Pharmacokinetics of noroxacin and lomeoxacin
in domestic rabbit aqueous humor analyzed by microdialysis. J. Jpn.
Ophthalmol., 7:513.
9. Ohtori, R., Sato, H., Fukuda, S., Ueda, T., Koide, R., Kanda, Y., Kiuchi, Y.,
and Oguchi, K. (1998). Pharmacokinetics of topical beta-adrenergic antagonists in rabbit aqueous humor evaluated with the microdialysis method. Exp.
Eye Res., 66:487.
10. Rittenhouse, K. D., Peier, R. L., and Pollack, G. M. (1999). Microdialysis
evaluation of the ocular pharmacokinetics of propanolol in the conscious
rabbit. Pharm. Res., 16:736.
11. Ben-Nun, J., Cooper, R. L., Cringle, S. J., and Constable, I.J. (1988). A new
technique for in vivo intraocular pharmacokinetic measurements. Arch.
Ophthalmol, 106:254.
12. Rittenhouse, K. D., Peier Jr., R. L., and Pollack, G.M. (1998). Evaluation of
microdialysis sampling of aqueous humor for in vivo models of ocular absorption and disposition. J. Pharm. Biomed. Anal., 16:951.
13. Hughes, P. M., Krishnamoorthy, R., and Mitra, A. (19xx.) Vitreous disposition
of two acycloguanosine antivirals in the albino and pigmented rabbit models: a
novel ocular microdialysis technique. J. Ocul. Pharmacol. Ther., 12:209.
14. Schoenwald, R. D. (1990). Ocular drug delivery. Pharmcokinetic considerations. Clin. Pharmacokinet., 18:255.
15. Plazonnet, B., Grove, J., Durr, M., Mazuel, C., Quint, M., and Rozier, A.
(1977). Pharmacokinetics and biopharmaceutical aspects of some anti-glaucoma drugs. In Ophthalmic drug Delivery Biopharmaceutical Technological and
Clinical Aspects (M. F. Saettone, M. Bucci, and P. Speiser, eds.). SpringerVerlag, Berlin, pp. 118139.
16. Lee, V. H. L., and Robinson, J. R. (1986). Review: topical ocular drug delivery: recent developments and future challenges. J. Ocular Pharmacol. 267.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

171

17. ODay, D. M., Head, W. S., Foulds, G., Robinson, R. D., Williams, T. E.,
and Ferraina, R. A. (1994). Ocular pharmacokinetics of orally administered
azithromycin in rabbits. J. Ocul. Pharmacol., 10:633.
18. Berthe, P., Baudouin, C., Garrao, R., Hofmann, P., Taburet, A.-M., and
Lapalus P. (1994). Toxicologic and pharmacokinetic analysis of intravitreal
injections of foscarnet, either alone or in combination with ganciclovir. Invest.
Ophthalmol. Vis. Sci., 35:1038.
19. Oh, C., Saville, B. A., Cheng, Y.-L., and Rootman, D. S. (1995). A compartmental model for ocular pharmacokinetics of cyclosporine in rabbits. Pharm.
Res., 12:433.
20. Schoenwald, R. D., Harris, R. G., Turner, D., Knowles, W. and Chien, D. S.
(1987). Ophthalmic bioequivalence of steroid/antibiotic combination formulations. Biopharm. Drug Disposit., 8:527.
21. Tang-Liu, D. D. S., and Burke S. S. (1988). The eect of azone on ocular
levobunolol absorption: calculating the area under the curve and its standard
error using tissue sampling compartments. Pharm. Res., 5:238.
22. Pettersson, K. J., and Nordgren, T. (1989). Determination of phosphonoformate (foscarnet) in biological uids by ion pair reversed-phase liquid chromatography. J. Chromatogr., 488:447.
23. Chrai, S. S., Makoid, M. C., Eriksen, S. P., and Robinson, J. R. (1974). Drop
size and initial dosing frequency problems of topically applied ophthalmic
drugs. J. Pharm. Sci., 63:333.
24. Chrai, S. S., and Robinson, J. R. (1974). Ocular evaluation of methylcellulose
vehicle in albino rabbits. J. Pharm. Sci., 63:1218.
25. Chrai, S. S., Patton, T. F., Mehta, A. and Robinson, J. R. (1973). Lacrimal
and instilled uid dynamics in rabbit eyes. J. Pharm. Sci., 62:1112.
26. Burstein, N. L., and Anderson, J. A. (1985). Corneal penetration and ocular
bioavailability of drugs. J. Ocular Pharmacol., 1:309.
27. Makoid, M. C., and Robinson, J. R. (1979). Pharmacokinetics of topically
applied pilocarpine in the albino rabbit eye. J. Pharm. Sci., 68:435.
28. Chiang C. H., and Schoenwald, R. D. (1986). Ocular pharmacokinetic models
of clonidine-3 H hydrochloride. J. Pharmacokin. Biopharm., 14:175.
29. Gibaldi, M., and Perrier, D. (1982). Pharmacokinetics, 2nd ed. Marcel
Dekker, New York, pp. 145166.
30. Hamalainen, K.M., Kananen, K., Auriola, S., Kontturi, K., and Urtti, A.
(1997). Characterization of paracellular and aqueous penetration routes in
cornea, conjunctiva, and sclera. Invest. Ophthalmol. Vis. Sci., 38:627.
31. Schoenwald, R. D., Deshpande, G. S., Rethwisch, D. G., and Barfknecht, C.
F. (1997). Penetration into the anterior chamber via the conjunctival/scleral
pathway. J. Ocular Pharm. Ther., 13:41.
32. Wagner, J. G. (1975). Fundamentals of Clinical Pharmacokinetics. Drug
Intelligence Publications, Hamilton, IL, pp. 102106.
33. Kalsi, G. S., Gudauskas, G., Bussanich, N., Freeman, D. J., and Rootman, J.
(1991). Ocular pharmacokinetics of subconjunctivally administered cyclosporine in the rabbit. Can. J. Ophthalmol., 26:200.

Copyright 2003 Marcel Dekker, Inc.

172

Schoenwald

34. Acheampong, A. A., Shackleton, M., Tang-Liu, D. D.-S., Ding, S., Stern, M.
E., and Decker, R. (1998). Distribution of cyclosporin A in ocular tissues after
topical administration to albino rabbits and beagle dogs. Curr. Eye Res.,
18:91.
35. Miller, S. C., Himmelstein, K. J., and Patton, T. F. (1981). A physiologically
based pharmacokinetic model for the intraocular distribution of pilocarpine in
rabbits. J. Pharmacokin. Biopharm., 9:653.
36. Ettl, A., Hofmann, U., Daxer, A., Dietrich, H., Schmid, E., and Eichelbaum,
M., (1998). Ocular pharmacokinetics of verapamil in rabbits. NaunynSchmiedebergs Arch. Pharmacol., 357:331.
37. Schoenwald, R. D., Gadiraju, R. R., and Barfknecht, C. F. (1997). Serotonin
antagonists for use as antiglaucoma agents and their ocular penetration. Eur.
J. Pharm. Biopharm., 43:273.
38. Palmero, M., Bellot, J. L., Alcoriza, N., and Garc a-Cabanes, C. (1997). The
ocular pharmacokinetics of topical dicloenac is aected by ocular inammation. Ophthalmic Res., 31:309.
39. Bauer, N.J. C., Motamedi, N., Wicksted, J. P., March, W. F., Webers, C. A.
B., and Hendrikse, F. (1999). Non-invasive assessment of ocular pharmacokinetics using confocal raman spectroscopy. J. Ocular Pharmacol. Ther.,
15:123.
40. Yamamura, K., Sasaki, H., Nakashima, M., Ichikawa, M., Mukai, T.,
Nishida, K., and Nakamura, J. (1999). Characterization of ocular pharmacokinetics of beta-blockers using a diusion model after instillation. Pharm.
Res., 16:1595.
41. Sasaki, H., Yamamura, K., Mukai, T., Nishida, K., Nakamura, J.,
Nakashima, M., and Ichikawa, M. (1999). Characterization of ocular pharmacokinetics of tilisolol after instillation into anesthetized rabbits. Biol.
Pharm. Bull., 22:1253.
42. Velpandian, T., Gupta, S. K., Gupta, Y. K., Agarwal, H. C., and Biswas, N.
R. (1999). Comparative studies on topical lomeoxacin and ciprooxacin on
ocular kinetic and experimental corneal ulcer. J. Ocular Pharmacol. Ther.,
15:505.
43. Desai, S. (1993). Ocular pharmacokinetics of tobramycin. Int. Ophthalmol.,
17:201.
44. Lee, V. H. L., and Robinson, J. R. (1979). Mechanistic and quantitative
evaluation of precorneal pilocarpine disposition in albino rabbits. J. Pharm.
Sci., 68:673.
45. Himmelstein, K. J., Guvenir, I., and Patton, T. P. (1978). Preliminary pharmacokinetic model of pilocarpine uptake and distribution in the eye. J.
Pharm. Sci., 67:603.
46. Patton, T. F., and Robinson, J. R. (1976). Quantitative precorneal disposition of topically applied pilocarpine ntirate in rabbit eyes. J. Pharm. Sci.,
65:1295.
47. Conrad, J. M., and Robinson, J. R. (1977). Aqueous chamber drug distribution volume measurements in rabbits. J. Pharm. Sci., 66:219.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

173

48. Kitaura, T, Tsukiai, S., Arai, S., Miyake, K., Kimura, M., and Fukuchi, H.
(1998). Ocular pharmacokinetics of latamoxef and cefaclor in humans.
Penetration into aqueous humor. J. Pharmacobio-Dyn., 12:60.
49. Aldana, I., Fos, D., Penas, E. G., Gazzaniga, A., Gianesello, V., Monti, N. C.,
Figini, P. G., Zato, M. A., Bruseghini, L., and Esteras, A. (1992). Ocular pharmacokinetics of thiamphenicol in rabbits. Asneim. Forsch. Drug Res., 42:1236.
50. Ben-Nun, J., Joyce, D. A., Cooper, R. L., Cringle, S. J., and Constable, I. J.
(1989). Pharmacokinetics of intravitreal injection. 30:1055.
51. Rao, C. S., Schoenwald, R. D., Barfknecht, C. F., and Laban S. L. (1992).
Biopharmaceutical evaluation of ibufenac, ibuprofen, and their hydroxyethoxy analogs in the rabbit eye. J. Pharmcokin. Biopharm., 230:357.
52. Schoenwald, R. D., and Zhu, J. (2000). The ocular pharmacokinetics of ketanserin and its metabolite, ketansreinol, in albino rabbits. J. Ocular Pharmacol.
Ther., 16:479.
53. Huang, A. J. W., Tseng S. C. G., and Kenyon, K. R. (1989). Paracellular
permeability of corneal and conjunctival epithelia. Invest. Ophthalmol. Vis.
Sci., 30:684.
54. Occhipinti, J. R., Mosier, M. A., LaMotte, J., and Monji G. T. (1988).
Fluorophotometric measurement of human tear turnover rate. Curr. Eye
Res., 7:995.
55. Palestine, A. G., and Brubaker, R. F. (1981). Pharmacokinetics of uorescein
in the vitreous. Invest. Ophthalmol. Vis. Sci., 21:542.
56. Igarashi, H., Sato, Y., Hamada, S., and Kawasaki, T. (1984). Studies on
rabbit corneal permeability of local anesthetics. Jpn. J. Pharmacol., 34:429.
57. Schoenwald, R. D., and Huang, H. S. (1983). Corneal penetration behavior of
beta-blocking agents I: Physicochemical factors. J. Pharm. Sci., 72:1266.
58. Huang, H. S., Schoenwald, R. D., and Lach, J. L. (1983). Corneal penetration
behavior of beta-blocking agents II: Assessment of barrier contributions. J.
Pharm. Sci., 72:1272.
59. Chiang, C. H., Huang, H. S., and Schoenwald, R. D. (1986). Corneal permeability of adrenergic agents potentially useful in glaucoma. J. Taiwan Pharm.
Assoc., 38:67.
60. Grass, G. M., and Robinson, J. R. (1988). Mechanisms of corneal drug penetration II: Ultrastructural analysis of potential pathways for drug movement.
J. Pharm. Sci., 77:15.
61. Camber, O., Edman, P., and Olsson, L. I. (1986). Permeability of
Prostaglandin F2alpha and prostaglandin F2alpha and prostaglandin F2alpha
esters across cornea in vitro. Int. J. Pharm., 29:259.
62. Corbo D. C., Liu, J. C., and Chien, Y. W. (1990). Characterization of the
barrier properties of mucosal membranes. J. Pharm. Sci., 79:202.
63. Gibaldi, M., and Perrier, D. (1982). Pharmacokinetics, 3nd ed. Marcel
Dekker, New York, pp. 475477.
64. ODay, D, Head, W. S., Robinson, R. D., Williams, T. E., and Wol, R.
(1992). Ocular pharmacokinetics of saperconazole in rabbits. Arch.
Ophthalmol., 110:550.

Copyright 2003 Marcel Dekker, Inc.

174

Schoenwald

65. Baeyens, V., and Gurney, R., (1997). Chemical and physical parameters of
tears relevant for the design of the ocular drug delivery formulations. Pharm.
Acta Helvetica, 72:191.
66. Le Bourlais, C., Acar, L., Zia, H., Sado, P. A., Needham, T.,and Leverage, R.
(1998). Ophthalmic drug delivery systemsrecent advances. Prog. Ret. Eye
Res., 17:33.
67. Tang-Liu, D. D.-S., and Sandri, R. (1989). Ocular biodistribution of clonidine
after topical application with ophthalmic rods or solution. J. Ocular
Pharmacol., 5:133.
68. Ito,Y., Cai, H., Koizumi, Y., Hori, R., Terao, M., Kimura, T., Takagi, S., and
Tomohiro, M. (2000). Eects of lipid composition on the transcorneal penetration of liposomes containing disulram, a potential anti-cataract agent, in
the rabbit. Biol. Pharm. Bull., 23:327.
69. Zane, P. A., Brindle, S. D., Gause, D. O., OBuck, A. J., Raghavan, P. R., and
Tripp, S. L. (1990). Physiological factors associated with binding and retention of compounds in ocular melanin or rats: correlations using data from
whole body autoradiography and molecular modeling for multiple linear
regression analyses. Pharm. Res., 9:935.
70. Kahn, M., Barney, N. P., Briggs, R. M., Bloch, K. J., and Allansmith, M. R.,
(1990). Penetrating the conjunctival barrier. Invest. Ophthalmol. Vis. Sci.,
31:258.
71. Ahmed, I., Gokhale, R. D., Shah, M. V., and Patton, T. F. (1987).
Physicochemical determinants of drug diusion across the conjunctiva, sclera,
and cornea. J. Pharm. Sci., 76:583.
72. Maurice, D. (1997). Characterization of paracellular penetration routes.
Invest. Ophthalmol. Vis. Sci., 38:2177.
73. Ling, T. L., and Combs, D. L. (1987). Ocular bioavailability and tissue
distribution of [14C]ketorolac tromethamine in rabbits. J. Pharm. Sci.,
76:289.
74. Wagner, J. G. (1993). Pharmacokinetics for the Pharmaceutical Scientist.
Technomic Publishing Co., Lancaster, PA, pp. 2733.
75. Gibaldi, M., and Perrier, D. (1982). Pharmacokinetics, 2nd ed. Marcel
Dekker, New York, pp. 409417.
76. Eller, M. G., Schoenwald, R. D., Dixson, J. A., Segarra, T., and Barfknecht,
C. F. (1985). Topical carbonic anhydrase inhibitors IV: relationship between
excised corneal permeability and pharmacokinetic factors. J. Pharm. Sci.,
74:525.
77. Morlet, N., Graham, G. G., Gatus, B., McLachlan A. J., Salonikas, C.,
Naidoo, D., Goldberg, I., and Lam, C. M. (2000). Pharmacokinetics of ciprooxacin in the human eye; a clinical study and population pharmacokinetic
analysis. Antimicrob. Agents Chemother., 44:1674.
78. Putnam, M. L., Schoenwald, R. D., Duel, M. W., Barfknecht, C. F.,
Segarra, T. M., and Campbell, D. A. (1987). Ocular disposition of aminozolamide in the rabbit eye. Invest. Ophthalmol. Vis. Sci., 28:1373.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

175

79. Chien, D. S., Homsy, J. J., Gluchowski, C., and Tang-liu, D. D. S. (1990).
Corneal and conjunctival/scleral penetration of p-aminoclonidine, AGN
190342, and clonidine. Curr. Eye Res., 9:1051.
80. Schoenwald, R. D., Tandon, V., Wurster, D. E., and Barfknecht, C. F. (1998).
Signicance of melanin binding and metabolism in the activity of 5-acetoxyacetylimino-4-methyl-
2 -1,3,4-thiadiazoline-2-sulfonamide. Eur. J. Pharm.
Biopharm., 46:39.
81. Hitoshi, S., Bundgaard, H., and Lee, V.H. L. (1989). Design of prodrugs to
selectively reduce timolol absorption on the basis of the dierential lipophilic
characteristics of the cornea and the conjunctiva. Invest. Ophthalmol. Vis. Sci.,
30(Suppl.):25.
82. Edelhauser, H. F., and Maren, t. H. (1988). Permeability of human cornea
and sclera to sulfonamide carbonic anhydrase inhibitors. Arch. Ophthalmol.,
106:110.
83. Camras, C. B., Bito, L. Z., and Toris, C. B. (19xx). Prostaglandins and prostaglandiin analogues. In Textbook of Ocular Pharmacology (T. J. Zimmerman,
K. S. Kooner, M. Sharir, and R. D. Fechtner, eds.). Lippincott-Raven
Publishers, Philadelphia, pp. 315328.
84. Friedland, B. R., and Muther, T. F. (1978). Autoradiographic localization of
carbonic anhydrase in the ciliary body. Invest. Ophthalmol. Vis. Sci.,
(suppl.):162.
85. Maurice, D. M., and Mishima, S. (1984). Ocular pharmacokinetics. In
Pharmacology of the Eye (M. L. Sears, ed.). Springer-Verlag, Berlin, pp. 19116.
86. Ahmed, I. Francoeur, M. L., Thombre, A. G., and Patton, T. F. (1989). The
kinetics of timolol in the rabbit lens: implications for ocular drug delivery.
Pharm. Res., 6:772.
87. Maurice, D. (1987). Kinetics of topically applied ophthalmic drugs. In
Ophthalmic Drug Delivery: Biopharmaceutical, Technological and Clinical
Aspects (M. F. Saettone, M. Bucci, and P. Speiser, eds.). Springer-Verlag,
Berlin, pp. 1926.
88. Liu, W., Liu, Q. F., Perkins, R., Drusano, G., Louie, A., Madu, A., Mian, U.,
Mayers, M., and Miller, M. H. (1998). Pharmacokinetics of sparoxacin in the
serum and vitreous humor of rabbits: physicochemical properties that regulate
penetration of quinolone antimicrobials. Antimicrob. Agents Chemother.,
42:1417.
89. Perkins, R. J., W. Liu, Drusano, G., Madu, M., Mayers, M., Madu, and Miller,
M. H. (1995). Pharmacokinetics of ooxacin in serum and vitreous humor of
albino and pigmented rabbits. Antimicrob. Agents Chemother., 39:1493.
90. Lee, V.H. L., Chang, S. C., Oshiro, C. M., and Smith, R. E. (1985). Ocular
esterase composition in albino and pigmented rabbits: possible implications in
ocular prodrug design and evaluation. Curr. Eye Res., 4:1117.
91. Chien, D. S., Bundgaard, H., and Lee, V. H. L. (1988). Inuence of corneal
epithelial integrity on the penetration of timolol prodrugs. J. Ocular
Pharmacol., 4:137.

Copyright 2003 Marcel Dekker, Inc.

176

Schoenwald

92. Lee, V. H. L., Morimoto, K. M., and Stratford, Jr., R. E. (1982). Esterase
distribution in the rabbit cornea and its implications in ocular drug bioavailability. Biopharm. Drug Disposit., 3:291.
93. Lee, V. H. (1983). Esterase activities in adult rabbit eyes. J. Pharm. Sci.,
72:239.
94. Chien, D. S., and Schoenwald, R. D. (1986). Improving the ocular absorption
of phenylephrine. Biopharm. Drug Disposit., 7:45.
95. Schoenwald, R. D., Folk, J. C., Kumar, V., and Piper, J. G. (1987). In vivo
comparison of phenylephrine and phenylephrine oxazolidine instilled in the
monkey eye. J. Ocular Pharmacol., 3:333.
96. Chang, S. C., Bundgaard, H., Buur, A., and Lee, V. H. L. (1987). Improved
corneal penetration of timolol by prodrugs as a means to reduce systemic drug
load. Invest. Ophthalmol. Vis. Sci., 28:487.
97. Bundgaard, H., Falch, E., Larsen, C., and Mikkelson, T. (1986). Pilocarpine
prodrugs I. Synthesis, physicochemical properties and kinetics of lactonization
of pilocarpic acid esters. J. Pharm. Sci., 75:36.
98. Bundgaard, H., Falch, E., Larsen, C., Mosher, G. L., and Mikkelson, T.
(1986). Pilocarpine prodrugs II. Synthesis, stbility, bioconversion, and physicochemical properties of sequentially labile pilocarpine acid diesters. J.
Pharm. Sci., 75:775.
99. Narukar, M. M., and Mitra A. K. (1989). Prodrugs of 5-iodo-20 -deoxyuridine
for enhanced ocular transport. Pharm. Res., 6:887.
100. Bodor, N., and Buchwald, P. (2000). Soft drug design: general principles and
recent applications. Med. Res. Rev., 20:58.
101. Buchwald, A., Browne, C. E., Wu, W. M., and Bodor, N. (2000). Soft cannabinoid analogues as potential anti-glaucoma agents. Pharmazie, 55:196.
102. Farag, H. H., Wu, W. M., Barros, M. D., Somogyi, G., Prokai,L., and Bodor,
N. (1997). Drug Des. Disc., 15:117.
103. Bodor, N., El-Koussi, A., Kano, M., and Khalifa, M. M. (1988). Soft drugs.
7. Soft -blockers for systemic and ophthalmic use. J. Med. Chem., 31:1651.
104. Bodor, N., Farag, H. H., Somogyi, G., Wu, W.M., and Prokai, L. (1997).
Ocular-specic delivery of timolol by sequential bioactivation of its oxime and
methoxime analogs. J. Ocular Pharmacol. Ther., 13:389.
105. Kumar, G. N., Hammer, R. H., Wu, W. M., and Bodor, N. S. (1993).
Mydriatic activity and transcorneal penetration of phenylsuccinic soft analogs
of methscopolamine as short acting mydriatics. Curr. Eye. Res., 12:501.
106. Campbell, D. A., Schoenwald, R. D., Duel, M. W., and Barfknecht, C. F.
(1991). Characterization of arylamine acetyltransferase in the rabbit eye.
Invest. Ophthalmol. Vis. Sci., 32:2190.
107. Lee, V. H. L., Chien, D. S., and Sasaki, H. (1988). Ocular ketone reductase
distribution and its role in the metabolism of ocular applied levobunolol in the
pigmented rabbit. J. Pharmacol. Exp. Ther., 246:871.
108. Shimada, S., Mishima, H., Kitamura, S., and Tatsumi, K. (1987).
Nicotinamide N-oxide reductase activity in bovine and rabbit eyes. Invest.
Ophthalmol. Vis. Sci., 28:1204.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

177

109. Masferrer, J. L., Dunn, M. W., and Schwartzman, M. L. (1990). 12(R)Hydroxyeicosatetraenoic acid, an endogenous corneal arachidonate metabolite, lower intraocular pressure in rabbits. Invest. Ophthamol. Vis. Sci.,
31:535.
110. OHara, K. (1977). Eects of cholinergic agonists on isolated iris sphincter
muscles: a pharmacodynamic study. Jpn. J. Ophthalmol., 21:516.
111. Mishima, S. (1981). Clinical pharmacokinetics of the eye. Invest. Ophthalmol.
Vis. Sci., 21:504.
112. Van Haeringen, N. J., Van Sorge, A. A., and Carballosa Core-Bodelier, V. M.
W. (2000). Constitutive cyclooxygenase-1 and induced cyclooxygenase-2 in
isolated human iris inhibited by S(+) urbiprofen. J. Ocular Pharmacol.
Ther., 16:353.
113. Gabrielsson, J., and Weiner, D. (1997). Dose-response-time analysis I. In
Pharmacokinetic and Pharmacodynamic Data Analysis Concepts and
Applications, 2nd ed. Swedish Pharmaceutical Society, Swedish
Pharmaceutical Press, Stockholm, Sweden, pp. 717723.
114. Smolen, V. F., and Schoenwald, R. D. (1971). Drug absorption analysis from
pharmacological data I. The method and its conrmation exemplied for a
mydriatic drug, tropicamide. J. Pharm. Sci. 60:96.
115. Schoenwald, R. D., and Smolen, V. F. (1971). Drug absorption analysis from
pharmacological data. II Transcorneal biophasic availability of tropicamide.
J. Pharm. Sci., 60:1039.
116. Smolen, V. F., and Schoenwald, R. D. (1974). Drug absorption analysis from
pharmacological data. III. Inuence of polymers and pH on transcorneal
biophasic availability and myriatic response of tropicamide. J. Pharm. Sci.,
63:1582.
117. Smolen, V. F. (1972). Applications of a pharmacological method of drug
absorption analysis to the study of the bioavailability characteristics of
mydriatic drugs. Can. J. Pharm. Sci., 7:7.
118. Smolen, V. F. (1975). Drug bioelectrometric study of the mechanisms of
carbachol interactions with the cornea and its relation to miotic activity. J.
Pharm. Sci., 64:526.
119. Chien, D. S., and Schoenwald, R. D. (1990). Ocular pharmacokinetics and
pharmacodynamics of phenylephrine and phenylephrine oxazolidine in rabbit
eyes. Pharm. Res., 7:476.
120. Schoenwald, R. D., and Chien D. S. (1988). Ocular absorption and disposition
of phenylephrine and phenylephrine oxazolidine. Biopharm. Drug Disposit.,
9:527.
121. Eng. I. S. (1986). Ph.D. thesis, University of Iowa College of Pharmacy.
122. Valeri, P., Palmery, M., Severini, G., Piccinelli, D, and Catanese, B.
(1986). Ocular pharmacokinetics of dapiprazole. Pharmacol. Res. Comm.,
18:1093.
123. Brazzell, R. K., Wooldridge, C. B., Hackett, R. B., and McCue B. A. (1990).
Pharmacokinetics of the aldose reductase inhibitor inhibitor imirestat following topical ocular administration. Pharm. Res., 7:192.

Copyright 2003 Marcel Dekker, Inc.

178

Schoenwald

124. Gudauskas, G., Kumi, C., Dedhar, C., Bussanich, N., and Rootman, J.
(1985). Ocular pharmacokinetics of subconjunctivally versus intravenously
administered 6-mercaptopurine. Can. J. Ophthalmol., 20:110.
125. Taylor P. B., Burd, E. M., and Tabbara, K. (1987). Corneal and intraocular
penetration of topical and subconjunctival fusidic acid. Bri. J. Ophthalmol.,
71:598.
126. Kleinberg, J., Dea, F. J., Anderson, J. A., and Leopold I. H. (19xx).
Intraocular penetration of topically applied lincomycin hydrochloride in rabbits, Arch. Ophthalmol, 97:933.
127. Hui, H. W., Zeleznick, L., and Robinson J. R. (1984). Ocular disposition of
topically applied histamine, cimetidine and pyrilamine in the albino rabbit.
Curr. Eye Res., 3:321.
128. Maren T. H., and Jankowska, L. (1985). Ocular pharmacology of sulfonamides: the cornea as barrier and depot. Curr. Eye Res., 4:399.
129. Walstad, R. A., Hellum K. Blika S., Dale L. G., Fredriksen, T., Myhre, K. I.,
and Spencer, G. R. (1983). Pharmacokinetics and tissue penetration of ceftazidime: studies on lymph, aqueous humor, skin blister, cerebrospinal and
pleural uid. J. Antimicrob. Chemother., 12 (Suppl. A):275.
130. Barza, M., Kane, A., and Baum, J. L. (1979). Intraocular levels of cefamandole compared with cefazolin after subconjunctival injection in rabbits. Invest.
Ophthalmol. Vis. Sci., 18:250.
131. Auclair, E., Laude, D., Wainer, I. W., Chaoulo, F., and Elghozi, J. L. (1988).
Comparative pharmacokinetics of D- and L-alphamethyldopa in plasma, aqueous humor, and cerebrospinal uid in rabbits. Fundam. Clin. Pharmacol.,
2:283.
132. Grove, J., Gautheron, P., Plazonnet, B., and Sugrue, M. F. (1988). Ocular
distribution studies of the topical carbonic anhydrase inhibitors L-643,799
and L-650,719 and related alkyl prodrugs. J. Ocular Pharmacol., 4:279.
133. Agata, M., Tanaka, M., Nakajima, A., Fujii, A., Kuboyama, N., Tamura, T.,
and Araie, M. (1984). Ocular penetration of topical diclofenac sodium, a nonsteroidal anti-inammatory drug, in rabbit eye. Nippon Ganka Gakkai Zashi,
88:991.
134. Anderson, J. A., Chen, C. C., Vita, J. B., and Shackleton, M. (1982).
Disposition of topical urbiprofen in normal and aphakic rabbit eyes. Arch.
Ophthalmol., 100:642.
135. Barza, M., and McCue, M. (1983). Pharmacokinetics of aztreonam in rabbit
eyes. Invest. Ophthalmol. Vis. Sci., 24:468.
136. Huuopponen, R., Kaila, T., Salminen, L., and Urtti, A. (1987). The pharmacokinetics of ocularly applied timolol in rabbits. Acta Ophthalmol., 65:63.
137. Huang, H. S., Schoenwald, R. d., and Lach, J. L. (1983). Corneal penetration
behavior of beta-blocking agents III: in vitro-in vivo correlations. J. Pharm.
Sci., 72:1279.
138. Leibowitz, H. M., Ryan, W. J., Kupferman, A., and DeSantis, L. (1986).
Bioavailability and corneal anti-inammatory eect of topical suprofen.
Invest. Ophthalmol. Vis. Sci., 27:628.

Copyright 2003 Marcel Dekker, Inc.

Ocular Pharmacokinetics and Pharmacodynamics

179

139. Sugrue, M. F., Gautheron, P., Mallorga, T. E., Nolan, S. L., Graham, H.,
Schwam, H., Shepard, K. L., and Smith R. L. (1990). L-662,583 is topically
eective ocular hypotensive carbonic anhydrase inhibitor in experimental animals. Br. J. Pharmacol., 99:59.
140. Mester, U., Krasemann, C., and Werner, H. (1982). Cefsulodine concentrations in rabbit eyes after intravenous and subconjunctival administration.
Ophthalmol. Res., 14:129.
141. Rootman, J., Ostry, A., and Gudauskas, G. (1984). Pharmacokinetics and
metabolism of 5-uorouracil following subconjunctival versus intravenous
administration. Can. J. Ophthalmol., 19:187.
142. Lee, v. H.L., and Robinson, J. R. (1982). Disposition of pilocarpine in the
pigmented rabbit eye. Int. J. Pharm., 11:155.
143. Vigo, J. F., Rafart, J., Concheiro, A., Martinez, R., and Cordido, M. (1988).
Ocular penetration and pharmacokinetics of cefotaxime: an experimental
study. Curr. Eye Res., 7:1149.
144. Sugrue, M. F. (1996). The preclinical pharmacology of dorzolamide hydrochloride, a topical carbonic anhydrase inhibitor. J. Ocular Pharmacol. Ther.,
12:363.

Copyright 2003 Marcel Dekker, Inc.

6
Mathematical Modeling of Drug
Distribution in the Vitreous Humor
Stuart Friedrich, Bradley Saville, and Yu-Ling Cheng
University of Toronto, Toronto, Canada

I.
A.

INTRODUCTION
Vitreous Physiology

The vitreous humor is a viscous uid occupying the space between the lens
and retina of the eye. A number of diseases that can aect the vitreous or the
surrounding retina can be treated by administration of various therapeutic
drugs. Due to physiological barriers within the eye that prevent drug in the
systemic circulation from entering the vitreous, the most common method of
treating diseases aecting the vitreous or retina is a direct intravitreal injection of drug (1). Many of the drugs used to treat vitreous and retinal disorders have a narrow concentration range in which they are eective, and
they may be toxic at higher concentrations (26). Therefore, knowledge of
the drug distribution following intravitreal administration is important if
the disease is to be properly treated and damage to tissues by high concentrations of drug is to be avoided.
There are three main tissues that bound the vitreous humor: the retina,
lens, and hyaloid membrane. The retina covers the posterior portion of the
interior of the globe and is immediately adjacent to the vitreous. In a rabbit
eye, the retina is supplied with nutrients via the choroid, the tissue layer
directly outside the retina. The choroid has a vast network of capillaries,
while the retina is avascular. The retina is made up of several cellular layers,
some of which form the blood-retinal barrier that prevents extraneous compounds from entering the vitreous from the bloodstream. The vitreous to
blood permeability of the blood-retinal barrier depends on the physicochem181

Copyright 2003 Marcel Dekker, Inc.

182

Friedrich et al.

ical properties of the drug. Some drugs cannot penetrate the barrier, while
others may be actively transported between the vitreous and the blood. The
lens forms the majority of the anterior boundary of the vitreous humor. The
lens is composed of highly compacted cellular material and is therefore
highly impermeable to most drugs. The hyaloid membrane is composed of
loosely packed collagen bers and spans the gap between the lens and the
ciliary body. Although the hyaloid membrane forms a boundary between
the stagnant vitreous and the owing aqueous humor, it does not form a
limiting barrier to the transport of low molecular weight compounds. The
aqueous humor is continuously produced by the ciliary body and drained
from the anterior chamber of the aqueous humor after it passes between the
iris and the lens. Therefore, once drug passes through the hyaloid membrane
from the vitreous it is eliminated by the ow of aqueous humor.

B.

Properties of Intravitreal Injected Drug Solutions

The distribution of a drug solution within the vitreous immediately following intravitreal injection may be dependent on several factors. Needle gauge,
needle length, penetration angle of the needle, speed of the injection, rheology of the injected solution, and rheology of the vitreous could all aect
how the drug solution is initially distributed in the vitreous. The shape that
the drug solution assumes immediately after injection may range from globular to irregular shapes that vary in the extent of ngering. Such variations
in shapes would inuence the diusional surface area and hence drug distribution within the vitreous (7).

C.

Experimental Measurements of Drug Distribution in the

Vitreous and Retinal Permeability

Due to the diculty in measuring solute concentrations within the vitreous,

there has only been one published report that shows experimentally measured concentration proles of model compounds across an entire cross
section of the rabbit vitreous (8). The compounds used in the study were
uorescein, uorescein glucuronide, and uorescein isothiocyanate. The
main elimination route for uorescein is across the retina; conversely, uorescein glucuronide has a very low retinal permeability and is eliminated
mainly across the hyaloid membrane. As a result of their contrasting elimination characteristics, the experimental uorescein and uorescein glucuronide concentration proles observed by Araie and Maurice (8) are ideal for
comparing to concentration proles calculated by a mathematical model as
a test of the models robustness.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

183

The in vitro permeability of an excised rabbit retina to uorescein and

uorescein glucuronide has been measured by Koyano et al. (9). Retinal
permeabilities can also be determined by tting vitreous concentrations
calculated by a mathematical model to experimental data, such as the
data observed by Araie and Maurice (8). Therefore, the in vitro retinal
permeability found by Koyano et al. can be compared to model calculated
retinal permeabilities as a method of model validation.

II.

EARLY MATHEMATICAL MODELS OF THE VITREOUS

HUMOR

Several models have been developed to simulate the distribution and elimination of drugs from the vitreous (8,1013). A number of other models
have been used to determine the retinal permeability from the blood to
the vitreous (1419). In each of the former models, a simplied geometry
and set of boundary and initial conditions were used so that the mathematical expression of the model would be easier to develop and solve. Some of
these simplications aect the generality of the model and may aect model
calculations and estimates of the retinal permeability.
A.

Araie and Maurice Model

Araie and Maurice (8) used the simplest approach to model distribution
and elimination in the vitreous by representing the vitreous as a sphere
with the entire outer surface representing the retina (Fig. 1). The predicted
concentration prole within the vitreous will be the same for any cross
section that passes through the center of the sphere, with the highest
concentration in the center and the lowest concentration next to the
outer surface. In a rabbit eye, the center of curvature of the retina is
immediately next to the lens, on the symmetry axis of the vitreous.
Qualitatively, the concentration prole calculated by a spherical model
will be correct for the posterior hemisphere of the vitreous, which is
behind the center of curvature of the retina. In a spherical geometry
model like the Araie and Maurice model, the concentration prole in
the anterior hemisphere will be the same as in the posterior hemisphere,
since the two hemispheres are the same. Therefore, the concentration prole calculated by a spherical model for the portion of the vitreous that is
in front of the center of curvature of the retina will not accurately reect
the actual prole that would be present in vivo. A spherical model also
assumes that there is no ux across the plane that passes through the
center of curvature of the retina and is perpendicular to the symmetry

Copyright 2003 Marcel Dekker, Inc.

184

Friedrich et al.

Figure 1 Comparison of models developed by Araie and Maurice (8), Yoshida et

al. (12,13), and Ohtori and Tojo (10,11).

axis. For this assumption to be true, the loss across the portion of the
retina located behind its center of curvature must equal the sum of the loss
across the hyaloid membrane plus the loss across the portion of the retina
in front of the center of curvature of the retina. This condition will only be
true for a particular retinal permeability.

B.

Yoshida Model

Yoshida et al. (12,13) extended the model used by Araie and Maurice by
dividing the vitreous into an anterior and posterior hemisphere (Fig. 1).
Each hemisphere was further subdivided into eight compartmental shells,

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

185

and a separate permeability was used for the outer surface of each hemisphere. Within each concentric compartmental shell, the concentration
calculated by the model would be uniform since each compartment was
assumed to be perfectly mixed. Yoshidas model is a more accurate representation of the true vitreous than a spherical model, since a separate
permeability is calculated for the outer surface of the anterior and posterior hemispheres. However, if this model is used to estimate retinal permeabilities using vitreous concentration data, an accurate estimate of the
retinal permeability will only be obtained for compounds that are primarily eliminated across the retina. As discussed in further detail (in Sec. 3.2),
compounds that are eliminated mainly across the hyaloid membrane will
have concentration contours within the vitreous which are perpendicular
to the retina, and, therefore, the concentric compartmental shells will not
accurately calculate the correct concentration prole. The prole calculated by a model that uses concentric compartmental shells will always
have concentration contours that are parallel to the retina since the
concentration within each compartmental shell must be uniform.
Furthermore, the concentric shell compartmental model will always predict
a uniform vitreal concentration over the entire inner surface of the retina;
the model will be unable to account for conditions that result in nonuniform concentrations adjacent to the retina (such as displaced or irregular
injections or diseases that cause the retinal permeability to vary over the
retinal surface).

C.

Ohtori and Tojo Model

The geometry and boundary conditions of the vitreous were more accurately modeled by Ohtori and Tojo (10,11). The model was cylindrical in
shape with one end of the cylinder and the curved surface representing the
retina, and the opposite end of the cylinder divided into an outer section
representing the hyaloid membrane and an inner section representing the
lens (Fig. 1). This model design allowed the use of dierent permeability
values for the boundaries representing the lens, hyaloid membrane, and
retina. The main advantage of this model is a more accurate representation
of drug loss from the anterior portion of the vitreous. The boundary
conditions representing the lens and hyaloid membrane are dierent
than for the retina, thereby allowing more accurate calculated concentrations within the region of the vitreous close to these boundaries. The main
limitations of this model are the use of a cylinder to approximate the
spherical shape of the vitreous and the inability to study complex and
nonsymmetric initial conditions.

Copyright 2003 Marcel Dekker, Inc.

186

III.

Friedrich et al.

FINITE ELEMENT MODEL BASED ON RABBIT EYE

Modelling drug distribution and elimination in the vitreous is particularly

suitable to nite element analysis, which allows the complex geometry and
boundary conditions of the vitreous to be more accurately incorporated into
the model. Finite element analysis also allows complex initial conditions to
be studied.
A nite element model has been developed (20), and the model was
initially based on the physiological dimensions of a rabbit eye, using the
cross sections of the eye shown by Araie and Maurice as a guide (8). The
rabbit eye was initially chosen rather than the human eye due to the availability of data for conrmation of model calculations.
A.

Model Development

A detailed description of the model equations and solution methods are

presented elsewhere (20). The tissues included in the model were the vitreous, retina, posterior surface of the lens, and the posterior chamber of the
aqueous humor (Fig. 2). Mass transport by both convection and diusion
was accounted for in the aqueous humor. Due to the viscous nature of the
vitreous, especially in the rabbit eye, only diusive mass transport was
considered in the vitreous.
The only unknown variables in the model were the initial distribution of
drug solution in the vitreous following injection (initial condition) and the
retinal permeability. The initial distribution of the injected drug solution is a
variable that could aect the elimination of drug from the vitreous. Therefore,
rather than simulating only a central injection, four extreme initial conditions
were considered. These conditions covered a range of possible positions where
the injected drug solution may have been placed in the in vivo experiments
performed by Araie and Maurice (8): (a) a central injection, (b) an injection
placed next to the lens on the symmetry axis, (c) an injection placed next to the
retina on the symmetry axis, and (d) an injection placed close to the hyaloid
membrane (Fig. 3). The model simulated the shape of the injected solution as
a sphere in cases 13 and a cylinder of equal height and diameter for case 4. In
reality, a drug solution injected into the vitreous will not take the shape of a
perfect sphere or a cylinder. Complex and poorly characterized shapes that
depend on drug solution and vitreous properties as well as the injection procedure may be observed (7). These complex shapes would be dicult to
incorporate in the nite element model, so for the purpose of studying the
eect of injection position, a simpler shape was selected. The concentration
proles produced by the model using each initial condition were t to the
concentration proles observed by Araie and Maurice (8), resulting in four

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

187

Figure 2 Cross section view of the model. In addition to the vitreous, the model
includes the posterior aqueous compartment and the surrounding retina layer. The
aqueous compartment was included to properly account for drug loss across the
hyaloid membrane.

possible values of retinal permeability that best t the experimental proles.

By simulating extreme injection positions, the range of retinal permeabilities
calculated by the model should include the actual retinal permeability. If the
actual injection position used by Araie and Maurice was precisely known, a
single retinal permeability could be calculated using the model.

B.

Comparison of Model Calculated and Experimental Data

Figures 4 and 5 show the model calculated concentration proles for uorescein at 15 hours and for uorescein glucuronide at 24 hours, respectively,
after a spherical central injection (A), and a cylindrical injection displaced
towards the hyaloid membrane (B). The concentration proles for the injection displaced towards the lens and the injection displaced towards the
retina were qualitatively similar to the concentration prole produced by
the central injection.
Qualitatively, the contour proles calculated using the model are similar to those found experimentally by Araie and Maurice (8). In Figure 4, the

Copyright 2003 Marcel Dekker, Inc.

188

Friedrich et al.

Figure 3 Injection positions studied using the model. Four extreme and distinct
injection positions were studied to determine the sensitivity of the model calculated
retinal permeability to the initial location of the injected drug.

concentration contour lines are parallel to the retina as expected, since the
ux of uorescein across the retina was the dominant elimination mechanism. For each injection position along the symmetry axis, the maximum
model calculated concentrations were next to the lens, on the symmetry
axis as shown in Figure 4A. In the case where uorescein was injected closer
to the hyaloid membrane (Fig. 4B), the maximum model calculated concentration was next to the lens; however, it is displaced slightly, closer to the site
of the injection.
In Figure 5, the model calculated concentration contour lines are
perpendicular to the retina since uorescein glucuronide is eliminated
mainly across the hyaloid membrane. Araie and Maurice (8) found that
the concentration of uorescein glucuronide in the vitreous was approximately the same next to the retina and next to the lens on the symmetry axis
of the vitreous at 24 hours. However, the model calculated that the concentration next to the retina (12.1 mg mL1 ) was slightly higher than the concentration next to the lens (10.4 mg mL1 ) for all injection positions. For the
hyaloid-displaced injection, the maximum concentration was shifted slightly
towards the injection site, similar to that noted for the hyaloid-displaced

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

189

Figure 4 Model calculated concentration prole for uorescein at 15 hours for a

spherical central injection (A) and a hyaloid-displaced injection (B). The dominant
elimination route for uorescein is across the retina; therefore, the concentration
contour lines are parallel to the retina surface. The concentration prole for the
lens-displaced injection and the retina-displaced injection were qualitatively similar
to the central injection.

Copyright 2003 Marcel Dekker, Inc.

190

Friedrich et al.

Figure 5 Model calculated concentration prole for uorescein glucuronide at 24

hours for a spherical central injection (A) and a hyaloid-displaced injection (B).
Fluorescein glucuronide has a very low retinal permeability; therefore, the concentration contour lines are perpendicular to the retina. The concentration prole for
the lens-displaced injection and the retina-displaced injection were qualitatively similar to the central injection.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

191

injection of uorescein. If the hyaloid membrane was the only elimination

route, theory would suggest that the maximum concentration would be next
to the retina on the symmetry axis, since this is the point where a drug
molecule must travel furthest to be eliminated.
In Figure 6, the model calculated concentration gradients of uorescein
between the lens and the retina along the symmetry axis are compared with
the experimental data of Araie and Maurice (8). In each case, the concentrations have been normalized with respect to the concentration found next to
the lens. Considering that the concentration gradients were calculated by
tting only the experimental concentration measured 1 mm adjacent to the

Figure 6 Concentration gradient between the lens and retina 15 hours after an
intravitreal injection of uorescein. Concentrations have been normalized with
respect to the concentration next to the lens. The experimental bars represent the
minimum and maximum distance from the center of curvature of the retina that each
specic experimental concentration contour line was observed (8). The model calculated proles were produced using the retinal permeabilities shown in Table 1. The
retinal permeabilities were calculated by tting only the experimental concentration
observed 1 mm adjacent to the lens measured at 15 hours. However, the model was
able to accurately t the entire prole.

Copyright 2003 Marcel Dekker, Inc.

192

Friedrich et al.

lens, the fact that the model-calculated concentration gradients follow the
experimentally observed gradient provides strong validation of the model
and its inherent assumptions. Of particular note is that even 15 hours after
the intravitreal injection, signicant variations in the concentration proles
were observed when dierent injection locations were considered. At earlier
times after the injection, the concentration variations would be much larger,
which suggests that injection position is an important variable that must be
controlled when performing animal experiments and in clinical treatment.
For each of the simulated injection positions, Table 1 shows the model
calculated retinal permeabilities of uorescein and uorescein glucuronide
that were determined by tting the experimental data and compares them to
values calculated by other published models and found in vitro by Koyano
(9) using an excised rabbit retina. As mentioned earlier, the retinal permeability for a compound is a constant. The dierent values obtained from the
various simulations occurred because the retinal permeability was the only
parameter with an unknown value that could be adjusted to t the experimental data. Since the injection position aects the concentration gradients
in the vitreous, dierent estimates for retinal permeabilities were obtained.
The estimated uorescein retinal permeabilities were much lower when the
injection was displaced towards the retina or the hyaloid membrane. In the
former case, uorescein was placed closer to the retina, producing a higher
initial concentration gradient of uorescein across the retina than obtained
from a central injection. Since the concentration gradient is higher, a lower

Table 1 Comparison of Finite Element Model Calculated Retinal

Permeabilities and Other Published Values
Retinal permeability
(cm/s)
Position of injection and other published values

Fluorescein

Fluorescein
glucuronide

Central Injection
Injection Displaced Towards Lens
Injection Displaced Towards Retina
Injection Displaced Towards Hyaloid Membrane
Araie and Maurice (8)
Koyano et al. (9)
Yoshida et al. (12,13)a

2:88  105
3:50  105
2:03  105
1:94  105
2:33  105
0:61:8  105
1:27  105

6:41  107
3:89  107
7:62  107
0
NC
6:3  106
1:5  106

These data were collected using monkey eyes but were included for comparison.
NC = Not calculated.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

193

retinal permeability is required to t the experimental data. In the latter

case, the injection position is closer to both the retina and the hyaloid
membrane than with a central injection. Retinal penetration is the dominant
elimination mechanism for uorescein; however, when the injection is placed
closer to the hyaloid membrane, more uorescein will be eliminated across
the hyaloid membrane than if the injection was placed in the center of the
vitreous. The combination of a greater loss across the hyaloid membrane
and a higher initial concentration gradient across the retinal results in a
lower retinal permeability required to t the experimental data.
Displacement of the injection towards the lens led to the highest estimate
of the retinal permeability, because this injection position places uorescein
furthest from the retina compared to the other three injection positions.
A retinal permeability of zero was obtained when the injection of
uorescein glucuronide was displaced towards the hyaloid membrane. A
low retinal permeability was expected since penetration across the hyaloid
membrane is the main elimination route for uorescein glucuronide, and
an injection placed closer to the hyaloid membrane will increase the
amount of uorescein glucuronide eliminated across the hyaloid membrane. Therefore, a lower retinal permeability will be required to t the
experimental data. Since uorescein glucuronide is known to penetrate the
retina at least to a small degree and the model has estimated a retinal
permeability of zero, it is reasonable to conclude that the extreme position
of the hyaloid-displaced injection is signicantly dierent than the actual
experimental injection position, leading to an error in the estimation of the
retinal permeability. This result was also due to the low sensitivity of the
uorescein glucuronide concentration at 24 hours to the retinal permeability. For the other three injection positions, as the injection site was moved
further from the hyaloid membrane, the net elimination across the hyaloid
was reduced. To compensate, the simulated amount of drug transferred
across the retina must increase, and, therefore, the estimated retinal permeability increases as the injection site was placed further from the hyaloid
membrane.
The retinal permeability of uorescein calculated by the models of
Araie and Maurice (8) and Yoshida et al. (12,13) and found in vitro by
Koyano (9) are close to the range of retinal permeabilities calculated by
the nite element model. Since the uorescein retinal permeability calculated
by Araie and Maurice (8) with their spherical model agrees with the retinal
permeability calculated with the nite element model, the uorescein permeability value is coincidentally the value that is required to balance the anterior and posterior losses. The retinal permeability calculated with a spherical
model for any compound that does not have an actual retinal permeability
similar to uorescein will be in error.

Copyright 2003 Marcel Dekker, Inc.

194

Friedrich et al.

Yoshida et al. (12,13) used their model to calculate the retinal permeability of uorescein and uorescein glucuronide in monkey eyes.
Therefore, any dierences between the retinal permeability calculated by
their model and the nite element model could be due to dierences
between the models and also due to dierences between the physiology of
the rabbit and monkey retina. However, a comparison of the two models
on a theoretical basis can still be made. Yoshida et al. (12,13) divided the
vitreous into an anterior and posterior hemisphere. Each hemisphere was
further subdivided into eight compartmental shells, and a separate permeability was used for the outer surface of each hemisphere. Within each
concentric compartmental shell, the concentration calculated by the
model would be uniform since each compartment was assumed to be perfectly mixed. As noted experimentally by Araie and Maurice (8), and with
the nite element model, the concentration contours of uorescein form
concentric rings that are parallel to the retina in the rabbit eye. The concentration contours in a monkey eye would be similar, due to the high
permeability of uorescein through the monkey retina. The concentration
contours calculated using Yoshidas model (12,13) for the posterior portion
of the vitreous, therefore, would be qualitatively correct. Since the anterior
portion of Yoshidas model (12,13) is also dened by concentric compartmental shells and the concentration contours in the region close to the
hyaloid membrane are not parallel to the retina, the concentration prole
calculated for the anterior portion of the vitreous would be incorrect.
Yoshidas model (12,13) is a more accurate representation of the true vitreous than a spherical model, since a separate permeability is calculated for
the outer surface of the anterior and posterior hemispheres. However, an
accurate estimate of the retinal permeability will only be obtained for compounds that are primarily eliminated across the retina. Compounds that are
eliminated mainly across the hyaloid membrane will have concentration
contours perpendicular to the retina, and, therefore, the concentric compartmental shells will not accurately calculate the correct concentration
prole. The prole calculated by a model that uses concentric compartmental shells will always have concentration contours that are parallel to the
retina since the concentration within each compartmental shell must be
uniform. Furthermore, since the concentration in the vitreous (adjacent
to the retina) calculated by a concentric shell compartmental model will
be always uniform over the entire inner surface of the retina, the model will
be unable to account for conditions that result in nonuniform concentrations adjacent to the retina (such as displaced or irregular injections or
diseases that have a local eect on retinal permeability).
Although the retinal permeability of uorescein found in vitro by
Koyano et al. (9) is similar to the values found using the nite element

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

195

model, their retinal permeability of uorescein glucuronide is signicantly

dierent from the value found using the nite element model. To test the
value that was found by Koyano et al. (9), a simulation was performed
with the nite element model using their permeability value and a central
injection of uorescein glucuronide with the same injected concentration
used by Araie and Maurice (8). As mentioned previously, the concentration
contour lines of uorescein glucuronide were found to be perpendicular to the
retina at 24 hours, experimentally by Araie and Maurice (8), and by the nite
element model for any of the injection positions. However, when the retinal
permeability value found by Koyano et al. (9) was used in the model, the
concentration contour lines at 24 hours were found to be parallel to the retina,
similar to the proles for uorescein. If the proles found experimentally by
Araie and Maurice (8) are assumed to be correct, then it must be concluded
that the retinal permeability values found in vitro by Koyano et al. (9) are not
accurate. This may be due to the diculty of excising a retina from the eye
and maintaining its viability while permeation experiments are performed.

IV.

USING FINITE ELEMENT MODELING TO INVESTIGATE

CLINICAL CONDITIONS

A.

Modication of Finite Element Model to Anatomy and

Physiology of Human Eye

To determine the implications of changing conditions which may aect drug

distribution in the vitreous in a more clinically relevant setting, the nite
element model can be modied to match the geometry of the human eye (21)
(Fig. 7). The most signicant dierences between the posterior segment of a
rabbit and human eye are the size of the lens and the volume of the vitreous
humor. In a human eye, the lens occupies a smaller portion of the vitreous
than in the rabbit eye. The volume of the vitreous humor in the human eye is
approximately 4 mL, and the volume of the rabbit vitreous is approximately
1.5 mL. Another dierence between the human and rabbit eye is the ow
dynamics in the posterior aqueous humor chamber. The volumetric ow
rate of aqueous humor is similar in human and rabbit eyes; however, the
cross-sectional area for ow is approximately four times larger in human
eyes, resulting in a lower ow velocity in human eyes. Similar to the rabbit
eye model, uorescein and uorescein glucuronide were used as test compounds to study the eects of injection position and volume in the human
eye model. In humans, the exact retinal permeabilities of uorescein and
uorescein glucuronide are not known, so the values used were 2:6  105
cm/s and 4:5  107 cm/s, respectively, which are the average of the values
that were found using the rabbit eye nite element model. For the purpose

Copyright 2003 Marcel Dekker, Inc.

196

Friedrich et al.

Figure 7 Cross-section view of human eye model. The model was based on the
physiological dimensions of the human eye. The posterior chamber of the aqueous
humor was included to account for the loss of drug from the vitreous across the
hyaloid membrane. Transport of drug by convection and diusion was accounted
for in the posterior chamber of the aqueous humor, and only diusive transport was
accounted for in the vitreous humor.

of examining the eects of changing intravitreal injection variables, the

accuracy of the retinal permeability is not of prime importance. The retinal
permeability value will impact the quantitative results of each injection
variable case studied, but qualitative intercase comparisons using a constant
retinal permeability will be valid.
B.

Effects of Injection Position and Volume on Drug

Distribution in the Vitreous

1. Injection Conditions Studied

The actual position and shape of an intravitreal injection will most likely not
be the same as any of the initial conditions that are described in Sec. III.A;
however, the model results indicated the variability that can occur when the
injection is not placed in the same position each time. Not only is knowledge

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

197

of the actual injection position and shape required to calculate the correct
retinal permeability, it is also very important for calculating the correct
concentrations within the vitreous. Dierent injection positions and shapes
will produce dierent concentrations within the vitreous, and, therefore, the
ecacy of the treatment produced by the drug will change. Although other
authors have suggested that the location of an intravitreal injection is signicant with respect to drug toxicity (16), a detailed study has not been
performed due to the diculty of experimentally measuring drug distribution within the vitreous.
Similar to the rabbit eye nite element model, four extreme injection
positions were studied using the human eye nite element model: (a) a central
injection, (b) an injection displaced towards the lens, (c) an injection displaced towards the retina, and (d) an injection displaced towards the hyaloid
membrane. The most common intravitreal injection volume in human
patients is 100 mL; however, dierent volumes may be used for clinical trials
and in animal studies (8,12,13). The two injection volumes compared using
the model were 15 and 100 mL. The mass injected in each case was the same
(30 mg), resulting in drug solution concentrations of 2000 and 300 mg/mL for
the 15 and 100 mL injections, respectively. In each of the 100 mL injection
cases, the distance between the outer boundary of the injected drug and the
adjacent tissue surface was the same as for the corresponding 15 mL injection.
Therefore, the overall displacements of the 100 mL injections from the central
injection were not as large as the displacements of the 15 mL injections due to
the larger diameter sphere that was used to represent the 100 mL injection.
At any time following an intravitreal injection, there may be signicant
concentration gradients within the vitreous due to the localized initial distribution and the fact that in the nite element model, mass transfer within
the vitreous occurs by diusion alone. Changing the variables of an intravitreal injection dramatically aects the concentration gradients and local
concentrations. Therefore, the concentrations within the vitreous were monitored at three dierent sites (Fig. 8): (a) adjacent to the lens on the symmetry axis, (b) adjacent to the retina on the symmetry axis, and (c) adjacent
to the hyaloid membrane. For the case where the injection was placed close
to the hyaloid membrane, the concentration was also monitored at the
opposite side of the vitreous adjacent to the hyaloid membrane.
2.

Results of Changing Injection Conditions

Figures 9 and 10 show the model calculated concentrations in the vitreous

adjacent to the retina on the symmetry axis of the vitreous following 15 mL
injections of uorescein or uorescein glucuronide, respectively, at the four
dierent vitreous locations. Tables 2 and 3 contain the peak concentrations

Copyright 2003 Marcel Dekker, Inc.

198

Friedrich et al.

Figure 8 Sites within vitreous where concentrations were monitored using the
nite element model. In addition to studying the eects of injection position and
volume on mean concentrations within the vitreous, concentrations were also monitored at several distinct points within the vitreous. The variation of drug concentrations within the vitreous will be revealed using this method of analysis.

and area-under-the-curve (AUC) values for all the vitreous sites and injection locations of uorescein and uorescein glucuronide.
In general, the data illustrate that injecting uorescein or uorescein
glucuronide at dierent locations within the vitreous may produce signicantly dierent local concentrations and concentration proles. The peak
concentrations that were calculated by the model depended on the location
of the injection relative to the site where the concentration was monitored.
The largest variations in the peak concentrations of uorescein were
observed at the vitreous site adjacent to the retina, where the retina-displaced injection produced a peak concentration over three orders of magnitude higher than that obtained from the hyaloid-displaced injection (Table
2). At the vitreous site adjacent to the hyaloid membrane, the peak concentration of uorescein produced by the hyaloid-displaced injection was also
over three orders of magnitude higher than the concentration produced by
the retina-displaced injection. Peak concentrations of uorescein at the vitr-

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

199

Figure 9 Concentration of uorescein at the vitreous site adjacent to the retina

following 15 mL injections at four dierent vitreous locations. Peak concentrations of
uorescein at this vitreous site varied by up to three orders of magnitude, depending
on the initial injection position. Similar variations were noted for the other sites
within the vitreous where the concentrations were monitored.

eous site adjacent to the lens varied by up to two orders of magnitude,

depending on the location of the injection (Table 2). Similar variations
were also noted for uorescein glucuronide; peak concentrations produced
by the dierent injection sites varied by up to two orders of magnitude at the
dierent vitreous sites (Table 3).
It is also important to note that there were signicant variations in the
model calculated peak concentrations at each vitreous site produced by an
individual injection location. For example, for the hyaloid-displaced injection, the peak concentration next to the injection (adjacent hyaloid) was over
three orders of magnitude higher than the peak concentration next to the
hyaloid membrane opposite the injection location (Table 3). Furthermore,
the time to reach the maximum uorescein concentration varied from
approximately 6 minutes to 12 hours, depending upon the proximity of the
monitoring site to the injection site. Similarly, the time to reach the peak
concentration of uorescein glucuronide ranged from approximately 10 minutes to 24 hours. The variations observed for the AUC-time curve followed

Copyright 2003 Marcel Dekker, Inc.

200

Friedrich et al.

Figure 10 Concentration of uorescein glucuronide at the vitreous site adjacent to

the retina following 15 mL injections at four dierent vitreous locations. Peak concentrations of uorescein glucuronide varied by up to three orders of magnitude at
this site within the vitreous, depending on the initial position of the injection. Similar
variations in concentrations were noted for the other sites within the vitreous.

trends similar to those observed for the peak concentration values. For
uorescein, AUC values for a specic vitreous site varied by over two orders
of magnitude, depending on the location of the injection. AUC values for
uorescein glucuronide for a specic vitreous site also varied by over an
order of magnitude. The AUC represents the cumulative exposure of a particular tissue to a drug over time and may therefore be related to the ecacy
and/or toxicity of a drug at a particular site.
The data in Table 3 also further emphasize the importance of the
injection location. After 24 hours there is one-quarter the amount of uorescein left in the vitreous following the 15 mL hyaloid-displaced injection
compared to after the 15 mL central injection. Similarily, there is one-third
the amount of uorescein glucuronide left in the vitreous following the 15
mL hyaloid-displaced injection compared to after the 15 mL retina-displaced
injection. For the larger 100 mL injections, the dierence between the
amount of drug remaining at 24 hours for the dierent injection locations

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

201

Table 2 Peak Concentrations and Area Under Curve Values at Various Vitreous Sites
Following 15 mL Intravitreal Injections of Fluorescein at Four Different Vitreous
Locations
Vitreous site where concentration was monitored
Injection
location

Adjacent
retina

Adjacent
lens

Peak
concentration
(mg=mL

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

9.53 (3.8)
6.77 (3.2)
0.989 (7.9) 628 (0.13)
1.52 (9.9)
563 (0.10)
5.73 (3.3)
0.166 (12)

AUC
(mg h/mL)
024 h

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

56.7
13.9
548
2.44

104
800
22.4
49.6

Adjacent
hyaloid

Adjacent
opposite
hyaloid

0.673 (6.9)
2.97 (2.8)
0.154 (11)
210 (0.10)
0.084 (9.3)
8.72
23.6
2.22
217

1.22

Values in parentheses indicate the time (hours) to reach the peak concentrations.

Table 3 Peak Concentrations and Area Under Curve Values at Various Vitreous Sites
Following 15 mL Intravitreal Injections of Fluorescein at Four Different Vitreous
Locations
Vitreous site where concentration was monitored
Injection
location
Peak
concentration
(mg=mL

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

AUC
(mg h/mL)
024 h

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

Adjacent
retina

Adjacent
lens

9.54 (3.8)
14.6 (4.6)
4.68 (21) 628 (0.13)
680 (0.14)
4.14 (21)
6.38 (3.4)
2.15 (24)
244
77.1
1290
28.6

138
843
66.6
79.9

Adjacent
hyaloid

0.904 (11)
3.04 (3.1)
0.742 (24)
210 (0.10)
0.188 (22)
17.3
32.7
10.4
260

Values in parentheses indicate the time (hours) to reach the peak concentrations.

Copyright 2003 Marcel Dekker, Inc.

Adjacent
opposite
hyaloid

3.15

202

Friedrich et al.

Table 4 Mean Vitreous Concentrations of Fluorescein and Fluorescein

Glucuronide 24 Hours After 30 mg Injections
Intravitreal injection volume
Injection location

15 mL

100 mL

Mean concentration of
uorescein in vitreous
(mg=mL)

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

0.603
0.564
0.339
0.159

0.527
0.535
0.361
0.214

Mean concentration of
uorescein glucuronide
in vitreous (mg=mL

Central
Lens-displaced
Retina-displaced
Hyaloid-displaced

5.16
3.73
5.71
1.87

5.11
4.09
5.60
2.49

is not as large; however, variations of up to a factor of 2.5 are still observed.

As the injection volume increases, the eect of injection position would be
dampened, with the upper limit of injection volume being a single injection
that completely replaces the whole vitreous.
Table 4 lists the mean model calculated concentrations of uorescein
and uorescein glucuronide in the vitreous as a function of the injection
position, following either a 15 or a 100 mL injection. As mentioned earlier,
the simulated mass of drug injected in each case was the same (30 mg). As
indicated by the lower values of the mean uorescein concentration at 24
hours, elimination of uorescein was faster for the 100 mL central and lensdisplaced injections than for the 15mL injections. However, when the uorescein was injected closer to the retina or hyaloid membrane, elimination
was higher for the 15 mL injections. These results are consistent with the fact
that uorescein is primarily eliminated across the retina. When the 15 mL
injection is placed next to the retina, the ux of drug across the retina
immediately after the injection is faster than for the 100 mL injection, due
to the higher concentration within the 15 mL injection. This is illustrated by
Figure 11, which shows the concentration of uorescein adjacent to the
retina on the symmetry axis of the vitreous following both the 15 and 100
mL retina-displaced injections. For the central and lens-displaced injections
of uorescein, elimination was higher for the 100 mL injection than for the
15 mL injection. In these cases, the 100 mL injection produces a higher
concentration at the surface of the retina than the 15 mL injection, resulting
in a higher rate of elimination. For injections of uorescein glucuronide, the
same trends were observed. Since uorescein glucuronide is eliminated

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

203

Figure 11 Concentration of uorescein at the vitreous site adjacent to the retina

following a 15 or 100 mL injection adjacent the retina on symmetry axis of vitreous.
The mass of uorescein injected in each case was identical, resulting in higher peak
concentrations adjacent to the retina following the 15 mL injection case and, therefore, a higher initial loss of uorescein across the retina.

mainly across the hyaloid membrane, the 15 mL injections placed closer to

the hyaloid membrane (hyaloid-displaced and lens-displaced) resulted in
lower mean concentrations at 24 hours than the 100 mL injections at the
same locations, due to a higher initial rate of elimination across the hyaloid
membrane. Figure 12 shows the concentration adjacent to the hyaloid membrane for the 15 and 100 mL hyaloid-displaced injections of uorescein
glucuronide. Similar to uorescein, when the injection of uorescein glucuronide was not placed next to its elimination surface (central and retinadisplaced), higher elimination is produced by the 100 mL injection.
3.

Clinical Implications of Changes in Injection Conditions

From a clinical perspective, the results of changes in injection conditions are

very signicant. Retinal damage from excessive drug concentrations is
observed periodically following an intravitreal injection. The results of this

Copyright 2003 Marcel Dekker, Inc.

204

Friedrich et al.

Figure 12 Concentration of uorescein glucuronide at the vitreous site adjacent to

the hyaloid membrane following a 15 or 100 mL injection adjacent to the hyaloid
membrane. The mass of uorescein glucuronide injected in each case was identical,
resulting in a higher peak concentration, adjacent to the hyaloid membrane following
the 15 mL injection case and, therefore, a higher initial loss of uorescein across the
hyaloid membrane.

modeling work show that enormous variations in local concentrations can

arise due to variations in injection positions, suggesting that small deviations
from a central injection position may contribute to retinal damage. For
treatment of infectious diseases, a specic minimum inhibitory level of
drug must be maintained for a specied length of time to eradicate the
infectious agent. Due to the potentially devastating eects of a vitreoretinal
infection, the antibiotics are used at the highest possible nontoxic doses (22);
however, these studies indicate that a nontoxic dose injected at one location
of the vitreous may be toxic if injected in a dierent location. Furthermore,
the time interval the antibiotic drug remains above its therapeutic concentration at a specic site in the vitreous is dependent on the injection location.
The results are also relevant to early clinical trials or experiments with
animal models, when the ecacy of new drugs is tested. Signicant
variability in the results may occur if care is not taken to ensure that
the conditions of the intravitreal injections are kept constant. Although

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

205

the injection positions that were examined in this study are extremes within
the anatomy of the eye, a variation of only 58 mm from a central injection
will produce these extremes. Slight changes in the injection conditions can
easily produce these variations. Knowledge of concentration variations that
are present at dierent sites within the vitreous will facilitate the optimization of administration techniques for diseases that aect the posterior segment of the eye.
C.

Effects of Aphakia and Changes in Retinal Permeability

and Vitreous Diffusivity on Drug Distribution in the
Vitreous

Posterior segment infections that result in endophthalmitis most often occur

as a complication following cataract extraction, anterior segment procedures, and traumatic eye injuries (2325). Vitreoproliferative disease, a disorder in which there is uncontrolled proliferation of nonneoplastic cells,
accounts for the majority of failures following retinal detachment surgery
(26). A common result of both of these diseases states is inammation of the
retina, which results in a breakdown of the blood-retinal barrier (27). Longterm diabetes is also known to result in a breakdown of the blood-retinal
barrier (28). The permeability of the retina will be aected as a result of
these disorders and will depend on the extent to which the blood-retinal
barrier has been compromised. The retinal permeability of compounds normally unable to cross the blood-retinal barrier will be increased; however,
the retinal permeability of compounds that are normally actively transported across the retina may actually decrease due to a disruption in the
active transport processes. Another transport parameter that may change
indirectly with changes in the pathophysiology of the eye is the diusivity of
drugs in the vitreous. Changes in drug diusivity will be most signicant
when drugs of dierent molecular weight are used to treat dierent pathological conditions. The developed human eye nite element model was used
to estimate how the pathophysiology of the posterior eye segment aects the
distribution and elimination of drug from the vitreous (29). In particular,
the eect of three conditions were examined: changes in the diusivity of
drugs in the vitreous, changes in retinal permeability, and, since it is common to inject drugs into aphakic eyes, the presence or absence of the lens.
1.

Range of Vitreous Diffusivity and Retinal Permeability

Values Considered

In order to cover a large number of drugs with a wide range of physicochemical properties, retinal permeabilities between 1  107 and 1  104

Copyright 2003 Marcel Dekker, Inc.

206

Friedrich et al.

cm/s were considered. Retinal permeabilities have been estimated for only a
small number of compounds, including uorescein (2:6  105 cm/s), uorescein glucuronide (4:5  107 cm/s), and dexamethasone sodium m-sulfobenzoate (4:9  105 cm/s) (1,9,1517,30). All of the reported values fall
within the range of permeabilities that were studied.
The vitreous is composed of water and low concentrations of collagen
and hyaluronic acid. As the vitreous ages, the concentration of collagen and
hyaluronic acid increases; however, even when elevated, the concentrations
are still relatively low, at 0.13 mg/mL and 0.4 mg/mL, respectively (31). It has
long been accepted that the diusivity of solutes in the vitreous is unrestricted
(32). An empirical relationship developed by Davis (33) can be used to determine if the concentration of collagen and hyaluronic acid would aect drug
diusivity in the vitreous. The diusivity of a substance in a hydrogel can be
estimated relative to its free aqueous diusivity using the following equation:



DP
exp  5 104 Mw Cp
Do
where DP and Do represent the hydrogen (vitreous) diusivity and the diffusivity in a polymer-free aqueous solution, respectively, MW represent the
molecular weight of the diusing species, and CP represents the concentration of polymer (collagen and hyaluronic acid) in the hydrogel in units of
grams of polymer per gram of hydrogel. Using the sum of the maximum
concentration of collagen and hyaluronic acid 5:3  104 g/g) as CP and the
molecular weight of uorescein (330 Da) gives a DP to Do ratio of 0.997.
This value indicates that the diusivity of a small molecule like uorescein in
the vitreous is virtually identical to the diusivity of uorescein in a polymer-free aqueous solution. Even if a molecular weight of 100,000 Da is used,
the ratio of DP to Do is still 0.992, indicating that for virtually all drugs of
interest, the diusivity in a free aqueous solution is an accurate representation of vitreous diusivity. This conclusion will hold for any molecule that
does not have some form of binding interaction with collagen and hyaluronic acid. The diusivity of molecules that do not interact with hyaluronic
acid and collagen is simply a function of the molecular weight of the diusing species. The molecular weight of drugs administered to the vitreous fall
within a range of approximately 10010,000. Davis (33) estimated the diffusivity of Na125 I (125 Da), [3H]prostaglandin F2/ (354 Da), and 125Ilabeled bovine serum albumin (67,000 Da) in water. Although these compounds would not be administered therapeutically to the vitreous, their
diusivities represent a reasonable range of values for testing the sensitivity
of drug distribution and elimination using the model. Therefore, the diusivities used in the model simulations are 2:4  105 cm2 /s (125 Da), 5:6 
106 cm2 /s (354 Da), and 5:4  107 cm2 /s (67,000 Da).

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

207

The eects of changing the retinal permeability or vitreous diusivity

were studied using the phakic eye model. When the sensitivity to the vitreous
diusivity was studied, the retinal permeability was held constant at
5  105 cm/s. Likewise, when the sensitivity to the retinal permeability
was studied, the vitreous diusivity was held constant at 5:6  106 cm2 /s.
When the eects of changing the vitreous diusivity and retinal permeability
were studied in the phakic eye model, only a central injection was considered
to reduce the number of variables that were changed.
2.

Modications to Finite Element Model to Simulate Aphakic

Eyes

Although cataract extractions previously involved removal of the entire

lens, it is more common today to leave the posterior lens capsule intact in
order to reduce postoperative complications such as vitreous changes and
retinal detachment (34). To study elimination in an aphakic eye, the human
phakic eye model was modied so that the curved barrier formed by the lens
(Fig. 7) was replaced by the posterior capsule of the lens (Fig. 13). All of the
other tissues of the aphakic eye model were assumed to be in the same

Figure 13

Cross-section view of aphakic human eye model.

Copyright 2003 Marcel Dekker, Inc.

208

Friedrich et al.

conguration as in the phakic eye model. The values noted earlier for the
retinal permeability of uorescein and uorescein glucuronide were also
used in the aphakic model to study the eects of removing the lens on the
elimination of compounds that have either a high or a low retinal permeability. The diusivity of uorescein and uorescein glucuronide used for
the vitreous and hyaloid membrane was 6:0  106 cm2 /s, which is the same
as the diusivity in free solution (35). Kaiser and Maurice (30) studied the
diusion of uorescein in the lens and concluded that the mass transfer
barrier formed by the posterior capsule of the lens was the same as an
equal thickness of vitreous. The drug diusivity used within the posterior
lens capsule, therefore, was also 6:0  106 cm2 /s.
3. Results of Changes in Vitreous Diffusivity and Retinal
Permeability
The eects of changing the retinal permeability and vitreous diusivity are
summarized in Table 5. The results agree with what would be expected based
on mass transfer principles. The eect of vitreous diusivity was examined
with the retinal permeability set to an intermediate value of 5:0  105 cm/s,
such that both the hyaloid membrane and the retina are expected to be
important elimination routes. Decreasing the drug diusivity through the
vitreous increases the time required for drug molecules to travel from the
injection site to an elimination boundary. Accordingly, the mean concentrations in the vitreous, calculated at 4, 12, and 24 hours after injection,
increased as the drug diusivity was reduced. Furthermore, the rate of
drug elimination, which is inversely related to the drugs elimination halflife, decreased signicantly as the drug diusivity was reduced. (Note: The
half-life noted in these studies is not the terminal phase half-life normally
quoted for a drugs pharmacokinetic properties, but rather the time required
for the average concentration in the vitreous to drop by a factor of two
immediately following injection.) At the lowest diusivity considered
(5:4  107 cm2 /s), the mean intravitreal concentration at 24 hours was
only 7.5% lower than the concentration at 4 hours. In contrast, at the highest
diusivity examined 2:36  105 cm2 /s), the mean vitreal concentration
decreased by more than 99% between 4 and 24 hours. Consequently, drug
diusivity can have a drastic eect upon drug distribution and elimination.
Table 5 shows the peak concentrations in the vitreous adjacent the lens
were only slightly aected by changes to the drug diusivity. However, the
time at which the peak concentration occurred increased as the drug diusivity decreased because the average time required for a drug molecule to
reach the lens increased. In the regions adjacent to the retina and hyaloid
membrane, the peak concentrations increased as the drug diusivity

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

209

Table 5 Sensitivity of Half-Life, Mean Vitreous Concentration, and Peak Vitreous

Concentration to Drug Diffusivity and Retinal Permeability
Cmean in vitreous
(mg=mL
Dvit
(cm2 /s)
107
5.4
56
236
56
56
56
56

Pretina
(cm/s)
107

t1=2
(hr)a

4h

12 h

24 h

500
63.9
500
7.64
500
2.69
1.0 44.4
10
31.1
100
12.3
1000
6.94

7.97
6.49
2.54
7.88
7.80
7.30
6.21

7.93
2.22
0.264
7.0
6.53
4.07
1.90

7.37
0.435
0.0169
5.69
4.79
1.54
0.323

Cpeak in vitreous
(mg=mL
Adjacent
lens
9.51
9.54
9.48
9.54
9.54
9.54
9.53

Adjacent
retina

(38.6)
0.591 (26.5)
(3.89)
4.36 (2.97)
(0.897) 9.26 (0.897)
(3.89) 14.8 (5.44)
(3.89) 14.1 (4.67)
(3.89)
9.86 (3.89)
(3.89)
2.59 (2.66)

Adjacent
hyaloid
0.366
0.621
0.775
0.915
0.876
0.742
0.587

(61.7)
(7.03)
(1.75)
(11.3)
(11.3)
(7.78)
(6.25)

All values were obtained using a central injection into the vitreous.
Values in parentheses indicate the time (hours) to reach the peak concentration.
a
The half-life noted in these studies is not the terminal phase half-life, but rather the time required for
the average concentration in the vitreous to drop by a factor of 2 immediately following injection.

increased, and the time to reach the peak concentration decreased as the drug
diusivity increased. When the vitreous diusivity is high, the retina limits
the rate of elimination. Therefore, the drug rapidly diuses to the retina, but
cannot be rapidly transferred across the retina, leading to a high concentration adjacent to the retina. When the diusivity is low, the retina is no longer
a rate-limiting barrier, and molecules are eliminated almost as soon as they
reach the retina surface, and the concentration adjacent to the retina is low.
The diusivity within the hyaloid membrane is equal to the diusivity within
the vitreous, and, therefore, the hyaloid membrane is never a rate-limiting
transport barrier. The observation that peak concentrations adjacent to the
hyaloid membrane decrease as the vitreous diusivity decreases is due to the
proximity of the hyaloid membrane to the retina.
Changing the retinal permeability had similar eects on the mean and
peak concentrations within the vitreous (Table 5). The eect of retinal permeability was examined with the vitreous diusivity held constant at an
intermediate value (5:6  106 cm2 /s). Increasing the retinal permeability
increases the rate of elimination; consequently, the mean concentration in
the vitreous at 4, 12, and 24 hours decreases as the retinal permeability
increases. The increase in elimination rate is also apparent by the fact that
the half-life dramatically decreases as the permeability increases. At the lowest retinal permeability considered 1:0  107 cm/s), the mean concentration

Copyright 2003 Marcel Dekker, Inc.

210

Friedrich et al.

in the vitreous at 24 hours was approximately 27% lower than at 4 hours. In

contrast, when the retinal permeability was 1:0  104 cm/s, the mean vitreal
concentration at 24 hours was 95% lower than the concentration at 4 hours.
Peak concentrations and peak times in the vitreous adjacent to the lens
were virtually unaected by changes to the retinal permeability. The largest
changes in the peak concentrations were noted adjacent to the retina, where
changing the retinal permeability by four orders of magnitude caused a
sixfold variation in peak concentrations. As the retinal permeability
increases, it is less likely to be a rate-limiting barrier. Therefore, where the
permeability is high, drugs are eliminated faster, leading to a lower concentration adjacent to the retina.
Figure 14 contains a plot of the half-life of a drug within the vitreous
as a function of either its vitreous diusivity or its retina permeability.

Figure 14 Dependence of half-life on vitreous diusivity or retinal permeability.

Note the half-life noted in these studies is not the terminal phase half-life, but rather
the time required for the average concentration in the vitreous to drop by a factor of
two immediately following injection.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

211

Similar relationships between retinal permeability, vitreous diusivity, molecular weight, and half-life have been shown by Maurice (32,36). Within the
range studied, half-life is inversely dependent on the vitreous diusivity and
retinal permeability. The half-life has a greater dependence on the vitreous
diusivity than on the retinal permeability, although neither relationship is
linear. As the retinal permeability either decreases towards zero or increases
to a high value, the half-life approaches either a high or a low limit, respectively. This is consistent with expectations because all drug is eliminated
across the hyaloid membrane when the retinal permeability is zero.
Therefore, the half-life will be dependent on the rate at which drug reaches
the hyaloid membrane, which is determined by the drug diusivity through
the vitreous. Likewise, when the retinal permeability is high, the rate of
elimination will be limited by the rate of diusion across the vitreous.
Although the range of drug diusivities considered is not large enough to
show the eect of extreme values of diusivity on half-life, it is expected that
as the vitreous diusivity decreases, the half-life should increase without
bound. However, as the vitreous diusivity increases, drug elimination
would occur primarily through the hyaloid membrane into the aqueous
humor and ultimately through the aqueous/blood barrier. Since diusivity
in the aqueous humor should be at the same as in the vitreous and hyaloid,
the owing aqueous humor should not represent a limiting mass transfer
barrier. Although the nite element model did not account for the aqueous/
blood barrier, the properties of this barrier would dictate the lower limit of
vitreous half-life when vitreous diusivity increases to large values.
Most drugs administered intravitreally have molecular weights ranging from 300 to 500 Da; therefore, Figure 14 (for a vitreous diusivity of
5:6  106 cm2 /s, 354 Da) will be representative of most drugs. However, for
smaller or larger compounds, the quantitative relationship between half-life
and the permeability will be dierent, as will the limiting values.
Nevertheless, the same qualitative relationship should still be observed,
regardless of the vitreous diusivity. Consequently, Figure 14 permits qualitative comparisons between the elimination of dierent drugs (molecular
weight aects diusivity). Furthermore, Figure 14 demonstrates the importance of dose adjustment if a drug is administered into an eye compromised
by retinal inammation or other disease that alter the permeability of the
blood-retinal barrier.
4.

Results of Aphakia on Drug Distribution in the Vitreous

Figure 15 shows the model calculated concentration prole of uorescein on

half of a cross section of the vitreous 24 hours after a central intravitreal
injection in the phakic and aphakic eye models. The concentration contours

Copyright 2003 Marcel Dekker, Inc.

212

Friedrich et al.

Figure 15 Model calculated concentration prole of uorescein on half of a crosssection of the vitreous 24 hours after a central intravitreal injection in the phakic eye
model (A) and the aphakic eye model (B).

are parallel to the retina, as expected, since uorescein is eliminated mainly

across the retina. Table 6 lists the half-lives, mean concentrations, and peak
concentrations of uorescein within the vitreous as a function of injection
position for both the phakic eye model or the aphakic eye model. Dierent

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

213

Table 6 Half-Life and Peak and Mean Vitreous Concentrations of Fluorescein

Calculated Using the Aphakic and Phakic Eye Models Following Intravitreal
Injections at Different Locations
Cmean in vitreous
(mg=mL
Injection
location

t1=2
(h)a

4h

Phakic
Central

8.36

6.61 2.61

0.60

8.08

6.49 2.49

0.564

3.54

3.79 1.27

0.339

1.39

2.11 0.695

0.158

8.38

6.61 2.47

0.646

3.54

3.72 1.26

0.312

3.75

3.84 1.35

0.303

2.29

2.41 0.626

0.146

Lens
Displaced
Retina
Displaced
Hyaloid
Displaced

Aphakic
Central
Lens
Displaced
Retina
Displaced
Hyaloid
Displaced

12 h

24 h

Cpeak in vitreous
(mg=mL
Adjacent
lens

Adjacent
retina

Adjacent
hyaloid

9.53
(3.78)
628
(0.128)
1.52
(9.89)
5.73
(3.28)

6.77
(3.17)
0.989
(7.94)
563
(0.119)
0.166
(12.1)

0.673
(6.89)
2.97
(2.83)
0.154
(11.1)
210
(0.104)
0.084b
(9.31)b

3.34
(4.33)
328
(0.093)
0.421
(11.2)
3.98
(2.03)

4.13
(4.33)
0.430
(10.3)
563
(0.131)
0.163
(12.9)

0.873
(5.22)
3.58
(2.01)
0.144
(11.2)
238
(0.137)
0.102b
(8.44)b

Values in parentheses indicate the time (hours) to reach the peak concentrations.
a
The half-life noted in these studies is not the terminal phase half-life, but rather the time
required for the average concentration in the vitreous to drop by a factor of 2 immediately
following injection. The terminal phase half-life would not be expected to change with
changes in injection position since the terminal phase occurs after a pseudo equilibrium has
been achieved in the vitreous. After this point only vitreous diffusivity and retinal permeability would govern the rate of elimination.
b
Peak concentration in vitreous adjacent hyaloid opposite the location of the intravitreal
injection.

trends were noted when comparing the half-life of uorescein in the phakic
versus aphakic eye model. In both cases, the longest half-life was found for a
central injection and the shortest half-life was found for a hyaloid-displaced
injection. The half-life for the lens-displaced injection, however, was much

Copyright 2003 Marcel Dekker, Inc.

214

Friedrich et al.

lower in the aphakic eye model than in the phakic eye model. Placing the
injected drug closer to the lens capsule in the aphakic eye model would
initially produce a rapid loss of drug to the posterior chamber of the aqueous humor. However, in the phakic eye model, since there is no loss across
the lens, injecting the drug closer to the lens has little eect. The initial drug
loss across the lens capsule in the aphakic eye model is conrmed by comparing, in the aphakic and phakic eye models, the ratio between the mean
concentrations at 4 and 24 hours for the central and lens-displaced injections. In the aphakic eye model, the mean concentration 4 hours following a
central injection is 1.75 times greater than the mean concentration from a
lens-displaced injection; this ratio increases slightly at 24 hours. In the
phakic eye model, however, this ratio is only approximately 1.02, despite
the fact that the mean concentration in the vitreous is the same for the
phakic and aphakic eye models 4 hours following a central injection. The
higher ratio in the aphakic eye model is therefore due to increased transport
across the lens capsule, much of which occurs within the rst 4 hours following an injection.
The mean vitreous concentrations in the phakic and aphakic eye models dier by less than 10% following central, retinal-displaced, and hyaloiddisplaced injections, regardless of the sample time considered. However, the
peak concentrations of uorescein adjacent to the lens and retina were
higher in the phakic eye model than in aphakic eye model for all the injection positions. Adjacent to the lens, the peak concentrations were higher in
the phakic eye model because there is no loss across the lens. Adjacent to the
retina, the peak uorescein concentrations were only signicantly higher in
the phakic eye model for the central and lens-displaced injections. This is
due to increased loss across the lens capsule in the aphakic eye model and
the fact that the distance between the injection site and the recording site is
slightly larger in the aphakic eye model than in the phakic eye model. The
peak concentrations adjacent to the hyaloid membrane were higher in the
aphakic eye model than in the phakic eye model for the central and lensdisplaced injections. This is due to the fact that, in the aphakic eye model,
the injection sites are slightly closer to the site adjacent to the hyaloid where
the concentrations were recorded.
Figure 16 shows the model calculated concentration prole of uorescein glucuronide in half of a cross section of the vitreous 36 hours after a
central injection in the phakic and aphakic eye models. In this case, since
uorescein glucuronide has a low retinal permeability and is eliminated
primarily across the hyaloid membrane, the concentration contours are
perpendicular to the surface of the retina. Table 7 lists the half-lives,
mean concentrations, and peak concentrations of uorescein glucuronide
within the vitreous as a function of injection position for both the phakic

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

215

Figure 16 Model calculated concentration prole of uorescein glucuronide on

half of a crosssection of the vitreous 24 hours after a central intravitreal injection
in the phakic eye model (A) and the aphakic eye model (B).

and aphakic eye models. It should be noted that the dierent half-lives
reported in Table 6 or Table 7, with respect to injection position, are due
to dierences in elimination rates immediately following the injection that
depend upon the injection position.

Copyright 2003 Marcel Dekker, Inc.

216

Friedrich et al.

Table 7 Half-Life and Peak and Mean Vitreous Concentrations of Fluorescein

Calculated Using the Aphakic and Phakic Eye Models Following Intravitreal
Injections at Different Locations
Cmean in vitreous
(mg=mL
Injection
location

t1=2
(h)a

4h

12 h

24 h

Phakic
Central

35.4

7.73

6.71

5.16

21.5

6.99

5.15

3.74

40.4

7.73

7.13

5.71

3.76

2.62

1.87

7.45

6.13

4.60

4.01

2.60

1.83

7.59

6.97

5.49

3.96

2.46

1.71

Lens
Displaced
Retina
Displaced
Hyaloid
Displaced

Aphakic
Central
Lens
Displaced
Retina
Displaced
Hyaloid
Displaced

3.33

30.4
4.08
38.1
3.97

Cpeak in vitreous
(mg=mL
Adjacent
lens

Adjacent Adjacent
retina
hyaloid

9.54
(3.77)
628
(0.128)
4.14
(21.0)
6.38
(3.4)

14.6
(4.61)
4.68
(21.0)
680
(0.138)
2.15
(24.0)

0.904
(11.0)
3.04
(3.11)
0.742
(24.0)
210
(0.104)
0.188b
(22.2)b

3.36
(4.78)
328
(0.093)
1.39
(21.6)
4.23
(2.11)

10.0
(6.53)
2.21
(21.7)
678
(0.130)
2.0
(24.0)

1.02
(6.53)
3.60
(2.01)
0.702
(21.6)
238
(0.137)
0.180b
(19.6)b

Values in parentheses indicate the time (hours) to reach the peak concentrations.
a
The half-life noted in these studies is not the terminal phase half-life, but rather the time
required for the average concentration in the vitreous to drop by a factor of 2 immediately
following injection. The terminal phase half-life would not be expected to change with
changes in injection position since the terminal phase occurs after a pseudo equilibrium has
been achieved in the vitreous. After this point only vitreous diffusivity and retinal permeability would govern the rate of elimination.
b
Peak concentration in vitreous adjacent hyaloid opposite the location of the
intravitreal injection.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

217

The rate of elimination from the vitreous at longer times (in the terminal phase) should be independent of the injection position. In general, the
half-life of uorescein glucuronide is higher than that for uorescein.
However, the elimination behavior observed with the phakic model and
the aphakic model is dierent for uorescein and uorescein glucuronide.
These dierences are due to the fact that uorescein glucuronide is eliminated mainly across the hyaloid membrane, rather than across the retina. In
both the aphakic model and the phakic model, the highest half-life occurred
for the retina-displaced injection and the lowest half-life occurred for the
hyaloid-displaced injection, which is consistent with the fact that the hyaloid
is the main elimination pathway. Similar to uorescein, the half-life following a lens-displaced injection was much lower in the aphakic model than in
the phakic model due to transport of drug across the lens capsule in the
aphakic eye model. Mean intravitreal concentrations of uorescein glucuronide at 12 and 24 hours are lower in the aphakic model for all the injection
locations considered.
A comparison of peak concentrations (Table 7) shows that uorescein
glucuronide concentrations adjacent to the lens and retina were consistently
lower in the aphakic eye model. However, concentrations adjacent to the
hyaloid membrane were typically higher following injection in the aphakic
eye model. Similar trends are observed for the peak uorescein concentrations (Table 6). The aphakic model calculated lower peak concentrations
near the retina and lens, for all the injection positions, but calculated higher
concentrations near the hyaloid membrane. Thus, this comparison of elimination in the aphakic and phakic eye models has indicated that not only does
the presence of the lens aect elimination, but the dierence in elimination
from an aphakic eye and a phakic eye is highly dependent on the injection
location and the retinal permeability of the drug. If the drug has a low retinal
permeability, then the half-life of the drug in an aphakic eye is highly dependent on the distance between the injection location and the lens capsule.

V.

SUMMARY

Finite element modeling has been shown to be a useful tool to study drug
distribution within the vitreous humor, with fewer limitations than previously developed mathematical models. Using a nite element model of
the vitreous, the site of an intravitreal injection was shown to have a substantial eect on drug distribution and elimination in the vitreous. The
retinal permeability of uorescein and uorescein glucuronide in rabbit
eyes calculated by the model ranged from 1.94 to 3:5  105 and 0 to 7:62 
107 cm/s, respectively, depending on the assumed site of the injection. The

Copyright 2003 Marcel Dekker, Inc.

218

Friedrich et al.

actual physiological retinal permeability will be a constant that is expected

to lie within these ranges. If the exact initial location and distribution of the
drug following injection were known, a single retinal permeability value
could be calculated using the model.
By using a nite element model that matched the geometry and physiology of the human eye, it was shown that variations in intravitreal injection conditions can produce radically dierent levels of drug exposure in
dierent sites within the vitreous. Variations in the injection location
resulted in peak concentrations that varied by over three orders of magnitude. These variations are very important to consider if toxicity to the retina
and other tissues is to be avoided. The mean calculated vitreous concentration 24 hours after an intravitreal injection varied by up to a factor of 3.8,
depending on the initial location of the injected drug. Changing the volume
of the injection from 15 to 100 mL dampened the eects of the initial injection location; however, mean concentrations at 24 hours still varied by up to
a factor of 2.5.
Using nite element modeling, it has also been shown that the rate of
drug elimination from the vitreous is highly dependent on diusivity through
the vitreous and retinal permeability. For a constant retinal permeability of
5:0  105 cm/s, increasing the vitreous diusivity from 5:4  107 to 2:4 
105 cm2 /s decreased the calculated half-life from 64 hours to 2.7 hours. For
a constant drug diusivity of 5:6  106 cm2 /s, increasing the retinal permeability from 1:0  107 to 1:0  104 cm/s decreased the calculated half-life
of drug from 44 to 7 hours. Therefore, the drug diusivity and retinal permeability are key factors that aected elimination from the vitreous and must be
considered when selecting drugs and doses, particularly if the blood-retinal
barrier has been compromised. Drug elimination was higher in an aphakic
eye model than in a phakic eye model, especially for drugs with a low retinal
permeability and if the injection was close to the lens capsule.
In the modeling work presented in this chapter, injection solutions
have been assumed to be spherical or cylindrical in shape. It is known,
however, that the distribution or shape of a drug solution within the
vitreous immediately following intravitreal injection may vary depending
on factors such as needle gauge and length, injection speed, solution viscosity, and vitreous rheology. Such variations in shape would inuence the
diusional surface area and hence drug distribution within the vitreous.
An attempt has been made to quantitate the eect of shape by using the
extent of ngering as a quantitative indicator of shape irregularity and
simulating intravitreal drug distribution using various shapes as the initial
condition (7). Although such simulations provide some insight into the
eect of shape, given the spurious nature of injections, it is dicult to
relate the results to any given injection.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

219

Another limitation of the models discussed in this chapter is that

transport of drug in the vitreous was assumed to occur only by diusion.
Vitreous liquefaction as a result of age or disease may result in pockets of
liqueed vitreous where there may be convective transport of drug. This
convection would dampen both the concentration gradients calculated by
the model and the eects of using dierent intravitreal injection conditions.
However, knowledge of exactly where liquefaction occurs or how much
convection occurs in liqueed pockets was not available at the time the
modeling was performed. When better knowledge of vitreous liquefaction
becomes available, this could be incorporated into the models.

REFERENCES
1. Lee, V. H. L., Pince, K. J., Frambach, D. A., et al. Drug delivery to the
posterior segment. In: Retina, T. E. Ogden and A. P. Schachat, eds. St.
Louis: C. V. Mosby, 1989, pp. 483498.
2. Forster, R. K., Abbott, R. L., and Gelender, H. Management of infectious
endophthalmitis. Ophthalmology 87:313319, 1980.
3. Pugfelder, S. C., Hernandez, E., Fliesler, S. J., Alvarez, J., Pugfelder, M. E.,
and Forster, R. K. Intravitreal vancomycin. Retinal toxicity, clearance, and
interaction with gentamicin. Arch. Ophth. 105:831837, 1987.
4. Stainer, G. A., Peyman, G. A., Meisels, H., and Fishman, G. Toxicity of
selected antibiotics in vitreous replacement uid. Ann. Ophth. 9:615618, 1977.
5. Tabatabay, C. A., DAmico, D. J., Hanninen, L. A., and Kenyon, K. R.
Experimental drusen formation induced by intravitreal aminoglycoside injection. Arch. Ophth. 105:826830, 1987.
6. Talamo, J. H., DAmico, D. J., Hanninen, L. A., and Kenyon, K. R., and
Shanks, E. T. The inuence of aphakia and vitrectomy on experimental retinal
toxicity of aminoglycoside antibiotics. Am. J. Ophth. 100:840847, 1985.
7. Lin, H.-H. Finite element modelling of drug transport processes after an
initravitreal injection. MASc. thesis, University of Toronto, 1997.
8. Araie, M., and Maurice, D. M. The loss of uorescein, uorescein glucuronide
and uorescein isothiocyanate dextran from the vitreous by the anterior and
retinal pathways. Exp. Eye Res. 52:2739, 1991.
9. Koyano, S., Araie, M., and Eguchi, S. Movement of uorescein and its glucuronide across retinal pigment epithelium-choroid. Invest. Ophth. Vis. Sci.
34:531538, 1993.
10. Ohtori, A., and Tojo, K. In vivo/in vitro correlation of intravitreal delivery of
drugs with the help of computer simulation. Biol. Pharm. Bull. 17:283290,
1994.
11. Tojo, K., and Ohtori, A. Pharmacokinetic model of intravitreal drug injection. Math. Biosci. 123:359375, 1994.

Copyright 2003 Marcel Dekker, Inc.

220

Friedrich et al.

12. Yoshida, A., Ishiko, S., and Kojima, M. Outward permeability of the bloodretinal barrier. Graef. Arch. Clin. Exp. Ophth. 230:7883, 1992.
13. Yoshida, A., Kojima, M., Ishiko, S., et al. Inward and outward permeability of
the blood-retinal barrier. In: Ocular Fluorophotometry and the Future, J. CunhaVaz and E. Leite, eds. Amsterdam: P Kugler & Ghedini Pub., 1989, pp. 8997.
14. Hosaka, A. Permeability of the blood-retinal barrier in myopia. An analysis
employing vitreous uorophotometry and computer simulation. Acta Ophth.
Suppl. 185:9599, 1988.
15. Larsen, J., Lund-Andersen, H., and Krogsaa, B. Transient transport across
the blood-retina barrier. Bull. Math. Bio. 45:749758, 1983.
16. Lund-Andersen, H., Krogsaa, B., la Cour, M., and Larsen, J. Quantitative
vitreous uorophotometry applying a mathematical model of the eye. Invest.
Ophth. Vis. Sci. 26:698710, 1985.
17. Lund-Andersen, H., Krogsaa, B., and Larsen, J. Calculation of the permeability of the blood-retinal barrier to uorescein. Graef. Arch. Clin. Exp.
Ophth. 222:173176, 1985.
18. Oguru, Y., Tsukahara, Y., Saito, I., and Kondo, T. Estimation of the permeability of the blood-retinal barrier in normal individuals. Invest. Ophth. Vis.
Sci. 26:969976, 1985.
19. Palestine, A. G., and Brubaker, R. F. Pharmacokinetics of uorescein in the
vitreous. Invest. Ophth. Vis. Sci. 21:542549, 1981.
20. Friedrich, S. W., Cheng, Y.-L., and Saville, B. A. Finite element modelling of
drug distribution in the vitreous humour of the rabbit eye. Ann. Biomed. Eng.
25(2):303314, 1997.
21. Friedrich, S. W., Saville, B. A., and Cheng, Y.-L. Drug distribution in the
vitreous humor of the human eye: the eects of intravitreal injection position
and volume. Curr. Eye Res. 16(7):663669, 1997.
22. Baum, J. Therapy for ocular bacterial infection. Trans. Ophthalmol. Soc. U.K.
10:569577, 1986.
23. Kattan, H. M., Flynn, H. W., Pugfelder, S. C., et al. Nosocomial
endophthalmitis survey. Current incidence of infection after intraocular surgery. Ophthalmology 98:227228, 1991.
24. Peyman, G. A., and Schulman, J. A. Intravitreal drug therapy. In: Intravitreal
Surgery. Norwalk: Appleton-Century-Crofts, pp. 407455.
25. Speaker, M. G., and Meniko, J. A. Postoperative endophthalmitis: pathogenesis, prophylaxis, and management. Int. Ophthalmol. Clin. 33:5179, 1993.
26. The Retina Society Terminology Committee. The classication of retinal
detachment with proliferative vitreoretinopathy. Ophthalmology 90:121125,
1983.
27. Goldberg, M. F. Diseases aecting the inner blood-retinal barrier. In: The
Blood-Retinal Barriers, Cunha-Vaz, J. G., eds. New York: Plenum Press, 1979,
pp. 309363.
28. Frank, R. N. The mechanism of blood-retinal barrier breakdown in diabetes.
Arch. Ophthalmol. 103:13031304, 1985.

Copyright 2003 Marcel Dekker, Inc.

Drug Distribution in Vitreous Humor

221

29. Friedrich, S. W., Saville, B. A., and Cheng, Y.-L. Drug distribution in the
vitreous humour of the human eye: the eects of aphakia and changes in
retinal permeability and vitreous diusivity. J. Ocular Pharm. Therap.
13(5);445459, 1997.
30. Kaiser, R., and Maurice, D. M. The diusion of uorescein in the lens. Exp.
Eye Res. 3:156165, 1964.
31. Sebag, J. Aging of the vitreous. Eye 1:254262, 1987.
32. Maurice, D. M., and Mishima, S. Ocular pharmacokinetics. In: Pharmacology
of the Eye. Vol. 69, Handbook of Experimental Pharmacology, Sears, M. L.,
ed. New York: Springer-Verlag, 1984, p. 72.
33. Davis, B. K. Diusion in polymer gel implants. Proc. Natl. Acad. Sci. USA
71:31203123, 1974.
34. Blinkhorst, C. Corneal and retinal complications after cataract extraction.
The mechanical aspect of endophthalmodonesis. Ophthalmology 87:609617,
1980.
35. Algvere, P. V., Hallnas, K., Dafgard, E., et al. Panretinal photocoagulation
aggravates experimental proliferative vitreoretinopathy. Graef. Arch. Clin.
Exp. Ophthalmol. 228:461466, 1990.
36. Tsukahara, Y., and Maurice, D. M., Local pressure eects on vitreous
kinetics. Exp. Eye Res. 60:563574, 1995.

Copyright 2003 Marcel Dekker, Inc.

7
Anterior Segment Microdialysis
Kay D. Rittenhouse
Pzer Inc., San Diego, California, U.S.A.

I.

INTRODUCTION

Preservation of sight is the major objective of many scientic studies.

Topically administered drugs eectively treat many important ocular diseases. Short-term ecacy endpoints are sometimes dicult to assess following these treatment approaches. Hence, it is important to determine
whether therapeutically relevant concentrations reach the site of action, as
in other regions of the body. The anterior chamber of the eye is a relatively straightforward region for sampling. Anatomically accessible by
paracentesis procedures, it is possible to obtain a single sample for measurement of drug concentrations. However, challenges are encountered
when time-course or steady-state data are collected. Repeat sampling of
this region is not possible by conventional methods in general.
Traditionally, rabbits or other mammal species have been used for the
assessment of intraocular concentrations of topically administered drugs.
In order to obtain time-course data in aqueous humor, many animals are
required, with each time point requiring multiple individual aqueous
humor samples following sacrice. These procedures present a number
of challenges to be managed.
The anterior segment is an interesting and important ocular region for
exploration with research tools such as microdialysis. More than 20 papers
describing microdialysis approaches for assessment of ocular drug delivery
and endogenous substrate characterization have been published, which
include both vitreous and aqueous humor sampling.

223

Copyright 2003 Marcel Dekker, Inc.

224

II.

Rittenhouse

PHYSIOLOGICAL CONSIDERATIONS OF THE

ANTERIOR SEGMENT

Aqueous humor, the watery solvent produced by the ciliary body in the
posterior chamber, is, in part, an ultraltrate of plasma (1). However, a
number of the electrolytes are present in higher concentration in aqueous
humor than in blood, providing evidence of active secretory and metabolic
components to aqueous formation. For example, ascorbate and lactate are
20-fold and 2-fold higher in concentration in aqueous relative to plasma,
respectively (2). Aqueous humor serves a nutritive role for avascularized
ocular tissues such as the cornea, trabecular meshwork, and lens (2).
A.

Aqueous Humor Formation and Turnover

1. Inow Dynamics
Blood, presented at the ciliary body arterioles at relatively high hydrostatic
pressure ( 30 mmHg) (35), is converted into aqueous humor through
complex and not completely characterized ways. The protein concentration
in aqueous humor is less than 1% of that present in plasma (1). Plasma
proteins are prevented from entry into aqueous humor by the tight junctions
located at the nonpigmented ciliary epithelium, a component of the so-called
blood-aqueous barrier, analogous to the blood-brain barrier (1). Active
secretion of electrolytes such as sodium, deposited at the intercellular clefts
of the tight junction regions of the nonpigmented ciliary epithelium, provide
for a concentration gradient favoring uid ow from the ciliary processes to
the posterior chamber (6). A number of active secretory pathways have been
identied (7,8) with specic active transport systems such as Na /K ATPase and others providing a major contribution. The formed aqueous
humor ows into the posterior chamber, down a pressure gradient, and is
transported via convective bulk ow through the pupil into the anterior
chamber, where the pressure is  16 mmHg (6).
2. Outow Dynamics
Return of aqueous humor to the systemic circulation is facilitated by the
lower pressure of the episcleral venous system ( 9 mmHg) relative to the
anterior chamber ( 16 mmHg), as aqueous percolates through the trabecular meshwork and collects into the canal of Schlemm (1). A second pressure-independent pathway, called the uveoscleral route, provides an
important contribution to aqueous outow in humans. In contrast, rabbits
have virtually no aqueous outow by this route (9). Resistance to ow, or
aqueous humor outow facility, is used to describe the passive resistance of

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

225

the trabecular meshwork to the passage of aqueous humor (10,11). The

pressure-independent ow pathway behaves like a constant-rate pump;
however, no metabolically dependent process has been identied as a driving force for pressure-independent ow (11). The uveoscleral pathway is
described as the slow entry of aqueous humor through the face of the ciliary
body just posterior to the scleral spur, with movement by bulk ow through
the tissue and absorption into the uveal vessels or into periocular orbital
tissues (10). There is considerable discussion concerning whether or not a
signicant energy-dependent component of the outow pathway exists (10).
The cells of the trabecular meshwork have phagocytic activity (1214),
which may contribute to increased facility of outow. Trabecular meshwork
outow is biologically active, providing biochemical modulation of a passive
physical process (10).
The relationship between inow and outow provides a means for
estimating the intraocular pressure (IOP). This relationship is described as:
IOP

F U
Pv
C

where F is aqueous humor formation, or ow, U is pressure-insensitive ow,

C is the facility of inow or pressure sensitive ow, and Pv is the episcleral
venous pressure (2).

III.

AQUEOUS HUMOR DYNAMIC IMPACT ON ANTERIOR

SEGMENT DRUG DISPOSITION

The pharmacokinetics of drugs in aqueous humor is complex. Aqueous

turnover, as well as availability of unbound substrate (i.e., tissue binding),
complicate the assessment of ocular clearance. Anterior chamber volume in
rabbit and humans is estimated to be  250300 mL. Aqueous humor turnover is  1% of anterior chamber volume  2:5 mL/min) (15). In the anterior chamber environment, volume and clearance are not independent in the
sense that drug clearance is a function of aqueous turnover and turnover
rate is a function, in part, of anterior chamber volume (16). The nature of
aqueous humor turnover and the pharmacodynamics of drugs that aect
aqueous formation can also complicate the characterization of drug disposition. As drug is absorbed and exerts the pharmacological eect resulting in
decreased aqueous humor formation, for example, the resulting aqueous
concentrations are elevated relative to that of substrates that would not
exert this eect (17). Tissue binding and drug lipophilicity, for example,
provide input into the dispositional characteristics of the drug. Systemic
eects can also inuence the ocular disposition of drugs. Analgesia may

Copyright 2003 Marcel Dekker, Inc.

226

Rittenhouse

result in decreased aqueous humor turnover (17), which in turn results in

elevated aqueous humor drug concentrations.

IV.

MICRODIALYSIS SAMPLING OF AQUEOUS HUMOR

A.

Important Problems in the Anterior Chamber

1. Anterior Segment Pharmacokinetics

The ocular pharmacokinetics of ophthalmic drugs has been evaluated for
many years by paracentesis sampling of anterior chamber aqueous. Lee et
al. (18) examined the systemic disposition of a series of beta-adrenergic
antagonists following topical administration to the pigmented rabbit (18)
in order to establish the relationship between the physicochemical drug
properties and absorption pathways. The eciency of nasolacrimal punctum occlusion for minimization of systemic exposure and increased local
absorption also was examined. Ross et al. (19) reported a propranolol
aqueous humor Cmax of  50005500 ng/mL ( 1011 ng/mL/mg, dose
normalized) in anesthetized rabbits with paracentesis sampling. Others
have examined the aqueous humor disposition of propranolol and other
beta-adrenergic antagonists using this sampling technique (1921).
Disadvantages to this approach include the large number of animals
required for evaluation and that paracentesis sampling is usually a terminal
procedure.
Rabbits are the species of choice for most ocular pharmacokinetic
experiments, although work in the cat, dog, and primate has been reported
(22,23). The rabbit eye is similar to the human eye in size and aqueous
humor volume. The rabbit eye has a thinner corneal thickness (0.35 mm
vs. 0.52 mm in humans) (9), slower blink reex (9), a nictitating membrane
(absent in humans) (24), and virtually no uveoscleral outow pathway (9).
2. Approaches to the Assessment of Modulation of Aqueous
Humor Inow and Outow
In order to study pharmacodynamics of drugs that aect aqueous humor
formation and turnover, a number of techniques have been developed.
Approaches such as uorophotometry have been used (25,26). In essence,
uorophotometry is a noninvasive technique that uses sophisticated instrumentation for the evaluation of the anterior chamber time course of topically or systemically administered uorescing compounds such as
uorescein or uorescein conjugates; the dilution of a topically or systemically administered dye in aqueous is measured without direct assay of aqueous humor contents. This procedure is advantageous for use in the clinical

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

227

setting due to its lack of invasive sampling. Measurement of uorescein in

the eye using uorophotometry is a somewhat complex procedure with a
number of possible sources of error (25). Tonography, also a noninvasive
approach, can be used to assess aqueous humor formation indirectly.
Briey, tonography tests the ability of the eye to recover from the elevation
of IOP induced by a tonometer. Such recovery primarily occurs through
increased outow of aqueous humor (27). Tonography depends on the
assumption that aqueous humor formation is insensitive to moderate
changes in IOP; the facility of trabecular meshwork outow is estimated.
This method neglects the pseudofacility component. With this approach, it
is dicult to separate out the dierent contributions to facility (27).
Constant pressure perfusion techniques have been used to estimate outow
facility (2830). A phenomenon described as a washing-out eect is commonly observed with the use of this method; the perfusion results in the
clearing of macromolecules (30) usually present at the trabecular meshwork
that partially occlude these outow channels (10). Inaccuracy in ow estimates can result since time dependent changes in facility are observed (29).
An invasive approach used by Miichi and Nagataki (31) to estimate aqueous
humor formation involves the assessment of the time course of dilution of
a nondiusable compound or dye, which is perfused at a constant rate into
aqueous humor. Perturbation of the rate of dilution of the dye can be
estimated via a change in the time course of the dye following administration of the pharmacological agent. The time-course data are approximated
mathematically employing nonlinear least-squares regression analysis in
order to obtain aqueous humor ow parameters perturbed by drugs that
aect inow. This method oers some attractive features in the quantitation
of the physiological eect. However, the technical procedures are quite
involved, with numerous intrusions simultaneously to the same eye (3136).
B.

Principles of Microdialysis: Probe Design and

Recovery

Microdialysis oers a novel means for obtaining samples of biological uids

while providing a relatively clean matrix, which may require little or no
sample preparation prior to analysis. However, microdialysis, in general,
does not provide meaningful information concerning endogenous or exogenous compounds implicitly. Microdialysis is a means of collecting the
sample for further analysis. The dialysate must be analyzed by other conventional analysis techniques. Such analytical techniques used in conjunction with microdialysis include high-performance liquid chromatography
(HPLC) (37,38), capillary electrophoresis (39,40), UV-visible spectrophotometry (41), and liquid scintillation spectroscopy (42).

Copyright 2003 Marcel Dekker, Inc.

228

Rittenhouse

1. Principles of Dialysis
Dialysis involves the separation of two compartments containing diering
concentrations of a solute in solution by a semi-permeable membrane. This
membrane allows passage of solutes of suciently small size from one compartment to the other along a concentration gradient. Theoretically, the
solute concentration in both compartments will establish equilibrium such
that there is no net ux of solute; the concentration of solute not bound to
nonpermeable macromolecules then will be equal in both compartments.
The solute diusion rate, as described by Ficks law, is a function of membrane surface area, thickness, concentration gradient, compartment volume,
and ligand diusion coecient (43).
Tissue and plasma proteins often bind drugs and other low molecular
weight compounds. Hypothesis regarding mechanisms of binding include
the generally held view that a reaction occurs between two oppositely
charged ions (essentially salt formation). Negatively charged drugs bind to
the positively charged amino acid groups, such as histidine or lysine, of
plasma proteins. Additional contributions to binding phenomena include
hydrophobic interactions (44). Nonpolar functional groups of drug and
protein or tissue interact via van der Waals forces.
Pharmacodynamic eects of drugs are considered to be a function of
the unbound concentration in plasma (45). For this reason, it is important
to determine the unbound (i.e., therapeutically relevant) concentration of
pharmacological agents. Dialysis techniques are well suited to make these
determinations. In the anterior chamber, low concentrations of proteins are
encountered (4). However, under conditions of compromised blood-aqueous barrier, an increased inux of proteins from plasma may result in
elevated aqueous protein concentrations (46). Under these conditions, the
assessment of unbound concentrations in aqueous humor may become more
important in the establishment of the pharmacodynamics arising from
intraocular exposures to the substrate in question.
Microdialysis is a dynamic process. Perfusion medium is perfused
through the probe. Analyte concentrations in perfusate and in the surrounding medium are not in equilibrium (41). This introduces a number of technical problems that must be overcome in creative ways. Microdialysis is a
relatively sophisticated tool. There are a number of challenges to appropriate use of this technique. Although nonspecic binding to the microdialysis
membrane is minimized as compared to other dialysis methods, plastic
tubing is used to deliver perfusate to the probe and to deliver the dialysate
from the probe to the collection vessel. Nonspecic binding to the tubing is
possible (47). This situation can be exacerbated when coupling microdialysis
directly to other instrumentation since longer tubing usually is required. In

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

229

experiments examining plasma protein binding of drug in vivo, microdialysis requires sucient time to achieve stable concentrations. This process
requires more time (than ultraltration, for example), and recovery of substrate across the membrane can be time- and temperature-dependent.
2.

Microdialysis Probe Design Issues

The unique and dynamic environment of the anterior chamber provides

features amenable to microdialysis sampling. Continuous ow of aqueous
humor about the probe tip prevents the creation of microenvironments near
the probe membrane. This is an important advantage for microdialysis use
in this organ as opposed to placement in other extracellular spaces. Special
problems that develop for specic placement into the anterior chamber
include brin formation (17), which must be circumvented in order to prevent reduced recovery of substrate in probe euent. Additionally, due to
possible disruption of the blood-aqueous barrier, protein inux may alter
the drug disposition of highly protein-bound substrates (17).
Specic advantages to microdialysis use for anterior chamber sampling are:
1. There is no extraction of internal aqueous humor uids; IOP is
not adjusted articially since aqueous humor volume is not
altered by inux of uids.
2. Microdialysis sampling of both pharmacological agent and endogenous substrate is performed simultaneously.
3. Although this method involves some intrusion via surgical placement of the microdialysis probe into the anterior chamber, a
more quantitative approach to the estimation of aqueous
humor formation rates and pharmacokinetic experimentation is
possible than with many conventional noninvasive approaches.
Fukada et al. (4850) used a linear probe design that involved both entry
and exit ports through the anterior chamber, a similar approach to that
taken by Macha and Mitra (51,52). For their work in conscious animals,
Rittenhouse et al. (19,53,54) modied a concentric microdialysis probe
design according to the scheme presented in Figure 1, incorporating a 908
bend in the probe shaft for each of anchoring the probe to the sclera of the
rabbit eye. In Figure 2, a photograph of an intact rabbit eye with a microdialysis probe in the anterior chamber is presented.
3.

Microdialysis Probe Recovery

A major concern in using microdialysis as a tool for the determination of

unbound drug concentrations in the in vivo as well as in vitro settings is the

Copyright 2003 Marcel Dekker, Inc.

230

Rittenhouse

Figure 1 Custom-designed microdialysis probe for posterior or anterior chamber

placement into the rabbit eye.

assessment of recovery. Recovery may be dened as the proportion of solute

extracted from the medium surrounding the probe (55). Recovery is dependent on the following parameters: dialysis membrane length, perfusion ow
rate, diusion rate of the solute through the compartment (the usual ratelimiting step in the process), and membrane properties (47). Recovery can

Figure 2 Rabbit eye with a microdialysis probe in the anterior chamber (magnication  4).

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

231

also be time- and temperature-dependent. Typical recovery values observed

in the literature range from a low of  10 up to 100%. By maximizing the
dialysis membrane length, signicant increases in recovery can be realized.
Decreases in perfusion ow rate also increase the relative recovery (although
they also decrease the available sample volume). The recovery of solute can
be dicult to ascertain. Ideally, probe perfusate composition should closely
match the environment of the medium in which it is placed. The probe also
can create a microenvironment near the probe surface, which may be different than the medium more distant from the membrane (47). Several
dierent types of recoveries are evaluated in microdialysis studies; these
include relative recovery (or concentration recovery) and absolute recovery
(mass recovery). Relative recovery is the fraction of solute obtained in the
dialysate relative to the actual concentration in the medium in which the
probe is placed. Absolute recovery is the total amount of solute collected
over a specied time period. A number of approaches are used to estimate
the recovery of a solute by the microdialysis probe, including water recovery, no-net-ux (or dierence method), perfusion rate, and relative loss
(55,56).
The water-recovery method is of limited use for in vivo settings
because drug diusion characteristics are usually dierent in articial aqueous physiological buers or solutions than in the dynamic in vivo environment. Where a solid in vitrotoin vivo correlation is established, the waterrecovery method has utility. For this method, the microdialysis probe is
placed in a reservoir (usually stirred) containing a known concentration of
solute. The perfusion medium, an aqueous solution of similar composition
to the medium in which it is placed but without solute, is delivered through
the probe at a constant rate. Dialysate is collected and the amount of solute
determined via appropriate analytical methods. The ratio of the dialysate
concentration of the known concentration of the medium in which the probe
is placed is the relative recovery. This method is known to underestimate the
concentration in the medium sampled (55).
The point of no-net-ux or dierence method is used for in vitro and
in vivo studies. By varying the concentration of solute in the perfusion
medium and xing the solute concentration in the surrounding medium,
the dialysate solute concentration is assessed. The direction of the concentration gradient of solute depends on whether the concentration in the
perfusion medium is higher or lower than the concentration in the surrounding medium (55). A plot of the perfusate solute concentration versus the
dierence in concentration between perfusate and dialysate is constructed;
the x-intercept identies the concentration at which no net ux of solute
occurs (55). In theory, this value will be the concentration of the surrounding medium. This method is very time-consuming.

Copyright 2003 Marcel Dekker, Inc.

232

Rittenhouse

The perfusion rate method is based on the principle that recovery is

dependent on the rate of perfusate transit through the probe. With an
increase in perfusate transit, there is a corresponding decrease in relative
recovery. Conversely, the lower the perfusion rate, the higher the relative
recovery. For the perfusion rate method, the initial surrounding medium
contains no solute and the probe is perfused with a xed concentration of
solute (55). This method is the most exhaustive in that several dierent
surrounding media concentrations must be assessed separately, each at different perfusion rates (in vitro). A typical experiment might evaluate four
dierent concentrations for the medium at three dierent perfusion rates
each over an extended period. Frequently regression models are employed
to provide the best estimates of probe performance. Typically for in vivo
determinations, the lowest possible perfusion rate (e.g., 0.1 mL=min) is
selected. This maximizes the relative recovery to nearly 100% in some
cases. At such low ow rates, longer collection times are required to obtain
sucient sample for further analysis.
The relative loss method is similar to the water-recovery method, but is
operated in reverse. Rather than placing a known concentration of solute in
the medium surrounding the probe, the solute concentration of the perfusate
is xed. The surrounding medium, which in most situations contains small
quantities of solute (i.e., sink conditions), then provides a negative concentration gradient of the solute. The net loss of solute reects the relative loss
of solute to medium. This method, which is based on the premise that
recovery is the same in both directions across the membrane (47), is by
far the simplest to use in the in vivo setting and provides a reliable estimate
of recovery. Relative loss is the ratio of the dierence in perfusate to dialysate solute concentrations to the perfusate concentration (56). This method
is often referred to as retrodialysis recovery. Under nonsink conditions of
the surrounding medium, an internal standard is sometimes employed.
C.

Anterior Versus Posterior Chamber Sampling

Anterior chamber aqueous microdialysis sampling approaches have been

explored by a number of researchers (16,17,4854). Challenges that were
managed using this approach include sensitivity of the eye to immunoprotective cascades following manipulation (17) and the requirement of the
protection of visual function, also a major concern for any procedures
proposed for observation of ocular pathophysiology or ocular pharmacokinetic/pharmacodynamic experimentation.
In published reports as early as the 1940s, researchers attempted to
obtain information regarding aqueous endogenous substrate concentrations
in the posterior versus the anterior chambers. Becker (57) and Kinsey and

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

233

Palmer (58,59) examined posterior chamber versus anterior chamber aqueous humor ascorbate concentrations. Rittenhouse et al. (53) used a microdialysis approach to estimate posterior versus anterior chamber ascorbate
aqueous concentrations; the probe tip was introduced into the anterior
chamber and directed through the pupil towards the posterior chamber.
The posterior chamber is a much smaller region ( 55 mL Vs  250 mL
for the anterior chamber) and provides additional challenges due to size
constraints.

V.

CASE STUDIES OF MICRODIALYSIS USE IN THE

ANTERIOR SEGMENT

A.

Ocular Pharmacokinetics

Recently, drug disposition in the anterior segment has been explored using
microdialysis. Fukuda et al. (4850) were the rst to examine the utility of
microdialysis sampling of anterior chamber aqueous humor. In their studies,
linear probes inserted into the temporal cornea through the anterior chamber and exteriorized out of the nasal cornea were used to examine intraocular disposition of uoroquinolones following oral or topical
administration of ooxacin, noroxacin, or lomeoxacin in the anesthetized
rabbit (48,49). Fukada et al. (48) characterized the ocular pharmacokinetics
(Cmax , Tmax , T1=2 ) of ooxacin. Sato et al. [of the same laboratory as Fukada
(49)] were able to conclude that lomeoxacin penetrated into aqueous
sooner and was eliminated faster than noroxacin. In later experiments,
Ohtori et al. [of the same laboratory as above (50)] examined the ocular
pharmacokinetics of timolol and carteolol in rabbits shortly after recovery
from anesthesia. A 5 mm cellulose membrane (50 kDa) linear probe of fused
silica (0.2 mm o.d., 23 g tubing) was used. In vitro recoveries of 1620% for
noroxacin/lomeoxacin and  1722% for timolol and carteolol were
reported. Pigmented rabbits (1.53.0 kg) were studied. The surgery involved
stitching the nictitating membrane in order to immobilize the eye followed
by the insertion of a 23 gauge needle attached to one end of the probe in the
temporal cornea and passing the needle through anterior chamber and out
of nasal side. The exteriorized tubing was glued at the puncture sites with
epoxy resin. The polyethylene tubing was taped to the face of the rabbit.
Rittenhouse et al. (16) developed an animal model for the evaluation
of microdialysis sampling of aqueous humor to assess the ocular absorption
and disposition of beta-adrenergic antagonist drugs. For this study using
anesthetized dogs (n 3 and rabbits n 3, microdialysis probes (10 mm
CMA/20) were implanted in the anterior chamber. Immediately following
probe implantation ( 30 min), a single dose of 3 HDL-propranolol was

Copyright 2003 Marcel Dekker, Inc.

234

Rittenhouse

administered topically or intracamerally in order to estimate intraocular

bioavailability of [3 HDL-propranolol. [3 HDL-Propranolol collected from
probe euent was assayed by liquid scintillation spectroscopy. The results
of this study indicated a 10-fold higher intraocular exposure to propranolol
in the rabbit relative to the dog (FAH  0:55 vs.  0:056). Time to peak was
longer in the dog relative to the rabbit ( 87 vs.  54 min), and the terminal
rate constant for the dog was  twofold higher than the rabbit  0:0189 vs.
 0:00983). Propranol recoveries of  3245% were reported. The results
obtained in this initial examination of propranolol disposition in aqueous
humor using microdialysis were highly variable. In general, aqueous humor
protein concentrations would have minimal inuence on ocular exposure (3)
due to the low concentrations present. However, since propranolol is a
highly protein-bound substrate (45), Rittenhouse et al. (17) examined the
possibility that time-dependent protein binding might have been a contributing factor to variability in parameter estimates, due to surgical insult
from probe implantation and subsequent increased inux of proteins into
aqueous humor. In addition, anesthesia is a known contributor to alterations in the pharmacokinetics and pharmacodynamics of drugs (60). Thus,
development of relevant experimental techniques for use in conscious animals was imperative.
Following redesign of the microdialysis probes for anterior or posterior chamber placement (4 mm, CMA/20 with 908 bend) (Fig. 1) in the
conscious rabbit, studies were conducted with propranolol (17) to estimate
the intraocular exposure (AUCAH ), time to peak Tmax , and aqueous
humor peak concentrations Cmax following a >5-day recovery. This
minimum recovery period was established by following the time course
of ocular wound healing and anterior segment resorption of brin, a
phenomenon that could result in reduced substrate recovery via microdialysis aqueous humor sampling. Briey, the surgical probe implantation
procedure for New Zealand white rabbits (2.350 kg) proceeded as follows: A limbal-based conjunctival ap was created superior nasally or
temporally  3 mm from the limbus. A 1012 mm conjunctival pocket
was prepared, and the probe inlet/outlets were exteriorized to the top of
head. A 20 gauge needle was inserted  23 mm from limbus into the
anterior chamber and removed. The microdialysis probe was then placed
into the opening and the anchor of probe sutured to the sclera and
covered with conjunctiva. Propranolol ocular pharmacokinetic parameter
estimates obtained from a previous study (16) were compared to those
obtained in the present study (17). It was observed that reduced dosenormalized AUCAH and Cmax were obtained in the previous study relative
to the present study ( 1:9-fold relative to anesthetized results with >5-day
recovery period). It was hypothesized that time-dependent aqueous humor

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

235

protein concentrations may have been present immediately postsurgery,

but that protein concentrations returned to basal levels given a sucient
recovery period. Increased aqueous humor protein concentrations have
been reported in vivo shortly after cannulation of the eye (46). In order
to examine this possible explanation for the apparent decrease in dosenormalized AUCAH observed in the previous study (16), the time course of
aqueous humor protein concentrations after microdialysis probe implantation in the anterior chamber was examined in 16 rabbits (17). Immediately
following probe implantation, aqueous humor protein concentrations were
comparable to control. At 30 minutes postimplantation, aqueous humor
protein concentrations were maximal ( 30 mg/mL) and were maintained
for up to 90 minutes. Aqueous humor protein concentrations were halfmaximal at 150 minutes. A simulation approach was used to examine
the hypothesis that altered protein concentrations were responsible for
dierences in propranolol exposure between the two experiments.
Results of the simulations indicated that time-dependent binding of
propranolol in aqueous humor was probably the major contributor to
the reduced aqueous humor intraocular exposure to propranolol observed
in rabbits with a minimal recovery period postimplantation (2.4-fold for
simulation results vs.  1:9-fold dierence observed in vivo). Another
salient observation in this study was the appreciable dierence between
the ocular pharmacokinetics of propranolol in the conscious versus the
anesthetized rabbit. Dose-normalized AUCAH was eight-fold lower in
conscious rabbits as compared to anesthetized rabbits. Propranolol dosenormalized Cmax values for the conscious rabbits were appreciably lower
than those reported in the literature for conscious animal experimentation
(20,21). A careful examination of this question resulted in the hypothesis
that traditional sampling procedures (euthanasia of conscious rabbits
following topical administration of drug, with paracentesis sampling of
aqueous humor as the last step) may result in artifactually higher intraocular exposures to topically administered xenobiotics. This work
provided a framework for examination of ocular pharmacokinetics in a
more physiologically relevant model.
Although a number of researchers have examined the blood-to-aqueous transport of ascorbate in ciliary body tissue and cell culture in vitro
(6168), the transport kinetics information derived from these studies, in
most instances, does not correlate to in vivo determinations. In vivo investigations have been limited due to diculties inherent in studies of ascorbate
transport kinetics; Km , the blood concentration of ascorbate at half-maximal transport, is reputed to be at or below physiological blood concentrations (58,63). Rittenhouse et al. (53), following the development of an
analytical procedure for assay of ascorbate in blood and aqueous humor,

Copyright 2003 Marcel Dekker, Inc.

236

Rittenhouse

examined the transport kinetics of ascorbate using the recently developed

conscious animal model with microdialysis sampling of aqueous humor.
Microdialysis probes were placed in the anterior chamber of one eye
and the posterior chamber of the fellow eye (53). Basal bloodtoaqueous
transport of 14C-ascorbate was established by the examination of aqueous
humor ascorbate corrected for specic activity. Following a 30-day recovery period, the rabbits n 4) were placed in restraining devices, the marginal ear veins of respective ears were cannulated, and ascorbate was
administered via an i.v. bolus loading dose followed by maintenance incremental infusions in order to characterize the linear-to-nonlinear kinetic
prole in blood to aqueous humor transport. Blood and probe euent
were analyzed via UV spectrophotometry at 265 nm. A nonlinear leastsquares regression analysis assessment of the transport kinetics of ascorbate was performed. Contrary to previous reports (58,63), ascorbate blood
concentrations, which were increased in a stepwise fashion (an overall
twofold increase), did not result in saturable ascorbate uptake into aqueous (blood concentrations from  14 to  21 to  30 mg/L). Nonlinear
least-squares regression analysis of a model that incorporated nonsaturable
uptake into aqueous with rst-order translocation from the posterior to
the anterior chamber and rst-order eux from the anterior chamber, with
an incorporated lag time of  1 hour, appeared to describe the data best.
The model ts to the serum, anterior, and posterior aqueous ascorbate
concentration-time data are presented in Figure 3. Physiologically relevant
parameter estimates were obtained with this approach. The analysis provided indications that reduced aqueous humor turnover occurred in this
group of rabbits (translocation rate constants were  0:005 min1 as compared to 0.01 min1 in intact animals). The parameter estimates were also
in agreement with the model independent ascorbate ocular clearance determinations  39 mL/min or  0:003 min1 , when divided by the estimated
aqueous humor volume of 200 mL) (53). It is possible that the apparent
transport of ascorbate was perturbed by surgery. Surgical trauma can
result in increased peroxide generation as a result of the inammatory
cascade (69). There are no reports of studies evaluating basal ascorbate
transport as a function of the degree of intraocular inammation. It also is
possible that time-dependent changes to ascorbate blood to aqueous transport were observed. In order to examine this possibility, the relationship
between aqueous humor ascorbate concentrations and time postprobe
implantation was examined (Fig. 4) (17,53,54). At 0 minutes, physiologically relevant ascorbate aqueous humor concentrations are observed ( 1:4
mM). An appreciable decrease ( 50% was observed from day 1 to day
12. Hence, recovery periods were lengthened for subsequent experiments in
order to examine ascorbate transport kinetics (>30 days). However, from

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

237

Figure 3 Four-compartment model and associated ascorbate disposition in serum

(A) and aqueous humor (B). Lines represent t of the kinetic model to the data.
(Adapted from Ref. 53.)

the data presented in Figure 5, it appeared that as early as  21 days,

ascorbate aqueous humor concentrations returned to physiologically relevant concentrations ( 1:3 mM). The relationship between basal blood to
aqueous transport and time postsurgery is presented in Figure 5. Basal
transport rates comparable to the expected value ( 27 mg/hr) were
observed for rabbits following a minimum of 21 days of recovery. These
data were examined retrospectively (53,54).
Macha and Mitra (51,52) explored an innovative approach in ocular
microdialysis experimentation. A dual microdialysis probe experimental
design provided the unique opportunity to examine intraocular drug disposition in both vitreous and aqueous humors simultaneously. Using this
approach, the pharmacokinetics of systemically versus intravitreously
administered uorescein (51) was examined in the anesthetized rabbit.
Concentric probes (CMA/20 with 0:5  10 mm, polycarbonate membrane
and 14 mm shaft) were used for vitreous sampling. Linear probes (MD2000, 0:32  10 mm, polyacrylonitrile membrane and 0.22 mm tubing)

Copyright 2003 Marcel Dekker, Inc.

238

Rittenhouse

Figure 4 Relationship between aqueous humor ascorbate concentrations and

recovery period in rabbits with intraocular microdialysis probes placed into the
anterior or posterior chamber of the eye. Area between dashed lines represents
reported steady-state aqueous humor ascorbate concentrations.

were used for aqueous sampling. For the implantation of the probe in the
vitreous, a 22 gauge needle was inserted into the eye about 3 mm below
the corneal-scleral limbus through the pars plana. The needle was then
removed and the probe adjusted such that the membrane was in the mid-

Figure 5 Relationship between basal ascorbate blood aqueous transport versus

recovery period in rabbits with intraocular microdialysis probes in the anterior
and posterior chambers. Dashed line represents theoretical basal transport. (Bars
represent  SD).

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

239

vitreous region as ascertained by microscopic examination. The linear

probe was implanted in the aqueous humor using a 25 gauge needle.
The needle was inserted across the cornea just above the corneal scleral
limbus so that it traversed through the center of the anterior chamber to
the other end of the cornea. The sample-collecting end of the linear probe
was inserted into the bevel edge end of the needle. The needle was then
retracted, leaving the probe with the membrane in the middle of the anterior chamber. The outlets of both probes were then xed. The rabbits were
maintained under general anesthesia throughout the experiment. A permeability index was calculated by taking the ratio of the area under the
aqueous or vitreous uorescein concentration curve relative to plasma
following intravenous administration. This assessment provided experimental documentation of the pathway for uorescein entry into the eye
primarily via the ciliary body; the aqueous permeability index was
vefold higher than vitreous. The plasma, aqueous, and vitreous disposition of uorescein is presented in Figure 6.

Figure 6 Concentration-time proles of plasma, anterior chamber, and vitreous

uorescein after systemic administration (10 mg/kg): (*) plasma concentrations;
(~) aqueous concentration; (^) vitreous concentration. The line drawn represents
the nonlinear least-squares regression t of the model to the concentration-time data.
(Ref. 52.)

Copyright 2003 Marcel Dekker, Inc.

240

B.

Rittenhouse

Ocular Pharmacology and Pharmacodynamic

Experimentation

A second study using the dualprobe approachsimultaneous sampling of

aqueous and vitreous humor using microdialysiswas conducted (52) for
the examination of cephalosporin ocular pharmacokinetics and the pharmacodynamics of inhibitory drugs on the intraocular disposition of cephalosporins. New Zealand albino rabbits (22.5 kg) were kept under
anesthesia throughout the experiment. The concentric and linear microdialysis probes were implanted into the vitreous and aqueous chambers,
respectively, as described above (51). Microdialysate samples were collected every 20 minutes over a period of 10 h. The animals were allowed
to stabilize for 2 hours prior to initiation of each experiment. The ocular
pharmacokinetics of cephalosporins were investigated following intravitreal administration of 500 mg dose of cephalexin, cephazolin, and cephalothin, respectively. In vivo inhibition experiments were conducted by
coadministration of one of two dipeptides, gly-pro or gly-sar, with a 50
mg dose of cephalexin or cefazolin. The dipeptides were administered by a
bolus injection into the vitreous 30 minutes prior to administration of the
drugs, as well as by continuous perfusion through the vitreous probes to
maintain the study state dipeptide concentrations throughout the experiment. The intravitreal elimination half-lives of cephalexin, cefazolin, and
cephalothin after intravitreal administration were found to be 185:38  27
:25 min, 111:40  17:17 min, and 146:68  47:52 min, respectively. Higher
aqueous cephalexin concentrations were observed in comparison to cefazolin concentrations. With respect to the pharmacokinetic parameters of
cephalexin in the presence of gly-pro, increased AUC (3-fold), decreased
clearance ( 3-fold), and increased terminal elimination half-life ( 3:5fold) was observed. The cephalexin intravitreal concentration time course
with or without inhibitor is presented in Figure 7. For cefazolin, no
change in the pharmacokinetic parameters was observed except for an
fourfold increase in terminal elimination half-life in the presence of
gly-pro. Gly-sar had no signicant eect on the pharmacokinetics of either
drug.
An important rst step toward the ultimate endpoint, the assessment
of the pharmacodynamics of beta-adrenergic antagonists, involved characterization of the disposition of the proposed endogenous marker for aqueous humor turnover, ascorbate, discussed in the previous section (53). The
utility of this approach was examined initially in the pioneering work of
Becker (57,70,71,72) for the pharmacodynamics of a systemically administered carbonic anhydrase inhibitor (CAI), acetazolamide. CAIs decrease
IOP via reduction in aqueous humor production (73). Ascorbate concentra-

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

241

Figure 7 Vitreous concentration-time prole of cephalexin (50 mg) in the presence

of inhibitors after intravitreal administration. (Ref. 51.)

tions in aqueous humor were observed to increase following acetazolamide

administration (70). This observation provided the impetus for examining
ascorbate as an endogenous substrate for the assessment of aqueous humor
production inhibition.
Examination of the pharmacodynamics of aqueous humor production
by beta-adrenergic antagonists was performed with a novel approach,
microdialysis sampling of aqueous humor ascorbate. The time course for
ascorbate in aqueous humor was used in order to determine the relative
changes in aqueous humor ow as a result of beta-blocker pharmacodynamics. The volume dilution technique (31,32) involves assessment of the
time course of an exogenously administered dye in aqueous humor following pharmacodynamic modulation of aqueous humor production (Fig. 8A),
as described previously. Due to the intrusiveness of the procedure, studies of
this nature are possible only in anesthetized animals. Rittenhouse et al. (54)
selected an approach that would permit examination in a conscious animal
model and would minimize the number of exogenous substrates required for
this examination. By using an endogenous substrate, ascorbate, a less intrusive method for the examination of in vivo aqueous humor turnover, was
developed (Fig. 8B).

Copyright 2003 Marcel Dekker, Inc.

242

Rittenhouse

Figure 8 Schematic of volume dilution method (A) and modied volume dilution
method: microdialysis probe placement (B).

Following anterior chamber placement of microdialysis probes into

each eye of six rabbits and a >14-day recovery period, each rabbit
received a tracer i.v. bolus of 14C-ascorbate (53). Basal ascorbate blood
to aqueous transport, Ro , and ascorbate ocular clearance, CLAH were
estimated. After a 1-hour washout, each rabbit received a series of three
doses of 3H-propranolol (7503000 mg, 16.5 mCi/mg) every 60 minutes into

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

243

the lower cul-de-sac of each eye. Microdialysis probe euent was analyzed
for ascorbate with a spectrophotometric assay (53,54), and the time course
of aqueous humor ascorbate was determined (Fig. 9) (54). Nonlinear leastsquares regression analysis of the ascorbate time-course in aqueous humor
was performed in order to estimate the altered CLAH following propranolol perturbation of aqueous humor production. The relative dierence in
the basal CLAH and the CLAH determined subsequent to propranolol
administration was calculated. The major assumption for this approach
was that this dierence in CLAH could be attributed solely to inhibition of
aqueous humor production. A  47% reduction in ow was observed

Figure 9 Ascorbate aqueous humor concentrations following three doses of topically administered propranolol: (A) AH ascorbate as % baseline (bars indicate  SE;
n 3) versus time prole for the 1500 mg dose (^); (B) AH ascorbate as % baseline
(bars indicate  SE; n 3 versus time proles for the 750 mg (^) and for the 3000 mg
(&) doses. Arrows represent time for administration of 14C-ascorbate i.v. dose and
the hourly propranolol topical doses. (From Ref. 57.)

Copyright 2003 Marcel Dekker, Inc.

244

Rittenhouse

following topical administration of propranolol (1500 mg). No modulation

in the ascorbate aqueous humor time-course was observed for the 750 and
3000 mg groups. The model t to three individual ascorbate aqueous concentrationtime data is presented in Figure 10. Examination of iris-ciliary
body tissue concentrations of ascorbate and propranolol provided evidence
that a spectrum of intraocular responses to propranolol was observed in
this experiment. It appeared that the 750 mg dose provided aqueous humor

Figure 10 Nonlinear least-squares regression t of aqueous humor ascorbate concentration versus time proles following three 1500 mg doses of topically administered propranolol in three individual rabbits. (From Ref. 54.)

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

245

propranolol concentrations below those that could inhibit aqueous humor

production. Reduction in aqueous humor production was observed for the
1500 mg group, while the simultaneous inhibition of ascorbate ciliary tissue
accumulation and aqueous humor formation resulting in no change in the
ascorbate aqueous humor time course was observed for the 3000 mg group
(54). Anterior segment microdialysis provided an excellent approach to
examine complex ocular physiology and pharmacology.

VI.

FUTURE CHALLENGES, MILESTONES, AND

OPPORTUNITIES

The anterior segment microdialysis technique has provided the framework

for novel approaches to the examination of the mechanisms involved in
ophthalmic drug ocular pharmacokinetics and the pharmacodynamics of
aqueous humor formation and modulation. Anesthetized and conscious
animal models, in some cases involving the long-term placement of microdialysis probes into the anterior segment (i.e., anterior and posterior chambers), were established, which have provided a substantial advance in
experimental approaches to ocular pharmacokinetic/pharmacodynamic
experimentation. Reduction in the number of animals required for these
studies (n 48 reduced to n 3  6) provides a clear market advantage
via reduction in costs and in animal-sparing research approaches. Animal
experimentation with probe placement in the anterior segment, pioneered
by Sato et al. (49), Fukada et al. (48), and Ohtori et al. (50), up to papers
describing conscious animal experimental with microdialysis probes in the
anterior chamber for 5 to > 30 days (17,53,54), demonstrate the rapid
development and increased utility of the microdialysis approach in the
study of ocular drug delivery and disposition. Unique applications of
the microdialysis approach in the examination of drug ocular disposition
such as a dual probe implantation technique involving simultaneous examination of aqueous and vitreous drug disposition have been performed
(51,52). The importance and broad-based applicability of the microdialysis
sampling approach for the examination of ocular pharmacokinetics and
dynamics of ophthalmics is of large impact. Reuse of animals is possible
because of the long-term tolerability of the animals to probe implantation.
A series of substrates could be examined using the same subject to assess
phenomenon such as tolerance development. Possible reduction in intersubject responses to ophthalmic drugs is possible via this approach. This
approach also provided the framework for a detailed examination of in
vivo basal blood to aqueous transport of ascorbate, a labile substrate of
utmost importance to intraocular health and homeostasis. Moreover, the

Copyright 2003 Marcel Dekker, Inc.

246

Rittenhouse

pharmacodynamics of beta-adrenergic antagonistassociated modulation

of both aqueous humor formation and ascorbate ciliary accumulation
was examined in detail using anterior segment microdialysis.
Pharmacokinetic modeling was facilitated with this technique; by using
alterations in the ascorbate aqueous humor time-course as a means of
estimating aqueous humor ow, the modulatory eects of a model substrate, propranolol, on aqueous humor turnover, was examined.

REFERENCES
1. Morrison, J. C., and Freddo, T. F. (1996). Anatomy, microcirculation, and
ultrastructure of the ciliary body. In The Glaucomas, 2nd ed., Ritch, R., Sields,
M. B., Krupin, T., eds. C. V. Mosby Publishers, St. Louis, pp. 125138.
2. Krupin, T., and Civan, M. M. (1996). Physiologic basic of aqueous humor
formation. In The Glaucomas, 2nd ed., Ritch, R., Sields, M. B., Krupin, T.,
eds. C. V. Mosby Publishers, St. Louis, pp. 252253.
3. Cole, D. F. (1974). Comparative aspects of intraocular uids. In The Eye:
Comparative Physiology, H. Dason, L. T. Graham, eds. Academic Press, New
York.
4. Cole, D. F. (1977). Secretion of the aqueuos humor. Exp. Eye Res. 25
(suppl):161.
5. Bill, A. (1973). The role of ciliary body blood ow and ultraltration in
aqueous humor formation. Exp. Eye Res. 16:287.
6. Freddo, T. F. (1989). Ocular anatomy and physiology related to aqueous
production and outow. In The Glaucomas, Ritch, R., Sields, M. B.,
Krupin, T., eds. C. V. Mosby Publishing Co., St. Louis, pp. 2346.
7. Sears, M. L. and Kondo, K. (1986). Drugs eects upon aqueous production.
Trans. Ophthalmol. Soc. U.K. 105:171179.
8. Krupin, T., and Nichols, P. F. (1984). In Glaucoma: Applied Pharmacology in
Medical Treatment. Drance, S. E., and Neufeld, A. H., eds. Grune and
Stratton, New York, pp. 7185.
9. Maurice, D. M., and Mishma, S. (1984). Ocular pharmacokinetics. In
Handbook Experimental Pharmacology, Sears, M. L., ed. Springer-Verlag,
New York, p. 32.
10. Kaufman, P. L. (1996). Pressure-dependent outow. In The Glaucomas, 2nd
ed., Ritch, R., Sields, M. B., Krupin, T., eds. C. V. Mosby Publishing Co., St.
Louis, pp. 307311.
11. Hart, W. M. (1992). Intraocular pressure. In Adlers Physiology of the Eye,
Clinical Application, 9th ed., Hart, W. M. ed., Mosby Year Book, St Louis,
pp. 248267.
12. Johnson, D. H., Richardson, T. M., and Epstein, D. L. (1989). Trabecular
meshwork recovery after phagocytic challenge. Curr. Eye Res. 8:11211130.

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

247

13. Epstein, D. L., Freddo, T. F., Anderson, P. J., Patterson, M. M., and BassettChu, S. (1986). Experimental obstruction to aqueous outow by pigment
particles in living monkeys. Invest. Ophthalmol. Vis. Sci. 27:387395.
14. Crean, E. V., Tyson, S. L., and Richardson, T. M. (1986). Establishment of
calf trabecular meshwork cell cultures. Exp. Eye Res. 43:503517.
15. Brubaker, R. F. (1991). Flow of aqueous humor in humans. Invest.
Ophthalmol. Vis. Sci. 32:31453166.
16. Rittenhouse, K. D., Peier, Jr., R. L. and Pollack, G. M., (1998). Evaluation
of microdialysis for in vivo models of ocular absorption and disposition. J.
Pharm. Biomed. Anal. 16:951959.
17. Rittenhouse, K. D., Peier, Jr., R. L. and Pollack, G. M. (1999). Microdialysis
evaluation of the ocular pharmacokinetics of propranolol in the conscious
rabbit. Pharm. Res. 16:736742.
18. Lee, Y. H., Kompella, U. B., and Lee, V. H. L. (1993). Systemic absorption
pathways of topically applied adrenergic antagonists in the pigmented rabbit.
Exp. Eye Res. 57:341349.
19. Ross, F. E., Innemee, H. C., and van Zwieten, P. A. (1980). Ocular penetration of beta-adrenergic blocking agents. An experimental study with atenolol,
metoprolol, timolol and propranolol. Docum. Ophthalmol. 48:291301.
20. Hussain, A., Hirai, S., and Sieg, J. (1980). Ocular absorption of propranolol in
rabbits. J. Pharm. Sci. 69:738739.
21. Schmitt, C., Lotti, V. J., and Le Douarec, J. C. (1981). Penetration of ve
beta-adrenergic antagonist into the rabbit eye after ocular instillation.
Albrecht Graefes Arch. Klin. Ophthalmol. 217:167174.
22. Ben-Nun, J., Joyce, D. A., Cooper, R. L., and Cringle, I. J. (1989).
Pharmacokinetics of intravitreal injection. Invest. Ophthalmol. Vis. Sci.
30:10551061.
23. Urtii, A. (1993). Animal pharmacokinetics studies. In Ophthalmic Drug
Delivery Systems, Mitra, A. K., ed. Marcel Dekker, New York, pp. 83108.
24. Prince, J. H. (1964). The Rabbit in Eye Research. Charles C Thomas,
Springeld IL.
25. Brubaker, R. F. (1996). Measurement of aqueous ow by uorophotometry.
In The Glaucomas, 2nd ed., Ritch, R., Sields, M. B., Krupin, T., eds. Mosby
Publishing Co., St. Louis, pp. 447453.
26. Jones, R. F., and Maurice, D. M. (1966). New methods of measuring the rate
of aqueous ow in man with uorescein. Exp. Eye Res., 5:208220.
27. Dueker, D. (1996). Tonography. In The Glaucomas, 2nd ed., Ritch, R., Sields,
M. B., and Krupin, T., eds. Mosby Publishing Co., St. Louis, pp. 399401.
28. Peier, R. L. Jr., Gelatt, K. N., and Gum, G. G. (1976). Determination of the
facility of aqueous humor outow in the dog, comparing in vivo and in vitro
tonographic and constant pressure perfusion techniques. Am. J. Vet. Res.
12:14731477.
29. Peier, R. L. Jr., Gum, G. G., Grimson, R. C., and Gelatt, K. N. (1980).
Aqueous humor outow in beagles with inherited glaucoma: constant pressure
perfusion. Am. J. Vet. Res. 41:18081813.

Copyright 2003 Marcel Dekker, Inc.

248

Rittenhouse

30. Johnson, M., Chen, A., Epstein, D. L., and Kamm, R. D. (1991). The pressure
and volume dependence of the rate of wash-out in the bovine eye. Curr. Eye
Res. 10:373375.
31. Miichi, H., and Nagataki, S. (1982). Eects of cholinergic drugs and adrenergic drugs on aqueous humor formation in the rabbit eye. Jpn. J. Ophthalmol.
26:425436.
32. Miichi, H., and Nagataki, S. (1983). Eects of pilocarpine, salbutamol, and
timolol on aqueous humor formation in cynomolgus monkeys. Invest.
Ophthalmol. Vis. Sci. 24:12691275.
33. Green, K., and Padgett, D. (1979). Eect of various drugs on pseudofacility
and aqueous humor formation in the rabbit eye. Exp. Eye Res. 28:239246.
34. Green, K., and Elija, D. (1981). Drug eects on aqueous humor formation
and pseudofacility in normal rabbit eyes. Exp. Eye Res. 33:239245.
35. Liu, H. K., Chiou, G. C., and Garg, L. C. (1980). Ocular hypotensive eects
of timolol in cat eyes. Arch. Ophthalmol. 98:14671469.
36. Oppelt, W. W. (1967). Measurement of aqueous humor formation rates by
posterior-anterior chamber perfusion with inulin: normal values and the eect
of carbonic anhydrase inhibition. Invest. Ophthalmol. 6:7683.
37. Evrard, P. A., Deridder G., and Verbeeck, R. K. (1996). Intravenous microdialysis in the mouse and the rat: Development and pharmacokinetic application of a new probe. Pharm. Res. 13:1217.
38. Deguchi Y., Terasaki, T., Yamada, H., and Tsuji, A. (1992). An application of
microdialysis to drug tissue distribution study: in vivo evidence for free-ligand
hypothesis and tissue binding of Beta-Lactam antibiotics in interstitial uids.
J. Pharmacobiol. Dyn. 15:7989.
39. Hogan, B., and Lunte, S. (1994). On-line coupling of in vivo microdialysis
sampling with CE. Anal. Chem. 66:596602.
40. Tellez, S., Forges, N., and Roussin, A. (1992). Coupling of microdialysis with
CE: a new approach to the study of drug transfer between two compartments
of the body in freely moving rats. J. Chromatography 581:257266.
41. Eisenberg, E. J., and Eicho, W. M. (1993). A method for estimation of
extracellular concentration of compounds by microdialysis using urea as a
endogenous recovery marker in vitro validation. JPM 30:2731.
42. Dubey, R. K., McAllister C. B., Inoue, M., and Wilkinson, G. R. (1989).
Plasma binding and transport of diazepam across the BBB. No evidence for
in vivo enhanced dissociation. J. Clin. Invest. 84:11551159.
43. Sebille, B., Zini, R., Madjar, C. V., Thuaud, N., and Tillement, J. P. (1990).
Review: separation procedures used to reveal and follow drug protein binding.
J. Chromatography 531:5177.
44. Rietbrock, N., and Woodcock, B., eds. (1981). Progress in drug protein binding: proceedings of lectures presented at a symposium in Frankfurt am Main,
April 1980. Methods in Clinical Pharmacology, Vieweg, Braunschweig/
Wiesbaden.
45. Rowland, M., and Tozer, T. N. (1995). Clinical Pharmacokinetics: Concepts
and Applications, 3rd ed. Williams and Wilkins, Media, PA.

Copyright 2003 Marcel Dekker, Inc.

Anterior Segment Microdialysis

249

46. Kulkarni, P. S., and Srinvasan, B. D. (1987). Nonsteroidal anti-inammatory

drugs in ocular inammatory conditions. In Nonsteroidal Anti-inammatory
Drugs, A. J. Lewis, D. E. Furst, eds., Marcel Dekker, New York.
47. Robinson, T. E., and Justice J. B., Jr. eds. (1991). Microdialysis in the
Neurosciences. Elsevier Science Publishers, Amsterdam.
48. Fukuda, M., Mikitani, M., Ueda, T., Inatomi, M., Koide, Y., Kurata, T., et
al. (1995). Application of microdialysis on analysis of pharmacokinetics in
domestic rabbits aqueous humor. J. Jpn. Ophthalmol. Soc. 99:400405.
49. Sato, H., Fukuda, S., Inatomi, M., Koide, R., Uchida, N., Kanda, Y., Kiuchi,
Y., and Oguchi, K. (1996). Pharmacokinetics of noroxacin and lomeoxacin
in domestic rabbit aqueous humor analyzed by microdialysis. J. Jpn.
Ophthalmol. Soc. 100:513519.
50. Ohtori, R., Sato, H., Fukuda, S., Ueda, T., Koide, R., Kanda, Y., Kiuchi, Y.,
and Oguchi, K. (1988). Pharmacokinetics of topical beta-adrenergic antagonists in rabbit aqueuos humor evaluated with the microdialysis method. Exp.
Eye Res. 66:487494.
51. Macha, S., Mitra, A. K. (2001). Ocular pharmacokinetics of cephalosporin
using microdialysis. J. Ocul. Pharmacol Ther. 17:485498.
52. Macha S., Mitra A. K. (2001). Ocular pharmacokinetics in rabbits using a
novel dual probe microdialysis technique. Exp. Eye Res., 72:289299.
53. Rittenhouse, K. D., Peier, R. L., Jr. and Pollack, G. M. (2000). Assessment
of ascorbate ocular disposition in the conscious rabbit: an approach using the
microdialysis technique. Curr. Eye Res. 20:351360.
54. Rittenhouse, K. D., and Pollack, G. M. (2000). Pharmacodynamics of betablocker modulation of aqueous humor. Exp. Eye Res. 70:429439.
55. Stahle, L., Segersvard, S., and Ungerstedt, U. (1991). A comparison between
three methods for estimation of extracellular concentrations of exogenous and
endogenous compounds by microdialysis. J. Pharmacol. Methods 25:4152.
56. Scheller, D., and Kolb, J. (1991). The internal reference technique in microdialysis: a practical approach to monitoring dialysis eciency and to calculating tissue concentration from dialysate samples. J. Neurosci. Methods 40:31
38.
57. Becker, B. (1955). The eects of the carbonic anhydrase inhibitor, acetazolamide, on the composition of aqueous humor. Am. J. Ophthalmol. 40:129136.
58. Kinsey, V. E. (1947). Transfer of ascorbic acid and related compounds across
the blood-aqueous barrier. Am. J. Ophthalmol. 30:12621266.
59. Kinsey, V. E. and Palmer, E. (1955). Posterior and anterior chamber aqueous
humor formation. AMA Arch. Ophthalmol. 53:330.
60. Ausinsch, B., Rayborn, R. L., Munsen, E. S., and Levy, N. S. (1976)
Ketamine and IOP in children. Anesthesia Analgesia 55:773775.
61. Delamere, N. A., Coca-Prados, M., and Shelinder, A. (1993). Studies on
regulation of the ascorbic acid transporter in a cell line derived from rabbit
non-pigmented ciliary epithelium. Biochem. Biophys. Acta 1149:102108.
62. Candia, O. A., Shi, X. P., and Chu, T. C. (1991). Ascorbate-stimulated active
Na+ transport in rabbit ciliary epithelium. Curr. Eye Res. 10:197203.

Copyright 2003 Marcel Dekker, Inc.

250

Rittenhouse

63. Becker, B. (1967). Ascorbate transport in guinea pig eyes. Invest. Ophthalmol.
51:410415.
64. Socci, R. R., and Delamere, N. A. (1988). Characteristics of ascorbate transport in the rabbit iris-ciliary body. Exp. Eye Res. 46:853861.
65. Chu, T., and Candia, O. A. (1988). Active transport of ascorbate across the
isolated rabbit ciliary epithelium. Invest. Ophthalmol Vis. Sci. 29:594599.
66. Helbig, H., Korbmacher, C., Wohlfarth, J., Berweck, S., Kuhner, D., and
Wiederholt, M. (1989). Electrogenic Na+-ascorbate cotransport in cultured
bovine pigmented ciliary epithelial cells. Am. J. Physiol. 256:C44C49.
67. Mead, A., Sears, J. and Sears, M. (1996). Transepithelial transport of ascorbic
acid by the isolated intact ciliary epithelial bilayer of the rabbit eye. J. Ocular
Pharmacol. Ther. 12:253258.
68. Hou, Y., Pierce, W. M., Jr., and Delamere, N. A. (1998). The inuence of
ascorbic acid on active sodium transport in cultured rabbit nonpigmented
ciliary epithelium. Invest. Ophthalmol. Vis. Sci. 39:143149.
69. Slater, T. F. (1972). Free Radical Mechanisms in Tissue Injury. Pion
Publishing, London.
70. Becker, B. (1956). The eects of acetazolamide on ascorbic acid turnover. Am.
J. Ophthalmol. 41:522529.
71. Becker, B. (1957). Chemical composition of human aqueous humor. AMA
Arch. Ophthalmol. 57:793799.
72. Becker, B. (1959). Carbonic anhydrase and the formation of aqueous humor.
Am. J. Ophthalmol. 47:342359.
73. Hurvitz, M., Kaufman, P. L., Robin, A. L., Weinreb, R. N., Crawford, K.,
and Balke, S. (1991). New developments in the drug treatment of glaucoma.
Drugs 41(4):514532.

Copyright 2003 Marcel Dekker, Inc.

8
Posterior Segment Microdialysis
Sreeraj Macha
and Ashim K. Mitra
University of MissouriKansas City, Kansas City, Missouri, U.S.A.

I.

INTRODUCTION

The eye is a specialized sensory organ that is relatively secluded from

systemic access by the blood-retinal, blood-aqueous, and blood-vitreous
barriers (1). Ocular infections can involve the lids, orbits and ocular
adnexia, anterior segment (cornea and conjunctiva), or the eye interior,
i.e., retina-choroid. While often benign and self-limiting, certain infections
can involve delicate structures of the eye and destroy visual function (2).
Postoperative endophthalmitis, which is bacterial in origin, is one of the
most important complications following intraocular surgery. It is a sightthreatening condition and occurs with an incidence of 0.050.33% (1,3).
Unfortunately, the outcome of clinical diagnosed postoperative endophthalmitis is far from satisfactory. More than 50% of all such cases suer irreparable visual loss, despite presently available antimicrobial agents (4).
Human cytomegalovirus retinitis is another most common cause of visual
loss among patients suering from acquired immunodeciency syndrome
(AIDS), occurring in about 3040% of patients (58).
Management of these ocular diseases demands immediate diagnosis
and treatment to maximize recovery of vision. Because of blood-ocular
barriers, therapeutic intravitreal concentrations of drugs cannot be readily
achieved by periocular or parenteral administration (911). Although standard therapy consists of both intravitreal and parenteral delivery, recent
clinical evidence suggests that intravitreal administration alone is as eective
in restoring vision as the combined therapy (3). Experimental pharmacoki

Current aliation: Boehringer Ingelheim Inc., Ridgeeld, Connecticut, U.S.A.

251

Copyright 2003 Marcel Dekker, Inc.

252

Macha and Mitra

netic studies have repeatedly conrmed the need to administer drugs intravitreally because systemic or subconjunctival therapy does not deliver adequate concentrations of the drug into the vitreous body (9).
The major constraint on the development and assessment of posterior
segment pharmacokinetics of drugs is the inaccessibility of the vitreous for
continuous serial sampling. Imprecise or nonexistent pharmacokinetic information on vitreous drug disposition has prevented the delineation of the
fundamental mechanisms of vitreal clearance and retinal uptake of drugs
(12). More often, ocular pharmacokinetics are determined by obtaining
single sample of the ocular uids from individual animals (1315) or by
direct serial sampling (16,17).
Microdialysis has gained wide recognition as a valuable tool for sampling the extracellular space of the living tissue. It has become a standard
technique in the eld of neurochemistry, as well as in sampling other tissues
and uids including liver, kidney, skin, tumor, bile, and blood (1821).
Recently, microdialysis has been employed for sampling the vitreous
(10,2227) and aqueous humors (28) in rabbits. The present chapter deals
with the methodological aspects to the use of this technique to study ocular
pharmacokinetics. This chapter also presents a brief overview of the vitreal
microdialysis studies reported recently.

II.

ANATOMY OF THE POSTERIOR SEGMENT

The posterior segment of the eye is comprised of lens, sclera, choroid, vitreous, and retina. Vitreous is a clear gel composed almost entirely of water
(99%) and collagen brils. Vitreous humor occupies a volume of 4 mL and
comprises about 80% of the internal volume of the human eye. Its pH is
about 7.5, and the water turnover rate in vitreous is reported to be about
1015 minutes. It is separated from the lens and posterior chamber by the
anterior vitreous membrane (anterior hyaloid membrane), which is
anchored by ligaments to the retina. Vitreous chamber has three important
anatomical adhesions: (a) ligamentum hyaloideo-capsulare, a pseudomembrane consisting of collagen tissue, (b) circular attachment around the optic
disc, and (c) the vitreous base, which is annular and lies against posterior
surface of the ciliary body and pars plana. Vitreous in humans can be
divided into two zones: the cortical zone and the medullary zone. Cortical
vitreous is closest to the retina and is gel-like, mainly composed of collagen
brils and hyaluronic acid. Medullary zone mainly consists of interstitial
substance and tissue. Special cells called hyalocytes are dispersed in the
outer parts of the cortical vitreous, particularly in the vicinity of pars plana.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

253

Vitreous shows a brillar structure, which can be dierentiated into

two types: Coarse branching bers originate at the base of the vitreous and
disappear behind the equator after having divided into two or three brils.
They are arranged in the cortex of the vitreous in two or more planes and
are approximately 30 mm long. Fine parallel bers are brittle with a diameter
of 3 mm and are more closely joined. They divide behind the equator.
Retina, a transparent tissue, is the innermost coating of the eye that lines
the posterior two thirds of the eyeball. It is rmly attached to the ciliary body
at its anterior termination (ora serrata) and at the margins of the optic nerve.
It consists of two major functional parts: the neural layer and the pigment
epithelium. Pigment epithelium cells are connected by tight junctions and
form a tight barrier between the choroid and retina. It selectively transfers
nutrients to the retina from the choroid. There also exists a uid pump
mechanism in the direction of the choroid through the pigment epithelium.
Choroid is a highly vascularized tissue present between the retina and
sclera. It forms the posterior part of the uvea, the anterior part consisting of
the iris and ciliary body. Choroid mainly consists of three parts: (a) a vessel
layer, consisting of veins and arteries; (b) a choriocapillary layer, consisting
of a very ne and dense network of fenestrated capillaries, which is very
permeable to plasma proteins and colloids; and (c) Bruchs membrane,
which is in direct contact with the retinal pigment epithelium.
Sclera is the outermost layer of the eye, which is mainly protective in
function. It is about 0.51.0 mm thick, consisting mainly of collagen bundles
and elastic bers.

III.

OCULAR BARRIERS TO VARIOUS ROUTES OF

ADMINISTRATION

Drug delivery to the posterior segment of the eye is of high clinical signicance in treating various ocular disorders, i.e., endophthalmitis, viral retinitis, proliferative vitreoretinopathy, uveitis, etc. The problem of
administering potent drugs to the posterior segment of the eye, which is a
relatively closed and well-dened compartment, may be approached in different ways. The most common and patient-acceptable route is the topical
instillation of drugs. Many studies have shown that approximately 5% or
less of the topically applied dose is absorbed across the cornea, which forms
the major barrier to drug penetration to anterior segment tissues. Amounts
of drug absorbed into the posterior segment of the eye will only be a minute
fraction of the amounts achieved in the anterior segment. The main constraints aorded by the eye upon topical delivery are the protective mechanisms, which include solution drainage, lacrimation, diversion of exogenous

Copyright 2003 Marcel Dekker, Inc.

254

Macha and Mitra

chemicals into the systemic circulation via conjunctiva, and a highly selective corneal barrier to exclude these compounds from the eye (29). In addition, there is a nite limit to the size of the dose that can be applied and
tolerated by the cul-de-sac (usually 710 mL) and the contact time of the
drug with the absorptive surfaces of the eye. The drugs usually disappear in
about 510 minutes in rabbits and 12 minutes in humans following topical
instillation (30,31) and at an even faster rate at the pH where most of the
ophthalmic drugs are formulated (32).
Intraocular drug concentrations achieved after systemic administration
depend primarily on the ocular blood circulation. With most drugs, very low
intraocular concentrations were achieved after systemic administration, due
mainly to the presence of blood-aqueous barrier, which restricts substances
from entering into the aqueous humor, and blood-retinal barrier, which
prevents the drugs entering into the vitreous chamber (3337). Higher levels
are generally found in the aqueous humor compared to vitreous, as the
blood-aqueous barrier is known to be leakier than the blood-retinal barrier.
Subconjunctival administration of drugs can also generate elevated
intraocular concentrations compared to topical and systemic administration, with minimal systemic adverse eects. The aqueous and vitreous concentrations of various drugs have been determined after subconjunctival
administration (3841). Even in this case the intraocular concentrations
achieved were found to be only a fraction of the dose administered, as
conjunctival epithelium constitutes a relatively tight barrier. The sclera
does not oer a very tight barrier to the penetration of even relatively
large molecular weight drugs (42).
The only possible way to achieve therapeutic concentrations in the
posterior segment of the eye was found to be the intravitreal administration
(4351). This route of administration has become the recommended therapy
for the treatment of endophthalmitis and cytomegalovirus retinitis. Drugs
injected into the vitreous may be eliminated by two routes. Drugs eliminated
by diusion into the posterior chamber with subsequent removal by normal
egress of uid from the anterior chamber (52,53) generally exhibit half-lives
within a range of 2030 hours (53). The second route is through the retina
via penetration of the blood-retinal barrier (52,53). Drugs eliminated by this
route usually exhibit half-lives in the range of 510 hour (53).

IV.

CONVENTIONAL TECHNIQUES TO STUDY OCULAR

PHARMACOKINETICS

Delineation of ocular pharmacokinetics is complicated due to the complex

nature of the retina and the presence of blood-ocular barriers. In addition,

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

255

the small volume of the intraocular uids aggravates the problem. Limited
volumes of the aqueous and vitreous humors does not allow multiple and
continuous sampling. Most of the studies were carried out with an objective
to decide a dosage regimen that can provide therapeutic concentrations of
drugs in the eye. These studies were conducted in human subjects by collecting vitreous and/or aqueous humor samples prior to surgery (such as cataract, paracentesis, vitrectomy, etc.) after administration of the drug by the
route of interest (48). This method allows for collection of only one sample
per subject at a specic time period. The data obtained can be interpreted
only to determine whether the therapeutic concentrations of the drug have
been achieved with the dose administered.
Ocular pharmacokinetic studies have been carried mainly in rabbits,
the ocular physiology of which is most similar to the human eyes. Initially,
the studies were carried out by collecting single samples from each animal at
a specied time point (13,14). The pooling of the data from individual
rabbits requires at least 100 animals for each prole (12,13). Moreover, it
introduces a signicant amount of intersubject variability.
Serial sampling technique has been developed to obtain more reliable
data with a signicantly reduced number of animals, compared to the single
sample approach, needed for the estimation of pharmacokinetic parameters
(1517). Miller et al. (15) have developed and validated on animal model by
measuring the protein concentrations of the intraocular uids for serial
sampling of the aqueous and/or vitreous humors in New Zealand albino
rabbits. Animals were anesthetized and vitreous samples of about 20 mL
were collected using a 28 gauge needle inserted into the vitreous chamber,
4 mm below the limbus. For serial sampling of the aqueous humor a 30
gauge needle fused to a calibrated 25 mL capillary tube was gently inserted
into the anterior chamber near the limbus, and approximately 7 mL of
aqueous humor was withdrawn.

V.

MICRODIALYSIS

The in vivo microdialysis technique has gained signicant importance in the

eld of the neurochemistry and neurophysiology. This technique has been
successfully adapted to sample various tissues and uids such as skin, liver,
kidney, blood, etc. Recently, microdialysis has found important applications
in the elds of pharmacokinetics, (especially in the area of drug distribution
and metabolism) and pharmacodynamics. It has also been used to study the
in vitro protein and melanin binding of drugs (54).
Microdialysis is based on the principle of diusion. It involves perfusion of a probe, containing a semipermeable membrane, implanted in a

Copyright 2003 Marcel Dekker, Inc.

256

Macha and Mitra

tissue under nonequilibrium conditions. The driving force for the diusion
of drugs across the semipermeable membrane is the concentration gradient.
Endogenous compounds (harmones, neurotransmitters, etc.) and exogenous
compounds (drugs and metabolites) diuse into the probe, whereas compounds added to the perfusate diuse out into the tissue. Therefore, the
technique can be used not only to monitor the extracellular concentrations
of the analyte, but also to deliver drugs to a specic tissue region (55,56).
Microdialysis oers a number of advantages: (a) it permits continuous
monitoring of the tissue concentration of a drug with limited interference
with the normal physiology; (b) no uid is introduced nor is any withdrawn
from the tissue, which is particularly important in sampling tissues/organs
with limited volume; (c) concentration versus time proles can be obtained
from individual animals; (d) the method provides protein free samples, thus
eliminating clean up procedures and ex vivo enzymatic degradation; (e) the
samples can be analyzed by any analytical technique, which contributes to
the selectivity and sensitivity of the method. Its disadvantages include a need
to determine the recovery of the probe (which is still controversial), the
diluting eect of dialysis, which requires sensitive analytical methods to
measure small concentrations, and the invasive nature of the probe implantation.
A.

Probe Design and Selection

A variety of probes have been employed to study posterior segment pharmacokinetics. Probes are selected mainly on the basis of the drug under
investigation, surgical accessibility, type, and length. A vertical probe with
a concentric design, either as such or with modication, is most commonly
used, both in xed and repeated models, by reinsertion via a guide cannula.
The molecular mass cut-o range of the commercially available probes is 5
50 kDa, and selection of a particular probe is based on the molecular weight
of the drug and/or metabolites being studied. The probe membrane type is
one more important factor to be considered in microdialysis. Tao and
Hjorth (57) demonstrated the dierence in the recoveries of three dierent
probe types: GF (regenerated cellulose cuprophan), CMA (polycarbonate
ether), and HOSPAL (polyacrilonitril/sodium methallylsulfonate copolymer). GF and CMA probes exhibited maximal recovery immediately after
the introduction of 5-HT in the articial CSF, whereas it took about 2 hours
for the HOSPAL probe. Landolt et al. (58) have shown that the CMA probe
recovery of cysteine and glutathione varied with concentration.
Ben-Nun et al. (10) used a sampling catheter for simultaneous sampling of the vitreous of both eyes for short-term experiments. Waga et al.
(25,59,60) designed a new probe consisting of soft protecting tube (outer

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

257

diameter 0.6 mm) with a long opening toward one side and a dialysis membrane mounted inside for long-term implantation in the vitreous chamber.
The dialysis membrane consisted of a tube of polycarbonate-polyether copolymer, with an outer diameter of 520 mm and an inner diameter of 400 mm. In
the later experiments the commercially available vertical probe (CMA 20),
with a sti plastic shaft, was used. The shaft length was set to 9 mm and was
bent 60908. Probes with a molecular weight cut-o of 20 and 100 kDA were
selected for small and large molecules, respectively. Stempels et al. (23) used
CMA probe with a shaft diameter of 0.6 mm, length 3 mm, semipermeable
membrane diameter of 0.52 mm, and a cut-o value of 20 kDa.
Probe recovery is directly proportional to the dialysis membrane surface area (6163). By increasing the area, low drug concentrations can be
detected with reasonably high perfusion ow rates while maintaining an
adequate time resolution. To obtain optimal recovery, Macha and Mitra
(26,27) selected a commercially available CMA probe with a membrane
length of 10 mm, shaft 14 mm, and a cut-o value of 20 kDa.
Recovery is shown to be independent of the extracellular analyte concentration (6163). A concentration gradient across the dialysis membrane
changes in unison with the extracellular analyte concentration, thus maintaining a constant recovery.
B.

Composition and Temperature of the Perfusate

Solution

Intraocular uid homeostasis is maintained by the highly perfused retina

and iris-ciliary body. A perfusion uid isosmotic with the plasma is preferred, as it has direct access to the vitreous humor. Previous studies have
revealed that perfusion with anisosmotic uid changes the dialysate concentration of taurine in brain (6466) and muscle (64) microdialysis. In addition, problems arise when a compound already present in the perfusion
medium is also measured using dialysis.
In vitro recovery is dependent on the temperature of the standard
solution. Wages et al. (67) have shown that the in vitro recovery of 3,4dihydroxyphenylacetic acid increased by approximately 30% when the solution temperature was raised from 23 to 378C. Therefore, the in vitro probe
calibration is always carried out at a constant temperature usually the physiological temperature.
C.

Perfusion Flow Rate

Recovery has been shown to improve with a decrease in the perfusion ow

rate. It is fairly high at the beginning but rapidly decelerates after 3060

Copyright 2003 Marcel Dekker, Inc.

258

Macha and Mitra

minutes of perfusion. The high dialysate concentrations immediately after

probe implantation may be due to the traumatic tissue response and also
due to the steep concentration gradient across the dialysis membrane when
the probe is rst inserted into the medium. Although this may play an
important role, it is usually neglected as the similar time-dependent decrease
in recovery is observed in aqueous solutions (68,69).
D.

Recovery

Dialysate concentration of an analyte of interest is only a measure of its

concentration in the extracellular space. The ratio between the concentration of a substance in the outow solution following microdialysis of a tissue
or a biological uid and the undisturbed concentration of the same substance in the solution outside the probe is dened as recovery, expressed
either as a ratio or as a percentage (70).
Recovery factor of the probes is an important parameter in determining the extracellular concentrations of the drug. In vitro recovery was found
to be not only simple and time consuming, but also very appropriate for
ocular microdialysis. In case of in vitro recovery technique, the recovery of a
drug is usually determined by placing the probe in a standard solution. The
probe is continuously perfused at a constant ow rate with saline containing
no analyte. Samples are collected during xed time intervals. The recovery
of the substance of interest is calculated as follows:
Recoveryin vitro

Cout
Ci

where Cout is the concentration in the outow solution and Ci is the concentration in the medium.
The dialysate substances concentrations are transformed into tissue
concentrations as follows:
Ci

Cout
Recoveryin vitro

where C i is the substance concentration in the tissue and C out is the concentration of the dialysate.
Several other techniques have been used to assess probe recovery in
vivo, especially for tissue microdialysis. Jacobsen et al. (71) calculated the
extracellular concentrations by varying the perfusion ow during an in vivo
experiment, measuring the change in the analyte concentration exiting the
probe, and then extrapolating to zero ow rate. This method is called as
ow-rate or stop-ow method. Lonnroth et al. (72) developed a method
where in vivo recovery is estimated by perfusing the probes with varying

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

259

concentrations of the test analyte and then calculating the equilibrium concentration, i.e., the concentration at which the analyte in the perfusate does
not change during the perfusion because it has the same concentration inside
the probe as in the extracellular uid. This is called as the concentration
dierence method or zero-net ux method. The two in vivo methods require
that the drug concentration in the tissue remains constant during the experiment. Probe recovery has also been calculated using a reference substance in
the perfusate (73). The method is based on the fact that recovery across the
membrane is same in both directions. The percentage loss of the reference
substance from the perfusate is used to calculate the concentration of the
compound of interest in the tissues.
Although the recovery of a drug as determined in saline solution was
used to calculate the drug concentrations in the extracellular space in ocular
microdialysis, several studies have been carried out to determine the eect of
extracellular milieu on probe recovery. In the case of brain microdialysis, in
order to account for factors that may aect mass transport from brain ECF
to membrane, in vivo recovery techniques have gained more popularity. In
addition, complex solutions like agar gel or red blood cell media have been
used to simulate the brain ECF conditions (74). However, these systems
have limitations, since it is unlikely that a relatively simple solution like
agar gel accurately reects the complexity of the in vivo physiology.
Vitreal microdialysis appears to be less complicated in terms of assessment
of the actual vitreal drug concentration compared to other tissues/organs.
Vitreous humor consists of almost 99% water. As a result, diusion of drugs
in the vitreous humor has been shown to be similar to that in water. As the
microdialysis probe is surrounded by the vitreous humor without any direct
contact with the tissue, in vitro recovery appears to be a good approximation of in vivo recovery.
The readers are referred to a review article by de Lange et al. (56) for a
detailed description of microdialysis recovery methods.

E.

Surgical Trauma and Blood-Retinal Barrier Integrity

Probe-induced inammation at the site of implantation and subsequent

healing is of major concern in in vivo microdialysis. Such physiological
changes may aect the intraocular pharmacokinetics signicantly. The
inammatory response of the eye mainly depends on the precision of the
surgical procedure; therefore, proper precautions should be taken during
probe implantation. The time interval between the surgery and the onset
of an experiment must be carefully determined, allowing the animal to
completely recover.

Copyright 2003 Marcel Dekker, Inc.

260

Macha and Mitra

Stempels et al. (23) reported that the scleral ports (internal diameter of
0.6 mm), implanted 23 mm from the limbus, were well tolerated during the
observation period. A transient are or minimal cell count was observed
during the rst few days following implantation at or near the entry port,
but it was not considered to be due to intolerance. Endophthalmitis, the
most common inammatory response of the eye, and uveitis were not
observed for up to 6 months following probe implantation.
Endophthalmitis was detected in 4 of the 23 insertions (17%) in which
probes were reused without sterilization; uveitis was not observed when
dialysis was conducted with new probes or probes treated with 25% ethanol.
According to Waga et al. (59) the probes were well tolerated for up to
30 days. Topical antibiotics eectively controlled the purulent discharge
observed in few cases. Clinical observations and histopathological analysis
demonstrated that the probes were well tolerated in majority of the cases.
The inserted probe did not elicit any vitreous reactions and the retina in the
posterior fundus remained normal. In a few cases when the probe touched
the lens due to improper implantation, cataract formation was noticed.
Macha and Mitra (26) selected intraocular pressure (IOP) to determine
the eect of microdialysis probe implantation in the anterior and vitreous
chambers. The baseline IOP prior to any probe implantation was 10:6  1:9
mmHg. A sharp fall in the IOP was observed immediately after the implantation of the dialysis probes. IOP reverted (10:9  1:4 mmHg) to the basal
level within 2 hours after the implantation and remained constant throughout the duration of an experiment. The steady IOP after 2 hours following
probe implantation suggests that there was no long-term eect of probe
implantation on the aqueous humor dynamics.
Blood-aqueous and blood-retinal barriers restrict the passage of serum
proteins into the aqueous and vitreous humors. Elevated protein levels and
high enzymic activities in the ocular uids indicate either a breakdown of the
respective barrier or a leakage from the injured ocular tissue. Paracentesis
(75) and vitrectomy (76) cause breakdown of the blood-ocular barriers,
thereby producing elevated protein levels in the intraocular uids.
Integrity of the blood ocular barriers must be maintained following probe
implantation, and this issue has been addressed in detail by Macha and
Mitra (26). A change in the total protein concentration in the aqueous
and vitreous humors was measured. Vitreal protein concentrations measured at 2 (0:5619  0:3085 mg/mL) and 12 hours (0:2696  0:0897 mg/
mL) after the probe implantation was not signicantly dierent from the
basal concentration (0:3045  0:1712 mg/mL at 2 hours and 0:2087  0:1050
mg/mL). Although the aqueous humor total protein concentration was
higher at 2 hours (1:6971  0:3766 mg/mL) compared to the control
(0:3895  0:1183 mg/mL), the basal level was reached during the course of

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

261

the experiment (0:7986  0:3460 mg/mL in the probe implanted and 0:5876
0:2336 mg/mL in the control eye at 12 hr). A transient rise observed in case
of aqueous humor protein level was assumed to be mainly due to the trauma
caused during probe implantation.
The blood-ocular barriers maintain the homeostasis of the intraocular environment by restricting the movement of compounds from the systemic circulation to the retinal tissue and vitreous cavity. Several reports
discussed the measurement of blood-ocular barrier integrity with the aid of
posterior vitreous uorophotometry (PVF) using uorescein. Penetration
of the dye depends on its concentration in the blood as well as its permeability across the blood-ocular barriers. Macha and Mitra (26) evaluated
the blood-ocular barrier integrity by studying the uorescein kinetics after
probe implantation. The rate constant for uorescein penetration into the
anterior chamber was found to be signicantly higher than into the vitreous, indicating that tighter barrier surrounds the vitreous compartment
compared to the anterior chamber. Integrity of the blood-retinal and
blood-aqueous barriers was ascertained by determining the permeability
index (PI). PI of the anterior (9.48%) and the vitreous chamber (1.99%)
determined using ocular microdialysis was found to be similar to the
values reported using PVF (77).

VI.

INTRAOCULAR PHARMACOKINETICS USING

MICRODIALYSIS

Gunnarson et al. (22) rst utilized in vivo dialysis technique to sample the
vitreous chamber. Studies were carried out to measure endogenous amino
acids in the preretinal vitreous space. The eects of high potassium and
nipecotic acid, a potent gamma-aminobutyric acid (GABA) inhibitor, on
amino acid concentrations were measured. A dialysis probe was implanted
in the vitreous of the eye of albino rabbits (Fig. 1). The integrity of the
blood-retinal barrier was demonstrated by measuring the concentrations of
3
HOH and [14C] j mannitol in the vitreous euent following intracarotid
injections. 3HOH was detected in the vitreous within a few minutes, whereas
[14C]mannitol was mostly excluded. Among the amino acids, glutamine had
a concentration similar to that in the plasma and cerebrospinal uid (CSF).
Vitreous concentration of all amino acids was lower than in plasma, the
majority being below 50% of the plasma concentrations. The taurine level
was approximately 70% that of plasma. A comparison with CSF shows that
all amino acids except for glutamine and phosphoethanolamine (PEA) are
present at higher concentrations in vitreous. Taurine was signicantly elevated (fourfold) in the vitreous, as are valine and alanine. Perfusion with 125

Copyright 2003 Marcel Dekker, Inc.

262

Macha and Mitra

Figure 1 (Top) The positioner device for the dialysis probe on the rabbit eye. The
Perspex cylinder (a) is sewn to the sclera. The probe (b) is placed in the guide channel
(c). The prismatic lens (d) used for biomicroscopy is in contact with the cornea.
(Bottom) The biomicroscopic view of the fundus and the inserted dialysis loop (e).
The iris (f) and the retina (g) can also be observed through the lens.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

263

mM KCl-containing media for 30 minutes, 4 hours after probe implantation

raised the taurine content by sixfold and PEA content by twofold. Other
amino acids remained unchanged (Fig. 2). Perfusion with 60 mM nipecotic
acid increased GABA concentration by 60 times and taurine levels by
almost 10 times, while other amino acids remained fairly constant (Fig. 3).
Ben-Nun et al. (10) evaluated the intraocular pharmacokinetics of
gentamicin after intravitreal administration. The experiments were carried
out for a short duration in domestic cats weighing 2.55 kg. The animals
were anesthetized and the pupils were dilated with tropicamide 0.5% and
phenylephrine 10%. Lateral canthotomy was performed in both eyes and
the area of the upper part of the sclera was exposed. The superior rectus
muscle was divided and cotton wool was inserted into the gap between the
posterior sclera and the superior margin of the orbit to stop bleeding. The
cotton wool was xed with a drop of cyanoacrylate glue. A rubber disk (5
mm in diameter and 1 mm thick) was glued to the sclera over the pars plana
region in the superotemporal quadrant of each eye. A 1 mm diameter hole
was made through the rubber disks and a 20 gauge needle was then passed
through each hole into the eye. A sampling catheter for ocular dialysis was

Figure 2 Change of amino acid concentration with time. (*) Amino acid level on
perfusion with Krebs-Ringer buer. (*) Level on perfusion with high potassium.

Copyright 2003 Marcel Dekker, Inc.

264

Macha and Mitra

Figure 3 Change of concentration of taurine, phosphoethanolamine, and GABA

in response to 60 mM nipecotic acid. The change from Krebs-Ringer buer to
nipecotic acid perfusion media is indicated by an arrow on the time axis.

passed into the sclerotomy site in each eye and glued by a drop of cyanoacrylate glue, followed by a drop of rapid setting epoxy adhesive. The catheters were connected in parallel to a double barrel Harvard pump and
perfused at a rate of 3.5 mL=min (Fig. 4). Gentamicin was administered at
a site away from the sampling site. The experiments were carried in two
groups of cats: normal and bacterial endophthalmitis (Staphylococcus

Figure 4 Diagram representing the bilateral simultaneous sampling of vitreous

humor by ocular dialysis.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

265

Figure 5 Plots of vitreal gentamicin concentrations tted with the pharmacokinetic

model sampled over 8 hours from the time of injection. The top curve represents the
gentamicin concentrations in control eye and the bottom curve the concentrations in
the eye with endophthalmitis.

aureus)induced eyes. Perfusate was collected over 30-minute periods for 3

hours and then hourly to 8 hours. Concentration-time data tted into a onecompartment model that incorporated the diusion of drug within the vitreous and its elimination from the vitreous (Fig. 5). The elimination rate
constants were greater in infected eyes (0.107 hr1 ) than in controls (0.055
hr1 ), which might be due to increased permeability of the blood-retinal
barrier. Aqueous humor gentamicin concentrations in control eyes were
three to six times those in the infected eyes at the end of the experiment.
Waga et al. (59) developed the ocular microdialysis technique for longterm pharmacokinetic studies in rabbits (Fig. 6). A probe (CMA 20) with a

Figure 6 Diagrammatic representation of the microdialysis probe in the rabbit eye.

Copyright 2003 Marcel Dekker, Inc.

266

Macha and Mitra

membrane length of 4 mm and the shaft bent at 60908 was used. Adult
pigmented rabbits were anesthetized with Hypnorm vet1, and a small opening was made in the sclera by conjunctival dissection, about one quarter of
the circumference around the limbus. The beginning was at the nasal end of
the superior rectus muscle, and the end was at the temporal side, before the
lateral rectus. Sling sutures at the superior rectus and a U-shaped suture was
made intrasclerally temporal to the rectus superior muscle. The tip of the U
was pulled out and a loop was formed. At the loop the sclera was punctured
with a 0.9 mm cannula, the probe was inserted, and the sutures were xed.
The tubes of the probe were led under the skin out between the ears.
Ceftazidime was injected intramuscularly (1 mg/kg) (Fig. 7) or intravitreally
(1 mg) (Fig. 8) in two groups: normal and the inammation-induced eyes.
The penetration of ceftazidime into the vitreous was higher (42%) in
inamed than in normal eyes (20%), suggesting an interference with the
blood-retinal barrier. The vitreal half-life of ceftazidime after intravitreal
administration was 8.1 hours and 11.7 hours in normal and inamed eyes,
respectively.
Microdialysis was also used to administer drugs into the vitreous
chamber. Waga and Ehinger (78) investigated the ability of 125I-labeled
NGF to cross a previously implanted probe. The probes were perfused
for dierent time periods with a solution containing NGF. With an inlet
NGF concentration of 8  1011 M, the vitreous concentrations were found
to be 0:08  1012 , 0:87  1012 , and 0:86  1012 M when the solution was
perfused for 1, 4, and 6 hours, respectively. When 9  1010 and 7:6  109
M concentrations of NGF were perfused for 4 hours, the vitreous concentrations were 012  1011 and 3:8  1011 M, respectively. The same model
was used to delivery ganciclovir into the rabbit vitreous (60). Ganciclovir
concentration in the vitreous after microdialysis infusion of 120 ml 3:4  1
04 M solution was 10:5  107  0:99 M. Microdialysis probe was also
used to administer 5-uorouracil, benzyl penicillin, daunomycin, and dexamethasone into the vitreal space of rabbits (25). The vitreal concentrations
achieved after perfusion were 5  105 M, 2 mM, 1.2 mM, and 1:2  107 M,
respectively.
Stempels et al. (23) developed a removable ocular microdialysis system
using scleral port for the rst time for measuring the vitreous levels of
biogenic amines. This model allowed long-term experiments using microdialysis. Dutch pigmented rabbits were equipped with a scleral entry port
(internal diameter 0.6 mm) with a removal closing plug. The scleral port was
sutured bilaterally about 23 mm from the limbus in the temporal superior
quadrant and covered with conjunctiva. The light-adapted rabbits were
intubated and maintained under halothane anesthesia with spontaneous
breathing. The pupils were dilated with one drop of homatropine 1%.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

267

Figure 7 Vitreous and blood concentration of ceftazidime after an intramuscular

injection of 1 mg/kg in (a) healthy rabbit eye and (b) mildly inamed eye.

The conjunctiva was reopened, the closing plug was removed and a microdialysis probe, with a shaft diameter of 0.6 mm and a cut-o value of 20
kDa, was inserted into the midvitreous. The position of the probe tip was
conrmed by direct illumination through the pupil. The perfusion uid used
was Ringers solution with a Ca2+ concentration of 0.75 mM. The perfusion
uid was pumped at a ow rate of 2 mL/min, and the samples were collected
every 20 minutes. Using this model, the concentration dihydroxyphenyl
acetic acid was found to be three times higher than in the bovine vitreous.
No signicant dierence was observed between simultaneously taken left

Copyright 2003 Marcel Dekker, Inc.

268

Macha and Mitra

Figure 8 Vitreous concentration of ceftazidime after an intravitreal injection of 1

mg ceftazidime in (a) healthy eye and (b) mildly inamed eye.

and right eye samples nor between days 1, 7, 11, and 14 for dopamine,
dihydroxyphenyl acetic acid (Fig. 9) and noradrenaline. This study proved
that ocular microdialysis could be carried out over several hours and repeatedly in the same animal.
Macha and Mitra (27) have used the technique to study the ocular
pharmacokinetics of cephalosporins after intravitreal administration and
also investigated the presence of peptide transporters on the retina. New

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

269

Figure 9 Concentrations of dihydroxyphenyl acetic acid in the dialysates of rabbit

vitreous on days 1, 7, 11, and 14.

Zealand albino male rabbits, weighing 22.5 kg, were kept under anesthesia throughout the experiment. A concentric microdialysis probe was
implanted into the midvitreous chamber using a 22 gauge needle about
3 mm below the limbus through the pars plana. Another linear microdialysis probe was implanted across the cornea in the aqueous humor using a
25 gauge needle (Fig. 10). The probes were perfused with isotonic phosphate buer saline (pH 7.4) at a ow rate of 2 mL/min and the samples
were collected every 20 minutes over a period of 10 hours. Animals were
allowed to stabilize for 2 hours prior to initiation of a study. Ocular
pharmacokinetics of cephalosporins were investigated following intravitreal administration of 500 mg of cephalexin, cephazolin, and cephalothin.
Inhibition experiments were carried in vivo using two dipeptides, gly-pro
and gly-sar. The dipeptides were administered by a bolus injection intravitreally 30 minutes prior to the administration of cephalosporins, followed
by continuous perfusion through the vitreous probe to maintain the

Copyright 2003 Marcel Dekker, Inc.

270

Macha and Mitra

Figure 10 Diagrammatic representation of the microdialysis probes implanted in

the anterior chamber and vitreous of the eye.

steady-state dipeptide concentration during an experiment. Vitreal elimination half-lives of cephalexin, cefazolin, and cephalothin after intravitreal
administration were found to be 185:38  27:25 min, 111:40  17:17 min,
and 146:68  47:52 min, respectively. Cephalexin (224:39  84:56 mg.min/
mL) was found to generate higher concentrations in the aqueous humor
compared to cefazolin (85:37  45:11 mg.min/mL). The pharmacokinetic
parameters of cephalexin in the presence of gly-pro, i.e., AUC
(44452:06  3326:55 mg.min/mL), clearance (0:0013  0:0004 mL/min),
and terminal elimination half-life (825:12  499:95 min), were found to
be signicantly dierent from that of the control 14612:83  4036:47
mg.min/mL, 0:0036  0:0011 mL/min, and 187:96  65:12 min, respectively) (Fig. 11). In the case of cefazolin, the control parameters
(30199:06  8819:07 mg.min/mL, 0:0018  0:0006 mL/min, and 229:53  3
7:31 min, respectively) were found to be similar, except the terminal elimination half-life, to those in the presence of gly-pro (26648:31  4156:56
mg.min/mL, 0:0019  0:0003 mL/min, and 881:03  469:67 min, respectively) (Fig. 12). Gly-sar was found to have no signicant eect on the
pharmacokinetics of both drugs. These studies indicated the involvement

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

271

Figure 11 Vitreous concentration-time prole of cephalexin (50 mg) in the presence

of inhibitors after intravitreal administration. The line drawn represents the nonlinear least-squares regression t of the model to the concentration-time data.

of a peptide carrier in the transport of cephalosporins across the retina.

Although gly-pro inhibited the elimination of cephalexin from the vitreous,
the eect of the a-amino group on the specicity of cephalosporins
towards peptide carriers was not clearly established.
Furthermore, Macha and Mitra have utilized the microdialysis technique to delineate the ocular pharmacokinetics of ganciclovir (GCV) and
its ester prodrugs (acetate, propionate, butyrate, and valerate). The prodrugs generated sustained therapeutic concentrations of GCV over a prolonged period of time after intravitreal administration. Drugs were
administered (0.2 mmol) intravitreally and the samples were collected
every 20 minutes over a period of 10 hours. The representative anterior
and vitreous chamber concentration-time proles of GCV following intravitreal administration are shown in Figure 13. The vitreal terminal phase
elimination half-life (t1=2 b) of GCV was found to be 426  109 minutes.
The proportion of GCV eliminating through the anterior chamber pathway was about 1%. The representative vitreous concentration-time prole

Copyright 2003 Marcel Dekker, Inc.

272

Macha and Mitra

Figure 12 Vitreous concentration-time prole of cefazolin (50 mg) in the presence

of inhibitors after intravitreal administration. The line drawn represents the nonlinear least-squares regression t of the model to the concentration-time data.

of the GCV butyrate is depicted in Figure 14. The hydrolysis rate and
clearance of the prodrugs increased with the ascending ester chain length.
Vitreal elimination half-lives t1=2 k10 ) of GCV, monoacetate, monopropionate, monobutyrate, and valerate esters of GCV were 170  37, 117  50,
122  13, 55  26, and 107  14 minutes, respectively. A parabolic relationship was observed between the vitreal elimination rate constant k10
and the ester chain length. The Cmax for the regenerated GCV after the
prodrug administration was found to be 2:75  0:431, 6:66  0:570,
8:03  1:19, and 8:26  1:76 mg for acetate, propionate, butyrate, and valerate esters, respectively. The mean residence time of the regenerated GCV
after prodrug administration was found to be three to four times the value
obtained after GCV injection. The low proportions of aqueous levels of
GCV indicate the retinal pathway as the major route of elimination. These
studies have shown that the ester prodrugs generated therapeutic concentrations of GCV in vivo and the MRT of GCV could be enhanced threeto-fourfold through prodrug modication.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

273

Figure 13 Concentration-time prole of GCV (50.0 mg) following intravitreal

administration: (*) vitreal and (~) anterior chamber concentrations. The line
drawn represents the nonlinear least-squares regression t of the model to the concentration-time data.

Figure 14 Concentration-time prole of GCV monobutyrate (63.3 mg) following

intravitreal administration: (*) vitreal GCV monobutyrate and (~) vitreal GCV
concentrations. The line drawn represents the nonlinear least-squares regression t of
the model to the concentration-time data.

Copyright 2003 Marcel Dekker, Inc.

274

VII.

Macha and Mitra

CONCLUSIONS

Microdialysis has been shown to be a very useful tool to study ocular

pharmacokinetics. The major strengths of the technique appear to be its
simplicity and its ability to monitor drug and metabolite concentration
and deliver the drugs. Despite its increasing popularity, microdialysis is
still far from being a routine method in eye research. Until now, much
work has gone into adapting and improving the technology involved.
Future studies need to be focused on the methodological problems and
limitations that could lead to erroneous data interpretation and conicting
results.

ACKNOWLEDGMENTS
Supported by NIH grants R01 EY09171-08 and R01 EY10659-07.

REFERENCES
1. G. Raviola. (1977). The structural basis of the blood-ocular barriers. Exp. Eye
Res. 25:2763.
2. S. P. Donahue, J. M. Khoury, and R. P. Kowalski. (1996). Common ocular
infections. A prescribers guide. Drugs 52:526540.
3. P. R. Pavan, E. E. Oteiza, B. A. Hughes, and A. Avni. (1994). Exogenous
endophthalmitis initially treated without systemic antibiotics. Ophthalmology
101:12891297.
4. L. IH. (1984). Anti-infective agents. In: Pharmacology of the Eye. M. Sears,
ed. New York: Springer-Verlag, pp. 385446.
5. B. D. Kuppermann, J. G. Petty, D. D. Richman, W. C. Mathews, S. C.
Fullerton, L. S. Rickman, and W. R. Freeman. (1993). Correlation between
CD4+ counts and prevalence of cytomegalovirus retinitis and human immunodeciency virus-related noninfectious retinal vasculopathy in patients with
acquired immunodeciency syndrome. Am. J. Ophthalmol. 115:575582.
6. J. G. Gross, S. A. Bozzette, W. C. Mathews, S. A. Spector, I. S. Abramson, J.
A. McCutchan, T. Mendez, D. Munguia, and W. R. Freeman. (1990).
Longitudinal study of cytomegalovirus retinitis in acquired immune deciency
syndrome. Ophthalmology 97:681686.
7. D. A. Jabs, C. Enger, and J. G. Bartlett. (1989). Cytomegalovirus retinitis and
acquired immunodeciency syndrome. Arch. Ophthalmol. 107:7580.
8. D. A. Jabs. (1992). Treatment of cytomegalovirus retinitis1992. Arch.
Ophthalmol. 110:185187.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

275

9. M. J. Daily, G. A. Peyman, G. Fishman. (1973). Intravitreal injection of

methicillin for treatment of endophthalmitis. Am. J. Ophthalmol. 76:343350.
10. J. Ben-Nun, D. A. Joyce, R. L. Cooper, S. J. Cringle, and I. J. Constable.
(1989). Pharmacokinetics of intravitreal injection. Assessment of a gentamicin
model by ocular dialysis. Invest. Ophthalmol. Vis. Sci. 30:10551061.
11. A. G. Palestine, M. A. Polis, M. D. DeSmet, B. F. Baird, J. Falloon, J. A.
Kovacs, R. T. Davey, J. J. Zurlo, K. M. Zunich, M. Davis, et al. (1991). A
randomized controlled trial of foscarnet in the treatment of cytomegalovirus
retinitis in patients with AIDS. Ann. Intern. Med. 115:665673.
12. L. B. Sheiner, B. Rosenberg, and V. V. Marathe. (1977). Estimation of population characteristics of pharmacokinetic parameters from routine clinical
data. J. Pharmacokinet. Biopharm. 5:445479.
13. G. L. Drusano, W. Liu, R. Perkins, A. Madu, C. Madu, M. Mayers, and M.
H. Miller. (1995). Determination of robust ocular pharmacokinetic parameters in serum and vitreous humor of albino rabbits following systemic
administration of ciprooxacin from sparse data sets by using IT2S, a population pharmacokinetic modeling program. Antimicrob. Agents Chemother.
39:16831687.
14. M. Mayers, D. Rush, A. Madu, M. Motyl, and M. H. Miller. (1991).
Pharmacokinetics of amikacin and chloramphenicol in the aqueous humor
of rabbits. Antimicrob. Agents Chemother. 35:17911798.
15. M. H. Miller, A. Madu, G. Samathanam, D. Rush, C. N. Madu, K.
Mathisson, and M. Mayers. (1992). Fleroxacin pharmacokinetics in aqueous
and vitreous humors determined by using complete concentration-time data
from individual rabbits. Antimicrob. Agents Chemother. 36:3238.
16. R. J. Perkins, W. Liu, G. Drusano, A. Madu, M. Mayers, C. Madu, and M.
H. Miller. (1995). Pharmacokinetics of ooxacin in serum and vitreous humor
of albino and pigmented rabbits. Antimicrob. Agents Chemother. 39:1493
1498.
17. W. Liu, Q. F. Liu, R. Perkins, G. Drusano, A. Louie, A. Madu, U. Mian, M.
Mayers, and M. H. Miller. (1998). Pharmacokinetics of sparoxacin in the
serum and vitreous humor of rabbits: physicochemical properties that regulate
penetration of quinolone antimicrobials. Antimicrob. Agents Chemother.
42:14171423.
18. L. Bito, H. Davson, E. Levin, M. Murray, and N. Snider. (1966). The concentrations of free amino acids and other electrolytes in cerebrospinal uid in
vivo dialysate of brain, and blood plasma of the dog. J. Neurochem. 13:1057
1067.
19. M. Telting-Diaz, D. O. Scott, and C. E. Lunte. (1992). Intravenous microdialysis sampling in awake, freely-moving rats. Anal. Chem. 64:806810.
20. J. E. Robinson. (1995). Microdialysis: a novel tool for research in the reproductive system. Biol. Reprod. 52:237245.
21. D. G. Maggs, W. P. Borg, and R. S. Sherwin. (1997). Microdialysis techniques
in the study of brain and skeletal muscle. Diabetologia 40(Suppl. 2):S7582.
(1987).

Copyright 2003 Marcel Dekker, Inc.

276

Macha and Mitra

22. G. Gunnarson, A. K. Jakobsson, A. Hamberger, and J. Sjostrand. (1987).

Free amino acids in the pre-retinal vitreous space. Eect of high potassium
and nipecotic acid. Exp. Eye Res. 44:235244.
23. N. Stempels, M. J. Tassignon, and S. Sarre. (1993). A removable ocular
microdialysis system for measuring vitreous biogenic amines. Graefes Arch.
Clin. Exp. Ophthalmol. 231:651655.
24. P. M. Hughes, R. Krishnamoorthy, and A. K. Mitra. (1996). Vitreous disposition of two acycloguanosine antivirals in the albino and pigmented rabbit
models: a novel ocular microdialysis technique. J. Ocul. Pharmacol. Ther.
12:209224.
25. J. Waga and B. Ehinger. (1997). Intravitreal concentrations of some drugs
administered with microdialysis. Acta Ophthalmol. Scand. 75:3640.
26. S. Macha and A. K. Mitra. (2001). Ocular pharmacokinetics in rabbits using a
novel dual probe microdialysis technique. Exp. Eye Res. 72:289299.
27. S. Machal and A. K. Mtira. Ocular pharmacokinetics of cephalosporins using
microdialysis. J. Ocul. Pharmacol. Ther. (in press).
28. K. D. Rittenhouse, R. L. Peier, Jr., and G. M. Pollack. (1999). Microdialysis
evaluation of the ocular pharmacokinetics of propranolol in the conscious
rabbit. Pharm. Res. 16:736742.
29. V. H. Lee and J. R. Robinson. (1986). Topical ocular drug delivery: recent
developments and future challenges. J. Ocul. Pharmacol. 2:67108.
30. S. S. Chrai, M. C. Makoid, S. P. Eriksen, and J. R. Robinson. (1973). Drop
size and initial dosing frequency problems of topically applied ophthalmic
drugs. J. Pharm. Sci. 63:333338.
31. S. S. Chrai, T. F. Patton, A. Mehta, and J. R. Robinson. (1973). Lacrimal and
instilled uid dynamics in rabbit eyes. J. Pharm. Sci. 62:11121121.
32. A. K. Mitra and T. J. Mikkelson. (1988). Mechanism of transcorneal permeation of pilocarpine. J. Pharm. Sci. 77:771775.
33. R. Abel, Jr., G. L. Boyle, M. Furman, and I. H. Leopold. (1974).
Intraocular penetration of cefazolin sodium in rabbits. Am. J. Ophthalmol.
78:779787.
34. S. J. Faigenbaum, G. L. Boyle, A. S. Prywes, R. Abel, Jr., and I. H.
Leopold. (1976). Intraocular penetrating of amoxicillin. Am. J.
Ophthalmol. 82:598603.
35. B. M. Faris and M. M. Uwaydah. (1974). Intraocular penetration of semisynthetic penicillins: methicillin, cloxacillin, ampicillin, and carbenicillin studies in experimental animals with a review of the literature. Arch. Ophthalmol.
92:501505.
36. G. A. Peyman, D. R. May, P. I. Homer, and R. T. Kasbeer. (1977).
Penetration of gentamicin into the aphakic eye. Ann. Ophthalmol. 9:871880.
37. L. Salminen. (1978). Ampicillin penetration into the rabbit eye. Acta
Ophthalmol. (Copenh.) 56:977983.
38. M. Barza, A. Kane, and J. L. Baum. (1979). Intraocular levels of cefamandole
compared with cefazolin after subconjunctival injection in rabbits. Invest.
Ophthalmol. Vis. Sci. 18:250255.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

277

39. M. Barza, A. Kane, and J. Baum. (1980). Oxacillin for bacterial endophthalmitis: subconjunctival, intravenous, both or neither? Invest. Ophthalmol. Vis.
Sci. 19:13481354.
40. W. M. Jay, R. K. Shockley, A. M. Aziz, M. Z. Aziz, and J. P. Rissing. (1984).
Ocular pharmacokinetics of ceftriaxone following subconjunctival injection in
rabbits. Arch. Ophthalmol. 102:430432.
41. F. P. Furgiuele, J. P. Smith, and J. G. Baron. (1978). Tobramycin levels in
human eyes. Am. J. Ophthalmol. 85:121123.
42. D. M. Maurice, and S. Mishima. (1984). Ocular pharmacokinetics. In:
Pharmacology of the Eye II. M. L. Sears, ed. New York: Springer-Verlag,
p. 19.
43. M. Barza, A. Kane, and J. Baum. (1983). Pharmacokinetics of intravitreal
carbenicillin, cefazolin, and gentamicin in rhesus monkeys. Invest.
Ophthalmol. Vis. Sci. 24:16021606.
44. V. N. Reddy, B. Chakrapani, and C. P. Lim. (1977). Blood-vitreous barrier to
amino acids. Exp. Eye Res. 25:543545.
45. A. G. Schenk and G. A. Peyman. (1974). Lincomycin by direct intravitreal
injection in the treatment of experimental bacterial endophthalmitis. Albrecht
Von Graefes Arch. Klin. Exp. Ophthalmol. 190: 281291.
46. S. C. Pugfelder, E. Hernandez, S. J. Fliesler, J. Alvarez, M. E. Pugfelder,
and R. K. Forster. (1987). Intravitreal vancomycin. Retinal toxicity, clearance, and interaction with gentamicin. Arch. Ophthalmol. 105:831837.
47. F. M. Ussery, 3rd, S. R. Gibson, R. H. Conklin, D. F. Piot, E. W. Stool, and
A. J. Conklin. (1988). Intravitreal ganciclovir in the treatment of AIDS-associated cytomegalovirus retinitis. Ophthalmology, 95:640648.
48. K. Henry, H. Cantrill, C. Fletcher, B. J. Chinnock, and H. H. Balfour, Jr.
(1987). Use of intravitreal ganciclovir (dihydroxy propoxymethyl guanine) for
cytomegalovirus retinitis in a patient with AIDS. Am. J. Ophthalmol. 103:1723.
49. G. A. Peyman, D. R. May, E. S. Ericson, and D. Apple. (1974) Intraocular
injection of gentamicin. Toxic eects of clearance. Arch. Ophthalmol. 92:42
47.
50. R. K. Shockley, W. M. Jay, T. R. Friberg, A. M. Aziz, J. P. Rissing, and M. Z.
Aziz. (1984). Intravitreal ceftriaxone in a rabbit model. Dose- and time-dependent toxic eects and pharmacokinetic analysis. Arch. Ophthalmol. 102:1236
1238.
51. A. G. Schenk, G. A. Peyman, and J. T. Paque. (1974). The intravitreal use of
carbenicillin (Geopen) for treatment of pseudomonas endophthalmitis. Acta
Ophthalmol. 52:707717.
52. S. K. Gardner. (1987). Ocular drug penetration and pharmacokinetic principles. In: Clinical Ophthalmic Pharmacology. D. W. Lamberts and D. E. Potter,
eds.
53. T. S. Lesar and R. G. Fiscella. (1985). Antimicrobial drug delivery to the eye.
Drug Intell. Clin. Pharm. 19:642654.
54. J. D. Wright, F. D. Boudinot, and M. R. Ujhelyi. (1996). Measurement and
analysis of unbound drug concentrations. Clin. Pharmacokinet. 30:445462.

Copyright 2003 Marcel Dekker, Inc.

278

Macha and Mitra

55. U. Ungerstedt. (1991). Microdialysisprinciples and applications for studies

in animals and man. J. Intern. Med. 230:365373.
56. E. C. de Lange, M. Danhof, A. G. de Boer, and D. D. Breimer. (1997).
Methodological considerations of intracerebral microdialysis in pharmacokinetic studies on drug transport across the blood-brain barrier. Brain Res.
Brain Res. Rev. 25:2749.
57. R. Tao and S. Hjorth. (1992). Dierences in the in vitro and in vivo 5-hydroxytryptamine extraction performance among three common microdialysis
membranes. J. Neurochem. 59:17781785.
58. J. Landolt, H. Langemann, T. H. Lutz, and O. Gratzl. (1991). Non-linear
recovery of cysteine and glutathione in microdialysis. In: H. Rollema, et al.,
eds. Meppel, Netherlands: Krips Repro, pp. 6365.
59. J. Waga, I. Nilsson-Ehle, B. Ljungberg, A. Skarin, L. Stahle, and B. Ehinger.
(1999). Microdialysis for pharmacokinetic studies of ceftazidime in rabbit
vitreous. J. Ocul. Pharmacol. Ther. 15:455463.
60. J. Waga. (2000). Ganciclovir delivery through an intravitreal microdialysis
probe in rabbit. Acta Ophthalmol. Scand. 78:369371.
61. A. Hamberger, C. H. Berthold, B. Karlsson, A. Lehmann, and B. Nystrom.
(1983). Extracellular GABA, glutamate and glutamine in vivo perfusion dialysis of the rabbit hyppocampus. In: Glutamine, Glutamate and GABA in the
Central Nervous System. New York: Alan R. Liss Inc., pp. 473492.
62. M. Sandberg and S. Lindstrom. (1993). Amino acids in the dorsal lateral
geniculate nucleus of the catcollection in vivo. J. Neurosci. Methods. 9:6474.
63. U. Tossman and U. Ungerstedt. (1986). Microdialysis in the study of extracellular levels of amino acids in the rat brain. Acta Physiol. Scand. 128:914.
64. A. Lehmann. (1989). Eects of microdialysis-perfusion with anisoosmotic
media on extracellular amino acids in the rat hippocampus and skeletal muscle. J. Neurochem. 53:525535.
65. J. M. Solis, A. S. Herranz, O. Herreras, J. Lerma, and R. Martin del Rio.
(1988). Does taurine act as an osmoregulatory substance in the rat brain?
Neurosci. Lett. 91:5358.
66. J. V. Wade, J. P. Olson, F. E. Samson, S. R. Nelson, T. L. Pazdernik. (1988).
A possible role for taurine in osmoregulation within the brain. J. Neurochem.
51:740745.
67. S. A. Wages, W. H. Church, and J. B. Justice, Jr. (1986). Sampling considerations for on-line microbore liquid chromatography of brain dialysate. Anal.
Chem. 58:16491656.
68. J. M. Delgado, J. Lerma, R. Martin del Rio, and J. M. Solis. (1984).
Dialytrode technology and local proles of amino acids in the awake cat
brain. J. Neurochem. 42:12181228.
69. G. Amberg, and N. Lindefors. (1989). Intracerebral microdialysis: II.
Mathematical studies of diusion kinetics. J. Pharmacol. Methods 22:157183.
70. T. Zetterstrom, L. Vernet, U. Ungerstedt, U. Tossman, B. Jonzon, and B. B.
Fredholm. (1982). Purine levels in the intact rat brain. Studies with an
implanted perfused hollow bre. Neurosci. Lett. 29:111115.

Copyright 2003 Marcel Dekker, Inc.

Posterior Segment Microdialysis

279

71. I. Jacobson, M. Sandberg, and A. Hamberger (1985). Mass transfer in brain

dialysis devicesa new method for the estimation of extracellular amino acids
concentration. J. Neurosci. Methods 15:263268.
72. P. Lonnroth, P. A. Jansson, and U. Smith. (1987). A microdialysis method
allowing characterization of intercellular water space in humans. Am. J.
Physiol. 253:E228231.
73. D. Scheller and J. Kolb. (1991). The internal reference technique in microdialysis: a practical approach to monitoring dialysis eciency and to calculating tissue concentration from dialysate samples. J. Neurosci. Methods 40:31
38.
74. H. Benveniste, A. J. Hansen, and N. S. Ottosen. (1989). Determination of
brain interstitial concentrations by microdialysis. J. Neurochem. 52:1741
1750.
75. G. Raviola. (1974). Eects of paracentesis on the blood-aqueous barrier: an
electron microscope study on Macaca mulatta using horseradish peroxidase as
a tracer. Invest. Ophthalmol. 13:828858.
76. F. P. Killey, H. F. Edelhauser, and T. A. Aaberg. (1980). Intraocular uid
dynamics. Measurements following vitrectomy and intraocular sulfur hexauoride administration. Arch. Ophthalmol 98:14481452.
77. L. L. Knudsen, T. Olsen, and F. Nielsen-Kudsk (1988). Anterior chamber
uorescein kinetics compared with vitreous kinetics in normal subjects. Acta
Ophthalmol. Scand. 76:561567.
78. J. Wagar and B. Ehinger. (2000). NGF administered by microdialysis into
rabbit vitreous. Acta Ophthalmol. Scand. 78:154155.

Copyright 2003 Marcel Dekker, Inc.

9
Ocular Penetration Enhancers
Thomas Wai-Yip Lee and Joseph R. Robinson
School of Pharmacy, University of Wisconsin-Madison, Madison,
Wisconsin, U.S.A.

I.

INTRODUCTION

Drug delivery to the eye is not an easy assignment. The cornea, being a very
important component in the visual pathway, is well protected by a number
of very eective defense mechanisms, e.g., blinking, high tear secretion rate
ushing its surface, induced lacrimation and tear protein production in
response to foreign substances, etc. These protective mechanisms provide
a challenge for pharmaceutical scientists to design drug delivery systems that
can deliver therapeutic agents in sucient concentrations to target sites.
After topical instillation of an eye drop, the drug is subject to a number of very ecient elimination mechanisms such as drainage, binding to
proteins, normal tear turnover, induced tear production, and nonproductive
absorption via the conjunctiva. Typically, drug absorption is virtually complete in 90 seconds due to the rapid removal of drug from the precorneal
area. To make matters worse, the cornea is poorly permeable to both hydrophilic and hydrophobic compounds. As a result, only approximately 10% or
less of the topically applied dose can be absorbed into the anterior segment
of the eye.
Basically, the two major barriers encountered in ocular drug delivery
are (a) short residence time in the precorneal area and (b) poor permeability
of the cornea.
Various eorts have been made to prolong the drug solution residence
time via vehicle modication (1,2), bioadhesives (3), inserts (4), etc. Another
approach to improve ocular bioavailability, which is less well understood, is
penetration enhancement. Penetration enhancement can be achieved via pro281

Copyright 2003 Marcel Dekker, Inc.

282

Lee and Robinson

drugs, penetration enhancers, etc. Prodrugs will be covered elsewhere in this

book. The main focus of this chapter will be on the use of penetration enhancers to improve ocular drug delivery. Fundamental aspects of ocular penetration enhancers will be covered, and recent advances will be presented as well.

II.

KINETIC BASIS OF THE NEED FOR PENETRATION

ENHANCEMENT

The simplest model for ocular pharmacokinetics is shown in Figure 1 (5). It

is well known that for most drugs the true absorption rate constant is much
smaller than the elimination rate constant. This will normally give rise to a
ip-op model. However, when the parallel elimination pathway is introduced (Fig. 2) (5), the apparent absorption rate constant is dened as:
Apparent kabs kabs kloss;pp
Thus, the model is not a ip-op model and drug concentration can be
described as
C FD=Vd k=k  KeKt  ekt

where F is the fraction of dose absorbed, D is the dose, k and K are

absorption and elimination rate constants, respectively, and Vd is the apparent volume of distribution. Obviously, K kelim , k kabs kloss;pp :
For many drugs, kloss; pp is of the order of 0.50.7 min1 , being several
orders of magnitude larger than kabs , which is typically of the order of 0.001
min1 . As a result, the peak time, which is controlled by kloss;pp and kabs , is
similar (2040 min) for a wide range of compounds since kloss;pp , which is
mainly due to drainage, induced lacrimination, etc., predominates over kabs
in controlling the peak time.
In order to improve the bioavailability F kabs =kabs kloss;pp  signicantly, it is essential to increase kabs by one or two order of magnitudes
or reduce kloss;pp to a similar extent.
Several approaches have attempted to reduce the magnitude of kloss;pp .
However, it has its limit. Keister et al. (6) showed that reducing the dose
volume from 25 mL to zero brings only a fourfold improvement in bioavailability for a poorly permeable compound. However, it is practically impos-

Figure 1

A one-compartment model for ocular absorption.

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

283

Figure 2 A two-compartmental model for ocular absorption.

sible to have zero dosing volume. Therefore, small dose volumes will give an
improvement in bioavailability that is less than fourfold. Similarly, theoretical calculations showed that use of bioadhesives does not necessarily give
any benet if the cornea is poorly permeable to that compound (7). These
calculations are as follows: steady-state, at which the rates of delivery and
elimination are equal, is approached after about ve drug half-lives, and the
amount of the drug in a particular ocular compartment can be expressed as
dA=dt R  KA
where A is the amount of the drug (in mg), R is the rate of drug input, and K
is the elimination rate constant. At steady state, dA=dt 0, leading to
ASS R=K Mp t1=2 =0:693T
where Mp is the mass penetrating and t1=2 is the half-life of the drug. It is
clearly shown that prolonging the contact time T via the use of bioadhesives will lower ASS provided that Mp and t1=2 are kept constant. For drugs
where permeability is not a problem, the use of bioadhesives is benecial
since the therapeutic drug level can be sustained. On the contrary, for a
poorly permeable compound, the use of adhesives may lower ASS below the
therapeutic level. In order to bring ASS back to a therapeutic level, Mp has to
be increased. This requires the use of penetration enhancers.
Methods of penetration enhancement such as prodrugs, ion pairing,
etc. are beyond the scope of this chapter and will not be discussed. The main
focus of this chapter will be ocular penetration enhancers.

III.

TRANSPORT CHARACTERISTICS OF EPITHELIAL

TISSUES

The normal expected mechanisms of corneal penetration is shown in Table 1

(8). Transport across epithelia occurs via two pathways: transcellular and
paracellular. The former involves cell/tissue partitioning/diusion, channel

Copyright 2003 Marcel Dekker, Inc.

284
Table 1

Lee and Robinson

Expected Mechanisms of Corneal Penetration

Drug type

Apparent rate-limiting
membrane

Water soluble

Epithelium

Water and oil

soluble
Oil soluble

Epithelium-stroma

Ionizable

Stroma
Epithelium + stroma or
leaky channel

Mechanisms
Low o/w partition into epithelium
Slow diusion through epithelium
High partition rate + rapid
diusion through stroma/
endothelium
Via leaky channels
Solute movement may be
intercellular and/or transcellular
Both mechanisms operate
High o/w partition into epithelium
Rapid diusion through epithelium
Mechanism not solely dependent
upon partition coecient

Source: Adapted from Ref. 8.

diusion, and carrier-mediated transport. In contrast, the latter represents

diusive and convective transport occurring through intercellular spaces
and tight junctions. Due to its aqueous nature, hydrophilic solutes would
preferably adopt the paracellular pathway. However, there are three forms
of junctional complexes that form between cells which hinder transport of
hydrophilic molecules, namely, tight junctions (zonula occludens), intermediate junctions (belt desmosome or zonula adherens), and spot desmosomes (macula adherens) (Fig. 3) (9). Among them, the tight junction is the
uppermost and tightest, and it gives the greatest resistance for hydrophilic
molecules to go between cells. The barrier property of the tight junction can
be reected by the transepithelial electrical resistance (TEER). The higher
the TEER, the tighter the junctions that give a higher resistance for transport of molecules. Generally, epithelia with resistances in the range of 10
100  cm2 are considered leaky, whereas those with resistance ranging from
300 to 10,000  cm2 are tight. The cornea is generally classied as a
moderately tight or moderately leaky tissue (4001000  cm2). A comparison of the electrophysiology and permeability of the cornea with other
tissues is shown in Table 2 and 3, respectively (10).
The cornea also shows permselectivity (11). It has an isoelectric point
(pI) of 3.2. At pHs above the pI, it carries a negative charge and is selective
to positively charged molecules. On the other hand, at pHs below the pI, it

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

285

Figure 3 Summary of the various cell junctions found in animal cell epithelia.
(From Ref. 9.)

Table 2 Comparative Permeability Coefcients of Several Drugs Between

Cornea and Other Tissues
Permeability coecient (cm/s)
Permanent
Water
Glycerol
Benzylamine
Octanol
Amphetamine
salicylic acid
Estradiol
Progesterone
Ouabain

MW

Rabbit/dog buccal

Rabbit cornea

18

3:7  105 (rabbit)

2:6  105 (dog)
6:0  107
1:4  105
2:2  105
1:5  105
9:3  107
6:6  106
8:9  106
6:5  106

1:5  104

92
107
130
135
138
272
314
584

Source: Adapted from Ref. 10.

Copyright 2003 Marcel Dekker, Inc.

4:5  106
1:76  105
7:4  106
1:8  105

Human skin
1:4  107

1:4  105
3:9  109
4:2  107
1:1  109

286

Lee and Robinson

Table 3 Electrical Resistances and Permselective of Various

Epithelia
Tissue
Proximal renal tube
Gall bladder
Small intestine
Duodenum
Jejunum
Ileum
Colon
Gastric mucosa
Antrum
Fundus
Urinary bladder
Amphibian skin
Cornea
Free solution

Species

Resistances
( cm2 )

PK

PNA

PCl

Dog
Rabbit

67
20

1.10
2.30

1.00
1.00

0.72
0.23

Rat
Rat
Rabbit
Rabbit

98
51
100
385

1.60
1.14
1.00

1.00
1.00
1.00

0.20
0.20

Necturus
Necturus
Toad
Toad
Frog
Rabbit

1,730
2,230
3,755
763
8,700
989

1.00

0.86

1.40

1.00

0.72

1.33
0.12
1.47

1.00
1.00
1.00

4.46
0.1
1.52

Source: Ref. 10.

carries a net positive charge. As a result, a positively charged molecule can

pass across the cornea more eectively at physiological pH (7.4).
The transcellular pathway is a path that a solute uses to diuse
through the apical lipid matrix of the epithelial membrane continuing
through the cytoplasm and across the basolateral membrane. The ability
of a solute to pass across the cell using this pathway depends on the interaction of the solute with plasma membrane components, e.g., lipids, cell
surface receptors. For a molecule that adopts a passive transport mechanism, partitioning is a crucial step since this is a prerequisite for a molecule to
enter the cell. As a result, the partition coecient becomes a key factor in
determining its transport across the epithelium. The optimum partition
coecients for corneal absorption has been reported to be in the range of
10100 (12), indicating lipophilic molecules are preferred. However, other
factors such as size, charge, etc. may also play a role. Theoretically, a small
and lipophilic molecule can pass across the cornea eectively. However, this
may not always be the case, as can be easily understood when the histology
of the cornea is taken into account. The cornea is composed of three layers
(Fig. 4) (8). The outermost layer is the epithelium, which is lipophilic in
nature and has tight junctions. The middle layer is an acellular matrix,
which contains about 85% water and is therefore hydrophilic in nature.

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

Figure 4

287

A simplied diagram of histology of the cornea. (Modied from Ref. 8.)

The innermost layer is the endothelium. Although the endothelium is lipophilic, it is leaky and does not give any signicant resistance to the transport
of molecules. It is believed that the epithelium provides the major resistance
for hydrophilic/charged molecules and gives minimal resistance to small
lipophilic molecules. However, after passing across the epithelium, further
movement of these lipophilic molecules is limited by the matrix, which is
hydrophilic in nature. As a result, in order to pass across the whole cornea,
the molecule has to have a balance between its lipophilic and hydrophilic
character.
Other transport mechanisms such as carrier-mediated transport, endocytosis, etc. may also be involved in transcellular transport but they are
poorly understood.

IV.

MECHANISMS OF OCULAR PENETRATION

ENHANCERS

A detailed mechanistic description of penetration enhancement is beyond

the scope of this chapter and can be found elsewhere (13). Our main focus is
on ocular penetration enhancement (14). Ideally, penetration enhancers
should have the following characteristics (15):
1. The absorbing-enhancing action should be immediate and unidirectional, and the duration should be specic and predictable.
2. There is immediate recovery of the tissue after removing the
absorption enhancers.
3. There is no systemic and local eect associated with the enhancers.

Copyright 2003 Marcel Dekker, Inc.

288

Lee and Robinson

4. The enhancers should be physically and chemically compatible

with a wide range of drugs and excipients.
However, currently available penetration enhancers are far from satisfying
the above requirements. None have yet been approved by the FDA presumably because of safety concerns. In order to design an ecient and safe
penetration enhancer, it is necessary to have a thorough understanding of
the mechanisms of penetration enhancement. Basically, penetration enhancers work by one or more of the following mechanisms (13):
1. Altering membrane structure and enhancing transcellular transport by extracting membrane components and/or increasing
uidity.
2. Enhancing paracellular transport:
Chelating calcium ions leads to opening of tight junctions;
Inducing high osmotic pressure that transiently opens tight
junctions;
Introducing agents to disrupt the structure of tight junctions.
3. Altering mucus structure and rheology so that this diusion barrier is weakened
4. Modifying the physical properties of the drug-enhancer entity
5. Inhibiting enzyme activity
A summary of ocular penetration enhancers is shown in Table 4 (14).
Typically, ocular penetration enhancement falls into two categories: paracellular and transcellular.
A.

Enhanced Paracellular Transport

As mentioned earlier, tight junctions are the major determinant of paracellular transport. In other words, tight junctions are the primary targets for a
penetration enhancer to act on in order to improve paracellular transport.
The most well-known penetration enhancer to improve paracellular transport is EDTA, which is a calcium chelator commonly used as a preservative.
It is well known that proper functioning of tight junctions depends on
calcium ions. In the absence of calcium ions, there is a widening of tight
junctions, resulting in an increase in paracellular permeability (8). EDTA
can remove divalent ions by its chelating action. Therefore, there is no
surprise that it has a permeabilizing eect on biological membranes (16).
However, its action on the cornea is believed to be much more complicated.
Rojanasakul et al. (17) showed that severe membrane damage is evident in
corneas treated with EDTA, bile salts, and surfactants. This disruption of
plasma membrane structures by EDTA is somewhat unexpected since it is
believed that EDTA only interferes with the ability of calcium to maintain

Copyright 2003 Marcel Dekker, Inc.

Table 4

A Summary of Ophthalmic Penetration Enhancers

Surfactants
Spans 20, 40 and 85,
Tweens 20, 40 and
81, Aptet 100, G
1045, Brji 35 and 58,
Myrj 52 and 53

Concentration

Drugs

Animal

Eect

1%

Fluorescein

Human

Enhanced aqueous humor concentration;

Tween 20 and Brij 35 at HLB 1617 are
most eective and dose dependent

BL-9

0.1%

Atenolol, Timolol,
Levobunolol,
Betaxolol

Rabbit

Enhanced Papp 3.4 times for atenolol and

7.3 times for timolol

Brji 35, 48

0.05%

Atenolol, Timolol,
Levobunolol,
Betaxolol

Rabbit

Enhanced Papp 3.910.5 times for atenolol

and 1.53.9 times for timolol

Brji 98

0.05%

Atenolol, Timolol,
Levobunlool,
Betaxolol

Rabbit

Enhanced Papp 2 times for betaxolol

Bile Acids
Deoxycholic acid

0.05%

Atenolol, Timolol,
Levobunolol,
Betaxolol
Timolol

Rabbit

Enhanced Papp, 1.9 times for atenolol, 5.3

times for timolol, 1.4 times for
levobunlool, and 2.2 times for betaxolol
Enhanced Papp 2.58.3 times

0.0250.1%
Taurocholic acid

1%

Copyright 2003 Marcel Dekker, Inc.

Rabbit

Rabbit

Enhanced 2.4 times for atenolol, 1.5 times

for carteolol, 1.4 times for tilisolol, and
2.1 times for timolol
Enhanced Papp 4.5 times for FD-4 and
7.1 times for FD-10

289

1%

Atenol, Carteolol,
Tilisolol, Timolol,
Befunolol
FD-4, FD-10

Rabbit

Ocular Penetration Enhancers

Enhancers

290

Table 4

Continued

Enhancers

Taurodeoxycholic acid

Concentration

Enhanced penetrated amount 7.2 times

Enhanced penetrated amount slightly

0.05%

Atenolol, Timolol,
Levobunolol,
Betaxolol
Timolol
6-Carboxyuorescein
FD-4

Rabbit

Enhanced Papp 5.8 times for atenolol and

1.6 times for timolol

Rabbit
Rabbit
Rabbit

Enhanced Papp 5.25.5 times

Enhanced penetrated amount 593 times
Enhanced penetrated amount 30.961.5
times

Rabbit

Enhanced Papp 2.1 times for timolol and

1.6 times for betaxolol

Rabbit

Enhanced Papp 8.311.0 times

Rabbit

Enhanced Papp 3.0 times for atenolol and

1.5 times for betaxolol

Rabbit

Enhanced Papp 3.3 times at 0.1%

Rabbit

Enhanced Papp 20.3 times for atenolol,

8.9 times for carteolol, 5.1 times for
tilisolol, and 3.0 times for timolol

0.05%

0.05%

Copyright 2003 Marcel Dekker, Inc.

0.5%

Atenolol, Timolol,
Levobunolol,
Betaxolol
Timolol
Atenolol, Timolol,
Levobunol,
Betaxolol
Timolol
Atenolol, Carteolol,
Tilisolol, Timolol,
Befunolol

Lee and Robinson

Rabbit
Rabbit

0.0750.1%
Fatty acids
Capric acid

Eect

6-Carboxyuorescein
FD-4

0.0750.1%
Tauroursodeoxycholic
acid

Animal

10 mM
210 mM

0.0750.1%
10 mM
210 mM
Urodeoxycholic acid

Drugs

Concentration

Preservatives
Benzalkonium
chloride

0.01%

Prostaglandin F2a ,
Pilocarpine,
Dexamethasone

Pig

0.01%

Tilisolol, FD-4, FD10

Rabbit

0.02%

Rabbit

0.05%

Atenolol, Timolol,
Levobunolol,
Betaxolol
FD-4, FD-10

Rabbit

0.005-0.02%
0.010.03%
0.025%

Fluorescein
Carbachol
Titmolol

Rabbit
Rabbit
Rabbit

0.01%

Pilocarpine,
Dexamethasone

Pig

0.00250.05%

Fluorescein

Benzyl alcohol

0.5%

Tilisolol, FD-4, FD10

Rabbit,
Human
Rabbit

Enhanced permeability signicantly over

at 0.005%
Enhanced Papp 2.6 times for FDA and
8.1 times for FD-10

Chlorbutanol

0.5%

Pilocarpine,
Dexamethasone

Pig

Enhanced Papp 1.8 times for pilocarpine

and 4.7 times for dexamethasone

Chlorhexidine
digluconate

Drugs

Animal

Eect

Enhanced Papp 7.2 times for

prostaglandin F2a , 1.7 times for
pilocarpine, and 3.3 times for
dexamethasone
Enhanced Papp 3.5 times for tilisolol, 28.8
times for FD-4, and 37.1 times for FD10
Enhanced Papp 5.2 times for atenolol, 2.7
times for timolol, and 1.3 times for
betaxolol
Enhanced Papp 43.6 times for FDA and
60.6 times for FD-10
Increased permeability 42.5 times 0.02%
Enhanced miotic response about 20 times
Enhanced the ocular absorption about
80% and the systemic absorption about
40%
Enhanced Papp 1.5 times for
dexamethasone

Ocular Penetration Enhancers

Enhancers

291

Copyright 2003 Marcel Dekker, Inc.

292

Table 4

Continued
Concentration

2-Phenylethanol

0.5%

Tilisolol, FD-4, FD10

Rabbit

Enhanced Papp 2.7 times for tilisolol, 5.6

times for FD-4, and 4.8 times for FD-10

Paraben

0.04%

Tilisolol, FD-4, FD10

Rabbit

Enhanced Papp 1.9 times for FD-10

Propyl paraben

0.02%

Dexamethasone

Pig

Enhanced Papp 1.5 times

Chelating Agents
EDTA

0.5%

Atenolol, Timolol,
Levobunolol,
Betaxolol
Atenolol, Carteolol,
Tilisolol, Timolol,
Befunolol
FD-4, FD-10

Rabbit

Enhanced Papp 1.4 times for atenolol

Rabbit

Enhanced Papp 1.7 times for atenolol, 2.9

times for carteolol, 2.3 times for timolol,
and 1.6 times for befunolol
Enhanced Papp 15.5 times for FDA and
39.0 times for FD-10
Enhanced Papp 31 times for atenolol and
1.9 times for timolol at 0.5%

0.5%

0.5%
0.10.5%

Others
Azone

Copyright 2003 Marcel Dekker, Inc.

Drugs

Animal

Rabbit
Rabbit

Eect

0.05%

Atenlool, Timolol,
Levobunolol,
Betaxolol
Timolol

Rabbit

Enhanced ocular and systemic absorption

signicantly

0.0251.0%

Cimetidine

Rabbit

Enhanced Papp 14.187.0 times

Lee and Robinson

Enhancers

Enhancers

Concentration

Animal

Eect
Enhanced Papp 29.1 times for
acetazolamide, 16.3 times for
guanethidine, >87.3 times for
guanethidine, 31.3 times for cimetidine,
2.2 times for bunolol, and 2.2. times for
prednisolone

Rabbit

5%

Cyclosporine

Rabbit

0.0251.0%

Cimetidine

Rabbit

Enhanced ocular bioavailability 3.922.0

times after instillation in rabbits
Enhanced penetration into the cornea and
rapidly achieved state-steady state drug
level
Enhanced Papp 17.4-64.3 times

0.0251.0%

Cimetidine

Rabbit

Enhanced Papp 5.7100.3 times

0.0251.0%
0.05%

Cimetidine
Atenolol, Timolol,
Levobunolol,
Betaxolol
Timolol

Rabbit
Rabbit

Enhanced Papp 2577 times

Enhanced Papp 16.5 times for atenolol,
11.0 times for timolol, 1.3 times for
levobunolol, 2.0 times for betaxolol
Enhanced Papp 2.1 times at 0.01%, 3.3
times at 0.015%, and 8.3 times at
0.025%

0.010.025%

Rabbit

Ocular Penetration Enhancers

Rabit

0.0250.1%

Acetazolamide,
Sulfacetamide,
Guanethidine,
Cimetidine,
Bunolol,
Predisolone,
Flurbiprofen
Amide
Cimetidine

0.10.5%

Hexamethylene
Lauramide
Hexamethylene
Octanamide
Decylmethylsulfoxide
Saponin

Drugs

293

Copyright 2003 Marcel Dekker, Inc.

294

Table 4 Continued
Enhancers

Concentration

Drugs

Animal

0.5%

Atenolol, Carteolol,
Tilisolol, Befunolol

Rabbit

0.5%

FD-4, FD-10

Rabbit

Eect
Enhanced Papp 31.9 times for atenolol,
13.2 times for carteolol, 7.6 times for
tilisolol, 3.3 times for timolol, 2.7 times
for befunolol
Enhanced Papp 100 times for FD-4 and
114 times for FD-10.

FD-4: FITC-dextran (average molecular weight 4400); FD-10: FITC-dextran (average molecular weight 9400); Papp: apparent permeability
coefcient.
Source: Modied from Ref. 14.

Lee and Robinson

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

295

intercellular integrity, but the eect may be due to concentration and contact time. Nishihata et al. (18) showed that EDTA caused leakage of cell
proteins from rectal epithelia. Therefore, there is a possibility that EDTA
can exert multiple eects on biological membranes. However, it is believed
that its primary action is still on the integrity of tight junctions since it fails
to improve delivery of progesterone, which appears to penetrate the cornea
primarily by the transcellular route (16).
Cytochalasins are a group of small molecules that bind specically to
actin microlaments, the major component of the cytoskeleton. It has been
shown that the cytoskeleton participates in regulation of epithelial permeability in a variety of conditions (19). Therefore, it is a reasonable strategy to
design a penetration enhancer to act specically on the cytoskeleton in order
to improve paracellular transport. Rojanaskul et al. (17) showed that cytochalasin B decreases TEER of the cornea in a dose-dependent manner. They
also studied the safety prole of cytochalasin B in vitro. Confocal microscopy showed that cytochalasin B produced negligible damage eect on the
cell membrane. Moreover, replacement of cytochalasin B after 30-minute
treatment with drug-free GBR results in a complete restoration of TEER, a
process that is completed within 30 minutes after solution replacement.
However, prolonged exposure time (e.g., > 1 hour) results in permanent
damage with incomplete recovery within the time frame of the experiment.
It is obvious that cytochalasin B is a relatively specic and safe ocular
penetration enhancer compared with other classical penetration enhancers
such as bile salts, surfactants, etc.
Another strategy to improve paracellular transport is to make use of
active transport systems. Active transport of glucose or amino acids, which
is coupled to sodium transport, across the intestinal mucosa into the intercellular lateral spaces creates an osmotic force for uid ow, and this in
turns triggers contraction of the perijunctional actomyosin, resulting in
increased paracellular permeability (decreased TEER) (20). MartinezPalomo (21) showed that a hypertonic lysine solution induced a reversible
opening of the tight junction of the toad urinary bladder without gross
deformation of tight junctions. The decreased TEER was reversed when
an isotonic solution was replaced on the apical side of the epithelium.
Due to complete reversibility, it appears that increased paracellular transport by applying a hypertonic solution works in isolated tissues. However, a
hypertonic solution may irritate the eye and induce tear production, which
ushes away the applied drug. Therefore, the practicality of this approach is
questionable and has yet to be conrmed.

Copyright 2003 Marcel Dekker, Inc.

296

B.

Lee and Robinson

Enhanced Transcellular Transport

Enhancers that increase transcellular permeability to drugs probably do so

by aecting membrane lipids and protein components. It is shown that fatty
acids and their derivatives have been found to act primarily on the phospholipid component of membranes thereby creating disorder, resulting in
increased permeability (22).
Membrane cholesterol is another target for enhanced transcellular
delivery. It was postulated that extraction of cholesterol out of the epithelial
membrane by medium chain monoglycerides, glyceryl-monooctanoate, glyceryl-1-monodecanoate, and glyceryl-monododecanoate promoted rectal
cefoxitin absorption. However, some fatty acids act on the protein component in membranes. This is certainly the case for caprylate (22).
Nonprotein thios are another membrane component where certain
enhancers can act. The good correlation between reduced nonprotein thiols
and enhanced transport of hydrophilic compounds suggests an important
role for nonprotein thiols in preventing the transport of hydrophilic compounds (22). Murakami et al. (23) showed that depletion of these nonprotein thiols by treating with SH-modifying agents like diethyl maleate, diethyl
ethoxymethylenemalonate, ethanol, or alicylates enhanced the mucosal to
serosal transport of many hydrophilic compounds including cefoxitin and
phenol red in rat intestinal tissue.

V. NEW PENETRATION ENHANCERS

Typically, classical penetration enhancers have a nonspecic action on biological membranes. They work by reversibly or permanently damaging the
membranes so that their safety is questionable. Newer penetration enhancers have been introduced in ocular drug delivery recently with an aim to
solving this problem.
A.

Cyclodextrin

Cyclodextrins are a group of homologous cyclic oligosaccharides consisting

of six, seven, and eight glucose units, namely, a-, b-, and g-cyclodextrin (Fig.
5) (24), respectively. Typically, cyclodextrins act as true carriers by keeping
hydrophobic molecules in solution by their hydrophobic cores, i.e., complexation between hydrophobic molecules and the inner hydrophobic core
of cyclodextrins. In other words, they are not capable of modifying the
permeability of a biological barrier. On the other hand, drug absorption
may be limited by the release of the drug from the drug-cyclodextrin com-

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

297

Figure 5 Structures of a-, b-, and g-cyclodextrins. (From Ref. 24.)

plex. As a result, addition of a cyclodextrin in an ophthalmic formulation

does not necessarily increase the ocular bioavailability. It may adversely
aect drug absorption (25,26).
However, a recent study showed that a-cyclodextrin has a signicant
penetration-enhancing eect (10-fold) on the corneal permeability of pilocarpine (27). It was speculated that such an eect was achieved by direct
interaction of the a-cyclodextrin with the corneal epithelium, which led to
subsequent destabilization of the cell membrane since the same eect was
also observed when the cornea was pretreated with the a-cyclodextrin before
the transport study. Although a-cyclodextrin only produced minimal damaging eect on the cornea at low concentration (8%), the toxic eect at higher
concentration (13% in this study) is still not known.
B.

1-Dodecylazacycloheptan-2-one (Azone1)

Although Azone1 was extensively studied as a percutaneous penetration

enhancer (2830), its use in the ocular route is still in its infancy. The
exact mechanism of the penetration enhancer eect is not known.
However, it can be speculated in terms of its lipophilicity (31). Azone1 is
highly lipophilic (log P>7), which can incorporate into lipoidal cell membrane and exert a penetration enhancing eect. It is interesting to note that
0.1% can enhance corneal penetration of hydrophilic compounds by at least
20-fold but inhibit corneal penetration of lipophilic compounds such as
urbiprofen. This was further supported by an earlier study that demonstrated that Azone1 did not improve the ocular bioavailability of levobunolol (32). It was speculated that insertion of Azone1 into the corneal
epithelium changed the structure and uidity of the cell membrane.

Copyright 2003 Marcel Dekker, Inc.

298

Lee and Robinson

Loosening of tight junctions and subsequent water (corneal swelling was

observed in the study) and drug inux might also be one of the possible
mechanisms. It was believed that retardation of absorption of lipophilic
molecules was due to creation of a more hydrated barrier, which retards
entry of lipophilic molecules. However, this is a pure speculation, and
further experiments need to be done to clarify the exact mechanism.
Safety is still the major concern in using Azone1 as an ocular penetration enhancer. It is necessary to keep the Azone1 concentration to a
minimum (<0.1%) since higher concentrations might cause ocular discomfort, conjunctival hyperemia, and epithelial thinning as a result of erosion
and/or atrophy (33). Fortunately, in vitro experiments showed that 0.1%
was enough to enhance penetration for most of the compounds under investigation to a signicant extent (33).
C.

Saponin

Saponin is a type of polysaccharide isolated from the bark of the Quillaja

saponaria tree. Saponin is an amphiphilic compound that has surface activity. As a result, although the exact mechanism of the penetration enhancing
eect is not well understood, it is believed that the penetration enhancing
eect relies solely on its detergent action. 0.5% Saponin increased corneal
permeability of atenolol by two- to threefold but only slightly improved the
delivery of relatively lipophilic timolol and befunolol. Saponin was also
extensively studied as a penetration promoter for systemic delivery of
macromolecules via the eye. Saponin demonstrated improvement in the
systemic delivery of insulin and glucagon via the ocular route in various
species (3439). It was shown that the penetration-enhancing eect of saponin was simply a direct detergent action since there did not exist a linear
correlation between ecacy and surfactant strength. Therefore, the exact
mechanism is still a mystery. The major concern in using saponin in an
ophthalmic product is still safety. Concentrations higher than 0.5% are
irritating to the eye (35).

VI.

NOVEL PENETRATION ENHANCERS

A.

a-Amino Acid

Most of classical penetration enhancers improve the paracellular transport

across a biological membrane by damaging the tight junctions to varying
degrees by their nonspecic actions. As a result, few of them are approved
by the FDA because of safety concern. Emisphere Technologies synthesized
a series of small molecular weight a-amino acids, which are used to promote

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

299

oral delivery (Fig. 6) (40,41). These delivery agents successfully increased

absorption of several macromolecules in vivo in rats and primates, including
humans, such as salmon calcitonin (42), interferon-a (43), heparin (44), and
human growth hormone (hgH) (45). Wu (41) showed that these carriers can
increase the permeability coecient of human growth hormone across
Caco-2 monolayers by 10-fold. Although it did have some eects on paracellular transport, the major pathway was observed to be transcellular.
Failure of these carriers to improve the transport of hydrocortisone, a
transcellular marker, in Caco-2 monolayers showed that these carriers
have a specic interaction with hGH, which makes the hGH more transportable, and such an interaction does not exist in the case of hydrocortisone. It was clearly established that these carriers do not damage cell
membranes and thus are not classical penetration enhancers. Moreover,
the carrier-drug complex is not absorbed by an active transport process.

Figure 6 Chemical structures of various amino acid derivative carriers. (From Ref.
41.)

Copyright 2003 Marcel Dekker, Inc.

300

Lee and Robinson

The working mechanistic assumption is that the carrier shields those hydrophilic groups on the molecule that restrain absorption.
Previous work in our laboratory showed that these carriers increase
the permeability coecient of hGH across the cornea of a rabbit by 10-fold
(46). Further study is ongoing in our laboratory to conrm the ecacy and
toxicity of these carriers as delivery agents/carriers for ocular drug delivery.
B.

Pz-Peptide

Pz-peptide (4-phenylazobenzoxycarbonyl-Pro-Leu-Gly-Pro-D-Arg) is a
hydrophilic collagenase-labile pentapeptide with a molecular weight of
777 daltons, which is capable of triggering opening of tight junctions in a
transient, reversible manner. As a result, it can facilitate paracellular transport in rabbit intestinal segments and Caco-2 monolayers (47,48).
Interestingly, it also facilitates its own transport.
The enhancement eect of Pz-peptide on permeability of drug across
the cornea and conjunctiva was studied by Chung et al. (49). Pz-peptide
increases penetration across the cornea and conjunctiva for a wide range of
compounds such as atenolol, propranolol, mannitol, uorescein, FITCDextran 4000, etc. Compared with other traditional penetration enhancers
such as a cytochalasin B and EDTA, Pz-peptide is less potent in facilitating
paracellular transport since it fails to improve the penetration of FITCDextran 10000 across the cornea.
The mechanism of enhancement is believed to involve stimulation of
transepithelial Na+ ux at the level of the amiloride-sensitive Na+ channel
and then triggering biochemical changes, which result in opening of tight
junctions. This was demonstrated in colonic segments of rabbits and Caco-2
cell monolayers (50). However, this may not be the case in ocular tissues.
Amiloride (a Na+ channel blocker), hexamethylene amiloride (Na+/H+
exchange blocker), ouabain (a Na+/K+ ATPase inhibitor), and replacement of Na+ with choline chloride fails to inhibit Pz-peptide penetration
(49). In addition, Pz-peptide can unexpectedly enhance the penetration of
propranolol, which transports across a biological membrane solely via a
transcellular pathway. This may be due to inhibition of Pgp 170 drug eux
pump, which was found in the conjunctiva (51), since propranolol is a
substrate for this eux system. Further investigation has to be carried out
to clarify its exact mechanism of enhancement.
Although the enhancement eect of Pz-peptide is promising in vitro
(2.45.1 for cornea, 1.82.0 for the conjunctiva in the case of atenolol
and propranolol), its eect is much less pronounced in vivo. Pz-peptide fails
to enhance ocular absorption of propranolol and only improves the absorption of atenolol by 1.42.0 times. This may be due to the dilution eect of

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

301

resident tears on Pz-peptide and the applied dose or binding of Pz-peptide to

mucin in other tears proteins.
C.

Multifunctional Approach (Polymeric Penetration

Enhancers)

1.

Colloidal Systems

Colloidal systems have been extensively studied as carriers for ocular drug
delivery (52). The mechanism of enhancement is generally believed to be
related to prolonged residence time in the cul-de-sac. However, enhanced
penetration may also be one of the explanations for improved ocular delivery. Poly-e-caprolactone nanoparticles, nanocapsules, and submicron emulsions improved ocular bioavailability of indomethacin when compared with
aqueous solutions and with a suspension of microparticles (53). It is believed
that the colloidal nature, rather than the inner structure or the specic
composition of the colloidal carriers, plays a key role in the enhancement
since all three colloidal carriers improve the ocular bioavailability of indomethacin to a similar extent. Confocal microscopy showed that the colloidal
carriers penetrate into the epithelial cells of the cornea without causing
damage to the cell membrane. This suggests that these carriers enter the
epithelium via endocytosis. Therefore, these carriers act as a penetration
enhancer or an endocytotic stimulator.
2.

Bioadhesives

A mentioned earlier, in order to improve ocular bioavailability, either keli or

kabs have to be increased two- to threefold. Various means have been
attempted to prolong residence time (14). Obviously, it is desirable to
have a delivery system that can stay in the precorneal area for an extended
period of time but at the same time enhance corneal penetration. A number
of bioadhesive polymers have such properties. Typically, these are macromolecules that have already been approved by FDA for other purposes.
Therefore, safety should not be a big problem.
3.

Polyacrylates

Poly(acrylic acid) derivatives such as Carbomer and Polycarbol are used

extensively as bioadhesives (54). They do have a membrane-penetrating
enhancing eect, although the exact mechanism is not well understood. It
was demonstrated that polyacrylic acid gel signicantly increased the inux
of water in the rat rectum (55). It was speculated that this solvent drag was
responsible for the enhanced absorption of low molecular weight com-

Copyright 2003 Marcel Dekker, Inc.

302

Lee and Robinson

pounds. Reduction of mucus on the microvillus and dilatation of the intercellular space was also observed 510 minutes after administration of polyacrylic acid gel into the rat rectum. However, these changes were reversible
and returned to normal relatively soon. It was not likely that polyacrylic
acid gel enhanced penetration solely by its detergent action since it inhibited
rather than induced hemolysis, which is commonly observed with surfactants. Another possible mechanism is related to its chelating activity.
Polycarbopol and other polyacrylic acidbased polymers are able to chelate
calcium (56), which is an essential component for proper functioning of tight
junctions. In addition, chelation of cations that are essential for normal
activity of enzymes further improves bioavailability. However, this inhibitory eect may be too weak to account for the improved bioavailability (57).
In the case of ocular drug delivery, there is reported to be only a minimum
amount of metabolizing enzymes in the precorneal area. As a result, there is
not likely to be a drastic improvement in ocular bioavailability because of
this enzyme inhibitory eect.
4. Chitosan and Derivatives
Chitosan (poly[b-(1-4)-2-amino-2-deoxy-D-glucopyranose]) (58) is a hydrophilic, biocompatible, biodegradable polymer of low toxicity. It is widely
used as a pharmaceutical excipient for direct compression of tablets, controlled release rate of drugs from a dosage form, enhanced dissolution, etc.
(58). It also shows strong mucoadhesive properties (59). Chitosan was evaluated as a delivery system to increase precorneal drug residence times (60).
The positively charged chitosan can reduce the elimination rate from the
precorneal area by increasing viscosity and by its interaction with negatively
charged mucus (mucoadhesive). The presence of chitosan with tobramycin
can bring an improvement in AUC and t1=2 in the precorneal area.
Moreover, this preparation is well tolerated with minimal toxicity.
Besides increasing residence time of a drug in the precorneal area,
chitosan can be used as a potential penetration enhancer to improve delivery
across the cornea. Dodane et al. (61) showed that chitosan caused a reversible, time and dose-dependent decrease in TEER in Caco-2 cell monolayers.
The increase in permeability was further conrmed by increased mannitol
permeability. They suggested that the above eects might be due to partial
alteration of the cytoskeleton, but the exact mechanism is not known. The
slight perturbation of the plasma membrane was evident by the rise in
extracellular LDH release. However, complete recovery was observed 24
hours after exposure to a low concentration of chitosan (0.005%) for a
short period of time (<60 min). Moreover, the chitosan did not aect cell
viability as shown by the trypan blue exclusion test.

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

303

Conjugation of an enzyme inhibitor to chitosan is another strategy to

improve drug delivery (62,63). However, due to the limited amounts of
enzymes in the precorneal area, this strategy may not be benecial in ocular
drug delivery.
It was shown that mucus may inhibit the binding of chitosan to an
epithelia surface and hence decreases its absorption-enhancing eect in
intestinal epithelia. The absorption-enhancing eect of chitosan has not
been evaluated in the eye. The extent of inhibition of chitosan binding to
mucus in the precorneal area has yet to be determined.

VII.

CONCLUSIONS

The use of penetration enhancers in ocular drug delivery has been studied
for more than a decade, but none has been approved by the FDA mainly
due to their nonspecic actions, which often give an unfavorable safety
prole. In order to design a more specic penetration enhancer, it is necessary to have a better understanding of membrane transport, physiology of
tight junctions, etc. Another alternative is to reversibly modify physicochemical properties of a drug so that it becomes more transportable (e.g.,
prodrugs or carriers).
Obviously, penetration enhancement has its limit. It is not possible to
increase drug bioavailability indenitely by use of penetration enhancement
alone. Other approaches such as increased residence time and inhibition of
metabolizing enzymes should be used in conjunction with penetration
enhancement. We hope that the coming biomaterial era will bring us such
a drug delivery system.

REFERENCES
1. Sieg, J. W., and Robinson, J. R. Vehicle eects on ocular drug bioavailability.
1. Evaluation of uorometholone. J. Pharm. Sci. 64:931936, 1975.
2. Patton, T. F., and Robinson, J. R., Ocular evaluation of polyvinyl alcohol
vehicle in rabbits. J. Pharm. Sci. 64:13121316, 1975.
3. Robinson, J. R., and Mlynek, G. M. Bioadhesive and phase-change polymers
for ocular drug delivery. Adv. Drug. Delivery Rev. 16:4550, 1995.
4. Salminen, L., and Urtti, H. Prolonged pulse-entry of pilocarpine with a soluble drug insert. Graefes Arch. Clin. Exp. Ophthalmol. 221:96, 1983.
5. Worakul, N., and Robinson, J. R. Ocular pharmacokinetics/pharmacodynamics. Eur. J. Pharm. Biopharm. 44:7183, 1997.

Copyright 2003 Marcel Dekker, Inc.

304

Lee and Robinson

6. Keister, J. C., Cooper, E. R., Missel, P. J., Lang, J. C., and Hager, D. F.
Limits on optimizing ocular drug delivery. J. Pharm. Sci. 80:5053, 1990.
7. Harris, D., and Robinson, J. R. Bioadhesive polymers in peptide drug delivery. Biomaterials 11:652658, 1990.
8. Grass, G. M. Mechanisms of Corneal Drug Penetration. Ph.D. thesis, School
of Pharmacy, University of WisconsinMadison, 1985.
9. Alberts, B., Bray, D., Lewis, J., Ra, M., Roberts, K., and Watson, J. D.
Molecular Biology of the Cell, 3rd ed., Garland Publishing, Inc., New York,
1994.
10. Liaw, J., and Robinson, J. R. Ocular penetration enhancers. In: Ophthalmic
Drug Delivery Systems (Mitra A., ed.), Marcel Dekker, New York, 1993, pp.
369381.
11. Rojanasakul, Y., and Robinson, J. R. Transport mechanisms of the cornea:
Characterization of barrier permselectivity. Int. J. Pharm. 55:237246, 1989.
12. Lee, V. H. L., and Robinson, J. R. Review: Topical ocular drug delivery:
Recent developments and future challenges. J. Ocular Pharmacol. 2:67108,
1986.
13. Robinson, J. R., and Yang, X. Absorption enhancers. In: Encyclopedia of
Pharmaceutical Technology, Vol. 18 (Swarbrick J. and Boylan J. C. (eds.),
Marcel Dekker, New York, 1999, pp. 127.
14. Sasaki, H., Yamamura, K., Mukai, Nishida, K., Nakamura, J., Nakashima,
M., and Ichikawa, M. Enhancement of ocular drug penetration, Crit. Rev.
Drug Carrier Syst. 16:85146, 1999.
15. Junginger, H. E., and Verhoef, J. C. Macromolecules as safe penetration
enhancers for hydrophilic drugsa ction? Pharmaceut. Sci. Technol.
Today 1:370376, 1998.
16. Grass, G. M., Wood, R. W., and Robinson, J. R. Eects of calcium chelating
agents on corneal permeability. Invest. Ophthalmol. Vis. Sci. 26:110113,
1985.
17. Rojanasakul, Y., Liaw, J. and Robinson, J. R. Mechanisms of action of some
penetration enhancers in the cornea: laser scanning confocal microscopic and
electrophysiological studies. Int. J. Pharm. 66:131142, 1990.
18. Nishihata, T., Tomida, H., Frederick, G., Rytting, J. H., and Higuchi, T.
Comparison of the eects of sodium salicylate, disodium ethylenediamine
tetraacetic acid and polyoxylene-23-lauryl ether as adjuvants for the rectal
absorption of sodium cefoxitin. J. Pharm. Sci. 37:159, 1985.
19. Frederiksen, O., and Leyssac, P. P. Eects of cytochalasin B. and dimethylsulsulfoxide on isoosmotic transport by rabbit gall-bladder in vitro. J. Facial
(Lund.) 265:103118, 1977.
20. Madara, J. L., et al. Eects of cytochalasin D on occluding junctions of
intestinal absorptive cells: Further evidence that the cytoskeleton may inuence paracellular permeability and junctional charge selectivity. J. Cell Biol.
102:2125, 1986.
21. Martinez-Palomo, A. Structure of tight junctions in epithelia with dierent
permeability. Proc. Natl. Acad. Sci. USA 72:44874491, 1975.

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

305

22. Lee, V. H. L., ed. Peptides and Proteins Drug Delivery. Marcel Dekker, New
York, 1990.
23. Murakami, M., et al. Intestinal absorption enhanced by unsaturated fatty
acids: Inhibitory eect of sulfhydryl modiers. Biochim. Biophys. Acta
293:238, 1988.
24. Cyclodextrins in Pharmacy (Fromming, K.-H., and Szejtli, J., eds.), Kluwer
Academic Publishers, Boston, 1994.
25. Davies, N. M., Wang, G., and Tucker, I. G. Evaluation of a hydrocortisone/
hydroxypropyl-beta-cyclodextrin solution for ocular drug delivery. Int. J.
Pharm. 156:201209, 1997.
26. Jarho, P., Jarvinen, K., Urtti, A., Stella, V. J., and Jarvinen, T. Use of cyclodextrins in ophthalmic formulations of dipivefrin. Int. J. Pharm. 153:225233,
1997.
27. Siefert, B., and Keipert, S. Inuence of alpha-cyclodextrin and hydroxyalkylated b-cyclodextrin derivatives on the in vitro corneal uptake and permeation
of aqueous pilocarpine-HCl solutions. J. Pharm. Sci. 86:716720, 1997.
28. Michniak, B. B., Player, M. R., Godwin, D. A., Lockhart, C. C., and Sowell,
J. W. In vitro evaluation of azone analogs as dermal penetration enhancers.
Part 5. Miscellaneous compounds. Int. J. Pharm. 161:169178, 1998.
29. Valenta, C., and Wedenig, S. Eects of penetration enhancer on the in vitro
percutaneous absorption of progesterone. J. Pharm. Pharmacol. 49:955959,
1997.
30. Chatterjee, D. J., Li, W. Y., and Koda, R. T. Eect of vehicles and penetration enhancers on the in vitro and in vivo percutaneous absorption of methotrexate and edatrexate through hairless mouse skin. Pharm. Res. 14:1058
1065, 1997.
31. Tang-Liu, D. D-S., Richman, J. B., Weinkam, R. J., and Takruri, H. Eects
of four penetration enhancers on corneal permeability of drugs in vitro. J.
Pharm. Sci. 83:8590, 1994.
32. Tang-Liu, D. D.-S., and Burke, J. The eect of Azone1 on ocular levobunolol
absorption: Calculating the area under the curve and its standard error using
tissue sampling compartments. Pharm. Res. 5:238241, 1988.
33. Ismail, I. M., Chen, C.-C., Richman, J. B., Andersen, J. S., and Tang-Liu, D.
D.-S. Comparison of azone and hexamethylene lauramide in toxicological
eects and penetration enhancement of cimetidine in rabbit eyes. Pharm.
Res. 9:817821, 1992.
34. Chiou, G. C. Y., Chang, C. Y., and Chang, M. S. Systemic delivery of insulin
through eyes to lower the glucose concentration. J. Ocular Pharmacol. 5:81
91, 1989.
35. Pillion, D. J., Atchison, J. A., Stott, J., McCracken, D. L., Gargiulo, C., and
Meezan, E. Ecacy of insulin eyedrops. J. Ocular Pharmacol. 10:461470,
1994.
36. Pillion, D. J., McCracken, D. L., Yang, M., and Atchison, J. A. Glucagon
administration to the rat via eye drops. J. Ocular Pharmacol. 4:349358, 1992.

Copyright 2003 Marcel Dekker, Inc.

306

Lee and Robinson

37. Pillion, D. J., Amsden, J. A., Kensil, C. R., and Recchia, J. Structure-function
relationship among Quillaja saponins serving as excipients for nasal and ocular delivery of insulin. J. Pharm. Sci. 85:518524, 1996.
38. Morgan, R. V. Delivery of systemic regular insulin via the ocular route in cats.
J. Ocular Pharmacol. 11:565573, 1995.
39. Morgan, R. V., and Huntzicker, M. A. Delivery of systemic regular insulin via
the ocular route in dogs. J. Ocular Pharmacol. 12:515526, 1996.
40. Steiner, S., and Rosen, R. Delivery System for Pharmacological Agents
Encapsulated Proteinoids. U.S. Patent, 1990.
41. Wu, S.-Y. Mechanistic studies on the enhanced mucosal transport of human
growth hormone by certain amino acid derivatives. Ph.D. thesis, School of
Pharmacy, University of WisconsinMadison, 1999.
42. Leone-Bay, A., McInnes, C., Wang, N., DeMorin, F., Achan, D., Lercara, C.,
Sarubbi, D., Haas, S., Press, J., Barantsevich, E., OBroin, B., Milstein, S.,
and Patron, D. Microsphere formation in a series of derivatized a-amino
acids: Properties, molecular modeling, and oral delivery of salmon calcitonin.
J. Med. Chem. 38:42574262, 1995.
43. Leone-Bay, A., Santiago, N., Achan, D., Chaudhary, K., DeMorin, F.,
Falzarano, L., Haas, S., Kalbag, D., Leipold, H., Lercara, C., OToole, D.,
Rivera, T., Rosado, C., Sarubbi, D., Vuocolo, E., Wang, N., Milstein, S., and
Baughman, R. A. N-acylated a-amino acids as novel oral delivery agents for
proteins. J. Med. Chem. 38:42634269, 1995.
44. Brayden D., Creed E., OConnell A., Leipold, H., Lercara C., Agarwal R.,
and Leone-Bay A. Heparin absorption across the intestine: Eects of sodium
N-[8(2-hydroxybenzoyl)amino]caprylate in Caco-2 monolayers and in rat in
situ intestinal instillations. Pharm. Res. 14:17721779, 1997.
45. Leone-Bay, A., Ho, K., Agarwal, R., Baughman, R. A., Chaudhary, K.,
DeMorin, F., Genoble, L., McInnes, C., Lercara, C., Milstein, S., OToole,
D., Sarubbi, D., Variano, B., and Paton, D. R. 4-[4-[(2Hydroxybenzoyl)amino]phenyl]butyric acid as a novel oral delivery agent
for recombinant human growth hormone. J. Med. Chem. 39:25712578, 1996.
46. Robinson, J. R., unpublished data, 1998.
47. Yen, W.-C., and Lee, V. H. L. Paracellular transport of a proteolytically labile
pentapeptide across the colonic and other intestinal segments of the albino
rabbit: Implications for peptide drug design. J. Contr. Rel. 28:97109, 1994.
48. Yen, W.-C., and Lee, V. H. L. Penetration enhancement eect of Pz-peptide, a
paracellularly transported peptide, in rabbit intestinal segments and Caco-2
cell monolayers. J Contr. Rel. 36:2537, 1995.
49. Chung, Y. B., Han, K., Nishiura, A., and Lee, V. H. L. Ocular absorption of
Pz-peptide and its eect on the ocular and systemic pharmacokinetics of
topically applied drugs in the rabbit. Pharm. Res. 15:18821887, 1998.
50. Yen, W.-C., and Lee, V. H. L. Role of Na+ in the asymmetric paracellular
transport of 4-phenylazobenzyloxycarbonyl-L-Pro-L-Leu-Gly-L-Pro-D-Arg
across rabbit colonic segments and Caco-2 cell monolayers. J. Pharmacol.
Exp. Ther. J. Contr. Rel. 36:2537, 275:114119, 1995.

Copyright 2003 Marcel Dekker, Inc.

Ocular Penetration Enhancers

307

51. Saha, P., Yang, J., and Lee, V. H. L. Existence of a P-glycoprotein drug eux
pump in cultured rabbit conjunctival epithelial cells. Invest. Ophthalmol. Vis.
Sci. 39:12211226, 1998.
52. Zimmer, A., and Kreuter, J. Microcapsules and nanoparticles used in ocular
delivery systems. Adv. Drug Delivery Rev. 16:6173, 1995.
53. Calvo, P., Alonson, M. J., Vila-Jato, J. L., and Robinson, J. R. Improved
ocular bioavailability of indomethacin by novel ocular drug carriers. J.
Pharm. Pharmacol. 48:11471152, 1996.
54. Yang, X., and Robinson, J. R. Bioadhesion in mucosal drug delivery. In:
Biomaterials for Drug Delivery (Okano, T., ed.), Elsevier, Amsterdam, 1998.
55. Morimoto, K., Iwamoto, T., and Morisaka, K. Possible mechanisms for the
enhancement of rectal absorption of hydrophilic drugs and polypeptides by
aqueous polyacrylic acid gel. J. Pharmacobio-Dyn. 10:8591, 1987.
56. Kriwet, B., and Kissel, T. Interactions between bioadhesive poly(acrylic acid)
and calcium ions. Int. J. Pharm. 127:135145, 1996.
57. Luen, H. L., Bohner, V., Perard, D., Langguth, P., Verhoef, J. C., de Boer,
A. G., Merkle, H. P., and Junginger, H. E. Mucoadhesive polymers in peroral
peptide drug delivery: V. Eect of poly(acrylates) on the enzymatic degradation of peptide drugs by intestinal brush border membrane vesicles. Int. J.
Pharm. 141:39, 1996.
58. Illum, L. Chitosan and its use as a pharmaceutical excipient. Pharm. Res.
15:13261331, 1998.
59. Yoshiaki, K., Yamamoto, H., Takeuchi, H., and Kuno, Y. Mucoadhesive dllactide/glycolide copolymer nanospheres coated with chitosan to improve oral
delivery elcatonin. Pharm. Dev. Tech. 5:7785, 2000.
60. Felt, O., Furrer, P., Mayer, J. M., Plazonnet, B., Buri, P., and Gurny, R.
Topical use of chitosan in ophthalmology: tolerance assessment and evaluation of precorneal retention. Int. J. Pharm. 180:185193, 1999.
61. Dodane, V., Khan, A. M., and Merwin, J. R. Eect of chitosan on epithelial
permeability and structure. Int. J. Pharm. 182:2132, 1999.
62. Bernkop-Schnurch, A., Paikl, C., and Pasta, M. Novel bioadhesive chitosanEDTA conjugate protects leucine enkephalin from degradation by aminopeptidase N. Pharm. Res. 14:917922, 1997.
63. Bernkop-Schnurch, A., and Pasta, M. Intestinal peptide and protein delivery:
Novel bioadhesive drug-carrier matrix shielding from enzymatic attack. J.
Pharm. Sci. 87:430434, 1998.

Copyright 2003 Marcel Dekker, Inc.

10
Corneal Collagen Shields for
Ocular Drug Delivery
Shiro Higaki, Marvin E. Myles, Jeannette M. Loutsch, and
James M. Hill
LSU Eye and Vision Center of Excellence, Louisiana State University
Health Science Center, New Orleans, Louisiana, U.S.A.

I.
A.

HISTORICAL ASPECTS
Soft Contact Lens for Ocular Drug Delivery

Bandage soft contact lenses made of hydrophilic polymers are widely used
to protect eyes with various problems, including recurrent corneal erosions
and epithelial defects after corneal transplantation or refractive surgery.
Although these bandage soft contact lenses may enhance healing while
allowing the eye to remain open, they can be inserted and removed only
in the ophthalmologists oce. Additionally, soft contact lenses may harbor
pathogens, which can cause ocular infection.
The idea of using bandage soft contact lenses to deliver drugs to the
cornea was proposed as far back as 1971 by Kaufman (1). In this procedure,
the hydrophilic lens was placed on the cornea and the drug was administered
topically onto the surface of the lens. The contact lens was thought to act as
a carrier vehicle, binding the drug and releasing it slowly, thereby increasing
retention of the therapeutic agent in the tear lm and at the corneal surface.
However, Busin and Spitznas (2) and Matoba and McCulley (3) showed
that hydrogen contact lenses hydrated with drug are nearly devoid of drug
after only 1 or 2 hours on the cornea. These soft contact lenses, therefore,
are not the ideal approach for sustained, continuous ocular drug delivery.

309

Copyright 2003 Marcel Dekker, Inc.

310

B.

Higaki et al.

Collagen Used to Enhance Drug Delivery

Bloomeld et al. (4) were the rst to suggest that collagen might provide a
suitable carrier for sustained ocular drug delivery. They showed that wafershaped collagen inserts impregnated with gentamicin produced higher levels
of drug in the tear lm and tissue in the rabbit eye compared to drops,
ointment, or subconjunctival injection.
Fyodorov et al. (5) suggested substituting collagen for hydrophilic
polymer in a contact lens shape. His purpose, however, was not drug delivery but the creation of a temporary protective device to enhance healing of
the cornea. In the mid-1980s, Fyodorov and colleagues (5,6) introduced the
collagen shield for use as a bandage lens and showed that the shields
enhance corneal epithelial healing after radial keratotomy and other anterior segment surgical procedures.
Numerous vision researchers saw the collagen shield as an extension of
and improvement on Bloomelds collagen inserts (4)a potential new
vehicle for the sustained administration of drugs to the cornea. Over the
next several years, various drugs were incorporated into the collagen shield
matrix during manufacture, absorbed into the shields during rehydration,
and/or applied in topical drops directly onto shields in situ. Studies in
animal models (described below) showed that as the drug dissolved in the
shield, it was released gradually into the tear lm, resulting in increased
contact time with the cornea and increased penetration into both the cornea
and the aqueous humor. Clinical studies demonstrated that the collagen
shield is easy to use in the ophthalmologists oce, prevents delay in beginning therapy, and maintains therapeutic concentrations of drug in the eye
without the need for frequent topical instillation of drops.

II.

CORNEAL COLLAGEN SHIELD: PHYSICAL

PROPERTIES

A.

Properties of Collagen

The safety of collagen for human use is evidenced by its diverse uses and
biomedical applications. Collagen is a common constituent in soaps, shampoos, facial creams, body lotions, and food-grade gelatin. In medicine, collagen has been used in cardiovascular surgery, plastic surgery, orthopedics,
urology, neurosurgery, and ophthalmology. The major medical application
of collagen is catgut suture, which is derived from intestinal collagen (7).
Twenty-ve percent of the total body protein in mammals is collagen; it is
the major protein of connective tissue, cartilage, and bone. The secondary
and tertiary structures of human, porcine, and bovine collagen are very

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

311

similar, making it possible to use collagen derived from animal sources in

humans. Biologically, collagen is suggested to promote wound healing (7).
Nearly all studies on collagen have shown very low or no immunogenicity
(7). Of the 10 collagen types that have been characterized, types 1, 3, and 5
are the most desirable for biomedical applications because of their high
biocompatibility and low immunogenicity.
A collagen molecule consists of three polypeptide chains, called achains, which form a helix connected by interchain hydrogen bonds. This
domain of collagen, called tropocollagen, forms a rodlike unit, 260026,000
A in length and 15 A in diameter. The molecular weight of the tropocollagen
is 300,000 daltons. Collagen has a characteristic amino acid sequence: glycine appears in approximately every third position. Proline and hydroxyproline make up approximately 25% of the total amino acid content. The
hydroxyproline residues are from interchain, noncovalent cross-linkages.
Newly synthesized collagen contains only a few cross-linked tropocollagen
bers. However, with increased age, there is an increase in the percentage of
cross-linking (7).
In the manufacture of corneal collagen shields, the ability to control
the amount of cross-linking in the collagen subunits by exposure to ultraviolet (UV) light is an important physicochemical property, because the
amount of cross-linking is related to the dissolution time of the shield on
the cornea. Kuwano et al. (8) investigated the eect of collagen cross-linking
on ooxacin bioavailability. In this study, the collagen shields were not
impregnated with drug, but drops were instilled after application of the
collagen shield. They found the dissolution times for the cross-linked collagen shield were longer than those of the noncross-linked type, thereby
prolonging drug delivery times. They concluded that cross-linked collagen
shields might be useful ocular drug delivery devices because they can allow
drug concentrations to achieve high levels in the cornea and aqueous humor.
B.

Properties of Commercial Corneal Collagen Shields

The collagen shield was designed to be a disposable, short-term therapeutic

bandage lens for the cornea. It conforms to the shape of the eye, protects the
corneal surface, and provides lubrication as it dissolves. Unlike the hydrophilic plastic bandage lenses, the collagen shield oers no refractive benet;
in fact, because it is not optically clear, it reduces visual acuity to the 20/80
20/200 range. Also, the collagen shield causes some discomfort.
Bio-Cor (Bausch & Lomb Surgical, Inc., Claremont, CA) was the rst
commercially available shield that was introduced in 1986. The diameter,
base curve, oxygen permeability, thickness, water content, and other physicochemical characteristics of Bio-Cor collagen shield have been described

Copyright 2003 Marcel Dekker, Inc.

312

Higaki et al.

Table 1
Shields

Comparison of Characteristics of Two Collagen

Brand name
Origin of collagen
Dissolution time (h)
Diameter (mm)
Base curve (mm)
Dry weight (mg)
Wet weight (mg)
Water content (% H2 O)
Surface pH

ProShielda
Porcine sclera
Rapid dissolution,
12, 24, 72
14.0
9.0
5.17.9
14.052.0
75
ND

SurgiLensb
Bovine corium
12
14.5, 16.0
ND (8.79.5)
9.011.0
ND
80
5.57.5

ND = not determined.
a
Data obtained from Alcon Surgical (Fort Worth, TX). Allergan, Inc.
(Irvine, California) and Oasis have a similar product called
KeraShield and SoftShield, respectively.
b
Data obtained from Bausch & Lomb Surgical, Inc. (Claremont, CA).

elsewhere (911). Bausch & Lomb Surgical, Inc. is selling only SurgiLens
now. The shields are derived from bovine collagen and are 14.5 mm in
diameter. Dissolution time, determined by UV irradiation during manufacture, is about 12 hours (Table 1). The shields are sterilized by gammairradiation, then dehydrated and individually packaged for storage and
shipping (12). Alcon Laboratories, Inc. (Fort Worth, TX) is selling
ProShield. The rapid dissolution as well as 12-, 24-, and 72-hour shields
are available. The shields have a diameter of 14 mm and a compound
base curve that is approximately 9 mm when hydrated. The water content
is approximately 75% (Table 1).

III.

DRUG DELIVERY BY COLLAGEN SHIELDS:

EXPERIMENTAL STUDIES

A variety of studies have described the pharmacokinetics of ocular delivery

of dyes and drugs by collagen shields as well as the use of the shields in the
chemotherapy of various disorders. These studies are reviewed below and
summarized in Tables 2 and 3 (13,14).
A.

Fluorescein, a Water-Soluble Dye

To determine the ocular penetration of water-soluble compounds delivered

by collagen shields, Reidy et al. (15) applied shields hydrated in a solution of

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

313

sodium uorescein to normal eyes of volunteers and measured the uorescence in the anterior chamber by uorophotometry. The shields delivered
signicantly larger amounts of dye to the aqueous humor at 2 and 4 hours
compared with drops of the same concentration instilled every 30 minutes
over 4 hours, as well as in comparison with daily wear soft contact lenses
presoaked in 0.01% uorescein. The collagen shields did not induce any
damage to the corneal epithelium over a 2-hour period. These results
demonstrate that the collagen shield is superior to topical drops and some
soft contact lenses in delivering uorescein to the cornea and aqueous
humor. The collagen shields might also successfully deliver other watersoluble compounds, such as antibiotics, to the eye in amounts comparable
to or greater than the amounts delivered by drops over the same period of
time.
B.

Antibacterial Agents

Ideally, chemotherapy for bacterial keratitis would delivery antibiotics

rapidly to both the cornea and aqueous humor, produce concentrations of
antibiotic signicantly above the minimum inhibitory concentration (MIC)
or minimum bactericidal concentration (MBC) of ocular pathogens, and
sustain this high concentration for many hours. However, there are numerous problems associated with achieving this ideal, and numerous approaches
have been taken to solve these problems (13,16,17).
Various investigators have examined the utility of the collagen shield
for the delivery of antibiotics to the cornea and aqueous humor. In one of
the earliest pharmacokinetic studies, Unterman et al. (18) assessed the pharmacokinetics of tobramycin delivered to rabbit eyes by means of collagen
shields hydrated in solutions of either 40 or 200 mg/mL of tobramycin.
Tobramycin concentrations in the cornea and aqueous humor were determined at 2, 4, and 8 hours after application. No toxicity was observed with
shields hydrated in the 40 mg/mL solution at any time. Eight hours after
application, the corneas with shields hydrated in the 200 mg/mL solution of
tobramycin had some epithelial defects. At all times and with either hydration solution, the concentration of tobramycin in the cornea and aqueous
humor exceeded the MIC for most aminoglycoside-sensitive strains of
Pseudomonas.
OBrien et al. (19) compared collagen shields with soft contact lenses in
pharmacokinetic studies of the ocular penetration of tobramycin. Three
groups were compared: (a) eyes with collagen shields rehydrated in 3 mg/
mL of tobramycin, (b) eyes with therapeutic soft contact lenses, and (c) eyes
with neither lenses nor shields. Topical tobramycin (3 mg/mL) was applied
to all eyes every 5 minutes for a total of six doses. Aqueous humor samples

Copyright 2003 Marcel Dekker, Inc.

314

Table 2 Studies of Collagen Shield Drug Delivery

Ref.

Drug

Compared with Collagen

shield (CS)

Gentamicin

Loading dose + frequent

drops

Phinney et al., 1988 (30)

Vancomycin

Loading dose + frequent

drops

OBrien et al., 1988 (18)

Unterman et al., 1988
(17)
Hwang et al., 1989 (32)

Tobramycin
Tobramycin

Soft contact lens

Subconjunctival injection

Dexamethasone

Single drop

Hwang et al., 1989 (32)

Dexamethasone

Frequent drops

Hwang et al., 1989 (32)

Dexamethasone

CS + frequent drops
versus frequent drops

Sawusch et al., 1989 (33)

Prednisolone

Single drop

Copyright 2003 Marcel Dekker, Inc.

Tears
Cornea
Aqueous
Tears
Cornea
Aqueous
Aqueous
Cornea
Aqueous
Cornea
Aqueous
Iris
Vitreous
Cornea
Aqueous
Iris
Vitreous
Cornea
Aqueous
Iris
Vitreous
Cornea
Aqueous

Overall result with collagen

shield (CS)
CS comparable at all sites

CS comparable at all sites

CS superior
CS comparable at both sites
CS superior at all sites

CS superior at all sites

CS superior at all sites

CS superior at both sites

Higaki et al.

Phinney et al., 1988 (30)

Assay site

Fluorescein

Schwartz et al., 1990 (40)

Amphotericin B

Frequent drops
Soft contact lens
Frequent drops

Reidy et al., 1990 (44)

Cyclosporin A

Frequent drops

Murray et al., 1990 (42)

Gussler et al., 1990 (47)

Heparin
Triuorothymidine

Subconjunctival injection
Drops
CS + drops

Chen et al., 1990 (46)

Tobramycin

Drops

Taravella et al., 1998 (65)

Kuster et al., 1998 (42)

Tobramycin
Triuorothymidine

Three drops
Drops

Taravella et al., 1999 (64)

Ooxacin

Three drops

Anterior
chamber
Cornea
Aqueous
Cornea
Aqueous
Aqueous
Aqueous

Cornea
Aqueous
Conjunctiva
Aqueous
Cornea
Aqueous
Aqueous

CS superior to both
CS comparable at both
sites
CS superior
CS comparable
CS Superior
Cornea with epithelial
defect:
CS + drops superior
CS superior from 02 h
Drops superior from 48 h
CS superior at all sites

No dierence
CS higher
CS superior

Corneal Collagen Shields for Ocular Drug Delivery

Reidy et al., 1990 (13)

315

Copyright 2003 Marcel Dekker, Inc.

316

Table 3 Studies of Collagen Shield Drug Delivery in Rabbit Models of Disease

Ref.

Experimental model

Drug

Compared with collagen

shield (CS)

Pseudomonas keratitis

Tobramycin

Pseudomonas keratitis

Tobramycin

Hobden et al., 1990 (29)

Aminoglycosideresistant Pseudomonas
keratitis

Ciprooxacin
Noroxacin
Tobramycin

CS with vehicle
CS with water

Chen et al., 1990 (19)

High-risk keratoplasty

Cyclosporine
A

Drops
Drops

Hagenah et al., 1990

(52)

Epithelial wound
healing

EGF, aFGF

CS alone
Untreated corneas

Clinch et al., 1992 (22)

Pseudomonas keratitis

Tobramycin

CS + frequent drops
versus frequent drops

Copyright 2003 Marcel Dekker, Inc.

CS + frequent drops
versus frequent drops
Frequent drops
CS + delayed drops
versus second CS

Enhanced antimicrobial
eect with CS
CS comparable
antimicrobial eect
CS comparable
antimicrobial eect
Ciprooxacin >
noroxacin for
antimicrobial eect
Tobramycin, vehicle,
waterno eect
CS superior preventive
eect on graft
rejection
CS superior therapeutic
eect on graft
rejection
aFGF and CS alone
superior to untreated
EGF and CS
comparable to
untreated
Enhanced antimicrobial
eect with CS

Higaki et al.

Sawusch et al., 1988

(21)
Hobden et al., 1988 (20)

Result

Pseudomonas keratitis

Gentamicin

CS + frequent drops
versus frequent drops

Baziuk et al., 1992 (28)

Lensectomy and
vitrectomy

Gentamicin

Frequent eye drops (every

30 minutes)

Murray et al., 1992 (43)

Anterior chamber brin

clot

tPA-hydrated CS versus
control CS

Assil et al., 1992 (23)

Pseudomonas keratitis

Tissue
plasminogen
activator
Tobramycin

Pleyer et al., 1992 (41)

Candida albicans
keratitis

Amphotericin
B

Eye drops (hourly)

Callegan et al., 1994

(31)

Staphylococcus aureus
keratitis

Tobramycin

CS + frequent drops
versus frequent drops

Eye drops

CS with topical drops

every 3 hours was
eective, but less
eective than the
topical drops every
half hour
CS: gentamicin
concentrations were
lower than frequent
eye drops in aqueous
humor
CS + tPA signicantly
shortened the time to
brin clot lysis
No signicant dierence
in ecacy
CS group had
signicantly lower
fungal counts
Enhanced antimicrobial
eect with CS

Corneal Collagen Shields for Ocular Drug Delivery

Silbiger et al., 1992 (24)

317

Copyright 2003 Marcel Dekker, Inc.

318

Higaki et al.

were taken 15 and 60 minutes following the last dose. At both times, the eyes
with the collagen shields had a signicantly greater concentration of tobramycin than the eyes with soft contact lenses or the eyes that received topical
drops only.
Chen et al. (20) compared the ocular bioavailability in rabbits of 0.3%
tobramycin applied with a collagen shield with topical drop application of
tobramycin. Groups of rabbits received either a collagen shield presoaked in
tobramycin with a tobramycin drop before and after shield application or
three drops of tobramycin. The collagen shield group had higher tobramycin levels in the cornea, aqueous humor, and conjunctiva than the second
group. They concluded that the use of collagen shields together with standard ophthalmic concentrations of tobramycin is useful in achieving higher
concentrations of topically delivered drugs into the anterior segment of the
eye.
Hobden et al. (21), Sawusch et al. (22), and Clinch et al. (23) reported
the ecacy of collagen shields rehydrated with tobramycin in the therapy of
experimental Pseudomonas keratitis in rabbit eyes. Hobden et al. (21)
demonstrated that collagen shields hydrated in 4% tobramycin were as
ecacious as 4% topical drops given every 30 minutes over a 4-hour period;
the number of colony-forming units in both the shield-treated and droptreated corneas were reduced 45 log. Also, eyes with antibiotic-hydrated
collagen shields plus one topical application of tobramycin drops over the
shield halfway through the 9-hour experimental period were compared to
eyes with shields in which the shield was replaced half way through the
experimental period. No dierence in the number of bacteria was seen.
Additionally, these studies showed that the shield alone does not enhance
bacterial growth; the number of bacteria was no greater in infected corneas
treated with collagen shields hydrated in distilled water (or balanced saline
solution) than in untreated control corneas. The overall results provided
support for the ecacy and convenience of collagen shields rehydrated in
a water-soluble antibiotic such as tobramycin for the treatment of
Pseudomonas keratitis.
Assil et al. (24) compared the ecacy of a fortied (14 mg/mL) tobramycin-soaked collagen shield to the ecacy of a single loading dose (four 50
mL drops) of fortied tobramycin eyedrops in the treatment of rabbits with
Pseudomonas aeruginosainduced keratitis. Six hours after a single treatment, signicantly fewer colony-forming units of Pseudomonas were present
in the corneas of all three drug-treated groups as compared to the number of
colonies in the corneas of balanced salt solutiontreated control rabbits.
However, no signicant dierence was found between a collagen shield
presoaked in tobramycin and a single loading dose of tobramycin eyedrops
in terms of the ability to reduce Pseudomonas.

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

319

Silbiger and Stern (25) studied the eectiveness of topical gentamicin

treatment, with and without the use of corneal collagen shields, in a rabbit
model of Pseudomonas keratitis. A 13.6 mg/mL solution of gentamicin was
topically administered, and collagen shields were soaked in 13.6 mg/mL
gentamicin for 5 minutes before being placed on the cornea. One hour
after the end of the treatment, the corneas were obtained and cultured.
The use of an antibiotic-impregnated collagen shield supplemented with
topical therapy was more eective than the use of topical treatment alone.
However, the collagen shield augmented with topical treatment every 3
hours was signicantly less eective than the topical treatment every half
hour. They concluded that the use of antibiotic-impregnated collagen shields
should not replace the use of topical treatment every half hour with fortied
antibiotics as a mainstay of initial drug treatment. However, Liang et al.
(26) reported that shield therapy provided signicantly higher gentamicin
levels in the cornea and aqueous humor than the hourly drop treatment at
0.5 and 2 hours after the end of the treatment in uninfected rabbit eyes.
Callegan et al. (27) compared the ecacy of topical fortied tobramycin (1.36%) administered by collagen shields or in topical drop form to
rabbit corneas infected with Staphylococcus aureus. Eyes were treated with
shields hydrated in and supplemented with fortied tobramycin drops
applied every 1, 2, 5 or 10 hours after infection. For topical treatment
alone, tobramycin was applied following the identical regimen. Shields supplemented with tobramycin drops applied every 1, 2, or 5 hours and topical
delivery of tobramycin ever hour sterilized all corneas. Collagen shield delivery of tobramycin with supplemental topical drops can eradicate staphylococci in rabbits with less frequent dosing intervals than required with topical
therapy alone.
Dorigo et al. (28) studied collagen shield delivery of netilmicin, an
aminoglycoside antibiotic, in rabbits. Collagen shields were immersed for
10 minutes in a commercially available solution of netilmicin, at the standard concentration of 3 mg/mL. The drug levels remained above the MIC
for the usual pathogens for 18 hours in the cornea and for 6 hours in the
aqueous humor. The study showed that a very concentrated drug solution is
not required to obtain high and persistent levels of netilmicin in the cornea.
Baziuk et al. (29) investigated the intraocular drug delivery of collagen
shield and fortied eye drops in rabbits that had undergone bilateral lensectomy and vitrectomy. The left eyes were tted with collagen shields that
had been soaked for 5 minutes in 2.0 mL of gentamicin solution (40 mg/mL)
and compared with the right eyes treated with fortied gentamicin drops
(13.6 mg/mL) every 30 minutes for 12 hours. The gentamicin concentration
was higher in the aqueous humor of all eyes treated with fortied gentamicin
drops.

Copyright 2003 Marcel Dekker, Inc.

320

Higaki et al.

Hobden et al. (30) reported the use of collagen shields hydrated with
various uoroquinolones for chemotherapy of aminoglycoside-resistant
Pseudomonas. The uoroquinolones used were noroxacin (40 mg/mL)
and ciprooxacin (25 mg/mL), and the aminoglycoside control was tobramycin (40 mg/mL). In these experiments, Pseudomonas was made aminoglycoside-resistant by conjugal transfer of a plasmid. The MICs were 31.25
mg/mL for tobramycin, 0.25 mg/mL for ciprooxacin, and 0.48 mg/mL for
noroxacin. The colony-forming units from rabbit corneas treated with
ciprooxacin were reduced by 4 log compared to corneas treated with collagen shields containing tobramycin or untreated corneas. Noroxacin,
which decreased the colony-forming units approximately 2 log, was not as
eective as ciprooxacin.
Phinney et al. (31) were the rst to report the delivery of two antibiotics in combination (gentamicin and vancomycin) to uninfected rabbit
eyes using the collagen shield. Tear, corneal, and aqueous humor concentrations of each of the two antibiotics were generally higher than, or at least
similar to, those achieved by frequent topical application. Combinations of
antibiotics have the potential to cover a broad spectrum of infectious agents,
but care must be taken to test for pharmacological compatibility to avoid
potential therapeutic interference and/or toxicity.
In conclusion, these results suggested that collagen shields containing
an antibiotic could serve as a vehicle for drug delivery and could be used for
preoperative and postoperative antibiotic prophylaxis and initial treatment
of bacterial keratitis (32).

C.

Anti-Inammatory Agents

Hwang et al. (33) and Sawusch et al. (34) used collagen shields to enhance
the penetration of anti-inammatory agents. Hwang et al. (33) compared
the deliver of dexamethasone to the cornea and aqueous humor in normal
rabbit eyes by four methods: single 0.1% dexamethasone drop, hourly
drops, collagen shields hydrated in 0.1% dexamethasone, and collagen
shields hydrated in 0.1% dexamethasone followed by hourly topical 0.1%
drops. Treatment with the drug-hydrated collagen shields plus hourly drops
resulted in both peak and cumulative drug concentrations in the cornea and
aqueous humor that were two- to fourfold greater than the concentration
achieved by hourly drops alone. Collagen shields without accompanying
drops yielded drug concentrations either equal to or greater than the peak
and cumulative drug concentrations produced by hourly drops. The authors
concluded that collagen shields signicantly enhance dexamethasone penetration and would be useful for maximizing the delivery of this anti-inam-

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

321

matory agent. They also suggested that the use of collagen shields would
decrease the requirement for frequent topical drops.
Sawusch et al. (34) compared (a) collagen shields hydrated in 1%
prednisolone, (b) collagen shields receiving topical drops in situ, and (c)
topical application of 1% prednisolone drops alone. Cornea and aqueous
humor were assessed for prednisolone acetate at 30 and 120 minutes after
drug application. Both collagen shield delivery systems produced signicantly greater drug levels than topical drops alone at both times. Thus,
both these reports support the potential for collagen shield delivery of corticosteroid anti-inammatory agents (33,34).

D.

Combination Therapy

In clinical cases, combinations of drugs are often used. Milani et al. (35)
investigated the ability of collagen shields impregnated with gentamicin
sulfate and dexamethasone to deliver medication to rabbit eyes. They compared collagen shields with subconjunctival injection therapy. The collagen
shields produced aqueous humor levels of gentamicin and dexamethasone
that were lower than those produced by subconjunctival injection therapy at
30 and 60 minutes, respectively, but that were comparable to subconjunctival injection at 3, 6, and 10 hours. They concluded that collagen shield
deliver of gentamicin-dexamethasone might be comparable to subconjunctival injections and provide an alternative therapy after intraocular surgery.
Renard et al. (36) used collagen shields soaked in gentamicin and dexamethasone for patients after cataract surgery. No adverse eect was
reported.
Mahlberg et al. (37) studied the aggregate formation of tobramycin
sulfate in combination with methylprednisolone acetate or dexamethasone
sodium phosphate on collagen shields. Aggregates were formed on the surface of the shields when they were immersed in methylprednisolone acetate
and tobramycin. Dexamethasone sodium phosphate and tobramycin
resulted in a completely transparent shield. To avoid undesired side eects,
such as epithelial sloughing and corneal edema after collagen shield application, antibiotics and steroids must be carefully selected. Combinations of
gentamicin and methylprednisolone sodium succinate or gentamicin and
cefazolin on collagen shield also result in precipitates (38). Additionally,
care must be taken when mixing drugs to prevent adverse reactions. An
aminoglycoside such as gentamicin and a penicillin such as mezlocillin
should not be combined because of inactivation of the aminoglycoside by
the penicillin (39).

Copyright 2003 Marcel Dekker, Inc.

322

Higaki et al.

E. Antifungal Agents
Schwartz et al. (40) compared the delivery of amphotericin B in collagen
shields hydrated in a 0.5% drug solution with frequent topical drops
(0.15%) in uninfected rabbit eyes. Drops were applied every 5 minutes for
the rst half hour and at hourly intervals thereafter. The corneas and aqeous
humor were assessed at 1, 2, 3, and 6 hours following the initiation of drug
delivery. Drug levels in the shield-treated corneas were signicantly higher
than levels in the drop-treated corneas at 1 and 2 hours after therapy began.
At 3 hours, the concentrations of the antifungal drug in the corneal tissues
were similar for both delivery methods. At 6 hours, both groups had signicant concentrations of the antifungal agent, but the amount of the drug
was greater in the drop-treated corneas than in the shield-treated corneas.
Drug levels in the aqueous humor did not dier between the two groups at
any time. The results suggest that amphotericin B can be delivered to the
cornea via collagen shields at a rate that is comparable with frequent drop
deliver.
Pleyer et al. (41) evaluated the eect of collagen shields presoaked with
amphotericin B on the treatment of experimental Candida albicansinduced
keratitis. Treatment results were compared to those of amphotericin B eyedrops instilled hourly in rabbits. Treatment groups were (a) hourly instillation of 0.15% amphotericin B drops, (b) application of a collagen shield
presoaked in 0.5% amphotericin B for one hour, and (c) hourly instillation
of saline drops. Rabbit eyes treated with amphotericin Bsoaked collagen
shields had signicantly lower fungal counts compared with eyes receiving
hourly amphotericin B drops at days 1 and 3 after the beginning of treatment. They concluded that collagen shields soaked in amphotericin B could
be a useful and convenient treatment device in keratomycosis such as that
caused by Candida albicans.

F. Anticoagulant Therapy
Heparin, a large molecule with a molecular weight between 6000 and 20,000
daltons, has been studied experimentally as a possible agent for the reduction of postoperative brin formation after vitrectomy. The intravenous
route of administration, however, can be associated with increased postoperative hemorrhage. In an attempt to discover a vehicle that would permit
more localized drug delivery, Murray et al. (42) examined the pharmacokinetics and anticoagulation ecacy of heparin delivered by collagen shields.
Collagen shields hydrated with radiolabeled heparin were applied to rabbit
eyes, and the amount of heparin in aqueous humor, cornea, and iris was
measured at intervals from 15 minutes to 6 hours. The peak of radioactivity

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

323

was detected in the cornea and aqueous humor 1 hour after application.
Also in this study, 12-hour collagen shields hydrated with heparin were
compared to subconjunctival heparin injection. The highest biological activity was seen 30 minutes after application of the collagen shield, and there
was a signicant amount of anticoagulant activity in the aqueous humor 6
hours after application. At no time was any anticoagulant activity seen in
the aqueous humor following subconjunctival injection. The results of this
study demonstrate that a high molecular weight compound such as heparin
can be delivered by 12- or 24-hour collagen shields, producing signicant
levels in the aqueous humor. This suggests that collagen shields hydrated
with heparin might be eective in the prevention of treatment of brin
formation in the aqueous humor.
Murray et al. (43) also studied collagen shield delivery of tissue plasminogen activator (tPA) to the anterior segment and vitreous of rabbit eyes.
Time to complete lysis of anterior chamber brin clots in the eyes treated
with balanced salt solution-hydrated collagen shields was 149  18 hours.
Treatment with tPA-hydrated collagen shields shortened the mean time to
clot clearance by approximately 50% (clearance time 49  23 hours).
Elevated aqueous tPA levels were rst measured 18 hours after application
of tPA-hydrated collagen shields. Aqueous tPA levels peaked at 36 hours
and remained elevated throughout the 48-hour study period. Vitreous tPA
levels were elevated by 2 hours, peaked at 24 hours, and also remained
elevated throughout the study period. These results document the ecacy
and safety of tPA delivery to the aqueous and vitreous humor via a hydrated
collagen shield in the rabbit.
G.

Immunosuppressive Agents

Cyclosporine has been used successfully to prevent graft rejection in many

types of transplantation. The side eects of systemic administration, however, are considerable, and the systemic dose needed to provide sucient
drug in the cornea makes this route less than useful to prevent rejection of
corneal transplants. Also, the drug penetrates the cornea poorly, frustrating
the eorts at topical administration. Reidy et al. (44) demonstrated that
collagen shields with 4 mg of cyclosporin A incorporated during manufacture delivered signicantly higher concentrations of drug to the cornea and
aqueous humor than an equivalent amount of cyclosporin A prepared in
olive oil and given as drops at 15-minute intervals. Four hours after application of the collagen shield, the corneas contained almost 2500 ng of the
drug and the aqueous humor contained about 250 ng. At all times, the
concentration of the drug in the aqueous humor was higher in rabbits
receiving cyclosporin A via collagen shields compared to drops.

Copyright 2003 Marcel Dekker, Inc.

324

Higaki et al.

Kanpolat et al. (45) also investigated the penetration of cyclosporin A

into the rabbit cornea and aqueous humor after topical drop and collagen
shield administration. The rabbits were divided into three groups. The rst
group received 6 mg of cyclosporin A in castor oil and the second group
received 6 mg of cyclosporin A in olive oil applied as topical drops to rabbit
eyes within 12 hours. In the third group 12-hour collagen shields soaked in 6
mg of cyclosporin A in olive oil were applied. The cyclosporin A of castor
oil drops were higher than those obtained with olive oil drops. In eyes with
collagen shields, cyclosporin A levels were higher than olive oil drops but
nearly equal to the castor oil drops. In an extension of the studies, Chen et
al. (46) showed that cyclosporin Acontaining collagen shields suppress
corneal allograft rejection. Rabbit eyes with penetrating keratoplasty grafts
were placed in vascularized beds to enhance the possibility of graft rejection.
The grafts were treated with equivalent amounts of cyclosporin A in the
collagen shields or in olive oil drops. The mean survival time of shieldtreated grafts was signicantly longer than that of drop-treated grafts.
Grafts showing early signs of graft reaction treated with cyclosporincontaining shields showed reversal of the rejection process. The results of these
two studies indicate that the collagen shield is an eective delivery system
for cyclosporin A and that the drug delivered in this manner can both
suppress the initiation of graft rejection and reverse a graft reaction in
progress.

H.

Antiviral Agent

Gussler et al. (47) investigated the delivery of triuorothymidine (TFT) in

collagen shields and topical drops in normal rabbit corneas and corneas
with experimental epithelial defects. Collagen shields hydrated in 1% TFT
and 1% topical drops were used. The eyes were treated with either collagen
shields hydrated with TFT, TFT drops, or a combination of collagen shields
and drops. Rabbits with normal corneas showed no dierence among the
treatment groups in terms of TFT levels in the cornea or aqueous humor 30
minutes and 2, 4, and 8 hours after application of the antiviral. Among the
rabbits with experimental epithelial defects, the highest drug concentrations
were found in the eyes treated with the combination of shields and drops,
and the second highest tissue concentrations were seen in the eyes treated
with collagen shields hydrated in TFT. Treatment with drops alone produced lower concentrations of TFT than either treatment involving the
collagen shields. The authors suggested that this method of drug delivery
may be useful to enhance the eradication of herpes simplex virus in eyes with
epithelial defects. The authors noted, however, the need for studies of cor-

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

325

neal toxicity and ecacy in herpes-infected eyes before a denitive therapeutic regimen can be established.
I.

Wound Healing

After Fyodorov et al. (5) rst described the clinical use of collagen shields
for the protection and enhancement of epithelial healing, a number of
experimental studies were published conrming their ndings. Frantz et
al. (48) showed that rabbit eyes with 6 mm supercial keratectomies treated
with collagen shields healed signicantly faster than untreated eyes. Simsek
et al. (49) studied the eects of collagen shields and therapeutic contact
lenses on corneal wound healing in rabbits. A corneal wound was created
by mechanical removal of the central 6 mm zone of the corneal epithelium
and basement membrane. The healing rate was found to be 0:52  0:08
mm2 /hour with the collagen shield, 0:54  0:05 mm2 /hour with the therapeutic lens, and 0:43  0:06 mm2 /hour in the control group. Comparison of
the study groups revealed no statistically signicant dierence between the
collagen shield and the therapeutic lens group at any time, whereas a signicantly larger wound size was observed in the control group compared
with the treatment groups. In conclusion, these results indicated that both
collagen shields and therapeutic lenses enhance wound healing in rabbit
eyes.
On the other hand, the results of some studies suggested that more
evidence would be needed to establish eectiveness of collagen shields in
promoting wound healing (50,51). Shaker et al. (50) reported that cat eyes
treated with noncross-linked porcine collagen shields had a signicantly
greater healing response than untreated eyes. However, there was no dierence in the slope of the healing curve, suggesting that the shield did not
increase the speed of epithelial cell migration. Callizo et al. (51) suggested
that collagen shields might be ineective or have adverse eects in deeply
injured rabbit corneas. They performed bilateral keratectomies, then treated
only the left eye with a collagen shield. No dierences were seen in the time
course of the healing process between control (untreated) and treated eyes.
More polymorphonuclear inltration, mainly composed of eosinophils, was
shown in treated eyes. They concluded that the usefulness of collagen shields
should be reappraised, especially in injured corneas.
Additional studies combined the healing properties of collagen shields
with delivery of growth factors to inuence the rate of reepithelialization.
Hagenah et al. (52) used rabbit eyes with supercial keratectomies to examine the eect of collagen shields alone or in combination with epidermal
growth factor (EGF) or acidic broblastic growth factor on epithelial
wound healing. Eyes treated with collagen shields alone or collagen shields

Copyright 2003 Marcel Dekker, Inc.

326

Higaki et al.

hydrated with a broblastic growth factor healed signicantly faster than

untreated controls. Shields containing EGF had no enhanced eect.
Although it is apparent from this and other studies that the shields alone
promote healing, the utility of growth factors in and of themselves, including optimal dosage, timing, and duration of application, is still uncertain.
Therefore, even if the shields enhance the delivery of growth factors, it is not
clear at this time how such enhanced delivery can be used to improve
epithelial healing.
J. Other Uses
Wentworth et al. (53) determined the impact of collagen shields on ulceration of rabbit corneas after alkali burn. After a 60-second sodium hydroxide
burn to rabbit corneas, 24-hour collagen shields were replaced once daily for
21 days. They showed daily use of 24-hour collagen shields after a severe
alkali burn to the rabbit cornea exacerbates the progression of corneal
ulceration and perforation. They thought that the harmful eects of collagen shields might have resulted from the repeated removal and insertion of
the shields. The trapping of activated polymorphonuclear leukocytes in the
collagen shields that were in contact with the corneal surface caused a
locally high concentration of metalloproteinases.

IV.

CLINICAL STUDIES

The eect of collagen shields on healing and antibacterial or other chemotherapeutic eects in human eyes has been reported.
A.

Therapy to Assist Wound Healing

Aquavella et al. (54) reported the use of porcine collagen shields hydrated
with ophthalmic drugs to treat patients following penetrating keratoplasty
and cataract extraction. The drugs included tobramycin, gentamicin, pilocarpine, dexamethasone, and urbiprofen. No adverse eects were noted. In
another study by this group (55), collagen shields used as bandage lenses
appeared to accelerate corneal reepithelialization after keratoplasty or other
types of anterior segment surgery.
Poland and Kaufman (56) reported the use of collagen shields
hydrated with tobramycin in patients who had cataract extraction, penetrating keratoplasty, epikeratophakia, or nonsurgical epithelial healing
problems. All surgical patients showed more rapid healing of epithelial
defects. Acute nonsurgical epithelial problems with impaired healing also

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

327

beneted from the use of collagen shields. In contrast, chronic epithelial

defects responded poorly. No infections occurred in any of the patients.
Similarly, Groden and White (57) found that 24-hour porcine collagen
shields did not contribute to healing of persistent (chronic) epithelial defects
following penetrating keratoplasty. They dened a persistent defect as an
epithelial erosion (noninfectious) that did not heal in 2 weeks with patching,
frequent lubrication, and/or temporary tape tarsorrhaphy. Patients with
such defects were assigned to either collagen shield treatment or treatment
with a hydrophilic bandage soft contact lens. When none of the collagen
shieldtreated defects healed, this approach was abandoned and bandage
contact lens treatment instituted for all patients. The authors suggested that
a longer-lasting corneal shield (72 hours dissolution time) might be more
eective.
Marmer (58) described postsurgical healing in human eyes with porcine collagen shields after radial keratotomy. The patients who were treated
with the collagen shields reported less glare and discomfort than patients
who did not receive shields. Also, the eyes with shields showed less inammatory reaction and edema.
Palmer and McDonald (59) used a disposable soft contact lens piggybacked onto a medicated, 12-hour corneal collagen shield after corneal
surgery in three patients known to have poor corneal epithelial wound
healing characteristics. This piggyback lens/shield system delivered sustained high levels of medication and promoted epithelial healing in the
acute postoperative period without unnecessary postoperative orbital
manipulation.
In a dierent kind of healing study, Fourman and Wiley (60)
reported the results of treating a glaucoma lter bleb with a 24-hour
collagen shield hydrated in 4 mg/mL of gentamicin. The shield was placed
over the leaking site; the bleb leak was sealed within 2 days and remained
sealed 2 months later. The collagen shield was helpful in the management
of lter bleb leak.
B.

Antibacterial or Antiviral Efcacy

Lois and Molino (61) treated a case of Mycobacterium chelonae keratitis

with topical fortied amikacin, with no response. They then debrided the
epithelial and stromal lesions, and applied an amikacin-soaked corneal collagen shield, supplemented every 4 hours with topical fortied amikacin
drops. After this treatment, clinical and laboratory examinations showed
that no infectious organisms could be detected.
Kuster et al. (62) investigated the ability of collagen shields to deliver
TFT to human cornea and aqueous humor. Collagen shields were soaked

Copyright 2003 Marcel Dekker, Inc.

328

Higaki et al.

in commercially prepared TFT. Patients undergoing penetrating keratoplasty wore a presoaked collagen shield for at least 30 minutes preoperatively. Control patients received drops of TFT only. Cornea and aqueous
samples were obtained during surgery. Collagen shields did not enhance
delivery of TFT to the cornea with an intact epithelium. In corneas with
poor epithelium, drug penetration was higher but variable. They concluded
that the role of collagen shields as a drug delivery system for the treatment
of herpes simplex keratitis remains to be determined.

C.

Postoperative Application

Renard et al. (36) compared the eectiveness of subconjunctival injections

and collagen shields in delivering anti-inammatory agents and antibiotics
after cataract surgery. The occurrence of folds in Descemets membrane was
less frequent and aqueous are less severe in the collagen shieldtreated
group than in those treated with subconjunctival injections.
Haaskjold et al. (63) compared the ecacy of collagen shields with
that of peribulbar/retrobulbar injection after cataract surgery. Collagen
shields were saturated with an antibiotic and a steroid and placed over
the cornea postoperatively. The second group received the same drugs
through a peribulbar/retrobulbar injection. One day after surgery, the
shield group had signicantly less corneal edema, conjunctival hemorrhage, postoperative pain, and fewer corneal opacities. They suggested
that using collagen shields for drug delivery after cataract surgery
decreases tissue damage and increases patient comfort without adverse
side eects.
Taravella et al. (64) investigated whether collagen shields are more
eective than topical eye drops for infection prophylaxis after cataract
surgery. In their studies, the patients were divided into two groups: the
rst received three postoperative drops of commercially available topical
ooxacin (0.3%) given 10 minutes apart; the second had a collagen shield
soaked in the same medication applied to the eye before surgery. Aqueous
humor was extracted immediately before surgery for analysis. Aqueous
concentration of ooxacin in the shield group was signicantly higher
than that in the drop group. The MICs of ooxacin for many common
ocular pathogens were reached or exceeded in the shield group. However,
in a similar study using tobramycin, the ocular penetration of the antibiotic into the anterior chamber of the human eye in the shield group was
not dierent from that in the drop group (65). The aqueous concentration
did not approach the MIC of tobramycin for many common ocular pathogens.

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

D.

329

Dry Eye

Though collagen shields have a lubricating eect, they are not useful as a
treatment for keratoconjunctivitis sicca because they must be applied in the
physicians oce and they are not transparent. Kaufman et al. (66) tested a
similar system, with the addition of therapeutic agents during manufacturing (Collasomes), for drug delivery to the ocular surface. The size of
Collasomes was 1 mm  0.1 mm  2 mm. Ease of application and minimal
interference with vision were the advantages of this system.
E.

Clinical Uses for Corneal Collagen Shields

At the LSU Eye Center, our ocular ophthalmologists performing cataract

surgeries, corneal transplants, and other invasive corneal procedures use the
corneal collagen shields as a bandage. Most often they will use antibiotics
and anti-inammatories and hydrate the shields in the solution for 3 minutes. This is the most common use of the corneal collagen shields following
many surgical procedures. Since the eye is bandaged, there is no problem
involving visual acuity and often only one shield is used until the bandage is
removed (2448 hours) and the patients eye is evaluated.
Another use for the collagen shield is when an ophthalmologist has
decided to hospitalize a patient with suspected ocular infection and there is
concern about compliance for administering antibiotic drops. In the oce,
the ophthalmologist can hydrate the shield in a solution of fortied tobramycin or commercially available uoroquinolone and then apply the corneal
collagen shield. Patients are instructed to apply drops every 5 or 10 minutes
until they are admitted and are being administered antibiotics as a single
entity or in combinations in a hospital setting. The placing of the collagen
shield hydrated with fortied antibiotics assures that a reliable amount of
antibiotic will be in contact with the corneal surface over the next 24 hours.
The shield should be removed after 35 hours if antibiotics are not administered. It is important not to have any oxygen deprivation of a cornea suspected to have an infection.

V.

CONCLUSIONS

Collagen is a protein that can be safely applied to the body for a variety of
medical and cosmetic purposes. The creation of the corneal collagen shield
has provided a means to promote wound healing and, perhaps more importantly, to deliver a variety of medications to the cornea and other ocular
tissues. There are many indications that the shields deliver drugs as well as,

Copyright 2003 Marcel Dekker, Inc.

330

Higaki et al.

if not better than, topical drops. The simplicity of use and convenience
aorded by shields make them an attractive delivery device. Although collagen shields produce some discomfort and interfere with vision, corneal
collagen shields could become a commonly employed technological
improvement in ophthalmic drug delivery.

ACKNOWLEDGMENTS
The authors wish to acknowledge the editorial assistance and research collaboration of A. K. Mitra, Ph.D. Specic research from the laboratory of
Dr. Hill was supported in part by U.S. Public Health Service grant EY06311
and EY08871 (JMH); EY02377 (Eye Center Core Grant); an unrestricted
grant from Research to Prevent Blindness (RPB); Dr. Hill is an RPB
Scientic Investigator Award recipient. The authors wish to acknowledge
the secretarial assistance of Mrs. Carole Hoth. The authors also wish to
acknowledge special thanks to Ms. Kristina Braud and Ms. Kathy Vu.
None of the authors have any nancial or proprietary interest in any agents
or devices mentioned in this review.

REFERENCES
1. Kaufman, H. E. (1988). Guest editorial: Collagen shield symposium, J.
Cataract Refract. Surg., 14:487488.
2. Busin, M., and Spitznas, M. (1988). Sustained gentamicin release by presoaked medicated bandage contact lense, Ophthalmology, 95:796798.
3. Matoba, A. Y., and McCulley, J. P. (1985). The eect of therapeutic soft
contact lenses on antibiotic delivery to the cornea, Ophthalmology, 92:9799.
4. Bloomeld, S. E., Miyata, T., Dunn, M. W., Bueser, N., Stenzel, K. H., and
Rubin, A. L. (1978). Soluble gentacin ophthalmic inserts as a drug delivery
system, Arch. Ophthalmol., 96:885887.
5. Fyodorov, S. N., Moroz, Z. I., Kramskaya, Z. I., Bagrov, S. N.,
Amstislavskaya, T. S., and Zolotarevsky, A. V. (1985). Comprehensive conservative treatment of dystrophia endothelialis et epithelialis cornea, using a
therapeutic collagen coating, Vestn. Oftalmol., 101:3336.
6. Ivashina, A. I. (1987). Radial keratotomy as a method of surgical correction
of myopia. In: Microsurgery of the Eye: Main Aspects (S. N. Fyodorov, ed.),
Mir Publishers, Moscow, p. 45.
7. Chvapil, M., Kronenthal, R. L., and van Winkle, Jr., W. (1973). Medical and
surgical applications of collagen, Int. Rev. connect. Tissue Res., 6:1.

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

331

8. Kuwano, M., Horibe, Y., and Kawashima, Y. (1997). Eect of collagen crosslinking in collagen corneal shields on ocular drug delivery, J. Ocul. Pharmacol.
Ther., 13:3140.
9. Weissman, B. A., and Lee, D. A. (1988). Oxygen transmissibility, thickness,
and water content of three types of collagen shields, Arch. Ophthalmol.,
106:17061708.
10. Weissman, B. A., Brennan, N. A., Lee, D. A., and Fatt, I. (1990). Oxygen
permeability of collagen shields, Invest. Ophthalmol. Vis. Sci., 31:334338.
11. Hill, J. M., OCallaghan, R. J., Hobden, J. A., Kaufman, H. E. (1993).
Corneal collagen shields for ocular deliver. In: Ophthalmic Drug Delivery
Systems (A. K. Mitra, ed.), Marcel Dekker, Inc., New York, pp. 261273.
12. Artandi, C. (1964). Production experiences with radiation sterilization, Bull.
Parenteral Drug Assoc., 18:2.
13. Hill, J. M., OCallaghan, R. J., and Hobden, J. A. (1993). Ocular iontophoresis. In: Ophthalmic Drug Delivery Systems (A. K. Mitra, ed.), Marcel Dekker,
Inc., New York, pp. 331354.
14. Friedberg, M. L., Pleyer, U., and Mondiono, B. J. (1991). Device drug delivery to the eye. Collagen shields, iontophoresis, and pumps, Ophthalmology,
98:725732.
15. Reidy, J. J., Limberg, M., and Kaufman, H. E. (1990). Delivery of uorescein
to the anterior chamber using the corneal collagen shield, Ophthalmology,
97:12011203.
16. Myles, M. M., Loutsch, J. M., Higaki, S., and Hill, J. M. Ocular iontophoresis. In: Ophthalmic Drug Delivery Systems. (A. K. Mitra, ed.), Marcel
Dekker, Inc., New York, Chapter 12.
17. Zheng, X., Marquart, M. E., Mitra, A. K., and Hill, J. M. (2002). Unique
ocular drug delivery systems. In: Textbook of Ophthalmology (S. Agarwal,
et al., eds.), Vaybee Brothers Medical Publishers, Ltd., New Delhi, India,
pp. 20762086.
18. Unterman, S. R., Rootman, D. S., Hill, J. M., Parelman, J. J., Thompson, H.
W., and Kaufman, H. E. (1988). Collagen shield drug delivery: Therapeutic
concentrations of tobramycin in the rabbit cornea and aqueous humor, J.
Cataract Refract. Surg., 14:500504.
19. OBrien, T. P., Sawusch, M. R., Dick, J. D., Hamburg, T. R., and Gottsch, J.
D. (1988). Use of collagen corneal shields versus soft contact lenses to enhance
penetration of topical tobramycin, J. Cataract Refract. Surg., 14:505507.
20. Chen, C. C., Taarui, H., and Duzman, E. (1993). Enhancement of the ocular
bipability of topical tobramycin with use of a collagen shield, J. Cataract
Refract. Surg., 19:242245.
21. Hobden, J. A. Reidy, J. J., OCallaghan, R. J., Hill, J. M., Insler, M. S., and
Rootman, D. S. (1988). Treatment of experimental Pseudomonas keratitis
using collagen shields containing tobramycin, Arch. Ophthalmol.,
106:16051607.
22. Sawusch, M. R., OBrien, T. P., Dick, J. D., and Gottsch, J. D. (1988).
Quinolones in collagen shields to treat aminoglycoside-resistant
Pseudomonas keratitis, Invest. Ophthalmol. Vis. Sci., 31:2241.

Copyright 2003 Marcel Dekker, Inc.

332

Higaki et al.

23. Clinch, T. E., Hobden, J. A., Hill, J. M., OCallaghan, R. J., Engel, L. S., and
Kaufmann H. E. (1992). Collagen shields containing tobramycin for sustained
therapy (24 hours) of experimental Pseudomonas keratitis, CLAO J., 18:245
247.
24. Assil, K. K., Zarnegar, S. R., Fouraker, B. D., and Schanzlin, D. J. (1992).
Ecacy of tobramycin-soaked collagen shields vs tobramycin eyedrop loading
dose for sustained treatment of experimental Pseudomonas aeruginosa-induced
keratitis in rabbits, Am. J. Ophthalmol., 113:418423.
25. Silbiger, J., and Stern, G. A. (1992). Evaluation of corneal collagen shields as
a drug delivery device for the treatment of experimental pseudomonas keratitis,
Ophthalmology, 99:889892.
26. Liang, F. Q., Viola, R. S., del Cerro, M., and Aquavella, J. V. (1992).
Noncross-linked collagen discs and cross-linked collagen shields in the delivery of gentamicin to rabbits eyes, Invest. Ophthalmol. Vis. Sci., 33:21942198.
27. Callegan, M. C., Engel, L. S., Clinch, T. E., Hill, J. M., Kaufman, H. E., and
OCallaghan, R. J. (1994). Ecacy of tobramycin drops applied to collagen
shields for experimental staphylococcal keratitis, Curr. Eye Res., 13:875878.
28. Dorigo, M. T., De Natale, R., and Miglioli, P. A. (1995). Collagen shields
delivery of netilmicin: a study of ocular pharmacokinetics, Chemotherapy,
41:1.
29. Baziuk, N., Gremillion, C. M., Peyman, G. A., and Cho, H. K. (1992).
Collagen shields and intraocular drug delivery: concentration of gentamicin
in the aqueous and vitreous of a rabbit eye after lensectomy and vitrectomy,
Int. Ophthalmol., 16:101107.
30. Hobden, J. A., Reidy, J. J., OCallaghan, R. J., Insler, M. S., and Hill, J. M.
(1990). Quinolones in collagen shields to treat aminoglycoside-resistant pseudomonal keratitis, Invest. Ophthalmol. Vis. Sci., 31:22412243.
31. Phinney, R. B., Schwartz, S. D., Lee, D. A., and Mondino, B. J. (1988).
Collagen-shield delivery of gentamicin and vancomycin, Arch. Ophthalmol.,
106:15591604.
32. Callegan, M. C., OCallaghan, R. J., and Hill, J. M. (1994). Pharmacokinetic
considerations in the treatment of bacterial keratitis, Clin. Pharmacokin.,
27:129.
33. Hwang, D. G., Stern, W. H., Hwang, P. H., and MacGowan-smith, L. A.
(1989). Collagen shield enhancement of topical dexamethasone penetration,
Arch. Ophthalmol., 107:13751380.
34. Sawusch, M. R., OBrien, T. P., and Updegra, S. A. (1989). Collagen corneal
shields enhance penetration of topical prednisolone acetate, J. Cataract
Refract. Surg., 15:625628.
35. Milani, J. K., Verbukh, I., Pleyer, U., Sumner, H., Adamu, S. A., Halabi, H.
P., Chou, H. J., Lee, D. A., and Mondino, B. J. (1993). Collagen shields
impregnated with gentamicin-dexamethasone as a potential drug delivery
device, Am. J. Ophthalmol., 116:622627.

Copyright 2003 Marcel Dekker, Inc.

Corneal Collagen Shields for Ocular Drug Delivery

333

36. Renard, G., Bennani, N., Lutaj, P., Richard, C., and Trinquand, C. (1993).
Comparative study of a collagen corneal shield and a subconjunctival injection at the end of cataract surgery, J. Cataract Refract. Surg., 19:4851.
37. Mahlberg, K., Krootila, K., and Uusitalo, R. (1997). Compatibility of corticosteroids and antibiotics in combination, J. Cataract Refract. Surg., 23:878
882.
38. Phinney, R. B., Schwartz, S. D., Lee, D. A., and Mondino, B. J. (1988).
Collagen-shield delivery of gentamicin and vancomycin, Arch. Ophthalmol.,
106:15991604.
39. Mondino, B. J. (1991). Collagen shields, Am. J. Ophthalmol., 112:587590.
40. Schwartz, S. D., Harrison, S. A., Engstrom, R. E., Bawdon, R. E., Lee, D. A.,
and Mondino, B. J. (1990). Collagen shield delivery of amphotericin B, Am. J.
Ophthalmol., 109:701704.
41. Pleyer, U., Legmann, A., Mondino, B. J., and Lee, D. A. (1992). Use of
collagen shields containing amphotericin B in the treatment of experimental
Candida albicans-induced keratomycosis in rabbits, Am. J. Ophthalmol.,
113:303307.
42. Murray, T. G., Stern, W. H., Chin, D. H., and MacGowan-Smith, E. A.
(1990). Collagen shield heparin delivery for prevention of postoperative brin,
Arch. Ophthalmol., 108:104106.
43. Murray, T. G., Jae, G. J., McKay, B. S., Han, D. P., Burke, J. M., and
Abrams, G. W. (1992). Collagen shield delivery of tissue plasminogen activator: functional and pharmacokinetic studies of anterior segment delivery,
Refract. Corneal Surg., 8:4448.
44. Reidy, J. J., Gebhardt, B. M., and Kaufman, H. E. (1990). The collagen
shield: A new vehicle for delivery of cyclosporin A to the eye, Cornea,
9:196199.
45. Kanpolat, A., Batioglu, F., Yilmaz, M., and Akbas, F. (1994). Penetration of
cyclosporin A into the rabbit cornea and aqueous humor after topical drop
and collagen shield administration, CLAO J., 20:119122.
46. Chen, Y. F., Gebhardt, B. M., Reidy, J. J., and Kaufman, H. E. (1990)>
Cyclosporin-containing collagen shields suppress corneal allograft rejection,
Am. J. Ophthalmol., 109:132137.
47. Gussler, J. R., Ashton, P., VanMeter, W. S., and Smith, T. J. (1990). Collagen
shield delivery of triuorothymidine, J. Cataract Refract. Surg., 16:719722.
48. Frantz, J. M., Dupuy, B. M., Kaufman, H. E., and Beuerman, R. W. (1989).
The eect of collagen shields on epithelial wound healing in rabbits, Am. J.
Ophthalmol., 108:524528.
49. Simsek, N. A., Ay, G. M., Tugal-Tutkun, I., Basar, D., and Bilgin, L. K.
(1996). An experimental study on the eect of collagen shields and therapeutic
contact lenses on corneal wound healing, Cornea, 15:612616.
50. Shaker, G. J., Ueda, S., LoCascio, J., and Aquavella, J. V. (1989). Eect of a
collagen shield on cat corneal epithelial wound healing, Invest. Ophthalmol.
Vis. Sci., 30:15651568.

Copyright 2003 Marcel Dekker, Inc.

334

Higaki et al.

51. Callizo, J., Cervello, I., Mayano, E., and Mallol, J. (1996). Inecacy of collagen shields in the rabbit corneal wound-healing process, Cornea, 15:258
262.
52. Hagenah, M., Lopez, J. G., and Insler, M. S. (1990). Eect of EGF, FGF, and
collagen shields on corneal epithelial wound healing following lamellar keratectomy, Invest. Ophthalmol. Vis. Sci. Suppl., 31:225.
53. Wentworth, J. S., Paterson, C. A., Wells, J. T., Tilki, N., Gray, R. S., and
McCartney, M. D. (1993) Collagen shields exacerbate ulceration of alkaliburned rabbit corneas, Arch. Ophthalmol., 111:389392.
54. Aquavella, J. V., Runi, J. J., and LoCascio, J. A. (1988). Use of collagen
shields as a surgical adjunct, J. Cataract Refract. Surg., 14:492495.
55. Aquavella, J. V., Musco, P. S., Ueda, S., and LoCascio, J. A. (1988).
Therapeutic applications of collagen bandage lens: A preliminary report,
CLAO J., 14:4750.
56. Poland, D. E., and Kaufman, H. E. (1990). Clinical uses of collagen shields, J.
Cataract Refract. Surg., 14:489491.
57. Groden, L. R., and White, W. (1990). Porcine collagen corneal shield treatment of persistent epithelial defects following penetrating keratoplasty, CLAO
J., 16:9597.
58. Marmer, R. H. (1988). Therapeutic and protective properties of the corneal
collagen shield, J. Cataract Refract. Surg., 14:496499.
59. Palmer, R. M., and McDonald, M. B. (1995). A corneal lens/shield system to
promote postoperative corneal epithelial healing, J. Cataracat Refract. Surg.,
21:125.
60. Fourman, S., and Wiley, L. (1989). Use of a collagen shield to treat a glaucoma lter bleb leak, Am. J. Ophthalmol., 107:673674.
61. Lois, N., and Molino, M. L. P. (1994). Mycobacterium chelonae keratitis:
resolution after debridement and presoaked collagen shields, Cornea,
14:536539.
62. Kuster, P., Taravella, M., Gelinas, M., and Stepp, P. (1998). Delivery of
triuridine to human cornea and aqueous using collagen shields, CLAO J.,
24:122124.
63. Haaskjold, E., Ohrstrom, A., Uusitalo, R. J., Krootila, K., Sandvig, K. U.,
Sonne, H., and Mahlberg, K. (1994). Use of collagen shields in cataract surgery, J. Cataract Refract. Surg., 20:150153.
64. Taravella, M. J., Balentine, J., Young, D. A., and Stepp, P. (1999). Collagen
shield delivery of ooxacin to the human eye, J. Cataract Refract. Surg.,
25:562565.
65. Taravella, M. J., Stepp, P., and Young, D. A. (1998). Collagen shield delivery
of tobramycin to the human eye, CLAO J., 24:166168.
66. Kaufman, H. E., Steinemann, T. L., Lehman, E., Thompson, H. W., Varnell,
E. D., Jacob-Labarre J. T., and Gebhardt, B. M. (1994). Collagen-based drug
delivery and articial tears, J. Ocul. Pharmacol., 10:1727.

Copyright 2003 Marcel Dekker, Inc.

11
The Noncorneal Route in Ocular
Drug Delivery
Imran Ahmed
Pzer Global R&DGroton Laboratories, Groton, Connecticut, U.S.A.

I.

INTRODUCTION

It is now generally accepted that topically applied drugs can enter the eye by
both a corneal and a noncorneal pathway. Although the corneal pathway is
the primary route of intraocular entry for most drugs, penetration through
the conjunctiva and sclera can also be important (13) and, under some
circumstances, could be more important than the corneal route (422).
Ahmed and Patton (5,6), using blocking techniques, showed that the noncorneal pathway was the dominant route for intraocular entry of inulin, a
molecule that is poorly absorbed across the cornea. This group was also the
rst to suggest that it might be possible to exploit the conjunctival/scleral
absorption route to promote site-specic delivery of drugs to intraocular
tissues in the back of the eye (23). The increasing interest in intraocular
delivery of peptide and protein drugs, drug therapy targeting posterior segment eye disease, and the advent of new chemical entities and novel ophthalmic drug delivery systems necessitates a thorough understanding of the
noncorneal pathway and approaches of exploiting this route in ocular therapy.
Recently there has been considerable progress in eorts to characterize
the permeability properties and permselectivity of the conjunctiva and the
sclera. A greater mechanistic understanding of drug transport in the ocular
epithelia has further enhanced this research. It has been established that
compared with the cornea, the conjunctiva is generally more permeable to
drugs (24). This is primarily due to a leakier epithelium in the conjunctiva
335

Copyright 2003 Marcel Dekker, Inc.

336

Ahmed

(2532). Furthermore, the conjunctiva plays an active role in drug transport

facilitated by the existence of a variety of drug transporters and ion transport processes (3337). There has been extensive work on elucidating the
diusional barrier properties of the sclera (3847) and conjunctiva (4854).
Systemic loss of drug via absorption into the ocular blood vessels is believed
to signicantly diminish the contribution of the scleral/conjunctival route in
ocular drug delivery. Factors aecting the systemic absorption of ocularly
applied drugs (5565) have been studied, as well as approaches to minimize
non-productive loss (6674) and the possibility of using the eye as a portal
for systemic drug delivery (7585). Several excellent reviews discussing the
pharmacokinetic and pharmacodynamic considerations in ocular drug
delivery have been published over the years (8694).
This chapter reviews the recent progress in the understanding of the
noncorneal penetration route towards creating a framework for rational
design of drug delivery systems for therapeutic agents that are poorly
absorbed across the cornea or require delivery to posterior segment of the
eye. Drug delivery across the conjunctiva and sclera is examined with reference to (a) physicochemical drug properties that may make the noncorneal
route signicant, (b) approaches of optimizing this route by understanding
the mechanism by which this occurs, and (c) opportunities for taking advantage of this route in the design and evaluation of ocular drug delivery
systems.

II.

OCULAR ANATOMY AND PHYSIOLOGY RELEVANT TO

THE NONCORNEAL ROUTE

The ocular anatomy and physiology relevant to ocular drug delivery was
reviewed by Robinson (95). Some of the main points relevant to noncorneal
drug delivery are summarized below.

A.

Noncorneal Drug Penetration and Tissue Distribution

The shape of the eye in humans and rabbits, the primary animal model used
in ocular research, approximates that of a globe, as shown in Figure 1. From
the perspective of ocular drug delivery, the eye can be regarded as consisting
of two parts: an anterior and a posterior segment. The tear uid, the cornea,
anterior chamber lled with aqueous humor, iris-ciliary body, and the lens
comprise the anterior segment. The posterior segment of the eye consists of

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

337

Figure 1 Anatomy of the eye. (Adapted from Ref. 154.)

four structuresthe conjunctiva, sclera, choroid, and retinasurrounding

the vitreous cavity that contains the vitreous humor. Taking into consideration the geometry of the eyeball and the diusional pathways, drug penetrating via the corneal route has direct access to the anterior segment tissues,
providing high levels in the cornea, aqueous humor, and the iris-ciliary
body. In contrast, drug entering via the noncorneal route traverses the
conjunctiva and sclera entering the choroid, retina, and eventually the vitreous humor. This selective tissue distribution of drugs entering the eye via
the noncorneal route may be promising for drug delivery targeting the
posterior eye (23,96).

Copyright 2003 Marcel Dekker, Inc.

338

B.

Ahmed

Barriers to Noncorneal Drug Penetration

Intraocular entry via the noncorneal route requires drug penetration across
the conjunctiva and the sclera. There are are three factors that can be
barriers to the noncorneal penetration of ocularly applied drugs: (a) drug
removal from the precorneal areas due to lacrimation and tear drainage, (b)
the barriers to drug diusion oered by the structures making up the outer
coat of the eye, and (c) drug loss to the systemic circulation via the ocular
vasculature.
1. Precorneal Fluid Dynamics
Topically applied drugs are commonly administered as an eye drop formulated as a solution, suspension, gel, ointment, and occasionally as a solid
insert (8694). The rst barrier to intraocular penetration of topically
applied drugs is tear turnover in the precorneal area. Under normal, unanesthetized conditions, the human tear volume averages 7 mL, with the
estimated maximum volume that the cul-de-sac can momentarily contain
with eye drop administration at about 30 mL (97). The human tear lm is a
lightly buered aqueous uid with a pH of approximately 7.27.5 and with
an estimated thickness of 49 mm (98,99). The average tear turnover rate in
humans is about 16% per minute under basal conditions but may be
increased to 30% per minute due to stimulation resulting from drop instillation. The restoration of normal tear volume in the human requires an estimated 23 minutes, with 80% or more of the administered eye drops lost to
drainage in the rst 1530 seconds after instillation (98). The resulting short
contact times of drugs with absorbing membranes of the eye is the primary
reason that typically less than 5% of a topically applied drug reaches the
intraocular tissues (8789).
2. Diffusion Across Ocular Membranes
a. Conjunctiva The conjunctiva is a thin, transparent mucous membrane that starts at the corneoscleral junction (limbus) and extends to the
eyelid margin. The portion of the conjunctiva loosely attached to the anterior surface of the globe is referred to as the bulbar conjunctiva. The
more rmly adhering segment lining the inside of the eyelids is called the
tarsal or palpebral conjunctiva. The conjunctiva can be divided into three
layers: (a) an outer epithelium, forming a permeability barrier, (b) the
substantia propria, containing structural and cellular elements, nerves,
lymphatics and blood vessels, and (c) the submucosa, providing a loose
attachment to the underlying sclera. The conjunctival epithelium is a stratied epithelium, squamous at the lids and columnar towards the cornea.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

339

It is nonkeratized and has tight junctions that can present a permeability

barrier to the diffusion of drugs. The thickness of the conjunctiva varies
from region to region, being 1015 layers thick towards the cornea and 5
6 layers thick at the eyelids. The conjunctival epithelium possesses dense
microvilli covered with glycoclayx and a mucus layer (100102). The area
of the conjunctival sac in humans has been estimated at 16 cm2 , of which
the cornea constitutes about 10%, whereas that in rabbits is approximately 1213 cm2 , of which the cornea accounts for about 20% (103,104).
Therefore, the conjunctival surface area exceeds that of the cornea by over
fourfold. Unlike the cornea the conjunctiva is highly vascularized. Systemic loss can signicantly reduce the fraction of drug available for penetration via the scleral/conjunctival route.
b. The Sclera The sclera, which is the white, tough outer coat of
the eye, is composed largely of connective tissue (105). It has a protective
function and maintains the shape of the eyeball by resisting intraocular
pressure (106). The sclera is continuous with the cornea and extends posteriorly from the limbus, representing nearly 80% of the total surface area
of the globe. The relatively softer, outer layer of the sclera is called the
episclera. In rabbits, the sclera adjacent to the limbus is about 0.5 mm
thick, thinning to as little as 0.2 mm in some areas (107). Structurally, the
sclera is very similar to the corneal stroma and is made up of primarily
collagen and mucopolysaccharides (108). Numerous channels through
which uid drainage occurs perforate the sclera. Blood vessels enter the
uvea and the retina, but the sclera is itself poorly vascularized (109).
3.

Systemic Loss via the Ocular Vasculature

Although the blood-ocular barrier prevents most systemically administered

drugs from penetrating into the eye (65), ocularly applied drugs can enter
the systemic circulation with relative ease. In excess of 70% of the instilled
dose of an eye drop can enter the systemic circulation via absorption into
the vasculature of the conjunctival and nasal mucosae (6265). The nasal
route is believed to be the primary contributor to the systemic loss of eye
drops. For example,the nasal mucosa was 2.5 times more ecient than the
conjunctival mucosa in contributing the systemic level of timolol (62).
Whereas tight junctions between the endothelial cells of the microvessels
result in a very low permeability to most solutes in the retina, the capillaries
in the choroid and the ciliary processes there are fenestrated and the permeability to low molecular weight compounds is high. By injecting labeled
proteins intravenously in rabbits and maintaining a steady-state concentration, Bill et al. (59) showed that the permeability of proteins in the choroid
and the ciliary body were high compared to other ocular tissues.

Copyright 2003 Marcel Dekker, Inc.

340

Ahmed

Table 1

Blood Flow in Ocular Tissues of Primates

Tissue

Blood ow in whole
tissue (mean  S.E.)
(mg/min)

Retina
Iris
Ciliary body
Ciliary processes
Ciliary muscle

34  2a N 15
8  1 N 15
81  6 N 15
677  67 N 15

Blood ow (mean  S.E.)

(g/min/tissue)
56  8 N 15
150  12 N 13
227  17 N 8
163  26 N 8

N = Number of determinations.
a
Standard Error of Mean.
Source: Ref. 59.

Radiolabeled microspheres and indicator-dilution techniques have been

used to measure ocular hemodynamics (5557). The blood ow rate through
various parts of the eye in primates is reported in Table 1. The kinetic
constant of drug transfer from the eye to the circulation usually has a
value of 2050 h1 from conventional dosage forms (64). Systemic loss
via the ocular vasculature is a major deterrent to noncorneal drug delivery
and must be factored into calculating the fraction of drug available for
intraocular absorption. Minimizing the systemic loss there can also help
to reduce adverse systemic eect following ocular application of potent
compounds (6675).

III.

NONCORNEAL ROUTES IN OCULAR DRUG DELIVERY

Ahmed and Patton (5) proposed a schematic for the pathways for the
intraocular penetration of topically applied drugs, as shown in Figure 2.
Modeling of the intraocular penetration routes and implication on ocular
pharmacokinetics and pharmacodynamics was recently reviewed by
Worakul and Robinson (14).
Although several investigators had reported a minor route of intraocular drug entry via the sclera and the conjunctiva (13), Bito and Baroody
were the rst to present evidence that this noncorneal route may, at least
under some circumstances, be more important than the corneal route (4).
They showed that after topical 3H-PGF2a application, the choroid, anterior
sclera, and the ciliary body contained higher drug concentrations than the
aqueous, indicating that the drug was entering the eye by some route other
than through the cornea.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

341

Figure 2 Intraocular penetration routes.

Over the past two decades there have been several investigations to
further examine the conjunctival/scleral pathway for intraocular entry of
drugs. The preferred method has been to mechanically block the cornea
from the conjunctiva and sclera in situ with the use of a cylindrical well
and introducing a drug solution either inside or outside the well. Hence, the
disappearance of drug from the reservoir as well as the appearance of drug
in intraocular tissues can be determined to compare the rate and extent of
drug penetration via the corneal versus the conjunctival/scleral pathway.
This method was employed by Schoenwald et al. (15) to study the ocular
penetration pathway for methazolamide analogs, 6-carboxyuorescein and
rhodamine. The conjunctival/scleral route of entry produced higher iris/
ciliary body concentrations for all compounds except for the lipophilic rhodamine. Confocal microscopy results suggested that drug gained entry into
the ciliary body through uptake into the blood vessels of the sclera. The
clinical implication of the scleral/conjunctival pathway may be important
for antiglaucoma drugs where a quicker route to the iris-ciliary body via the
blood vessels may result in a faster onset of action. Sasaki et al. (11) used the
in situ technique to show that while the b-blocker tilisolol entered the
aqueous humor primarily via the corneal route, the access to the vitreous

Copyright 2003 Marcel Dekker, Inc.

342

Ahmed

body was four times more eective through the sclera than through the
cornea. The application of tilisolol in the conjunctiva or the sclera also
showed a high concentration in plasma whereas corneal application produced no systemic levels.
The physicochemical drug properties important to noncorneal penetration of topically applied drugs appear to be lipophilicity and molecular
size. Using a series of b-blockers Sasaki et al. (16) showed that the permeability of penetrants is strongly dependent on lipophilicity for the cornea but
less so for the conjunctiva and the sclera (Fig. 3). Chien et al. (10) studied a2 adrenergic agents of varying lipophilicity and observed that the conjunctival/scleral pathway was the predominant route for delivery of least lipophilic molecule, p-aminoclonidine. The investigators also reported evidence of
lateral diusion of drug from the conjunctiva to the cornea. Pech et al. (18)
evaluated a series of amphiphilic timolol prodrugs and observed that the
transcleral absorption was the highest with the longest aliphatic chain prodrugs, which also had the most amphiphilic/lipophilic character. Hence, the
in vitro studies suggest that solute lipophilicity is less important for noncorneal drug penetration than it is for transcorneal drug penetration.
However, the eect of lipophilicity on the extent of noncorneal penetration

Figure 3 Relationship between logarithmic values of the octanol/water partition

coecient (PC) and permeability coecient (Kp ). (0) Cornea; (
) conjunctiva; (&)
sclera; (*) scraped cornea.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

343

of topically applied drugs in vivo may be dicult to predict from in vitro

studies alone. This requires information on the permeability of the drug
across the conjunctiva, sclera, and ocular blood vessels, as well as drug
binding to ocular tissues. A suitable predictive model that accounts for all
these factors is not yet available.
There is also strong evidence that the noncorneal route may be the
preferred pathway for intraocular entry of large, polar molecules that have
poor corneal permeability. Ahmed and Patton (5,6) used corneal blocking
techniques to demonstrate that the noncorneal pathway was the primary
route of intraocular entry for inulin, a molecule that was poorly absorbed
across the cornea (Table 2). The permeability of large molecules in conjunctiva and sclera is typically higher than in the cornea, suggesting that the
noncorneal route may contribute more to the intraocular absorption of
large molecules than the corneal route. The permeability of the sclera and
conjunctiva will be discussed later in the chapter.

Table 2 Concentration of Inulin in Various Ocular Tissues 20 Minutes

Following the Topical Instillation of 25 mL of a 0.65% Inulin Solution, in the
Presence and Absence of Corneal Access
Concentration (mg/g)

Aqueous humor
Cornea
Lens
Vitreous humor
Iris-ciliary body
Sclera
Bulbar
conjunctiva

With corneal
access

Without corneal
access

2.10a
(0.420,9)
22.8
(4.00,9)
ND
0.03
(0.008,9)
0.79
(0.162,9)
7.82
(2.354,6)
177
(34.9,6)

0.03
(0.020,6)
1.87
(0.210,9)
ND
0.02
(0.003,9)
0.63
(0.087,9)
8.45
(2.280,6)
273
(24.9,6)

Percent
(with/without)  100
1.4
8

67
80
108
154

The mean; standard error of the mean and the number of eyes in parentheses.

ND = Not detectable.
Source: Ref. 5.

Copyright 2003 Marcel Dekker, Inc.

344

Ahmed

Systemic loss via drug absorption into the ocular blood vessels of the
conjunctiva is a nonproductive pathway that diminishes the fraction of
drug available for intraocular penetration via the noncorneal route.
Accordingly, delivery methods that maximize the drug concentration at
the conjunctival surface and minimize nonproductive systemic loss are
also expected to improve noncorneal drug penetration. This hypothesis
has been supported by some recent studies. Urtti and coworkers (7) presented evidence of application sitedependent noncorneal entry of timolol
in rabbit from a topical device that released timolol at 7.2 mg/h. When the
device was placed in the inferior conjunctival sac, the resulting timolol
concentration in the aqueous humor was nearly 100-fold lower than in
the iris-ciliary body. Romanelli et al. (20) showed that bendazac was
absorbed into the retina-choroid via the scleral/conjunctival route when
delivered topically in polysaccharide vehicles. It was noted that transcorneal penetration of bendazac was hindered not only by the epithelial barrier but also by the strong binding of the drug to the stroma. In another
study, Lehr et al. (21) reported formulating gentamicin in the mucoadhesive polymer, polycarbophil, facilitated the noncorneal penetration of gentamicin probably by intensied contact between the polymer and the
underlying bulbar conjunctiva. Dosage design considerations for noncorneal delivery will be discussed in a subsequent section.

IV.

PERMEABILITY OF THE CONJUNCTIVA AND THE

SCLERA

The recognition that noncorneal penetration may be a productive route of

intraocular entry for some drugs has triggered extensive research to understand the barrier properties of the conjunctiva and the sclera and the
physicochemical determinants of drug permeability across these membranes.
In vitro and in situ permeability studies have proven to be a useful approach
in estimating drug transport for in vivo conditions. In general, permeability
characteristics of ocular membranes correlate well with the intraocular
absorption of drugs.

A.

Methodology

Four methods have been commonly used for measuring permeability across
ocular membranes. The rst method is based on the measurement of steadystate permeability of drugs across the isolated ocular membrane using a
side-by-side diusion cell or a modied two-chamber Ussing apparatus.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

345

Accordingly, the membrane permeability coecients may be calculated using the following equation:
P VR
C=
t:A:CD
where
C=
t is the change in the concentration with time represented by
the slope of the linear part of the amount of drug accumulated versus time
curve, VR is the volume of the receiver chamber, A is the surface area of the
mounted membrane, and CD is the initial concentration of the drug in the
donor chamber. This method assumes sink conditions. As pointed out by
Maurice (32) in his critique of such techniques, it is important to ensure that
the in vivo characteristics of the membranes are maintained as closely as
possible. This requires care during the dissection to avoid folding or damage
to the surfaces of the membranes and preservation of the tissue during the
course of the experiment using physiologically relevant buer systems (e.g.,
oxygenated glutathione bicarbonate Ringers solution) to maintain tissue
viability. It is also important to measure, conrm, and report the integrity
of the barrier based on electrical properties, hydration level, and diusion of
marker substances. Another alert is interspecies dierences and caution in
overinterpreting data obtained from animals to humans.
The second method is in situ perfusion technique that enables quantitation of drug uptake in live animals. This method was originally utilized for
measuring corneal permeation and uptake of drugs (110,111) and subsequently applied to study the conjunctival and scleral uptake of various
compounds (10,11,16). The technique involves axing a cylindrical well
on the surface of the eye near the corneoscleral junction using surgical
adhesives (Fig. 4). A drug solution is then placed either inside the well
bathing the cornea or outside the well bathing the remainder of the conjunctival sac. The rate of mass ux in the system can be described by:
dCs Vs =dt Cls :Cs
where Cs is the concentration of drug in the systeml, Vs is the volume of uid
in the reservoir and Cls is a clearance parameter describing the loss of drug
from the system. This method can be utilized to study the eect of physicalchemical drug properties while avoiding some of the problems associated
with in vitro studies on isolated membranes as previously described. The
primary disadvantage is the diculty in extracting a true permeability coefcient or mechanistic information from such experiments.
A third is a ow-through permeation chamber where the excised tissue
is mounted horizontally method (43,46,112). A small volume of the drug
solution is applied on the external surface of the membrane and the internal
surface is perfused with a physiological receptor solution. In this method the
ux across the membrane can be calculated using the following equation:

Copyright 2003 Marcel Dekker, Inc.

346

Ahmed

Figure 4

In situ corneal blocking technique.

V@C=@t JS  QC
where V is the volume of the receiver compartment, C is the concentration
in the receiver compartment, S is the exposed surface area, J is the ux at
the membrane surface on the receptor side, Q is the ow rate out of the
receptor compartment, and t is the time. This method is appropriate for the
measurement and prediction of transient transport across the ocular membrane simulating more realistic ocular drug delivery scenarios. For example,
it has been shown that the predictions of transscleral transport based on
steady-state measurements may signicantly overpredict the amount of drug
delivered into the eye because the lag time for transport across the sclera is
similar or longer than the drug-sclera contact time from an eye drop (43).
Drug binding to the sclera may also prolong the lag time. A disadvantage of
nonsteady-state experiments is that the data treatment and mathematics is
complicated.
The fourth method is the use of cell culture and physical models.
During the past decade advances in cell culture techniques have resulted
in the development of a primary culture model of rabbit conjunctival epithe-

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

347

lial cells exhibiting tight barrier properties (112115). There has been an
attempt to develop physical models to describe the transport of molecules
through the cornea and the sclera by taking into account the ultrastructure
of these tissues (44). The use of intestinal tissues to predict ocular permeability and ex vivo models based on isolated perfused tissue has also been
reported (118). These techniques may oer an alternative to the use of
animals in research and an opportunity for gaining a greater mechanistic
understanding of transport processes. However, the practical application
and scope of these methods remain to be determined.

B.

Permeability of the Sclera and Conjunctiva

1.

Sclera

The permeability of the sclera to small molecules is comparable to that of

the corneal stroma. There was no signicant dierence between the corneal
stroma and the sclera of the beef eye in their permeability to a variety of
small solute (24). Even large molecules can move across the sclera by the
way of perivascular spaces but also by diusion through interbrillary
spaces (38). Maurice and Polgar measured the movement of a number of
anionic and cationic dyes, as well as small ions, proteins, and biologically
active molecules across the isolated beef sclera using a two-chamber diusion cell (39). Only anionic dyes of small molecular weights were able to
cross the sclera, but the sclera oered little obstacle to the penetration of
drugs into the eye.
Prausnitz and Noonan (24) recently published a comprehensive database of ocular tissue permeability measurements found in a review of the
literature. There was no apparent dependence on distribution coecient but
a strong dependence of penetrant size on scleral permeability. The scleral
permeability to tracers of various molecular sizes in rabbit and humans is
shown in Table 3. Ambati et al. (47) showed that molecular radius was a
better predictor of scleral permeability than molecular weight, and large
molecules, such as IgG, diuse across the sclera in a manner consistent
with porous diusion through a ber matrix. Edwards and Prausnitz supported this assertion with a ber matrix model to predict the permeability of
the sclera to water and solute (44). Unlu and Robinson (45) observed that
the scleral permeability of radiolabeled hydrocortisone and mannitol across
the isolated rabbit sclera was ve times greater than the corneal permeability. The low activation energy of scleral transport of hydrocortisone suggested an aqueous pore pathway. Hamalainen (30) used a modied twochamber Ussing apparatus to characterize quantitatively the paracellular
permeation routes in the rabbit cornea, conjunctiva, and sclera using poly-

Copyright 2003 Marcel Dekker, Inc.

348

Ahmed

Table 3 Permeability of Sclera to Tracers of Various Molecular Sizes in

Rabbits and Humans
Tracer

Molecular
weight (D)

Water
Mannitol
Sucrose
Hydrocortisone
Dexamethasone

18
182
342
362
396

Sodium uorescein
Carboxyuorescein
Dextran, 4 kDa
Inulin
Dextran, 10 kDa
Dextran, 20 kDa
Dextran, 40 kDa
Dextran, 70 kDa
Dextran, 150 kDa

376
4,400
50005,250
10,000
19,600
38,900
71,200
150,000

Molecular
radius (nm)

Permeability coecient
106 cm/s (SD)
Rabbit

Human

44.6 (13)a

54.4
28.3
42.2
21.8
12.7
0.5
1.3

3.2
4.5
6.4
8.25

919)
(3.7)d
(13.7)e
(4.3)d
(2.3)a

84.5 (16.1)c
13.0 (3.4)a
25.2 (5.1)c
2.54 (0.35)e
6.79
2.79
1.39
1.34

(4.18)c
(1.58)c
(0.88)c
(0.88)c

21.6 (6.0)b
18.2 (5.8)a
23.5 (7.7)b
11.8 (1.37)a
9.0 (2.2)b
6.4 (1.7)b
4.9 (2.4)b
1.9 (0.4)b

SD = Standard deviation.
a

From Ref. 46.

From Ref. 42.
c
From Ref. 47.
d
From Ref. 45.
e
From Ref. 40.
b

ethylene glycol (PEG) oligomers. The scleral permeability was 1525 times
more than that of the cornea but about half that of the conjunctiva. The
permeability of PEGs decreased linearly with increasing molecular weight,
for example, decreasing threefold from a value of 8:80  106 cm/s for PEG
238 to 3:08  106 cm/s for PEG 942.
There have been several reports comparing the permeability of the
sclera, conjunctiva, and cornea to penetrant lipophilicity. The comparative
permeability of a series of b-blockers across isolated rabbit ocular membranes is shown in Table 4. Ahmed et al. (40) showed that the conjunctival
and corneal permeabilities for timolol were comparable but fourfold lower
than the scleral permeability. Sasaki and coworkers (54) reported that the in
vitro permeability of a series of b-blockers across the isolated rabbit sclera
and conjunctiva was not aected by penetrant lipophilicity. Edelhauser and
Maren (41) compared the scleral versus corneal permeability of the carbonic

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

349

Table 4 Differential Permeability of b-Blockers Across Isolated Ocular

Membranes in Rabbit
Permeability (cm/s  106 )
Compound

MW

Log PC

Atenolol
Sotalol
Nadolol
Metroprolol
Timolol
Acetobutolol
Pindolol
Oxyprenolol
Betaxolol
Levobunolol
Labetalol
Alprenolol
Propranolol

266
272
309
267
316
336
248
265
307
291
328
249
259

0.11
0.23
0.23
1.20
1.61
1.63
1.67
1.69
2.17
2.26
2.50
2.65
2.75

Penbutolol

291

4.04

Cornea

Conjunctiva

1.1, 0.68

52

14, 7, 1.6
28, 24
12, 18, 8, 12
1.1, 0.85
10
28
27, 54
29, 23, 17
14
29
31, 34, 46,
58
22, 60

53
62
52
50
15
9
42
51
60
24
20

Sclera
68
39
41

58
71

Source: Adapted from Ref. 24.

anhydrase inhibitors ethoxzolamide and methazolamide and concluded that

the permeability was six times greater in the sclera than in the cornea for the
less lipophilic methazolamide but similar for the more lipophilic ethoxzolamide.
2.

Conjunctiva

Unlike the sclera, the conjunctiva shares an important attribute with the
cornea in that both contain an outer lining of stratied squamous epithelium
continuous with each other at the corneoscleral limbus. However, the cornea
is avascular while the conjunctival epithelium overlies a loose, highly vascular connective tissue, the substantia propria.
Characteristic of any epithelial tissue, both paracellular and transcellular transport is possible across the conjunctiva. In both the cornea and the
conjunctiva, lipophilic drugs prefer the transcellular route, while the paracellular route is preferred by hydrophilic drugs (36). The ratio of the permeability coecients of a series of b-adrenergic blockers in the cornea and
those in the conjunctiva exhibited a sigmoidal correlation with log partition

Copyright 2003 Marcel Dekker, Inc.

350

Ahmed

coecient, as shown in Figure 5 (51). This indicates that while the conjunctiva is leakier than the cornea, the dierential permeability of the conjunctiva to the cornea is higher for hydrophilic compounds than for lipophilic
compounds. The data presented in Table 5, which shows that the permeability of peptides in conjunctiva is higher than in the cornea, support this
conclusion (52,53).
Several authors have studied the eect of molecular size on conjunctival permeability. Horibe et al. (31) characterized the conjunctival permeability to polar solutes ranging from 182 to 167,000 daltons in molecular
weight and concluded that solutes up to 40 kDa traverse the conjunctival
epithelial barrier primarily by restricted diusion through equivalent pores
of 5.5 nm. Polar solutes of greater than 70 kDa may cross the barrier
primarily via nondiusional pathways, such as nonspecic endocytosis.
Huang et al. (27) and Kahn et al. (28) also studied the solute size eect
on conjunctival penetration. They estimated the limiting size of solutes that
can pass the conjunctival barrier by paracellular transport at between 20
and 40 kDa. Hamalainen et al. (29) reported that the conjunctival epithelia
in the rabbit have 2 times larger pores and 16 times higher pore density than
the cornea. They estimated that intercellular space in the cornea is 0:3 
103 mm2 , whereas it is 7  103 mm2 in rabbit conjunctiva. Finally, there is
some evidence that human conjunctiva may be more permeable to hydrophilic solutes than rabbit conjunctiva (32)

Figure 5 Permeability coecient ratio (cornea/conjunctiva) as a function of the log

partition coecient

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

Table 5

351

Permeability Ratio of Peptides Across Ocular Membranes in Rabbit

Compound

MS

Log PC

Diglycine
Triglycine
Tetraglycine
Pentaglycine
TRH
PNP
LHRH

132
189
246
303
362
950
1182

1.89
1.74
1.70
0.77
1.59

Permeability ratio
Conjunctiva/Cornea
10
93
5
9
22
17
47

Source: Adapted from Refs. 36, 53.

V.

DESIGN OF DRUG DELIVERY SYSTEMS TARGETING

THE NONCORNEAL ROUTE

There are limited reports of practical examples of noncorneal drug delivery

dosage forms and drug delivery systems. As noted previously, the noncorneal pathway may be best suited to deliver large, polar molecules, which
have poor permeability across the cornea and which are less likely to be
extensively cleared by systemic loss via absorption into the ocular blood
vessels. However, eective noncorneal drug delivery is likely to require special considerations in dosage form design and methods of administration
that can minimize precorneal and retain high concentrations of drug at the
absorptive surfaces or the conjunctiva or sclera. Delivery system options
may include bioadhesive polymeric vehicles, controlled release inserts, prodrugs, microparticulates, and scleral implants. The method of administration that is most likely to maximize the opportunity to deliver drugs via the
noncorneal route is suprachoroidal delivery via subconjunctival or subTenons injection or implantation, as these methods are best suited to
administer the drug in close proximity of the absorbing surface as a depot
form. Iontophoretic delivery is an option that has not yet been explored for
noncorneal deliver, as is the approach of minimizing systemic drug loss via
absorption into the conjunctival blood vessels by using vasoconstrictors.
A.

Custom Vehicles

Romanelli et al. (20) studied the absorption, distribution, and elimination of

bendazac in ocular tissues after topical administration as eye drops formulated with dierent polysaccharide vehicles. Based on spatial and temporal
distribution of drug in ocular tissues, they concluded that bendazac entered

Copyright 2003 Marcel Dekker, Inc.

352

Ahmed

the retina-choroid and the iris-ciliary body by a sclero-conjunctival route.

Lehr et al. (21) investigated the use of polycarbophil, a mucoadhesive polymer, to improve the ocular delivery of gentamicin formulated in eye drops.
A twofold increase in bulbar conjunctival levels was noted. Based on the
rank-order of peak concentrations and peak times in ocular tissues, the
authors proposed that gentamicin formulated in polycarbophil-containing
eyedrops reach the anterior chamber primarily via the noncorneal route.
B.

Conjunctival Inserts

Urtti et al. (7) showed that application sitedependent absorption of timolol

formulated in a silicone cylindrical device that released drug at 7.2 m/h. Very
low timolol concentrations in the aqueous humor following placement of
the device in the inferior conjunctival sac and high drug concentrations in
parts of each tissue that was closer to the site of application was presented as
evidence of noncorneal entry.
C.

Microparticulates

Ahmed et al. (23) showed evidence of site-specic, noncorneal delivery of

inulin to the posterior eye from topical application of multilamellar liposomes. It is reasonable to expect that noncorneal delivery of some drugs
using nanoparticulates may be feasible.
D.

Prodrugs and Enhancers

In a preliminary evaluation of a series of amphiphilic timolol prodrugs, Pech

et al. (18) presented possible evidence of transscleral absorption. The potential of a drug latentiation as a means to promote selective noncorneal entry
was also presented in the in vitro evaluation of polyethylene glycol esters of
hydrocortisone 21-succinate as ocular prodrugs (120). Chien et al. reported
improved permeability across the conjunctiva for prostaglandin F2a prodrugs (19). Noncorneal enhancers may be agents that reduce systemic loss
or increase conjunctival permeability. Epinephrine pretreatment did not
signicantly aect the concentrations of topically applied timolol in the
cornea, aqueous humor, iris-ciliary body, and conjunctiva of rabbits but
resulted in signicantly higher concentrations in the sclera (67). Although
not explicitly stated, this approach of minimizing systemic loss with vasoconstrictors may render noncorneal entry of selective drugs more favorable.
There have been some exciting leads in approaches and entities to transiently enhance the epithelial permeability of ocular membranes (122124).
The technology of enhancing the conjunctival permeability may become
available in the near future.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

E.

353

Devices and Novel Administration Methods

Arguably the most promising approach to noncorneal delivery is deposition

of drug, preferably as a depot or as a biodegradable implant at, or in the
near proximity of the episclera. Kunou et al. (119) formulated betamethasone phosphate in a biodegradable, polylactic glycolic acid scleral implant
and showed that the drug concentrations in the retina-choroid stayed in the
therapeutic range for one month. Further, the concentrations in the retinachoroid were consistently greater than in the vitreous, which is evidence of
noncorneal entry. Transcleral penetration of drugs following subconjunctival and sub-Tenons injection is precedented and is considered to be a viable
approach for delivering drugs to the posterior tissues of the eye (125127).
Advances in iontophoretic techniques present the possibility that transscleral iontophoresis may replace or supplement intravitreal injection of
antibiotics for the treatment of endophthalmitis (128130).

VI.

CONCLUSIONS/FUTURE DIRECTION

Based on the current understanding it is possible to put the conjunctival/

scleral pathway for intraocular entry of drugs in perspective vis-a-vis ocular
drug delivery. First, the noncorneal penetration pathway involves the permeation of drug across the conjunctiva and sclera and may contribute signicantly to drug penetration into intraocular tissues for some drugs.
Second, drug entering the eye via the cornea enters the aqueous humor
and provides high drug levels to the anterior segment tissues, as described
earlier. In contrast, the fraction of drug entering the eye via the noncorneal
route may bypass the anterior chamber and access tissues of the posterior
segment of the eye, such as the uveal tract, choroid, and retina and, to a
lesser extent, the vitreous humor. The dierential spatial distribution of
drug entering the eye via the corneal versus conjunctival/scleral pathway
has exciting implications in terms of ocular drug delivery. For example,
whereas the corneal route may be preferred for treating anterior segment
eye disease (e.g., glaucoma), the noncorneal route may be considered for
drug therapy targeting the posterior segment of the eye (e.g., uveitis, choroidal neovascular membrane formation, viral retinitis, age-related macular
degeneration). Third, the nonproductive loss of ocularly applied drugs to
the systemic circulation diminishes the fraction of drug that can enter the
eye via the noncorneal route. Since the cornea is nonvascularized, the conjunctival/scleral entry is a minor pathway for most small, semipolar heterocycles that represent the majority of commonly used ophthalmic drugs.
However, the noncorneal pathway may become signicant for large, polar

Copyright 2003 Marcel Dekker, Inc.

354

Ahmed

molecules, administration methods that can minimize precorneal and systemic loss, drugs susceptible to degradation during diusion across the
cornea, and for delivery systems that can retain high concentrations of
drug at the absorptive surfaces or the conjunctiva or sclera.
Recent advances in drug delivery systems that minimize precorneal
loss and can retain high concentrations of drug at the absorptive surfaces
of the conjunctiva or sclera may be particularly suited for noncorneal delivery. These include bioadhesive vehicles, microparticulates, and controlled
release conjunctival inserts. Suprachoroidal delivery of drugs via subconjunctival and sub-Tenons injection, scleral and suprachoroidal implants,
may be the most promising approach to noncorneal delivery. Prodrugs
and permeation enhancers and vasoconstrictors are plausible concepts,
but they require further investigation.
Much progress has been made over the past two decades towards
understanding the fundamental basis of drug penetration via the noncorneal
pathway. The challenge for the future is to creatively apply the available
knowledge to the practical design of drugs and drug delivery systems for
ocular therapy. Noncorneal delivery is not a panacea and will probably have
niche utility in ocular drug delivery. The greatest potential for the concept
appears to be in the area of intraocular delivery of polar molecules, peptides
and protein drugs, and directed drug delivery to treat posterior segment eye
disease.

REFERENCES
1. McCartney, H. J., Drysdale, I. O., Gornall, A. G., and Basau, P. K. (1965).
An auto-radiographic study of the penetration of subconjunctivally injected
hydrocortisone into the normal and inamed rabbit eyes, Invest. Ophthalmol.,
4:297.
2. Bienfang, D. C. (1973). Sector pupillary dilation with an epinephrine strip,
Am. J. Ophthalmol., 75:883.
3. Doane, M. G., Jensen, A. D., and Dohlman, C. H. (1978). Penetration routes
of topically applied eye medications, Am. J. Ophthalmol., 85:383386.
4. Bito, L. Z., and Baroody, R. A. (1981). The penetration of exogenous prostaglandin and arachidonic acid into, and their distribution within, the mammalian eye, Curr. Eye Res., 1:659669.
5. Ahmed, I., and Patton, T. F. (1985). Importance of the noncorneal absorption
route in topical ophthalmic drug delivery, Invest. Ophthal. Vis. Sci., 26:584
587.
6. Ahmed, I., and Patton, T. F. (1987). Disposition of timolol and inulin in the
rabbit eye following corneal versus non-corneal absorption, Int. J. Pharm.,
38:921.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

355

7. Urtti, A., Sondo, T., Pipkin, J. D., Rork, G., and Repta, A. J. (1988).
Application site dependent ocular absorption of timolol, J. Ocul.
Pharmacol., 4:335343.
8. Ashton, P., and Lee, V. H. L. (1989). Role of drug lipophilicity in determining
the contribution of noncorneal penetration in ocular drug absorption, Pharm.
Res., 6:S114.
9. Lee, V. H. L., Ashton, P., Bundgaard, H., and Heuer, D. K. (1990).
Noncorneal route of drug penetration role of drug lipophilicity in determining
its contribution to the ocular absorption of beta blockers and timolol prodrugs, Invest. Ophthal. Vis. Sci., 31:403.
10. Chien, D.-S., Homsy, J. J., Gluchowski, C., and Tang-Liu, D. (1990). Corneal
and conjunctival/scleral penetration of p-aminoclonidine, AGN 190342, and
clonidine in rabbit eyes, Curr. Eye Res., 9:10511059.
11. Sasaki, H., Ichikawa, M., Kawakami, S., Yamamura, K., Nishida, K., and
Nakamura, J. (1996). In situ ocular absorption of tilisolol through ocular
membranes in albino rabbits, J. Pharm. Sci., 85:940943.
12. Rabiah, P. K., Fiscella, R. G., and Tessler, H. H. (1996). Intraocular penetration of periocular ketorolac and ecacy in experimental uveitis, Invest.
Ophthal. Vis. Sci., 37:613618.
13. Kurmala, P., Wilson, G. C., Foulds, W. S., Dhillon, B., Kamal, A., and Rao,
L. S. (1997). Suprachoroidal route of drug delivery to the posterior segment of
the eye, J. Pharm. Pharmacol., 49:83.
14. Worakul, N., and Robinson, J. R. (1997). Review: Ocular pharmacokinetics/
pharmacodynamics, Eur. J. Pharm. Biopharm., 44:7183.
15. Schoenwald, R. D., Deshpande, G. S., Rethwisch, D. G., and Barfknecht, C.
F. (1997). Penetration into the anterior chamber via the conjunctival/scleral
pathway, J. Ocul. Pharmacol. Therap., 13:4159.
16. Sasaki, H., Icikawa, M., Kawakami, S., Yamamura, K., and Mukai, T.
(1997). In situ ocular absorption of ophthalmic b-blockers through ocular
membranes in albino rabbits, J. Pharm. Pharmacol., 49:140144.
17. Conroy, C. W., and Maren, T. H. (1998). The ocular distribution of methazolamide after corneal and sclera administration: Eect of ionization state, J.
Ocul. Pharmacol. Therap., 14:565573.
18. Pech, B., Chetoni, P., Saettone, M. F., Duval, O., and Benoit, J.-P. (1993).
Preliminary evaluation of a series of amphiphilic timolol prodrugs: Possible
evidence for transscleral absorption, J. Ocular Pharmacol., 9:141150.
19. Chien, D. S., TangLiu, D. D. S., and Woodward, D. F. (1997). Ocular penetration and bioconversion of prostaglandin F2 alpha prodrugs in rabbit cornea and conjunctiva, J. Pharm. Sci., 86:11801186.
20. Romanelli, L., Valeri, P., Morrone, L. A., Pimpinella, G., Graziani, G., and
Tita, B. (1994). Ocular absorption and distribution of bendazac after topical
administration to rabbits with dierent vehicles, Life Sci., 54:877885.
21. Lehr, C. M., Lee, Y.-H., and Lee, V. H. L. (1992). Eect of mucoadhesive
polymer polycarbophil on ocular penetration of gentamicin, Pharm. Weekbl.
Sci. Ed., 14: F31.

Copyright 2003 Marcel Dekker, Inc.

356

Ahmed

22. Shiue, M. H. I., Kim, K. J., and Lee, V. H. L. (1998). Modulation of chloride
secretion across the pigmented rabbit conjunctiva, Curr. Eye. Res., 14:927
935.
23. Ahmed, I., and Patton, T. F. (1986). Selective intraocular delivery of liposome-encapsulated inulin via the non-corneal absorption route, Int. J. Pharm.,
34:163167.
24. Prausnitz, M. R., and Noonan, J. S. (1998). Permeability of the cornea, sclera,
and conjunctiva: A literature analysis for drug delivery to the eye, J. Pharm.
Sci., 87:14791488.
25. Lee, V. H. L., Carson, L. W., and Takemoto, K. A. (1986). Macromolecular
drug absorption in the albino rabbit eye, Int. J. Pharm., 29:4351.
26. Ashton, P., Lee, V. H. L. (1989). Para- and transcellular pathways of drug
penetration across the cornea and conjunctiva of the pigmented rabbit,
Pharm. Res., 6: S590.
27. Huang, A. J. W., Tseng, S. C. G., and Kenyon, K. R. (1990). Paracellular
permeability of corneal and conjunctival epithelia, Invest. Ophthalmol. Vis.
Sci., 30:684689.
28. Kahn, M., Barney, N. P., Briggs, R. M., Bloch, K. J., and Allansmith, M. R.
(1990). Penetrating the conjunctival barrier: The role of molecular weight,
Invest. Ophthalmol. Vis. Sci., 31:258261.
29. Hamalainen, K. M., Kananen, K., Auriola, S., Kontturi, K., and Urtti, A.
(1997). Characterization of paracellular and aqueous penetration routes in
cornea, conjunctival and sclera, Invest. Ophthal. Vis. Sci., 38:627634.
30. Hamalainen, K. M. (1997). Characterization of the paracellular penetration
routereply, Invest. Ophthal. Vis. Sci., 38:21792180.
31. Horibe, Y., Hosoya, K., Kim, K. J., Ogiso, T., and Lee, V. H. L. (1997). Polar
solute transport across the pigmented rabbit conjunctiva: size dependence and
the inuence of 8-bromo cyclic ademosine monophosphate, Pharm. Res.,
14:12461251.
32. Maurice, D. M. (1997). Letter: Characterization of paracellular penetration
routes, Invest. Ophthalmol. Vis. Sci., 38:21772179.
33. Kompella, U., Kim, K. J., and Lee, V. H. L. (1992). Active ion and nutrient
transport mechanisms of the pigmented rabbit conjunctiva, Pharm. Res., 9
(Suppl. 3).
34. Kompella, U., Kim, K. J., and Lee, V. H. L. (1993). Active chloride transport
in the pigmented rabbit conjunctiva, Curr. Eye Res., 12:10411048.
35. Kompella, U., Kim, K. J., and Lee, V. H. L. (1995). Possible existence of Na+
coupled amino acid transport in the pigmented rabbit conjunctiva, Life Sci.,
57:14271431.
36. Lee, V. H. L. (1996). Ocular epithelial models, Pharm Biotech., 8:425436.
37. Kompella, U. B., and Lee, V. H. L. (1998). Barriers to drug transport in the
ocular epithelia, in Transport Processes in Pharmaceutical Systems (G. L.
Amidon, P. I. Lee, and E. M. Topp, eds.), Marcel Dekker, New York, pp.
317375.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

357

38. Bill, A. (1965). Movement of albumin and dextran through the sclera, Arch.
Ophthalmol., 74:248252.
39. Maurice, D. M., and Polgar, J. (1977). Diusion across the sclera, Exp. Eye.
Res., 25:577582.
40. Ahmed, I., Gokhale, R. D., Shah, M. V., and Pattom, T. F. (1987).
Physicochemical determinants of drug diusion across the conjunctiva, sclera
and cornea, J. Pharm. Sci., 76:583586.
41. Edelhauser, H. F., and Maren, T. H. (1988). Permeability of the human
cornea and sclera to sulfonamide carbonic anhydrase inhibitors, Arch.
Ophthalmol., 106:11101115.
42. Olsen, T. W., Edelhauser, H. F., Lim, J. I., and Geroski, D. H. (1995). Human
scleral permeability: eects of age, cryotherapy, transscleral diode laser and
surgical thinning, Invest. Ophthalmol. Vis. Sci., 36:18931903.
43. Prausnitz, M. R., Edwards, A., Noonan, J. S., Rudnick, D. E., Edelhauser, H.
F., and Geroski, D. H. (1998). Measurement and prediction of transient
transport across the sclera for drug delivery to the eye, Ind. Eng. Chem.
Res., 37:29032907.
44. Edwards, A., and Prausnitz, M. R. (1998). Fiber matrix model of sclera and
corneal stroma for drug delivery to the eye, AIChE J., 44:214225.
45. Unlu, N., and Robinson, J. R. (1998). Scleral permeability to hydrocortisone
and mannitol in the albino rabbit eye, J. Ocul. Pharm. Ther., 14:273281.
46. Rudnick, D. E., Noonan, J. S., Geroski, D. H., Praunitz, M. R., and
Edelhuser, H. F. (1999). The eect of intraocular pressure on human and
scleral permeability, Invest. Ophthal. Vis. Sci., 40:30543058.
47. Ambati, J., Canakis, C. S., Miller, J. W., Gragoudas, E. S., Edward, A.,
Weissgold, D. J., Kim, I., Delor, F. C., and Adamis, A. P. (2000). Diusion
of high molecular weight compounds through the sclera, Invest. Ophthal. Vis.
Sci., 41:11811185.
48. Maurice, D. M. (1973). Electrical potential and ion transport, Exp. Eye Res.,
15:527532.
49. Sorensen, T., and Jensen, F. T. (1979). Conjunctival transport of technetium99m pertechnetate, Acta Opthalmol., 57:691699.
50. Sasaki, H., Chien, D. S., and Lee, V. H. L. (1988). Dierential conjunctival
and corneal permeability to beta-blockers and its inuence on the ratio of
systemic to ocular drug absorption, Pharm. Res., 5:S98.
51. Wang, W., Sasaki, H., Chien, D.-S., and Lee, V. H. L. (1991). Lipophilicity
inuence on conjunctival drug penetration in the pigmented rabbit: A comparison with corneal penetration, Curr. Eye. Res., 10:571579.
52. Sasaki, H., Igarashi, Y., Nagano, T., Yamamura, K., Nishida, K., and
Nakamura, J. (1995). Penetration of beta-blockers through ocular membranes
in albino rabbits, J. Pharm. Pharmacol., 47:1721.
53. Sasaki, H., Icikawa, M., Yamamura, K., Nishida, K., and Nakamura, J.
(1997). Ocular membrane permeability of hydrophilic drugs for ocular peptide
delivery, J. Pharm. Pharmacol., 49:135139.

Copyright 2003 Marcel Dekker, Inc.

358

Ahmed

54. Sasaki, H., Masataka, I., Shigeru, K., Keno, Y., Takahiro, M., Koyo, N., and
Junzo, N. (1997). Ocular absorption of ophthalmic beta-blockers through
ocular membranes in albino rabbits, J. Pharm. Pharmacol., 49:140144.
55. Alm, A., and Bill, A. (1973). The eect of pilocarpine and neostigmine on
blood ow through the anterior uvea in monkeys. A study with radiolabelled
microspheres, Exp. Eye. Res, 15:3136.
56. Green, K., Wynn, H., and Padgett, D. (1978). Eects of 9Tetrahydrocannabinol on ocular blood ow and aqueous humor formation,
Exp. Eye Res., 26:6569.
57. Riva, C. E., and Ben-Sira, I. (1974). Injection method for ocular hemodynamic studies in man, Invest. Ophthalmol., 13:7779.
58. Lee, V. H. L., and Robinson, J. R. (1979). Mechanistic and quantitative
evaluation of precorneal pilocarpine disposition in albino rabbits, J. Pharm.
Sci., 68:673.
59. Bill, A., Tornquist, P., and Alm, A. (1980). Permeability of the intraocular
blood vessels, Trans. Ophthalmol. Soc. U.K., 100:332336.
60. Lee, V. H. L., Takemoto, K. A., and Iimoto, D. S. (1984). Precorneal factors
inuencing the ocular distribution of topically applied inulin, Curr. Eye Res.,
3:585592.
61. Ziada, G., El-Haddad, S., Fatouh, M., Mustafa, H., and Mahfouz, M. (1985).
Radionuclide study of the blood ocular barrier, Eur. J. Drug Met.
Pharmacokin., 10:325328.
62. Chang, S.-C., and Lee, V. H. L. (1987). Nasal and conjunctival contribution
to the systemic absorption of topical timolol in the pigmented rabbit implications in the design of strategies to maximize the ratio of ocular to systemic
absorption, J. Ocul. Pharmacol., 3:159170.
63. Maitani, Y., Yamamoto, T., Takayama, K., Peppas, N. A., and Nagai, T.
(1995). A modelling analysis of drug absorption and administration from
ocular, nasolacrimal duct, and nasal routes, Int. J. Pharm., 126:8994.
64. Yoshi, M., Nagai, T., Kollias, K., and Peppas, N. (1997). Design of ocular/
lacrimal and nasal systems through analysis of drug administration and
absorption, J. Contr. Rel., 49:185192.
65. Cunha-Vaz, J. G. (1997). The blood-ocular barriers: Past, present, and future,
Doc. Ophthalmol. 93:149157.
66. Kyyronen, K., and Urtti, A. (1990). Improved ocular-systemic absorption
ratio of timolol by viscous vehicle and phenylephrine, Invest. Ophthal. Vis.
Sci., 31:18271833.
67. Kyyronen, K., and Urtti, A., (1990). Eects of epinephrine pretreatment and
solution pH on ocular and systemic absorption of ocularly applied timolol in
rabbits, J. Pharm. Sci., 79:688691.
68. Finne, U., Vaisanen, V., and Urtti, A. (1990). Modication of ocular and
systemic absorption of timolol from ocular inserts by a buering agent and
a vasoconstrictor, Int. J. Pharm., 65:1927.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

359

69. Ohdo, S., Grass, G. M., and Lee, V. H. L. (1991). Improving the ocular to
systemic ratio of topical timolol by varying the dosing time, Invest. Ophthal.
Vis. Sci., 32:27902798.
70. Jarvinen, K., Vartianen, E., and Urtti, A. (1992). Optimizing the systemic and
ocular absorption of timolol from eye drops, STP Pharma Sci., 2:105110.
71. Urtti, A., and Salminen, L. (1993). Review: Minimizing systemic absorption
of topically administered ophthalmic drugs, Surv. Ophthalmol., 37:435456.
72. Lee, Y.-H., and Lee, V. H. L. (1993). Formulation inuence on ocular and
systemic absorption of topically applied atenolol in the pigmented rabbit, J.
Ocular Pharmacol., 9:4758.
73. Urtti, A. (1994). Delivery of antiglaucoma drugs: ocular versus systemic
absorption, J. Ocul. Pharmacol., 10:349357.
74. Jones, A. L., Keighley, J. E., Gold, W., and Good, A. M. (1996). Review: eye
dropsthe hidden poison, Scott. Med. J., 41:110112.
75. Jay, W. M., Aziz, M. J., and Green, K. (1985). The eect of retrobulbar
epinephrine injection on ocular and optic nerve blood ow, Curr. Eye. Res.,
4:5558.
76. Chast, F., Bardin, C., Sauvageon-Martre, H., Callaert, S., and Chaumeil, J. C.
(1991). Systemic morphine pharmacokinetics after ocular administration, J.
Pharm. Sci., 80:911917.
77. Losa, C., Alonson, M. J., Vila, J. L., Orallo, F., Martinez, J., Saavedra, J. A.,
and Pastor, J. C. (1992). Reduction of cardiovascular side-eects associated
with ocular administration of metipranolol by inclusion in polymeric nanocapsules, J. Ocul. Pharmacol., 8:191198.
78. Li, B. H. P., and Chiou, G. C. Y. (1992). Systemic administration of calcitonin
through the ocular route, Life Sci., 50:349354.
79. Chiou, G. C. Y., Shen, Z. F., Zheng, Y. Q., and Chen, Y. J. (1992).
Enhancement of systemic delivery of peptide drugs via the ocular route with
surfactants, Drug Dev. Res., 27:177183.
80. Harris, D., Liaw, J. H., and Robinson, J. R. (1992). Routes of delivery: case
studies. (7). Ocular delivery of peptide and protein drugs, Adv. Drug Delivery
Rev., 8:331339.
81. Chiou, G. C. Y. (1994). Systemic delivery of polypeptide drugs through ocular
route, J. Ocul. Pharmacol., 10:9399.
82. Morgan, R. V. (1995). Delivery of systemic regular insulin via the ocular route
in cats, J. Ocul. Pharmacol. Ther., 11:565573.
83. Morgan, R. V., and Huntzicker, M. A. (1996). Delivery of systemic regular
insulin via the ocular route in dogs, J. Ocul. Pharmacol. Ther., 12:515526.
84. Friedrich, S. W., Saville, B. A., Cheng, Y.-L., and Rootman, D. S. (1996).
Pharmacokinetic dierences between ocular inserts and eye drops, J. Ocul.
Pharmacol., 12:518.
85. Lee, Y. C., and Yalkowski, S. H. (1999). Eect of formulation on the systemic
absorption of insulin from enhancer-free ocular device, Int. J. Pharm.,
185:199204.

Copyright 2003 Marcel Dekker, Inc.

360

Ahmed

86. Mishima, S. (1981). Clinical pharmacokinetics of the eye, Invest. Ophthalmol.

Vis. Sci., 21:504541.
87. Shell, J. W. (1982). Pharmacokinetics of topically applied ophthalmic drugs,
Surv. Ophthalmol., 26:207218.
88. Mikkelson, T. J. (1984). Review: Ophthalmic drug delivery, Pharm. Tech.,
8:9098.
89. Lee, V. H. L. (1985). Review: Topical ocular drug deliveryrecent advances
and future perspectives, Pharm. Int., 6:135138.
90. Lee, V. H. L., and Robinson, J. R. (1986). Review: Topical ocular drug
delivery: Recent developments and future challenges, J. Ocul. Pharmacol.,
2:67108.
91. Schoenwald, R. D. (1990). Review: Ocular drug delivery-Pharmacokinetic
considerations, Clin. Pharmacokin., 18:255269.
92. Lee, V. H. L. (1990). Review: New directions in the optimization of ocular
drug delivery, J. Ocul. Pharmacol., 6:157164.
93. Ding, S. (1998). Review: Recent advances in ophthalmic drug delivery, Pharm.
Sci. Technol. Today., 1:328335.
94. Boutlais, C. L., Acar, L., Zia, H., Sado, P. A., Needham, T., and Leverge, R.
(1998). Ophthalmic drug delivery systemsrecent advances, Prog. Retinal Eye
Res., 17:3358.
95. Robinson, J. C. (1993). Ocular anatomy and physiology relevant to ocular
drug delivery. Drugs Pharm. Sci., 58:2957.
96. Hosoya, K., and Lee, V. H. L. (1997). Cidofovir transport in the pigmented
rabbit conjunctiva, Curr. Eye Res., 16:693697.
97. Mishima, S., Gasset, A., Klyce, S. D., and Baum, J. L. (1966). Determination
of the tear volume and tear ow, Invest. Ophthalmol., 5:264276.
98. Maurice, D. M. (1967). The use of uorescein in ophthalmic research, Invest.
Ophthalmol., 6:464.
99. Holly, F. J. (1973). Formation and stability of the tear lm, Int. Ophthalmol.
Clin., 13:73.
100. Pster, R. R. (1975). The normal surface of the conjunctival epithelium: A
scanning electron microscopic study, Invest. Ophthalmol., 14:267.
101. Kessing, S. V. (1968). Mucus gland system of the conjunctiva, Acta
Ophthalmol. (Suppl.), 95:133.
102. Nichols, B., Davson, C. R., and Togni, B. (1983). Surface features of the
conjunctiva and cornea, Invest. Ophthalmol. Vis. Sci., 24:570576.
103. Watsky, M. A., Jablonski, M. M., and Edelhauser, H. F. (1988). Comparison
of conjunctival and corneal surface areas in rabbit and human, Curr. Eye Res.,
7:483486.
104. Ehlers, N. (1965). On the size of the conjunctival sac, Acta Ophthalmol.,
43:205210.
105. Maurice, D. M. (1984). The cornea and sclera in The Eye (H. Davson, ed.),
Academic Press, New York, pp. 1158.
106. Battagliolo, J. L., and Kamm, R. D. (1984). Measurements of the compressive
properties of scleral tissue, Invest. Ophthalmol. Vis. Sci., 25:5965.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

361

107. Olsen, T. W., Aaberg, S. Y., Geroski, D. H., Edelhauser, H. F. (1998). Human
sclera: thickness and surface area, Am. J. Ophthalmol., 125:237241.
108. Keeley, F. W., Morin, J. D., and Vesely, S. (1984). Characterization of collagen from normal human sclera, Exp. Eye Res., 39:533542.
109. Kleinstein, R. N., and Fatt, I. (1977). Pressure dependency of transscleral
ow, Exp. Eye Res., 24:335340.
110. Francoeur, M., and Patton, T. F. (1979). Kinetics of corneal drug up-take
studied by corneal perfusion in situ I. Evaluation of system and up-take of
ethyl p-aminobenzoate in rabbits, Int. J. Pharmaceut., 2:337342.
111. Olejnik, O., Davis, S. S., and Wilson, C. G. (1981). A non-invasive perfusion
technique for measuring the corneal permeation of drugs, J. Pharm.
Pharmacol.
112. Krohn, D. L., and Breitfeller, J. M. (1974). Transport of pilocarpine by isolated cornea, Invest. Ophthalmol., 13:312316.
113. Gegge, H. S., and Gipson, I. K. (1985). Removal of viable sheets of conjunctival epithelium with dipase II, Invest. Ophthalmol. Vis. Sci., 26:1522.
114. Saha, P., Kim, K. J., and Lee, V. H. L. (1996). A primary culture model of
rabbit conjunctival epithelial cells exhibiting tight barrier properties, Curr.
Eye. Res., 15:11631169.
115. Goskonde, V. R., Khan, M. A., Hutak, C. M., and Reddy, I. K. (1999).
Permeability characteristics of novel mydriatic agents using an in vitro cell
culture model that utilizes sirc rabbit corneal cells, J. Pharm. Sci., 88:180184.
116. Yang, J. J. Ueda, H., Kim, K. J., and Lee, V. H. L. (2000). Meeting future
challenges in topical ocular drug delivery: Development of an air interfaced
primary culture of rabbit conjunctival epithelial cells on a permeable support
for drug transport studies, J. Contr. Rel., 65:111.
117. Sasaki, H., Igarashi, Y., Nishida, K., and Nakamura, J. (1994). Intestinal
permeability of ophthalmic beta-blockers for predicting ocular permeability,
J. Pharm. Sci., 83:13351338.
118. Zhu, Y. P., Wilson, W. S. (1996). An ex vivo model for the assessment of drug
delivery to the eye: isolated bovine perfusion system, Eur. J. Pharm.
Biopharm., 42:405410.
119. Kunou, N., Ogura, Y., Honda, Y., Hyon, S. H., and Ikada, Y. (2000).
Biodegradable scleral implant for controlled intraocular delivery of betamethasone phosphate, J. Biomed. Mat. Res., 51:634641.
120. Foroutan, S. M., and Watson, D. G. (1999). In vitro evaluation of polyethylene glycol esters of hydrocortisone esters of hydrocortisone 21-succinate as
ocular prodrugs, Int. J. Pharm., 182:7992.
121. Sasaki, H., Igarashi, Y., Nishida, K., and Nakamura, J. (1993). Ocular delivery of the beta-blocker, tilisolol, through the prodrug approach, Int. J.
Pharm., 93:4960.
122. Hamalainen, K. M., Ranta, V. P., Auriola, S., and Urtti, A. (2000). Enzymatic
and permeation barrier of [D-ala (2)]-met-enkephalinamide in the anterior
membranes of the albino rabbit eye, Eur. J. Pharm. Sci., 9:265270.

Copyright 2003 Marcel Dekker, Inc.

362

Ahmed

123. Madhu, C., Rix, P., Nguyen, T., Chien, D. S., Woodward, D. F., and
TangLiu, D. D. S. (1998). Penetration of natural prostaglandins and their
ester prodrugs and analogs across human ocular tissues in vitro, J. Ocul.
Pharm. Ther., 14:389399.
124. Sasaki, H., Nagano, T., Yamamura, K., Nishida, K., and Nakamura, J.
(1995). Ophthalmic preservatives as absorption promoters for ocular drug
delivery, J. Pharm. Pharmacol., 47:703707.
125. Sasaki, H., Yamamura, K., Tei, C., Nishida, K., and Nakamura, J. (1995).
Ocular permeability of FITC-dextran with absorption promoter for ocular
delivery of peptide drugs, J. Drug Targeting, 3:129135.
126. Chung, Y. B., Nishiura, A., and Lee, V. H. L. (1993). Pz-peptide as a novel
enhancer of ocular epithelial paracellular permeability in the rabbit, Pharm.
Res., 10:204.
127. Bok, C. Y., Kun, H., Akio, N., and Lee, V. H. L. (1998). Ocular absorption of
pz-peptide and its eect on the ocular systemic pharmacokinetics of topically
applied drugs in the rabbit, Pharm. Res., 15:18821887.
128. Peyman, G. A., and Ganiban, G. J. (1995). Delivery systems for intraocular
routes, Adv. Drug Delivery Rev., 16:107123.
129. Geroski, D. H., and Edelhauser, H. F. (2000). Review: Drug delivery for
posterior segment eye disease, Invest. Ophthal. Vis. Sci.: 961964.
130. Coltrust, M. J., Williams, R. L., Hiscott, P. S., and Grierson, I. (2000).
Review: Biomaterials used in posterior segment of the eye, Biomaterials.,
21:649665.
131. Peyman, G. A., and Ganiban, G. J. (1995). Review: Delivery systems for
intraocular routes, Adv. Drug Delivery Rev., 16:107123.
132. Velez, G., and Whitcup, S. M. (1999). New developments in sustained release
drug delivery for the treatment of intraocular disease, Br. J. Ophthalmol.,
83:12251229.
133. Conroy, C. W. (1997). Sulfonamides do not reach the retina in therapeutic
amounts after topical application to the cornea, J. Ocul. Pharmacol., 13:465
472.
134. Lim, J. I., Maguire, A. M., John, G., Mohler, M. A., and Fiscella, R. G.
(1993). Intraocular tissue plasminogen-activator concentrations after subconjunctival delivery, Ophthalmology, 100:373376.
135. Pakes, S. P. (1998). V. A., Grant: Implantation of a sub-tenon drug delivery
device loaded with a test article in rabbits and distribution of the test article in
ocular tissues. Dept. of Veterans Administration Grant, 1998.
136. Adamis, A. P., Gragoudas, E. S., and Miller, J. W. (1999). Patent: Targeted
trans-scleral controlled release drug delivery to the retina and choroid, WO
2000US207.
137. Sasaki, H., Kashiwagi, S., Mukai, T., Nishida, K., Nakamura, J., Nakashima,
M., and Ichikawa, M. (1999). Drug absorption behavior after periocular
injections, Biol. Pharm. Bull., 22:956960.

Copyright 2003 Marcel Dekker, Inc.

The Noncorneal Route in Ocular Drug Delivery

363

138. Blair, J. M., Gionfriddo, J. R., Polazzi, L. M., Sojka, J. E., Pfa, A. M., and
Bingaman, D. P. (1999). Subconjunctivally implanted micro-osmotic pumps
for continuous ocular treatment in forses, Am. J. Vet. Res., 60:11021105.
139. LaFranco, D. M., Tao, T. V., Yan, G. C., and Herbert, C. P. (1999). Posterior
sub-Tenons steroid injection for the treatment of posterior ocular inammation: Indications, ecacy and side-eects, Graefes Arch. Clin. Exp. Ophthal.,
237:289295.
140. Kunow, N., Ogura, Y., Honda, Y., Hyon, S. H., and Ikada, Y. (2000).
Biodegradable sclera implant for controlled intraocular delivery of betamethasone phosphate, J. Biomed. Res., 51:635641.
141. Schulman, J. A., Peyman, G. A., and Liu, J. (1987). The intraocular penetration
of acyclovir after subconjunctival administration, Ophthal. Surg., 18:111114.
142. Kurmala, P., Wilson, C. G., Dhillon, B., Kamal, A., and Rao, L. S. (1997). A
comparison of intraocular delivery routes for an acyclovir implant using an
arterially perfused sheep eye, Pharm. Res., 14:S46.
143. Callegan, M. C., Mobden, J. A., OCallaghan, R. J., and Hill, J. M. (1995).
Ocular drug delivery: A comparison of transcorneal iontophoresis to corneal
collagen shields, Int. J. Pharm., 123:173179.
144. Sarraf, D., and Lee, D. A. (1994). Review: The role of iontophoresis in ocular
drug delivery, J. Ocul. Pharmacol., 10:6981.
145. Frucht, P. J., Solomon, A., Doron, R., Ever-Hadani, P., Manor, O., and
Shapiro, M. (1996). Ecacy of iontophoresis in the rat cornea, Graefes
Arch. Clin. Exp. Ophthal., 234:765769.
146. Friedberg, M. L., Pleyer, U., and Mondino, B. J. (1991). Device drug delivery
to the eye. Collagen shields, iontophoresis, and pumps, Ophthalmology.,
98:725732.
147. Hill, J. M. (1991). Symposium on drug delivery. V. Ocular drug delivery;
corneal collagen shields and ocular iontophoresis, Proc. Soc. Exp. biol.
Med., 196:365.
148. Burstein, N., and Anderson, J. A. (1985). Review: Corneal penetration and
ocular bioavailability of drugs, J. Ocular Pharmacol., 1:309326.
149. Munjusha, M., and Majumdar, D. K. (1997). In vitro transcorneal penetration of ketorolac tromethamine from buered and unbuered aqueous ocular
drops, Indian J. Exp. Biol., 35:941947.
150. DeSantis, L. M., and Schoenwald, R. D. (1978). Lack of inuence of rabbit
nectitating membrane on miosis eect of pilocarpine, J. Pharm. Sci., 67:1189
1190.
151. Burstein, N. L., and Anderson, J. A. (1985). Review: Corneal penetration and
ocular bioavailability of drugs, J. Ocular Pharmacol., 3:309326.
152. Klyce, S. D., and Beuerman, R. W. (1988). Structure and function of the
cornea, in The Cornea (H. E. Kaufman, B. A. Barron, M. B. McDonald, S.
R. Waltman, eds.), Churchill Livingstone, New York, pp. 354.
153. Pepose, J. S., and Ubels, J. L. (1992). The cornea, in Alders Physiology of the
Eye (W. Hart, ed.), Mosby Year book, New York, pp. 2970.
154. A. K. Mitra. Ophthalmic drug delivery. In: P. Tyle, ed. Drug Delivery
Devices. New York: Marcel Dekker, 1988.

Copyright 2003 Marcel Dekker, Inc.

12
Ocular Iontophoresis
Marvin E. Myles, Jeannette M. Loutsch, Shiro Higaki, and
James M. Hill
LSU Eye and Vision Center of Excellence, Louisiana State University
Health Science Center, New Orleans, Louisiana, U.S.A.

I.
A.

IONTOPHORESIS
Introduction, Literature Reviews, and Citations

Iontophoresis is the use of a direct electrical current to drive topically

applied ionized substances into or through a tissue (1). Iontophoresis is
based on the physical principle that ions with the same charge repel (electrorepulsion) and ions with opposite charge attract (electroosmosis) (2).
Iontophoresis usually employs low voltage (10 V or less) to supply a continuous direct current of 0.5 mA/cm2 or less (1). These basic operational
guidelines have enabled iontophoresis to be used to enhance drug delivery in
a wide variety of conditions. The symmetry of the procedure also permits its
application to the noninvasive sampling of biologically important subcutaneous uids or for blood monitoring (3). Table 1 lists reviews on the topic
and selected citations that highlight some of the innovative ways in which
this modality is being used in the treatment and diagnosis of various conditions.

B.

Basic Concepts and Electrical Laws

Iontophoresis causes increased transport of ionized substances into or

through a tissue by application of an external electric current (14,15).
Iontophoretic transport results from the passage of a current from electrodes into an electrolyte solution and thus into the skin or body. The
365

Copyright 2003 Marcel Dekker, Inc.

366

Myles et al.

Table 1 Iontophoresis Reviews

Topic
Historical
General aspects
Principles

Dermatology
Anesthesia
Devices/technology
Peptides and proteins
Ocular therapy

Author(s)

Year

Ref.

Duke-Elder
Harris
Hughes and Maurice
Banga and Chien
Singh and Maibach
Nair et al.
Guy et al.
Kassan et al.
Zempsky and Ashburn
Tyle
Garrison
Hirvonen et al.
Shofner et al.
Sarraf and Lee
Sasaki et al.

1962
1967
1984
1988
1994
1999
2000
1996
1998
1986
1998
1996
1989
1994
1999

4
5
72
6
1
7
2
33
61
8
9
10
11
12
13

basic electrical principle that oppositely charged ions attract and same
charged ions repel is the central tenet of iontophoresis. Thus, positive ions
(cations) are attracted to the negative electrode (or cathode) and repelled by
the positive electrode (or anode). Conversely, negative ions (anions) are
attracted to the positive electrode and repelled by the negative electrode.
When iontophoresis is used therapeutically, the ions of importance are
charged molecules of the drug or other bioactive substances. The ionized
substances are driven into the tissues by electrorepulsion at either the anode
(if they carry a positive charge) or the cathode (if they carry a negative
charge) (2).
A few laws of physics and chemistry will help to delineate the important parameters in iontophoretic transport. Those parameters are the
amount of drug transported, the current rate, and the amount of time
that current is applied. The rst law is Ohms law:
V IR
where V is the electromotive force in volts, I is the current in milliamperes
(mA), and R is the resistance in ohms. At constant voltage, any change in
resistance results in a change in the current. With iontophoresis, resistance
decreases during the procedure. The result is that the current (mA) increases
and must be reduced to maintain a constant current over time.
A second law important for iontophoresis is Coulombs law:
Q IT

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

367

where Q is the quantity of electricity, I is the current in mA, and T is the

time in minutes. Thus, Q, which is the total current dosage, can be expressed
as mA-minutes. Precise conditions for specic iontophoretic applications
can be expressed as a minimum, maximum, or a range of mA-minutes.
Finally, a third important physical principle is Faradays law:
D

IT
IZIF

where D is the drug delivered in gram-equivalents, I is the current in mA, T

is the time in minutes, IZI is the valence of the drug, and F is Faradays
constant. Faradays constant is the electrical charge carried by 1 gramequivalent of a substance. The importance of this proportional relationship
is that if more current is applied (either by increasing the current rate or
increasing the time of application of a constant low current), more of the
drug enters the tissue.

C.

Design of Iontophoresis Devices

Iontophoretic devices vary in complexity, but the basic design is a unit with
a power source (either a battery or an on-line unit with a voltage regulator),
a milliampere meter to measure the current, a rheostat to control the
amount of current owing through the system, and two electrodes.
Platinum is the material of choice for the electrodes, since it releases almost
no ions, undergoes degradation at a slow rate, and is nontoxic.
A variety of iontophoretic apparatuses exist for use in ocular iontophoresis. They mainly consist of either an eyecup or an applicator probe.
Figure 1 shows a diagram of ocular iontophoresis of a positively charged
drug in a rabbit. The eyecup, with an internal diameter of  1 cm, is placed
over the cornea and lled with the drug solution. A metal electrode that is
connected to a direct current power supply is submerged in the solution in
the eyecup without making contact with the surface of the eye. The ground
electrode, connected to the other terminal of the power supply, is attached
to the ear of the rabbit via wet (0.9% NaCl) gauze to ensure a good connection. With the hand-held applicator probe, the metal (platinum) electrode extends into the eyecup that is lled with the drug solution. The
eyecup is placed against the eye and is held in place throughout the entire
iontophoresis procedure. Iontophoresis requires a complete electrical circuit
with direct current passing from the anode to the cathode and from the
cathode back to the anode. The two electrodes are placed as anatomically
close to each other as possible on the body, which is an excellent conductor
of electricity, to complete the circuit.

Copyright 2003 Marcel Dekker, Inc.

368

Myles et al.

Figure 1 Ocular iontophoresis in the rabbit. The drug is placed in a cylindrical eye
cup with a central diameter of 912 mm; the inner circumference of the eye cup ts
within the corneoscleral limbus. The current is controlled by a rheostat on the direct
current transformer. In general, the current should not exceed 2.0 mA and the time
be no longer than 10 min. In the case illustrated here, the drug molecules (cations)
have a positive charge. Therefore, the platinum electrode connected to the anode (the
positively charged pole) is placed in contact with the solution. The other electrode
(cathode) is connected to the ear or front leg of the rabbit to complete the circuit.
The positively charged anode drives the positively charged drug molecules from the
solution into the eye at a greater rate than would be observed with simple diusion.

The drug solution or preparation to be iontophoresed should be

devoid or have a minimum of extraneous ions. Drugs with one or more
pKa values either below pH 6 or above pH 8 are generally excellent candidates for iontophoresis into the eye because these drugs will be in the ionized
form at the physiological pH of the eye (12). The salt form of a drug is also
preferred for iontophoresis since the dissociated salt is highly soluble. The
drug is driven into the ocular tissue with an electrode carrying the same
charge as the drug while the ground electrode, which is of the opposite
charge, is placed elsewhere on the body (usually the ear) to complete the
circuit. The drug or bioactive substance serves as a conductor of the current
through the ocular tissue. The transported drugs or bioactive substance
either remain in the tissue until they are altered/metabolized or are carried
away by the blood vascular network.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

II.

MEDICAL APPLICATIONS OF IONTOPHORESIS

A.

Archival Studies

369

The use of the shock of the torpedo, an electric sh, for the treatment of
gout was described by Aetius, a Greek physician, more than 1000 years ago
(14). In 1747, Veratti enunciated the concept of applying an electric current
to increase the penetration of drugs into surface tissues (15). In 1898,
Morton demonstrated that nely powdered graphite could be driven into
his arm under the positive electrode and produced small black spots that
persisted for weeks (16). In 1900, Leduc reported the rst controlled studies
of iontophoresis as a therapeutic modality (17,18). Leduc showed that transcutaneous iontophoretic delivery of strychnine and cyanide ions into rabbits
produced fatal tetanic seizures and cyanide poisoning.
The earliest description of ocular iontophoresis was published in 1908
by the German investigator Wirtz (19), who performed iontophoresis of zinc
salts for the treatment of corneal ulcers. In 1927, Morisot (20) enumerated
many successful ophthalmological applications of iontophoresis including
iontophoresis of magnesium for treatment of glaucoma, iontophoresis of
ammonium chloride for treatment of cataract, and iontophoresis of phosphoric acid for treatment of optic atrophy. Erlanger was one of the rst
ophthalmologists to introduce iontophoresis to England and the United
States. In 1936, he delivered barium chloride iontophoretically into the
eyes of guinea pigs and observed cataract formation 48 hours later (21).
He went on to describe the usefulness of iontophoresis in the clinical treatment of corneal ulcers, conjunctivitis, scleritis, glaucoma, and cataract
(22,23).
During the 1940s in the United States, Ludwig von Sallmann, a prominent ophthalmologist, was one of the pioneers in the clinical use of ocular
iontophoresis. Von Sallmann showed that transcorneal iontophoresis of
penicillin was more eective than subconjunctival injection for the delivery
of penicillin into the aqueous humor (24,25) and demonstrated modest
success in the treatment of intraocular staphylococcal infection (26). In
1956, Witzel and his colleagues (27) published a report on the use of ocular
iontophoresis as a drug delivery system for a variety of antibiotics. They
found that iontophoresis was eective in the delivery of streptomycin, neomycin, and penicillin.
Despite its widespread use and study during the rst 60 years of the
twentieth century, iontophoresis was never fully adopted as a standard
procedure. The lack of carefully controlled trials and the paucity of toxicity
data were among the reasons that precluded its acceptance as a viable
alternative for drug delivery. However, over the past 3040 years, iontophoresis has been adapted for use in a variety of medical specialties, includ-

Copyright 2003 Marcel Dekker, Inc.

370

Myles et al.

ing anesthesiology, dermatology, dentistry, and ophthalmology. Table 1

provides a list of selected reviews and reports describing the recent evolution
of this procedure and highlighting some of its medical applications.

B.

Iontophoresis in General Medicine and Dentistry

1. Diagnosis of Cystic Fibrosis: Pilocarpine Iontophoresis

Iontophoresis of pilocarpine (28) to induce sweating to measure sweat
sodium and chloride concentration is the basis of the sweat test that is
used to diagnose cystic brosis (2931). High concentrations of both sodium
and chloride ions in the sweat are considered as unequivocal evidence of the
disease. Although essential criteria have been established for a positive test
result in both children (30) and adults (31), pilocarpine iontophoresis should
be used in conjunction with other clinical features to make a denitive
diagnosis. This procedure is particularly useful in children younger than 1
year old because it is essentially painless and takes only 35 minutes to
complete.
2. Treatment of Hyperhidrosis: Tap-Water Iontophoresis
The rst dermatological application of iontophoresis was to treat hyperhidrosis (32). Hyperhidrosis is a condition characterized by pathologically
excessive sweating due to abnormal secretion of the eccrine sweat glands
in various parts of the body [primarily the palms, soles and axillae (3335)].
Iontophoresis of tap water has been very eective (90% of patients) in
inhibiting palmar and plantar hyperhidrosis, but the results are less rewarding for axillary hyperhidrosis (33,35).
3. Treatment of Hypersensitive Teeth: NaF Iontophoresis
Iontophoresis of sodium uoride to treat teeth that have become thermosensitive is a very useful and successful therapy in dentistry (36,37). Teeth
previously shown to be hypersensitive to cold lose their sensitivity to heat
and cold immediately after iontophoresis of sodium uoride, and the eects
are long lasting. The mechanisms by which teeth become hypersensitive and
by which iontophoresis alleviates this condition are not fully understood.
One theory is that exposed dentin allows uid movement through microtubules that stretch from the pulp. Occlusion of the patent dentinal tubules
would appear to be one mechanism by which sodium uoride iontophoresis
could have a benecial eect.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

4.

371

Pediatric Anesthesia: Lidocaine Iontophoresis

Iontophoresis of anesthetics for induction of local anesthesia in pediatric

oce patients has been very successful (3840). Randomized studies have
found lidocaine iontophoresis to be more eective than EMLA (eutetic
mixture of local anesthetics) cream or topical lidocaine (38,41), and its
eectiveness has been documented for placement of pediatric intravenous
catheters (39). Iontophoresis is advantageous in pediatric patients because it
alleviates the pain and anxiety due to needle injections and works more
rapidly than topically applied anesthetics, which take a long time to become
eective (up to 1 h) and often provide incomplete anesthesia. Squire et al.
(41) demonstrated that the time to accomplish topical anesthesia was shorter
with iontophoresis of 2% lidocaine with epinephrine 1:100,000 (13 min)
compared to a surface-applied EMLA cream [60 min (41)]. Iontophoresis
of lidocaine with epinephrine is a safe, rapid, and eective method of local
anesthesia delivery for pediatric procedures.

C.

Iontophoretic Therapies in Ophthalmology

Table 2 provides a list of the drugs, dyes, and other charged molecules
summarized below. The cations (positive ions) are used in anodal (positive
electrode) iontophoresis, whereas the anions (negative ions) are used in
cathodal (negative electrode) iontophoresis. Iontophoresis of the various
classes of drugs (antibiotics, antivirals, antifungal, antimetabolite, adrenergic, steroid, anesthetic, and dyes) can be delivered by two approaches.
Transcorneal iontophoresis (described earlier and in diagrammatic form in
Fig. 1) delivers a high concentration of drug to the anterior segment of the
eye (cornea, aqueous humor, ciliary body, and lens). In phakic animals, the
lens-iris diaphragm limits penetration of a drug to the posterior tissues of
the eye such as posterior vitreous and retina. This barrier can be overcome
by applying the current through the pars plana (transscleral iontophoresis),
which can produce signicantly high and sustained drug concentration in
the vitreous and retina. For transscleral iontophoresis, the drug solution is
contained in a narrow tube within an eyecup held to the conjunctiva by
suction. The tube is placed over the pars plana to avoid current damage to
the retina. This technique circumvents the lens-iris barrier and delivers drugs
into the vitreous or retina. Figure 2 shows a diagram of a transscleral
iontophoresis device and setup.
Within the past 10 years, a number of excellent articles/chapters have
reviewed the application of iontophoresis in therapeutic approaches in
ophthalmology (11,12,4244). Current research in ocular iontophoresis is
aimed at resolving the delivery problems associated with newly developed

Copyright 2003 Marcel Dekker, Inc.

372

Table 2 Drugs Used in Ocular Iontophoresis

Class
Antibiotic

Antiviral

Antifungal
Antimetabolite
Adrenergic

Steroid
Generic

Copyright 2003 Marcel Dekker, Inc.

Site

Ion

Application

Ref.

Gentamicin
Tobramycin
Ciprooxacin
Vancomycin
Ticarcillin
Cefazolin
Foscarnet
Ara-AMP
Iododeoxyuridine
Acyclovir
Ketoconazole
5-Fluorouracil
Epinephrine
Timolol maleate
b-Propranolol
6-Hydroxydopamine
a-Methylparatyrosine
Dexamethasone
Hydrocortisone
Lidocaine
Lysophosphatidic acid
Fluorescein
Methylene blue
reactive black 5

Transcorneal
Transcorneal
Transscleral
Transcorneal
Transscleral
Transscleral
Transscleral
Transcorneal
Transcorneal
Transcorneal
Transcorneal
Transscleral
Transcorneal
Transcorneal
Transcorneal
Transcorneal
Transcorneal
Transscleral
Transcorneal
Transcorneal
Transcorneal
Transcorneal
Transscleral
Transscleral

Cation
Cation
Cation
Cation
Anion
Anion
Anion
Anion
Anion
Anion
Cation
Anion
Cation
Cation
Cation
Cation
Cation
Anion

Tx Pseudomonal keratitis
Tx Pseudomonal keratitis
Tx Pseudomonal keratitis
Tx Pseudomonal keratitis
Tx Endophthalmitis
Tx Endophthalmitis
Tx CMV retinitis
Tx Herpes keratitis; stromal disease
Tx Herpes keratitis
Tx Herpes stromal disease
Tx Fungal infection
Glaucoma surgery, Tx ocular tumors
HSV-1 shedding
HSV-1 shedding
HSV-1 shedding
Tx glaucoma
Tx glaucoma
Tx Inammation
Tx Inammation
Anesthesia
HSV-1 reactivation
Study aqueous humor dynamics
Laser sclerostomy; Tx glaucoma
Laser sclerostomy; Tx glaucoma

7078
7982
8385
86
75
75
91101
104106
107
107
87
59,60
110,115126
127129
130133
5457
58
6469
64
6163
163
4548
4952
53

Cation
Anion
Anion
Cation
Anion

Myles et al.

Dyes

Drug

Ocular Iontophoresis

Figure 2
Ref. 75.)

373

Diagram of apparatus used for transscleral iontophoresis. (Adapted from

drugs, techniques, and therapies. This section discusses the literature concerning the use of iontophoresis for delivery of anesthetics to the eye and in
the treatment/study of glaucoma, ocular infections, ocular inammation,
and ocular herpes infection.
1.

Glaucoma

a. Fluorescein for Studies of Aqueous Humor Dynamics Fluorescein

is an orange-colored dye that is soluble in water and has a negative charge
in solution. In 1966, Jones and Maurice (45) published a new procedure
for measuring the rate of ow of the aqueous humor in patients using iontophoresis of uorescein (10% solution) and a slit-lamp uorophotometer.
The solution was iontophoresed for 1015 seconds with a current of 0.2
mA. Other early investigators included Starr (46), who reported results
similar to those of Jones and Maurice (45) using iontophoresis of uorescein at 0.2 mA for 3060 seconds (s), and Holm (47), who observed pupillary ow of uorescein after saturating the anterior chamber by
iontophoresis (0.5 mA for 4 min) through a cotton-wick electrode.

Copyright 2003 Marcel Dekker, Inc.

374

Myles et al.

A review of the procedures and techniques used to study the ow of

aqueous humor in the eye was published by Brubaker in 1982 (48). He noted
that these methods were essentially equivalent to those described by Jones
and Maurice (45). After application of topical anesthesia, a gel 5 mm in
diameter containing 2% agar and 10% uorescein was placed on the central
cornea. The power source was a 45 V battery. The agar gel constituted the
negative electrode, and to complete the circuit, the patient held the positive
electrode in his hand. The current (0.2 mA) was applied for 57 seconds. In
more than 10,000 iontophoretic applications of uorescein, no obvious side
eects of the iontophoretic procedure were observed.
b. Dyes for Laser Sclerostomy Pulsed dye laser sclerostomy is an
adjunctive procedure used in the treatment of glaucoma (49). The pulsed
dye laser procedure uses a gonioscopic approach for the ab interno delivery of visible laser light. The procedure requires a full thickness penetration of a dye throughout a 12 mm2 area of scleral tissue for adequate
absorption of the visible light energy. The light beam is transmitted
through the cornea, crosses the anterior chamber, and ablates stained limbal scleral tissue with the formation of a stula [a ltration channel for intraocular pressure (IOP) release (49)].
Methylene blue dye is water soluble and has a positive electrical charge
in solution. It has an absorption peak of 668 nm and is used to enhance the
optical absorption of scleral tissue (49). Latina et al. (50) iontophoresed 1%
methylene blue at the limbal region of 35 glaucoma patients. The current
applied was 5.0 mA for a duration of 48 minutes. To create the sclerostomy,
the laser energy (200250 mJ), delivered by a slit-lamp, was focused onto the
dyed sclera, using a goniolens, so that only a light beam penetrated the eye.
The red wavelength of 660 nm generated by the laser was maximally
absorbed by the stained sclera and created a complete sclerostomy.
Successful sclerostomies were achieved in 21 of 35 patients (60%) with a
reduction in IOP from a mean preoperative value of 35 mmHg to a mean
postoperative value (at 9 months) of 22 mmHg. Melamed (51) obtained
similar results with a 58% success rate for complete sclerostomies and a
reduction in IOP from a mean preoperative IOP of 36.6 mmHg to a mean
postoperative IOP of 23.7 mmHg. Grossman et al. (52) examined the stability of iontophoresed methylene blue in rabbit eyes. Decreased dye concentration of over 50% within 2 hours and a complete disappearance of dye
within 24 hours were demonstrated. They also noted that the stain tended to
bleach from the laser dye exposure. This prevented further absorption of the
laser energy, resulting in incomplete scleral ablation and stula formation.
Melamed (51) reported blanching of the stained sclera after the rst laser
shots, which adversely aected the ecacy of subsequent shots.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

375

Reactive black-5 (RB5) is a water-soluble black dye that is negatively

charged at the physiological pH of the eye. RB5 has been proposed as an
alternative dye to stain the scleral tissue (53). RB5 stain was subjected to a
number of dierent conditions simulating laser treatment of the sclera. The
stain was stable over time (72 h) and stable when exposed to high temperature (1208C), to scleral breakdown products (collagen), to strong oxidants
(1.5% H2 O2 ), and to laser light energy (53). Optimal parameters for iontophoretic delivery of RB5 into limbal scleral tissue were also determined.
Ideal parameters for iontophoresis included a probe tip surface area
between 0.1 and 0.7 mm2, a current of 0.5 mA, and a duration of 5 minutes.
Using these parameters for iontophoresis, the maximum concentration of
RB5 achieved in sclera was 0.15% (53). This value is considerably greater
than the threshold for ablation of 0.001% RB5 using a laser energy of 250
mJ. Thus, iontophoresis can deliver an amount of RB5 stain to the sclera
that is more than sucient for laser ablation. This approach obviates conjunctival dissection and decreases the stimulus for episcleral scarring that
eventually could cause reelevation of intraocular pressure (49,52,53).
c. Adrenergic Agents for Treatment of Glaucoma 6-Hydroxydopamine and a-methylparatyrosine are two pharmacological agents that block
the synthesis of norepinephrine. 6-Hydroxydopamine, a congener of norepinephrine, causes the reversible destruction of nerve terminals in the
anterior segment. In the early to mid-1970s, a number of investigators
(5458) used iontophoresis to deliver these substances to the eyes of rabbits, normal volunteers, and glaucoma patients with primary open-angle
glaucoma. The theory behind the treatment involved the depletion of ocular norepinephrine, which would result in an increased sensitivity to glaucoma drugs such as epinephrine.
Kitazawa et al. (54,55) were the rst to report the results of iontophoresis of 6-hydroxydopamine in rabbits and human eyes. A 1% solution of 6hydroxydopamine was iontophoresed at 0.75 mA for 3 minutes. High concentrations of 6-hydroxydopamine were achieved in ocular tissues in rabbits, and intraocular pressure was reduced in normal human eyes.
Subsequently, Kitazawa et al. (56) treated patients with primary openangle glaucoma with a combination therapy of iontophoresed 6-hydroxydopamine and topical epinephrine. The results led them to conclude that
this combination therapy had clinical value in the management of openangle glaucoma.
Iontophoresis of 6-hydroxydopamine to treat primary open-angle glaucoma was also examined by Watanabe et al. (57). For the patients for whom
they had sucient data for analysis (49/100), this procedure was therapeutically eective in 41%, questionable in 31%, and ineective in 28%.

Copyright 2003 Marcel Dekker, Inc.

376

Myles et al.

Colasanti and Trotter (58) performed ocular iontophoresis of amethylparatyrosine in rabbits. A 4.0% a-methylparatyrosine solution was
iontophoresed at a current of 3 mA for 5 minutes. This drug is similar to 6hydroxydopamine, and it also produced a signicant decrease in the norepinephrine concentration in rabbit ocular tissues. No clinical studies with amethylparatyrosine were done in normal human eyes or in patients with
primary open-angle glaucoma.
These results demonstrated that iontophoresis of 6-hydroxydopamine
was a viable method of sensitizing the eyes to glaucoma drugs and that this
procedure had some clinical value in the management of this disease.
However, with the advent of long-acting antiglaucoma drugs such as timolol, levobunolol, and betaxolol, iontophoresis of 6-hydroxydopamine for the
therapy of glaucoma was discontinued.
d. 5-Fluorouracil for Control of Cellular Proliferation After Glaucoma
Surgery 5-Fluorouracil (5-FU) acts as an antiproliferative agent to prevent cellular replication. The concentration of 5-FU required for 50% inhibition of rabbit conjunctival broblasts in culture is 0.20.5 mg/mL (59).
5-FU is a small, negatively charged molecule with a pKa of  8. Kondo
and Araie (60) were the rst to report iontophoresis of 5-FU to the rabbit
eye. A 5% solution of 5-FU containing 8.47% tris(hydroxymethyl)aminomethane was delivered transsclerally at 0.5 mA for 30 seconds. An electrode 7 mm in diameter was placed on the bulbar conjunctiva 4 mm
posterior to the limbus in the superior temporal quadrant. Iontophoresis
of 30-second duration delivered enough 5-FU into ocular tissue such that
30 minutes later the drug concentration was 50 mg/g and 21 mg/g in conjunctival and scleral tissue, respectively. Over the next 10 hours, the
amount of 5-FU decreased to 0.6 mg/g in the conjunctiva and 1.2 mg/g in
the sclera. These concentrations are still high enough to have a therapeutic effect.
Transscleral iontophoresis of 5-FU would eliminate the need for subconjunctival injection and its unwanted complications (risk of bleeding,
infections, scarring, and drug penetration into other ocular tissues).
Iontophoretically delivered 5-FU may improve the ecacy of antiglaucoma
surgery (e.g., sclerostomy) by interfering with healing and thereby maintaining patency of the stulas. To date, however, no studies have been done in
experimental models of disease or in human eyes.
2. Ocular Anesthesia
The deliver of local anesthetics by iontophoresis has been very successful
(6163). Iontophoretically delivered anesthetics can provide topical anesthesia within 515 minutes with limited systemic absorption (61). The anes-

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

377

thetic solutions used most often are a combination of lidocaine and epinephrine.
Sisler (62) iontophoresed anesthetic solutions containing either 4%
lidocaine with 1:1000 epinephrine or 2% lidocaine with 1:2000 epinephrine
to patients with lesions of the tarsus and tarsal conjunctiva prior to surgical excision of conjunctival plaques. A current 0.5 mA was applied for 10
minutes. Twenty-seven patients were treated. None of the patients
reported any discomfort in the eyelids or the arm to which the negative
electrode was attached; only a mild sensation was described. Three
patients with lesions in the deeper portion of the tarsus reported pain
and were given an injection to achieve local anesthesia. This is the only
report describing iontophoretic delivery of an anesthetic agent to adnexal
areas for pain prevention.
Meyer et al. (63) iontophoresed a 4% lidocaine solution to eyelids of
patients for local anesthesia prior to blepharoplasty or ptosis repair. A
current of 2 mA was applied for 12 minutes. Ten normal volunteers were
used so as to compare pain sensations after anesthesia by iontophoresis or
topical application of the anesthetic. Both surgical patients and volunteers
reported signicantly less pain after iontophoresis of the anesthetic. No side
eects of this iontophoresis procedure were observed.
3.

Ocular Inammation

Corticosteroids are the most common drugs used in treating ocular inammatory disorders (6467). Topical drop application is preferred to avoid the
serious systemic side eects of steroids (68). However, this mode of administration does not allow for sucient drug delivery to the posterior segment
of the eye. Iontophoretic delivery of anti-inammatory drugs into the eye
has been examined in human and various animal models and oers a viable
alternative to topical or systemic administration.
Lachaud (65) iontophoresed hydrocortisone acetate (0.1% solution)
into rabbit eyes with a current of 3 mA for 10 minutes. He demonstrated
that iontophoresis could deliver higher concentrations of steroid to rabbit
ocular tissue than either topical drops (0.5%) or subconjunctival injection
(0.1 mL, 2.5%). In human studies, Lachaud iontophoresed dexamethasone
acetate (7 mg%, 12 mA, 20 min) to treat a variety of clinical conditions,
including idiopathic uveitis. He reported that a signicant proportion of the
patients with uveitis beneted in terms of more rapid recovery and/or
increased comfort. Lachaud (65) concluded that iontophoresis resulted in
therapeutic concentrations of the steroid(s) in ocular tissue. However, it
must be noted that this open clinical study did not involve comparisons
with eyes receiving other therapies or with untreated control eyes.

Copyright 2003 Marcel Dekker, Inc.

378

Myles et al.

Lam et al. (66) iontophoresed a 30% dexamethasone solution transsclerally into rabbit eyes using 1.6 mA for 25 minutes. The diameter of the
cylinder holding the drug solution in contact with the sclera was 0.7 mm.
They compared peak steroid concentrations in the choroid-retinal tissue
following iontophoresis, subconjunctival injection (1 mg) or retrobulbar
injection (1 mg). The peak steroid concentration (mg/g tissue) for iontophoresis was 122, for subconjunctival injection 18.1, and for retrobulbar injection 6.6. In the vitreous humor the values were 140, 0.2, and 0.3 mg/mL,
respectively. Even at 24 hours after iontophoresis, signicant therapeutic
levels of dexamethasone remained 3.3 mg/mL in the vitreous and 3.9 mg/g
in the choroid-retina.
Behar-Cohen et al. (67) investigated the ecacy of iontophoretic delivery of dexamethasone for the treatment of endotoxin-induced uveitis in the
rat. Dexamethasone was delivered by concurrent transcorneal-transscleral
iontophoresis (1% at 0.4 mA for 4 min) using a 1 mL reservoir electrode
that covered the cornea, the limbus, and the rst millimeter of the sclera.
They showed that administration of dexamethasone by iontophoresis inhibited anterior and posterior signs of intraocular inammation (protein exudation, cellular inltration) as eectively as systemic administration.
Cytokine (TNF-a) (69) expression was inhibited in the anterior as well as
the posterior segment of the eye. No clinical or histological damage was
caused by iontophoresis. Thus, iontophoresis can deliver therapeutic doses
of this anti-inammatory drug to the posterior as well as the anterior segment of the eye and may be a viable alternative to systemic administration
of corticoids in severe ocular inammation.
4. Ocular Infection
Transcorneal iontophoresis of antibiotics is an eective means of treatment
for bacterial keratitis and other anterior segment infections. Transcorneal
iontophoresis delivers therapeutic concentrations of antibiotics to the cornea and aqueous humor. In phakic animals, the lens-iris diaphragm limits
penetration of the drug into the posterior tissues of the eye such as posterior
vitreous and retina. Transscleral iontophoresis circumvents the lens-iris barrier and delivers drugs into the vitreous or retina in amounts high enough to
be therapeutic in the treatment of posterior segment infections, such as
endophthalmitis. Numerous studies have documented successful iontophoretic delivery of various antibiotics into ocular tissues of animal models.
Examples are summarized below.
a. Gentamicin: Transcorneal and/or Transscleral The aminoglycoside gentamicin has a molecular weight of approximately 430 daltons, is
lipid insoluble, and bears two positive charges at physiologic pH (70). It

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

379

possesses bactericidal properties against a wide variety of gram-negative

and gram-positive bacteria (71). Gentamicin concentrations of 5 mg/mL
inhibit over 90% of Pseudomonas (71), Proteus rettgeri, P. vulgaris, and
P. morganii strains, as well as 99% of Staphylococcus strains (70). These
molecular features, combined with the extreme sensitivity of the target organisms, make gentamicin an ideal drug for iontophoresis.
Hughes and Maurice (72) found that transcorneal iontophoresis of
gentamicin in the rabbit eye increased permeability to the antibiotic more
than 100-fold, compared with control eyes in which topical application was
performed under the same iontophoretic conditions but with no current
applied. Grossman et al. (70) iontophoresed 10% gentamicin in a 2%
agar solution into rabbit corneas. They demonstrated that signicantly
higher and longer-lasting gentamicin concentrations were achieved in the
cornea with iontophoresis as compared to subconjunctival injection of a 20
mg dose. Frucht-Pery et al. (73) showed that higher current density did not
signicantly enhance antibiotic penetration into rabbit cornea, but bactericidal concentrations of gentamicin could be obtained. Frucht-Pery et al.
(74) also described the distribution of transcorneally iontophoresed gentamicin in the rabbit cornea. They found that the highest concentrations of the
drug were in the central cornea, while the midperipheral cornea(s) had
higher levels than peripheral cornea(s).
Fishman et al. (71) described iontophoresis of gentamicin to uninfected aphakic rabbit eyes. Gentamicin iontophoresis (50 mg/mL at 0.75
mA for 10 min) yielded peak corneal (71 mg/g of tissue) and aqueous
humor (78 mg/mL) concentrations 30 minutes after treatment. The peak
vitreous concentration (10.4 mg/mL) was observed 16 hours after treatment.
Therapeutic concentrations of gentamicin were still present in the vitreous
24 hours after iontophoresis. This study suggests that even transcorneal
iontophoresis has the potential to deliver high concentrations of gentamicin
to the posterior segment in aphakic eyes. Since many patients with
endophthalmitis are aphakic, transcorneal iontophoresis could be a suitable
route of administration of antibiotics for therapeutic management of this
disease.
Barza et al. (75) modied the standard transcorneal procedures to
achieve direct delivery of high concentrations of gentamicin into the vitreous
by transscleral iontophoresis. First, a reservoir holding the drug was placed
over the pars plana, thereby bypassing the lens-iris barrier. Second, the
contact area of the uid that delivered both the antibiotic and the current
was kept small (approximately 1 mm in diameter). They reported that transscleral iontophoresis delivered therapeutic concentrations (94207 mg/mL)
of gentamicin to the vitreous humor of uninfected rabbit eyes, thus obviating the need for intraocular injections.

Copyright 2003 Marcel Dekker, Inc.

380

Myles et al.

Another study from the same laboratory (76) demonstrated that transscleral iontophoresis of gentamicin is a useful adjunct to intravitreal injections for the treatment of endophthalmitis caused by pseudomonas
aeruginosa. They showed that rabbits receiving both intravitreal injection
and transscleral iontophoresis of gentamicin had lower bacteria counts at
each treatment interval compared to rabbits that received gentamicin by a
single intravitreal injection of a 100 mg dose. These results support the use of
iontophoresis of gentamicin as a useful supplement to intravitreal injection
for the treatment of bacterial endophthalmitis.
Grossman et al. (70) reported that transscleral iontophoresis of gentamicin produced results similar to their ndings with transcorneal iontophoresis described above. Iontophoresis of 10% gentamicin in 2% agar solution
at 2 mA for 10 minutes with a contact area of 2 mm in diameter delivered
very high concentrations of gentamicin to the vitreous humor of rabbit eyes.
Vitreous concentrations peaked (53.4 mg/mL) 16 hours after iontophoresis
and remained at inhibitory levels even at 24 hours. As with transcorneal
iontophoresis (70), no ocular tissue toxicity or damage was observed with
transscleral iontophoresis.
Burstein et al. (77) reported that transscleral iontophoresis of gentamicin into uninfected rabbit eyes resulted in antibiotic concentrations of 10
20 mg/mL in the vitreous humor. These concentrations are signicantly
lower than the concentration range (94207 mg/mL) reported by Barza et
al. (75) in rabbit eyes. The authors found that the total surface area of the
electrode is inversely proportional to the amount of antibiotic delivered,
e.g., a 4.5 mm2 applicator delivers approximately 20 times more drug to
the vitreous than a 28 mm2 applicator. This result reinforced the conclusion
of Barza et al. (75,76) that a small area of contact in iontophoresis results in
a higher concentration of drug in the eye.
Barzas research group (78) also reported the rst use of a nonhuman
primate (the cynomolgus monkey) to study the pharmacokinetics of transsclerally delivered gentamicin and/or potential histological damage of
transscleral iontophoresis. High and sustained concentrations of antibiotics
were achieved in the vitreous. Although small burns were observed in the
area of the pars plana where the electrode was applied, all electroretinograms were normal. The results suggest that transscleral iontophoresis is
well tolerated in the primate eye and that investigations with human eyes
may yield alternative treatment options. The absence of side eects with
the agar-based delivery system of Grossman et al. (70) is noteworthy. This
formulation may facilitate the use of transscleral iontophoresis as a treatment of choice for patients with bacterial endophthalmitis or other clinical
conditions that require high concentrations of antibiotics in the posterior
segment.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

381

b. Tobramycin: Transcorneal Transcorneal iontophoresis of the

aminoglycoside tobramycin has been examined by our group in studies
with the normal rabbit eye. Rootman et al. (79) were the rst to demonstrate the efcacy of iontophoresed tobramycin for the treatment of experimental Pseudomonas keratitis. Rabbit corneas were infected with 103
colony-forming units of P. aeruginosa. Transcorneal iontophoresis of a
2.5% tobramycin solution at 0.8 mA for 10 minutes was performed at 22,
27, and 32 hours after inoculation. On average, the treated corneas had a
6 log reduction in colony-forming units relative to untreated corneas. At
32 hours postinoculation, 67% of the corneas had no viable bacteria (i.e.,
were sterile). Topically applied or subconjunctival injection of tobramycin
did not yield corneas free of viable (infectious) Pseudomonas.
In other studies, Hill and associates examined the pharmacokinetics of
iontophoresed tobramycin and/or potential toxicity to the corneal epithelium of transcorneal iontophoresis (80,81). In uninfected, mock-infected,
and P. aeruginosainfected rabbit corneas, transcorneal iontophoresis produced high and sustained concentration of the antibiotic in the corneal
epithelium, corneal stroma, and aqueous humor (80,81). Iontophoresis
delivered ve times more drug than bathing the cornea with a 2.5% tobramycin solution and 20 times more the applying 1.36% tobramycin as fortied drops (80). No permanent abnormalities were observed by slit-lamp
biomicroscopy, scanning electron microscopy, or light microscopy (81).
After 10 minutes of iontophoresis, the epithelium showed focal edema
and disruption of all cell layers. Histological specimens obtained 8 and 16
hours after iontophoresis showed no defects in the corneal epithelium.
The ecacy of iontophoretically delivered tobramycin was examined
using a tobramycin-resistant strain of Pseudomonas (82). A strain of P.
aeruginosa with a minimum inhibitory concentration (MIC) for tobramycin
of 31 mg/mL was injected into the corneal stroma in rabbit eyes.
Transcorneal iontophoresis of 2.5% tobramycin resulted in a 3 log reduction in the number of bacteria. These results show that transcorneal iontophoresis can deliver concentrations of tobramycin high enough to combat a
clinically tobramycin-resistant strain of Pseudomonas.
c. Ciprooxacin: Transcorneal/Transscleral Ciprooxacin, a very
potent uoroquinolone antibiotic, is active against a broad spectrum of
gram-positive and gram-negative bacteria (83). Hobden et al. (84) used
transcorneal iontophoresis to deliver ciprooxacin to rabbit corneas infected with an aminoglycoside-resistant strain of Pseudomonas. Iontophoresis of 1% or 2.5% ciprooxacin reduced the number of colony-forming
units by more than 5 log relative to untreated controls. This level of inhibition was signicantly greater than either topically applied drops (0.75%

Copyright 2003 Marcel Dekker, Inc.

382

Myles et al.

ciprooxacin) or corneal bathing with an eyecup containing 2.5% ciprooxacin for 10 minutes. The concentration of ciprooxacin in the aqueous
humor following iontophoresis (84 mg/mL) was signicantly higher than
that with drops (30 mg/mL) or corneal bathing (25 mg/mL). This was the
rst report of iontophoresis of a uoroquinolone for therapy of experimental Pseudomonas keratitis (84).
Yoshizumi et al. (85) used transscleral iontophoresis to deliver ciprooxacin to the aqueous humor and vitreous body of the rabbit eye.
Ciprooxacin molecules carrying either a positive or negative charge were
tested. Therapeutic concentrations of antibiotic were achieved in the anterior chamber only when the negatively charged drug molecule was used.
Transscleral iontophoresis did not deliver therapeutic concentrations of
either charged form (+ or ) of the molecule to the posterior segment.
Peak concentrations were obtained in the aqueous (0.62 mg/mL) and vitreous bodies (0.19 mg/mL) one hour after transscleral iontophoresis of negatively charged ciprooxacin at 5 mA for 15 minutes. These concentrations
reached the MIC90 for several organisms that commonly cause endophthalmitis: Staphylocccus aureus (0.5 mg/mL), Staphylcoccus epidermidis (0.5 mg/
mL), and Pseudomonas aeruginosa (0.5 mg/mL) (85).
d. Vancomycin: Transcorneal/Transscleral Vancomycin, a complex
glycopolypeptide antibiotic with a high molecular weight (1448 daltons),
is very active against gram-positive bacteria. Choi and Lee (86) were the
rst to report that vancomycin could be applied to the eye by transcorneal
iontophoresis. A 5% solution of the antibiotic was iontophoresed at 0.5
mA for 5 minutes with an area of contact between the drug solution and
the cornea of 30 mm2. Peak antibiotic levels in the cornea (12.4 mg/mL)
were obtained 30 minutes after transcorneal iontophoresis. By comparison, a 25 mg subconjunctival injection resulted in a peak concentration of
14.7 mg/mL. This study demonstrated that iontophoresis could be used to
deliver therapeutic concentrations of a high molecular weight antibiotic to
the cornea and is a viable alternative to subconjunctival injection.
These investigators also studied transscleral iontophoresis for delivery
of vancomycin to the vitreous humor (86). The contact area was approximately 2530 mm2 of the temporal sclera overlying the pars plana. A 5%
vancomycin solution was iontophoresed at 3.5 mA for 10 minutes. Peak
antibiotic concentration (13.4 mg/mL) in the vitreous was reached 2 hours
after iontophoresis. Even at 16 hours postiontophoresis, the concentration
of vancomycin in the vitreous was still relatively high (3 mg/mL). This study
reported that bactericidal concentrations for gram-positive organisms isolated from patients with endophthalmitis were less than or equal to 4 mg/mL
(86). Thus, iontophoresis clearly delivered enough antibiotic to be therapeu-

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

383

tically useful. This is the rst report of transscleral iontophoresis of a high

molecular weight glycopolypeptide antibiotic.
e. Cefazolin and Ticarcillin: Transscleral Barza et al. (75), using the
modications previously described, also reported that therapeutic concentrations of cefazolin and ticarcillin could be delivered to the vitreous humor of uninfected rabbit eyes by transscleral iontophoresis instead of
intraocular injection. Cefazolin, a rst-generation cephalosporin, has excellent gram-positive and gram-negative antibacterial activity. Ticarcillin,
the disodium salt of penicillin, is particularly useful in treating P. aeruginosa infections. Cefazolin and ticarcillin bear negative charges (one and
two, respectively) and were applied with the negative electrode in the antibiotic solution. Application of 2 mA for 10 minutes yielded mean vitreal
concentrations of 94207 mg/mL in the normal rabbit eye. Drug penetration correlated with strength of current (0.12 mA) and duration of iontophoresis (110 min). A two- to fourfold variation in the concentration of
the drug solution did not alter the total amount of drug delivered under
constant current and time. The authors concluded that the concentrations
produced in the rabbit eye are in the range expected from intravitreal injection in humans. However, the volume of vitreous humor in an adult
human eye is about three times that of the rabbit eye. Therefore, the expected peak concentrations in humans would be about one third of those
obtained in the rabbit.
f. Ketoconazole: Transcorneal/Transscleral Ketoconazole is a relatively nontoxic, broad-spectrum, imidazole antifungal agent that is used in
the treatment of various forms of ocular fungal keratitis, including Candida albicans, histoplasmosis, and Cryptococcus neoformans. Grossman et al.
(87) performed both transcorneal and transscleral iontophoresis of ketoconazole and measured the resulting drug concentrations in the aqueous and
vitreous humor as well as the cornea of the rabbit eye. Transcorneal iontophoresis (1.5 mA for 15 min) resulted in peak corneal ketoconazole concentrations of 27.6 mg/mL and peak aqueous concentrations of 1.4 mg/mL
1 hour posttreatment. These drug levels are inhibitory for many fungal
pathogens and transcorneal iontophoresis of ketoconazole may be an effective means of treating anterior segment fungal infections.
Transscleral iontophoresis (46 mA for 15 min) resulted in peak aqueous ketoconazole concentrations of 10.2 mg/mL 1 hour after iontophoresis
and fungicidal levels were present even after 8 hours. Both the aqueous and
corneal concentrations of ketoconazole were signicantly higher after iontophoresis than after subconjunctival injection. Transscleral iontophoresis
resulted in peak vitreal ketoconazole concentrations of 0.1 mg/mL 12 hours
posttreatment. By comparison, a subconjunctival injection of ketoconazole

Copyright 2003 Marcel Dekker, Inc.

384

Myles et al.

produced a peak concentration of 0.7 mg/mL 1 hour after injection. Since

neither of these values is sucient for therapeutic fungicidal activity, the
treatment of endophthalmitis by transscleral iontophoresis or subconjunctival injection of ketoconazole is questionable at best.
5. Antiviral Agent for Treatment of Cytomegalovirus Retinitis
Human cytomegalovirus is a member of the herpesvirus family, which also
includes the herpes simplex (discussed below), varicella zoster, and EpsteinBarr viruses (88,89). Cytomegalovirus (CMV) and herpes simplex are the
most studied of the microorganisms responsible for the opportunistic infections associated with acquired immunodeciency syndrome (AIDS) (90).
Cytomegalovirus infection in AIDS patients is most commonly manifested
as retinitis (CMV retinopathy) and is the most common cause of blindness
in AIDS patients (90).
Foscarnet (trisodium phosphonoformate) inhibits all human herpes
viruses in vitro, including CMV (91), and intravenously administered foscarnet is currently available for treatment of CMV infections and/or retinitis
(92,93). The treatment regimen for CMV retinitis consists of an initial induction dose (180 mg/kg of body mass per day for 2 weeks) followed by a
maintenance dose (120 mg/kg/day indenitely) (92,93). Since foscarnet prevents replication of the viral DNA but does not eliminate the virus from the
tissues, CMV eventually reactivates and the lesions enlarge so that higher
doses of the drug have to be administered. This mode of administering
foscarnet may result in serious systemic toxicity (92,93). Intravitreal injection of foscarnet has been reported but also suers from many of the same
complications (endophthalmitis, increased intraocular pressure, retinal
detachment, etc.) associated with other modes of ocular injection (9496).
Foscarnet is an ideal candidate for iontophoresis, since it is ionized at the
physiological pH of the eye, is soluble in water, and has a molecular weight
of 300.1 (91,97).
Sarraf et al. (97) studied the vitreous pharmacokinetics of foscarnet
(24 mg/mL solution) after transscleral iontophoresis into normal rabbit
eyes. The iontophoretic probe was applied perpendicular over the conjunctiva 13 mm posterior to the corneoscleral limbus. The surface area of the
probe was 0.2 mm2. The current applied was 1.0 mA for a 10-minute
duration. Intravitreal samples were obtained at 12 time points (15 and 30
min and 1, 2, 4, 8, 16, 24, 32, 40, 48, and 60 h) after iontophoresis or
subconjunctival injection and analyzed by HPLC. Mean vitreous concentrations of foscarnet were within the therapeutic range (25800 mM) (91) for
inhibition of CMV as early as 15 minutes posttreatment. A peak vitreal
foscarnet concentration of 200  311 mM was observed 4 hours after ionto-

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

385

phoresis. This concentration is well below the concentration reported to

cause retinal toxicity (>2000 mM) (94). Therapeutic levels of foscarnet for
CMV retinitis were maintained for up to 60 hours posttreatment. These
levels were well above the range (1025 mM) at which foscarnet is active
against HIV (93). No toxic eects to the sclera, cornea, anterior chamber, or
lens were observed by biomicroscopy.
Yoshizumi et al. (98) characterized the utility of foscarnet iontophoresis for the treatment of CMV retinitis. The possible retinotoxic eects of
transscleral iontophoresis of foscarnet were examined by slit-lamp biomicroscopy, indirect ophthalmicroscopy, light and electron microscopy, and electroretinography (ERG) (98). Slit-lamp examination (SLE) revealed no toxic
eects for any of the treated eyes. Indirect ophthalmicroscopy showed retinal and choroidal burns 13 mm in diameter at the site of iontophoresis.
Light and electron microscopy showed focal retinal, retinal pigment epithelial, and choroidal necrosis at the site of iontophoresis but no abnormalities
elsewhere. ERG studies showed no changes in the response between foscarnet-treated eyes and saline-treated control eyes. The lesions reported were
minimal and judged to be equivalent to or less than those that would have
been created by an intravitreal injection or surgical implantation of an
intraocular drug delivery device (99).
Another study from Yoshizumis group (100) demonstrated that ocular iontophoresis of foscarnet could supplement intravenous injection of
foscarnet in the management of CMV retinopathy. Foscarnet (120 or 180
mg/kg) was injected intravenously into one group of rabbits and peak concentrations in the serum and vitreous were determined at 1, 4, 8, 24, 60, and
120 hours after injection. A second group of rabbits, injected with the same
dose of foscarnet, received ocular foscarnet iontophoresis 1 hour after the
injection. Vitreous humor samples were obtained at the same time points as
for group one. Maximum serum and vitreous humor concentrations were
obtained 1 hour after each intravenous dose. Peak vitreous humor concentrations were obtained 4 hours after the 120 mg/kg intravenous dose plus
iontophoresis and 8 hours after the 180 mg/kg intravenous dose plus iontophoresis. Vitreous humor levels of foscarnet were signicantly higher in eyes
receiving intravenous foscarnet plus ocular foscarnet iontophoresis than in
those receiving intravenous foscarnet alone (126.77 mM vs. 7.63 mM for the
group receiving 120 mg/kg; p < 0:00001 and 86.92 mM vs. 7.43 mM for the
group receiving 180 mg/kg; p < 0:00001). The dierence in vitreous humor
levels observed in eyes receiving the intravenous dose plus iontophoresis
(either 120 or 180 mg/kg) was not signicant (p < 0:1). Although the intravenous dose was administered 1 hour prior to ocular iontophoresis, it did
not signicantly aect the levels obtained after ocular iontophoresis
p < 0:1. The authors concluded that the increased foscarnet concentra-

Copyright 2003 Marcel Dekker, Inc.

386

Myles et al.

tions obtained with ocular iontophoresis could be a valuable adjunct to

intravenous injection in the treatment of CMV retinitis.
A subsequent study from Yoshizumi and associates examined the
ocular toxicity of multiple applications of foscarnet iontophoresis (101).
They performed multiple ocular iontophoretic applications of foscarnet at
the same paralimbal site for a total of seven treatments over a period of 21
days (one treatment every 3 days). This was done to mimic the repeated
treatments that may be necessary for long-term iontophoretic therapy for
treatment of CMV retinitis and/or AIDS (92,93). The mean concentration
of foscarnet in the vitreous humor 4 hours after the seventh ocular iontophoresis was 189  50:6 mM. These levels are clearly within the therapeutic
range (25800 mM) needed for treatment of CMV retinitis (91). Moreover,
the values are comparable to those obtained in eyes treated with only a
single foscarnet iontophoresis. Electroretinography and slit-lamp examination revealed no evidence of ocular toxicity. Examination of the retinas and
choroid revealed a single small burn in the retina and choroid corresponding
to the iontophoresis probe contact site, which was similar to the lesion
resulting from a single iontophoretic application (98). These results suggest
that repeated ocular iontophoresis of foscarnet could be an eective means
of achieving enhanced, localized treatment of CMV retinitis in humans.
6. Herpes Simplex Virus Infection
Herpes viruses are among the most well-studied microorganisms that cause
chronic infection and are a well-described cause of ocular diseases (89,102).
Ocular degeneration due to infection with herpes simplex type 1 (HSV-1)
and herpes simplex type 2 (HSV-2) is a leading cause of blinding keratitis in
industrialized countries (102,103). Analysis of ocular viral isolates demonstrated that HSV-1 is responsible for about 85% of ocular HSV infections
(102,103). Ocular HSV-1 infection consists of an acute keratoconjunctivitis
followed by a lifelong cycle of latency, reactivation, and recurrent infection
(102,103). This section reviews research on herpetic eye infection and the use
of iontophoresis to deliver therapeutic doses of antiviral agents and/or
agents that cause experimental reactivation of latent HSV-1.
a. Antivirals for Treatment of Epithelial Keratitis Hill et al. (104)
were the rst to report the transdermal delivery of antiviral agents by iontophoresis. They demonstrated that the antiviral agents iododeoxyuridine
(IDU), phosphonoacetic acid (PPA), and vidarabine monophosphate
(Ara-AMP) could be iontophoresed across mouse epidermis with enough
efciency to achieve long-lasting and potentially therapeutic concentrations of drug.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

387

A subsequent study (105) examined the pharmacokinetics of iontophoresed Ara-AMP. Ara-AMP was chosen because the parent compound,
vidarabine, was shown to be eective in the treatment of HSV-1 infection
after topical or systemic administration. Ara-AMP is phosphorylated and
highly charged and thus an excellent candidate for iontophoresis. Tritiumlabeled Ara-AMP was applied either topically or iontophoresed into uninfected rabbit eyes. Transcorneal iontophoresis was performed at 0.5 mA for
4 minutes with the cathode in contact with the drug solution. Iontophoresis
resulted in drug concentrations in the cornea, aqueous humor, and iris
which were 312 times higher than those obtained with topical application.
No obvious corneal damage was observed. The results showed that iontophoresis increased corneal penetration of an antiviral agent in comparison
to topical application.
Kwon et al. (106) were the rst to demonstrate that iontophoresed
Ara-AMP retained its antiviral properties. They studied the virucidal eects
of iontophoretically applied Ara-AMP on experimental HSV keratitis in
rabbits infected with HSV-1 McKrae strain. Transcorneal iontophoresis
of Ara-AMP (0.3 M, 3.4%) was done in both eyes at 24, 48, and 72
hours postinoculation. A second group of infected rabbits served as a control for the procedure and were iontophoresed with a 0.9% sodium chloride
solution. Additional groups of rabbits received either topical 10% AraAMP, 0.5% IDU, or 0.9% sodium chloride ve times daily for 4 days.
Slit-lamp examination was performed daily for 10 consecutive days after
initiation of treatment and the severity of the disease was scored. Mean
lesion scores for the eyes treated by iontophoresis of Ara-AMP were signicantly lower (i.e., less severe disease of shorter duration) compared to the
scores of eyes treated with iontophoresis of sodium chloride or topically
with Ara-AMP, IDU, or sodium chloride.
b. Antivirals for Treatment of Stromal Disease Hill et al. (107) extended these studies by comparing the efcacy of transcorneal iontophoresis versus intravenous injection of either acyclovir or Ara-AMP (alone or
in combination) for the treatment of HSV-1 stromal infection in rabbits.
Herpetic infection was achieved by an intrasomal injection of puried
HSV-1 McKrae strain. Treatments (iontophoresis of either Ara-AMP,
acyclovir, or sodium chloride; or intravenous infusion of acyclovir or sodium chloride) were initiated 24 hours later. Two ophthalmologists performed slit-lamp examination daily for up to 22 days postinoculation and
recorded the severity of the disease.
Iontophoresis itself did not have any eect on the severity of the
disease. Iontophoresis of acyclovir or Ara-AMP signicantly shortened
the duration of the disease compared to iontophoresis of sodium chloride,

Copyright 2003 Marcel Dekker, Inc.

388

Myles et al.

and iontophoresed Ara-AMP was as eective as iontophoresed acyclovir.

Acyclovir signicantly reduced the severity of the disease whether it was
delivered iontophoretically or by intravenous injection. Intravenous administration of acyclovir resulted in signicantly higher levels of drug in ocular
tissues than that obtained with iontophoresis, but iontophoresis delivered a
dose that achieved the same therapeutic eect. These results suggested that
iontophoresis of acyclovir or Ara-AMP (with or without intravenous supplementation) could be a valuable treatment option for patients with HSV-1
stromal keratitis.
c. Adrenergic Agents for Experimental Reactivation of HSV-1 in the
Study of Ocular Herpetic Disease Primary HSV infection is an acute inammatory response. While the acute infection is occurring, HSV travels
along neuronal axons to regional sensory ganglia where it enters a state
of dormancy (i.e., latency) (108110). The exact stimulus that triggers
HSV reactivation is unknown, but certain environmental factors (stress,
irradiation, hypo- or hyperthermia) result in the appearance of infectious
virus at the site of primary infection and a recurrence of the disease (110
113). The rabbit eye model has allowed investigators to study the pathogenesis of HSV latency and reactivation (102,114). In this model, viral latency can be established by inoculation of the cornea with 17Syn+,
McKrae, or other strains of HSV-1 (102,115,116). Viral reactivation can
be induced by iontophoresis of adrenergic agents and monitored by viral
shedding on the ocular surface (102,116,117).
Clinicians had noted for some time the relationship between physical
stress and the appearance of fever blisters (cold sores) or genital herpetic
lesions. This suggested a role for epinephrine in the process, and various
investigators showed that topical or intravenous applications of epinephrine
could cause reactivation of HSV-1 in rabbits. Researchers from Hills group
(118,119) were the rst to report that transcorneal iontophoresis of this
agent could induce viral shedding in rabbits harboring latent HSV-1.
They inoculated rabbits with HSV-1 and 60 days later performed epinephrine iontophoresis on one eye from each rabbit. A 0.01% epinephrine solution was iontophoresed (0.8 mA for 8 min) once daily for 3 consecutive
days. Viral shedding was veried by ocular swabbing and culturing the
tear lm in cell culture. Antigenic and/or nucleic acid analysis was used to
conrm the identity of the viral isolate. Epinephrine iontophoresis resulted
in HSV-1 viral shedding in all treated eyes (118). A follow-up report from
this group (119) validated these results and demonstrated that the HSV-1
titers rose, peaked, and fell with time after iontophoresis. This reliable and
ecient means of inducing viral shedding served as the basis for the development of other animal models for the study of herpetic disease (120123).

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

389

These models have been used to study the kinetics, pathogenesis, and molecular biology of HSV-1 latency, reactivation, and recurrence of clinical disease.
A subsequent study from this group (124) characterized the appearance of HSV-1 in the nerves and ganglia after reactivation by epinephrine
iontophoresis. Hill et al. (124) showed that the presence of infectious virus
could be detected more rapidly from cultured neuronal tissue from rabbits
that had undergone epinephrine iontophoresis than from their untreated
counterparts.
Shimomura et al. (125) modied the epinephrine iontophoresis protocol so as to eliminate the need to anesthetize animals once daily for 3
consecutive days. They developed a method that was based on a single
iontophoresis of 6-hydroxydopamine followed by topical application of epinephrine over several days. The rationale for this came from studies using
iontophoresed 6-hydroxydopamine to treat glaucoma. As described above,
6-hydroxydopamine causes a selective and reversible degeneration of sympathetic nerve terminals in the anterior segment, after which the innervated
structures of the eye are exquisitely sensitive to extremely dilute solutions of
epinephrine. With this method, viral shedding was induced by a single iontophoresis of a 1% solution of 6-hydroxydopamine under various conditions (0.50.75 mA for 38 min). Two drops of 2% epinephrine were applied
6 hours after iontophoresis and twice daily for the next 4 days. All treated
eyes (17/17; 100%) shed HSV-1, regardless of the iontophoretic conditions.
Hill et al. (115,126) used this modied procedure to further characterize ocular HSV-1 shedding induced by 6-hydroxydopamine iontophoresis
plus topical epinephrine. They were the rst to quantify the number of
plaque-forming units of HSV-1 shed into the tear lm following 6-hydroxydopamine iontophoresis plus topical epinephrine-induced reactivation
(126). Titers ranged up to 105 plaque-forming units/eye. A subsequent study
(115) showed that dipivefrin hydrochloride could be used in place of epinephrine in this reactivation model. Dipivefrin hydrochloride is a prodrug of
epinephrine, which has increased corneal penetration compared to epinephrine. In clinical studies, 0.1% dipivefrin hydrochloride was as eective as
topical 1 or 2% epinephrine. Hill et al. (115) demonstrated that iontophoresed 6-hydroxydopamine followed by topical application of 0.1% dipivefrin
hydrochloride resulted in HSV-1 ocular shedding and recurrent HSV-1 corneal lesions. This was the rst report showing that an adrenergic drug could
induce both HSV-1 ocular shedding (reactivation) and HSV-1 corneal
epithelial lesions (recurrence) in rabbits harboring latent virus.
Rivera et al. (110) used epinephrine iontophoresis to study the temporal relationship between viral reactivation and the presence of viral particles or virions in corneal nerves or the cornea, respectively. Transmission

Copyright 2003 Marcel Dekker, Inc.

390

Myles et al.

electron microscopy revealed viral particles (in low abundance) in unmyelinated axons, but no enveloped virions were found. This study suggests that
ocular iontophoresis of epinephrine reactivates HSV-1 in the ganglia and
that the virus is translocated from the sensory ganglia (trigeminal and/or
superior cervical) to the cornea by anterograde axonal transport mechanisms.
Hill et al. (117) demonstrated that levo(-)epinephrine was signicantly
more potent than dextro(+)epinephrine for inducing HSV-1 ocular shedding. The data suggested that the mechanism of induction of HSV-1 ocular
shedding by epinephrine is correlated with the receptor potency of
levo(-)epinephrine. Epinephrine activates both a- and b-adrenergic receptors
in the eye. If the signal for viral reactivation was transmitted exclusively
through a b-adrenergic receptor, a b-adrenergic receptor antagonist should
inhibit epinephrine-induced reactivation.
Studies from Hills group (127129) assessed the eect of topically
applied timolol maleate, a nonspecic b1 ,b2 -adenergic receptor blocking
agent, on ocular HSV-1 reactivation in rabbit eyes. Timolol maleate was
applied following iontophoresis of 6-hydroxydopamine or by direct transcorneal iontophoresis. It was discovered that timolol iontophoresis for 3
consecutive days could cause corneal epithelial lesions (127). However, a
single direct transcorneal iontophoresis of timolol induced ocular HSV-1
shedding (128). The data suggest that both timolol (a b-adrenergic receptor
antagonist) and epinephrine (an a,b-adrenergic receptor agonist) could
induce ocular HSV-1 shedding. The results of these studies demonstrated
that specic receptor occupancy of epinephrine alone is not an exclusive
signal for viral reactivation. We are unable to explain why iontophoresis
of this agonist/antagonist pair of adrenergic agents induces viral reactivation in animals that harbor latent HSV-1 strain McKrae.
Studies with propranolol, also a b-adrenergic antagonist, have yielded
paradoxical results. Kaufman et al. (130) showed that propranolol suppressed both ocular HSV-1 recurrence and severity following spontaneous
reactivation in the rabbit. Gebhardt and Kaufman (131) demonstrated that
propranolol blocked the HSV-1 reactivation following hyperthermic induction in latently infected mice. Garza and Hill (132) examined the eect of
propranolol on HSV-1 reactivation in latently infected rabbits following
induction either by epinephrine iontophoresis (stress pathway) or by systemic immunosuppression (nonstress pathway). Immunosuppression was
produced by injection of cyclophosphamide and dexamethasone (133).
Propranolol was given intramuscularly by injection in doses of 5, 20, or
200 mg/kg twice daily. Propranolol had no eect at any concentration on
blocking HSV-1 induction, either by transcorneal iontophoresis of epinephrine or by systemic immunosuppression due to cyclcophosphamide plus

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

391

dexamethasone treatment. These results suggest that (a) propranolol can

specically block the induction pathway in the mouse model but not the
rabbit model, or (b) propranolol is not a broad/potent enough inhibitor to
prevent HSV-1 induction in the rabbit.
d. Epinephrine Iontophoresis and the Study of LATs Initial studies
in murine models of HSV infection demonstrated the presence of novel
HSV RNA transcripts in sensory neurons of mice harboring a latent HSV
infection (134136). These mRNA species resulted from transcription of
the only region of the HSV genome that is transcriptionally active during
the latent phase of HSV infection (134,137139). Latency-associated transcripts (LATs RNA) have also been demonstrated in rabbits (140,141)
and humans (135,142,143). Three major LATs (2, 1.55, and 1.45 kb) as
well as a minor transcript (8.3 kb) have been detected in ganglia
latently infected with HSV-1 (134,136,144146). The major LATs are
highly expressed and poly(A) , whereas the minor LAT has low level
expression and is poly(A)+. LATs are expressed at similar levels in sensory neurons of mice, rabbits, and humans (136,140,142). Although LAT
is not essential for latency, it remains the only known molecular marker
for latent HSV infection (134,139,147,148).
Epinephrine iontophoresis and/or the rabbit eye model have been used
in conjunction with various mutant strains of HSV to examine the role of
LATs in the process of establishment, maintenance, and reactivation of
latent HSV infection. HSV constructs that contain genetic alterations in
the LAT locus were compared to unmodied (wild-type) or marker rescued
(LAT-positive) viruses for their ability to establish latency and/or reactivate
from latency (134,136). Hill et al. (149,150) used mutant HSV-1 strains
(X10-13 or 17
Sty) that could not produce a latently associated transcript
(LAT-negative) and genetically engineered LAT-positive mutants (XC-20 or
17
Sty-Res) to demonstrate that the LAT-positive strains, which are similar to the parent HSV-1 strains 17Syn+, could be reactivated by iontophoresis with high eciency. The LAT-negative strains showed very limited or
signicantly reduced reactivation. The results suggested a role for the
latently associated transcript in viral reactivation. All the mutants (LATnegative or LAT-positive) were able to establish and maintain latency in
sensory ganglia (149,150).
Bloom et al. (151) demonstrated that HSV-1 mutants with deletions
encompassing the LAT promoter area have signicantly reduced ocular
reactivation following transcorneally iontophoresed epinephrine induction.
17
Pst (LAT-negative) has a 202 bp deletion in the LAT promoter in an
area encompassing the TATA box, cAMP response element (CRE),
upstream stimulatory factor, and transcription start sites, which signi-

Copyright 2003 Marcel Dekker, Inc.

392

Myles et al.

cantly reduces LAT transcription (147,151). They demonstrated that

although this virus established latency with the same kinetics as its parent,
17Syn+, spontaneous and adrenergically induced reactivation is signicantly reduced. This evidence suggested that expression of LATs was important in regulating HSV-1 reactivation. Bloom et al. (152) also examined the
LAT promoter region at a more precise level. HSV-1 mutants containing a
mutated CRE-1 element were shown to be signicantly impaired in their
ability to reactivate either spontaneously or by induction following epinephrine iontophoresis in the rabbit eye model (152). This evidence suggested a role for CRE-1 (CRE-1 binding factors) (134,136,153) in the
reactivation process.
Similar studies have provided evidence that viral strains exhibit varying sensitivities to genetic alterations within the LAT domain. Loutsch et al.
(154) used mutant strains of McKrae and 17Syn+ that had identical 371base-pair deletion mutations in the LAT genes to examine spontaneous in
vivo reactivation kinetics. They were able to demonstrate that an identical
deletion resulted in dierent in vivo reactivation phenotypes. The results
suggest a dierence in genetic background of McKrae and 17Syn+ resulting
in dierent in vivo reactivation phenotypes.
Zheng et al. (155) recently investigated the ability of LAT-positive and
LAT-negative strains of HSV-1 to establish latent infection in rabbit corneas
following penetrating keratoplasty. 17
Pst (LAT-negative, low reactivation) and 17Pr (LAT-positive, high reactivation recusant of 17
Pst) were
inoculated into rabbit corneas. Latently infected rabbits received corneal
allografts from naive rabbits, and naive rabbits received grafts from latently
infected rabbits. Penetrating keratoplasty of latently infected and naive
rabbits made it possible to study the migration of HSV from a site of latency
to other tissues. They demonstrated that corneas from latently infected
rabbits contain HSV-1 DNA that can replicate after viral reactivation by
transcorneal epinephrine iontophoresis. The LAT negative strain also had a
signicantly reduced ability to replicate. The data showed that HSV-1 in
latently infected rabbit sensory ganglia could be induced by epinephrine
iontophoresis and migrate into other uninfected tissues (i.e., the transplanted cornea of a naive rabbit). The LAT-negative strain had a reduced
ability to migrate. The results suggest that HSV-1 can migrate in both
anterograde and retrograde directions between the site of viral latency
(the trigeminal ganglion in the rabbit) and corneal tissues following epinephrine-induced reactivation. This is the rst report to provide evidence
to support the concept of corneal HSV-1 latency and reactivation (156,157).
Numerous other studies have used iontophoresis to study viral reactivation, as well as the treatment and prevention of recurrent herpetic disease
(158164). In summary, these studies validate the accuracy, reliability, and

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

393

reproducibility of iontophoresis as a noninvasive mode of drug delivery and

suggest that the iontophoretic model may be benecial in the development
of new, designer drugs to replace or supplement current herpetic disease
regimens. Experimentally, ocular iontophoresis and the rabbit eye model
are mainstays in the study of HSV-1 latency and reactivation.

III.

COMMERCIALLY AVAILABLE IONTOPHORETIC

DEVICES

A variety of iontophoretic devices are available with dierent types of power

sources that vary in complexity from the simple battery-rheostat type to
modern electronic circuitry (8,9). Table 3 provides a list of companies that
produce iontophoresis equipment, their addresses and phone numbers, as
well as a brief description of one specic model manufactured by each
company. The commercial devices are safe, portable, and work well for
the intended purpose.

IV.

CONCLUSIONS

The last 1020 years have seen the development and optimization of the
technology of iontophoresis as a noninvasive method of drug delivery.
Peptide and protein drugs (e.g., insulin and vasopressin) have been delivered
iontophoretically without the pain and side eects associated with traditional subcutaneous injection (10,165,166). Iontophoresis also has the
potential to revolutionize clinical diagnosis and monitoring procedures in
that it oers an avenue for extracting information from the body without
the need for invasive procedures (3,167,168). Iontophoresis is already used
clinically in physical therapy clinics, and an allergy patch test is close to
commercial reality (33,169).
Ocular iontophoresis is fast, painless, safe, and, in most cases, results
in the delivery of a high concentration of the drug to a specic ocular site.
Experimentally, iontophoresis has proven extremely useful as a reliable system for inducing reactivation of herpes simplex virus in various animal
models of this ocular disease. Clinically, ocular iontophoresis is potentially
a valuable adjunctive treatment for cytomegalovirus retinitis. In glaucoma
studies, iontophoresis of uorescein, adrenergic agents, and 5-uorouracil
has facilitated the study of aqueous humor dynamics, treatment of glaucoma, and the control of cellular proliferation of glaucoma surgery.
Anesthetics and antibiotic/antifungal agents can be delivered transcorneally
or transsclerally according to the site of infection. Ocular iontophoresis has

Copyright 2003 Marcel Dekker, Inc.

394
Table 3

Myles et al.
Iontophoresis Equipment and Suppliers

Company

Model number/description

IOMED, Inc.
2441 South 3850 West, Suite A
Salt Lake city, UT 84120
(801) 975-1191
(800) 621-3347
Fax (801) 975-7366

Phoresor II Auto model number PM900. This

model provides an eective, safe, and easy to
use system with a preset dose (40 mAmp-Mins)
for iontophoretic drug application to the skin.
This model is compact, easy to operate, fast to
set-up, and portable.
E-mail: es@iomed,com
Website: www.iomed.com

Life-Tech, Inc.
4235 Greenbrier Drive
Staord, TX 77477
(713) 495-9411
(800) 231-9841
Fax (281) 492-6646

Iontophor-II model number 6111 PM/DX. This

model provides a painless, safe, and eective
method of delivering high concentrations of
drug into a localized skin area. This unit is
battery operated, compact, and portable. It is
an excellent unit for maintaining constant
current. This model has dual site treatment
function that enables you to treat two sites
simultaneously.
E-mail: [email protected]
Website: www.life-tech.com

General Medical Company

1935 Armacost Avenue
Los Angeles, CA 90025
(310) 820-5881
(800) 432-5362
Fax (310) 826-5778

The Drionic unit. This compact iontophoresis

unit has FDA approval for home treatment of
hyperhidrosis. Battery-generated direct current.
E-mail: [email protected]
Website: www.drionic.com

Wescor, Inc.
459 South Main Street
Logan, UT 84321
(435) 752-6011
(800) 453-2725
Fax (435) 752-4127

Macroduct model 3700. This model is used for

sweat stimulation (pilocarpine iontophoresis),
collection and analysis (via a sweat analyzer). It
is easy to operate and provides accurate and
reliable sweat testing.
E-mail: [email protected]
Website: www.dataloggersinc.com

Fischer Co., Inc

517 Commercial Street
Glendale, CA 91203
(818) 241-1178
(800) 525-3467
Fax (323) 245-2748

Fischer Galvanic Unit, Model MD-1a is a

high-output iontophoretic device ideally suited
for treatment of large areas (hyperhidrosis
patients). It produces DC current in two ranges
(010mAmp and 050mAmp). Available to
physicians or by prescription only.
E-mail: info@rascher. com
Website: www.rascher. com

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis
Table 3

395

Continued

Company

Model number/description

Dagan Corporation
2855 Park Avenue
Minneapolis, MN 55407
(612) 827-5959
Fax (612) 827-6535

Dagan 6400 Advanced model. MicroIontophoresis current generator. Delivers

precise drug delivery without batteries. Power
is 100 to 115 volts AC. Current range is 01000
n/Amp.
E-mail: sales@dagan. com
Website: www.dagan.com

MedTherm Corporation
2604 Newby Road
Huntsville, AL 35805
(205) 837-2000

Electro-Medicator Model A1 (Distributed by

Pamir Electronics corp.: 603 N. Gordon Drive,
Exton PA 19341; (610) 594-8537). This model
is used for dental research and clinical practice.
It is easy to operate and has AC charging or
direct use of AC. E-mail: [email protected]
Website: www.pamir.com

Parkell
155 Schmitt Boulevard
Farmingdale, NY 11735
(631) 249-1134
(800) 243-7446
Fax (631) 249-1242

Desensitron II model number ID643DGC. This

model is used for treating hypersensitive teeth
by iontophoresis of uoride ions. Current range
is 00.5mAmp.
E-mail: [email protected]
Website: www.parkell.com

MedTronic, Inc.
7000 Central Ave., NE
Minneapolis, MN 55432
(612) 574-4000
(800) 525-5552
Fax (612) 785-6541

Model 9820. This iontophoretic device is part

of complete system (Model 9800) for sweat
analysis after pilocarpine iontophoresis.
E-mail
Website: www.medtronic.com

also been used to introduce genes into the anterior and posterior segments of
the rabbit eye (170).
There are several ongoing ocular iontophoresis studies that are particularly pertinent to this discussion and should be noted. Chapon et al. (171),
using rabbits, have shown that transscleral iontophoresis delivered 15 times
more of the anti-CMV agent ganciclovir to the retina and 6.5 times more to
the choroid (6 h posttreatment 370 and 828 ng/mg, respectively) than topical
administration. At 3 days posttreatment, retinal levels of ganciclovir
remained signicantly higher than those achieved by oral administration,
intravenous injection, topical application, or intravitreal injection.
Chauvaud et al. (172) presented initial ndings of a Phase II clinical trial

Copyright 2003 Marcel Dekker, Inc.

396

Myles et al.

for transscleral iontophoresis of the anti-inammatory corticosteroid

SoluMedrol (hemisuccinate methylprednisolone). Transscleral iontophoresis of SoluMedrol was safe, well tolerated, and easily applicable to the
treatment of severe ocular inammation thereby reducing the systemic
side eects of corticotherapy. Hayden et al. (173) studied transscleral iontophoresis of the anticancer drug carboplatin in transgenic mice with hereditary retinoblastoma. A dose-dependent inhibition of intraocular tumor
formation was observed following repetitive iontophoretic treatment. These
results suggest that transscleral iontophoresis of carboplatin may be useful,
as well as advantageous, in the treatment of intraocular pediatric retinoblastoma. Ocular iontophoresis is not a panacea for all eye disorders, but it
is a viable alternative delivery system for those substances that are not
amenable to topical application and/or require repeated administration
over an extended period of time.

ACKNOWLEDGMENTS
Louis P. Gangarosa, D.D.S., Ph.D., is acknowledged as the person who
introduced iontophoresis to one of the authors (JMH). The authors wish
to acknowledge the editorial assistance and research collaboration of A. K.
Mitra, Ph.D. Specic research from the laboratory of JMH was supported
in part by U.S. Public Health Service grant EY06311 and EY08871;
EY02377 (Eye Center Core Grant); an unrestricted grant from Research
to Prevent Blindness (RPB); Dr. Hill is an RPB Scientic Investigator
Award recipient. The authors wish to acknowledge the secretarial assistance
of Mrs. Carole Hoth. The authors also wish to acknowledge the student
workers with special thanks to Ms. Kristina Braud and Ms. Kathy Vu.
None of the authors have any nancial or proprietary interest in any agents
or devices mentioned in this review.

REFERENCES
1. Sing, P., and Maibach, H. I. (1994). Iontophoresis in drug delivery: Basic
principles and applications. Crit. Rev. Therp. Drug Carrier Systems, 11:161
213.
2. Guy, R. H., Kalia, Y. N., Delgado-Charro, M. B., Merino, V., Lopez, A., and
Marro, D. (2000). Iontophoresis: electrorepulsion and electroosmosis. J.
Control Release, 64:129132.
3. Merino, V., Kalia, Y. N., and Guy, R. H. (1997). Transdermal therapy and
diagnosis by iontophoresis. Trends Biotechnol, 15:288290.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

397

4. Duke-Elder, S. (1962). Iontophoresis. In The Foundations of Ophthalmology,

Vol. VII. C. V. Mosby, St. Louis, pp. 507514.
5. Harris, R. (1967). Iontophoresis, in Therapeutic Electricity and Ultraviolet
Radiation (S. Licht, ed.), Waverly Press, Baltimore, pp. 156178.
6. Banga, A. K., and Chien, Y. W. (1988). Iontophoretic delivery of drugs:
fundamentals, development and biomedical applications. J. Control Release,
7:124.
7. Nair, V, Pillai, O, Poduri, R., and Panchagnula, R. (1999). Transdermal
iontophoresis. Part I: Basic principles and considerations. Meth. Find. Exp.
Clin. Pharmacol., 21:139151.
8. Tyle, P. (1986). Iontophoretic devices for drug delivery. Review. Pharm. Res.,
3:318326.
9. Garrison, J. (1998). Iontophoresis: an alternative drug-delivery system. Med.
Device Technol., 9:3236.
10. Hirvonen, J., Kalia, Y. N., and Guy, R. H. (1996). Transdermal delivery of
peptides by iontophoresis. Nature Biotechnol., 14:17101713.
11. Shofner, R. S., Kaufman, H. E., and Hill, J. M. (1989). New horizons in
ocular drug delivery. Ophthalmol. Clin. North Am., 2:1524.
12. Sarraf, D., and Lee, D. A. (1994). The role of iontophoresis in ocular drug
delivery. J. Ocular Pharm., 10:6981.
13. Sasaki, H., Yamamura, K., Mukai, T., Nishida, K., Nakamura, J.,
Nakashima, M., and Ichikawa, M. (1999). Enhancement of ocular drug penetration. Crit. Rev. Ther. Drug Carrier Syst., 16:85146.
14. Shriber, W. J. (1975). The direct current and ion transfer. In: A Manual of
Electrotherapy (W. J. Shriber, ed.). Lea & Febiger, Philadelphia, p. 124.
15. Turrell, W. J. (1921). Therapeutic action of the constant current. Proc. R. Soc.
Med., 14:4152.
16. Morton, W. J. (1898). Cataphoresis of Electric Medicament Surgery. American
Technical Book, New York.
17. Leduc, S. (1900). Introduction of medicinal substances into the depth of tissues by electric current. Ann. Electrobiol., 3:545550.
18. Leduc, S. (1908). Electric Ions and Their Use in Medicine. Rebman Ltd.,
London.
19. Wirtz, R. (1908). Die Ionentherapie in der Augenheilkunde. Klin. Monatsbl.
Augenheilkd., 46:543579.
20. Morisot, L. (1927). Liontherapie ou ionisation appliquee au traitement des
aections ocu-laires. Clin. Ophthalmol., 31:516.
21. Erlanger, G. (1936). On the scientic and practical value of ionization in
ophthalmology. Recent advances and researchers. Br. J. Ophthalmol.,
20:213229.
22. Erlanger, G. (1939). Iontophoretic medication in ophthalmology. Theoretic
and practical aspects. Arch. Phys. Ther., 20:1624.
23. Erlanger, G. (1954). Iontophoresis: a scientic and practical tool in ophthalmology. Ophthalmologica, 128:232246.

Copyright 2003 Marcel Dekker, Inc.

398

Myles et al.

24. Von Sallmann, L., and Meyer, K. (1944). Penetration of penicillin into the eye.
Arch. Ophthalmol., 31:17.
25. Von Sallmann, L. (1944). Penicillin and sulfadiazine in the treatment of
experimental intraocular infections with Staphylococcus aureus and
Clostridium welchii. Arch. Ophthalmol., 31:5461.
26. Von Sallmann, L. (1945). Penetration of penicillin into the eye. Arch.
Ophthalmol., 34:195201.
27. Witzel, S. H., Fielding, I. Z., and Ormsby, H. L. (1956). Ocular penetration of
antibiotics by iontophoresis. Am. J. Ophthalmol., 42:8995.
28. Ichihashi, T. (1936). Eect of drugs on the sweat glands by cataphoresis and
an eective method for suppression of local sweating: Observation on the
eect of diaphoretics and adiaphoretics. J. Oriental Med., 25:101102.
29. Gibson, L. E., and Cooke, R. E. (1959). A test for concentration of electrolytes in sweat in cystic brosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics., 23:545549.
30. Carter, E. P., Barrett, A. D., Heeley, A. F., and Kuzemko, J. A. (1984).
Improved sweat test method for the diagnosis of cystic brosis. Arch. Dis.
Child., 59:919922.
31. Hall, S. K., Stableforth, D. E., and Green, A. (1990). Sweat sodium and
chloride concentrationsessential criteria for the diagnosis of cystic brosis
in adults. Ann. Clin. Biochem., 27:318320.
32. Sloan, J. B., and Soltani, K. (1986). Iontophoresis in dermatology: A review.
J. Am. Acad. Dermatol., 15:671684.
33. Kassan, D. G., Lynch, A. M., and Stiller, M. J. (1996). Physical enhancement
of dermatologic drug delivery: Iontophoresis and phonophoresis. J. Am.
Acad. Dermatol., 34:657666.
34. Stolman, L. P. (1998). Treatment of hyperhidrosis. Dermatol. Clin., 16:863
869.
35. Murphy, R., and Harrington, C.I. (2000). Treating hyperhidrosis:
Iontophoresis should be tried before other treatments (letter). BMJ,
321:702703.
36. Gangarosa, L. P., and Park, N. H. (1978). Practical considerations in iontophoresis of uoride for desensitizing dentin. J. Prosthet. Dent., 39:173178.
37. Krauser, J. T. (1986). Hypersensitive teeth. Part II: Treatment. J. Prosthet.
Dent., 56(3):307311.
38. Greenbaum, S. S., and Bernstein, E. F. (1994). Comparison of iontophoresis
of lidocaine with a eutetic mixture of lidocaine and prilocaine (EMLA) for
topically administered local anesthesia. J. Dermatol. Surg. Oncol., 20:579583.
39. Zempsky, W. T., Anand, K. J. S., Sullivan, K. M., Fraser, D., and Cucina, K.
(1998). Lidocaine iontophoresis for topical anesthesia before intravenous line
placement in children. J. Pediatr., 132:10611063.
40. DeCou, J. M., Abrams, R. S., Hammond, J. H., Lowder, L. R., and
Gauderer, W. L. (1999). Iontophoresis: a needle-free, electrical system of
local anesthesia delivery for pediatric surgical oce procedures. J. Pediatr.
Surg., 34:946949.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

399

41. Squire, S. J., Kirchho, K. T., and Hissong, K. (2000). Comparing two methods of topical anesthesia used before intravenous cannulation in pediatric
patients. J. Pediatr. Health Care., 14:6872.
42. Hill, J. M., OCallaghan, R. J., and Hobden, J. A. (1993). Ocular iontophoresis. In: Ophthalmic Drug Delivery Systems (A. K. Mitra, ed.). Marcel Dekker,
New York, pp. 331354.
43. Callegan, M. C., OCallaghan, R. J., and Hill, J. M. (1994). Pharmacokinetic
considerations in the treatment of bacterial keratitis. Clin. Pharmacokinet.,
27:129149.
44. Pillai, O., Nair, V., Poduri, R., and Panchagnula, R. (1999). Transdermal
iontophoresis. Part II: Peptide and protein delivery. Methods Find. Exp.
Clin. Pharmacol., 21:229240.
45. Jones, R. F., and Maurice, D. M. (1966). New methods of measuring the rate
of aqueous ow in man with uorescein. Exp. Eye Res., 5:208220.
46. Starr, P. A. J. (1966). Changes in aqueous ow determined by uorophotometry. Trans. Ophthalmol. Soc. U.K., 86:639646.
47. Holm, O. (1968). A photogrammetric method for estimation of the pupillary
aqueous ow in the living human eye. I. Acta Ophthalmol., 46:254277.
48. Brubaker, R. F. (1982). The ow of aqueous humor in the human eye. Trans.
Am. Ophthalmol. Soc., 80:391474.
49. Latina, M. A., Dobrogowski, M., March, W. F., and Birngruber, R. (1990).
Laser sclerostomy by pulsed-dye laser and goniolens. Arch. Ophthalmol.,
108:17451750.
50. Latina, M. A., Melamed, S., March, W. F., Kass, M. A., and Kolker, A. E.
(1992). Gonioscopic Ab Interno Laser Sclerostomy. A pilot study in glaucoma
patients. Ophthalmology, 99:17361744.
51. Melamed, S., Solomon, A., Neumann, D., Hirsch, A., Blumenthal, M., and
Belkin, M. (1993). Internal sclerostomy using laser ablation of dyed sclera in
glaucoma patients: a pilot study. Br. J. Ophthalmol., 77;139144.
52. Grossman, R. E., Sarraf, D., and Lee, D. A. (1993). Iontophoresis of methylene blue for gonioscopic pulsed dye laser sclerostomy. J. Ocul. Pharmacol.,
9:277285.
53. Sarraf, D., and Lee, D. A. (1993). Iontophoresis of reactive black 5 for pulsed
dye laser sclerostomy. J. Ocul Pharmacol. Ther., 9:2533.
54. Kitazawa, Y., Nose, H., and Horie, T. (1973). The eects of chemical sympathectomy on intraocular pressure of the normal human subjects. Acta Soc.
Ophthalmol. Jpn., 77:19011903.
55. Kitazawa, Y., and Horie, T. (1974). Denervation supersensitivity induced by
chemical sympathectomy with 6-hydroxydopamine. Jpn. J. Ophthalmol.,
18:109118.
56. Kitazawa, Y., Nose, H., and Horie, T. (1975). Chemical sympathectomy with
6-hydroxydopamine in the treatment of primary open-angle glaucoma. Am. J.
Ophthalmol., 79:98103.

Copyright 2003 Marcel Dekker, Inc.

400

Myles et al.

57. Watanabe, H., Levene, R. Z., and Bernstein, M. R. (1977). 6Hydroxydopamine therapy in glaucoma. Trans. Am. Acad. Ophthalmol.
Otolaryngol., 83:6977.
58. Colasanti, B. K., and Trotter, R. R. (1977). Enhanced ocular penetration of
the methyl ester of alpha-methyl-para-tyrosine after iontophoresis. Arch. Int.
Pharmacodyn. Ther., 228:171176.
59. Blumenkranz, M. S., Clan, A., and Hajek, A. S. (1984). Selection of therapeutic agents for intraocular proliferative disease. Arch. Ophthalmol.,
102:598604.
60. Kondo, M., and Araie, M. (1989). Iontophoresis of 5-uorouracil into the
conjunctiva and sclera. Invest. Ophthalmol. Vis. Sci., 30:583585.
61. Zempsky, W. T., and Ashburn, M. A. (1998). Iontophoresis: noninvasive drug
delivery. Am. J. Anesthesiol., 25:158162.
62. Sisler, H. A. (1978). Iontophoretic local anesthesia for conjunctival surgery.
Ann. Ophthalmol., 10:597598.
63. Meyer, D. R., Lindberg, J. V., and Vasquez, R. J. (1990). Iontophoresis for
eyelid anesthesia. Ophthalmic Surg., 21:845848.
64. Havener, W. H. (1978). Corticosteroid therapy. In: Ocular Pharmacology, 4th
ed. CV Mosby, St. Louis, pp. 143174.
65. Lachaud, J. P. (1965). Considerations on the use of corticosteroids by ionization in certain ocular diseases. Bull. Soc. Ophthalmol. France, 65:8489.
66. Lam, T. T., Edward, D. P., Zhu, X., and Tso, M. O. M. (1989). Transscleral
iontophoresis of dexamethasone. Arch. Ophthalmol., 107:13681371.
67. Behar-Cohen, F. F., Parel, J.-M., Pouliquen, Y., Thillaye-Goldenberg, B.,
Goureau, O., Heydolph, S., Courtois, Y., and DeKozak, Y. (1997).
Iontophoresis of dexamethasone in the treatment of endotoxin-induced-uveitis in rats. Exp. Eye Res., 65:533545.
68. Pagano, G., Bruno, A., Cavallo-Perin, P., Cesco, L., and Imbimbo, B. (1989).
Glucose intolerance after short-term administration of corticosteroids in
healthy subjects. Arch. Int. Med., 149:10981091.
69. Yoshida, M., Yoshimura, N., Hangai, M., Tanihara, H., and Honda, Y.
(1994). Interleukin-1 a, interleukin-1 b, and tumor necrosis factor gene expression in endotoxin-induced uveitis. Invest. Ophthalmol. Vis. Sci., 35:11071113.
70. Grossman, R. E., Chu, D. F., and Lee, D. A. (1990). Regional ocular gentamicin levels after transcorneal and transscleral iontophoresis. Invest.
Ophthalmol. Vis. Sci., 31:909916.
71. Fishman, P. H., Jay, W. M., Hill, J. M., Rissing, J. P., and Shockley, R. K.
(1984). Iontophoresis of gentamicin into aphakic rabbit eyes: sustained vitreal
levels. Invest. Ophthalmol. Vis. Sci., 25:343345.
72. Hughes, L., and Maurice, D. (1984). A fresh look at iontophoresis. Arch.
Ophthalmol., 102;18251829.
73. Frucht-Pery, J., Solomon, A., Doron, R., Ever-Hadani, P., Manor, O., and
Shapiro, M. (1996). Ecacy of iontophoresis in the rat cornea. Graefes Arch.
Clin. Exp. Ophthalmol., 234:765769.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

401

74. Frucht-Pery, J., Goren, D., Solomon, A., Siganos, C. S., Mechoulam, H.,
Ever-Hadani, P., and Shapiro, M. (1999). The distribution of gentamicin in
the rabbit cornea following iontophoresis to the central cornea. J. Ocular
Pharm. Therapeut., 15:251256.
75. Barza, M., Peckman, C., and Baum, J. (1986). Transscleral iontophoresis of
cefazolin, ticarcillin, and gentamicin in the rabbit. Ophthalmology, 93:133
139.
76. Barza, M., Peckman, C., and Baum, J. (1987). Transscleral iontophoresis as
an adjunctive treatment for experimental endophthalmitis. Arch. Ophthalmol.,
105:14181420.
77. Burstein, N. L., Leopold, I. H., and Bernaacchii, D. B. (1985). Transscleral
iontophoresis of gentamicin. J. Ocul. Pharmacol., 1:363368.
78. Barza, M. M., Peckman, C., and Baum, J. (1987). Transscleral iontophoresis
of gentamicin in monkeys. Invest. Ophthalmol. Vis. Sci., 28:10331036.
79. Rootman, D. S., Hobden, J. A., Jantzen, J. A., Gonzales, V. R., OCallaghan,
R. J., and Hill, J. M. (1988). Iontophoresis of tobramycin for the treatment of
experimental Pseudomonas keratitis in the rabbit. Arch. Ophthalmol., 106:262
265.
80. Hobden, J. A., Rootman, D. S., OCallaghan, R. J., and Hill, J. M. (1988).
Iontophoretic application of tobramycin to uninfected and Pseudomonas aeruginosa-infected rabbit corneas. Antimicrob. Agents Chemother., 32:978981.
81. Rootman, D. S., Jantzen, J. A., Gonzalez, J. R. Fischer, M., Beuerman, R.,
and Hill, J. M. (1988). Pharmacokinetics and safety of transcorneal iontophoresis of tobramycin in the rabbit. Invest. Ophthalmol. Vis. Sci., 29:1397
1401.
82. Hobden, J. A., OCallaghan, R. J., Hill, J. M., Reidy, J. J., Rootman, D. S.,
and Thompson, H. W. (1989). Tobramycin iontophoresis into corneas
infected with drug-resistant Pseudomonas aeruginosa. Curr. Eye Res.,
8:11631169.
83. Physicians Desk Reference. (1977). Medical Economics Company, Montvale,
N.J.
84. Hobden, J. A., Reidy, J. J., OCallaghan, R. J., and Hill, J. M. (1990).
Ciprooxacin iontophoresis for aminoglycoside-resistant Pseudomonas keratitis. Invest. Ophthalmol. Vis. Sci., 31:19401944.
85. Yoshizumi, M. O., Cohen, D., Verbukh, I., Leinwand, M., Kim, J., and Lee,
D. A. (1991). Experimental transscleral iontophoresis of ciprooxacin. J.
Ocular Pharm., 7:163167.
86. Choi, T. B., and Lee, D. A. (1988). Transscleral and transcorneal iontophoresis of vancomycin in rabbit eyes. J. Ocular Pharmacol., 4:153164.
87. Grossman, R., and Lee, D. A. (1989). Transscleral and transcorneal iontophoresis of ketoconazole in the rabbit eye. Ophthalmology, 96:724729.
88. Boniuk, I. (1972). The cytomegaloviruses and the eye. Int. Ophthalmol. Clin.,
12:169190.

Copyright 2003 Marcel Dekker, Inc.

402

Myles et al.

89. Miller, R. F., Howard, M. R., Frith, P., Perrons, C. J., Pecorella, I., and
Lucas, S. B. (2000). Herpes-virus infection of eye and brain in HIV. Sex.
Transm. Inf., 76:282286.
90. Friedman, A. H., Orellana, J., Freeman, W. R., Luntz, M. H., Starr, M. B.,
Tapper, M. L., et al. (1983). Cytomegalovirus retinitis: a manifestation of the
acquired immune deciency syndrome (AIDS). Br. J. Ophthalmol., 67:372
380.
91. Oberg, B. (1983). Antiviral eects of phosphonoformate (PFA foscarnet
sodium). Pharmacol. Ther., 19:387415.
92. Fannibg, M. M., Read, S.. E., Benson, M., Vas, S., Rachlis, A., Kozousek, V.,
et al. (1990). Foscarnet therapy of cytomegalovirus retinitis in AIDS. J.
Acquir. Immune Dec. Syndr., 3:472479.
93. Jacobson, M. A., and ODonnell, J. J. (1991). Approaches to the treatment of
cytomegalovirus retinitis: Ganciclovir and Foscarnet. J. Acquir. Immune
Dec., Syndr., 4 (Suppl. 1):S11S15.
94. She, S.-C., Peyman, G. A., and Schulman, J. A. (1988). Toxicity of intravitreal
injection of foscarnet in the rabbit eye. Int. Ophthalmol., 12:151154.
95. Diaz-Llopis, M., Chipont, E., Sanchez, S., Espana, E., Navea, A., and
Menezo, J. L. (1992). Intravitreal foscarnet for cytomegalovirus retinitis in
a patient with acquired immunodeciency syndrome. Am. J. Ophthalmol.,
114:742747.
96. Kreiger, A. E., Foos, R. Y., and Yoshizumi, M. O. (1992). Intravitreous
granulation tissue and retinal detachment following pars plana injection for
cytomegalovirus retinopathy. Graefes Arch. Clin. Exp. Ophthalmol., 230:197
198.
97. Sarraf, D., Equi, R. A., Holland, G. N., Yoshizumi, M. O., and Lee, D. A.
(1993). Transscleral iontophoresis of foscarnet. Am. J. Ophthalmol., 115:748
754.
98. Yoshizumi, M. O., Lee, D. A., Sarraf, D. A., Equi, R. A., and Verdon, W.
(1995). Ocular toxicity of iontophoretic foscarnet in rabbits. J. Ocul.
Pharmacol. Ther., 11:183189.
99. Anand, R., Nightingale, S. D., Fish, R. H., Smith, T. J., and Ashton, P.
(1993). Control of cytomegalovirus retinitis using sustained release of intraocular ganciclovir. Arch., Ophthalmol., 111:223227.
100. Yoshizumi, M. O., Roca, J. A., Lee, D. A., Lee, G., and Gomez, I. (1996).
Ocular iontophoretic supplementation of intravenous foscarnet therapy. Am.
J. Ophthalmol., 122:8690.
101. Yoshizumi, M. O., Dessouki, A., Lee, D. A., and Lee, G. (1997).
Determination of Ocular toxicity in multiple applications of foscarnet iontophoresis. J. ocular Pharmacol. Therapeut., 13:529536.
102. Hill, J. M., Wen, R., and Halford, W. P. (1998). Pathogenesis and molecular
biology of ocular HSV in the rabbit. In: Herpes Simplex Virus Protocols (MS
Brown, AR Maclean, eds.). Humana Press, Inc., Totowa, NJ, pp. 291315.
103. Kaufman, H. E., and Rayeld, M. A. (1988). In: The Cornea: Viral
Conjunctivitis and Keratitis (H. E. Kaufman, B. A. Barron, M. B.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

104.
105.

106.

107.

108.

109.
110.

111.

112.

113.

114.
115.
116.

117.

403

McDonald, and S. R. Waltman, eds.). Churchill Livingstone, New York, pp.

299331.
Hill, J. M., Gangarosa, L. P., and Park, N. H. (1977). Iontophoretic application of antiviral chemotherapeutic agents. Ann. NY Acad. Sci., 284:604612.
Hill, J. M., Park, N. H., Gangarosa, L. P., Hull, D. S., Tuggle, C. L.,
Bowman, K., and Green, K. (1978). Iontophoretic application of vidarabine
monophosphate into rabbit eyes. Invest. Ophthalmol. Vis. Sci., 17:473476.
Kwon, B. S., Gangarosa, L. P., Park, N. H., Hull, D. S., Fineberg, E.,
Wiggins, C., and Hill, J. M. (1979). Eects of iontophoretic and topical application of antiviral agents in treatment of experimental HSV-1 keratitis in
rabbits. Invest. Ophthalmol. Vis. Sci., 18:984988.
Hill, J. M., Kwon, B. S., Burch, K. D., deBack, J., Whang, I., Jones, G. T.,
Luke, B., Andrews, P., Harp, R., Shimomura, Y., Hull, D. S., and Gangarosa,
L. P. (1982). Acyclovir and vidarabine monophosphate: a comparison of
iontophoretic and intravenous administration for the treatment of HSV-1
stromal keratitis. Am. J. Med., 73:300304.
Stevens, J. G., Newburn, A. B., and Cook, M. L. (1972). Latent herpes simplex virus from trigeminal ganglia of rabbits with recurrent eye infection.
Nature (New Biology), 235:216217.
Stevens, J. G. (1994). Overview of herpesvirus latency. Semi. Virol., 5:191196.
Rivera, L., Beuerman, R. W., and Hill, J. M. (1988). Corneal nerves contain
intra-axonal HSV-1 after virus reactivation by epinephrine iontophoresis.
Curr. Eye Res., 7:10011008.
Schmidt, D. D., Zyzanski, S., Ellner, J., Kumar, M. L., and Arno, J. (1985).
Stress as a precipitating factor in subjects with recurrent herpes labialis. J.
Family Pract., 20:359366.
Varnell, E. D., Kaufman, H. E., Hill, J. M., and Thompson, H. W. (1995).
Cold stress-induced recurrences of herpetic keratitis in the squirrel monkey.
Invest. Ophthalmol. Vis. Sci., 36:11811183.
Hill, J. M., Garza, H. H., Helmy, M. F., Cook, S. D., Osborne, P. A.,
Johnson, E. M., Thompson, H. W., Green, L. C., OCallaghan, R. J., and
Gebhardt, B. M. (1997). Nerve growth factor antibody stimulates reactivation
of ocular herpes simplex virus type 1 in latently infected rabbits. J. Neurovirol.,
3:206211.
Stanberry, L. R. (1994). Animal models and HSV latency. Sem. Virol., 5:213
219.
Hill, J. M., Haruta, Y., and Rootman, D. S. (1987). Adrenergically induced
recurrent HSV-1 corneal epithelial lesions. Curr. Eye Res., 6:10651071.
Hill, J. M., Rayeld, M. A., and Haruta, Y. (1987). Strain specicity of
spontaneous and adrenergically induced HSV-1 ocular reactivation in latently
infected rabbits. Curr. Eye Res., 6:9197.
Hill, J. M., Shimomura, Y., Kwon, B. S., and Gangarosa, L. P. (1986).
Iontophoresis of epinephrine isomers to rabbit eyes induced HSV-1 ocular
shedding. Invest. Ophthalmol. Vis. Sci., 26:12991303.

Copyright 2003 Marcel Dekker, Inc.

404

Myles et al.

118. Kwon, B. S., Gangarosa, L. P., Burch, K. D., deBack, J., and Hill, J. M.
(1981). Induction of ocular herpes simplex virus shedding by iontophoresis of
epinephrine into rabbit cornea. Invest. Ophthalmol. Vis. Sci., 21:442449.
119. Kwon, B. S., Gangarosa, L. P., Green, K., and Hill, J. M. (1982). Kinetics of
ocular herpes simplex virus shedding induced by epinephrine iontophoresis.
Invest. Ophthalmol. Vis. Sci., 22:818821.
120. Willey, D. E., Trousdale, M. D., and Nesburn, A. B. (1984). Reactivation of
murine latent HSV infection by epinephrine iontophoresis. Invest. Ophthalmol.
Vis. Sci., 25: 945950.
121. Gordon, Y. J., Araullo-Cruz, T. P., Romanowski, E., Ruziczka, L., Balouris,
C., Oren, J., Cheng, K. P., and Kim, S. (1986). The development of an
improved murine iontophoresis reactivation model for the study of HSV-1
latency. Invest. Ophthalmol. Vis. Sci., 27:12301234.
122. Varnell E. D., Kaufman, H. E., Hill, J. M., and Wolf, R. H. (1987). A primate
model for acute and recurrent herpetic keratitis. Curr. Eye Res., 6:277279.
123. Rootman, D. S., Haruta, Y., and Hill, J. M. (1990). Reactivation of HSV-1 in
primates by transcorneal iontophoresis of adrenergic agents. Invest.
Ophthalmol. Vis. Sci., 31:597600.
124. Hill, J. M., Kwon, B. S., Shimomura, Y., Colborn, G. L., Yaghmai, F., and
Gangarosa, L. P. (1983). Herpes simplex virus recovery in neural tissues after
ocular HSV shedding induced by epinephrine iontophoresis to the rabbit
cornea. Invest. Ophthalmol. Vis. Sci., 24:243247.
125. Shimomura, Y., Gangarosa, L. P., Sr., Kataoka, M., and Hill, J. M. (1983).
HSV-1 shedding by iontophoresis of 6-hydroxydopamine followed by topical
epinephrine. Invest Ophthalmol Vis. Sci., 24:15881594.
126. Hill, J. M., Dudley, J. B., Shimomura, Y., and Kaufman, H. E. (1986).
Quantitation and kinetics of induced HSV-1 ocular shedding. Curr. Eye.
Res., 5:241246.
127. Haruta, Y., Rootman, D. S., and Hill, J. M. (1988). Recurrent HSV-1 corneal
epithelial lesions by timolol iontophoresis in latently infected rabbits. Invest.
Ophthalmol. Vis. Sci., 29:387392.
128. Hill, J. M., Shimomura, Y., Dudley, J. B., Berman, E., Haruta, Y., Kwon, B.
S., and Maguire, L. J. (1987). Timolol induces HSV-1 ocular shedding in the
latently infected rabbit. Invest. Ophthalmol. Vis. Sci., 28:585590.
129. Rootman, D. S., Haruta, Y., Hill, J. M., and Kaufman, H. E. (1989).
Triuridine decreases ocular HSV-1 recovery, not recurrent HSV-1 lesions
following timolol iontophoresis in the rabbit. Invest. Ophthalmol. Vis. Sci.,
30:678683.
130. Kaufman, H. E., Varnell, E. D., Gebhardt, B. M., Thompson, H. W., and
Hill, J. M. (1996). Propranolol suppression of ocular HSV-1 recurrence and
associated corneal lesions following spontaneous reactivation in the rabbit.
Curr. Eye Res., 9:10151021.
131. Gebhardt, B. M., and Kaufman, H. E. (1995). Propranolol suppresses reactivation of herpesvirus. Antiviral Res., 27:255261.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

405

132. Garza, H. H., Jr., and Hill, J. M. (1997). Eect of a beta-adrenergic antagonist, propranolol, on induced HSV-1 ocular recurrence in latently infected
rabbits. Curr. Eye Res., 16:453458.
133. Haruta, Y., Rootman, D. S., Xie, L., Kiritoshi, A., and Hill, J. M. (1989).
Recurrent HSV-1 corneal lesions in rabbits induced by cyclophosphamide and
dexamethasone. Invest. Ophthalmol. Vis. Sci., 30:371376.
134. Wagner, E. K., and Bloom, D. C. (1997). Experimental investigation of herpes
simplex virus latency. Clin. Microbiol. Rev., 10:419443.
135. Miller, C. S., Danaher, R. J., and Jacob, R. J. (1998). Molecular aspects of
herpes simplex virus I latency, reactivation, and recurrence. Crit. Rev. Oral
Biol. Med., 9:541562.
136. Millhouse, S., and Wigdhal, B. (2000). Molecular circuitry regulating herpes
simplex virus type 1 latency in neurons. J. Neurovirol., 6:624.
137. Stevens, J. G., Wagner, E. K., Devi-Rao, G. B., Cook, M. L., and Feldman,
L. T. (1987). RNA complementary to a herpesvirus a gene mRNA is prominent in latently infected neurons. Science, 235:10561059.
138. Spivak, J. G., and Fraser, N. W. (1987). Detection of herpes simplex virus
type 1 transcripts during latent infection in mice. J. Virol., 61:38413847.
139. Javier, R. T., Stevens, J. G., Dissette, V. B., and Wagner, E. K. (1988). A
herpes simplex virus transcript abundant in latently infected neurons is dispensable for establishment of the latent state. Virology, 166:254257.
140. Stevens, J. G. (1989). Human herpesviruses: a consideration of the latent state.
Microbiol. Rev., 53:318332.
141. Rock, D. L., Nesburn, A. B., Ghiasi, H., Ong, I., Lewis, T. L., Lokensgard, J.
R., and Wechsler, S. L. (1987). Detection of latency-related viral RNAs in
trigeminal ganglia of rabbits latently infected with herpes simplex virus type 1.
J. Virol., 61:38203826.
142. Cook, S. D., Hill, J. M., Lynas, C., and Maitland, N. J. (1991). Latencyassociated transcripts in corneas and ganglia of HSV-1 infected rabbits. Br.
J. Ophthalmol., 75:644648.
143. Croen, K. D., Ostrove, J. M., Dragovic, L. J., Smialek, J. E., and Straus, S. E.
(1987). Latent herpes simplex virus in human trigeminal ganglia. Detection of
an immediate early gene anti-sense transcript by in situ hybridization. N.
Engl. J. Med., 317:14271432.
144. Kaye, S. B., Lynas, C., Patterson, A., Risk, J. M., McCarthy, K., and Hart, C.
A. (1991). Evidence of herpes simplex viral latency in the human cornea. Br. J.
Ophthalmol., 75:195200.
145. Fraser, N. W., Block, T. M., and Spivak, J. G. (1992). The latency-associated
transcripts of herpes simplex virus: RNA in search of a function. Virology
191:18.
146. Wagner, E. K., Flanagan, W. M., Devi-Rao, G., Zhang, Y. F., Hill, J. M.,
Anderson, K. P., and Stevens, J. G. (1988). The herpes simplex virus latencyassociated transcript is spliced during the latent phase of infection. J. Virol.,
62:45774585.

Copyright 2003 Marcel Dekker, Inc.

406

Myles et al.

147. Block, T. M., and Hill, J. M. (1997). The latency associated transcripts (LAT)
of herpes simplex virus: still no end in sight. J. Neurovirol., 3:313321.
148. Farrell, M. J., Dobson, A. T., and Feldman, L. T. (1991). Herpes simplex
virus-latency-associated transcript is a stable intron. Proc. Natl. Acad. Sci.
USA, 88:790794.
149. Hill, J. M., Sedarati, F., Javier, R. T., Wagner, E. K., and Stevens, J. G.
(1990). Herpes simplex virus latent phase transcription facilitates in vivo reactivation. Virology, 174:117125.
150. Hill, J. M., Maggioncalda, J. B., Garza, H. H., Jr., Su, Y.-H., Fraser, N. W.,
and Block, T. M. (1996). In vivo epinephrine reactivation of ocular herpes
simplex virus type 1 in the rabbit is correlated to a 370-base-pair region
located between the promoter and the 50 end of the 2.0-kilobase latency-associated transcript. J. Virol., 70:72707274.
151. Bloom, D. C., Devi-Rao, G. B., Hill, J. M., Stevens, J. G., and Wagner, E. K.
(1994). Molecular analysis of herpes simplex virus type 1 during epinephrineinduced reactivation of latently infected rabbits in vivo. J. Virol., 68:1283
1292.
152. Bloom, D. C., Stevens, J. G., Hill, J. M., and Tran, R. K. (1997). Mutagenesis
of a cAMP response element within the latency-associated transcript promoter
of HSV-1 reduces adrenergic reactivation. Virology, 236:202207.
153. Colgin, M. A., Smith, R. L., and Wilcox, C. L. (2001). Inducible cyclic AMP
early repressor produces reactivation of latent herpes simplex virus type 1 in
neurons in vitro. J. Virol., 75:29122920.
154. Loutsch, J. M., Perng, G.-C., Hill, J. M., Zheng, X., Marquart, M. E., Block,
T. M., Ghiasi, H., Nesburn, A. B., and Wechsler, S. L. (1999). Identical 371base-pair deletion mutations in the LAT genes of herpes simplex virus type 1
McKrae and 17syn+ result in dierent in vivo reactivation phenotypes. J.
Virol., 73:767771.
155. Zheng, X., Marquart, M. E., Loutsch, J. M., Shah, P., Sainz, B., Ray, A.,
OCallaghan, R. J., Kaufman, H. E., and Hill, J. M. (1999). HSV-1 migration
in latently infected and naive rabbits after penetrating keratoplasty. Invest.
Ophthalmol. Vis. Sci., 40:24902497.
156. McGill, J. (1991). Herpes simplex latency and the eye. Br. J. Ophthalmol.,
75:641642.
157. Cook, S. D., Ophth, F. C., and Hill, J. M. (1991). Herpes simplex virus:
molecular biology and the possibility of corneal latency. Surv. Ophthalmol.,
36:140148.
158. Bloom, D. C., Hill, J. M., Devi-Rao, G., Wagner, E. K., Feldman, L. T., and
Stevens, J. G. (1996). A 348-base pair region in the latency-associated transcript facilitates herpes simplex virus type 1 reactivation. J. Virol., 70:2449
2459.
159. Perng, G.-C., Slanina, S. M., Ghiasi, H., Nesburn, A. B., and Wechsler, S. L.
(1996). A 371-nucleotide region between the herpes simplex virus type 1 (HSV1) LAT promoter and the 2-kilobase LAT is not essential for ecient spontaneous reactivation of latent HSV-1. J. Virol., 70:20142018.

Copyright 2003 Marcel Dekker, Inc.

Ocular Iontophoresis

407

160. Hill, J. M., Garza, H. H., Su, Y. H., Meegalla, R., Hanna, L. A., Loutsch, J.
M., Thompson, H. W., Varnell, E. D., Bloom, D. C., and Block, T. M. (1997).
A 437-base pair deletion at the beginning of the latency-associated transcript
promoter signicantly reduced adrenergically induced herpes simplex virus
type 1 ocular reactivation in latently infected rabbits. J. Virol., 71:65556559.
161. Devi-Rao, G. B., Aquilar, J. S., Rice, M. K., Garza, H. H., Jr., Bloom, D. C.,
Hill, J. M., and Wagner, E. K. (1997). Herpes simplex virus genome replication and transcription during induced reactivation in the rabbit eye. J. Virol.,
71:70397047.
162. Behar-Cohen, F. F., Savoldelli, M., Parel, J. M., Goureau, O., ThillayeGoldenberg, B., Courtois, Y., Pouliquen, Y., and de Kozak, Y. (1998).
Reduction of corneal edema in endotoxin-induced uveitis after application
of L-NAME as nitric oxide synthase inhibitor in rats by iontophoresis.
Invest. Ophthalmol. Vis. Sci., 39:897904.
163. Martin, R. E., Loutsch, J. M., Garza, H. H., Jr., Boedeker, D. J., and Hill, J.
M. (1999). Iontophoresis of lysophosphatidic acid into rabbit cornea induces
HSV-1 reactivation: evidence that neuronal signaling changes after infection.
Mol. Vis., 5:36, www.molvis.org/molvis/v5/p36/.
164. Dowd, N. P., Day, F., Timon, D., Cunningham, A. J., and Brown, L. (1999).
Iontophoretic vincristine in the treatment of postherpetic neuralgia: A doubleblind, randomized, controlled trial. J. Pain Symptom Manage., 17:175180.
165. Santi, P., Volpato, N. M., Bettini, R., Catellani, P. L., Massimo, G., and
Colombo, P. (1997). Transdermal iontophoresis of salmon calcitonin can
reproduce the hypocalcemic eect of intravenous administration. Farmaco,
52:445448.
166. Chang, S. L., Hofmann, G. A., Zhang, L., Deftos, L. J., and Banga, A. K.
(2000). Transdermal iontophoretic delivery of salmon calcitonin. Int. J.
Pharm., 200:107113.
167. Mize, N. K., Buttery, M., Daddona, P., Morales, C., and Cormier, M. (1997).
Reverse iontophoresis: monitoring prostaglandin E2 associated with cutaneous inammation in vivo. Exp. Dermatol., 6:298302.
168. Merino, V., Lopez, A., Hochstrasser, D., and Guy, R. H. (1999). Noninvasive
sampling of phenylalanine by reverse iontophoresis. J. Control Release, 61:65
69.
169. Brasch, J., Huttemann, M., and Proksch, E. (2000). Iontophoresis of nickel
elicits a delayed cutaneous response in sensitized individuals that is similar to
an allergic patch test reaction. Contact Dermatitis, 42:3641.
170. Asahara, T., Shinomiya, K., Naito, T., and Shiota, H. (1999). Induction of
genes into the rabbit eye by iontophoresis. Acta Soc. Ophthalmol. Jpn.,
103:178186.
171. Chapon, P., Voigt, M., Gautier, S., Behar-Cohen, F., OGrady, G., and Parel,
J.-M. (1999). Intraocular tissues pharmacokinetics of ganciclovir transscleral
Coulomb Controlled iontophoresis in rabbits (abstr). IOVS/ARVO,
40(4):S189.

Copyright 2003 Marcel Dekker, Inc.

408

Myles et al.

172. Chauvaud, D., Behar-Cohen, F., Parel, J. M., and Renard, G. (2000).
Transscleral iontophoresis of corticosteroids: Phase II clinical trial (abstr).
IOVS/ARVO 41(4):S79.
173. Hayden, B. C., Voigt, M., Murray, T. G., Hernandez, E., Parel, J.-M.,
Cicciarelli, N., Feuer, W., Fulton, L., and OBrien, J. M. (2000).
Iontophoretic delivery of carboplatin in the treatment of murine transgenic
retinoblastoma (abstr). IOVS/ARVO 41(4):S788.

Copyright 2003 Marcel Dekker, Inc.

13
Mucoadhesive Polymers in
Ophthalmic Drug Delivery
Thomas P. Johnston, Clapton S. Dias, and Ashim K. Mitra
University of MissouriKansas City, Kansas City, Missouri, U.S.A.

Hemant Alur
Murty Pharmaceuticals, Inc., Lexington, Kentucky, U.S.A.

I.
A.

MUCOADHESIVE DOSAGE FORMS

Rationale for the Use of Mucoadhesives

Mucoadhesive dosage forms can provide a localized delivery of medicinal

agents to a specic site in the body. The ability of mucoadhesive dosage
forms to provide an intimate contact of the delivery system with the absorbing corneal layer would undoubtedly improve ocular bioavailability. The
intimate contact may result in high drug concentration in the local area
and hence high drug ux through the absorbing tissue. The intimate contact
may also increase the local permeability of high molecular weight drugs such
as peptides and proteins (1).
Bioadhesion is a term that is widely used in the pharmaceutical literature. For drug delivery purposes, this term refers to the attachment of a drug
carrier to a specic biological tissue. The majority of bioadhesives studied
for drug delivery adhere to epithelial tissue and possibly to the mucosal
surface of these tissues. Coating the external surface of the globe of the
eye is a thin lm of glycoprotein referred to as mucin. Therefore, such
bioadhesion is also referred to as mucoadhesion. We shall take this opportunity to examine the structure and function of the mucus layer and its role
in the process of mucoadhesion.
409

Copyright 2003 Marcel Dekker, Inc.

410

Johnston et al.

1. Physiology of the Mucus Layer

Mucus is a highly viscous secretion, which forms a thin, continuous gel
blanket adherent to the mucosal epithelial surface. It is continually secreted
by either the goblet cells or specialized exocrine glands in various regions of
the body (2). The major constituents of mucus are water (95%) and high
molecular weight glucoproteins capable of forming slimy, viscoelastic gels
(3,4). The mean thickness of the mucus layer varies from 50 to 450 m in
humans and about half as much in the rat (5). The exact composition of the
mucus layer varies substantially depending on the anatomical location, the
species, and the pathophysiological state (6). Mucus contains some nonmucin components, which aid in its protective functions. Lipids and covalently
bound fatty acids are frequently found in the mucin layer.
The mucus in the eye (Fig. 1) is mainly produced by the conjunctival
goblet cells, which are most abundant in the inner canthral region and the
lower fornix. The maximum number of such cells per unit area is found in
the palpebral conjunctiva. On the surface of the mucosal tissue, the mucin
molecules are tightly packed. As one proceeds outward from the epithelial
layer, the mucus layer becomes less densely packed with a corresponding
lowering of viscosity and ion content. Neutral and acidic mucins are produced by the goblet cells. Once secreted onto the conjunctiva, the mucus is
spread over the surface of the cornea by the upper eyelid.
The principal functions of the mucus layer are lubrication and protection of the underlying epithelial cells from dehydration and other challenges.
Continuous secretion of mucus is necessary to compensate for the loss due
to digestion, bacterial degradation, and solubilization of mucin molecules.
Soluble mucus may form temporary unstirred layers atop the adherent
mucus gel (5).
2. Composition of Mucin
Characteristically, the mucus is composed of a number of components:
glycoproteins, proteins, lipids, electrolytes, inorganic salts, water, enzymes,
mucopolysaccharides, among others. The mucin molecule of a polypeptide
backbone, which is attached to the pendant sugar groups at periodic intervals on the peptide chain. The molecular weights of these glycoproteins vary
from 2  106 to 14  106 daltons (3). In general, a major portion of the
peptide backbone is covered with carbohydrates grouped in various combinations. Galactose, fucose, N-acetylglucosamine, N-acetylgalactosamine,
and N-acetylneuraminic acid (sialic acid) are typically found in the mucin
molecules. These carbohydrates may constitute as much as 7090% of the
total mucin weight (7). The sugar molecules can carry sulfate residues via
ester linkages. Each carbohydrate chain terminates in either a sialic acid

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

411

Figure 1 Structure and composition of the tear lm, illustrating the physicochemical properties of the mucus layer.

pKa 2:6) or with an l-fucose group. Hence, the mucin molecules behave
as anionic polyelectrolytes as neutral pH (8). Because of the rather large
number of sugar groups, the mucin molecule is capable of picking up almost
5080 times its own weight in water. The oligosaccharides form a protective
coat over the glycoprotein backbone by preventing the enzymatic action of
proteases (9). Numerous sugar hydroxyl groups of mucin molecules have the
potential to interact with other polymers by hydrogen bonds. The positions
and relative amounts of amino acids in the glycoprotein backbone are

Copyright 2003 Marcel Dekker, Inc.

412

Johnston et al.

important to the matrix structure of the mucus, since they confer to the
overall tertiary structure and folding of the glycoprotein.

B.

Mechanism of Mucoadhesion

The attachment of mucin to the epithelial surface may be considered as an

interaction of a number of charged and neutral polymer groups with the
mucin through noncovalent bonds. Understanding the mechanisms of
mucoadhesion is fundamental to the development of mucoadhesive. One
may view the entire process to be simply a physical entanglementa currently accepted mechanism for the attachment of cross-linked polyacrylates
to mucin (10). The polymer undergoes swelling in water, which permits
entanglement of the polymer chains with mucin on the epithelial surface
of the tissue (11). The un-ionized carboxylic acid residues on the polymer
form hydrogen bonds with the mucin molecule.
The mucoadhesion phenomenon has also been explained using the
mechanisms of nonbiological adhesion, such as electron transfer (12), wetting (1316), diusion (11,1720), adsorption (2123), fracture (2425), and
mechanical interlocking theories (26). Although these theories provide some
insights into the mechanisms of mucoadhesion, no one theory by itself has
successfully explained the phenomenon of mucoadhesion. Considering the
number of factors involved in this process, this is not surprising.
Understanding molecular interactions between mucin and mucoadhesive may provide a better hypothesis for mucoadhesion. When two molecules coalesce, the interaction is composed of attractive and repulsive forces.
The magnitudes of these two forces determine whether the molecules will
interact or not. For mucoadhesion to occur, the attractive interaction
should be larger than nonspecic repulsion. Attractive interactions result
from van der Waals forces, hydrogen bonding, electrostatic attractions, and
hydrophobic bonding. Repulsive interactions occur as the result of electrostatic and steric repulsions. The theories relating to these phenomena are
detailed elsewhere (27).
The bioadhesive process can be conceptualized as the establishment of
intimate contact, by diusion or network expansion, of the polymer chains,
with subsequent interpenetration (11). This physical model is depicted in
Figure 2. In swellable hydrogels, an expanded polymer is a necessary prerequisite for adhesion, and this process may be further enhanced by viscoelastic deformation of the bioadhesive and substrate tissue by applied force
or pressure.
When anionic polymers interact with anionic mucin, the maximum
adhesion occurs at an acidic pH, indicating that it is the protonated form

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

413

Figure 2 Schematic representation of the chain interpenetration during the bioadhesion of a polymer (A) with the mucus layer (B).

of the mucoadhesive that is responsible for the bioadhesion. Therefore,

hydrogen bonding plays an important role in bioadhesion (28).
In addition, the expanded nature of both mucin and polymer networks
permit mutual interpenetration. Interpenetration/interdiusion of mucin
and the adhesives results in increased contact and henceforth physical entanglement of the two dierent macromolecules. The physical entanglement is
time dependent and may be enhanced by promoting intermolecular interactions between specic functional groups on the two polymers. Strong
mucoadhesion depends on the moderate interactive forces between mucus
and mucoadhesives, which will allow diusion and subsequent entanglements among polymer chains. A number of factors may aect interpenetration; i.e., chain segment mobility, chain entanglement, cross-linking density
of the networks, swelling, porosity, and compatibility of adhesives and
mucin.

Copyright 2003 Marcel Dekker, Inc.

414

Johnston et al.

While a number of polymers will attach to mucin through noncovalent

and covalent bonds, the former is preferred, since the strength of attachment
is suciently strong. The removal occurs primarily through mucin turnover.
The strength of adhesion between polycarbophil (partial structure shown in
Fig. 3) and mucin is suciently stronger to resist rinsing. Forcible removal
leads to rupture of mucin-mucin bonds and polymer-mucin bonds. The
water-swellable yet water-insoluble systems are preferred as mucoadhesives,
since predictable drug release from such systems would be easier to obtain.
Moreover, toxicity concerns will also be less for an insoluble polymer. Table
1 lists some of the representative mucoadhesives.
C.

Factors Relevant to Ocular Mucoadhesion

A number of variables can aect the performance of an ocular delivery

system, especially when mucoadhesives are employed in the design of an
ophthalmic vehicle. Satisfactory performance of a topical dosage form
depends on a number of variables, which include experimental, physiological, and dosage form eects.
1. Experimental Variables
a. Mucoadhesive Polymer Characteristics. Choice of polymer plays
an important role in the release kinetics of the drug from a mucoadhesive
dosage form. Ocular bioavailability from a mucoadhesive dosage form
will depend on the polymers bioadhesion properties, which, in turn, are
affected by its swelling properties, hydration time, molecular weight, and
degree of crosslinking. Other factors, such as pH, mucin turnover, and
disease state, that affect bioadhesion will be discussed later. There are various new polymers now being introduced in the pharmaceutical market.
Some of these polymers have found a place in the ophthalmic drug development industry. Recently, polyethylene oxide (PEO) has been investi-

Figure 3 Partial structure of polycarbophil.

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

415

Table 1 Some Representative Mucoadhesives with Their Relative

Mucoadhesive Performance
Substance

Adhesive
performance

Carboxymethylcellulose
Carbopol
Carbopol and hydroxypropylcellulose
Carbopol base with white petrolatum/hydrophilic petrolatum
Carbopol 934 and EX 55
Poly(methyl methacrylate)
Polyacrylamide
Poly(acrylic acid)
Polycarbophil
Homopolymers and copolymers of acrylic acid and butyl acrylate
Gelatin
Sodium alginate
Dextran
Pectin
Acacia
Povidone
Poly(acrylic acid) cross-linked with sucrose

Excellent
Excellent
Good
Fair
Good
Excellent
Good
Excellent
Excellent
Good
Fair
Excellent
Good
Poor
Poor
Poor
Fair

gated extensively for ophthalmic purposes (29). Investigators have shown

that polymer molecular weight affects the in vitro and in vivo performance of the dosage form into which it is incorporated. It is reported that
for a given polymer there exists a hydration degree corresponding to an
optimum polymer concentration in the gel allowing maximal interaction
with the substrate (in this case, the mucous layer). Hydration time affects
the bioadhesive force, and hence it is essential to determine the optimal
hydration degree. At the optimum hydration, there exists an optimum
polymer concentration in the gel allowing maximal interaction between
the polymer and the substrate. Higher concentrations of the polymer will
favor polymer-polymer interactions, whereas, at lower concentrations, the
number of penetrating polymer chains per unit volume of the mucous
layer is decreased, leading to a weak bioadhesive force. Studies with PEO
showed that molecular weight affected the degree of hydration and, in
turn, affected bioadhesion. With PEOs it was found that the mucoadhesive potential of the lower molecular weight PEOs was greater than that
of the higher molecular weight PEOs, with a maximum for PEO 400. Another important factor to be considered is the affect of these polymers on
the viscosity of the tear uid. If the polymer enhances the viscosity of the
tear uid, precorneal clearance will decrease, allowing more contact time

Copyright 2003 Marcel Dekker, Inc.

416

Johnston et al.

between the drug and the cornea. Viscosity of the aqueous solution increases with an increase in the polymer molecular weight. However, when
formulating an ophthalmic dosage form, it is essential to strike a balance
between the hydration properties and the viscosity-enhancing effects of
the polymer. A polymer that enhances viscosity severalfold may not necessarily be ideal, as it may have poor hydration properties leading to weak
bioadhesion and hence might fail. Higher molecular weight polymers may
swell excessively in aqueous solutions leading to their limited use in the
dosage form. In addition to these factors, the toxicity and degree of irritation exerted by the polymer affect the precorneal residence time of the
dose. A polymer that irritates the corneal surface will cause increased tear
secretion leading to dilution and elimination of the drug.
b. pH. The pH of the medium employed in the mucoadhesion studies has a profound effect on the performance of the delivery system. The
effects of hydrophilicity and hydrogen bonding of polymers cannot be
overemphasized in mucoadhesion, considering the fact that common functional groups found in bioadhesive polymers, i.e., carboxyl, amide, and
sulfate groups, are polar and have the ability to form hydrogen bonds.
This property of the polymers in turn has been shown to correlate well
with the degree of hydration, which can be controlled by adjusting the pH
of the medium (30). The rst systematic investigation of pH effects on
mucoadhesive strength was undertaken using polycarbophil and rabbit
gastric tissue (Fig. 3). The experiments revealed maximum adhesive
strength at or below pH 3 and a complete loss in mucoadhesive property
above pH 5. The results indicated that the protonated carboxyl groups
rather than the ionized carboxylate anions interact with mucin molecules
through numerous hydrogen bonds. At higher pH values, the chains are
fully extended owing to electrostatic repulsion of the carboxylate anions.
Since the mucin molecules are negatively charged in this environment,
electrostatic repulsion also occurs. Similar postulations have been forwarded by Nagai and Machida (31), where the interpolymer complex between hydroxypropylcellulose and Carbopol 934 was observed below pH
4.5 (32).
At physiological pH (pH 7.4) mucin is negatively charged owing to the
presence of sialic acid groups at the terminal ends of the mucopolysaccharide chains (33). The preferential uptake of cationic liposomes by the cornea
is probably evidence supporting the hypothesis of electrostatic interactions
between the mucin and cationic mucoadhesives. In the case of anionic polymers, a hydrogen bonding mechanism is suggested for mucoadhesion.
At physiological pH, the hydration of the cross-linked polymers in the
precorneal uid is maximum, whereas the number of hydrogen bonds is

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

417

comparatively low. Such loss in hydrogen bonding ability somewhat lowers

the mucoadhesive strength in the precorneal uid. However, the attachment
is rm enough to provide some retention of the delivery system in the precorneal area.
c. Contact Time. Almost all bioadhesive polymers are solvated in
aqueous medium and owe their expanding networks to hydration and
subsequent swelling. Also, this swelling would increase the exibility and
mobility of the polymer chains. Swelling is an important prerequisite for
the interpenetration and entanglement, i.e., strong bioadhesion.
The initial contact time between mucoadhesives and the mucus determines the extent of swelling of the mucoadhesives and the interpenetration
of polymer chains. The mucoadhesive strength has been shown to increase
with an increase in the initial contact time (34,35). Nevertheless, one needs
to consider the optimum contact time based on the tissue viability. In the
case of mucoadhesive dosage forms, which needs to be polymerized at the
site of application, i.e., corneal, buccal, or nasal tissue, the initial contact
time is critical for successful mucoadhesion (36).
Recent reports (37) suggest that the adhesive strength increases as the
molecular weight of the bioadhesive polymer increases up to 100,000 daltons. For sodium carboxymethyl cellulose to function as an eective bioadhesive, the molecular weight should be in excess of 78,600 daltons (38). The
adhesive force may be related to the critical macromolecular length required
to produce entanglement and an interpenetrating layer.
d. Selection of Model Substrate. An abundance of literature exists
detailing the use of tissue samples for understanding the mechanism of
mucoadhesion and bioadhesion of new materials. Gastric mucosa obtained from rabbits is the most common model tissue specimen cited in
the literature. However, caution needs to be exercised in terms of extrapolation of these ndings to any human clinical studies.
The handling and treatment of biological substrates during the testing
of mucoadhesives is an important factor. Physical and biological changes
may occur in the mucus gels or tissues under the experimental conditions
(3840), which may be of major concern in optimizing the delivery systems.
The viability of the biological tissue needs to be conrmed by electrophysiology or histological examinations.
2.

Physiological Variables

a. Mucin Turnover. The mucin turnover is expected to limit the residence time of the mucoadhesives. This is especially signicant, since the
mucoadhesive will eventually be detached from the surface of the eye ow-

Copyright 2003 Marcel Dekker, Inc.

418

Johnston et al.

ing to mucin turnover. However, the turnover rate may increase in the
presence of the mucoadhesive dosage form. An increase in the rate of mucus production generates a substantial amount of the soluble mucin molecules, which will interact with mucoadhesives before they have a chance
to attach to the mucus layer. This phenomenon has been demonstrated to
be true and unavoidable (41). The exact turnover rate of the mucus layer
remains to be determined. Although the thickness of the mucus layer in
contact with the epithelial cells is quite small, it is of similar magnitude to
the estimated mean diffusional path on the corneal surface.
b. Choice of Animal Model. The most commonly used animal model for ocular studies has been the albino rabbit, because of its ease of
handling, low cost, comparable eye size to those of humans, and a vast
amount of available information on its anatomy and physiology. This animal model seems to be less sensitive to ocular availability alterations from
viscosity changes in topical vehicles than humans (42). Since the blink rate
of rabbits (4 times/h) (43) is signicantly less than that of humans (15
times/min), humans commonly require higher viscosities than rabbits to
retain the drug on the corneal surface. A very important consideration in
using the rabbit model for evaluation of mucoadhesives as an ophthalmic
delivery device is the size of the drainage apparatus. Rabbits have one
large punctum that is capable of accommodating a large particle, whereas
humans have two small punctae in each eye. Cross-linked mucoadhesives
must be cleared from the eye. Thus, the drainage opening is an important
consideration.
A recent report (44) documented the species dierences in the eect of
polymeric vehicles on the corneal membrane disrupting action of benzalkonium chloride. The report suggested that the rabbit and human corneas
diered in the mucin glycocalyx domains at their surfaces. In view of this
report, more work needs to be done to delineate these dierences. The
animal model, which may be predictive of behavior of the ocular delivery
systems in humans, plays an important role in the iterative process of design
and evaluation of such systems. The lack of absolute predictability of the
rabbit model relative to vehicle eects on ocular drug bioavailability may be
attributed in part to the dierences between rabbits and human subjects
with respect to the anatomy and physiology of the precorneal area and
the cornea itself.
c. Disease States. The physicochemical properties of the mucus are
known to change during various pathological conditions such as the common cold, bacterial and fungal infections, and inammatory conditions of
the eye (4547). The exact structural changes taking place in mucus under
these conditions are not clearly understood. The problems presented by

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

419

such a complex and changing biological milieu for potential adhesion represent a unique challenge to pharmaceutical scientists. If mucoadhesives
are to be used in the diseased states, the bioadhesion property needs to be
evaluated under identical experimental conditions.
Many physiological factors of normal and diseased eyes aect the
performance of the delivery system. The rate of tear turnover and composition of the preocular tear lm changes in various pathological conditions.
The vehicles used in the formulation may also contribute to direct stimulation of the epithelial layers of the cornea or conujunctiva and cause release
of enzymes, glycoproteins, or immunological factors. In pathologies involving mucus-secreting epithelial cells, hypersecretion is more common than
hyposecretion. Mucus hyposecretion results in disruption of the tear lm
and dry spot formation. An excess of mucus occurs in a number of disease
states such as neuroparaly tickeratitis and keratoconjunctivitis sicca. Under
these conditions, the degree of sulfation of the mucin layer is also known to
increase (7). Blinking mixes the secretions and removes tear lm debris. In
addition, the dosage form location will contribute to the composition and
rate of tear secretion in a variety of ways. Thus, in both normal and diseased
eyes, the eects of the adhesive dosage form on ocular physiology may
determine the ultimate therapeutic outcome.

3.

Dosage Form Effects

a. Extent of Drug Incorporation. The drug may be loaded onto the

polymer matrix in a variety of ways. The most common approach is to incorporate the drug into mucoadhesive drug delivery systems, as shown in
Figure 4. For water-soluble polymers, it is possible to employ the mucoadhesive as a typical polymer to coat or to laminate a device. The contact time in such cases is rate limited by the dissolution of the polymer.
Such systems, however, suffer from the disadvantage of having a short
shelf life, because of the undesirable release of the drug in the aqueous environment (moisture) of the storage container.
The cross-linked mucoadhesives need to become hydrated to function
as an eective mucoadhesive drug delivery device. In such cases, the adhesive often detaches itself from the rate-controlling drug delivery device and
causes a premature release of the drug, especially with water-soluble drugs.
One solution to such a problem is through incorporation of a sparingly
soluble drug inside the mucoadhesive polymer. The device can slowly provide drug release until dissolution is complete. This approach may be used
for sparingly soluble salts and lipophilic prodrugs of highly water-soluble
drugs.

Copyright 2003 Marcel Dekker, Inc.

420

Johnston et al.

Figure 4 Approaches to incorporate drug into mucoadhesive drug delivery systems. Top panels: (*) mucoadhesive. Bottom panels: (*) drug.

Achieving high consistent loading compounds into mucoadhesive

dosage forms continues to be a signicant problem. A combination of a
phase change and mucoadhesive concept may be used in some cases. For
example, polycarbophil exhibits large changes in viscosity with pH and ionic
strength. Thus, drugs either alone or with a small quantity of lipid can be
suspended with the polycarbophil at the proper pH and ionic strength.
When such a suspension is placed in the tear pocket of the eye, it rapidly
gels, trapping the drug. The newly formed gel then releases the drug slowly
over prolonged periods of time.
b. Vehicular Effects. A delivery system that would allow the drug
to remain associated with a vehicle possessing enhanced precorneal retention may, therefore, provide for an attractive ocular drug delivery system.
Most attention has been directed toward the inuence of solution viscosity, although it has been shown that the maximum ocular availability increase obtained in rabbits by increasing the viscosity of the vehicle is
approximately twice that from a simple aqueous solution (7). The viscosity effect of mucoadhesives on ocular delivery, therefore, needs to be addressed.
According to Swan (48), maintaining the lubricity and viscosity of the
precorneal lm is as important as maintaining the isotonicity and pH of the

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

421

ophthalmic solutions. When aqueous solutions are instilled in the eye, the
integrity of the precorneal lm is altered. However, when 1% methylcellulose solution is instilled into the eye, it spreads evenly over the surface of the
globe, imparts viscosity to the precorneal lm, and causes minimal alteration in the integrity of the precorneal lm. The viscosity of the methylcellulose solution prevents it from being washed rapidly from the eye and
maintains the normal physiology, i.e., lacrimation. A combination of
these eects increases contact time, which prolongs the absorption of
drugs like homatropine (49). Oechsner and Keipert (50) have altered a
polyacrylic acid (PAA) aqueous formulation for dry eyes by including a
second polymer. Since PAA solutions have the disadvantage that PAA
builds high viscous gels in the usual concentration of 0.2% and at physiological pH, the authors have included polyvinylpyrrolidone (PVP) in the
ocular formation (50). After full hydration of the PAA polymer, a dispersion containing 2% PA was prepared and combined with an aqueous solution of PVP at a temperature of 30 C while stirring (50). A 10% aqueous
solution of NaCl was then added, which resulted in a clear preparation and
the formulations then made isotonic with mannitol and stabilized with
0.01% EDTA (50). The net eect of the addition of the PVP to the PAA
solution was a signicant reduction in the apparent viscosity of the formulation such that the preparation was nonirritating to ocular tissues (50). The
mucoadhesion index (as a measure of bioadhesive strength) was determined
for the experimental PAA/PVP formulations and found to possess greater
mucoadhesivity compared to monopolymer formulations that employed
PAA alone. It was postulated that perhaps sustained or prolonged delivery
of both hydrophilic and lipophilic drugs may be possible by incorporation
of either in a PAA complex with PVP (50).
Increasing contact time with methylcellulose ophthalmic vehicles has
been found to be preportional to its viscosity for up to about 25 cps. This
eect has been found to level o at 55 cps (51). In humans, a signicant
reduction in the drainage rates was observed with higher concentrations of
polyvinyl alcohol (5.85%) and with 0.9% hydroxypropyl methylcellulose
(42). However, it appears that in order to achieve the substantial reduction
in drainage rate, abnormally high viscosities are required.
Physicochemical parameters of the viscosity-imparting agents, other
than those related to viscosity eects, may also inuence the corneal retention as well as ocular bioavailability from an ophthalmic product. Benedetto
et al. (52) examined this eect using an in vitro model of the corneal surface
and suggested that polyvinyl alcohol, but not hydroxypropyl methylcellulose, would signicantly increase the thickness of the corneal tear lm.
However, such aects are considered to be only minimal. Davies et al.,
using rabbits, demonstrated not only a signicant increase in the precorneal

Copyright 2003 Marcel Dekker, Inc.

422

Johnston et al.

clearance, but also a signicant increase in the bioavailability of pilocarpine

when administered as a mucoadhesive polymeric solution (Carbopol 934P)
as compared to an equivoscous, nonmucoadhesive polyvinyl alcohol (PVA)
solution (53). Studies conducted to evaluate vehicle-drug (Carbopol 934Ppilocarpine) association indicated no binding of the pilocarpine to the polymer at physiological pH (53). Huupponen et al., using albino rabbits, combined the myriatic and cycloplegic agent, cyclopentolate, with either
polygalacturonic or hyaluronic acid and determined whether the mydriatic
response was increased compared to cyclopentolate base alone (54).
Although treated eyes all demonstrated approximately the same time to
reach a maximum mydriatic response when compared to cyclopentolate
base alone, the cyclopentolate/polygalacturonic (CY-PGA) and cyclopentolate/hyaluronic acid (CY-HA) formulations demonstrated a signicant
increase in the maximal mydriatic response when compared to the base
alone. However, only the CY-PGA formulation demonstrated a signicant
(p < 0:05) increase in the ocular bioavailability of cyclopentolate (54).
Increasing the contact time in the precorneal area appears to be governed by both the mucoadhesive agent as well as the viscosity eects of the
polymer. Thus, in designing the ocular drug delivery systems using mucoadhesives, one needs to nd a vehicle that imparts good mucoadhesive strength
as well as high viscosity at a low concentration.

II.

CURRENT STATUS OF MUCOADHESIVES IN OCULAR

DRUG DELIVERY

The successful development of newer mucoadhesive dosage forms for ocular

delivery still poses numerable challenges. Particularly important among
these are the determination of the exact nature of the interactions occurring
at the tissue mucoadhesive interface and the development of an ideal, nontoxic, nonimmunogenic mucoadhesive for clinical application. Moreover, a
better understanding of the exact physical structure of mucin molecules by
computational chemistry may aid in the calculation of the mucoadhesive
strength.
The pioneering work of Hui and Robinson (55) illustrated the utilization of bioadhesive polymers in the enhancement of ocular bioavailability of
progesterone (Fig. 5). Subsequently, several natural and synthetic polymers
have been screened for their ability to adhere to mucin epithelial surfaces;
however, little attention has been paid to their use in ophthalmic drug
delivery.
Saettone et al. (56) undertook a study evaluating the ecacy of a series
of bioadhesive dosage forms for ocular delivery of pilocarpine and tropica-

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

423

Figure 5 Progesterone levels in aqueous humour following topical administration

of 0.3% suspension of progesterone-entrapped polymer (*) and 0.3% suspension of
progesterone without polymer (*).

mide. From this study, hyaluronic acid emerged as the most promising
mucoadhesive agent. The biological analysis data, however, revealed that
the physicochemical properties of the drug itself had an impact on the
ecacy of the delivery system.
To retard rapid drug loss from the precorneal area, various devices
have been tested. Some of the potential candidates have been:
1. Erodible inserts of polyvinyl alcohol lm or silicone rubber for
the ocular delivery of pilocarpine and oxytetracycline, respectively (57,58)
2. Poly(vinyl methyl ether-maleic anhydride) matrices containing
timolol (59)
3. Polycyanoacrylate nanoparticles to improve the corneal penetration of hydrophilic drugs (60)
4. An aqueous dispersion with limited water solubility (61)
5. In situ forming gel preparation (62)
6. Sustained-release liposomes coated with a mucoadhesive polymer (63)
7. Microsphere preparations (64)
8. Mucoadhesive polysaccharides (65)

Copyright 2003 Marcel Dekker, Inc.

424

Johnston et al.

Ocular inserts (cylindrical rods) fabricated from medical grade silicone

rubber have also been evaluated for sustained delivery of oxytetracycline
(OXT) when placed in the upper or lower conjunctival fornix (58).
Cylindrical rods (diameter 0.9 mm, length 612 mm, weight 38 mg) all
containing OXT were prepared from mixtures of silicone elastomer, OXT,
and sodium chloride as a release modier. A stable polyacrylic acid (PAA)
or polymethacrylic acid (PMA) interpenetrating polymer network (IPN;
30% or 46% w/w) was grafted onto the inserts surface by treatment with
a mixture of acrylic (or methacrylic) acid and ethylene glycol dimethacrylate
in xylene at 100 C. This grafting procedure was employed since the hydrophobic silicone rubber would not be mucoadhesive to the hydrophilic palpebral and scleral mucosae. The thickness of the IPN layer was positively
correlated with the strength of mucoadhesion in vitro with the PMA IPN
grafting causing a lower rate of release of OXT than the IPN graft comprised of PAA (58). Release of OXT in vitro was zero-order and spanned
nearly a week for some of the PMA IPN grafted silicone rods. When tested
in rabbits, some of the IPN grafted inserts maintained in the lacrimal uid
an OXT concentration of 2030 g/mL for several days; an OXT concentration sucient for killing microorganisms responsible for common ocular
infections (58).
The use of poly(vinyl methyl ether-maleic anhydride) (PVMMA) ocular inserts containing timolol was shown to have both advantages and limitations. Timolol, a nonselective -adrenergic antagonist widely used to treat
open-angle glaucoma, can produce unwanted respiratory and cardiovascular side eects when systematically absorbed following ocular instillation.
Using pigmented rabbits, Lee et al. (66) demonstrated that timolol/
PVMMA inserts released the drug relatively slowly ( 50% of the loading
dose in 6 h) in vitro but increased the extent of systemic timolol absorption
(AUC). The timolol/PVMMA inserts reduced the peak timolol concentration in plasma (Cmax ) and signicantly delayed the time at which the timolol
Cmax was attained, raising the possibility that delayed timolol absorption
occurred until the timolol/PVMMA inserts were discharged into the nasal
cavity (66).
Poly(alkylcyanoacrylate) nanoparticles, specically those formulated
with poly(isobutylcyanoacrylate), have been systematically evaluated in
vitro by Das et al. (67). Several formulation variables were assessed in a
factorial design, including the use of dextran T40 or T70 and PluronicTM F68 or TweenTM 20 acting as stabilizer and surfactant, respectively, and three
pH levels (2, 4, and 7). Signicant eects of pH, surfactant, and stabilizer
were noted on the molecular weight and size distribution of the timololcontaining nanoparticles (67). As an example, the greatest percent yield of
formulated timolol nanoparticles was the PluronicTM F-68 at pH 2. These

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

425

authors concluded that formulation variables are extremely important for

optimal in vivo performance of all poly(alkylcyanoacrylate)-based nanoparticles.
The aqueous dispersion upon instillation into the eye generated an
apparent opaque mass, which adhered to and stayed in the lower fornix
for extended periods of time. The slow dissolution of the polymer itself
and diusion of pilocarpine out of the polymeric matrix probably controls
the availability of the drug for ocular absorption (68).
Similar results of pilocarpine ocular availability enhancement due to
precorneal retention through mucoadhesion has been reported by Saettone
et al. (69). This group investigated the use of a series of bioadhesive polymers, including hyaluronic acid, polygalacturonic acid, mesoglycan, and
carboxymethylchitin.
In situ activated gel-forming systems can be described as viscous
liquids, which undergo a transition to the gel phase upon exposure to certain
physiological conditions like a pH change or temperature change. The systems utilize polymers that demonstrate transition from a sol to a gel phase
when the dosage form is applied to the corneal surface. The phase transition
occurs either due to a change in the pH (e.g., from 4.5 to 7.4) or due to a
change in the temperature (from a lower temperature to the temperature of
the corneal surface). Josh et al. were the rst to report the use of a combination of polymers in such systems (70). They reported that a combination of
polymers elicit the desired characteristics when subjected to changes in the
physicochemical environment. Another factor that may aect this transition
is the change in electrolyte composition. Several polymers such as carbopol,
methylcellulose, pluronics, tetronics, and cellulose acetophthalate (CAP)
have been used for this purpose. In order to reduce polymer content,
Kumar and Himmelstein (71) developed a combination of polymers,
which included polyacrylic acid (PAA) and hydroxypropylmethylcellulose
(HPMC) for the release of timolol maleate. Various studies involving several
dierent polymers have shown to increase ocular bioavailability. Rozier et
al. found an improvement in the ocular absorption of timolol in the albino
rabbit when administered in Gelrite, when compared with an equiviscous
solution of hydroxyethylcellulose (72). Sanzigiri et al. compared various
systems of methylprednisolone (MP): ester of Gelrite eyedrops, gellan-MP
lm, and gellan lm with dispersed MP. The authors showed that the gellan
eyedrops provided signicant MP precorneal levels over a period of 6 hours
(73).
Davies et al., using rabbits demonstrated that liposomes containing
the mydriatic tropicamide and coated with either Carbopol 934P or
Carbopol 1342 (a hydrophobic modied Carbopol resin) displayed a signicant increase in precorneal retention at pH 5 compared to noncoated

Copyright 2003 Marcel Dekker, Inc.

426

Johnston et al.

tropicamide liposomes (63). Both polymer coatings of tropicamide-containing liposomes failed to signicantly increase the bioavailability of the
entrapped drug relative to uncoated vesicles (63). However, in contrast to
previous work (53), size and zeta potential measurements of uncoated tropicamide liposomes and liposomes coated in both polymer solutions demonstrated an association between the Carbopols and the vesicles at both pH 7.4
and pH 5, as evidenced by an increase in size and a decrease in the corresponding zeta potential. It was suggested that the prolongation in precorneal residence for Carbopol 1342coated tropicamide liposomes was due to
the formation of a three-dimensional microgel structure, which interacted
with the phospholipid vesicles and subsequently increased their retention via
a mechanism of adhesion to the mucin network (63).
Microsphere formulations have been evaluated for their capacity to
retain 111 In as indium chloride in the preocular (precorneal) area of the
rabbit eye (64). Clearance of the radiolabeled compound was monitored
using gamma scintigraphy, and the inuence of pH and prehydration of
the microspheres on precorneal retention was assessed. These authors prepared microspheres of poly(acrylic acid) (Carbopol 907) cross-linked with
maltose by a water-in-oil (w/o) emulsication process. Precorneal clearance
of the microspheres at pH 5 and 7.4 were compared to an 111 In aqueous
suspension (64). Clearance of the microspheres demonstrated a biphasic
(
rapid initial phase and  slower phase) prole, and microspheres
buered at pH 5 exhibited a signicantly slower  phase than microspheres
buered at pH 7.4. Presumably, the neutralized Carbopol formulation (pH
7.4) did not possess the same degree of mucoadhesive strength as the microsphere preparation formulated at pH 5. In vitro tests of mucoadhesive
strength veried that the force of detachment of the microspheres formulated at pH 5 from mucus glycoprotein was signicantly greater than the
corresponding value for microspheres buered at pH 7.4 (64). A signicant
increase in the retention of prehydrated microspheres in the preocular area
was observed compared to microsphere formulations that were not hydrated
prior to instillation (64).
Albasini and Ludwig (65) evaluated a series of polysaccharides for
their potential inclusion in ocular dosage forms. The polysaccharides evaluated were carrageenan, locust bean gum, guar gum, xanthan gum, and
scleroglucan. Measurements of the dynamic surface tension, pH, refractive
index, and a visual clarity check comprised the physical measurements and
determination of viscosity, viscoelasticity, eect of ionic strength on the
resulting viscosity, and mucoadhesive strength (as assessed by an increase
in the viscosity of a solution of the polysaccharide and mucin) comprised the
rheological analysis of all the polysaccharides evaluated (65). All ocular
preparations were made by adding the required amount of each polysac-

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

427

charide to an aqueous iso-osmotic vehicle at 90 C and stirring the mixture

mechanically until the polysaccharide was completely dissolved (65). Only
scleroglucan and xanthan gum were found to demonstrate desirable viscoelastic and mucoadhesive properties suitable for instillation into the eye. The
two preparations were evaluated in human volunteers and found to possess
no ocular irritancy (65). Polysaccharides have also been shown to have
potential for drug delivery by other routes of administration. Using a
high molecular weight polysaccharide gum, Hakea gibbosa, isolated from
a tree, Alur et al. have shown that the gum possessed both the ability to
sustain the release of low molecular weight, organic-based drug substances
as well as therapeutic polypeptides (salmon calcitonin) both in vitro and in
vivo following application to the buccal mucosa of rabbits (7476). This
polysaccharide may hold promise for sustained delivery to the eye of both
conventional drugs and newer therapeutic polypeptides requiring retention
of biological activity.
The biomaterials for ocular use have been mainly synthetic polymers.
Some natural biopolymers, such as collagen and hyaluronic acid, have also
been examined. Of these, hyaluronic acid oers attractive possibilities (77
80). Some of the materials indicated as ocular mucoadhesives are mentioned below.
A.

Naturally Occurring Mucoadhesives

Collagen and brin have been used as erodible insets for the long-term
delivery of pilocarpine to the eye (81,82). The utility of these macromolecules in ophthalmic drug delivery depends largely upon their attachment
capability to the drug molecules and their interaction with the glycocalyx
domain of the corneal surface for maximum mucoadhesion. Among these,
lectins and bronectin are most promising.
The role of lectins as cellular-recognition mediators has been explored
in great detail in the eld of cellular biology. Lectins belong to a class of
proteins of nonimmune origin that bind carbohydrates specically and noncovalently (83). The most commonly studied lectin is the one derived from
tomatoes. This particular lectin has been found to be nontoxic, binds specically to the sialic acids (a major component of the mucus glycoproteins),
and is transported into the cells by endocytosis (84). Such properties could
be useful for the delivery of therapeutic agents into the ocular chambers.
Fibronectin is a glycoprotein and a component of the extracellular
matrix. The pentapeptide backbone of this substance has been identied
as having a cell-attachment property (85). Puried bronectin has been
reported to lessen the healing time of corneal ulcers (86). It has also been
used in conjunction with hyaluronic acid for decreasing the healing time.

Copyright 2003 Marcel Dekker, Inc.

428

B.

Johnston et al.

Synthetic Mucoadhesives

As discussed earlier, the potential of a mucoadhesive agent is determined by

a number of parameters; i.e., chain length, conguration, and molecular
weight. The extent of corneal adhesion of some neutral polymers has been
reported to be comparable to that of natural mucins. Lemp and Szymanski
(87) measured the extent of corneal adsorption of water-soluble polymers
onto the epithelial surface. Among the polymers, 1.4% polyvinyl alcohol,
0.5% hydroxypropyl methyl cellulose, and 2% hydroxyethyl cellulose vehicles have shown comparable corneal adhesion to that of mucin. The study
concluded that ocular therapeutic agents would be well absorbed from topical formulations containing such polymers. Since only marginal improvements (two- to threefold) in ocular bioavailability were seen with these
agents, the adsorbed polymers were either unable to hold the drug or
were being rapidly removed from the surface by the bathing tears. Other
water-soluble polymers like polyacrylic acid also improved the ocular bioavailability of pilocarpine, albeit by a factor of two.
A polymer most eective as a mucoadhesive will be the one that can
form an extended and hydrated network to allow for greater interpenetration and subsequent physical entanglement. These kinds of networks may be
formed by:
1. Physical intertwining of the polymers
2. Bridging of the polymer chains
3. Cross-linking of the polymer chains
Thus, cross-linked polyacrylic acid has been shown to have an excellent mucoadhesive property, causing signicant enhancement in ocular bioavailability (55). Using pigmented rabbits, Lehr et al. (88) demonstrated a
twofold increase in the uptake of gentamicin by the bulbar conjunctiva when
the aminoglycoside was delivered as a mucoadhesive, polycarbophil [a
poly(acrylic acid)based polymer] formulation. Two gentamicin formulations of this polymer (neutralized vs. nonneutralized) were evaluated and
compared to an aqueous control formulation. While both the neutralized
(pH 7.5) and nonneutralized (pH 2.5) gentamicin/polycarbophil formulations increased the uptake of the aminoglycoside antibiotic by the bulbar
conjunctiva, only the nonneutralized aminoglycoside formulation provided
drug penetration into the aqueous humor. Penetration of gentamicin into
the aqueous humor from the nonneutralized formulation was suggested to
result from drug absorption via the conjunctival-scleral pathway facilitated
by intensied contact between the mucoadhesive polymer and the underlying bulbar conjunctiva (88). A partially esteried acrylic acid polymer was
successful in prolonging the therapeutic eect of topically applied pilocar-

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

429

pine (89). Urtti et al. (90) also indicated that the use of a polyacrylamide and
a copolymer of acrylamide (N-vinyl pyrrolidone and ethyl acrylate) as a
matrix, which resulted in a threefold increase in the ocular bioavailability
of pilocarpine.
Similarly, cyanoacrylates have been used in the eld of opthalmology
to seal corneal perforations and ulcers, to stop leakage of aqueous or vitreous humor, and to protect against external contamination (91,92). These
agents have a potential as eective ocular mucoadhesive agents provided the
monomer polymerization could be controlled.

III.

WHAT IS IN STORE FOR THE FUTURE?

A multidisciplinary approach will be necessary to overcome the challenges

associated with the development of ocular mucoadhesives. A nonbiodegradable adhesive is adequate for topical use to treat perforations and ulcerations. However, the use of a nontoxic biodegradable surgical adhesive is
deemed necessary for long-term use in ocular drug delivery.
Mucoadhesives can make an important contribution in this area. An ideal
ocular mucoadhesive would be site specic, durable for the desired period of
time, biodegradable, and above all nontoxic, nonimmunogenic, and nonirritant. It would be even better if the adhesive could serve as absorption
enhancers (for therapeutic protein and peptide drugs) and/or as enzyme
inhibitors.

REFERENCES
1. Harris, D., and Robinson, J. R. (1990). Bioadhesive polymers in peptide drug
delivery, Biomaterials, 11:652.
2. Schacter, H., and Williams, D. (1982). Biosynthesis of mucus glycoproteins,
Adv. Exp. Med. Biol., 144:3.
3. Marriot, C., and Gregory, N. P. (1990). Mucus physiology and pathology. In:
Bioadhesive Drug Delivery Systems (V. Lenaerts and R. Gurny, eds.). CRC
Press, Boca Raton, FL, p. 21.
4. Allen, A. (1981). The structure and function of gastrointestinal mucus. In:
Basic Mechanisms of Gastrointestinal Mucosal Cell Injury and Protection (J.
W. Harmon, ed.). Williams & Wilkins, Baltimore, p. 351.
5. Allen, A., and Carrol, N. J. H. (1985). Adherent and soluble mucus in the
stomach and duodenum, Dig. Dis. Sci., 30:55s.
6. Gandhi, R. B., and Robinson, J. R. (1988). Bioadhesion in drug delivery,
Indian J. Pharm. Sci., 50:145.

Copyright 2003 Marcel Dekker, Inc.

430

Johnston et al.

7. Marriot, C., and Hughes, D. R. L. (1989). Mucus physiology and pathology.

In: Bioadhesion: Possibilities and Future Trends (R. Gurny and H. E.
Junginger, eds.). Wissenschaftliche Verlagsgesellschaft, Stuttgart, p. 29.
8. Johnson, P. M., and Rainsford, K. D. (1978). The physical properties of
mucus: Preliminary observations on the sedimentation behavior of porcine
gastric mucus, Biochim. Biophys. Acta, 286:72.
9. Phelps, C. F. (1978). Biosynthesis of mucus glycoproteins, Br. Med. Bull.,
34:43.
10. Robinson, J. R. (1989). Ocular drug delivery. Mechanism(s) of corneal drug
transport and mucoadhesive delivery systems, S.T.P. Pharma, 5:839.
11. Mikos, A. G., and Peppas, N. A. (1986). Systems for controlled release of
drug. V. Bioadhesive systems, S.T.P. Pharm., 2:705.
12. Dedaguin, B. V., Toporov, Y. P., Mueler, V. M., and Aleinikova, I. N. (1977).
On the relationship between the electrostatic and the molecular component of
the adhesion of elastic particles to a solid surface. J. Colloid Interface Sci.,
58:528.
13. Baier, R.E., Shafrin, I.G., and Zisman, W.A. (1968). Adhesion: Mechanism
that assist and impede it, Science, 162:1360.
14. Helfand, E., and Tagami, Y. (1971). Theory of the interface between immiscible polymers, Polym. Lett., 9:741.
15. Helfand, E., and Tagami, Y. (1972). Theory of the interface between immiscible polymers. II. J. Chem. Phys., 56:3592.
16. Helfand, E., and Tagami, Y. (1972). Theory of the interface between immiscible polymers, J. Chem. Phys., 57:1812.
17. Peppas, N. A., and Buri, P. A. (1985). Surface, interfacial and molecular
aspects of polymer adhesion on soft tissues, J. Controlled Release, 2:257.
18. Peppas, N. A., and Lustig, B. R. (1985). The role of crosslinks, entanglements,
and relaxations of the macromolecular carrier in the diusional release of
biologically active materials: Conceptual and scaling relationships, Ann. NY
Acad. Sci., 446:26.
19. Reinhart, C. T., and Peppas, N. A. (1984). Solute diusion in swollen membranes. II. Inuence of crosslinking on diusive properties. J. Membr. Sci.,
18:227.
20. Peppas, N. A., and Reinhart, C. T. (1983). Solute diusion in swollen membranes. Part I. New therapy, J. Membr. Sci., 15:275.
21. Tabor, D. J. (1977). Surface forces and surface interactions, J. Colloid
Interface Sci., 58:2.
22. Kinloch, A. J. (1980). The science of adhesion: I. Surface and interfacial
aspects, J. Mater. Sci., 15:2141.
23. Good, R. J. (1977). Surface free energy of solids and liquids:
Thermodynamics, molecular forces and structure, J. Colloid Interface Sci.,
58:398.
24. Ponchel, G., Touchard, F., Duchene, D., and Peppas, N. A. (1987).
Bioadhesive analysis of controlled release systems. I. Fracture and interpene-

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

25.
26.
27.
28.
29.

30.

31.
32.

33.
34.

35.

36.
37.

38.
39.

40.

431

tration analysis in poly(acrylic acid) containing systems, J. Controlled Release,

5:129.
Mikos, A. G., and Peppas, N. A. (1988). Polymer chain entanglements and
brittle fracture, J. Chem. Phys., 88:1337.
Wake, W. C. (1976). Theories of adhesion and adhesive action. In: Adhesion
and Formulation of Adhesives. Applied Science, London, p. 65.
Israelachvili, J. N. (1985). Intermolecular and Surface Forces. Academic Press,
New York.
Pritchard, W. H. (1971). The role of hydrogen bonding in adhesion, Aspects
Adhes., 6:11.
Colo, G. D., Bugalassi, S., Chetoni, P., Fiaschi, M. P., Zambito, Y. and
Sattone, M. F. (2001). Relevance of polymer molecular weight to the in
vitro/in vivo performances of ocular inserts based on poly(ethylene oxide),
Int. J. Pharmaceutics, 220:169.
Park, H., and Robinson, J. R. (1985). Physico-chemical properties of water
insoluble polymers important to mucin/epithelial adhesion, J. Controlled
Release, 2:47.
Nagai, T., and Machida, Y. (1985). Advances in drug delivery. Mucosal
adhesive dosage forms, Pharma. Int. Engl. Ed., August:196.
Satoh, K., Takayama, K., Machida, Y., Suzuki, Y., Nakagaki, M., and
Nagai, T. (1989). Factors aecting the bioadhesive properties of tablets consisting of hydroxypropyl cellulose and carboxy vinyl polymers, Chem. Pharm.
Bull., 37:1366.
Gottschalk, A. (1960). In: The Chemistry and Biology of Sialic Acid and
Related Substances. Cambridge University Press, London.
Leonard, F., Hodge, J. W., Jr., Houston, S., and Ousterhout, D. K. (1968).
Alpha cyanoacrylate adhesive bond strengths with proteinaceous and non
proteinaceous substances, J. Biomed. Mater. Res., 2:173.
Smart, J. D., Kellaway, I. W., and Worthington, H. E. C. (1984). An in vitro
investigation of mucosal adhesive materials for use in controlled drug delivery,
J. Pharm. Pharmacol., 36:295.
Leung, S. H. S., and Robinson, J. R. (1990). Polymer structure features contributing to mucoadhesion. II., J. Controlled Release, 12:187.
Chen, J. L., and Cyr, G. N. (1970). Compositions producing adhesion through
adhesion. In: Adhesive Biological Systems (R. S. Manly, ed.). Academic Press,
New York, p. 78.
Smart, J. D. (1991). An in vitro assessment of some muco-adhesive dosage
forms, Int. J. Pharm., 73:69.
Park, K., and Park, H. (1990). Test methods of bioadhesion. In: Bioadhesive
Drug Delivery Systems (V. Lenaerts and R. Gurny, eds.). CRC Press, Boca
Raton, FL, p. 43.
Wang, P. Y., and Forrester, D. H. (1974). Conditions for the induced adhesion of hydrophobic polymers to soft tissue, Trans. Am. Soc. Artif. Int.
Organs, 20:504.

Copyright 2003 Marcel Dekker, Inc.

432

Johnston et al.

41. Teng, C. L. C., and Ho, N. F. L. (1987). Mechanistic studies in the simultaneous ow and adsorption of polymer coated latex particles on intestinal
mucus. I: Methods and physical model development, J. Controlled Release,
6:133.
42. Zaki, I., Fitzgerald, P., Hardy, J. G., and Wilson, C. G. (1986). A comparison
of the eect of viscosity in the precorneal residence of solutions in rabbit and
man, J. Pharm. Pharmacol., 38:463.
43. Moss, R. A. (1987). The eyelids. In: Adlers Physiology of the Eye: Clinical
Applications, 8th ed. (R. A. Moses and W. M. Hart, Jr., eds.). Mosby, St.
Louis, p. 1.
44. Saettone, M. F., Giannaccinni, B., Guiducci, A., LaMarca, F., and Tota, G.
(1985). Polymer eects on ocular bioavailability. II: The inuence of benzalkonium chloride on the mydriatic response of tropicamide in dierent polymeric vehicles, Int. J. Pharm., 25:73.
45. Wright, P., and Mackie, I. A. (1977). Mucus in healthy and diseased eye,
Trans. Ophthalmol. Soc. UK, 91:1.
46. Hardy, J. G., Lee, S. W., and Wilson, C. G. (1985). Intranasal drug delivery
by sprays and drops, J. Pharm. Pharmacol., 37:294.
47. Tabachnik, N. F., Blackburn, P., and Cerami, A. (1981). Biochemical and
rheological characteristics of sputum mucus from a patients with cystic brosis, J. Biol. Chem., 256:7161.
48. Swan, K. C. (1945). Use of methylcellulose in ophthalmology, Arch.
Ophthalmol., 33:378.
49. Mueller, W. H., and Deardor, D. L. (1956). Ophthalmic vehicles: The eect
of methylcellulose on the penetration of Homatropine hydrobromide through
the cornea, J. Am. Pharm. Assoc., 45:334.
50. Oechsner, M., and Keipert, S. (1999). Polyacrylic acid/polyvinylpyrrolidone
biopolymeric systems. I. Rheological and mucoadhesive properties of formulations potentially useful for the treatment of dry-eye syndrome, Eur. J.
Pharm. Biopharm., 47:113118.
51. Saettone, M. F., Giannaccinni, B., Ravecca, S., La Marca, F., and Tota, G.
(1984). Polymer eects of ocular bioavailability: The inuence of dierent
liquid vehicles on the mydriatic response of tropicamide in humans and in
rabbits, Int. J. Pharm., 20:187.
52. Benedetto, D. A., Shah, D. O., and Kaufman, H. E. (1975). Instilled uid
dynamics and surface chemistry of polymers in the preocular tear lm, Invest.
Ophthalmol., 14:887.
53. Davies, N. M., Farr, S. J., Hadgraft, J., and Kellaway, I. W. (1991).
Evaluation of mucoadhesive polymers in ocular drug delivery. I. Viscous
solutions, Pharm. Res., 8:10391043.
54. Huupponen, R., Kaila, T., Saettone, M. F., Monti, D., Iisalo, E., Salminen,
L., and Oksala, O. (1992). The eect of some macromolecular ionic complexes
on the pharmacokinetics and dynamics of ocular cyclopentolate in rabbits, J.
Ocular Pharmacol., 8:5967.

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

433

55. Hui, H. W., and Robinson, J. R. (1985). Ocular delivery of progesterone using
a bioadhesive polymer, Int. J. Pharm., 26:203.
56. Saettone, M. F., Chetoni, P., Torracca, M. T., Burgalassi, S., and
Giannaccinni, B. (1989). Evaluation of mucoadhesive properties and in vivo
activity of ophthalmic vehicles based on hyaluronic acid, Int. J. Pharm.,
51:203.
57. Saettone, M. F., Giannaccini, B., Chetoni, P., Galli, G., and Chiellini, E.
(1984). Vehicle eects on the ophthalmic bioavailability: An evaluation of
polymeric inserts containing pilocarpine, J. Pharm. Pharmacol., 36:229.
58. Chetoni, P., Di Colo, G., Grandi, M., Morelli, M., Saettone, M. F., and
Darougar, S. (1998). Silicone rubber/hydrogel composite ophthalmic inserts:
Preparation and preliminary in vitro/in vivo evaluation, Eur. J. Pharm.
Biopharm., 46:125132.
59. Finne, U., Salivirta, J., and Urtti, A. (1991). Sodium acetate improves the
ocular/systemic absorption ratio of timolol applied ocularly in monoisopropyl
PVM-MA matrices, Int. J. Pharm., 75:R1.
60. Losa, C., Calvo, P., Castro, E., Vila-Jato, J. L., and Alono, M. J. (1991).
Improvement of ocular penetration of amikacin sulphate by association to
poly(butylcyanoacrylate) nanoparticles, J. Pharm. Pharmacol., 43:548.
61. Vanderho, J., El-Asser, E. R., and Urgerstad, J. (1977). U.S. Patent
Application #867031.
62. Kumar, S., Haglund, B. O., and Himmelstein, K. J. (1994). In situ-forming
gels for opthalmic drug delivery, J. Ocular Pharmacol. 10:47.
63. Davies, N. M., Farr, S. J., Hadgraft, J., and Kellaway, I. W. (1992).
Evaluation of mucoadhesive polymers in ocular drug delivery. II. Polymercoated vesicles, Pharm. Res., 9:11371144.
64. Durrani, A. M., Farr, S. J., and Kellaway, I. W. (1995). Precorneal clearance
of mucoadhesive microspheres from the rabbit eye, J. Pharm. Pharmacol.,
47:581584.
65. Albasini, M., and Ludwig, A. (1995). Evaluation of polysaccharides intended
for ophthalmic use in ocular dosage forms, Farmaco, 50:633642.
66. Lee, V. H. L., Li, S. Y., Sasaki, H., Saettone, M. F., and Chetoni, P. (1994).
Inuence of drug release rate on systemic timolol adsorption from polymeric
occular inserts in the pigmented rabbit, J. Ocular. Pharmacol., 10:421429.
67. Das, S. K., Tucker, I. G., Hill, D. J. T., and Ganguly, N. (1995). Evaluation of
poly(isobutylcyanoacrylate) nanoparticles for mucoadhesive ocular drug
delivery. I. Eect of formulation variables on physicochemical characteristics
of nanoparticles, Pharm. Res., 12:534540.
68. Robinson, J. R., and Li, V. H. K. (1984). Ocular disposition and bioavailability of pilocarpine from Piloplex1 and other sustained release drug delivery
systems. In: Recent Advances in Glaucoma (U. Ticho and R. David, eds.).
Excerpta Medica, Amsterdam, p. 231.
69. Saettone, M. F., Giannaccinni, B., Guidicci, A., and Savigni, P. (1986).
Semisolids ophthalmic vehicles. III. An evaluation of four organic hydrogels
containing pilocarpine, Int. J. Pharm., 31:261.

Copyright 2003 Marcel Dekker, Inc.

434

Johnston et al.

70. Joshi, A., Ding, S., and Himmelstein, K. J. Patent application U.S. 91/04104
(Publication No. WO 91/19481).
71. Kumar, S., and Himmelstein, K. J. (1995). Modication of in situ gelling
behavior of carbopol solutions by hydroxypropyl methylcellulose. J. Pharm.
Sci., 84:344.
72. Rozier, A., Mazuel, C., Grove, J., and Plazonnet, B. (1989). Gelrite: a novel,
ion activated, in situ gelling polymer for ophthalmic vehicles. Eect on bioavailability of timolol, Int. J. Pharmaceutics, 57:163.
73. Sanzigiri, Y. D., Maschi, S., Crescenzi, V., Callegaro, L., Topp, E. M., and
Stella, V. J. (1993). Gellan-based systems for ophthalmic sustained delivery of
methylprednisolone. J. Control Release, 26:195.
74. Alur, H. H., Pather, S. I., Mitra, A. K., and Johnston, T. P. (1999).
Transmucosal delivery of chlorpheniramine maleate from a buccal tablet containing a natural, mucoadhesive gum, Int. J. Pharm., 188:110.
75. Alur, H. H., Beal, J. D., Panther, S. I., Mitra, A. K., and Johnston, T. P.
(1999). Evaluation of a novel, natural oligosaccharide gum as a sustainedrelease and mucoadhesive component of calcitonin buccal tablets, J. Pharm.
Sci., 88:13131319.
76. Alur, H. H., Pather, S. I., Mitra, A. K., and Johnston, T. P. (1999).
Evaluation of the gum Hakea gibbosa as a sustained-release and mucoadhesive component in buccal tablets, Pharm. Develop. Technol., 4:347358.
77. Benditti, L. M., Kyyronen, K., Hume, L., Topp, E., and Stella, V. (1991).
Steroid ester of hyaluronic acid in ophthalmic drug delivery, Proc. Int. Symp.
Controlled Release Bioact. Mater., 18:497.
78. Saettone, M. F., Giannaccinni, B., Torracca, M. T., and Burgalassi, S. (1987).
An evaluation of the bioadhesive properties of hyaluronic acid. Proceedings of
the 3rd Eur. Congress on Biopharm. Pharmacokinetics, Vol. 1, Freiburg, April,
p. 413.
79. Saettone, M. F., Chetoni, P., Torracca, M. T., Giannaccinni, B., and Ordello,
G. (1986). Evaluation of hyaluronic acid as a vehicle for topical ophthalmic
drugs. Abstr. of Int. Symp. Ophthalmic Dosage Forms, Pisa, Oct. 1314.
80. Saettone, M. F., Chetoni, P., and Giannaccinni, B. (1985). Evaluation of
hyaluronic acid as a vehicle for topical ophthalmic drugs, Abstr. of 2nd Int.
Conference on Polymers in Medicine, Capri, June 37.
81. Miyazaki, S., Ishii, K., and Takada, M. (1982). Use of brin lm as a carrier
for drug delivery: A long acting delivery system for pilocarpine into the eye.
Chem. Pharm. Bull., 30:3405.
82. Bloomeld, S. E., Miyata, T., Dunn, M. W., Bueser, N., Stenzel, K. H., and
Rubin, A. L. (1978). Soluble gentamicin ophthalmic inserts as a drug delivery
system, Arch. Ophthalmol., 96:885.
83. McCoy, J. P., Jr. (1986). Contemporary laboratory applications of lectins,
Biotechniques, 4:252.
84. Kilpatrick, D. C., Pusztai, A., Grant, G., Graham, G., and Ewen, S. W. B.
(1985). Tomato lectins resist degradation in the mammalian alimentary canal
and binds to intestinal villi without deleterious eects, FEBS Lett., 185:299.

Copyright 2003 Marcel Dekker, Inc.

Mucoadhesive Polymers in Ophthalmic Drug Delivery

435

85. Hynes, R. O., and Yamada, K. M. (1982). Fibronectins: Multifunctional

modular glycoproteins, J. Cell Biol., 95:369.
86. Nishida, T., Ohashi, Y., Awanta, T., and Manabe, R. (1983). Fibronectin,
new therapy for corneal trophic ulcer, Arch. Ophthalmol., 101:1046.
87. Lemp, M. H., and Szymanski, E. S. (1975). Polymer adsorption at the ocular
surface, Arch. Ophthalmol., 99:134.
88. Lehr, C. M., Lee, Y. H., and Lee, V. H. L. (1994). Improved ocular penetration of gentamicin by mucoadhesive polymer polycarbophil in the pigmented
rabbit, Invest. Ophthalmol. Vis. Sci., 35:28092814.
89. Ticho, U., Blumenthal, M., Zonis, S., Gal, A., Blank, L., Mazor, Z. W. (1979).
A clinical trial with Piloplex1a new long-acting pilocarpine compound:
Preliminary report, Ann. Ophthalmol., April:555.
90. Urtti, A., Salminen, L., Kujari, H., and Jantti, V. (1984). Eect of ocular
pigmentation on pilocarpine pharmacology in the rabbit eye. II. Drug
response, Int. J. Pharm., 19:53.
91. Refojo, M. F. (1982). Current status of biomaterials in ophthalmology, Surv.
Ophthalmol., 26:257.
92. Refojo, M. F., Dohlman, C. H., and Koliopoulos, J. (1971). Adhesives in
ophthalmology: A review, Surv. Ophthalmol., 15:217.

Copyright 2003 Marcel Dekker, Inc.

14
Microparticles and Nanoparticles in
Ocular Drug Delivery
Murali K. Kothuri, Swathi Pinnamaneni,
Nandita G. Das, and
Sudip K. Das
Idaho State University, Pocatello, Idaho, U.S.A.

I.

INTRODUCTION

Controlled and sustained delivery of ophthalmic drugs continues to remain

a major focus area in the eld of pharmaceutical drug delivery with the
emergence of new, more potent drugs and biological response modiers
that may also have very short biological half-lives. The major objective of
clinical therapeutics is to provide and maintain adequate concentration of
drugs at the site of action. In ocular drug delivery, the physiological constraints imposed by the protective mechanisms of the eye lead to poor
absorption of drugs with very small fractions of the instilled dose penetrating the cornea and reaching the intraocular tissues. The reasons for inecient drug delivery include rapid turnover, lacrimal drainage, reex blinking,
and drug dilution by tears (1,2). The limited permeability of cornea also
contributes to the low absorption of ocular drugs. As shown in Figure 1,
tear drainage causes a major portion of the administered dose to be transported via the nasolacrimal duct to the gastrointestinal (GI) tract, where it
may be absorbed, leading to unwanted systemic side eects and occasional
toxicity due to the drug (3). The rapid elimination of administered eye drops
often results in a short duration of the ocular therapeutic eect, making a
frequent dosing regimen necessary.

Current aliation: Bristol-Myers Squibb Company, New Brunswick, New Jersey, U.S.A.

437

Copyright 2003 Marcel Dekker, Inc.

438

Kothuri et al.

Fig. 1 Schematic diagram of ocular distribution. (From Ref. 3.)

Three main factors have to be considered when drug delivery is

attempted to the intraocular tissues (4): (a) how to cross the blood-eye
barrier (systemic to ocular) or cornea (external to ocular) to reach the site
of action, (b) how to localize the pharmacodynamic action at the eye and
minimize drug action on other tissues, and (c) how to prolong the duration
of drug action such that the frequency of drug administration can be
reduced. There is a clear need for eective topical formulations providing
superior bioavailability of drugs along with a reasonable frequency of application, and the goals during the development of appropriate delivery systems should include increased contact time of the drug with the eye surface
and promotion or facilitation of transfer of drug molecules from the tear
phase into the eye tissue without causing any inconvenience to the patient.
Approaches to optimize drug residence time have included prolongation of
precorneal drug retention by the use of viscous gels, colloidal suspensions,
and erodible or nonerodible inserts (5,6). Besides the issue of bioavailability,
patient compliance and comfort considerations in drug instillation are very
important factors that may impact the drugs therapeutic ecacy. Patient

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

439

discomfort excludes inserts from the list of potentially popular drug delivery
systems. Although adding soluble polymers to ophthalmic solutions can
increase drug retention by increasing viscosity and decreasing the rate of
drainage, there are problems associated with viscous solutions during their
manufacture and administration that usually result in vision blurring, limiting their chances of becoming popular dosage forms. Liposomes have been
extensively investigated as ocular drug delivery vehicles for over a decade as
they oer potential benets of controlled and sustained drug release and
protection from metabolic processes while the therapeutic agent remains
sequestered within the vesicles. But the problems associated with liposomes
are possible toxicity and irritability. The primary factors that determine the
relative toxicity of liposomes appear to be the lipid composition and irritability associated with the charge of liposomes (710). These factors may
limit their chances of becoming popular ocular dosage forms of the future.
Attempts to reduce systemic absorption have been made based on the design
of prodrug derivatives with a higher lipophilic character (1113). However,
results obtained from these studies are inconclusive because most prodrugs
are unstable in an aqueous solution.
In order to address the above-stated problems, micro- and nanotechnology involving drug-loaded polymer particles has been proposed as an
ophthalmic drug delivery technique that may enhance dosage form acceptability while providing sustained release in the ocular milieu (14).
Particulate drug delivery consists of systems described as microparticles,
nanoparticles, microspheres, nanospheres, microcapsules, and nanocapsules. They consist of macromolecular materials and can be used therapeutically by themselves, e.g., as adjuvant in vaccines, or as drug carriers,
in which the active principle (drug or biologically active material) is dissolved, entrapped, encapsulated, and/or to which active principle is
absorbed, adsorbed, or attached. Particles ranging from 100 nm to the
order of several hundred micrometers are included in the microparticulate
category, which is divided into two broad groups (15): microcapsules are
almost spherical entities of the order of several hundred micrometers in
diameter where the drug particles or droplets are entrapped inside a polymeric membrane, and microspheres are polymer-drug combinations where
the drug is homogeneously dispersed in the polymer matrix. Nanoparticles
possess similar characteristics as microparticles, except their size is
approximately three orders of magnitude smaller (< 1 m).
Nanoparticles are also subdivided into two groups: nanospheres and nanocapsules (11). Nanospheres are small solid monolithic spheres constituted
of a dense solid polymeric network, which develops a large specic area
(16). The drug can be either incorporated or adsorbed onto the surface.
Nanocapsules are small reservoirs consisting of a central cavity (usually an

Copyright 2003 Marcel Dekker, Inc.

440

Kothuri et al.

oily droplet containing dissolved drug) surrounded by a polymeric membrane. Several studies have shown nanoparticles to be more stable in
biological uids and during storage compared to other carriers that are
similar in size distribution and controlled-release properties, such as liposomes. Furthermore, they can entrap and retain various drug molecules in
their stable state.
Polymers used for the preparation of microparticulates may be erodible, biodegradable, nonerodible, or ion exchange resins (17). Nanoparticles
made of nonbiodegradable polymers are neither digested by enzymes nor
degraded in vivo through a chemical pathway (18). The risk of chronic
toxicity due to the intracellular overloading of nondegradable polymers
would be a limitation of their systemic administration to human beings,
making these materials more suitable for removable inserts or implants.
Erodible systems have an inherent advantage over other systems in that
the self-eroding process of the hydrolyzable polymer obviates the need for
their removal or retrieval after the drug is delivered. Upon the administration of particle suspension in the eyes, particles reside at the delivery site and
the drug is released from the polymer matrix through diusion, erosion, ion
exchange, or combinations thereof (19).
Nanoparticles, when formulated properly, provide controlled drug
release and prolonged therapeutic eect. To achieve these characteristics,
particles must be retained in the cul-de-sac after topical administration,
and the entrapped drug must be released from the particles at an appropriate rate. As mentioned before, the utility of nanoparticles as an ocular
drug delivery system may depend on (19) (a) optimizing lipophilic-hydrophilic properties of the polymer-drug system, (b) optimizing rates of biodegradation in the precorneal pocket, and (c) increasing retention
eciency in the precorneal pocket. It is highly desirable to formulate the
particles with bioadhesive materials in order to enhance the retention time
of the particles in the ocular cul-de-sac. Without bioadhesion, nanoparticles could be eliminated as quickly as aqueous solutions from the precorneal site. Bioadhesive systems can be either polymeric solutions (20) or
particulate systems (21). With several pilot studies using natural bioadhesive polymers demonstrating promising improvements in ocular bioavailability, synthetic biodegradable and bioadhesive polyalkylcyanoacrylate
systems were developed, and these may prove to be the most
promising particulate ocular drug delivery systems of the future.
Polyalkylcyanoacrylates gained popularity because of their apparent lack
of toxicity, proven by decades of safe and successful use in surgery (22),
which from a toxicological point of view is a very favorable characteristic
for a preferred pharmaceutical drug delivery system.

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

II.

PARTICULATE SYSTEMS IN OCULAR DRUG DELIVERY

A.

Topical Systems

441

The isolation of the vitreous caused by the blood-retinal and blood-aqueous

barriers creates diculties for eective drug therapy in all eye diseases (23).
Systemic administration is often not feasible because only a small percentage of drugs can penetrate the ophthalmic barriers and their pharmacokinetics is complicated by fast-owing blood supply in the posterior segment
of the eye (e.g., ocular pharmacokinetics of chloramphenicol and tetracycline) (17). This necessitates large systemic doses to achieve therapeutic
concentrations in ocular tissues, which brings forth the problems of
unwanted side eects and potential toxicity from the drug. Therefore,
regardless of the dosage form and constituents thereof, the most common
route of drug delivery to the eye is topical. Drug-loaded microparticulate
systems are suspended in aqueous or nonaqueous medium and instilled
typically in the cul-de-sac of the eye, wherefrom drug is slowly released in
the lacrimal pool by dissolution and mixing, diusion, or mechanical disintegration and erosion of the polymer matrix (15). Microparticles for topical
administration are of several types, including polymer-drug complex systems, erodible microspheres, responsive particulates, in situ gelling systems,
ion-exchange systems, and nanoparticles (15).
B.

Local Injectable Systems

When the vitreous cavity is targeted for drug delivery, topical, systemic, or
subconjunctival drug delivery prove unsatisfactory due to their inability to
target the action site and maintain sucient, constant, and prolonged therapeutic levels of the drug (24,25). The treatment of some diseases like proliferative vitreoretinopathy, endophthalmitis, and recurrent uveitis requires
repeated injection of drugs in the vitreous cavity to maintain therapeutic
levels (2428). Such multiple injections may result in clinical complications
such as lens damage or retinal injury and may cause deleterious infections or
bleeding in the eye, not to mention the patient discomfort and noncompliance issues that may lead to failure of therapy. Development of biodegradable particulate dosage forms for local injections may circumvent limitations
of frequent intravitreal injections by providing a slow-releasing depot of
drug in the vitreous cavity and reducing the frequency of injections, thereby
increasing patient compliance. Microspheres made of poly(lactide/glycolide)
and loaded with 5-uorouracil (5-FU) were used for in vitro and in vivo
intravitreal kinetic studies (24). 5-FU was released from the microspheres
for at least 2 and up to 7 days, and microspheres were completely eroded in
about 7 weeks with no adverse eects on ocular tissues. The drug was

Copyright 2003 Marcel Dekker, Inc.

442

Kothuri et al.

released from the polymer matrix through simultaneous polymer hydrolysis

and drug diusion.

III.

METHODS OF PREPARATION OF NANOPARTICLES

A.

Emulsion Polymerization in a Continuous Aqueous

Phase

In this process monomers are dissolved in the aqueous phase and within
emulsier micelles. Additional monomers may be present as monomer droplets stabilized by emulsier molecules. Initiation of polymerization takes
place in the aqueous phase when the dissolved monomer molecules are hit
by a starter molecule or by high-energy radiation (2933). Polymerization
and chain growth is maintained by further monomer molecules, which originate from the aqueous phase, the emulsier micelles, or the monomer
droplets. The monomer droplets and the emulsier micelles therefore act
mainly as reservoirs for the monomers or for the emulsier, which later
stabilize the polymer particles after phase separation and prevent coagulation. Also, to prevent excessively rapid polymerization and promote the
formation of nanoparticles, emulsion polymerization is carried out at an
acidic pH (12) (34). The drugs may be added before, during, or after
polymerization and formation of particles. It has been demonstrated that
pilocarpine adsorbed onto blank nanoparticles induced longer miosis compared to drug incorporated in the particles (35). On the other hand, addition
before or during polymerization may lead to more drug being incorporated
into the nanoparticles.

B.

Emulsion Polymerization in a Continuous Organic

Phase

In this process, the conditions for the dierent phases are reversed compared
to aqueous phase emulsion polymerization, and very water-soluble monomers are used. Because of the high amounts of organic solvents and toxic
surfactants required as well as the toxic nature of the monomers, this process has limited utility (36). The drugs are dissolved in a small amount of
water and solubilized by the surfactants in the organic phase. For solubilization of drugs, higher amounts of surfactants are required than for polymerization in an aqueous phase. The monomers are then added either
directly to or dissolved in organic solvents (3739).

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

C.

443

Interfacial Polymerization

Interfacial polymerization of the polyalkylcyanoacrylate polymers allows

the formation of nanocapsules with a shell-like wall (40). This carrier type
can encapsulate drugs with lipophilic character, and the rate of encapsulation is generally related to the solubility of the drug in the oily compartment
(40). The technique involves dissolving the polyalkylcyanoacrylate (PACA)
monomers and lipophilic drug in an ethanolic solution or oil and slowly
injecting this mixture into a well-stirred solution of 0.5% poloxamer 338 in
water at pH 6 (may contain nonionic surfactant) (41). At the oil/water
interface, nanoparticles with a shell-like wall are formed spontaneously by
hydroxyl ioninduced polymerization, and the polymeric colloidal suspension occurs immediately.
D.

Polymerization by Denaturation or Desolvation of

Natural Proteins

Macromolecules such as albumin or gelatin can form nanoparticles through

the desolvation and denaturation processes. Desolvation of macromolecules
in aqueous solution can be induced by changes in pH, charge, or the addition of desolvating agents such as ethanol (31,41). The desolvation process
induces the swollen molecules to coil tightly. At this point, the tightly coiled
macromolecules can be xed and hardened by crosslinking with glutaraldehyde to form nanoparticles rather than nanocapsules. Nanoparticles are
then puried by gel ltration. The denaturation process involves preparing
an emulsion from an aqueous phase containing the drug, magnetite particles, and the macromolecule and oil (cottonseed oil) (4245). Polymerization
is carried out by heat denaturation at temperatures above 120 C or by
chemical crosslinking. Nanoparticles are precipitated out and washed with
ether (or in the case of gelatin, acetone) and stored in the dry form.
E.

Solvent Evaporation Method

Gurny et al. (46) were the rst to use this process for the production of
polylactic acid nanoparticles containing testosterone. In this method, the
polymer of interest is dissolved in an organic solvent, suspended in a suitable
water or oil medium, after which the solvent is extracted from the droplets.
The particles obtained after solvent evaporation are recovered by ltration,
centrifugation, or lyophilization. In general, the diameter of the particles
depends on the size of the microdroplets that are formed in the emulsion
before evaporation of the solvent. Chiang et al. used the solvent evaporation
technique with an oil-in-oil emulsion to prepare polylactide-co-glycolide

Copyright 2003 Marcel Dekker, Inc.

444

Kothuri et al.

microspheres of 5-uorouracil for ocular delivery (47). Microspheres containing cyclosporin A have been prepared with a mixture of 50 : 50 polylactic
and polyglycolic acid polymers using the solvent evaporation process. The
polymer and drug mixture was dissolved in a mixture of chloroform and
acetone, emulsied in an aqueous solution of polyvinyl alcohol, and stirred
for 24 hours to evaporate the organic solvent and yield the microparticle
dispersion (48).
F. Ionic Gelation Technique
De Campos et al. developed chitosan nanoparticles using the ionic gelation
technique (49). Nanoparticles were obtained upon the addition of sodium
tripolyphosphate aqueous solution to an aqueous polymer solution of chitosan under magnetic stirring at room temperature. The formation of nanoparticles was a result of the interaction between the negative groups of the
tripolyphosphate and the positively charged amino groups of chitosan. In
this technique, drug in an acetonitrile-water mixture can be incorporated
either into chitosan solution or the tripolyphosphate solution.
G.

Nanoprecipitation

Fessi et al. (50) developed nanoparticles using this method. In this technique
a polymer and a specied quantity of drug is dissolved in acetonitrile. The
organic phase is then added dropwise to the aqueous phase and stirred
magnetically at room temperature until complete evaporation of the organic
phase takes place. Drug-free nanoparticles may be prepared using the same
procedure by simply omitting the drug.
H.

Spray-Drying

In this technique (51), microparticles are prepared by dissolving the polymer

of interest in an organic solvent. The drug is added to this solution and
spray-dried using a spray-dryer. The process parameters and spray nozzle
size are set up as required. The spray-dried product is collected by a cyclone
separator.

IV.

POLYMERS USED IN THE PREPARATION OF

NANOPARTICLES

A successful nanoparticulate system may be one that has a high loading

capacity, thus reducing the quantity of carrier required for administration.

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

445

The drug can be either adsorbed onto the surface of performed particles or
incorporated into the nanospheres during the polymerization process.
Concerning the loading capacity of nanoparticles, it has been found that
both the nature and quantities of the monomer used inuences the absorption capacity of the carrier. Generally, the longer the chain length, the
higher is the anity of the drug to the polymer, i.e., the capacity of adsorption is related to the hydrophobicity of the polymer and to the specic area
of the polymer (52).
Several types of polymeric nanoparticles are used in ophthalmic drug
delivery
and
prepared
by
the
methods
described
earlier.
Polymethylmethacrylate (PMMA) nanoparticles, which are excellent adjuvants for vaccines, can be produced by the emulsion polymerization technique. In this process, monomeric methylmethacrylate is dissolved in a
concentration range of 0.11.5% in water or phosphatebuered saline or
a solution or suspension of drugs or antigens (33). The polymerization is
carried out by irradiation with -radiation source or by chemically initiated
polymerization using potassium peroxodisulfate and heating to high temperatures. PMMA nanoparticles are generally produced without emulsiers
(12). The biologically active substance, such as drug or antigen, may be
present during polymerization or can be added to previously produced
nanoparticles. Polymerization in the presence of heat-sensitive materials
can only be carried out by -radiation (53). Other polymers produced by
emulsion polymerization in continuous aqueous phase include acrylic copolymer nanoparticles (5456), polystyrene (5759), and polyvinyl pyridine
(54,60). Nanoparticles made of polyacrylamide or PMMA do not degrade
either biologically or enzymatically, which makes them less attractive for
ophthalmic use.
Cellulose acetate phthalate has been used for in situ gelling of latex
nanoparticles (61). The preparation of these latex particles involves emulsication of polymer in organic solvent followed by solvent evaporation. This
latex suspension, upon coming in contact with the lacrimal uid at pH 7.2
7.4, gels in situ, thus averting rapid washout of the instilled solution from
the eye. The disadvantage of these preparations is vision blurring.
PACA particles possess properties of biodegradation and bioadhesion,
making them of considerable interest as possible drug carriers for controlled
ocular drug delivery and drug targeting. Wood et al. (62) showed that
PACA nanoparticles were able to adhere to the corneal and conjunctival
surfaces, which represent their mucoadhesion property. This polymer has
the ability to entangle in the mucin matrix and form a noncovalent or ionic
bond with the mucin layer of the conjunctiva. PACA nanoparticles are
prepared in the same manner as previously described for PMMA. The
monomer at a concentration of 0.13% is added to an aqueous system or

Copyright 2003 Marcel Dekker, Inc.

446

Kothuri et al.

to the drug solution (33). Polymerization starts at room temperature and

does not require -radiation or addition of special chemical initiators. This
is the most signicant advantage over acrylic derivatives, as these particles
do not require high energy input for the polymerization process, and there is
no eect on the stability of the absorbed drug. Alkyl cyanoacrylate particles
polymerize according to an anionic mechanism in an aqueous medium using
hydroxyl ions as the initiator. Starting with the polymeric reaction in an
acidic medium and varying the pH of the medium during the polymerization
process, the velocity of the polymerization and molecular weight of resultant
polymer can be controlled, which in turn inuences the particle size of the
nanoparticles formed by this reaction (33,63). The main disadvantage of
these carriers is that PACA nanoparticles penetrate into the outer layers
of the corneal epithelium, causing a disruption of the cell membranes (64).
As an alternative to PACA nanoparticles, recent studies have shown
that poly--caprolactone (PECL) nanocapsules may serve as superior polymer systems for ocular drug delivery (65,66). Marchal-Heussler et al. (66)
compared nanoparticles prepared by using PACA, PECL, and polylacticco-glycolic acid with betaxolol as model drug. It was shown that the PECL
nanoparticles yielded the highest pharmacological eect. This was believed
to be due to the agglomeration of these nanoparticles in the conjunctival sac.
In previous studies it has been shown that regarding ocular administration, the surface charge of the nanoparticles and binding type of the drug
onto the nanoparticles were much more important parameters than the drug
adsorption percentage onto the nanoparticles (67). It was assumed that coating nanoparticles with positively charged bioadhesive polymers enhances the
interaction between nanoparticles and the negatively charged corneal surface, but there are indications that this assumption may not always hold true
(68). The bioavailability of nanoparticles coated with poly-l-lysine and chitosan (both have positive charge) were compared to that of noncoated nanoparticles. It was suggested that the specic nature of chitosan was responsible
for bioavailability improvement rather than the charge. Using uoresceinlabeled chitosan, it was revealed that chitosan nanoparticles interact favorably with rabbit corneal and conjunctival epithelia and remain associated to
these tissues for over 48 hours (68). By contrast, chitosan solutions were
washed o the eye within a much shorter period of time.

V. BIODEGRADATION AND DRUG RELEASE FROM

PARTICULATE SYSTEMS
In order to prolong the entrance of drugs into the intraocular structures, a
long residence time of the particles in the cul-de-sac and a total desorption

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

447

of the drug from the particles during that time have to be attained. Alkyl
cyanoacrylate polymers degrade relatively rapidly in vivo when compared to
other polymers used in particulate ocular drug delivery such as poly(lactic
acid) and its copolymers with glycolic acid (69). The degradation time is
dependent on the alkyl chain length and ranges from a few hours (when
methyl cyanoacrylate is used) to approximately 3 days (when 80% of isobutyl cyanoacrylate is used) (70). The degradation of cyanoacrylates leads to
the formation of alcohol together with formaldehyde and poly(cyanoacrylic
acid) compounds, which could be toxic in high concentrations (71). Too
rapid a degradation can lead to a burst release of degradation products,
possibly causing cytotoxic eects (70). In order to assess the relevance of the
degradation products regarding possible toxic eects in humans, their rate
of release during particle degradation and the maximum local concentrations must be considered (69).
The rate of degradation of cyanoacrylate particles is dependent on the
alkyl chain length, and the dominating mechanism of particle degradation
was found to be a surface erosion process (72,73). By this mechanism, the
polymeric chain remains intact, but it gradually becomes more and more
hydrophilic until it is water-soluble. Since the biodegradability of polyalkylcyanoacrylate particles depends on the length of the alkyl chain, it is
theoretically possible to choose a monomer whose polymerized form has
the desirable release characteristics (74). However, the release of a drug
cannot always be attributed to polymer degradation alone. Drug desorption
from the polymer surface and diusion of the drug through the polymer
matrix are other mechanisms by which a drug can be released from the
nanoparticles. By using a radiolabeling technique (14 C-labeled nanoparticles
loaded with 3 H-labeled actinomycin), drug release from polyisobutylcyanoacrylate nanoparticles was found to be a direct consequence of polymer
bioerosion (71). The main mechanism of the interaction of PACA particles
with cells in culture was found to be endocytosis, leading to intralysosomal
localization of the carrier (75). As shown in Figure 2, the particles with a low
drug payload and a lower negative surface charge (suspension A, suspension
C) trigger a better response compared to the particles containing a greater
amount of drug adsorbed onto their surface along with a more negatively
charged surface (suspension B) (values are shown in Table 1).

VI.

TRANSPORT PATHWAY OF NANOPARTICLES

Ocular transport of polybutylcyanoacrylate nanoparticles has been studied

by Zimmer et al. (54) using uoroscence microscopy. Nanoparticles were
labeled with rhodamine 6G or propidium iodide as uorescent dyes and

Copyright 2003 Marcel Dekker, Inc.

448

Kothuri et al.

Fig. 2 Adsorption isotherms of betaxolol chlorhydrate. ~, Betaxolol solution; &,

suspension A; ~, suspension B; &, suspension C. (From Ref. 80.)

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

449

Table 1 Physicochemical Characteristicsa of Polyalkylcyanoacrylate

Nanoparticles

Suspension A
without betaxolol chl.
with betaxolol chl.
Suspension B
without betaxolol chl.
with betaxolol chl.
Suspension C
without betaxolol chl.
with betaxolol chl.

Size (nm)
(S.D.)

Zeta-potential
(mV) (S.D.)

Adsorption
percentage
(S.D.)

246 (3.6)
240 (4)

15 (1)
11 (1.5)

30 (3)

276 (2.8)
220 (3.1)

45 (2.5)
30 (2.1)

65 (2)

242 (3)
241 (3.8)

2 (1.5)
1 (1.5)

22 (3)

a
Physicochemical characteristics of particles at 25 C and pH 7.4
S.D. = standard deviation; n 9;
Suspension A = Isobutylcyanoacrylate nanoparticles in acidic aqueous solution
(103 M HCl) with a stabilizer dextran 70000 (1%).
Suspension B = Isobutylcyanoacrylate nanoparticles in acidic aqueous solution
(103 M HCl) with a mixture of dextran 70000 (0.8%) and
dextran sulfate (0.3%)
Suspension C = Isobutylcyanoacrylate nanoparticles in acidic aqueous solution
(103 M HCl) with a mixture of dextran 70000 (0.5%) and Nacetylglucosamine (0.5%).
Source: Ref. 80.

incubated with freshly excised rabbit cornea and conjunctiva for about 30
minutes in standard perfusion cells. After incubation, it was found that there
is a uorescence signal inside the cells, which indicates uptake of nanoparticles by the cornea and conjunctiva. Fluorescent particles were visually
observed inside the cells in what appeared to be vesicles or granules.
Thus, either endocytosis of the nanoparticles by conjunctival tissue or
lysis of the cell membrane by the nanoparticle metabolic degradation products explained the results of their experiments. The authors also found that
the penetration was limited to the rst two cell layers of the cornea, and
further penetration did not occur.

VII.

OCULAR DISTRIBUTION OF NANOPARTICLES

The fate of nanoparticles in the body depends on the physicochemical properties of the nanoparticles (76). Properties such as pH, surfactant, and stabilizers inuence the mucoadhesion properties to the ocular membrane and

Copyright 2003 Marcel Dekker, Inc.

450

Kothuri et al.

thereby modify the precorneal retention of the nanoparticles. It has been

shown that the molecular weight of the polymer inuences the residence
time of nanoparticles in the precorneal area (Fig. 3) (77). As the molecular
weight increases, the polymer becomes poorly retained, whereas low molecular weight polymers are retained for a longer time. Experimental data on
the disposition of polyhexylcyanoacrylate nanoparticles in tears, the aqueous humor, cornea, and conjunctiva of albino rabbits clearly showed adhesion of nanoparticles to absorbing tissue (74). Polyalkylcyanoacrylate
colloidal carriers were eliminated from the tears with a half-life of about
1520 minutes (78). This is signicantly slower than the elimination rate of
aqueous drops, which show a half-life of 13 minutes. With pilocarpine, it
was found that PACA nanoparticles were able to prolong the intraocular
pressure-reducing eect of pilocarpine in rabbits for more than 9 hours (79).
Similar results were obtained with betaxolol as a model drug (80). Thus,
either prolonged drug release or increased contact time between the corneal
tissue and drug could improve the bioavailability and the therapeutic eciency.
Wood et al. (62) studied the disposition of 14 C-labeled polyhexylcyanoacrylate nanoparticles using radiotracer techniques. The concentration of
nanoparticles in the cornea, conjunctiva, and aqueous humor was found to
be three to ve times higher in rabbit eyes in which chronic inammation
had been induced. This is an important observation that suggests that cya-

Fig. 3. Precorneal drainage prole of poly(isobutylcyanoacrylate) nanoparticles:

M.W. (*) 4,275; (&) 13,178; (~) 72,030; () 128,865; (&) 607,439 in g/mol.
(From Ref. 77.)

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

451

noacrylates may have an enhanced bioadhesiveness to inamed tissues. In

addition, the ratio of nanoparticles between inated and normal tissue was
higher in the conjunctiva than in the cornea. This is especially favorable
since various anti-inammatory drugs are used in conjunctival inammation, but these have some side eects after diusion through the cornea into
the aqueous humor. Marchal-Heussler et al. (80) showed that the surface
charge and hydrophobicity of the drug play an important role in drug
adsorption onto the particles. It has been observed that polyethylene glycolcoated (PEG) nanospheres made of polyethyl-2-cyanoacrylate (PECA)
particles loaded with acyclovir showed signicant improvement in bioavailability (81). The improved bioavailability is due to better interaction of the
PEG-coated PECA nanospheres with the corneal epithelium, thereby
increasing ocular mucoadhesion.

VIII.

DRUGS USED IN PARTICULATE OCULAR DELIVERY

The rst nanoparticulate system (average size, 0:3 m) for pilocarpine was
introduced by Gurny and employed cellulose acetate hydrogen phthalate
(CAP) pseudolatex as the polymer (82). The formulation increased the miosis time (up to 10 h) as well as AUC of the drug by 50% compared to the
drug solution by decreasing the elimination rate of the drug. This was
possible by the dissolution of the polymer at pH 7.2 (pH of tears) forming
a viscous polymer solution when the formulation (pH 4.5) was administered
in the eye (83). It has been shown that nanodispersions made of anionic
lattices with low viscosity and containing large amount of polymeric material exhibit an important increase in viscosity when neutralized with a base
(84). Wesslau (85) described this eect as an inner thickening that is due to
the swelling of the nanoparticles from the neutralization of the acid groups
contained on the polymer chain and the absorption of water.
A polybutylcyanoacrylate nanoparticle delivery system for pilocarpine
nitrate has been evaluated in comparison to the solution of the drug for
pharmacokinetic and pharmacodynamic aspects (86). Emulsion polymerization technique was employed in preparing nanoparticles, and in vivo experiments were performed by application of the formulations to the eyes of New
Zealand white rabbits pretreated with betamethasone to create an elevated
intraocular pressure mimicking glaucoma conditions. The results indicated
an increase of 23% in pilocarpine levels in aqueous humor and prolonged
t1=2 for the polybutylcyanoacrylate nanoparticle preparation compared to
the aqueous control solution. It was possible to prolong the miosis with
nanoparticles with lower drug content compared to the control solution.
Betaxolol (80) and amikacin sulfate (87) loaded polyalkylcyanoacrylate

Copyright 2003 Marcel Dekker, Inc.

452

Kothuri et al.

nanoparticles have shown similar eects. The supercial charge and binding
type of the drug onto the nanoparticles are important factors playing a role
in the improvement of the therapeutic response. In another study, adsorption of pilocarpine onto polybutylcyanoacrylate nanoparticles enhanced the
miotic response by about 22% compared to the control aqueous drug solution (35).
Diepold et al. (79) incorporated pilocarpine into polybutylcyanoacrylate nanoparticles and evaluated the aqueous humor drug levels and the
intraocular pressure-lowering eects using three models (the water-loading
model, the alpha-chymotrypsin model, and the betamethasone model) in
rabbits. The miotic response was enhanced by about 33% while the miotic
time increased from 180 to 240 minutes for nanoparticles compared to the
control solution. Also, the intraocular pressure-lowering eects were prolonged to more than 9 hours in all three models mentioned. Vidmar et al.
(88) showed that poly(lactic acid) microcapsules of pilocarpine hydrochloride prepared by a solvent precipitation method prolonged miosis about 4
hours in comparison to control solution (< 2 hours) in rabbits (88). A
signicant improvement in the bioavailability of pilocarpine was attained
by co-administering the pilocarpine-loaded albumin nanoparticles with the
viscous bioadhesive polymer mucin (89).
In a clinical study with Piloplex1 (latex emulsion of pilocarpine
hydrochloride) a lower level of the drug with less uctuation compared to
the corresponding control solution was observed on the third day of treatment. This study involving nine subjects showed a reduction by 5.25 mmHg
of the average diurnal intraocular pressure value compared to the control.
Only one out of 30 patients complained of a local sensitivity reaction with
Piloplex in the yearlong study (90). Similar results were obtained in yet
another study involving 50 patients, where 67.6% of the eyes treated with
the formulation were under control, while only 45.2% were under control
with the pilocarpine solution (91).
Nanocapsules for topical ocular delivery of cyclosporin A (CyA) comprising an oily core (Miglyol 840) and a poly--caprolactone coating
increased the corneal levels of the drug by 5 times compared to the oily
solution of the drug when administered to the cul-de-sac of fully awake New
Zealand white rabbits (92). Also, the drug levels remained higher for up to 3
days with the nanocapsule preparation. More than 90% of CyA could be
encapsulated giving a maximum loading capacity of 50% (drug: polymer)
(92). The enhancement in the drug levels occurred due to the increased
uptake of nanocapsules by the corneal epithelial cells (93). Poly(acrylic
acid) gel of nanocapsules containing 1% CyA showed a better percent
absorption (7:92  2:55%) compared to 1% CyA solution in olive oil
(5:81  2:04% after 24-hour contact time on bovine cornea ex vivo owing

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

453

to the bioadhesive behavior of poly(acrylic acid) polymer and the encapsulated form. The nanocapsules incorporated in the gel were prepared by
interfacial polymer deposition method, with isobutyl-2-cyanoacrylate and
Miglyol 812 forming the polymer coating and the oily core, respectively. The
nanocapsule gel presents a potential ocular drug delivery system with higher
absorption rate and lower risk of toxicity to the cornea (94). In a study by
De Campos et al. (49), chitosan nanoparticles developed for intraocular
delivery of CyA (73% association eciency and 9% loading) by ionic gelation technique showed improved levels of the lipophilic drug in the cornea
and conjunctiva on topical administration to rabbit eyes compared to an
aqueous CyA suspension. The studies conrmed that CyA was preferentially accumulated on the external tissues from chitosan nanoparticles while
sparing the intraocular structures (49).
Polyalkylcyanoacrylate nanoparticles could also be used for antiinammatory drugs to target inamed ocular tissue as they have four
times more anity towards inamed tissue compared to healthy tissue
(95). Both indomethacin-loaded nanoparticles and nanocapsules performed
better in terms of bioavailability and drug levels in the cornea compared to
the commercial solution of the drug Indocollyre1 when administered to
rabbit eye (96). The nanocapsules were prepared by interfacial polymerization using PECL, lecithin, Miglyol 840 as oil, acetone, and poloxamer 188,
while the nanoparticles were prepared omitting the oil and lecithin using a
nano-precipitation method, and nanoemulsions were prepared by using
spontaneous emulsication technique. The authors suggest that the colloidal
particles are taken up by the corneal epithelium through an endocytic
mechanism, and additionally, the colloidal nature of these formulations
(nanocapsules and nanoparticles) aids in increasing the bioavailability
(Fig. 4). These formulations provide a potential for treating intraocular
inammatory diseases with reduced doses of indomethacin. In another
study, chitosan and poly-l-lysine (PLL)coated PECL nanocapsules
increased indomethacin levels in the cornea and aqueous humor of rabbits
four times and eight times, respectively, compared to Indocollyre, the commercial eye drops (67). The positively charged PLL and chitosan coatings
were employed in an attempt to increase interaction of the particles with the
negatively charged corneal epithelium. However, it was found that it was the
specic nature of CS and not the positive charge that was responsible for the
enhanced uptake. PLL coating failed to enhance the uptake of the drug
compared to the corresponding uncoated PECL nanocapsules.
Poly--caprolactone nanocapsules also showed good performance in
increasing the ocular availability of drugs such as metipranolol (65) and
betaxolol (66) while suppressing their systemic absorption. The PECL nanocapsules of metipranolol were prepared by interfacial polymerization tech-

Copyright 2003 Marcel Dekker, Inc.

454

Kothuri et al.

nique incorporating the drug in a Miglyol 840 oily core. When administered
to rabbits, the nanocapsules decreased the intraocular pressure similar to the
commercial ophthalmic solution of the drug, but the systemic side eects,
studied by evaluation of the cardiovascular eects, were signicantly suppressed with the nanocapsules. The heart rate reached normal values within
an hour of administration of nanocapsules versus the commercial eye drops,
which showed pronounced bradycardia for more than 2 hours (97).
Acyclovir-loaded PEG-coated polyethyl-2-cyanoacrylate (PECA)
nanospheres prepared by emulsion polymerization technique showed
increased drug levels in the aqueous humor compared to the free drug
suspension in the rabbits (83). Polylactide and polylactide-co-glycolide biopolymers in the molecular weight range of 3000109,000 have been
employed in the preparation of microparticulate systems for intravitreal
administration of acyclovir (98). Spray-drying technique was employed for
the preparation and the in vivo evaluation was performed by intravitreal
administration in rabbits. The poly-d, l-lactide microspheres of acyclovir
were more ecient compared to the free drug in providing a sustained
release of the drug in the vitreous humor in rabbits. Not only is the initial
drug concentration in vitreous humor attained with microspheres higher

Fig. 4 Permeation of indomethacin through isolated rabbit cornea: (~) PECL

nanoparticles; (*) PECL nanocapsules; (&) submicron emulsion; and (*) commercial eye drops (Indocollyre). (From Ref. 96.)

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

455

compared to the free drug, but the drug levels with the former remained
constant for 14 days, unlike the free drug. The drug concentration diminished to undetectable levels with free drug after 3 days.
PECL nanoparticles and nanocapsules (with a TiO5 oily core) have
been studied for the glaucoma drug carteolol (99). Both formulations
demonstrated a pronounced decrease in the intraocular pressure compared
to the commercial aqueous solution, Carteol1, in rabbits with induced
intraocular hypertension. The PECL carriers increase the residence time
of the drug, enhance the corneal uptake of the drug in unionized form,
and decrease the systemic side eects. Moreover, the studies show that
PECL nanocapsules demonstrate a better eect compared to the PECL
nanoparticles for carteolol. It was concluded that the drug entrapped in
the oily core is more available for corneal absorption.
3
H-Labeled hydrocortisone-17-butyrate-21-propionate (3 H-HBP)
loaded lipid microspheres have been shown to produce a signicant increase
in the drug levels in the cornea of rabbits compared to the control 3 H-HBP
suspension after 1 and 3 hours of administration (100). Kimura et al. (101)
prepared 75:25 lactide/glycolide microspheres for an antifungal agent, uconazole, used for the treatment of endophthalmitis. Microspheres of 110
m in diameter have been prepared with hyaluronate esters incorporating
methylprednisolone as a model drug and by using the spray-drying technique (102,103). The carboxyl group of methylprednisolone was esteried for
50% of the drug content, while the remaining half was present as the sodium
salt. From in vivo studies it was found that the drug could be delivered for
long time with a lower drug peak concentration, diminished side eects, and
enhanced bioavailability compared to the control suspension.
Pilocarpine-loaded albumin or gelatin microspheres (104) and acyclovir-loaded chitosan microspheres (105) have also been studied. Both microparticulate systems were found to be superior to conventional dosage forms
in terms of in vivo performance. Pilocarpine-loaded albumin or gelatin
microspheres with an average particle size of 30 m were prepared by emulsication of the aqueous solution of pilocarpine nitrate together with the
macromolecule in sunower oil and subsequent crosslinking with formaldehyde (for gelatin microspheres) or heating to 150 C (for albumin microspheres). Further, the oil was removed by washing with ether. The AUC
of the miosis versus time curve was increased by 2.3 and 3.3 times for the
gelatin and albumin microspheres, respectively, compared to the aqueous
solution of the drug (104).
Binding betaxolol to ion-resin exchange resin microparticles
(Betoptic1 S) increases the drug bioavailability by reducing the drug
release in the tear. Betoptic S 0.25% and Betoptic solution 0.5% are
considered bioequivalent. Since introduction in the market, Betoptic S

Copyright 2003 Marcel Dekker, Inc.

456

Kothuri et al.

has increased patient compliance by reducing the dose and frequency of

dosing (106). Marchal-Heussler et al. (66) studied betaxolol-loaded nanospheres or nanocapsules using three dierent polymers: polyisobutylcyanoacrylate, a copolymer of lactic and glycolic acid, and PECL. The
intraocular pressure-lowering eect was most pronounced for the PECL
nanocapsules or nanospheres compared to other two polymers as well as
the commercial eye drops, owing to the higher hydrophobicity exhibited
by PECL. The hydrophobic character of the polymer allows agglomeration of the nanoparticles in the eye and subsequent prolongation of the
drug residence time in the precorneal area. Further, PECL nanocapsules
performed better compared to the corresponding nanospheres in that the
drug entrapped in the unionized form in the oily core could penetrate
better into the cornea (66).
5-Fluorouracil and adriamycin loaded microspheres or poly(lactic
acid) and copolymer of lactic/glycolic acid have been prepared using the
solvent evaporation method (107,108). The formulations for both the antiproliferative drugs excelled in their in vivo intravitreal kinetics in rabbits
compared to their corresponding controls. A 10 g injection of adriamycin
solution led to severe toxic reaction in the retina, while the same amount of
drug incorporated in the microspheres substantially decreased the rate of
retinal detachment after 4 weeks with no detectable toxic eects. The drug
release from 5-FU loaded microspheres extended up to 7 days with no
subsequent adverse eects to the ocular tissue.
As shown in Fig. 5, piroxicam loaded in the pectin microspheres (M1,
M2) showed faster in vitro dissolution rates compared to the solid micronized drug (109). The precorneal retention of uorescein-loaded piroxicam
microspheres was evaluated in vivo in albino rabbits, and it was observed
that an aqueous dispersion of microspheres showed a signicantly increased
residence time in the eye (2.5 vs. 0.5 h) when compared to a control uorescein solution. This study also showed signicantly improved bioavailability of piroxicam from microspheres in aqueous humor when compared to
the commercial piroxicam eyedrops.
In a study involving pilocarpine (110), polyisobutylcyanoacrylate
nanocapsules containing 1% pilocarpine were dispersed in an aqueous medium (I) and compared to the same nanocapsule formulation incorporated
into a Pluronic1 F127 gel delivery system (II) and 1% pilocarpine incorporated into a Pluronic F127 gel containing 5% methyl cellulose (III), by
measuring the miotic response in the albino rabbit eye. As shown in
Figure 6, polyisobutylcyanoacrylate nanocapsules of pilocarpine dispersed
in the Pluronic F127 gel (II) showed extended release of pilocarpine compared to formulations (I) and (III) with respect to the length of miotic
response time. As shown in Table 2, statistical analysis indicated a rank

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

457

Fig. 5 In vitro release proles of piroxicam microspheres M1 and M2 compared to

the dissolution prole of micronized piroxicam powder. (From Ref. 109.)

Fig. 6 Miotic response to various formulations of 1% pilocarpine in the albino

rabbit eye. (From Ref. 110.)

Copyright 2003 Marcel Dekker, Inc.

458

Table 2

Mitotic Response to Various Pilocarpine Formulations in Albino Rabbit Eyea

Time to peak
response
(Tp , min)

Duration of
mitotic
response (min)

Intensity of
response
(AUC,b mm.
min)

Absolute

% change

60 L, dispersion of
PILO-loaded PIBCA-NC
N 4)
III
60 L, Pluronic1
F127 gel (N 6),
5% MC
II
60 L Pluronic1
F127 gel (N 6),
5% MC
containing-dispersion of
PILO-loaded PIBCA-NC

14.13
(0.24)

150.75
(3.11)

167.00
(6.52)

2.13
(0.13)

35.42
(2.08)

11.58
(1.25)

217.08
(4.61)

374.27
(11.26)

3.00
(0.13)

17.63
(0.55)

259.88
(2.30)

409.16
(3.59)

2.50
(0.00)

41.67
(0.00)

Overall probability (p)

0.0042

<0.0005

<0.0005

<0.0007

<0.0007

Treatment (L) 1% (w/v)

pilocarpine (PILO)

Peak change in the pupil

diameter (Imax , mm)

50.00
(2.15)

Values shown represent means and SEM (in parentheses).

Area under the temporal mitotic response intensity curve.
c
Vertical lines within a column join data pairs that are not signicantly different. All other values within a given column differ
signicantly (p < 0:05).
Source: Ref. 110.
b

Copyright 2003 Marcel Dekker, Inc.

Kothuri et al.

Microparticles and Nanoparticles in Ocular Drug Delivery

459

order for both the duration and intensity of miosis as II > III  I, with all
dierences being signicant.

IX.

CONCLUSIONS

Particulate systems have the potential to become promising systems for

ophthalmic drug delivery. To date, only one microparticulate ophthalmic
prescription product, Betoptic1 S, has been approved for marketing in the
United States. Betoptic S 0.25% is considered to be bioequivalent to
Betoptic 0.5% solution. By binding betaxolol to ion exchange resin particles
for the Betoptic S formulation, drug release was retarded in the tear and
drug bioavailability was enhanced (105). The ocular comfort of betaxolol
was also greatly enhanced by reducing the availability of free drug molecules
in the precorneal tear lm. Thus, microparticulate technology introduces the
advantage of superior patient acceptability in combination with extended
drug release and improved patient compliance. Particles are suspended in a
medium that can be administered topically or intraocularly as an injection.
The potential for success of nanoparticles in ophthalmic drug delivery has
been demonstrated in a number of studies of either hydrophilic or hydrophobic drugs. Formulation stability, control of particle size, control of the
rate of drug release, and large-scale manufacture of sterile preparations are
some of the major issues involved in the development of ophthalmic particulate formulations.

REFERENCES
1. Lee, V. H. L., and Robinson, J. R. (1979). Mechanistic and quantitative
evaluation of pilocarpine disposition in albino rabbits. J. Pharm. Sci., 68:673.
2. Chrai, S. S., and Robinson, J. R. (1984). Ocular evaluation of methylcellulose
vehicle in rabbits. J. Pharm. Sci., 68:1218.
3. Worakul, N., and Robinson, J. R. (1997). Ocular pharmacokinetics/pharmacodynamics. Eur. J. Pharm. Biopharm., 44:71.
4. Barber, R. F., and Shek, P. N. (1993). Liposomes as topical ocular drug
delivery system. In: Alain Rolland (ed.), Pharmaceutical Particulate
Carriers. Marcel Dekker, New York, pp. 120.
5. Patton, T. F., and Robinson, J. R. (1975). Ocular evaluation of polyvinyl
alcohol vehicle in rabbits. J. Pharm. Sci., 64:1312.
6. Saettone, M. F., Giannaccini, B., Teneggi, A., Savigni, P., and Tellini, N.
(1982). Vehicle eects on ophthalmic bioavailability: The inuence of dierent
polymers on the activity of pilocarpine in rabbit and man. J. Pharm.
Pharmacol., 34:464.

Copyright 2003 Marcel Dekker, Inc.

460

Kothuri et al.

7. Allen, T. M., McAllister, L., Mausolf, S., and Gyory, E. (1981). Liposomecell interactions. A study of the interactions of liposomes containing
entrapped anti-cancer drugs with the EMT6, S49 and AE1 (transport-decient) cell lines. Biochim. Biophys. Acta, 643:346.
8. Campbell, P. I. (1983). Toxicity of some charged lipids used in liposome
preparations. Cytobiosis, 37:21.
9. Yoshihara, E., and Nakae, T. (1986). Cytolytic activity of liposomes containing stearylamine. Biochim. Biophys. Acta, 854:93.
10. Chang, S. C., Bundgaard, H., Buur, A., and Lee, V. H. L. (1987). Improved
corneal penetration of timolol by prodrugs as a means to reduce systemic drug
load. Invest. Ophthalmol. Vis. Sci., 28:487.
11. Lebourlais, C. A., Treupel Acar, L. Rhodes, C. T., Sado, P. A., and Leverage,
R. (1995). New ophthalmic drug delivery systems. Drug. Dev. Ind. Pharm.,
21:19.
12. Bundgaard, H. (1989). Improved ocular delivery of pilocarpine and timolol
through prodrugs. 5th Int. Conf. Pharm. Tech. Paris, p. 52.
13. Candida, L., Maria, J. A., Jose, L. V., Francisco, O., Martinez, J., Saavedra,
A. S., Jose, C. C. (1992). Reduction of cardiovascular side eects associated
with ocular administration of Metipranolol by inclusion in polymeric nanocapsules. J. Ocul. Pharmacol., 8:191.
14. Joshi, A. (1994). Microparticulates for ophthalmic drug delivery. J. Ocul.
Pharmacol., 10:29.
15. Joshi, A. (1996). Microparticulates as an ocular drug delivery system. In:
Ocular Therapeutics and Drug Delivery, I. K. Reddy (ed.). Technomic, pp.
441459.
16. Rollot, J. M., Couvreur, P., Roblot Treupe, L., Devissaguet, J. P. H., and
Puisieux, F. (1986). Physico-chemical and morphological characterization of
polyisobutylcyanoacryate nanocapsules. J. Pharm. Sci., 75:361.
17. Schulman, J. A., and Peyman, G. A. (1993). Intracameral, intravitreal, and
retinal drug delivery. In: A. K. Mitra (ed.), Ophthalmic Drug Delivery Systems.
Marcel Dekker, New York, pp. 383425.
18. Kreuter, J., Tauber, U., and Illi, V. (1979). Distribution and elimination of
poly methyl-2-14-(methacrylate) nanoparticle radioactivity after injection in
rat and mice. J. Pharm. Sci., 68:1443.
19. Lee, V. H. K., Wood, R. W., Kreuter, J., Harima, T., and Robinson, J. R.
(1986). Ocular drug delivery of progesterone using nanoparticles. J. Microen.,
3:213.
20. Gurny, R., Ibrahim, H., Aebi, A., Buri, P., Wilson, C. G., and Washington,
N. (1987). Design and evaluation of controlled release systems for the eye. J.
Cont. Rel., 6:367.
21. Hui, H. W., and Robinson, J. R. (1985). Ocular delivery of progesterone using
a bioadhesive polymer. Int. J. Pharm., 26:203.
22. Collins, J. A., James, P. M., Levitski, S. A., Brendenburg, C. E., Anderson, R.
W., Leonard, F., and Hardway, R. M. (1969). Clinical use in severe combat

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

23.
24.

25.

26.

27.

28.
29.
30.
31.

32.
33.

34.

35.

36.
37.
38.

39.

461

casualties. Cyanoacrylate adhesive as topical homeostatic aids. Surgery,

65:260.
Cuncha-vaz, J. G. (1997). The blood ocular barriers: past, present and future.
Documenta Ophthalmol., 93:149.
Moritera, T., Ogura, Y., Yoshimura, N., Honda, Y., Wada, R., Hyon, S. H.,
and Ikada, Y. (1992). Biodegradable microspheres containing adriamycin in
the treatment of proliferative vetreonopathy. Invest Ophth. Vis. Sci., 33:3125.
Khoobehi, B., Stradtmann, M. O., Peyman, G. A., and Aly, O. M. (1991).
Clearance of uorescein incorporated into microspheres from the vitreous
after intravitreal injection. Ophthal. Surg., 21:175.
Khoobehi, B., Stradtmann, M. O., Peyman, G. A., and Aly, O. M. (1990).
Clearance of uorescein incorporated into microspheres from the cornea and
aqueous after subjunctival injection. Ophthal. Surg., 21:840.
Moritera, T., Ogura, Y., Yoshimura, N., Honda, Y., Wada, R., Hyon, S. H.,
and Ikada, Y. (1991). Microspheres of biodegradable polymers as a drug
delivery system in the vitreous. Invest. Ophth. Vis. Sci., 32:1785.
Martini, B. (1992). Proliferative vitreo-retinal disorders: Experimental models
in vivo and in vitro. Acta Ophthalmol., 201(Suppl.):1.
Kreuter, J. (1983). Evaluation of nanoparticles as drug delivery systems. I.
Preparation methods. Pharm. Acta. Helv., 58:196.
Fitch, R. M., and Tsai, C. (1970). Polymer colloids: Particle formation in
micellar systems. J. Polymer Sci., Polym. Lett., 8:703.
Fitch, R. M., Prenosil, M. B., and Sprick, K. J. (1993). The mechanism of
particle formation in polymer hydrosols. I. Kinetics of aqueous polymerization of methyl methacrylate. J. Polymer Sci. C, 27:467.
Fitch, R. M. (1993). The homogenous nucleation of polymer colloids. Br.
Poly. J., 5:467.
Kreuter, J. (1992). Nanoparticles-Preparation and applications. In:
Microcapsules and Nanocapsules in Medicine and Pharmacy, M. Don Brow
(ed.). CRC Press Inc., Boca Raton, pp. 126143.
Mezei, M., and Meisner, D. (1993). Liposomes and nanoparticles as ocular
drug delivery systems. In: Biopharmaceutics of Ocular Drug Delivery. CRC
Press, Inc., Boca Raton, pp. 91101.
Harima, T., Kreuter, J., Speiser, P., Boye, T., Gurny, R., and Kubis, A.
(1986). Enhancement of miotic response of rabbits with pilocarpine-loaded
polybutylcyanoacrylate nanoparticles. Int. J. Pharm., 33:187.
Kreuter, J. (1978). Nanoparticles and nanocapsulesnew dosage forms in the
nanometer size range. Pharm. Acta. Helv., 53:33.
El-Samaligy, M. S., Rhodewald, P., and Mahmoud, H. A. (1986).
Polyalkylcyanoacrylate nanocapsules. J. Pharm. Pharmacol., 38:216.
Krause, H. J., Schwarz, A., and Rohdewald, P. (1986). Interfacial polymerization, a useful method for the preparation of polymethyl cyanoacrylate
nanoparticles. Drug. Dev. Ind. Pharm., 12:527.
Gasco, M. R., and Trotta, M. (1986). Nanoparticles from microemulsions.
Int. J. Pharm., 29:267.

Copyright 2003 Marcel Dekker, Inc.

462

Kothuri et al.

40. Al Khouri Fallouh, N., Roblot-Treupel, L., Fess, H., Devissaguet, J. P. H.,
and Puissieux, F. (1986). Development of a new process for the manufacture
of polyisobutylcyanoacrylate nanoparticles. Int. J. Pharm., 28:125.
41. Marty, J. J., Oppenheim, R. C., and Speiser, P. (1978). Nanoparticles a new
colloidal drug delivery systems. Pharm. Acta. Helv., 53:17.
42. Scheel, U., Rhodes, B. A., Natarajan, T. K., Jr., and Wagner, H. N. (1972).
Albumin microspheres for the study of reticuloendothelial system. J. Nucl.
Med., 13:498.
43. Widder, K., Fluoret, G., and Senyei, A. (1979). Magnetic microspheres:
Synthesis of a novel parenteral drug carrier. J. Pharm. Sci., 68:79.
44. Kramer, P. A. (1974). Letter: Albumin microspheres as vehicles for achieving
specicity in drug delivery. J. Pharm. Sci., 63:1646.
45. Gallo, J. M., Hung, C. T., and Perrier, D. G. (1984). Analysis of albumin
microsphere preparation. Int. J. Pharm., 22:63.
46. Gurny, R., Peppas, N. A., Harrington, D. D., and Banker, G. S. (1981).
Development of biodegradable and injectable lattices for the controlled
release of potent drugs. Drug Dev. Ind. Pharm., 7:1.
47. Chiang, C. H., Tung, S. M., Lu, D. W., and Yeh, M. K. (2001). In vitro and in
vivo evaluation of an ocular delivery system of 5-uorouracil microspheres. J.
Ocul. Pharmacol. Ther., 17:545.
48. Harper, C. A. 3rd, Khoobehi, B., Peyman, G. A., Gebhardt, B. M., and
Dunlap, W. A. (199394). Bioavailability of microsphere-entrapped cyclosporine A in the cornea and aqueous of rabbits. Int. Ophthalmol., 17:337.
49. De Campos, A. M., Sanchez, A., and Alonso, M. J. (2001). Chitosan nanoparticles: a new vehicle for the improvement of the delivery of drugs to the
ocular surface. Application to cyclosporin A. Int. J. Pharm., 224:159.
50. Fessi, H., Puisieux, F., Devissaguet, J. P., Ammoury, N., and Benita, S.
(1989). Nanocapsule formation by interfacial polymer deposition following
solvent displacement. Int. J. Pharm., 55:R1.
51. OHara, P., and Hickey, A. J. (2000). Respirable PLGA microspheres containing rifampicin for the treatment of tuberculosis: manufacture and characterization. Pharm. Res., 17:955.
52. Illum, L., Khan, M. A., Mark, E., and Davis, S. S. (1986). Evaluation of
carrier capacity and release characteristics for poly (butyl 2-cyanoacrylate)
nanoparticles. Int. J. Pharm., 30:17.
53. Kreuter, J., and Zehnder, H. J. (1978). The use of Co--radiation for the
production of vaccines. Radiat. Eects, 35:161.
54. Rembaum, A., Yen, S. P. S., Cheong, E., Wallace, S., Molday, R. S., Gordon,
I. L., and Dreyer, W. J. (1976). Functional polymeric microspheres based on
2-hydroxyethyl methacrylate for immunochemical studies. Macromolecules,
9:328.
55. Rembaum, A., Yen, S. P. S., and Molday, R. S. (1979). Synthesis and reaction
of hydrophilic functional microspheres for immunological studies. J.
Macromol. Sci. Chem. A, 13:603.

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

463

56. Rembaum, A. (1980). Synthesis, properties, and biomedical applications of

hydrophilic, functional, polymeric immunomicrospheres. Pure Appl. Chem.,
52:1275.
57. Kreuter, J., Liehl, E., Berg, U., Soliva, M., and Speiser, P. P. (1988). Inuence
of hydrophilicity on the adjuvant eect of particulate polymeric adjuvants.
Vaccine, 6:253.
58. Ugelstad, J., Rembaum, J. T., Kemshead, K., Mustad, S., Fuderud, S., and
Schmid, R. (1984). Preparation and biomedical applications of polymeric
particles in microspheres and drug therapy. In: S. S. Davis, L. Illum, J. G.
Mcvie, E. Tomlinson (eds.). Elsevier, Amsterdam.
59. Shahar, M., Meshulam, H., and Margel, S. (1986). Synthesis and characterization of microspheres of polystyrene derivatives. J. Polym. Sci. Chem. Ed.,
24:203.
60. Schwartz, A., and Rembaum, A. (1985). Poly(vinyl pyridine) microspheres.
In: K. J. Widder, R. Green (eds.). Academic Press, Orlando.
61. Gurny, R. (1983). Latex systems. In: D. D. Breimer and P. Speiser (eds.).
Topics in Pharmaceutical Sciences. Elsevier Science Publishers B. V,
Amsterdam, pp. 277288.
62. Wood, R. W., Li, V. H. K., Kreuter, J., and Robinson, J. R. (1985). Ocular
disposition of poly-hexyl-2-cyano(3-14 C) acrylate nanoparticles in the albino
rabbit. Int. J. Pharm., 23:175.
63. Das, S. K., Tucker, I. G., Hill, D. J., and Ganguly, N. (1995). Evaluation of
poly(isobutylcyanoacrylate) nanoparticles for mucoadhesive ocular drug
delivery. I. Eect of formulation variables on physicochemical characteristics
of nanoparticles. Pharm. Res., 12:534.
64. Zimmer, A., Kreuter, J., and Robinson, J. R. (1991). Studies on the transport
pathway of PBCA nanoparticles in ocular tissues. J. Microen., 8:497.
65. Losa, C., Marchal-Heussler, L., Orallo, F., Vila Jato, J. L., and Alonso, M. J.
(1993). Design of new formulations for topical ocular administration: polymeric nanocapsules containing metipranolol. Pharm. Res., 10:80.
66. Marchal-Haussler, L., Fessi, H., Devissaguet, J. P., Homan, M., and
Maincent, P. (1992). Colloidal drug delivery systems for the eye. A comparison of the ecacy of three dierent polymers: polyisobutylcyanoacrylate,
polylactic-coglycolic acid, poly-epsilon-caprolactone. Pharm. Sci., 2:98.
67. Calvo, P., Vila-Jato, J. L., and Alonso, M. J. (1997). Evaluation of cationic
polymer-coated nanocapsules as ocular drug carriers. Int. J. Pharm., 153:41.
68. Janes, K. A., Calvo, P., and Alonso, M. J. (2001). Polysaccharide colloidal
particles as delivery systems for macromolecules. Adv. Drug Deliv. Rev., 47:83.
69. Grislain, L., Couvreur, P., Lenaerts, V., Roland, M., Deprez-Decampeneere,
D., and Speiser, P. (1983). Pharmacokinetics and distribution of bio-degradable drug carrier. Int. J. Pharm., 15:335.
70. Lherm, C., Muller, R. H., Puisieux, F., and Couvreur, P. (1992). Alkyl cyanoacrylate drug carriers: II. Cytotoxicity of cyanoacrylate nanoparticles with
dierent alkyl chain length. Int. J. Pharm., 84:13.

Copyright 2003 Marcel Dekker, Inc.

464

Kothuri et al.

71. Lenaerts, V., Couvreur, P., Christiaens-Leyh, D., Joiris, E., Ronald, M.,
Rollman, B., and Speiser, P. (1984). Degradation of polyisobutylcyanoacrylate nanoparticles. Biomaterials, 5:65.
72. Leonard, F., Kulkarni, R. K., Brandes, G., Nelson, J., and Mameron, J.
(1966). Synthesis and degradation of poly (alkyl cyanoacrylates). J. Appl.
Polym. Sci., 10:259.
73. Muller, R. H., Lherm, C., Herbort, J., and Couvreur, P. (1990). In vitro model
for the biodegradation of alkyl cyanoacrylate nanoparticles. Biomaterials,
11:590.
74. Vezin, W., and Florence, A. (1978). In vitro degradation rates of biodegradable poly-N-alkylcyanoacrylates. J. Pharm. Pharmacol., 30:5P.
75. Couvreur, P., Tulkens, P., Roland, M., Trouet, A., and Speiser, P. (1977).
Nanocapsules: a new type of lysosomotropic carrier. FEBS Lett., 84:323.
76. Davis, S. S., and Illum, L. (1983). The targeting of drugs using polymeric
microspheres. Br. Polym. J., 15:160.
77. Das, S. K., Tucker, I. G., and Davies, N. M. (1992). A gamma scintigraphic
evaluation of the eect of molecular weight of poly(isobutyl cyanoacrylate)
Nanoparticles on the precorneal residence in rabbit. Proc. Int. Symp. Control.
Rel. Matter. Controlled Release Society, Inc., 19:395.
78. Diepold, R., Kreuter, J., Guggenbuhl, P., and Robinson, J. R. (1989).
Distribution of poly-hexyl-2-cyano[3-14 C]acrylate nanoparticles in healthy
and chronically inamed rabbit eyes. Int. J. Pharm., 54:149.
79. Diepold, R., Kreuter, J., Himber, J., Gurny, R., Lee, V. H. L., Robinson, J.
R., Saettone, M. F., and Schnaudigel, O. E. (1989). Comparison of dierent
models for the testing of pilocarpine eyedrops using conventional eyedrops as
a novel depot formulation (nanoparticles). Graefes Arch. Clin. Exp.
Ophthalmol., 227:188.
80. Marchal-Heussler, L., Maincent, P., Homan, H., Spittler, J., and Couvreur,
P. (1990). Antiglaucomatous activity of betaxolol chlorhydrate sorbed onto
dierent isobutyl cyanoacrylate nanoparticles preparations. Int. J. Pharm.,
58:115.
81. Fresta, M., Fontana, G., Bucolo, C., Cavallaro, G., Giammona, G., and
Puglisi, G. (2001). Ocular tolerability and in vivo bioavailability of poly(ethylene glycol) (PEG)-coated polyethyl-2-cyanoacrylate nanosphere-encapsulated acyclovir. J. Pharm. Sci., 90:288.
82. Gurny, R. Preliminary study of prolonged acting drug delivery system for the
treatment of glaucoma. Pharma. Acta. Helv., 56:130.
83. Gurny, R. Ocular therapy with nanoparticles. In: Polymeric Nanoparticles and
Microspheres, P. Guiot and P. Couvreur (eds.). CRC Press, Boca Raton, pp.
127136.
84. Gurny, R., Ibrahim, H., and Buri, P. (1993). The development and use of in
situ formed gels, triggered by pH. In: Biopharmaceutics of Ocular Drug
Delivery, P. Edman (ed.). CRC Press, Boca Raton, pp. 8289.

Copyright 2003 Marcel Dekker, Inc.

Microparticles and Nanoparticles in Ocular Drug Delivery

465

85. Weslau, H. (1963). Zur Kenntnis von Acrylsaureenthaltenden Copolymerdispersionen. II. Die Verdichbarkeit Acrylsaureenthaltender Dispersionen.
Makromol. Chem., 69:220.
86. Zimmer, A., Mutschler, E., Lambrecht, G., Mayer, D., and Kreuter, J. (1994).
Pharmacokinetic and Pharmacodynamic aspects of an ophthalmic pilocarpine
nanoparticle-delivery-system. Pharm. Res., 11:1435.
87. Losa, C., Calvo, P., Castro, E., Vila-Jato, J. L., and Alonso, M. J. (1991).
Improvement of ocular penetration of amikacin sulphate by association to
poly(butyl cyanoacrylate) nanoparticles. J. Pharm. Pharmacol., 43:548.
88. Vidmar, V., Pepeljnjak, S., and Jals enjak, I. (1985). The in vivo evaluation of
poly(lactic acid) microcapsules of pilocarpine in hydrochloride. J. Microen.,
2:289.
89. Zimmer, A. K., Chetoni, P., Saettone, M. F. Zerbe, H., and Kreuter, J. (1995).
Evaluation of pilocarpine-loaded albumin particles as controlled drug delivery
systems for the eye. II. Co-administration with bioadhesive and viscous polymers. J. Cont. Rel., 33:31.
90. Ticho, U., Blumenthal, M., Zonis, S., Gal, A., Blank, I., and Mazor, Z. W.
(1979). Piloplex, a new long-acting pilocarpine polymer salt. A long-term
study. Br. J. Opthalmol., 63:48.
91. Ticho, U., Blumenthal, M., Zonis, S., Gal, A., Blank, I., and Mazor, Z. W.
(1979). A clinical trial with Piloplexa new long-acting pilocarpine compound: preliminary report. Ann. Ophthalmol., 11:555.
92. Calvo, P., Sanchez, A., Mart nez, J., Lopez, M. I., Calonge, M., Pastor, J. C.,
and Alonso, M. J. (1996). Polyester Nanocapsules as new topical ocular delivery systems for cyclosporin A. Pharm. Res., 13:311.
93. Calvo, P., Thomas, C., Alonso, M. J., Vila Jato, J. L., and Robinson, J. R.
(1994). Study of the mechanism of interaction of poly(-caprolactone) nanocapsules with the cornea by confocal laser scanning microscopy. Int. J.
Pharm., 103:283.
94. Le-Bourlais, C. A., Chevanne, F., Turlin, B., Acar, L., Zia, H., Sado, P. A.,
Needham, T. E., and Leverge, R. (1997). Eect of cyclosporine A formulations on bovine corneal absorption: ex-vivo study. J. Microencapsulation,
14:457.
95. Calvo, P., Alonso, M. J., Vila-Jato, J. L., and Robinson, J. R. (1996).
Improved ocular bioavailability of indomethacin by novel ocular drug carriers. J. Pharm. Pharmacol., 48:1147.
96. Calvo, P., Vila-Jato, J. L., and Alonso, M. J. (1996). Comparative in vitro
evaluation of several colloidal systems, nanoparticles, nanocapsules, and
nanoemulsions, as ocular drug carriers. J. Pharm. Sci., 85:530.
97. Losa, C., Alonso, M. J., Vila, J. L., Orallo, F., Martinez, J., Saavedra, J. A.,
and Pastor J. C. (1992). Reduction of cardiovascular side eects associated
with ocular administration of metipranolol by inclusion in polymeric nanocapsules. J. Ocul. Pharmacol., 8:191.

Copyright 2003 Marcel Dekker, Inc.

466

Kothuri et al.

98. Conti, B., Bucolo, C., Giannavola, C., Puglisi, G., Giunchedi, P., and Conte,
U. (1997). Biodegradable microspheres for the intravitreal administration of
acyclovir: in vitro in vivo evaluation. Eur. J. Pharm. Sci., 5:287.
99. Marchal-Heussler, L., Sirbat, D., Homan, M., and Maincent, P. (1993).
Poly(-caprolactone) nanocapsules in carteolol ophthalmic delivery. Pharm.
Res., 10:386.
100. Komatsu, A., Ohashi, K., Oba, H., Kakehashi, T., Mizushima, Y., Shirasawa,
E., and Horiuchi, M. (1988). Application of lipid microsphere drug delivery
system to steroidal ophthalmic preparation. Jpn. J. Ophthalmol., 32:41.
101. Kimura, H., Ogura, Y., Moritera, T., Honda, Y., Hyon, S. H., and Ikada, Y.
(1992). Biodegradable polymer microspheres for a sustained release of uconazole in the vitreous. Invest. Ophthalmol. Vis. Sci. 33:1013.
102. Joshi, H. N., Stella, V. J., and Topp, E. M. (1992). Drug release from membranes of hyaluronic acid and its esters. J. Control. Rel., 20:109.
103. Kyyronen, K., Hume, L., Urtti, A., Topp, E., and Stella, V. (1992).
Methylprednisolone esters of hyaluronic acid in ophthalmic drug delivery:
In vitro and in vivo release studies. Int. J. Pharm., 80:161.
104. Leucuta, S. E. (1989). The kinetics of in vitro release and the pharmacokinetics of miotic response in rabbits of gelatin and albumin microspheres with
pilocarpine. Int. J. Pharm., 54:71.
105. Genta, I., Conti, B., Perugini, P., Pavanetto, F., Spadaro, A., and Puglisi, G.
(1997). Bioadhesive microspheres for ophthalmic administration of acyclovir.
J. Pharm. Pharmacol., 49:737.
106. Jani, R., Gan, O., Ali, Y., Rodstrom, R., and Hancock, S. (1994). Ion
exchange resins for ophthalmic delivery. J. Ocul. Pharmacol. Ther., 10:57.
107. Ward, C., and Weck, S. (1990) Dilution and storage of recombinant tissue
plasminogen activator (Activase) in balanced salt solutions. Am. J.
Ophthalmol., 109:98.
108. Becker, G. W., Bowsher, R. R., Mackellar, W. C., Poor, M. L., Tackitt, P. M.,
and Riggin, R. M. (1987). Chemical, physical and biological characterization
of a dimeric form of biosynthetic human growth hormone. Biotechnol. App.
Biochem., 9:478.
109. Giunchedi, P., Conte, U., Chetoni, P., and Saettone, M. F. (1999). Pectin
microspheres as ophthalmic carriers for piroxicam: evaluation in vitro and
in vivo in albino rabbits. Eur. J. Pharm. Sci., 9:1.
110. Desai, S. D., and Blanchard, J. (2000). Pluronic1F127-based ocular delivery
system containing biodegradable polyisobutylcyanoacrylate nanocapsules of
pilocarpine. Drug Deliv., 7:201.

Copyright 2003 Marcel Dekker, Inc.

15
Dendrimers: An Innovative and
Enhanced Ocular Drug Delivery
System
Jeannette M. Loutsch, Desiree Ong, and James M. Hill
LSU Eye and Vision Center of Excellence, Louisiana State University
Health Science Center, New Orleans, Louisiana, U.S.A.

I.

INTRODUCTION

Current ocular drug delivery methods include such methods as topical

drops, ointments, and therapeutic lenses, soluble ocular drug inserts, collagen shields, liposomes, and transcorneal iontophoresis (reviewed in Refs.
1, 2, 3, 4, and 5, respectively). Topically applied drugs, such as suspensions,
ointments, and gels, are seldom maintained at the necessary therapeutic
concentration in the target tissues due to rapid tear turnover and drainage,
tear dilution, and/or an impermeable corneal surface epithelium. Topical
drops, currently the most common form of ocular drug delivery, quickly
lose their eectiveness and cannot reach the back of the eye because of their
rapid elimination. Reduced ocular bioavailability of the drug, in turn, could
result in excessive systemic toxicity, causing unwanted side eects such as
uncontrolled growth of nonsusceptible pathogens. Emulsifying the drug in
an ointment or viscous solution such as polyvinyl alcohol or methylcellulose
can decrease the clearance rate over solutions but decrease vision for a time
(1). Other current ocular drug delivery methods are also not devoid of
problems; for example, collagen shields reduce visual acuity due to the
translucent nature of collagen (3,6). Although advancements have been
made in these existing systems, ecient drug delivery to the eye remains
problematic.
467

Copyright 2003 Marcel Dekker, Inc.

468

Loutsch et al.

Dendrimers, synthetic spherical macromolecules named after their

characteristic tree-like or dendritic branching around a central core, are
a new class of polymers. The branching structure makes them a useful vector
capable of ecient drug and gene delivery. Their nanoscopic architecture
could provide a solution to current ocular drug delivery problems.

II.

HISTORICAL BACKGROUND

Many experimental models have suggested that life arose spontaneously;

simple inorganic building blocks or monomers combined to form larger,
organic clusters with new properties. For years, scientists have attempted
to copy the evolutionary process, endeavoring to assemble small, identical
subunits into complex macromolecules that could be manipulated to interact with and even mimic life. In this attempt, the rst major class of polymers was identied in the 1930s. Staudinger linked identical monomers into
a spaghetti-like string called the random-coil polymer. These polymers
yielded products such as Styrofoam, polyethylene plastic, nylon, and
Plexiglas, known as thermoplastics (7,8). The next decade showed progress
in the making of larger molecules that constituted the second class of polymersthe family of cross-linked polymers or thermosets. Flory (911) and
Stockmayer (12,13) successfully created three-dimensional structures by
bridging or cross-linking the random-coil polymers at various sites, forming an inexible molecule. These compounds are insoluble in most liquids
due to their rigidity and therefore are used as a component of epoxies,
urethanes, berglass, resins rubbers, and gels (7,8). The third class of polymers was developed in the 1960s. These randomly branched polymers
change the bulk property of a substance such as the dierences between
low-density polyethylene, which is highly branched, and high-density polyethylene, which is virtually unbranched (8). A change in the properties of
these polymers is accomplished by altering the types and combinations of
the reagents used in the polymerization process.
The problem with the rst three classes of polymers was the diculty
in predicting the precise internal structure of the polymer. In an attempt to
further dene the submolecular structure, a fourth class of polymers was
synthesized. This class was referred to as having a dendritic macromolecular
architecture due to the branching structure, similar to that of trees. Tomalia,
Vogtle, and Newkome independently reported the synthesis of dendrimers.
Buhleier et al. (14) and Newkome et al. (15) developed the rst dendritic
structures, called cascade molecules and arborols, respectively. Tomalia et
al. (16) developed the rst dendrimer, named the StarburstTM polyamidoamine (PAMAM) dendrimer because of its dendritic branches and controlled

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

469

starburst growth. This macromolecule is built on an ammonia core with

extending branches of alternating methyl acrylate and ethylene diamine
molecules (17). The cascade is continued by adding methyl acrylate moieties
onto the reactive ends of the ethylene diamine molecules and then ethylene
diamine moieties onto the methyl acrylate. Each addition creates another
branched layer, referred to as a generation. Each generation causes an
exponential increase in the surface reactive sites that may have functional
implications (7).
The prospects for the use of dendrimers were numerous and exciting.
Unlike the unpredictable length, size, and shape of the previous classes of
polymers, the architecture of these dendritic branched molecules can be
controlled. The critical molecular design parameters of size, shape, surface
chemistry, exibility, and topology can be carefully regulated to create the
complex molecule. In addition, the dendrimer possesses a remarkably celllike construction consisting of a low-density core and modiable internal
and external surfaces, making it a perfect container or scaolding for drugs,
DNA, and protein. Dendrimers show great promise for a variety of uses
such as drug and gene delivery systems, imaging agents, diagnostic kits,
tumor therapy, industrial catalysts, and sensors.

III.

CHEMICAL AND PHYSICAL CHARACTERISTICS

A.

General Chemical Structure and Synthesis

Dendrimers are composed of concentric, geometrically progressive layers

created through radial amplication from a single, central initiator core
molecule containing either three or four reactive sites such as ammonia or
ethylene diamine. Like the nucleus of the biotic cell, the core contains the
basic information of the dendrimer; it denes the nal size, shape, multiplicity, and functionality of the entire structure. Starting from the center,
each layer stores information, which is transferred to the next outer layer or
generation via covalent connections, determining that layers physical properties (17). This is accomplished by choosing reactants with additional reactive sites. Then through a series of protection/deprotection strategies, all the
reactive sites on the core molecules are converted without reacting with the
reactive sites of the additional reactants. As these steps are repeated, new
generations are added to the dendrimer. The radial transfer of information
distinguishes the synthetic dendrimer from the biotic cell that transmits
information in a linear fashion using DNA and RNA.
The dendrimer amplication process can be described by the simplied
synthesis of the basic PAMAM dendrimer (7,17). The rst step adds methanol to the mixture of ammonia (NH3 ), the initiator core (I), and methyl

Copyright 2003 Marcel Dekker, Inc.

470

Loutsch et al.

acrylate (reactant B), causing a molecule of B to substitute for each of the

three hydrogen atoms on I. The second step, ethylene diamine, a primary
amine is added to the end of each of the methyl acrylates, completing the
rst generation of the dendrimer. Each ethylene diamine moiety contains
two N-H bonds at the end of each branch of the structure, generating six
exposed reactivation sites. The second generation of the dendrimer is created by adding a methyl acrylate moiety at each hydrogen and then an
ethylene diamine molecule to the end of the branch, yielding 12 exposed
reactive sites (Fig. 1). Thus, with such successive generation, the number of
reactive sites increases depending on the substrate added to B. The amplication process can continue for about 9 or 10 generations, when the dendrimer is no longer able to form perfect branches due to steric hindrance
(7,17).
Dendrimer molecules can be assembled via two pathways. The original
pathway, the divergent process, involves working outward from the initiator
molecule with the addition of each generation (17). This method of assembly
can create a number of defects within the dendrimer. The defect can be in an
individual dendrimer or between dendrimers within the mixture. The intra-

Fig. 1 Schematic drawing of dendrimer structures. Dendrimers are synthesized by

two pathways, either the convergent or the divergent pathway. The convergent pathway consists of building the branches prior to addition to the core molecule (A). The
divergent pathway consist of building the branches in concentric layers (B). Each
layer is a generation (G).

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

471

dendrimer faults are usually in the branching and are caused by incomplete
reactions, branch-juncture fragmentation, or abnormal development or
sterically induced stoichiometry. The interdendrimer defects usually arise
form the incomplete removal of reagents that may function as a core or
dendrimer fragmentation, which may act as an initiator species that can act
with other dendrimers (Fig. 2). Optimization of the process and synthesis
strategies can eliminate these problems (17). In 1989, Hawker and Frechet
(18) introduced the second pathway, the convergent process, which builds
the branch wedge independently and then adds it to the root molecule (Fig.
1). This method allows for using ideal branches because they can be separated from the defective ones. This method may also allow for the creation
of dendrimers that have branches with dierent functions, such as one
branch that targets the dendrimer to a cell and a second branch that contains a drug (18). Although this process can generate sucient dendrimers
for the laboratory, it is not conducive to large-scale production (18).

Fig. 2 Defects within dendrimers. Interdendrimer faults can arise form incomplete
reactions creating active sites that interact with other dendrimers. The incomplete
branching in the top left dendrimer (intradendrimer) has a reactant B, which can
react with z from a second dendrimer to interlink the dendrimer structures, creating a
defective molecule.

Copyright 2003 Marcel Dekker, Inc.

472

B.

Loutsch et al.

Physical Properties

The physical properties of dendrimers depend on the chemical structure/

steric properties of internal and external functional groups. The shape may
range from an almost perfect sphere to an ellipsoid to a cylinder-like structure composed of intricately branched fans extending out of an elongated
base (19). Changes in the core molecule and/or the number of reactive sites
present on the core can also inuence the shape. Newer dendrimers use
phosphorus- or silicone-containing molecules, hydrocarbons, lysine, or
thiols as the internal cores (7,17). The dendrimer can also be adjusted
based on the selection of reactants. This also can aect the overall size of
the dendrimer. Usually size is dependant on the number of generations
assembled, but it can also be inuenced by the length of the monomers
used and the angles between the monomers (7). The average mass of a
dendrimer varies from about 500 to 1500 daltons, with the diameter of a
spherical dendrimer varying from about 10 A (rst generation) to a maximum of about 130 A. Each polymerization step increases the diameter by
about 10 A (17).
Dendrimers are relatively stable, covalent macromolecules. They exist
independently from each other and from their external environment. As the
dendrimer grows, steric congestion eventually prevents further growth. At
this point, the dendrimer converts from an open structure to a tight spheroid
with open cavities and a dense surface (20). A specic dendrimers stability
depends on the reactivity of the functional groups located on its surface,
which can be modied to achieve the necessary level of stability.

C.

Dendrimer Core (the Guest-in-Box System)

Naylor et al. (21) were the rst to describe the interior dendrimer space and
to demonstrate the large carrying capacity of PAMAM dendrimers.
Generations 4 through 7 were capable of encapsulating three times their
weight in aspirin. Tomalia and Esfand (19) showed that the dendrimer
interior also contains recognition properties specic to the shape and functionality of the guest molecule. The dissolving of water-insoluble molecules,
such as ibuprofen, increased in the presence of the dendrimer structure. The
ability to encapsulate drugs within the cavities of the dendrimers could
result in novel delivery methods.
Jansen et al. (22) reported the synthesis of a dendrimer box that is
based on a chiral shell of protected amino acids on the dendrimer core.
During synthesis of the box, guest molecules and small amounts of solvent become trapped within the cavities; the molecules dissolve as they
become tightly bound to functional groups on the interior wall of the

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

473

box. Compatible guests such as drugs, DNA, or other small molecules

cannot be extracted, and their spontaneous outward diusion is extremely
slow due to the close packing of the shell. Only medium- to high-generation
dendrimers possess dense enough shells to capture and retain guests (22).
The release of guest molecules can be controlled by changing the pH of the
reaction mix (23).

D.

Dendrimer Surface

The dendrimer surface ultimately determines the structures interactions

with its environment. The surface serves to protect the internal functional
groups via steric interactions, which shield these groups from large molecules while retaining their accessibility and reactivity to small molecules.
These interactions also maintain an outside versus inside position with
ionic (polar) chain ends containing a high density of positively charged
amino groups and a hydrophobic (nonpolar) interior (19). Dendrimer surface properties can be subdivided into endo-receptor properties and exoreceptor properties. Endo-receptor properties are the interior dendritic features such as size, chemical composition, exibility, and topology, which are
responsible for the so-called convergent recognition of guest molecules by
the internal dendrimer surface (17). Exo-receptor properties encompass the
exterior features, such as shape, reactivity, stoichiometry, exibility, and
fractal character, which similarly govern the divergent recognition of external associator molecules by the external dendrimer surface (17).
The dendrimer surface determines its solubility through chemical
recognition of dierent external reagents/solvents and aects the reactivity
of the molecule through the number of reactive surface groups (17). The
surface can serve as scaolding for up to thousands of reactive functional
groups such as carbohydrate residues and peptidyl epitopes. Functional
groups on the surface of dendrimers show higher chemical reactivity than
the same groups attached to other polymer molecules (24). This charging
allows for important covalent coupling reactions with DNA and other molecules. Furthermore, biologically active molecules, when complexed with
dendrimers, tend to retain their maximum activity even when surface groups
on the dendrimer are activated for a reaction.
Modications of the dendrimer surface, specically the addition of
functional groups, occur through the addition of either subnanoscopic or
nanoscaled reactants (19). Subnanoscopic reactants can be a variety of small
molecules, while nanoscaled reactants are composed of larger molecules
such as DNA, antibodies, and proteins that can complex with dendrimers.
Scott et al. (25) describe the rapid synthesis of a second-generation dendri-

Copyright 2003 Marcel Dekker, Inc.

474

Loutsch et al.

mer with primary alcohol groups at the periphery. (The use of modied
dendrimers in bioorganic chemistry is reviewed in Ref. 26.)

IV.

TISSUE DISTRIBUTION OF DENDRIMERS

For dendrimers to be a viable delivery system for drugs and genes, as well as
imaging contrast agents, it must be possible to administer them systemically
or orally with reliability. They also must have low cytotoxicity, low immunogenicity, and few adverse side eects.
To determine the potential for dendrimers as drug delivery molecules,
Wiwattanapatapee et al. (27) used various [125 I]-labeled dendrimers in an ex
vivo system, the everted rat intestinal sac model. The distribution of the
dendrimers was monitored through the wall of the gut by measuring the
radioactivity in the serous uid and tissues. Anionic PAMAM dendrimers
(generation 2.5 and 3.5) had faster, concentration-dependent serosal transfer
than cationic PAMAM dendrimers (generation 3 and 4). The anionic
PAMAM generation 5.5 and the cationic PAMAM dendrimers had high
levels of tissue uptake, thereby slowing the transfer rate (27). These results
indicate that these dendrimers have potential as an oral drug delivery system
and require an in vivo study.
Sakthivel et al. (28) and Florence et al. (29) gave Sprague-Dawley rats
an oral dose of a lipidic peptide dendrimer and monitored tissue distribution. The radiolabeled 2.5 nm dendrimer (generation 4) was given by oral
gavage. The rst study revealed that the maximum level of the dendrimer
was found in the small intestine, with decreasing amounts in the large intestines, blood, and other organs. The amount absorbed by the lymphoid tissue
of the intestinal tract was greater than the amount in the nonlymphoid
tissues, based on organ weight at 3 and 24 hours (28). In the second
study, Peyers patches were found to absorb more of the dendrimer than
normal enterocytes in the small intestine, while the opposite was true in the
large intestine, with enterocytes absorbing more (29). The uptake and transport of these lipidic dendrimers was lower than expected based on polystyrene particle data and may indicate that there is an optimum size for
dendrimer uptake in the intestines.
Malik et al. (30) investigated the relationship between dendrimer structure and biocompatibility in vitro. Dendrimers with NH2 termini had a
concentration-dependent hemolysis and red cell morphology after 1-hour
incubation with rat red blood cell suspension. The PAMAM dendrimers
also demonstrated generation-dependent hemolysis. Those with carboxylate
termini had no hemolytic or cytotoxicity in cell culture (30). In vivo analysis
of cationic and anionic PAMAM dendrimers in the Wistar rat revealed that

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

475

third- and fourth-generation cationic dendrimers are cleared faster than

generations 2.5, 3.5, and 5.5 of the anionic dendrimers. In the case of the
anionic dendrimers, the longer clearance times were generation dependent,
with the smaller dendrimers remaining in the blood longer (30). The systemic clearance correlated with the increased accumulation of the radiolabeled dendrimer in the liver. However, there was no dierence between
intravenous and intraperitoneal administration. Dendrimers with low toxicity and carefully tailored surfaces that remain in the circulation with minimal hepatic uptake, especially if they need to reach a target organ, are
needed if they are to be used for oral drug delivery. The eect of dendrimer
degradation products on tissue also needs to be addressed.
Wiener et al. (31) and Konda et al. (32) adapted dendrimers for use as
magnetic resonance (MR) imaging agents. Second- and sixth-generation
dendrimers were linked to a chelator that enhanced conventional MR imaging (31). These dendrimers showed enhanced contrast in the heart, liver,
kidneys, and blood vessels of Fisher 344 rats. The higher molecular weight
dendrimer may aid in the three-dimensional time-of-ight MR angioplasty
due to the prolonged clearance time. A fourth-generation dendrimer was
linked to folic acid and a chelating agent for use in targeting tumors with
folic acid receptor expression (32). Studies in nude mice with ovarian tumors
(positive for folic acid receptor expression) showed enhanced contrast 24
hours following administration of the agent while receptor-negative tumors
did not. These new contrast agents will aid in the imaging of tumors and
normal tissues.

V.

THERAPEUTIC USES OF DENDRIMERS

The use of dendrimers in drug and gene therapy is increasing. Table 1

provides a summary of studies that used dendrimers or other small synthetic
macromolecules. A discussion of dendrimer-related studies is given below.
A.

Dendrimer-Mediated Delivery of Drugs

Studies have shown dendritic drug delivery to tissues to be very promising.

Some of the advantages of dendrimers include their capacity to hold a large
variety of molecules and to protect these molecules from their surroundings
until they are released in a time-dependent, controlled manner. Dendrimers
that are synthesized to mimic micelles (hydrophobic inner layer and hydrophilic outer layer) can create a micro-container to transport drugs. The
hydroscopic drugs trapped within this space are protected from the environment and have a reduced rate of uptake by the liver and prolonged

Copyright 2003 Marcel Dekker, Inc.

Nanoparticle

Drug/Gene/Virus

PAMAM

Epstein-Barr virus plasmid vector

PAMAM
PAMAM
PAMAM (G4)
PAMAM, Linear
polyacrylamide polymers,
dendron and
dendrigraft
PAMAM (G0-4)

Epstein-Barr virus plasmid vector

Folate receptor
Cisplatin
Inuenza viruses

Biotinylated residues

Folate residues or methyltrexate

PAMAM, liposomes

Normal cystic brosis

transmembrane regulator
gene
Cholera toxin
Platelet-derived growth factor
inhibitor
IL-10
Neurotrophic factors and
antimitotic drugs

PAMAM
Poly(d,l-lactide)
nanoparticles
PAMAM (G5)
Poly(lactide-co-glycolide)
microspheres

Copyright 2003 Marcel Dekker, Inc.

Hepatocellular carcinoma
cells
Intratumoral injection
Ovarian tumors
Cancer chemotherapy
Block sialic acid receptor
and prevent viral
attachment
Pretargeting of cancer cells
to allow the increase of
radioactivity
Target tumor cells
Inhibit viral challenge in
vaginal model of HSV
Replacement gene therapy

Inhibition of adherence
Prevent restinosis
Cardiac graft rejection
Blood-brain barrier

Model

Ref.

In vitro, cell culture

49

In
In
In
In

69
32
40
35

vivo, SCID mice

vivo, mice
vivo, mice
vitro

In vivo, mouse

70

In vivo, mice

71
36

In vitro

61

In vitro
In vitro and in
vivo, rat
In vivo, mice
In vitro and in vivo

72,73
74
53
75

Loutsch et al.

PAMAM (G2)
PAMAM

Application/Target

476

Table 1 Use of Nanoparticles, Including Dendrimers, Liposomes, and Micelles, for Drug and Gene Delivery In Vivo and In
Vitro

Epidermal growth factor

Cationic dendrimer

Folate-conjugated

PAMAM

Magnetic resonance imaging

contrast agent
Monoclonal antibody
immunoconjugate
Antisense oligonucleotide

PAMAM (G2, G4)

PAMAM
Polyethyleneimine
n-(2-Hydroxypropyl)
methacrylamide
PAMAM
Dendritic unimolecular
micelles

Gene and oligonucleotide

transfer
Doxorubicin hydrochloride
Poly(ethylene glycol) grafts
and anticancer drugs
Indomethacin

Target tumor cells for neutron

capture therapy
Folate receptormediated gene
therapy
Tumor imaging

In vitro and in
vivo, rats
In vitro

52
50

In vivo

31

Neutron capture therapy

In vivo, mice

51

Gene expression, rescue

defective genes
Target brain tissue

In vitro

46
76

Chemotherapeutic agent

In vitro and in
vivo, mice
In vivo, humans

42

Chemotherapeutic agents

In vitro

41

Anti-inammatory

In vitro

33

Dendrimers

PAMAM (G4)

477

Copyright 2003 Marcel Dekker, Inc.

478

Loutsch et al.

systemic half-life (33). The anti-inammatory drug indomethacin could be

incorporated at a rate of nine to ten molecules per micelle, with sustained in
vitro release (33).
1. Antibacterials
Dendrimers could assist in counteracting bacterial infections by serving as
transport facilitators for antibiotics and antisense nucleotides or by binding
to receptors on the bacterial cell, thereby preventing infection. The antisense approach is based on the ability of oligonucleotides to interfere with
the biological activity of bacteria by interacting with a necessary protein or
hybridizing to the bacterial mRNA. The latter aptameric interaction blocks
the synthesis of important bacterial proteins.
Oligonucleotides function at the bacterial ribosome or nucleus and can
enter eukaryotic cells via endocytosis. However, the hydrophobic exterior of
prokaryotes serves as a barrier to oligonucleotides. Attia et al. (34) reported
the use of a small fourth-generation dendrimer combined with ethambutol
(a metabolic cell wall inhibitor), which acted as a transport facilitator for
oligonucleotides to interact with Mycobacterium tuberculos in vitro. Further
investigation is needed to determine the eect of the oligonucleotides on
treatment of M. tuberculosis.
2. Antiviral Agents
Dendrimers can be used to counter viral activity in several dierent ways.
They can be modied with external residues that inhibit virus-host cell
interactions (35), act as antivirals themselves (36), or serve as delivery vehicles.
Inuenza viruses infect human cells by binding the viral hemagglutinin
to the sialic acid receptor on the host cell and then undergoing receptormediated endocytosis of the virus into the cell. The inuenza virus can
modify the antigenicity of the hemagglutinin and therefore can evade neutralizing antibodies as well as antiviral compounds (35). Synthetic dendrimers containing sialic acid moieties on the surface layer can successfully
compete with hemagglutinin for the host cell. This interaction can prevent
adhesion to the host cell and subsequent infection. These dendrimers are
polyvalent inhibitors, meaning that one dendrimer can interact with many
hemagglutinin molecules on one virus at one time. This cooperative interaction of individual receptors increases the overall binding anity. Reuter et
al. (35) determined that binding of the inuenza virus hemagglutinin and
neuraminidase glycoproteins with sialic acid residues attached to a dendrimer resulted in a dose-dependent decrease in inuenza infection in vitro.

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

479

These dendrimers, which were nontoxic to cells, were the rst multivalent
sialic acid inhibitor shown to be eective in vitro.
Bourne et al. (36) investigated the potential of ve dierent polylysine
PAMAM dendrimers (containing a central benhydrylamine core) as topical
microbicides for herpes simplex virus (HSV) types 1 and 2. In vitro studies
showed that the dendrimer BRI-2999 could block HSV infection in cytopathic eect (CPE) inhibition assays. In vivo studies in the mouse vaginal
model of HSV infection determined that the application of a topical solution
of BRI-2999 before viral challenge could signicantly reduce the occurrence
of disease, with no noticeable signs of toxicity to the mice. The dendrimer
prevents normal virus-host cell adhesion/adsorption by complexing with the
virus. No eect was seen when the dendrimer was applied after the viral
challenge. Further evaluation is needed to determine if these compounds can
be used to reduce the spread of sexually transmitted diseases. Additionally,
the use of dendrimers has produced promising in vitro inhibitory activity
against respiratory syncytial virus. BRI-2999 had potency in the CPE inhibition assay but also had some cytotoxic eects on stationary phase cells
(37). In HIV studies, polyanionic dendrimers showed inhibition of HIV
replication in MT4 cells (38). The interference came during two phases of
the viral life cycle: adsorption and reverse transcription. In vitro ecacy was
also seen against cytomegalovirus, HSV-1 and HSV-2, thymidine kinase
decient HSV-1, vesicular stomatitis virus, yellow fever virus, reovirus type
1, and dengue fever virus (38).
3.

Antitumor Drug Therapy

Duncan and Malik (39) use doxorubicin and cisplatin as model drugs to
study the potential of dendrimer-anticancer therapeutic agents. They found
that these compounds retained biological activity in vitro and may be useful
in vivo. PAMAM dendrimers were conjugated to cisplatin to create a dendrimer-platinate molecule that was highly water-soluble and released the
drug slowly (40). This compound displayed antitumor activity and was
less toxic than cisplatin alone. Kojima et al. (41) report that PAMAM
dendrimers with poly(ethylene glycol) grafts possessed a space that could
carry anticancer drugs with a biocompatible surface, such as adriamycin and
methotrexate. Increasing generations and chain length of the poly(ethylene
glycol) grafts allowed for increasing amounts of drugs to be encapsulated
within the space. The dendrimer carrying methotrexate released the drug
slowly in an aqueous solution but rapidly in an ionic solution (41). Vasey et
al. (42) used a polymer linked to doxurubicin (adriamycin) in a phase 1
clinical trial. The drug-polymer complex (PK1) was given to patients with
refractory solid tumors such as colorectal, breast, ovary, pancreas, and

Copyright 2003 Marcel Dekker, Inc.

480

Loutsch et al.

others. PK1 was found to decrease the nonspecic organ toxicity and side
eects to the drug while maintaining antitumor activity without polymer
toxicity (42). Although this study did not employ a dendrimer, it demonstrates that polymers could play as signicant role in future cancer chemotherapy regimens.
B.

Dendrimer-Mediated Gene Therapy for Cancer and

Genetic Disease

Gene therapy is an exciting new eld focusing on the treatment of illnesses

via genetic modication. Conventional medicine often manages to increase
the patients degree of comfort by alleviating the symptoms of the disease
while not actually treating the disease. Gene therapy may hold the key to a
more permanent solution by seeking to correct the problem at the root of
the diseasethe defective gene. A reliable gene-delivery vehicle is an indispensable component of gene transfer therapy. Current gene-delivery systems
use viruses or liposomes as vectors. However, these vectors have many
problems such as immunogenicity or carcinogenicity and potential infectivity (43). The characteristics of the dendrimer (nontoxic, nonimmunogenic,
and nonviral) make it a very good candidate for gene therapy. (The use of
StarburstTM dendrimers in gene transfer is reviewed in Ref. 44.)
The electrostatic interactions between DNA (negative charge) and the
dendrimer (positive charge) yield a complex that forms quickly. The complex is stable over a wide range of pH (45). The DNA condenses when it
contacts the polymer, and the degree of condensation is dened by the
generation of the dendrimer, the concentration of the DNA, and the
DNA : dendrimer charge ratio (46). As the DNA : dendrimer charge ratio
increases, so does the biological function of the complex. Electron microscopy shows that DNA complexed with polylysine or whole dendrimer
aggregates in solution, whereas DNA complexed with polyethylenimine or
fractured dendrimer remains as single discrete units (47). Bielinska et al. (46)
and Kukowska-Latallo et al. (45) reported that DNA-dendrimer complexes
have a broader concentration range between transfection and cytotoxicity in
numerous cell lines.
The eciency of in vitro transfection increases exponentially with the
increase in the generation number of the dendrimer (45). The DNA is protected from nuclease degradation while bound to the dendrimer; however,
the DNA is unable to initiate transcription from the promoter. This may be
due to the inaccessibility of the promoter to the polymerase enzyme because
the secondary and tertiary structure of the DNA is altered when it is complexed with the dendrimer. Elongation of the RNA transcript and translation do not appear to be aected in these molecules (48). Additional in vivo

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

481

studies are needed to determine cellular localization and transfection eciency with these dendrimer-DNA complexes.
1.

Cancer Gene Therapy

One interesting method for treating certain cancers is to program the tumor
cells to self-destruct. This can be achieved by inserting a viral gene, usually
herpes simplex virus thymidine kinase (TK), directly into tumor cells,
enabling the cells to produce the enzyme. The tumor is then treated with
ganciclovir, the substrate of TK, creating the active form of the drug that
causes the cancer cells to commit suicide. Harda et al. (49) transfected
human hepatocellular carcinoma cells with an Epstein-Barr virusbased
plasmid vector containing the TK gene complexed with either PAMAM
dendrimers or cationic liposomes. Addition of therapeutic concentrations
of ganciclovir resulted in a signicant decrease in the number of viable
carcinoma cells in the dendrimer-treated cultures. Carcinoma cells transfected with the liposome vector alone survived the same dose of ganciclovir.
The dendrimer-based vectors showed increased eciency of gene transfer.
These results demonstrate the successful use of dendrimers in the transfer of
suicide genes in vitro (49).
Dendrimers are being considered for use in targeting therapy to solid
organ tumors. Reddy et al. (50) used a folate-targeted, cholesterol-stabilized
liposomal vector linked to a sixth-generation dendrimer to target tumor
cells. Certain cancer cells over-express folic acid receptor on their surface,
making this a viable mechanism for targeting therapeutic and imaging
agents to the tumor. The folic acidlinked dendrimer had a higher transfection eciency compared to the control dendrimer in cell culture (50).
Another method of cancer chemotherapy is to target boron-10, a
stable isotope, to the tumor and then activate it by the use of low-energy
irradiation, which excites the boron-10 and causes the death of the cancer
cell. Barth et al. (51) used boron-10 coupled to a Starburst dendrimer and
conjugated to an antibody against mouse B16 melanoma to treat mice with
tumors. They found that the immunoconjugate was localized largely in the
liver and spleen rather than in the tumor, even though the antibody is very
specic for the tumor in vitro (51). This novel approach to targeting and
treating tumor cells could be eective if antibody specic for the tumor in
vivo can be developed. Yang et al. (52) exploited the fact that gliomas
overproduce the epidermal growth factor receptor on the surface of the
cancer cell. They linked epidermal growth factor to a boronated fourthgeneration Starburst dendrimer and administered it either intravenously
or intratumorally to rats with glial tumors. The intratumor injection
resulted in 22% and 16% localization in the tumor at 24 and 48 hours,

Copyright 2003 Marcel Dekker, Inc.

482

Loutsch et al.

respectively, whereas the intravenous dose showed virtually no localization

to the tumor (52). Additionally, the intravenous dose had a greater localization to the liver and spleen than the intratumor injection. These results
indicate that intracerebellar injection could be the most eective treatment
delivery system in epidermal growth factor receptorpositive tumors.
2. Cardiac Gene Therapy
Qin et al. (53) used plasmid-mediated gene transfer and subsequent expression in a mouse cardiac transplant model to improve graft survival. A
plasmid encoding viral interleukin 10 (IL-10) under the control of
-myosin
heavy chain promoter (
-MHC-vIL-10) injected into the graft prolonged
allograft survival signicantly from 13.9 days to 21.4 days. When
MHC-vIL-10 was complexed with G5EDA dendrimer, a 60-fold decrease
in DNA resulted in signicantly prolonged graft survival to 38.6 days (53).
Increasing the amount and time of immunosuppressive cytokine expression
in more tissues allowed for further prolongation of graft survival. This novel
method of local immunosuppression could be benecial to patients undergoing solid organ transplantation.
A concern for dendrimer-mediated therapy is cytotoxicity at the site of
the dendrimer localization. Brazeau et al. (54) report that PAMAM dendrimers alone had high myotoxicity in an isolated rat muscle cell. The toxicity
decreased when the dendrimer was coupled to a DNA plasmid but not to the
level of saline or DNA alone (54). While this is an in vitro system, it is
imperative that toxicity studies be examined in vivo.
3. Genetic Disease Therapy
Oligonucleotide therapy has limiting factors that may reduce its usefulness.
The majority of these therapies involves the incorporation of antisense oligonucleotide sequences that bind to the mRNA of cells and interfere with
protein synthesis. Fomivirsen, and antiviral, is a successful antisense oligonucleotide therapy that is clinically available for the treatment of cytomegalovirus retinitis (5557). The factors that limit the usefulness of
oligonucleotide therapy are degradation, inecient cellular uptake and
intracellular transport, and binding to the RNA target. The use of dendrimers could alleviate these problems by the generation of cell-targeting dendrimers that could deliver the gene to the appropriate cell.
Yoo and Juliano (58) used fth-generation dendrimers labeled with
Oregon green 488 and an oligonucleotide labeled with TAMRA, a red
uorophore to determine localization in transfected HeLa cell by uorescence microscopy. The oligonucleotide would correct a splicing error in the
reporter gene, luciferase, if the transfection was successful and functionally

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

483

active. They found luciferase activity in the cells as well as the dendrimer
and oligonucleotide localized in close proximity to one another.
Interestingly, the dendrimers with Oregon green 488 enhanced the delivery
of the PAMAM dendrimer-oligonucleotide complex to the cells compared
to the unlabeled dendrimer-oligonucleotide complex (58). Helin et al. (59)
used FITC-labeled oligodeoxynucleotides (ODN) with and without dendrimers as carriers to determine cellular uptake and intracellular distribution in
various cell lines (Table 1). The FITC-ODN had increased cellular uorescence when complexed with dendrimers. Some cells had perinuclear uorescence and some cells had intense nuclear staining (59). These studies
indicate that oligonucleotide can reach the nucleus, the target for gene
therapy. Nilsen et al. (60) suggest that dendrimers can be constructed
from nucleic acids for use as a nucleic acid membrane. This structure
could be useful in gene therapy but to date has not been tested.
Cystic brosis is a genetic disease caused by a deletion mutation in the
cystic brosis transmembrane conductance regulator (CFTR) gene that
aects the pulmonary system. Patients, mostly children, suer from respiratory congestion and numerous lung infections. Goncz et al. (61) used dendrimers to deliver a fragment of DNA that could correct the mutated
sequence in CFTR. The small single-stranded DNA fragment (488 nucleotides long) was complexed with a Starburst dendrimer and used to transfect
primary normal airway epithelial cells (61). The DNA fragment corrected
the defective gene in 1 out of 100 cells, a range that could have therapeutic
potential. This method of small fragment homologous replacement therapy
could be developed into gene therapy strategies for disease caused by sitespecic mutations.
C.

Additional Therapeutic Uses of Dendrimers

Dendrimers linked with DNA can retain their ability to transfect cells after
drying, unlike viral vectors. Bielinska et al. (62) used this novel property of
dendrimer-DNA to coat a solid-phase, bioerodable support or to embed the
complex in the support. The DNA was a plasmid that expressed either
luciferase or green uorescent protein when transfected into cells.
Expression of luciferase or green uorescent protein was detected following
release of the complex from the support or while still attached (62). A
membrane that had a dendrimer-DNA complex on the surface was placed
on the skin of a hairless mouse. Chloramphenicol transacetylase activity was
found in the tissue under the membrane indicating that the DNA was delivered to the skin from the membrane (62).
Kawase et al. (63) created a high-performance, bio-articial liver support system using fructose-modied dendrimers immobilized on polystyrene

Copyright 2003 Marcel Dekker, Inc.

484

Loutsch et al.

dishes. The modied dishes were capable of increasing the number of hepatocytes attached to the surface, which in turn increased urea synthesis. The
number of cells remained high as the fructose-modied dendrimers appeared
to suppress apoptosis (63). Increasing the generation of the dendrimer led to
increased numbers of hepatocytes adherent to the support system. This may
be a novel way to treat fulminant hepatic failure, similar to renal dialysis.

VI.

OCULAR THERAPY

A.

Corneal Drug and Gene Delivery

The cornea is composed of epithelial, stromal, and endothelial layers that

provide a barrier to the penetration of most drugs used to treat ocular
disease. As such, the ocular surface presents a challenge for chemotherapy.
The epithelial surface is susceptible to bacterial, fungal, and viral infection
and inammation, as well as to mechanical injury. The stroma may be
aected by granular or lattice corneal dystrophy, which are diseases that
result in abnormal protein deposits in the stroma caused by autosomal
dominant mutation. The corneal endothelium function to maintain corneal
transparency and normal visual acuity. The endothelial cells cannot regenerate, and therefore corneal transplants are the only way to correct endothelial damage that results in blindness. The ability of dendrimers to be targeted
to specic cell types and to carry drugs or genes could allow for the treatment of these diseases and prevention of corneal graft rejection.
Hudde et al. (64) reported the eciency of gene transfer to the corneal
endothelium using PAMAM dendrimers in an ex vivo model. The dendrimer was complexed with a plasmid containing a reporter gene, -galactosidase, in a ratio of 18 : 1 and incubated with a quarter of a full-thickness
rabbit cornea. The expression of -galactosidase was assessed after 3 days:
610% of the endothelial cells were stained blue (64). The transfection of the
dendrimer-plasmid complex was also determined in human corneas but with
less eciency ( 2% of cells). There was no evidence of cellular toxicity
from the dendrimer. The staining was seen only in the endothelium and
not in the epithelium or stroma in both rabbit and human corneas (64).
In the same study, the introduction and expression of the gene, tumor
necrosis factor receptor fusion protein (TNFR-Ig) was assessed in this ex
vivo system. Tumor necrosis factor (TNF) is produced in high concentrations during transplant rejection and causes severe damage to the graft.
TNFR-Ig could prevent the TNF response and therefore protect the graft.
Rabbit corneas incubated with the dendrimer/plasmid secreted an active
form of TNFR-Ig into the supernatant that inhibited TNF-mediated toxicity in a bioassay (64).

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

485

Corneas that will be used for transplantation are stored in a special

medium prior to surgery. This corneal storage medium could be modied to
contain dendrimers complexed with genes expressing important proteins
that could prevent graft rejection. Hudde et al. (64) demonstrated that the
dendrimer/TNFR-Ig gene complex transfected into ex vivo corneas produces an active protein that inhibits TNF-mediated cytotoxicity. Qin et al.
(53) prolonged cardiac graft survival in mice by giving a dendrimer carrying
a plasmid coding for IL-10, an interleukin that enhances immunosuppression and hence improves graft survival. Combinations of these strategies
could prolong survival of the transplanted cornea. Dendrimer therapy
could also correct hereditary diseases such as lattice corneal dystrophy by
introducing the normal gene (i.e., transforming growth factor, -induced
gene) into the endothelial cells so a functioning protein could be expressed.
While no ocular dendrimer drug delivery studies have been performed to
date, the possibility of dendrimers being used carry drugs to corneal cells
should be investigated.
B.

Intraocular Drug and Gene Delivery

Treatment of the inner ocular tissues and the posterior region of the eye has
posed a signicant challenge to ophthalmologists. A major tissue to be
targeted is the retina, the transparent tissue lining the back of the eye that
is responsible for visual acuity, color discrimination, and peripheral vision.
Diseases of the retina include retinitis pigmentosa and macular degeneration; both result in blindness due to the death of retinal cells. Drugs applied
to the cornea usually do not reach the back of the eye. Therefore, direct
injection of the drugs into the vitreous is required to target the retina. While
this method provides adequate drug levels, it is not without signicant risk
of retinal detachment, endophthalmitis, vitreal hemorrhage, and/or cellular
toxicity (6567). The use of dendrimers to deliver drugs and gene therapy to
this area of the eye has great potential.
Urtti et al. (68) investigated the ecacy of in vitro gene delivery in fetal
primary human retinal pigment epithelial (RPE) cells using plasmids encoding the reporter gene -galactosidase or luciferase. The reporter gene plasmids were complexed with either degraded sixth-generation PAMAM
dendrimers, cationic lipids, or polyethylene imines (PEI; 25 or 750 kDa)
to determine the level of protein expression and cellular toxicity in RPE
cells. A degraded dendrimer has several branches removed by boiling in
butanol, which does not alter the size and shape but does increase the spaces
within the dendrimer and increases the exibility of the molecule. Naked
plasmid DNA was not as eective as complexed DNA in transfecting the
RPE cells. When the luciferase plasmid was linked with the dendrimer or

Copyright 2003 Marcel Dekker, Inc.

486

Loutsch et al.

PEI-25, the expression of luciferase was signicantly higher than with the
other carriers (68). The dendrimer/plasmid and PEI-750/plasmid were also
less toxic to the cells than the other vehicles. Although only a small number
of RPE cells were transfected, a large amount of protein was expressed from
the plasmid (68). This is especially signicant with a secreted protein product; a few cells making the protein could secrete enough to counteract the
deciency in all the cells. This method of delivering genes works well in
vitro, but additional investigations are needed before the in vivo impact
can be determined.

VII.

CONCLUSION

The ability of dendrimers to carry drugs and genes may allow for the
advancement of drug and gene therapy, not only in ophthalmology, but
in other areas of medicine. In the eld of ocular drug and gene delivery,
the controlled architecture and the nontoxic, nonimmunogenic nature of
dendrimers raise the possibility of future replacement of currently problematic delivery systems. The creation of dendrimers that can target specic
cells and carry a particular therapy will allow physicians to enhance treatment regimens.

ACKNOWLEDGMENTS
The authors wish to acknowledge Jean Jacobs, Ph.D., Kristina Braud, and
Kathy Vu for assistance with this review. None of the authors have any
nancial or proprietary interest in any of the agents or devices in this review.

REFERENCES
1. Olejnik, O. (1993). Conventional systems in ophthalmic drug delivery. In: A.
K. Mitra (ed.), Ophthalmic Drug Delivery Systems. Marcel Dekker, Inc., New
York, p. 177.
2. Bawa, R. (1993). Ocular insert. In: Ophthalmic Drug Delivery Systems. A. K.
Mitra (ed.). Marcel Dekker, Inc., New York, p. 223.
3. Hill, J. M., OCallaghan, R. J., Hobden, J. A., and Kaufman, H. E. (1993).
Corneal collagen shields for ocular drug delivery. In: Ophthalmic Drug
Delivery Systems. A. K. Mitra (ed.). Marcel Dekker, Inc., New York, p. 261.

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

487

4. Davies, N. M., Kellaway, I. W., Greaves, J. L., and Wilson, C. G. (1993).

Advanced corneal delivery systems: liposomes. In: Ophthalmic Drug Delivery
Systems. A. K Mitra (ed.). Marcel Dekker, Inc., New York, p. 289.
5. Hill, J. M., OCallaghan, R. J., and Hobden, J. A. (1993). Ocular iontophoresis. In: Ophthalmic Drug Delivery Systems. A. K. Mitra (ed.). Marcel Dekker,
Inc., New York, p. 331.
6. Hill, J. M., OCallaghan, R. J., Hobden, J. A., and Kaufman, H. E. Collagen
shields for ocular drug delivery. In: Ophthalmic Drug Delivery Systems. M. K.
Mitra (ed.). Marcel Dekker, Inc., New York, p. 261.
7. Tomalia, D. A. (1995). Dendrimer molecules. Sci. Am., 5:42.
8. Tomalia, D. A., and Brothers, H. M. I. (1998). Regiospecic conjugation to
dendritic polymers to produce nanodevices. In: Biological Molecules in
Nanotechnology: The Convergence of Biotechnology, Polymer Chemistry, and
Material Science. International Business Communications, Inc.,
Southborough, MA, pp. 107120.
9. Flory, P. J. (1941). Molecular size distribution in three dimensional polymers.
I. Gelation. J. Am. Chem. Soc., 63:30833090.
10. Flory, P. J. (1941). Molecular size distribution in three dimensional polymers.
II. Trifunctional branching units. J. Am. Chem. Soc., 63:30913096.
11. Flory, P. J. (1941). Molecular size distribution in three dimensional polymers.
III. Tetrafunctional branching units. J. Am. Chem. Soc., 63:30963100.
12. Stockmayer, W. H. (1943). Theory of molecular size distribution and gel
formation in branched-chain polymers. J. Chem. Phys., 11:4555.
13. Stockmayer, W. H. (1944). Theory of molecular size distribution and gel
formation in branched polymers. II. General cross linking. J. Chem. Phys.,
12:125131.
14. Buhleier, E., Wehner, W., and Vogtle, F. (1978). Cascade- and nonskidchain-like syntheses of molecular cavity topologies. Synthesis, 55:155158.
15. Newkome, G. R., Yao, Z.-Q., Baker, G. R., and Gupta, V. K. (1985). Cascade
molecules: A new approach to micelles. A [27]-arborol. J. Org. Chem.,
50:20032004.
16. Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Kallos, G., Roeck, J., Ryder,
J., and Smith, P. (1985). A new class of polymers: Starburst-dendritic macromolecules. Polym. J., 17:117132.
17. Tomalia, D. A., Naylor, A. M., and Goddard, W. A. I. (1990). Starburst
dendrimers: molecular-level control of size, shape, surface chemistry, topology, and exibility from atoms to macroscopic matter. Angew. Chem. Int. Ed.
Engl., 29:138175.
18. Hawker, C. J., and Frechet, J. M. J. (1990). Preparation of polymers with
controlled molecular architecture. A new convergent approach to dendritic
macromolecules. J. Am. Chem. Soc., 112:76387647.
19. Tomalia, D. A., and Esfand, R. (1997). Dendrons, dendrimers, and dendrigrafts. Chem. Ind., 6:416420.
20. Frechet, J. M. J. (1994). Functional polymers and dendrimers: Reactivity,
molecular architecture, and interfacial energy. Science, 263:17101715.

Copyright 2003 Marcel Dekker, Inc.

488

Loutsch et al.

21. Naylor, A. M., Goddard III, W. A., Kiefer, G. A., and Tomalia, D. A. (1989).
Starburst dendrimers. 5. Molecular shape control. J. Am. Chem. Soc.,
111:23392341.
22. Jansen, J. F. G. A., de Brabander-van den Berg, E. M. M., and Meijer, E. W.
(1994). Encapsulation of guest molecules into a dendritic box. Science,
266:12261229.
23. Stephan, H., Spies, H., Johannsen, B., Kaumann, C., and Vogtle. (2000).
pH-controlled inclusion and release of oxyanions by dendrimers bearing
methyl orange moieties. Org. Lett., 2:23432346.
24. Singh, P. (1998). Terminal groups in Starburst dendrimers: activation and
reactions with proteins. Bioconjug. Chem., 9:54.
25. Scott, D. A., Krulle, T. M., Finn, M., and Fleet, G. W. J. (2000). Bis(1,3dihydroxy-isopropyl)amine (BDI) as an AB4 dendritic building block: rapid
synthesis of a second generation dendrimer. Tetrahedron Lett., 41:3959.
26. Veprek, P., and Jezek, J. (1999). Peptide and glycopeptide dendrimers. Part II.
J. Peptide Sci., 5:203.
27. Wiwattanapatapee, R., Carreno-Gomez, B., Malik, N., and Duncan, R.
(2000). Anionic PAMAM dendrimers rapidly cross adult rat intestines in
vitro: A potential oral delivery system? Pharm. Res., 17:991.
28. Sakthivel, T., Toth, I., and Florence, A. T. (1999). Distribution of a lipid 2.5
nm diameter dendrimer carrier after oral administration. Int. J. Pharm.,
183:51.
29. Florence, A. T., Sakthivel, T., and Toth, I. (2000). Oral uptake and translocation of a polylysine dendrimer with a lipid surface. J. Controlled Release,
65:253.
30. Malik, N., Wiwattanapatapee, R., Klopsch, R., Lorenz, K., Frey, H., Weener,
J. W., Meijer, E. W., Paulus, W., and Duncan R. (2000). Dendrimers: relationship between structure and biocompatibility in vitro, and preliminary
studies on the biodistribution of 125 I-labelled polyamidoamine dendrimers in
vivo. J. Controlled Release, 65:133.
31. Wiener, E. C., Brechbiel, M. W., Brothers, H., Magin, R. L., Gansow, O. A.,
Tomalia, D. A., and Lauterbur, P. C. (1994). Dendrimer-based metal celates:
A new class of magnetic resonance imaging contrast agents. Magn. Reson.
Med., 31:1.
32. Konda, S. D., Aref, M., Brechbiel, M., and Wiener, E. C. (2000).
Development of a tumor-targeting MR contrast agent using the high-anity
folate receptor. Invest. Radiol., 35:50.
33. Liu, M., Kono, K., and Frechet, M. J. (2000). Water-soluble dendritic unimolecular micelles: Their potential as drug delivery agents. J. Controlled
Release, 65:121.
34. Attia, S. A., Shepherd, V. E., Rosenblatt, M. N., Davidson, M. K., and
Hughes, J. A. (1998). Interaction of oligodeoxynucleotides with mycobacteria:
Implications for new therapeutic strategies. Antisense Nucleic Acid Drug Dev.,
8:207.

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

489

35. Reuter, J. D., Myc, A., Hayes, M. M., Gan, Z., Roy, R., Qin, D., Yin, R.,
Piehler, L. T., Esfand, R., Tomalia, D. A., and Baker, J. R. Jr. (1999).
Inhibition of viral adhesion and infection by sialic-acid-conjugated dendritic
polymers. Bioconjug. Chem., 10:271.
36. Bourne, N., Stanberry, L. R., Kern, E. C., Holan, G., Matthews, B., and
Bernstein, D. I. (2000). Dendrimers, a new class of candidate topical microbicides with activity against herpes simplex virus infection. Antimicrob. Agents
Chemother., 44:2471.
37. Barnard, D. L., Sidwell, R. W., Gage, T. L., Okleberry, K. M., Mathews, B.,
and Holan, G. (1997). Anti-respiratory syncytial virus activity of dendrimer
polyanions. Antiviral Res., 34:A88.
38. Witvrouw, M., Pannecouque, C., Matthews, B., Schols, D., Andrei, G.,
Snoeck, R., Neyts, J., Leyssen, P., Desmyter, J., Ra, J., DeClerq, E., and
Holan, G. (1999). Dendrimers inhibit the replication of human immunodeciency virus (HIV) by a dual mechanism of action. Antiviral Res., 41:A41.
39. Duncan, R., and Malik, N. (1996). Dendrimers: biocompatibility and potential for delivery of anticancer agents. Proc. Int. Symp. Control Rel. Bioact.
Mater., 23:105.
40. Malik, N., Evagorou, E. G., and Duncan, R. (1999). Dendrimer-platinate: a
novel approach to cancer chemotherapy. Anticancer Drugs, 10:767.
41. Kojima, C., Kono, K., Maruyama, K., and Takagishi, T. (2000). Synthesis of
polyamidoamine dendrimers having poly(ethylene glycol) grafts and their
ability to encapsulate anticancer drugs. Bioconjugate Chem., 11:910.
42. Vasey, P. A., Kaye, S. B., Morrison, R., Twelves, C., Wilson, P., Duncan, R.,
Thomson, A. H., Murray, L. S., Hilditch, T. E., Murray, T., Burtles, S.,
Fraier, D., Frigerio, E., Cassidy, J., on behalf of the Cancer-Research
Campaign Phase I/II Committee. (1999). Phase I clinical and pharmacokinetic
study of PK1 [N-(2-hydroxypropyl)methacrylamide copolymer doxorubicin]:
rst member of a new class of chemotherapeutic agents drug-polymer conjugates. Clin. Cancer Res., 5:83.
43. Cheng, H., Zhou, R., Liu, L., Du B., and Zhou, R. (2000). Cyclic core dendrimer as a new kind of vector for gene transfer into mammalian cells.
Genetica, 108:53.
44. Kukowska-Latallo, J. F., Bielinska, A. U., Chen, C., Rymaszewski, M.,
Tomalia, D. A., and Baker, J. R. Jr. (1998). Gene transfer using
StarburstTM dendrimers. In Self-Assembling Complexes for Gene Delivery:
From Laboratory to Clinical Trial. A. V. Kabanov, P. L. Felgner, and L.
W. Seymour (eds.). John Wiley & Sons, Ltd., Chichester, p. 240.
45. Kukowska-Latallo, J. F., Bielinska, A. U., Johnson, J., Spindler, R., Tomalia,
D. A., and Baker, J. R. Jr. (1996). Ecient transfer of genetic material into
mammalian cells using Starburst polyamidoamine dendrimers. Proc. Natl.
Acad. Sci. USA, 93:4897.
46. Bielinska, A., Kukowska-Latallo, J. F., Johnson, J., Tomalia, D. A., and
Baker, J. R. Jr. (1996). Regulation of in vitro gene expression using antisense

Copyright 2003 Marcel Dekker, Inc.

490

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.
57.

58.

59.

Loutsch et al.
oligonucleotides or antisense expression plasmids transfected using starburst
PAMAM dendrimers. Nucleic Acids Res., 24:2176.
Tang, M. X., and Szoka, F. C. (1997). The inuence of polymer structure on
the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther., 4:823.
Bielinska, A. U., Kukowska-Latallo, J. F., and Baker, J. R. Jr. (1997). The
interaction of plasmid DNA with polyamidoamine dendrimers: mechanism of
complex formation and analysis of alterations induced in nuclease sensitivity
and transcriptional activity of the complexed DNA. Biochim. Biophys. Acta,
1353:180.
Harada, Y., Iwai, M., Tanaka, S., Okanoue, T., Kashima, K., MaruyamaTabata, H., Hirai, H., Satoh, E., Imanishi, J., and Mazda, O. (2000). Highly
ecient suicide gene expression in hepatocellular carcinoma cells by epsteinbarr virus-based plasmid vectors combined with polyamidoamine dendrimer.
Cancer Gene Ther., 7:27.
Reddy, J. A., Dean, D., Kennedy, M. D., and Low, P. S. (1999). Optimization
of folate-conjugated liposomal vectors for folate receptor-mediated gene therapy. J. Pharm. Sci., 88:1112.
Barth, R. F., Adams, D. M., Soloway, A. H., Alam, F., and Darby, M. V.
(1994). Boronated starburst dendrimer-monoclonal antibody immunoconjugates: evaluation as a potential delivery system for neutron capture therapy.
Bioconjug. Chem., 5:58.
Yang, W., Barth, R. F., Adams, D. M., and Soloway, A. H. (1997).
Intratumoral delivery of boronated epidermal growth factor for neutron capture therapy of brain tumors. Cancer Res., 57:4333.
Qin, L., Pahud, D. R., Ding, Y., Bielinska, A. U., Kukowska-Latallo, J. F.,
Baker, J. R. Jr., and Bromberg, J. S. (1998). Ecient transfer of genes into
murine cardiac grafts by Starburst polyamidoamine dendrimers. Human Gene
Ther., 9:553.
Brazeau, G. A., Attia, S., Poxon, S., and Hughes J. A. (1998). In vitro myotoxicity of selected cationic macromolecules used in non-viral gene delivery.
Pharm. Res., 15:680.
de Smet, M. D., Meenken, C. J., and van den Horn, G. J. (1999).
Fomivirsena phosphorothioate oligonucleotide for the treatment of CMV
retinitis. Ocul. Immunol. Inamm., 7:189.
Perry, C. M., and Balfour, J. A. (1999). Fomivirsen. Drugs, 57:375.
Bochot, A., Couvreur, P., and Fattal, E. Intravitreal administration of antisense oligonucleotides: potential of liposomal delivery. Prog. Retin. Eye Res.,
19:131.
Yoo, H., and Juliano, R. L. (2000). Enhanced delivery of antisense oligonucleotides with uorophore-conjugated PAMAM dendrimers. Nucleic Acids
Res., 28:4225.
Helin, V., Gottikh, M., Mishal, Z., Subra, F., Malvy, C., and Lavignon, M.
(1999). Uptake and intracellular distribution of oligonucleotides vectorized by
a PAMAM dendrimer. Nucleosides Nucleotides, 18:1721.

Copyright 2003 Marcel Dekker, Inc.

Dendrimers

491

60. Nilsen, T. W., Grayzel, J., and Prensky, W. (1997). Dendritic nucleic acid
structures. J. Theor. Biol., 187:273.
61. Goncz, K. K., Kunzelmann, K., Xu, Z., Gruenert, D. C. (1998). Targeted
replacement of normal and mutant CFTR sequences in human airway epithelial cells using DNA fragments. Hum. Mol. Genet., 7:1913.
62. Bielinska, A. U., Yen, A., Wu, H. L., Zahos, K. M., Sun, R., Weiner, N. D.,
Baker, J. R. Jr., and Roessler, B. J. (2000). Application of membrane-based
dendrimer/DNA complexes for solid phase transfection in vitro and in vivo.
Biomaterials, 21:877.
63. Kawase, M., Shiomi, T., Matsui, H., Ouji, Y., Higashiyama, S., Tsutsui, T.,
and Yagi, K. (2001). Suppression of apoptosis in hepatocytes by fructosemodied dendrimers. J. Biomed. Mater. Res., 54:519.
64. Hudde, T., Rayner, S. A., Comer, R. M., Weber, M., Isaacs, J. D.,
Waldmann, H., Larkin, D. F. P., and George, A. J. T. (1999). Activated
polyamidoamine dendrimers, a non-viral vector for gene transfer to the corneal endothelium. Gene Ther. 6:939.
65. Henry, K., Cantrill, H., Fletcher, C., Chinnock, B. J., and Balfour, H. H. Jr.
(1987). Use of intravitreal ganciclovir (dihydroxy propoxymethyl guanine) for
cytomegalovirus retinitis in a patient with AIDS. Am. J. Ophthalmol., 103:17.
66. Schwartz, D. M. (1996). New therapies for cytomegalovirus retinitis. Int.
Ophthalmol. Clin., 36:1.
67. Harris, M. L., and Mathalone, M. B. (1989). Intravitreal ganciclovir in CMV
retinitis: Case report. Br. J. Ophthalmol., 73:382.
68. Urtti, A., Polansky, J., Lui, G. M., and Szoka, F. C. (2000). Gene delivery and
expression in human retinal pigment epithelial cells: eects of synthetic carriers, serum, extracellular matrix, and viral promoters. J. Drug. Target, 7:413.
69. Maruyama-Tabata, H., Harada, Y., Matsumura, T., Satoh, E., Cui, F., Iwai,
M., Kita, M., Hibi, S., Imanishi, J., Sawada, T., and Mazda, O. (2000).
Eective suicide gene therapy in vivo by EBV-based plasmid vector coupled
with polyamidoamine dendrimer. Gene Ther., 7:53.
70. Wilbur, D. S., Pathare, P. M., Hamlin, D. K., Buhler, K. R., and Vessella, R.
L. (1998). Biotin reagents for antibody pretargeting. 3. Synthesis, radioiodination, and evaluation of biotinylated starburst dendrimers. Bioconjug. Chem.,
9:813.
71. Kono, K., Liu, M., and Frechet, J. M. (1999). Design of dendritic macromolecules containing folate or methotrexate residues. Bioconjug. Chem., 10:1115.
72. Thompson, J. P., and Schengrund, C. L. (1998). Inhibition of the adherence of
cholera toxin and the heat-labile enterotoxin of Escherichia coli to cell surface
GM1 by oligosaccharide-derivatized dendrimers. Biochem. Pharmacol.,
56:591.
73. Thompson, J. P., and Schengrund, C. L. (1997). Oligosaccharide-derivatized
dendrimers: Dened multivalent inhibitors of the adherence of the cholera
toxin B subunit and the heat labile enterotoxin of E. coli to GM1.
Glycoconj. J., 14:837.

Copyright 2003 Marcel Dekker, Inc.

492

Loutsch et al.

74. Fishbein, I., Chorny, M., Rabinovich, L., Banai, S., Gati, I., and Golomb, G.
(2000). Nanoparticle delivery system of a tyrphostin for the treatment of restenosis. J. Controlled Release, 65:221.
75. Benoit, J.-P., Faisant, N., Venier-Julienne, M.-C., and Menei, P. (2000).
Development of microspheres for neurological disorders: From basics to clinical applications. J. Controlled Release, 65:285.
76. Bousif, O., Lezoulch, F., Zanta, M. A., Mergny, M. D., Scherman, D.,
Demeneix, B., and Behr, J.-P. (1995). A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: Polyethylenimine. Proc.
Natl. Acad. Sci. USA, 92:7297.

Copyright 2003 Marcel Dekker, Inc.

16
Ocular Delivery and Therapeutics
of Proteins and Peptides
Surajit Dey and Ashim K. Mitra
University of MissouriKansas City, Kansas City, Missouri, U.S.A.

Ramesh Krishnamoorthy
Inspire Pharmaceuticals, Durham, North Carolina, U.S.A.

I.

INTRODUCTION

Recent advances in molecular biology and biotechnology have resulted in a

signicant increase in the number of protein and peptide drugs that have
been approved for clinical use. Recent use of recombinant DNA technology
has enabled us not only to produce these peptides and proteins in large
quantities but also to modify these peptides for enhanced therapeutic eect.
The application of peptide- and protein-based drugs in the treatment of a
variety of ailments has been well documented. Table 1 lists some of the
therapeutic agents and their clinical use (1). Along with a wide range of
clinical and therapeutic applications, these macromolecular compounds
exhibit a wide range of structures and sizes (Table 2).
The barriers associated with the delivery of peptides and proteins as
therapeutic agents are primarily associated with their physicochemical and
pharmacokinetic properties (size, charge, adsorption characteristics, etc.).
Due to their molecular size and hydrophilicity, peptides and proteins do
not cross membranes very readily. For these compounds to cross biological
membranes readily, the molecular size should be small enough to undergo
paracellular transport or be actively transported across the membrane. So
far, most proteins have been delivered by invasive routes, such as parenteral
administration. Repeated injections, particularly in the treatment of chronic
ailments, deter many patients from undergoing intensive therapy. Phlebitis
493

Copyright 2003 Marcel Dekker, Inc.

494

Dey et al.

Table 1 Selected Proteins and Peptides and Their Possible Therapeutic

Applications
Peptide/Protein
Growth hormone
Insulin
LHRH
TRH
Gastrin antagonists
Somatostatin
Neurotensin
Nerve growth factor
-Endorphin
Enkephalins
Cholecystokinin
Salmon calcitonin
Bradykinin
Angiotensin antagonists
Tissue plasminogen factors
Cyclosporin A
Interferon
Thymopoietin
Epidermal growth factor
Vasoactive intestinal peptide (VIP)
Substance P
Fibronectin

Indication
Increasing the height of children decient in
growth hormone.
Lowering blood glucose levels
Inducing ovulation in women
Prolonging infertility and lactation in women
Reducing secretion of gastric acid
Reducing bleeding of gastric ulcers
Inhibiting gastric juice secretion
Simulating nerve growth and repair
Relieving pain
Pain suppressor and reducing inammation
Suppressing appetite
Osteoporosis
Improving peripheral circulation
Lowering blood pressure
Dissolving blood clots
Inhibiting function of T lymphocytes
Enhancing activity of natural killer cells
Selective T-celldierentiating hormone
Wound healing
Inducing relaxation of smooth muscles
Reducing inammation, inducing analgesia
Helping in wound healing

and tissue irritation are some of the complications of parenteral delivery.

Oral delivery of peptides and proteins has been attempted in the past. The
gastrointestinal tract, which is rich in proteolytic enzymes, digests therapeutic peptides and proteins. Although oral bioavailability of most peptides and
proteins is less than 1%, a few peptides (e.g., cyclosporine) have high bioavailability when administered in a suitable vehicle. Arginine and lysine vasopressin (AVP and LVP) and the synthetic analog 1-deamino-8-d-arginine
vasopressin (DDAVP) are well absorbed following oral administration (3
5). Numerous attempts have been made to utilize alternative noninvasive
routes for peptide and protein administration (610). Numerous approaches
to overcome the problems of low absorption and rapid metabolism of proteins at the site of administration have been pursued. Although buccal (11),
nasal (1214), transdermal (15,16), tracheal (17), rectal (18), and vaginal (19)
administrations have been attempted, none so far has been found to be

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

Table 2

495

Peptide Sizes and Structural Features

Peptide
Sandostatin
Oxytocin
Vasopressin
Desmopressin
Gonadorelin
Tetracosactide
Glucagon
Calcitonin
Corticotrophin
Insulin
Aprotinin
Interferons
Trypsin
Gonadotrophin
Pepsin
Erythropoietin

Size

Structural features

8 amino acids
9 amino acids
9 amino acids
9 amino acids
10 amino acids
24 amino acids
29 amino acids
32 amino acids
39 amino acids
51 amino acids
6 kDa
1530 kDa
24 kDa
30 kDa
35 kDa
38 kDa

One-chain structure
Cyclic nanopeptides
Cyclic nanopeptides
Cyclic nanopeptides
One-chain structure
One-chain structure
One-chain structure
One-chain structure
One-chain structure
Two-chain structure
Single-chain polypeptide
Complex structure
Globular (650 amino acids)
Globular (
 heterodimer)
Globular
Globular (native hormone composed
of 165 amino acids)

Single chain, NH2 -terminal

gla-domain, two EGF domains
Oblate ellipsoid (140  40 A)
55%
-helix and 45% random
conformation
5 Recurring domains

Streptokinase
Urokinase
Factor IX

46 kDa
54 kDa
60 kDa

Albumin

67 kDa

Tissue plasminogen
activator
IgG

72 kDa
150 kDa

Factor VIII

330 kDa

Light and dark chains 2 Fab

and Fc portion
One heavy chain (90220 kDa)
and one light chain (76 kDa)

Source: Adapted from Ref. 2.

completely satisfactory. Studies have been carried out to compare the

transport barriers of several epithelia. The rank order of intrinsic
membrane permeability has been found to be as follows: intestine  nasal 
bronchial  tracheal  vaginal  rectal > corneal > buccal > skin (20).
The ocular route may be considered for systemic delivery of proteins
and peptides. This route can be especially benecial when delivering local
therapeutic agents to the anterior segment of the eye.

Copyright 2003 Marcel Dekker, Inc.

496

Dey et al.

Several peptides have been shown to traverse the cornea. Moreover,

systemic delivery of polypeptide and protein drugs through the ocular route
has several advantages:
1. The mode of delivery is convenient, i.e., eye drops.
2. Systemic absorption is extremely rapid.
3. The absorbed peptide can bypass the hepatic circulation and thus
avoid rst-pass metabolism.
4. The formulation can be designed to prolong drug action and/or
reduce drug concentrations to achieve consistent drug action
with least side eects.
5. Drug delivery can be controlled precisely.
While this route is well accepted by patients, it yields low systemic bioavailability. Size, charge, and hydrophilic nature of proteins are the primary
determinants of their extent of absorption. Because the systemic delivery
of peptides depends on their transport and contact time with mucous membrane of the conjunctiva and the nasolacrimal system (21), the following
factors need to be considered:
1. The eye drops used must be devoid of any local irritation and/or
side eects.
2. Inclusion of permeation enhancers may be necessary when dealing with agents with molecular weight larger than 1000.
3. The absolute quantity of polypeptide that can be instilled in the
eye cannot exceed 2.5 mg because the eye cannot tolerate more
than 25 L of a 10% (w/v) ophthalmic solution.
Systemic delivery of peptides via the ocular route may exert a local toxic
response. This chapter will summarize the work done in this area and will
review the experimental methods currently being employed.

II.

MECHANISM OF PEPTIDE TRANSPORT

Protein transport across biological membranes depends on a number of

physical properties, among which the size and polarity of the compound
are most signicant. From the changes in permeability coecients under
dierent experimental conditions, the major pathways of protein permeation can be hypothesized. The kinetic approach provides indirect evidence of
the permeability characteristics and can be a powerful tool in deducing
pathways.
Electrophysiological measurements and uorescent confocal microscopy have been used to understand the mechanism of corneal drug permea-

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

497

tion (22,23). The latter method provides a sensitive approach to monitoring

ux across the corneal tissue and characterizing tissue damage. Confocal
microscopy utilizes a laser computer system that permits visualization of a
tissue specimen under light microscopy.
Insulin, polylysine, and thyrotropin-releasing hormone have been studied for their corneal permeability behavior using confocal microscopy (24).
Based on the results obtained, the movement of these peptides across the
cornea appears to be governed by the following factors:
1. A paracellular route is favored irrespective of the charge or size
of the peptide.
2. The outermost two cell layers of the corneal epithelium oer the
maximum resistance; the intercellular spaces then widen, making
the transport across the corneal layer easier.
3. Charge plays an important role in the transport of the small- to
medium-sized peptides. The cornea oers greater resistance to
negatively charged compounds than it does to positively charged
ones.
Minor inammation of the cornea and calcium chelation may cause a
widening of the intercellular spaces, thereby allowing a substantial ux of
the peptide compounds. If polypeptide absorption in therapeutic amounts
across the cornea is to be achieved, it will be necessary to maintain prolonged contact of the peptide with the corneal surface and also to employ a
proper penetration enhancer to potentiate ux across the intercellular
spaces.

III.

METHODS FOR ENHANCING PEPTIDE AND PROTEIN

DELIVERY

Irrespective of the noninvasive route employed to deliver the peptide and

protein drugs, some inherent delivery problems must be overcome. Table 3
lists some of the general methods that can be employed to negotiate such
formulation diculties.
Physical methods for increasing absorption such as iontophoresis and
phonophoresis have been examined extensively (2528). Such methods of
enhancement are quite complex and may lead to chemical and physical
instability of the permeating protein (2931). Erratic results may also be
obtained due to dierent degrees of surface absorption.
Prolongation of the biological half-life of proteins and enhancement of
their systemic bioavailability are also necessary. An attempt has been made
by coadministering proteins with known enzyme inhibitors (32). The cova-

Copyright 2003 Marcel Dekker, Inc.

498

Dey et al.

Table 3 General Methods for Enhancing Protein Delivery

1.

2.

3.

By increasing absorption through:

(a) Application of physical methods like iontophoresis or phonophoresis
(b) Coadministration with permeation enhancers
(c) Incorporation into liposomes or other carriers
(d) Chemical modication of primary structure and development of
prodrugs
By minimizing metabolism through:
(a) Covalent attachment to a polymer
(b) Chemical modication of the primary structure
(c) Targeting to specic tissues
(d) Coadministration with an enzyme inhibitor
By prolonging blood levels through:
(a) Use of bioadhesives
(b) Protection using liposomes, polymers, or other carrier

lent attachment of polymers has also been shown to protect a number of

proteins from enzymatic hydrolysis (3337). Thus, azopolymers may be used
to deliver proteins orally whereby the system could bypass the digestive
enzymes of the small intestine; however, in the ora of the colon the polymer
will release the protein. In any event, the problem of low bioavailability still
exists, and permeation enhancers may have to be used to overcome this
obstacle.
The use of chemical permeation enhancers has been reviewed extensively relative to nasal, oral, and rectal absorption of proteins and peptides
(3841). Three major mechanisms of action are possible for these enhancers:
perturbation of membrane integrity, expansion of the paracellular pathway,
and increase in the thermodynamic activity of the permeating species.
The following section will briey discuss the studies performed in
other noninvasive routes of protein and peptide delivery, the ndings of
which may be useful in designing ocular peptide delivery systems.
A.

Oral Route

The oral route, though the most convenient, is the least likely to be successful because of extensive degradation of proteins and peptides in the gut.
Based on the concentration-dependent absorption of 1-desamino-8-d-arginine-vasopressin (DDAVP), Lundin and Artursson (42) suggested the process to be mediated by passive transport. Other investigators have shown
that the permeability of peptides can be enhanced through prodrug modication designed to utilize the peptide transport system of the digestive
tract (4346). Coadministration of enzyme inhibitors may oer some pro-

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

499

tection in conjunction with a delivery system that can target the protein to
the site of optimal absorption (38,47). The incorporation of absorption
enhancers to improve the oral bioavailability of proteins has been well
documented (48). Lundin et al. (49) showed that sodium taurodihydrofusidate (STDHF) enhanced both the in vitro and in vivo absorption of
DDAVP. Schilling and Mitra (50) used the everted gas sac technique to
evaluate the optimal site of insulin absorption. The addition of sodium
glycocholate and linoleic acid enhanced insulin absorption in the duodenum
and jejunum by eight- and threefold, respectively.
B.

Nasal Route

The extensive network of blood capillaries underneath the nasal mucosa

could provide eective systemic absorption of drugs. The nasal route is
capable of providing a rapid absorption with a bioavailability relatively
similar to that following subcutaneous injection. Nasal delivery of peptides
and proteins has been reviewed (51,52). The polypeptides intended for delivery by this route should be readily soluble in a low mucosal irritant vehicle.
It must also be absorbed in eective amounts to make this mode of administration both economical and acceptable (53). This route has been shown to
be acceptable for peptides with 10 residues or less (54,55). When the number
of amino acids in the peptide approaches 20 or more, satisfactory bioavailability is obtained only with a permeation enhancer (56). The nasal mucosa
contains enzymes capable of hydrolyzing peptides such as leucine enkephalin. It appears that the human nasal passage contains a variety of peptidases
with wide specicities. The enhancing eect of two enzyme inhibitors, amastatin and bestatin, a mucolytic agent, N-acetyl-l-cysteine, and the permeation enhancers palmitoyl-d,l-carnitine and l-
-lysophosphatidylcholine
(LPC) on the nasal absorption of human growth hormone (HGH) was
studied by OHagan et al. (57). The highest bioavailability relative to the
subcutaneous injection was found with amastatin, followed by LPC and
palmitoyl-d,l-carnitine. Tengammuay and Mitra (58,59) found that mixed
micelles of sodium glycocholate and fatty acids were more eective in
enhancing the nasal delivery of peptides than the bile salt itself. Vadnere
et al. (60) found that ethylenediaminetetraacetic acid (EDTA) and
-cyclodextrin were capable of increasing the bioavailability of leuprolide when
given intranasally.
C.

Buccal Route

Delivery of macromolecules through the buccal membrane has also received

considerable attention in recent years (6163). Both keratinized and non-

Copyright 2003 Marcel Dekker, Inc.

500

Dey et al.

keratinized mucosae have been used in studying the in vitro rate of penetration of drugs through the buccal tissue. In vivo absorption of peptides/
proteins from the buccal cavity is likely to be inuenced by the presence
of mucosal secretions and immunological reactions among other factors.
Molecular size may not be the limiting factor in the buccal delivery of
peptides (64). Gandhi and Robinson (65) reported that amino acid penetrate
the buccal membrane by an active process, whereas peptide drugs permeate
passively. The buccal cavity exhibits greater proteolytic enzyme activity than
the nasal or vaginal mucosa (64). The metabolic activity is shown to reside
primarily in the epithelium (67). Aungst and Rogers (8,68) studied a variety
of absorption enhancers to determine their eects on buccal absorption and
showed that signicant changes in the morphology of this mucosal barrier
take place following exposure to the absorption enhancers.
D.

Pulmonary

Delivery of protein and peptide drugs via the pulmonary route has also
received signicant attention in recent years. The walls of the alveoli are
thinner than the epithelial/mucosal membrane; the surface area of the lung
is much greater and the lungs receive the entire blood supply from the heart,
all of which work in favor for the absorption of protein drugs more rapidly
and to a greater extent. Of course, the lungs are rich in enzymes, and overcoming this barrier is no easy task. Peptide hydrolases, peptidases, and a
wide variety of proteinases are present in the lung cells (69). However, some
proteinases inhibitors are also present at concentrations varying with the
disease state, which might work to prevent the destruction of administered
peptides (70). Liposomal delivery of peptide and protein drugs through the
pulmonary route have been attempted (71). Molecular modications have
also been undertaken to explore this route of protein and peptide delivery
(72).
E. Ocular Route
Lee reviewed the factors aecting corneal drug penetration (73).
Rojanasakul et al. showed that polylysine permeated through epithelial surface defects via an intracellular pathway when administered to the eye,
whereas insulin predominates in the surface cells of the cornea (23). They
noted that there was a signicant amount of aminopeptidase activity present
in the ocular uids and tissues. Figure 1 summarizes the results of the
metabolism of topically applied enkephalins to the eye (74). Pretreatment
with the peptidase inhibitor bestatin had a signicant protease inhibitory
eect, albeit in the tears only.

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

501

Fig. 1 Concentration of intact (open bar) and degraded (marked bar) leucine
enkephalin (===), methionine enkephalin (\ \ \) or [D-Ala2]Met-enkephalinamide
(lled bar) recovered in each part of the rabbit eye. (From Ref. 74.)

Studies have been conducted with absorption enhancers to improve

the delivery of peptides and proteins into the systemic circulation via the
ocular route (7577). Table 4 lists some penetration enhancers that have
been used in the ocular delivery of peptide-like drugs. Ocular delivery of
insulin to generate a therapeutic glucose-lowering response requires a penetration enhancer (78). Yamamoto et al. (79) reported that the bioavailability

Copyright 2003 Marcel Dekker, Inc.

502

Dey et al.

Table 4 Penetration Enhancers Used to Improve Ocular Absorption

Enhancer
Azone
Cetrimide, cytochalasin B
EDTA
Taurocholate, taurodeoxycholate

Eect
Threefold increase in cyclosporine absorption
Increased absorption of inulin
Threefold increase in glycerol absorption
Increased permeation of insulin and
FITC-dextran

of insulin could be improved in the following descending order by coadministration of the permeation enhancers: polyoxyethylene-9-lauryl
ether > sodium deoxycholate > sodium glycocholate  sodium taurocholate.

IV.

OCULAR DELIVERY OF PEPTIDE AND PROTEIN

DRUGS

Peptides and proteins may be instilled into the eye for local/topical use.
Instillation of a topical dose of a drug to the eye leads to absorption of a
drug mainly through the conjunctival and corneal epithelia. For drugs
meant for topical use, it must be minimally absorbed systemically as it
can lead to undesirable side eects. Absorption into the systemic circulation
may occur across the conjunctiva and sclera. However, for local delivery the
cornea presents a signicant barrier to the introcular penetration of peptide
drugs in view of their high molecular weight and low lipophilicity. Lee et al.
(75) reported that the penetration of inulin through the rabbit cornea was
probably occurring via a paracellular route rather than a transcellular route.
Systemic absorption of peptide and protein drugs following topical
administration to the eye could occur through contact with the conjunctival
and nasal mucosae, the latter occurring as a result of drainage through the
nasolacrimal duct. When systemic eects are desired, absorption through
the conjunctival and nasal mucosae needs to be maximized. One also must
consider other competing processes present in the ocular tissues. Of these
processes, absorption by the avascular cornea is important, since a large
portion of the drug thus absorbed is distributed to adjacent ocular tissues.
Ahmed and Patton (80) found that noncorneal (scleral) absorption
accounted for about 80% absorption of inulin, a highly hydrophilic macromolecule, into the iris-ciliary body. This observation is important, since
most therapeutic peptides act locally in the iris-ciliary body, which is con-

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

503

tiguous with the sclera. Therefore, macromolecular drug absorption would

benet from scleral absorption.
Beside the transport barrier, another factor severely limiting the ocular
absorption of peptide drugs is metabolism by ocular enzymes, specically
peptidases. Endopeptidases, like plasmin and collagenase, and exopeptidases, like aminopeptidases, are present in the ocular uids and tissues.
The endopeptidase levels are usually low unless the eye is inamed (81,82)
or injured (83) and are of little concern relative to the stability of topically
applied doses. Lee et al. (74) reported that within 5 minutes postinstillation,
about 90% of leucine enkephalin and almost 100% of methionine enkephalin (pentapeptides) was recovered in the rabbit corneal epithelium in a
hydrolyzed form. Therefore, aminopeptidase activity must be inhibited to
facilitate ocular peptide absorption. Controlling these enzymes in the target
tissues may not be practical given the fact that the same enzymes might be
necessary for the homeostasis in the eye.
Cyclosporin A has been shown to improve the prognosis for corneal
allograft rejection. It was found that when administered by nonocular routes
in rabbits, it was detected in the systemic circulation but not in the ocular
tissues (20,84,85). Also, topical administration of cyclosporin A did not
produce any signicant penetration within the eye beyond the cornea or
the conjunctiva. This may be because cyclosporin A was bound to corneal
and conjunctival epithelial cell membranes. Cyclosporin A eyedrops formulated in absolute ethanol did produce higher levels in intraocular tissues,
which may be due to damage to corneal epithelium by alcohol.
Growth factors, especially epidermal growth factor (EGF), have been
found to stimulate cell proliferation in the corneal epithelium, thus stimulating epithelialization during wound healing. Growth factors are mostly
used in accelerating the wound-healing process, and it would be of great
importance in corneal wounds since the cornea is an avascular organ. Many
in vitro corneal preparations have been used to demonstrate the woundhealing process. Human EGF promotes endothelial wound healing (84).
Many other growth factors also play a major role in corneal wound healing,
including transforming growth factor  (TGF-) (87) and platelet-derived
growth factor (PDGF). Basic broblast growth factor (bFGF) and insulinlike growth factor I (IGF-I) have been found in higher levels in patients
suering from diabetic retinopathy (8890). IGF-I and bFGF can also
induce brovascular changes in the retinal vessels.
A more practical strategy for circumventing the enzymatic barrier
would be to administer peptide analogs that are resistant to the principal
peptidases but possess equivalent biological activity [D-Ala2 ]methionine
enkephalinamide (DAMEA), which resists aminopeptidase-mediated cleavage, falls in this category of peptide analogs (74). The permeation and

Copyright 2003 Marcel Dekker, Inc.

504

Dey et al.

metabolic degradation of DAMEA in the albino rabbit cornea, conjunctiva,

and sclera has been studied (91). DAMEA was administered with and without peptidase inhibitors bestatin (aminopeptidase inhibitor) and SCH 39370
(enkephalinase inhibitor). It was found that sclera was the most permeable
membrane to DAMEA, while cornea was almost impermeable to DAMEA.
Without inhibitors, the permeability coecients of DAMEA were 2:7
108 cm/s, 3:1  106 cm/s, and 12:5  106 cm/s across the cornea, conjunctiva, and sclera, respectively. When inhibitors were co-administered with
DAMEA, the corneal permeability of intact DAMEA increased 15 times,
conjunctival permeability increased 5.5 times, while scleral permeability
remained practically unaltered.
The corneal and conjunctival penetration of 4-phenylazobenzyloxycarbonyl-l-Pro-l-Leu-Gly-l-Pro-d-Arg (Pz-peptide) and its eect on the
corneal and conjunctival penetration of hydrophilic solutes as well as on
the ocular and systemic absorption of topically applied atenolol and propranolol in the rabbit have been evaluated (92). The conjunctiva was 29
times more permeable than the cornea to 3 mM Pz-peptide. Conjunctival
Pz-peptide transport was 1.7 times greater in the mucosal-to-serosal than
in the opposite direction, whereas corneal Pz-peptide transport showed no
directionality. The apparent permeability coecients of Pz-peptide across
the cornea and the conjunctiva increased over the 15 mM range, which
suggests that Pz-peptide enhanced its own transport across both epithelial
tissues. The cornea was more sensitive than the conjunctiva to the penetration-enhancement eect of Pz-peptide. Pz-peptide elevated the corneal
transport of mannitol, uorescein, and FD4 by 50, 57, and 106%, respectively, but it did not aect the conjunctival transport of mannitol and
uorescein. While Pz-peptide enhanced the ocular absorption of topically
applied hydrophilic atenolol, it did not aect the ocular absorption of
lipophilic propranolol. Interestingly, Pz-peptide did not aect the systemic
absorption of either -adrenergic antagonist. Pz-peptide appeared to facilitate its own penetration across the cornea and the conjunctiva and
increase the ocular absorption of topically applied hydrophilic but not
lipophilic drugs, while not aecting the systemic absorption of either
type of drug.
In addition, the presence of sites beyond the absorbing epithelia that
are capable of degrading peptides and protein and the availability of multiple peptidases in a given site further decrease the absorption potential of
such compounds. While the ocular route has been widely accepted for the
use of topical application, its use in systemic delivery of peptides and proteins will be rather limited.

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

V.

505

SYSTEMIC ADMINISTRATION OF PEPTIDES AND

PROTEINS THROUGH THE OCULAR ROUTE

Systemic absorption of polypeptides and proteins primarily occur through

contact with the conjunctival and nasal mucosae. Table 5 lists some of the
peptides that could be administered through the ocular route (93). Almost
all the studies involving the absorption of peptides and proteins in animal
models have been carried out using labeled peptide samples (9496). Apart
from monitoring the blood concentrations for pharmacokinetic evaluation,
pharmacodynamic studies have also been extensively pursued. Some of the
biological response parameters include reduction in blood sugar by insulin,
increase in blood glucose by glucagon, analgesic eects by enkephalins, and
increase in blood pressure by vasopressin.
Systemic peptide availability following ocular administration has been
related to biological response. The study by Christie and Hanzal (97)
showed that insulin instilled into the conjunctiva is absorbed rapidly, giving
rise to a fairly constant and consistent lowering of blood sugar levels in
rabbits. Another study with somatostatin and its analog revealed that
there was an attenuation of the miotic response to noiceptive stimuli by
these agents, whereas intracameral injection of 150 mg met-enkephalin
had no eect on the miotic response (98).
Lee et al. (99) found that enkephalinamide and inulin are absorbed
into the blood stream following topical ocular administration, the former to
a greater extent than the latter. The authors proposed that depending on the

Table 5 Therapeutically Useful Peptides that Could Be

Administered Through the Ocular Route
Peptide
ACTH
-Endorphin
Calcitonin
Glucagon
Insulin
Leu-enkephalin
Met-enkephalin
Oxytocin
Somatostatin
TRH
Vasopressin
VIP

Application
Antiallergic, decongestant anti-inammatory
Analgesic
Pagets disease, hypercalcemia
Hypoglycemic crisis
Diabetes mellitus
Analgesic
Immunostimulant
Induce uterine contractions
Attenuate miotic responses
Diagnosis of thyroid cancer
Diabetes insipidus
Secretion of insulin

Copyright 2003 Marcel Dekker, Inc.

506

Dey et al.

molecular size, lipophilicity, and susceptibility to proteolysis, other peptides

and proteins may also be absorbed to varying extents. Similarly, Chiou and
Chuang (94) demonstrated the feasibility of eective systemic delivery of
topically instilled peptides in the eye. Their ndings suggest that systemic
delivery of peptide drugs is superior to the parenteral route, especially when
the drug is potent and doses required are low. Enkephalin could eectively
be absorbed systemically through the eye with the use of an absorption
enhancer (95). This ocular route was found to be superior to administering
the peptide by an intravenous route. Similar results have been obtained with
other peptides like thyrotropin-releasing hormone (TRH), luteinizing hormonereleasing hormone (LHRH), glucagon, and insulin (94). Spantide, a
tachykinin antagonist, is readily taken up into the rabbit eye following
topical application. Measurable concentrations of the peptide were observed
in the aqueous humor as well as in the general circulation. Similarly, insulin
could be absorbed eectively into the systemic circulation through ocular
instillation (100). The systemic absorption of 1% insulin through the eyes
can be enhanced at least sevenfold when 1% saponin, a surfactant, was
added to the solution. This absorption enhancement was not aected by
aminopeptidase inhibition. Recently, calcitonin, a polypeptide hormone,
was found to be poorly absorbed into the systemic circulation through the
ocular rote (101). Inclusion of permeation enhancers like Brij-78 and BL-9
markedly improved its systemic absorption.
In summary, small polypeptides such as TRH (MW 300), enkephalins
(MW  600), LHRH (MW 1200), and glucagon (MW 3500) are absorbed to
a signicant extent through the eyes, almost to the extent of 99% (94).
Polypeptides with larger molecular weight such as -endorphin
(MW  5000) and insulin (MW  6000) are also absorbed, but to a much
lesser extent. The absorption of such large molecular weight compounds
can, however, be improved by simultaneous use of absorption enhancers
(78).

VI.

ENHANCED SYSTEMIC ABSORPTION WITH

PERMEATION ENHANCERS

One of the major problems associated with the ocular delivery of peptide
drugs is their poor systemic bioavailability. This may be overcome by using
penetration enhancers. Most permeation enhancers need to be evaluated
with caution, since most of these agents cause local irritation to the eye.
Among them the most eective are Brij-78 and BL-9, because these compounds have been shown to enhance insulin absorption to a signicant
extent without causing any noticeable irritation (78). Table 6 lists the pene-

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

507

Table 6 Effects of Various Permeation Enhancers on Systemic

Absorption of Insulin Following Ocular Administration

Permeation enhancer
Saponin
Fusidic acid

Polyethylene-9-laurylether (BL-9)

Polyethylene-20-stearylether (Brij 78)

Polyethylene-20-oleorylether (Brij 99)

Concentration
of enhancer
(%)

Insulin absorption
enhanced
()

0.5
1.0
0.25
0.5
1.0
2.0
0.25
0.5
1.0
2.0
0.5
1.0
0.5

4.0
7.0
2.3
2.7
3.9
7.5
2.6
4.5
6.0
7.6
6.8
6.3
4.0

tration enhancers that have been examined for enhancing insulin absorption
by the ocular route and their relative performance (93). Saponin, fusidic
acid, and Brij-99 possess high irritation potential and therefore cannot be
used. As indicated earlier, the same surfactants are also capable of enhancing the absorption of calcitonin (101).
The permeability barrier of the corner to hydrophilic molecules is
presumed to reside in the epithelial layer. The presence of tight junctions
render the epithelium almost impermeable to all but the smallest molecules.
Grass and Robinson have shown that the aqueous channel of the cornea has
a cut-o of around 90 D (102). Other studies have suggested a close relationship between epithelial permeability (or tight junction integrity) and the cell
cytoskeleton (103,104). Treatment with cytochalasins or removal of extracellular calcium ions cause opening of tight junctions and increase tight
junction permeability (105,106). Cytoskeletal modulators have been
explored as corneal penetration enhancers (107,108). The ecacy of
EDTA, bile salts, and cytochalasin B (109) in enhancing the ocular permeability of hydrophilic compounds has been examined. Cytochalasin B has
the ability to increase corneal permeability with minimum membrane
damage, indicating its potential as an ocular penetration enhancer.
Further development of novel penetration enhancers with highly specic
foci of action, good reversibility, and minimal toxicity is needed for any
realistic delivery of peptides and proteins by the ocular route.

Copyright 2003 Marcel Dekker, Inc.

508

VII.

Dey et al.

CONCLUSIONS

With breakthroughs in biotechnology, newer and more potent peptide and

protein drugs are emerging in the market. The majority of these polypeptides require special delivery systems. However, since most of these compounds are very potent, require low doses, and are well absorbed from the
mucous membrane, their delivery via the ocular route may be viable.
However, one of the principal problems in the ocular delivery of peptide
and protein drugs is that of relatively low bioavailability to the ocular
tissues. This problem may be circumvented by the use of penetration enhancers. The conjunctival administration of this class of compounds to achieve
therapeutic levels in the systemic circulation may well be possible in the near
future. We hope that novel drug delivery systems will be developed to
deliver potent polypeptide drugs through the ocular route.

REFERENCES
1. Lee, V. H. L. (1987). Ophthalmic delivery of peptides and proteins, Pharm.
Tech., 11:26.
2. Bristow, A. F. (1991).
3. Akerlund, M., Stromberg, P., Forsling, M. L., Melin, P., and Vilhardt, H.
(1983). Inhibition of vasopressin eects on the uterus by a synthetic analogue,
Obstet. Gynecol., 62:309.
4. Akerlund, M., Kostrzewska, A., Laudanski, T., Melin, P., and Vilhardt, H.
(1983). Vasopressin eects on isolated non-pregnant myometrium and uterine
arteries and their inhibition by deamino-ethyl-lysine-vasopressin and deamino-ethyl-oxytocin, Br. J. Obstet. Gynaecol., 90:732.
5. Vilhardt, H., and Bie, P. (1983). Antidiuretic response in conscious dogs
following peroral administration of arginine vasopressin and its analogues,
Eur. J. Pharmacol., 93:201.
6. Tobey, N., Heizer, W., Yeh, R., Huang, T. I., and Honer, C. (1985). Human
intestinal brush border peptidases, Gastroenterology, 88:913.
7. Ziv, E., Lior, O., and Kidron, M. (1987). Absorption of protein via the intestinal wall, Biochem. Pharmacol., 36:1035.
8. Aungst, B. J., Rogers, N. J., and Shefter, E. (1988). Comparison of nasal,
rectal, buccal, sublingual and intramuscular insulin ecacy and the eects of
bile salt absorption promoter, J. Pharmacol. Exp. Ther., 244:23.
9. Aungst, B. J., and Rogers, N. J. (1988). Site dependence of absorption promoting actions of laureth-9, Na salicylate, Na2 EDTA, and apoprotinin on
rectal, nasal and buccal insulin delivery, Pharm. Res., 5:305.
10. Moore, J. A., Pletcher, S. A., and Ross, M. J. (1986). Absorption enhancement of growth hormone from the gastrointestinal tract of rats, Int. J. Pharm.,
34:35.

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

509

11. Anders, R., Merkle, H. P., Schurr, W., and Ziegler, R. (1983). Buccal absorption of protirelin: An eective way to stimulate thyrotropin and prolactin, J.
Pharm. Sci., 72:1481.
12. Salzman, R., Manson, J. E., Gring, G. T., Kimmerle, R., and Ruderman, N.
(1985). Intranasal aerosolized insulin: Mixed meal studies and long term use in
Type I diabetes, N. Engl. J. Med., 312:1078.
13. Moses, A. C., Gordon, G. S., Carey, M. C., and Flier, J. S. (1983). Insulin
administration intranasally as an insulin-bile salt aerosol, eectiveness and
reproducibility in normal and diabetic subjects, Diabetes, 32:1040.
14. Hirai, S., Ikenaga, T., and Matsuwaza, T. (1978). Nasal absorption of insulin
in dogs, Diabetes, 27:296.
15. Siddiqui, O., Sun, Y., Liu, J. C., and Chien, Y. W. (1987). Facilitated transdermal transport of insulin, J. Pharm. Sci., 76:341.
16. Kari, B. (1986). Control of blood glucose levels in alloxan-diabetic rabbits by
iontophoresis of insulin, Diabetes, 35:217.
17. Wigley, F. M., Londono, J.H., Wood, S. H., and Shipp, J. C. (1971). Insulin
across respiratory mucosae by aerosol delivery, Diabetes, 20:522.
18. Yamasaki, Y., Shichiri, M., Kawamori, R., Kikuchi, M., and Yagi, T. (1981).
The eectiveness of rectal administration of insulin suppository on normal
and diabetic subjects, Diabetes Care, 4:454.
19. Fisher, N. F. (1923). The absorption of insulin from the intestine, vagina and
scrotal sac, Am. J. Physiol, 67:65.
20. BenEzra, D., Maftzir, G., de Courten, C., and Timonen, P. (1990). Ocular
penetration of cyclosporin A. III: The human eye, Br. J. Ophthalmol., 74:350.
21. Fraunfelder, F. T., and Meyers, S. M. (1987). Systemic side eects from
ophthalmic timolol and their prevention, J. Ocul. Pharmacol., 3:177.
22. Robinson, J. R. (1989). Ocular drug delivery. Mechanism(s) of corneal drug
transport and mucoadhesive systems, S.T.P. Pharma., 5:839.
23. Rojanasakul, Y., Paddock, S. W., and Robinson, J. R. (1990). Confocal laser
scanning microscopic examination of transport pathways and barriers of some
peptides across the cornea, Int. J. Pharm., 61:163.
24. Harris, D., and Robinson, J. R. (1990). Bioadhesive polymers in peptide drug
delivery, Biomaterials, 11:652.
25. Green, P., Hinz, R., Cullander, C., Yamane, G., and Guy, R. H. (1989).
Iontophoretic delivery of amino acids and analogs, Pharm. Res., 6:S148.
26. Miller, L., Kolaskie, C. J., Smith, G. A., and Riviere, J. (1990). Transdermal
iontophoresis of gonadotrophin releasing hormone (LHRH) and two analogues, J. Pharm. Sci., 79:490.
27. Sun, Y., Xue, H., and Liu, J. C. (1990). A unique iontophoresis system
designed for transdermal protein drug delivery, Pharm. Res., 7:S113.
28. Chien, Y. W., Lelawong, P., Siddiqui, O., Sun, Y., and Shi, W. M. (1990).
Facilitated transdermal delivery of therapeutic peptides and proteins by iontophoretic delivery devices, J. Controlled Rel., 7:1.
29. Dill, K. A. (1990). Dominant forces in protein folding, Biochemistry, 29:7133.
30. Creighton, T. E. (1990). Protein folding, Biochem. J., 270:1.

Copyright 2003 Marcel Dekker, Inc.

510

Dey et al.

31. Horbett, T. A., and Brash, J. L. (1987). Proteins at interfaces: Current issues
and future prospects. In: Proteins at Interfaces: Physicochemical and
Biochemical Studies. T. A. Horbett and J. L. Brash, (eds.). American
Chemical Society, Washington, DC, Chap. 1.
32. Okumura, K., Kiyohara, Y., Komade, F., Mishima, Y., and Fuwa, T. (1990).
Protease inhibitor potentiates the healing eect of epidermal growth factor in
wounded or burned skin, J. Controlled Rel., 13:310.
33. Katre, N. V., Knauf, M. J., and Laird, W. J. (1987). Chemical modication of
recombinant interleukin 2 by polyethylene glycol increase its potency in the
murine Meth A sarcoma model, Proc. Natl. Acad. Sci. USA, 84:1487.
34. Yoshihiro, I., Casolaro, M., Kono, K., and Imanishi, Y. (1989). An insulin
releasing system that is responsible to glucose, J. Controlled Rel., 10:195.
35. Hori, T., Komada, F., Iwakawa, S., Seino, Y., and Okumura, K. (1989).
Enhanced bioavailability of subcutaneously injected insulin coadministered
with collagen in rats and humans, Pharm. Res., 6:813.
36. Fuertges, F., and Abuchowski, A. (1990). The clinical ecacy of poly(ethylene
glycol)-modied proteins. J. Controlled Rel., 11:139.
37. Saran, M., Kumar, G. S., Neckers, D. C., Pena, J., Jones, R. H., and Field, J.
(1990). Biodegradable copolymer coating for oral delivery of peptide drugs,
Biochem. Soc. Trans., 18:752.
38. Lee, V. H. L. (1990). Protease inhibitors and penetration enhancers as
approaches to modify peptide absorption, J. Controlled Rel., 13:213.
39. Lee, V. H. L., and Yamamoto, A. (1990). Penetration and enzymatic barriers
to peptide and protein absorption, Adv. Drug. Deliv. Rev., 4:171.
40. De Boer, A. G., Van Hoogdalem, E. J., Heijligers-Feigen, C. D., Verhoef, J.
C., and Breimer, D. D. (1990). Rectal absorption enhancement of peptide
drugs, J. Controlled Rel., 13:241.
41. Pontiroli, A. E. (1990). Intranasal administration of calcitonin and of other
peptides: Studies with dierent promoters, J. Controlled Rel., 13:247.
42. Lundin, S., and Atursson, P. (1990). Absorption of vasopressin analogue, 1desamino-8-D-arginine-vasopressin (dDAVP), in human intestinal epithelial
cell line, CaCO-2, Int. J. Pharm., 64:181.
43. Amidon, G. L., Sinko, P. J., Hu, M., and Leesman, G. D. (1989). Absorption
of dicult drug molecules: Carrier mediated transport of peptides and peptide
analogues. In: Novel Drug Delivery and Therapeutic Application. L. F. Presscot
and W. S. Ninmo (eds.). Wiley, New York, Chap. 5.
44. Sinko, P. J., Hu, M., and Amidon, G. L. (1987). Carrier mediated transport of
amino acids, small peptides and their drug analogs, J. Controlled Ref., 6:115.
45. Hu, M., Subramaniam, P., Mosberg, H. I., and Amidon, G. L. (1989). Use of
the peptide carrier system to improve the intestinal absorption of L-
-methyldopa: Carrier kinetics, intestinal permeabilities and in vitro hydrolysis of
dipeptidyl derivatives of L-
-methyldopa, Pharma. Res., 6:66.
46. Friedman, D. I., and Amidon, G. L. (1990). Characterization of the intestinal
transport parameters for small peptide drugs, J. Controlled Rel., 13:141.

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

511

47. Ungell, A., and Andreasson, A. (1990). The eect of enzymatic inhibition
versus increased paracellular transport of vasopressin peptides, J. Controlled
Rel., 13:313.
48. Drewe, J., Vonderscher, J., Hornung, K., Munzer, J., Reinhardt, J., Kissel, T.,
and Beglinger, C. (1990). Enhancement of oral absorption of somatostatin
analog Sandostatin in man. J. Controlled Rel., 13:315.
49. Lundin, S., Pantzar, N., Hedin, I., and Westron, B. R. (1990). Intestinal
absorption by sodium taurodihydrofusidate of a peptide hormone analogue
(dDAVP) and a macromolecule (BSA) in vitro and in vivo, Int. J. Pharm.,
59:263.
50. Schilling, R. J., and Mitra, A. K. (1990). Intestinal mucosal transport of
insulin, Int. J. Pharm., 62:53.
51. Su, K. S. E. (1991). Nasal route delivery of peptide and protein drug delivery.
In: Peptide and Protein Drug Delivery. V. H. L. Lee (ed.). Marcel Dekker, New
York, Chap. 13.
52. Harris, A. S. (1986). Biopharmaceutical aspects of the intranasal administration of peptides. In: Delivery Systems for Peptide Drugs. S. S. David, I. Illum,
and E. Tomlinson (eds.). Plenum Press, New York, p. 191.
53. Sandow, J., and Petri, W. (1985). Intranasal administration of peptides, biological activity and therapeutic ecacy. In: Transnasal Systemic Medications.
Y. W. Chien (ed.). Elsevier, New York, Chap. 7.
54. Solbach, H. G., and Wiegelmann, W. (1973). Intranasal application of luteinizing hormone releasing hormone, Lancet, 1:1259.
55. Dashe, A. M., Kleeman, C. R., Czarczkes, J. W., Rubino, H., and Spears, I.
(1964). Synthetic vasopressin nasal spray in the treatment of diabetes insipidus. JAMA, 190:113.
56. Flier, J. S., Moses, A. C., Gordon, G. S., and Silver, R. S. (1985). Intranasal
administration of insulin ecacy and mechanism. In: Transnasal Systemic
Medications. Y. W. Chien (ed.). Elsevier, New York, Chap. 9.
57. OHagan, D. T., Critchley, H., Farraj, N. F., Fisher, A. N., Hohansen, B. R.,
David, S. S., and Illum, L. (1990). Nasal absorption enhancers for synthetic
human growth hormone in rats, Pharm. Res., 7:772.
58. Tengamnuay, P., and Mitra, A. K. (1990). Bile salt-fatty acid mixed micelles
as nasal absorption promoters of peptides. I. Eects of ionic strength, adjuvant composition and lipid structure on the nasal absorption of [DArg2 ]kyotrophin, Pharm. Res., 7:127.
59. Tengamnuay, P., and Mitra, A. K. (1990). Bile salt-fatty acid mixed micelles
as nasal absorption promoters of peptides. II. In vivo nasal absorption of
insulin in rats and eects of mixed micelles on the morphological integrity
of the nasal mucosa, Pharm. Res., 7:370.
60. Vadnere, M., Adjei, A., Doyle, R., and Johnson, E. (1990). Evaluation of
alternative routes for delivery of leuprolide, J. Controlled Rel., 13:322.
61. Merkle, H. P., Anders, R., Sandow, J., and Schurr, W. (1985). Self adhesive
patches for buccal delivery of peptides, Proc. Int. Symp. Controlled Rel.
Bioact. Mater., 12:85.

Copyright 2003 Marcel Dekker, Inc.

512

Dey et al.

62. Squier, C. A., and Hall, B. K. (1985). The permeability of skin and oral
mucosa to water and horseradish peroxidase as related to the thickness of
the permeability barrier, J. Invest. Dermatol., 84:176.
63. Yokosuka, T., Omori, Y., Hirata, Y., and Hirai, S. (1977). Nasal and sublingual administration of insulin in man, J. Jpn. Diabetic Soc., 20:146.
64. Tolo, K., and Jonsen, J. (1975). In vitro penetration of tritriated dextrans
through rabbit oral mucosa, Arch. Oral Biol., 20:419.
65. Gandhi, R. B., and Robinson, J. R. (1990). Mechanism of transport of
charged compounds across rabbit buccal mucosa, Pharm. Res., 7:S116.
66. Dodda-Kashi, S., and Lee, V. H. L. (1986). Enkephalin hydrolysis in homogenates of various absorptive mucosae of the albino rabbit: Similarities in rate
and involvement of aminopeptidases, Life Sci., 38:2019.
67. Garren, K. W., and Repta, A. J. (1988). Buccal absorption. III. Simultaneous
diusion and metabolism of an aminopeptidase substrate in the hamster cheek
pouch, Pharm. Res., 6:966.
68. Aungst, B. J., and Rogers, N. J. (1989). Comparison of the eects of various
transmucosal absorption promoters on buccal insulin delivery, Int. J. Pharm.,
53:277.
69. Crooks, P. (1990). Lung peptidases and their activities. Presented at
Respiratory Drug Delivery II, Keystone, Colorado.
70. Schankar, L. S., Mitchel, E. W., and Brown, R. A. (1986). Species comparison
of drug absorption from the lung after aerosol inhalation or intratracheal
injection, Drug Metab. Dispos., 14:79.
71. Maruyama, K., Homberg, E., Kennel, S. J., Klibanov, A., Forchlin, V. P., and
Huang, L. (1990). Characterization of in vivo immunoliposome targeting to
pulmonary endothelium, J. Pharm. Sci., 79:978.
72. ODonnell, M. (1990). Novel approaches to the development of peptide
bronchodilator drugs. Presented at Respiratory Drug Delivery II, Keystone,
Colorado.
73. Lee, V. H. L. (1990). Mechanisms and facilitations of corneal drug penetration, J. Controlled Rel., 11:79.
74. Lee, V. H. L., Carson, L. W., Dodda-Kashi, S., and Stratford, R. E. (1986).
Metabolic permeation barriers to the ocular absorption of topically applied
enkephalins in albino rabbits, J. Ocul. Physiol., 2:345.
75. Lee, V. H. L., Carson, L. W., and Takemoto, K. A. (1986). Macromolecular
drug absorption in the albino rabbit eye, Int. J. Pharm., 29:43.
76. Newton, C., Gebhardt, B. M., and Kaufman, H. E. (1988). Topically applied
cyclosporine in zone prolongs corneal allograft survival, Invest. Ophthalmol.
Vis. Sci., 29:208.
77. Morimoto, K., Nakai, T., and Morisaka, K. (1987). Evaluation of permeability enhancement of hydrophilic compounds and macromolecular compounds
by bile salts through rabbit corneas in vitro, J. Pharm. Pharmacol., 39:124.
78. Chiou, G. C., and Ching, Y. C. (1989). Improvement of systemic absorption
of insulin through the eyes with absorption enhancers, J. Pharm. Sci., 78:815.

Copyright 2003 Marcel Dekker, Inc.

Peptides and Proteins as Therapeutic Agents

513

79. Yamamoto, A., Luo, A. M., Dodda-Kashi, S., and Lee, V. H. L. (1989). The
ocular route for the systemic insulin delivery in the albino rabbit, J. Pharm.
Exp. Ther., 249:249.
80. Ahmed, I., and Patton, T. F. (1985). Importance of the noncorneal absorption
route in topical ophthalmic drug delivery, Invest. Ophthalmol. Vis. Sci.,
26:584.
81. Pandol, M., Astedt, B., and Dyster-Aas, K. (1972). Release of brinolytic
enzymes from the human cornea, Acta. Ophthalmol., 50:199.
82. Berman, M., Manseau, E., Law, M., and Aiken, D. (1983). Ulceration is
correlated with degradation of brin and bronectin at the corneal surface,
Invest. Ophthalmol. Vis. Sci., 24:1358.
83. Hayasaka, S., and Hayasaka, I. (1979). Cathepsin B and collagenolytic cathepsin in the aqueous humor of patients with Bechets disease, Albrecht v.
Graefes Arch. Klin. Exp. Ophthal., 206:103.
84. BenEzra, D., and Maftzir, G. (1990). Ocular penetration of cyclosporin A.
The rabbit eye, Invest. Ophthalmol. Vis. Sci., 31:1362.
85. BenEzra, D., and Maftzir, G. (1990). Ocular penetration of cyclosporine A in
the rat eye, Arch. Ophthalmol., 108:584.
86. Hoppenreijs, V. P., Pels, E., Vrensen, G. F., Oosting, J., and Treers, W. F.
(1992). Eects of human epidermal growth factor on endothelial wound healing of human corneas, Invest. Ophthalmol. Vis. Sci., 33:1946.
87. Pasquale, L. R., Dorman-Pease, M. E., Lutty, G. A., Quigley, H. A., and
Jampel, H. D. (1993). Immunolocalization of TGF-beta 1, TGF-beta 2, and
TGF-beta 3 in the anterior segment of the human eye, Invest. Ophthalmol. Vis.
Sci., 34:23.
88. Grant, M. B., Caballero, S., and Millard, W. J. (1993). Inhibition of IGF-I
and b-FGF stimulated growth of human retinal endothelial cells by the somatostatin analogue, octreotide: a potential treatment for ocular neovascularization, Regul. Pept., 48:267.
89. Grant, M. B., Mames, R. N., Fitzgerald, C., Ellis, E. A., Caballero, S.,
Chegini, N., and Guy, J. (1993). Insulin-like growth factor I as an angiogenic
agent. In vivo and in vitro studies, Ann. NY Acad. Sci., 692:230.
90. Grant, M. B., Mames, R. N., Fitzgerald, C., Ellis, E. A., Aboufriekha, M.,
and Guy, J. (1993). Insulin-like growth factor I acts as an angiogenic agent in
rabbit cornea and retina: Comparative studies with basic broblast growth
factor, Diabetologia, 36:282.
91. Hamalainen, K. M., Ranta, V. P., Auriola, S., and Urtti, A. (2000). Enzymatic
and permeation barrier of [D-Ala(2)]-Met-enkephalinamide in the anterior
membranes of the albino rabbit eye, Eur. J. Pharm. Sci., 9:265.
92. Chung, Y. B., Han, K., Nishiura, A., and Lee, V. H. (1998). Ocular absorption of Pz-peptide and its eect on the ocular and systemic pharmacokinetics
of topically applied drugs in the rabbit, Pharm. Res., 15:1882.
93. Chiou, G. C. (1991). Systemic delivery of polypeptide drugs through ocular
route, Ann. Rev. Pharmacol. Toxicol., 31:457.

Copyright 2003 Marcel Dekker, Inc.

514

Dey et al.

94. Chiou, G. C., and Chuang, C. Y. (1988). Systemic delivery of polypeptides

with molecular weights between 300 and 3500 through the eye, J. Ocul.
Pharmacol., 4:165.
95. Chiou, G. C., Chuang, C. Y., and Chang, M. S. (1988). Systemic delivery of
enkephalin peptide through eyes, Life Sci., 43:509.
96. Chiou, G. C., and Chuang, C. Y. (1988). Treatment of hypoglycemia with
glucagon eye drops, J. Ocul. Pharmacol., 4:179.
97. Christie, C. D., and Hanzal, R. F. (1931). Insulin absorptin by the conjunctival membranes in rabbits, J. Clin. Invest., 10:787.
98. Stjernschantz, J., Sears, M. L., and Oksala, O. (1985). Eects of somatostatin,
a somatostatin analog, neurotensin, and metenkephalin in the eye with special
reference to the irritative response, J. Ocul. Pharmacol., 1:59.
99. Lee, V. H. L., Carson, L. W., Dodda-Kashi, S. D., and Stratford, R. E. Jr.
(1988). Systemic absorption of ocularly administered enkephalinamide and
inulin in the albino rabbit: Extent, pathways and vehicle eects, J. Pharm.
Sci., 77:838.
100. Chiou, G. C., Chuang, C. Y., and Chang, M. S. (1989). Systemic delivery of
insulin through eyes to lower the glucose concentration, J. Ocul. Pharmacol.,
5:81.
101. Li, B. H. P., and Chiou, G. C. Y. (1992). Systemic administration of calcitonin
through ocular route, Life Sci., 50:349.
102. Grass, G. M., and Robinson, J. R. (1988). Mechanisms of corneal drug penetration I: In vivo and in vitro kinetics, J. Pharm. Sci. 77:3.
103. Meza, I., Ibarra, G., Sabanero, M., and Cereijido, M. (1980). Occluding
junctions and cytoskeletal components in cultures transporting epithelium,
J. Cell Biol., 87:746.
104. Madara, J. L., Barenberg, D., and Carlson, S. (1986). Eects of cytochalasin
D on occluding junctions of intestinal absorptive cells. Further evidence that
the cytoskeleton may inuence paracellular permeability and junctional
charge selectivity, J. Cell Biol., 102:2125.
105. Bentzel, C. J., Hainau, B., Ho, S., Hui, S. W., Edelman, A., Anagnostopoulos,
T., and Benedetti, E. L. (1980). Cytoplasmin regulation of tight junction
permeability: eect of plant cytokinins, Am. J. Physiol., 239:C75.
106. Martinez-Palomo, A., Meza, I., Beaty, G., and Cereijido, M. (1980).
Experimental modulation of occluding junctions in a cultured epithelium, J.
Cell Biol., 87:736.
107. Aldridge, D. C., Armstrong, J. J., Speake, R. N., and Turner, W. B. (1967).
The cytochalasins, a new class of biologically active mold metabolites, Chem.
Commun., 1:26.
108. Rothweiler, W., and Tamm, C. (1966). Isolation and structure of phomin,
Experientia, 22:750.
109. Binder, M., and Tamm, C. (1973). The cytochalasins: A new class of biologically active microbial metabolites, Agnew. Chem. Int. Edit., 12:370.

Copyright 2003 Marcel Dekker, Inc.

17
Retinal Disease Models for
Development of Drug and Gene
Therapies
Leena Pitkanen
University of Kuopio and Kuopio University Hospital, Kuopio, Finland

Lotta Salminen
University of Tampere and Tampere University Hospital, Tampere,
Finland

Arto Urtti
University of Kuopio, Kuopio, Finland

I.

INTRODUCTION

In humans the retina is the innermost layer of the eye, which consists of
retinal pigment epithelium (RPE) and neural retina. The neural retina has
several layers and various cell types, which are illustrated in Figure 1. RPE is
a single layer of hexagonal cells that maintains the homeostasis of neural
retina. It has essential biochemical, physiological, physical, and optical functions in maintaining the visual system, including phagocytosis of rod outer
segments, transport of substances between photoreceptors and choriocapillaries, and uptake and conversion of the retinoids, which are needed in
visual cycle. Together with endothelial cell linings of retinal capillaries,
RPE forms the blood-retinal barrier. The neural retina is a complicated
and delicate multilayer. The thickness of neural retina varies from 0.4 mm
near the optic nerve to about 0.1 mm anteriorly at the ora serrata. The
photoreceptors are the light-sensing part of retina. The electric impulses
are amplied and integrated by bipolar, horizontal, amacrine, and ganglion
cells. The principal glial cell of the retina is the Muller cell. The bipolar cells
515

Copyright 2003 Marcel Dekker, Inc.

516

Pitkanen et al.

Fig. 1 Light photomicrograph of a rat retinal section. The inner nuclear layer
includes the nuclei of bipolar, amacrine, horizontal, and Muller cells. The nuclei
of rods and cones are in the outer nuclear layer.

are the rst and ganglion cells the second neuron of the visual pathway from
photoreceptors to brain. Macula is the central part of retina located temporally of optic nerve head between the upper and lower temporal vessels.
Fovea is the central, approximately 1.5 mm wide sloping part of macula.
Visual acuity is decreased quickly in the paramacular areas. Of the photoreceptors, the cones take care of photoptic and color vision and are located
mainly in the macula. Rods are the main photoreceptor type in the periphery; they are specialized to scotopic vision.
The human retina may be aected by many vascular diseases such as
occlusions, vasculitis, and anomalies. Retinopathy of prematurity (ROP) is
a retinal disease aecting premature children, where the growth of developing retinal vasculature is interrupted. Diabetic retinopathy is a common
cause of blindness, while arteriosclerosis, hypertension, and other cardiovascular diseases may cause changes in the retinal vasculature.
Neovascularization is also associated in macular degeneration.
In retinal detachment, uid is collected in the potential space between
the neural retina and RPE. In rhegmatogenous detachment, the uid comes

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

517

from the vitreous cavity through a retinal hole or tear. Extravasation may
originate from choroid or retina and results in secondary retinal detachment. Retinal detachment caused by the traction of brous bands in vitreous
is called traction retinal detachment. Traumas, intraocular inammations,
retinal or vitreal degeneration, or vitreal bleeding are etiological factors of
retinal detachment. Proliferative vitreoretinopathy (PVR) is found in about
5% of retinal detachments. It is characterized by the formation of vitreal,
epiretinal, or subretinal membranes after retinal reattachment surgery or
ocular trauma. In some cases the membranes cause traction and distortion
of retina. Severe postoperative PVR is the most common cause of failed
retinal detachment surgery.
Retinoblastoma is a malignant retinal tumor with an incidence of
about 1 : 20,000. The genetic abnormality of this disease located to 13q14.
Both genes in this locus must be abnormal before this malignancy develops.
In the nonhereditary form, mutation occurs only in the retinal cells. In the
hereditary form the patient has inherited the rst mutation from his or her
parents, and 90% of these patients develop a clinical retinoblastoma.
In this chapter we present some recent development in the retinal
disease models of animals. Models of retinal degeneration, proliferative
diseases, and neovascularization are presented. These models are important
tools in current research, since various growth factors, gene therapies, and
transplantation strategies have demonstrated possibilities for treating severe
retinal diseases.

II.

RETINAL DEGENERATIONGENETIC MODELS

Retinal degeneration leads to impaired function of the photoreceptors and

consequently gradual loss of vision., Many types of degeneration are based
on genetic factors, e.g., retinitis pigmentosa, a common term for various
mutations causing retinal degeneration. In addition to genetic factors, environmental factors (e.g., light exposure) may lead to retinal degeneration.
Macular degeneration is the most common type of retinal degeneration,
being the leading cause of vision loss in the industrial world. In the following
sections we present some genetic and environmental animal models of retinal degeneration.
A.

Natural Mutation Mouse Models

The naturally occurring mouse rd (retinal degeneration) and rds (retinal

degeneration slow) photoreceptor dystrophies are recessively inherited.
The mice have defects in the cGMP phosphodiesterase beta subunit gene

Copyright 2003 Marcel Dekker, Inc.

Pitkanen et al.

518

(1,2) and in the peripherin gene (3,4). The rd mouse is a model of retinitis
pigmentosa in which a mutation of a rod-specic photophodiesterase leads
to the rapid loss of photoreceptors during early postnatal life. Very little is
known about the associated changes in the inner retinal neurons. Bipolar
and horizontal cells of the rd mouse retina undergo dramatic morphological
changes accompanying photoreceptor loss, demonstrating a dependence of
second-order neurons on photoreceptors (5).
The rds phenotype is considered to be an appropriate model for peripherin 2/rds-mediated retinitis pigmentosa. Peripherin 2 glycoprotein is
needed for the formation of photoreceptor outer discs. The photoreceptor
cell is the primary site of the genetic defect that results in retinal dystrophy
in the rds mouse model (6).
The protective eect of a number of survival factors on degenerating
photoreceptors in mutant mice with naturally occurring inherited retinal
degenerations, including retinal degeneration (rd/rd), retinal degeneration
slow (rds/rds), nervous (nr/nr), and Purkinje cell degeneration (pcd/pcd), in
three dierent forms of mutant rhodopsin transgenic mice and in light
damage in albino mice were examined by La Vail et al. (7). The slowing
of degeneration in the rd/rd and Q344ter (a naturally occurring stop codon
mutation that removes the last ve amino acids of rhodopsin) mutant mice
demonstrated that intraocularly injected survival factors can protect photoreceptors from degenerating. Importantly, these animal models have the
same or similar genetic defects as those in human inherited retinal degenerations (7). Such models have also been used to improve the condition of
photoreceptors by adeno-associated virus-mediated peripherin 2 gene therapy (8). The outcome of the gene therapy was dependent on the timing of
the therapy (9).
B.

Transgenic Mouse Models

To generate transgenic animals, whole genes are injected into a fertilized egg
pronucleus. The genes associate randomly into the genome, and their
expression is controlled by their own regulatory sequences. Due to the
complexity of the photoreceptor biology, several genes can be used to generate transgenic mouse models of retinal degeneration.
The VPP mouse carries three mutations (P23H, V20G, P27L) near the
N-terminus of opsin, the apoprotein of rhodopsin, the rod photopigment.
These animals have slowly progressive degeneration of the rod photoreceptors and subsequent changes in retinal function. These changes mimic autosomal dominant retinitis pigmentosa of humans, which results from a point
mutation (P23H) in opsin (10). The rate of photoreceptor degeneration in
VPP mice seems to be adversely aected by the existence of the albino

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

519

phenotype (11). Light deprivation aects the rate of degeneration in pigmented transgenic VPP mice (12).
To establish a transgenic mouse line with a mutated mouse opsin gene
in addition to the endogenous opsin gene, a mutated mouse opsin gene was
introduced into the germ line of a normal mouse. Simultaneous expression
of mutated and normal opsin genes induces a slow degeneration of both rod
and cone photoreceptors. The time course mimics the course of human
autosomal dominant retinitis pigmentosa (13).
The biochemical, morphological, and physiological analyses of a
transgenic mouse model for retinal degeneration slow (RDS) retinitis pigmentosa have been carried out. RDS retinitis pigmentosa is caused by a
substitution of proline 216 to leucine (P216L) in rds/peripherin. The phenotype in P216L-transgenic mice probably caused by a combination of two
genetic mechanisms: a dominant eect of the P216 substituted protein and a
reduction in the concentration of normal rds/peripherin. The expression of
the normal and mutant genes is similar to that predicted for humans with
RDS-mediated autosomal-dominant retinitis pigmentosa. These mice may
be used as an animal model for this disease (14).
The W70A transgenic mouse carries a point mutation (W70A) in the
gene that encodes for the gamma-subunit of rod cGMP phosphodiesterase.
This mouse represents a new model of stationary nyctalopia that can be
recognized by its unusual ERG (electroretinogram) features (15).
Another transgenic mouse model with defective expression of the
alpha subunit of the rod cGMP-gated channel was reported recently (16).
Expression was reduced by antisense RNA. The low expression of the rod
cGMP-gated channel causes a disease model that can be used to test therapies designed to slow down or cure retinal degenerations (16).
Mice (Pdegtm1/Pdegtm1) that are homozygous for a mutant allele of
the gamma subunit of retinal cyclic guanosine monophosphate phosphodiesterase (PDE gamma) have a severe photoreceptor degeneration.
Interestingly, the transgene that encodes the BCL2 protein was introduced
by mating into the mutant background. Antiapoptotic transgene BCL2
delayed temporarily the degeneration of photoreceptors in this murine
model of retinal degeneration (17).

C.

Knockout Mouse Models

Knockout mutation is created by transferring a gene that is inactivated by

mutation to pluripotent embryonal stem cells. They often nd their copy in
the genome and settle beside it and then change places by recombination.
The cells with wanted recombination are transferred to blastocysts to pro-

Copyright 2003 Marcel Dekker, Inc.

Pitkanen et al.

520

duce chimeric animals. Homozygous animals with the mutation can be

produced by mating.
A retinitis pigmentosa GTPase regulator-decient mouse model for Xlinked retinitis pigmentosa has been created by gene knockout. In the
mutant mice, cone photoreceptors exhibit ectopic localization of cone
opsins. Rod photoreceptors have a reduced level of rhodopsin, and subsequently photoreceptors degenerate (18).
Likewise, rhodopsin knockout (opsin/) mice have been generated
as an animal model of retinitis pigmentosa. In that case a gene encoding
ciliary neurotrophic factor (CNTF) was delivered subretinally with adenoassociated virus-vector. CNTF gene therapy delayed the death of photoreceptors (19).
Homozygous rhodopsin knockout (Rho/) mice have a mutation in
exon 2 of the rhodopsin gene. They show a complete absence of functional
rhodopsin and do not build rod outer segments. The Rho(/) mice can
serve during postnatal weeks 46 as a model for pure cone function (20).
These mice do not elaborate rod outer segments, and the photoreceptors are
lost in 3 months. No rod ERG response is seen in 8-week-old animals. In
contrast, Rho= animals retain most of their photoreceptors, although the
inner and outer segments of the cells display some structural disorganization. These animals may be a useful genetic background on which other
mutant opsin transgenes can be expressed. (21).
Knockout mice with arrestin gene defect have been generated.
Excessive light accelerated the cell death in pigmented arrestin knockout
mice. Human patients with mutations leading to nonfunctional arrestin
and rhodopsin kinase have Oguchi disease. This disease is a form of stationary night blindness (22).
D.

Rat Models

The Royal College of Surgeons (RCS) rat is the rst animal model with
inherited retinal degeneration. Although the genetic defect is actually not
known, the RCS rat is widely used as a model of photoreceptor degeneration with relevance to retinitis pigmentosa and hereditary retinal dystrophies
(23,24). Experiments with RCS rats have been used to demonstrate the
benecial eects of growth factors (like basic broblast growth factor,
bFGF) on retinal degeneration (25).
Adenovirus-mediated gene transfer has been used to develop a rat
model for photoreceptor degeneration. Recombinant adenovirus-mediated
downregulation of cathepsin S (CatS) in the retinal pigment epithelium and/
or neural retina was achieved. These results demonstrate that the transient
modulation of gene expression in RPE cells induced changes in the retina.

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

521

Despite the low expression of endogenous CatS in RPE cells, this enzyme
appears to play an important role in the maintenance of normal retinal
function (26).
Transgenic rat P23H have been used as a model of autosomal dominant retinitis pigmentosa. Substitution of proline by histidine in position 23
in rhodopsin (P23H) is the most common human mutation in RP in the
United States, with a prevalance of 15%. Several sublines of this strain have
been developed. These lines have a similar genotype, but the rate of retinal
degeneration varies. In line 1, almost complete degeneration is seen in 2
months, but in line 2 similar degeneration develops in one year. Similarly,
there are many sublines of transgenic rats that carry a rhodopsin mutation
S334ter with dierent rates of retinal degeneration. Ribozyme-directed cleavage of mutant mRNAs slows the rate of photoreceptor degeneration in this
rat model (27). d-cis-Diltiazem did not rescue photoreceptors of Pro23His
rhodopsin mutation line 1 rats treated according to the protocol used in rd
mouse (28). Extended photoreceptor viability by light stress has been
detected in RCS rats but not in opsin P23H mutant rats (29). The photoreceptors of transgenic rats expressing either a P23H or an S334ter rhodopsin mutation were protected from apoptosis by recombinant adenoassociated virus-mediated production of broblast growth factors fgf-2,
fgf-5, and fgf-18 (30,31), while lens epithelium-derived growth factor promoted photoreceptor survival in light-damaged and RCS rats, but not in
P23H rats (32).
In addition to biochemical measures, these disease state models can be
monitored on the basis of retinal morphology (number of outer nuclear
layers) and ERG (a and b waves).
E.

Cat Models

Abyssinian cats with recessively inherited rod-cone degeneration have been

introduced (33). Photoreceptor allografts were examined to determine the
viability and inuence of such transplants on the host retina of the cats.
Also, clinical and pathological features, light and electron microscopy, and
the electrophysiology of an autosomal dominant, early-onset feline model of
rod/cone dysplasia (Rdy cats) have been documented (3436). The immunohistochemical changes in the retina and photoreceptor cell death of this
model have also been studied (37).
F.

Transgenic Pig Model

Transgenic pigs that express a mutated rhodopsin gene (Pro347Leu) were

generated (38). These transgenic pigs provide a large animal model to study

Copyright 2003 Marcel Dekker, Inc.

Pitkanen et al.

522

the protracted phase of cone degeneration in retinitis pigmentosa and for

preclinical treatment trials.
G.

Dog Models

Canine rcd1 model of retinitis pigmentosa is caused by a null mutation in

the PDE6B gene. Treatment of rcd1-aected dogs with d-cis-diltiazem did
not modify the photoreceptor disease (39).
Rod-cone dysplasia types 1 (rcd1; Irish setter) and 2 (red2; collie) in
dogs are early-onset forms of progressive retinal atrophy, which serve as
models of retinitis pigmentosa in humans (40).
Swedish Briard dogs have a very slowly progressive retinal dystrophy
that is inherited in an autosomal recessive manner. The lipid and fatty acid
compositions of plasma, retina, and retinal pigment epithelium were analyzed in this model (41). These studies provide evidence for yet another
animal model of inherited retinal degeneration with a defect in retinal polyunsaturated fatty acid metabolism. The fatty acid pattern in aected dogs
resembles that in the retina in n-3 fatty acid deciency.

III.

RETINAL DEGENERATIONLIGHT-INDUCED MODELS

Retinal damage by light has two distinct action spectra. One peaks in the
ultraviolet A (UVA) and the other in the midvisible wavelength. It was
shown in the Long Evans rat that UVA and green light can produce histologically dissimilar types of damage. UVA light in particular produces
severe retinal damage at low irradiation levels (42).
Albino rats were continuously exposed to blue light for 17 days.
Continuous exposure of albino rats to moderate blue light for 25 days
selectively eliminated most of the photoreceptors while leaving the RPE
intact (43).
Monocularly aphakic gray squirrels were exposed for 10 minutes to
monochromatic near-ultraviolet radiation to determine if their yellow pigmented lens protected retinal tissue from photochemical damage. In aphakic
eyes the retinas revealed irreversible lesions to the photoreceptors. Eyes
exposed to ultraviolet radiation with their lenses intact were devoid of signicant retinal lesions. This study represents a model system for studying
the potential damaging eects of near-UV radiation to the aphakic eyes of
humans (44).
Constant uorescent light can also be used to generate light-induced
degeneration model. Albino rats of the F344 strain were exposed to 1 or 2
weeks of constant light, either with or without intravitreal or subretinal

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

523

bFGF solution injected 2 days before the start of light exposure. Constant
light exposure causes a decrease in the thickness of the outer nuclear layer
and blocks ERG responses. The results indicated that the photoreceptor
rescue activity of bFGF is not restricted to inherited retinal dystrophy in
the rat. The light damage is an excellent model for studying the normal
function of bFGF and its survival-promoting activity (45). It has been
shown that, in the retina, basic broblast growth factor delays photoreceptor degeneration in Royal College of Surgeons rats with inherited retinal
dystrophy. bFGF also reduces or prevents the rapid photoreceptor degeneration produced by constant light in the rat. This light-damage model was
used to assess the survival-promoting activity in vivo of a number of growth
factors and other molecules. Photoreceptors can be signicantly protected
from the damaging eects of light by intravitreal injection of eight dierent
growth factors, cytokines, and neurotrophins. They act through several
distinct receptor families. In addition to basic broblast growth factor,
eective photoreceptor rescue was obtained with brain-derived neurotrophic
factor, ciliary neurotrophic factor, interleukin 1 beta, and acidic broblast
growth factor. Less activity was seen with neurotrophin 3, insulin-like
growth factor II, and tumor necrosis factor alpha, while nerve growth factor, epidermal growth factor, platelet-derived growth factor, insulin, insulinlike growth factor I, heparin, and laminin did not show any protection (25).

IV.

PROLIFERATIVE VITREORETINOPATHY

Proliferative vitreoretinopathy (PVR) is found in about 5% of retinal

detachments. The cellular evens of PVR include migration of glial cells,
pigment epithelial cells, and brocytes into the vitreous cavity, where they
proliferate and transform and dedierentiate. The cells may interact with
endogenous membranous components of the vitreous. This leads to the
formation of vitreal, epiretinal, and subretinal membranes and traction retinal detachment (47). Severe postoperative PVR is the most common cause
of failed retinal detachment surgery. Animal models of PVR are based on
environmental injuries.
A.

Cell Injection

Injury can be caused by intraocular injection of broblasts into rabbit eye.

This was used as a model to test treatment of PVR by gene therapy. A
classication of severity of PVR in this model has been published (4850).
The extent of PVR by cells that do or do not express the receptors for
platelet-derived growth factor (PDGF) was investigated. Mouse embryo

Copyright 2003 Marcel Dekker, Inc.

Pitkanen et al.

524

broblasts was derived from PDGF receptor knock-out embryos. They do

not express PDGF receptors and induced PVR poorly when injected into
the eyes of rabbits. PDGF made an important contribution to the development of PVR in this animal model. Furthermore, there was a marked difference between the two receptors from PDGF. PDGF
receptor was
capable of inducing PVR (51).
B.

Dispase

PVR can be induced by injecting dispase intravitreally to rabbits (Dutch

belted, New Zealand white). Proliferative vitreoretinopathy developed in
response to subretinal or intravitreal dispase, with or without retinal
break. Severity of PVR was correlated with increasing doses of dispase.
The dispase model of PVR is easy to perform, and it permits a clear view
of the retina. This model showed a high success rate in development of PVR
(52), and intravitreally administered prinomastat decreased development of
PVR in this experimental model (53).
C.

Combined Models

A proliferative vitreoretinopathy model was generated in albino rabbits by

combing some factors that probably cause the disease. The eyes were
injected with platelet-rich plasma, and in addition they underwent cryotherapy or vitrectomy or both procedures. Total retinal detachment and giant
holes were obtained more often in experimental eyes than in controls.
Microscopic investigation showed intravitreal or preretinal proliferation
of broblast-like cells (54).
Another combination model involves retinotomy with removal of vitreous, cryotherapy, and platelet-rich plasm injection. This is an ecient
model of PVR: retinal detachments were produced in 100% of rabbit eyes
(55).
In a similar model (56) combined therapy of systemic methylprednisolone, sodium diclofenac, and colchicine was combined with topical atropine, adrenaline, and dexamethasone phosphate. The therapies were useful
in treating experimental PVR.
D.

Laser

Laser-induced retinal injury can be used to provoke PVR formation. For

example, pigmented rabbits underwent argon laser panretinal photocoagulation in one eye. Then cultured broblasts were implanted into the intact

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

525

vitreous of both eyes. More severe PVR developed in the eye with prior
panretinal photocoagulation than in the controls (57).
E.

Other Models

Platelet-rich plasma causes PVR after injection into the vitreous in rabbit
eyes (59). It contributed more eectively to the development of an experimental porcine PVR than PDGF. The ecacy depends on the platelet concentration of the plasma. It seems that other growth factors and plasma
components may interact synergistically with PDGF in the pathogenesis of
PVR (58).

V.

NEOVASCULARIZATION

Neovascularization is involved in various diseases (e.g., cancer, psoriasis),

and the mechanism of angiogenesis has intensely been studied.
Neovascularization complicates the treatment of many retinal diseases,
and, therefore, appropriate animal models are needed.
A.

Laser-Induced Neovascularization

Subretinal neovascularization (NV) can be induced by intense laser photocoagulation in monkey eyes (60). In pigs, the laser-induced branch of retinal
venous obstruction with rose bengal develops neovascularization of the
optic nerve head and retina (612). This process was assisted by photodynamic thrombosis. A model of retinal ischemia and associated NV established by venous thrombosis was produced. After anesthesia, eyes of
pigmented rats received an intraperitoneal injection of sodium uorescein
prior to laser treatment. With a blue-green argon laser, selected venous sites
next to the optic nerve head were photocoagulated (64).
B.

Angiogenic Factor

Several ocular NV models are based on exposing the retina to excess of

angiogenic compounds. The eect of increased vascular endothelial growth
factor (VEGF) expression in the retina was investigated using transgenic
mice in which bovine rhodopsin promoter is coupled with the gene for
human VEGF. This study demonstrated that overexpression of VEGF in
the retina was sucient to cause intraretinal and subretinal NV and provided a valuable new animal model (62).

Copyright 2003 Marcel Dekker, Inc.

Pitkanen et al.

526

Controlled-release systems have been developed in order to provide a

long-term supply of angiogenic factors to the retina at dened levels.
Ethylenevinyl acetate copolymer pellets release VEGF slowly into the vitreous cavity of rabbits and primates. This induces neovascularization.
Sustained intravitreal release of VEGF caused widespread retinal vascular
dilation and breakdown of the blood-retinal barrier. Retinal NV seems to
require persistent high levels of VEGF at the retinal surface. This can be
achieved in rabbits more easily than in primates (63).
In alternative controlled-release models, subretinal implantation of
bFGF-impregnated gelatin microspheres is used to induce subretinal neovascularization in the rabbit (65).
C.

Ischemia

Relative hypoxia triggers formation of neovessels in the retina. When oneweek-old C57BL/6J mice were exposed to 75% oxygen for 5 days and then
to room air, the retinal neovascularization occurs between postnatal days 17
and 21. This model can then be used to study the therapeutic strategies (66).
In a similar ischemia-induced ocular neovascularization model, the expression of Flk-1 and neuropilin-1 was restricted on neovascularized vessels,
suggesting that these molecules may play important roles in retinal NV (67).
NV studies with ischemic models suggest that PaO2 uctuation is more
important than extended hyperoxia for retinal neovascular response in rats
(68). Indeed, a cycled hypoxia/hyperoxia (1050% O2 ) protocol followed by
normoxia (20% O2 ) has been used as a retinal model of retinopathy of
prematurity to induce neovascularization in rat pups (69,70).
The time course and degree of proliferative vascular response after
hyperoxic insult were examined in dogs after oxygen-induced retinopathy.
In the neonatal dog, revascularization after hyperoxic insult involves a period of marked vasoproliferation peaking 310 days after a return to room
air. Oxygen-induced changes in the extravascular milieu probably aect the
pattern of reforming vasculature and possibly restrict the growth anteriorly
(71).
D.

Genetic Models

NV of the RPE occurs earlier in a line of P23H mutant rhodopsin transgenic

mice than in most other mice and rats. The temporal course of RPE NV in
P23H mice was compared with that of two other retinal degeneration
mutants with a similar time course of photoreceptor cell loss. The ndings
suggest that the P23H mutant rhodopsin transgenic mouse may be a useful
model for studying the regulation of NV in the outer retina (72).

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

E.

527

Other Models

NH4Cl gavage in the neonatal rat produced a metabolic acidosisinduced

retinopathy that may be a model for retinopathy of prematurity. Acidosis is
induced by high-dose acetazolamide. Independently of hyperoxemia or
hypoxemia, the treatment is associated with preretinal neovascularization
in the neonatal rat (73).
A consistent model of preretinal NV in the rabbit was developed by
partially digesting the posterior virtreous with repeated injection of hyaluronidase. Then 250,000 homologous dermal broblasts were injected intravitreally (74). Neovascular events observed in this model agree with those
previously described for diabetic retinopathy and retinopathy of prematurity in humans (75).

REFERENCES
1. Bowes, C., Li, T., Danciger, M., Bacter, L. C., Applebury, M. L., and Farber,
D. B. (1990). Retinal degeneration in the rd mouse is caused by a defect in the
beta subunit of rod cGMP-phosphodiesterase, Nature, 347:677.
2. Pittler, S. J., and Baehr, W. (1991). Identication of a nonsense mutation in
the rod photoreceptor cGMP phosphodiesterase beta-subunit gene of the rd
mouse, Proc. Natl. Acad. Sci., 88:8322.
3. Travis, G. H., Sutclie, J. G., and Bok, D. (1991). The retinal degeneration
slow (rds) gene product is a photoreceptor disc membrane-associated glycoprotein, Neuron, 6:61.
4. Connell, G., Bascom, R., Molday, L., Reid, D., McInnes, R. R., and Molday,
R. S. (1991). Photoreceptor peripherin is the normal product of the gene
responsible for retinal degeneration in the rds mouse, Proc. Natl. Acad. Sci.,
88:723.
5. Strettoi, E., and Pignatelli, V. (2000). Modications of retinal neurons in a
mouse model of retinitis pigmentosa, Proc. Natl. Acad. Sci., 97:11020.
6. Li, L., Sheedlo, H. J., and Turner, J. E. (1993). Retinal pigment epithelial cell
transplants in retinal degeneration slow mice do not rescue photoreceptor
cells, Invest. Ophthalmol. Vis. Sci., 34:2141.
7. LaVail, M. M., Yasumura, D., Matthes, M. T., Lau-Villacorta, C., Unoki, K.,
Sung, C. H., and Steinberg, R. H. (1998). Protection of mouse photoreceptors
by survival factors in retinal degenerations, Invest. Ophthalmol. Vis. Sci.,
39:592.
8. Ali, R. R., Sarra, G. M., Stephens, C., Alwis, M. D., Bainbridge, J. W.,
Munro, P. M., Fauser, S., Reichel, M. B., Kinnon, C., Hunt, D. M.,
Bhattacharya, S. S., and Thrasher, A. J. (2000). Restoration of photoreceptor
ultrastructure and function in retinal degeneration slow mice by gene therapy,
Nat. Genet., 25:306.

Copyright 2003 Marcel Dekker, Inc.

528

Pitkanen et al.

9. Sarra, G. M., Stephens, C., de Alwis, M., Bainbridge, J. W., Smith, A. J.,
Thrasher, A. J., and Ali, R. R. (2001). Gene replacement, therapy in the
retinal degeneration slow (rds) mouse: the eect on retinal degeneration following partial transduction of the retina, Hum. Mol. Genet., 10:2353.
10. Wu, T. H., Ting, T. D., Okajima, T. I., Pepperberg, D. R., Ho, Y. K., Ripps,
H., and Naash, M. I. (1998). Opsin localization and rhodopsin photochemistry in a transgenic mouse model of retinitis pigmentosa, Neuroscience,
87:709.
11. Naash, M. I., Ripps, H., Li, S., Goto, Y., and Peachey, N. S. (1996). Polygenic
disease and retinitis pigmentosa: albinism exacerbates photoreceptor degeneration induced by the expression of a mutant opsin in transgenic mice, J.
Neurosci., 16:7853.
12. Naash, M. L., Peachey, N. S., Li, Z. Y., Gryczan, C. C., Goto, Y., Blanks, J.,
Milam, A. H., and Ripps, H. (1996). Light-induced acceleration of photoreceptor degeneration in transgenic mice expressing mutant rhodopsin, Invest.
Ophthalmol. Vis. Sci., 370:775.
13. Naash, M. I., Hollyeld, J. G., al-Ubaidi, M. R., and Baehr, W. (1993).
Simulation of human autosomal dominant retinitis pigmentosa in transgenic
mice expressing a mutated murine opsin gene, Proc. Natl. Acad. Sci., 90:5499.
14. Kedzierski, W., Lloyd, M., Birch, D. G., Bok, D., and Travis, G. H. (1997).
Generation and analysis of transgenic mice expressing P216L-substituted rds/
peripherin in rod photoreceptors, Invest. Ophthalmol. Vis. Sci., 38:498.
15. Salchow, D. J., Gouras, P., Doi, K., Go, S. P., Schwinger, E., and Tsang, S.
H. (1999). A point mutation (W70A) in the rod PDE-gamma gene desensitizing and delaying murine rod photoreceptors, Invest. Ophthalmol. Vis. Sci.,
40:3262.
16. Leconte, L., and Barnstable, C. J. (2000). Impairment of rod cGMP-gated
channel alpha-subunit expression leads to photoreceptor and bipolar cell
degeneration, Invest. Ophthalmol. Vis. Sci., 41:917.
17. Tsang, S. H., Chen, J., Kjeldbye, H., Li, W. S., Simon, M. I., Gouras, P., and
Go, S. P. (1997). Retarding photoreceptor degeneration in Pdegtm1/
Pdegtm1 mice by an apoptosis suppressor gene, Invest. Ophthalmol. Vis.
Sci., 38:943.
18. Hong, D. H., Pawlyk, B. S., Shang, J., Sandberg, M. A., Berson, E. L., and Li,
T. (2000). A retinitis pigmentosa GTPase regulator (RPGR)-decient mouse
model for X-linked retinitis pigmentosa (RP3), Proc. Natl. Acad. Sci., 97:3649.
19. Liang, F. Q., Dejneka, N. S., Cohen, D. R. Krasnoperova, N. V., Lem, J.,
Maguire, A. M., Dudus, L., Fisher, K. J., and Bennett, J. (2001). AAVmediated delivery of ciliary neurotrophic factor prolongs photoreceptor survival in the rhodopsin knockout mouse, Mol. Ther., 3:241.
20. Jaissle, G. B., May, C. A., Reinhard, J., Kohler, K., Fauser, S., LutjenDrecoll, E., Zrenner, E., and Seeliger, M. W. (2001). Evaluation of the rhodopsin knockout mouse as a model of pure cone function, Invest. Ophthalmol.
Vis. Sci., 42:506.

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

529

21. Humphries, M. M., Rancourt, D., Farrar, G. J., Kenna, P., Hazel, M., Bush,
R. A., Sieving, P. A., Sheils, D. M., McNally, N., Creighton, P., Erven, A.,
Boros, A., Gulya, K., Capecchi, M. R., and Humphries, P. (1997).
Retinopathy induced in mice by targeted disruption of the rhodopsin gene,
Nat. Genet., 15:216.
22. Chen, J., Simon, M. I., Matthes, M. T., Yasumura, D., and LaVail, M. M.
(1999). Increased susceptibility to light damage in an arrestin knockout mouse
model of Oguchi disease (stationary night blindness), Invest. Ophthalmol. Vis.
Sci., 40:2978.
23. Yu, D. Y., Cringle, S. J., Sue, E. N., and Yu, P. K. (2000). Intraretinal oxygen
levels before and after photoreceptor loss in the RCS rat, Invest. Ophthalmol.
Vis. Sci., 41:3999.
24. Strauss, O., Stump, F., Mergler, S., Wienrich, M., and Wiederholt, M.
(1998). The Royal College of Surgeons rat: An animal model for inherited
retinal degeneration with a still unknown genetic defect, Acta Anat., 162:101.
25. LaVail, M. M., Unoki, K., Yasumura, D., Matthes, M. T., Yancopoulos, G.
D., and Steinberg, R. H. (1992). Multiple growth factors, cytokines, and
neurotrophins rescue photoreceptors from the damaging eects of constant
light, Proc. Natl. Acad. Sci., 89:11249.
26. Lai, C. M., Shen, W. Y., Constable, I., and Rakoczy, P. E. (2000). The use of
adenovirus-mediated gene transfer to develop a rat model for photoreceptor
degeneration, Invest. Ophthalmol. Vis. Sci., 41:580.
27. Lewin, A. S., Drenser, K. A., Hauswirth, W. W., Nishikawa, S., Yasumura,
D., Flannery, J. G., and LaVail, M. M. (1998). Ribozyme rescue of photoreceptor cells in a transgenic rat model of autosomal dominant retinitis pigmentosa, Nat. Med., 4:967.
28. Bush, R. A., Kononen, L., Machida, S., and Sieving, P. A. (2000). The eect
of calcium channel blocker diltiazem on photoreceptor degeneration in the
rhodopsin Pro213His rat, Invest. Ophthalmol. Vis. Sci., 41:2697.
29. Nir, I., Harrison, J. M., Liu, C., and Wen, R. (2001). Extended photoreceptor
viability by light stress in the CS rats but not in the opsin P23H mutant rats,
Invest. Ophthalmol. Vis. Sci., 42:842.
30. Green, E. S., Rendahl, K. G., Zhou, S., Ladner, M., Coyne, M., Srivastava,
R., Manning, W. C., and Flannery, J. G. (2001). Two animal models of retinal
degeneration are rescued by recombinant adeno-associated virus-mediated
production of fgf-5 and fgf-18, Mol. Ther., 3:507.
31. Lau, D., McGee, L. H., Zhou, S., Rendahl, K. G., Manning, W. C.,
Escobedo, J. A., and Flannery, J. G. (2000). Retinal degeneration is slowed
in transgenic rats by AAV-dilated delivery of FGF-2, Invest. Ophthalmol. Vis.
Sci., 41:3622.
32. Machida, S., Chaudhry, P., Shinohara, T., Singh, D. P., Reddy, V. N.,
Chylack, L. T. Jr, Sieving, P. A., and Bush, R. A. (2001). Lens epitheliumderived growth factor promotes photoreceptor survival in light-damaged and
RCS rats, Invest. Ophthalmol. Vis. Sci. 42:1087.

Copyright 2003 Marcel Dekker, Inc.

530

Pitkanen et al.

33. Ivert, L., Gouras, P., Naeser, P., and Narfstrom, K. (1998). Photoreceptor
allografts in a feline model of retinal degeneration, Graefes Arch. Clin. Exp.
Ophthalmol., 236:844.
34. Curtis, R., Barnett, K. C., and Leon, A. (1987). An early-onset retinal dystrophy with dominant inheritance in the Abyssinian cat. Clinical and pathological ndings, Invest. Ophthalmol. Vis. Sci., 28:131.
35. Leon, A., and Curtis, R. (1990). Autosomal dominant rod-cone dysplasia in
the Rdy cat. 1. Light and electron microscopic ndings, Exp. Eye Res., 51:361.
36. Leon, A., Hussain, A. A., and Curtis, R. (1991). Autosomal dominant rodcone dysplasia in the Rdy cat. 2. Electrophysiological ndings, Exp. Eye Res.,
53:489.
37. Chong, N. H., Alexander, R. A., Barnett, K. C., Bird, A. C., and Luthert, P. J.
(1999). An immunohistochemical study of an autosomal dominant feline rod/
cone dysplasia (Rdy cats), Exp. Eye Res., 68:51.
38. Petters, R. M., Alexander, C. A., Wells, K. D., Collins, E. B., Sommer, J. R.,
Blanton, M. R., Rojas, G., Hao, Y., Flowers, W. L., Banin, E., Cideciyan, A.
V., Jacobson, S. G., and Wong, F. (1997). Genetically engineered large animal
model for studying cone photoreceptor survival and degeneration in retinitis
pigmentosa, Nat. Biotechnol., 15:965.
39. Pearce-Kelling, S. E., Aleman, T. S., Nickle, A., Laties, A. M., Aguirre, G. D.,
Jacobson, S. G., and Acland, G. M. (2001). Calcium channel blocker D-cisdiltiazem does not slow retinal degeneration in the PDE6B mutant rcd1 canine
model of retinitis pigmentosa, Mol. Vis., 7:42.
40. Wang, W., Acland, G. M., Ray, K., and Aguirre, G. D. (1999). Evaluation of
cGMP-phosphodiesterase (PDE) subunits causal association with rod-cone
dysplasia 2 (rcd2), a canine model of abnormal retinal cGMP metabolism,
Exp. Eye Res., 69:445.
41. Anderson, R. E., Maude, M. B., Narfstrom, K., and Nilsson, S. E. (1997).
Lipids of plasma, retina, and retinal pigment epithelium in Swedish briard
dogs with a slowly progressive retinal dystrophy, Exp. Eye Res., 64:181.
42. Rapp, L. M., Tolman, B. L., and Dhindsa, H. S. (1990). Separate mechanisms
for retinal damage by ultraviolet-A and mid-visible light, Invest. Ophthalmol.
Vis. Sci., 31:1186.
43. Seiler, M. J., Liu, O. L., Cooper, N. G., Callahan, T. L., Petry, H. M., and
Aramant, R. B. (2000). Selective photoreceptor damage in albino rats using
continuous blue light. A protocol useful for retinal degeneration and transplantation research, Graefes Arch. Clin. Exp. Ophthalmol., 238:59.
44. Collier, R. J., Waldron, W. R., and Zigman, S. (1989). Temporal sequence of
changes to the gray squirrel retina after near-UV exposure, Invest.
Ophthalmol. Vis. Sci., 30:631.
45. Faktorovich, E. G., Steinberg, R. H., Yasumura, D., Matthes, M. T., and
LaVail, M. M. (1992). Basic broblast growth factor and local injury protect
photoreceptors from light damage in the rat, J. Neurosci., 12:3554.

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

531

46. Organisciak, D. T., Darrow, R. M., Barsalou, L., Darrow, R. A., Kutty, R.
K., Kutty, G., and Wiggert, B. (1998). Light history and age-related changes
in retinal light damage, Invest. Ophthalmol. Vis. Sci., 39:1107.
47. Berman, E. R. (1991). Biochemistry of the Eye. Plenum Press, New York.
48. Hida, T., Chandler, D. B., and Sheta, S. M. (1987). Classication of the stages
of proliferative vitreoretinopathy in a rened experimental model in the rabbit
eye, Graefes Arch. Clin. Exp. Ophthalmol., 225:303.
49. Murata, T., Kimura, H., Sakamoto, T., Osusky, R., Spee, C., Stout, T. J.,
Hinton, D. R., and Ryan, S. J. (1997). Ocular gene therapy: experimental
studies and clinical possibilities, Ophthalmic Res., 29:242.
50. Fastenberg, D. M., Diddie, K. R., Dorey, K., and Ryan, S. J. (1982). The role
of cellular proliferation in an experimental model of massive periretinal proliferation, Am. J. Ophthalmol., 93:565.
51. Andrews, A., Balciunaite, E., Leong, F. L., Tallquist, M., Soriano, P., Refojo,
M., and Kazlauskas, A. (1999). Platelet-derived growth factor plays a key role
in proliferative vitreoretinopathy, Invest. Ophthalmol. Vis. Sci., 40:2683.
52. Frenzel, E. M., Neely, K. A., Walsh, A. W., Cameron, J. D., and Gregerson,
D. S. (1998). A new model of proliferative vitreoretinopathy, Invest.
Ophthalmol. Vis. Sci., 39:2157.
53. Ozerdem, U., Mach-Hofacre, B., Cheng, L., Chaidhawangul, S., Keefe,
McDermott, C. D., Bergeron-Lyn, G., Appelt, K., and Freeman, W. R.
(2000). The eect of prinomastat (AG3340), a potent inhibitor of matrix
metalloproteinases, on a subacute model of proliferative vitreoretinopathy,
Curr. Eye Res., 20:447.
54. Pinon, R. M., Pastor, J. C., Saornil, M. A., Goldaracena, M. B., Layana, A.
G., Gayoso, M. J., and Guisasola, J. (1992). Intravitreal and subretinal proliferation induced by platelet-rich plasma injection in rabbits, Curr. Eye. Res.,
11:1047
55. Goldaracena, M. B., Garcia-Layana, A., Pastor, J. C., Saornil, M. A., de la
Fuente, F., and Gayoso, M. J. (1997). The role of retinotomy in an experimental rabbit model of proliferative vitreoretinopathy, Curr. Eye Res. 16:422.
56. Pastor, J. C., Rodriguez, E., Marcos, M. A., and Lopez, M. I. (2000).
Combined pharmacologic therapy in a rabbit model of proliferative vitreoretinopathy (PVR), Ophthalmic Res., 32:25.
57. Algvere, P. V., Hallnas, K., Dafgard, E., and Hoog, A. (1990). Panretinal
photocoagulation aggravates experimental proliferative vitreoretinopathy,
Graefes Arch. Clin. Exp. Ophthalmol., 228:461.
58. Garcia-Layana, A., Pastor, J. C. Saornil, M. A., and Gonzalez, G. (1997).
Porcine model of proliferative vitreoretinopathy with platelets, Curr. Eye Res.,
16:556.
59. Baudouin, C., Khosravi, E., Pisella, P. J., Ettaiche, M., and Elena, P. P.
(1998). Inammation measurement and immunocharacterization of cell proliferation in an experimental model of proliferative vitreoretinopathy,
Ophthalmic Res., 30:340.

Copyright 2003 Marcel Dekker, Inc.

532

Pitkanen et al.

60. Ishibashi, T., Inomata, H., Sakamoto, T., and Ryan, S. J. (1995). Pericytes of
newly formed vessels in experimental subretinal neovascularization, Arch.
Ophthalmol., 113:227.
61. Danis, R. P., Yang, Y., Massicotte, S. J., and Bold, H. C. (1993). Preretinal
and optic nerve head neovascularization induced by photodynamic venous
thrombosis in domestic pigs, Arch. Ophthalmol., 111:539.
62. Okamoto, N., Tobe, T., Hackett, S. F., Ozaki, H., Vinores, M. A.,
LaRochelle, W., Zack, D. J., and Campochiaro, P. A. (1997). Transgenic
mice with increased expression on vascular endothelial growth factor in the
retina: A new model of intraretinal and subretinal neovascularization, Am. J.
Pathol., 151:281.
63. Ozaki, H., Hayashi, H., Vinores, S. A., Moromizato, Y., Campochiaro, P. A.,
and Oshima, K. (1997). Intravitreal sustained release of VEGF causes retinal
neovascularization in rabbits and breakdown of the blood-retinal barrier in
rabbits and primates, Exp. Eye Res., 64:505.
64. Saito, Y., Park, L., Skolik, S. A., Alfaro, D. V., Chaudhry, N. A., Barnstable,
C. J., and Liggett, P. E. (1997). Experimental preretinal neovascularization by
laser-induced venous thrombosis in rats, Curr. Eye Res., 16:26.
65. Kimura, H., Sakamoto, T., Hinton, D. R., Spee, C., Ogura, Y., Tabata, Y.,
Ikada, Y., and Ryan, S. J. (1995). A new model of subretinal neovascularization in the rabbit. Invest. Ophthalmol. Vis. Sci., 36:2110.
66. Smith, L. E., Wesolowski, E., McLellan, A., Kostyk, S. K., DAmato, R.,
Sullivan, R., and DAmore, P. A. (1994). Oxygen-induced retinopathy in the
mouse, Invest. Ophthalmol. Vis. Sci. 35:101.
67. Ishihama, H., Ohbayashi, M., Kurosawa, N., Kitsukawa, T., Matsuura, O.,
Miyake, Y., and Muramatsu, T. (2001). Colocalization of neuropilin-1 and
Flk-1 in retinal neovascularization in a mouse model of retinopathy, Invest.
Ophthalmol. Vis. Sci., 42:1172.
68. Penn, J. S., Henry, M. M., Wall, P. T., and Tolman, B. L. (1995). The range of
PaO2 variation determines the severity of oxygen-induced retinopathy in newborn rats, Invest. Ophthalmol. Vis. Sci., 36:2063.
69. Lukiw, W. J, Gordon, W. C., Rogaev, E. I., Thompson, H. and Bazan, N. G.
(2001). Presenilin-2 (PS2) expression up-regulation in a model of retinopathy
of prematurity and pathoangiogenesis, Neuroreport 12:53.
70. Zhang, S., Leske, D. A., and Holmes, J. M. (2000). Neovascularization grading methods in a rat model of retinopathy of prematurity, Invest. Ophthalmol.
Vis. Sci., 41:887.
71. McLeod, D. S., Crone, S. N., and Lutty, G. A. (1996). Vasoproliferation in
the neonatal dog model of oxygen-induced retinopathy, Invest. Ophthalmol.
Vis. Sci., 37:1322.
72. Nishikawa, S., and LaVail, M. M. (1998). Neovascularization of the RPE:
temporal dierences in mice with rod photoreceptor gene defects, Exp. Eye
Res., 67:509.

Copyright 2003 Marcel Dekker, Inc.

Retinal Disease Models

533

73. Zhang, S., Leske, D. A., Lanier, W. L., Berkowitz, B. A., and Holmes, J. M.
(2001). Preretinal neovascularization associated with acetazolamide-induced
systemic acidosis in the neonatal rat, Invest. Ophthalmol. Vis. Sci., 42:1066.
74. Antoszyk, A. N., Gottlieb, J. L., Casey, R. C., Hatchell, D. L., and
Machemer, R. (1991). An experimental model of preretinal neovascularization
in the rabbit, Invest. Ophthalmol. Vis. Sci., 32:46.
75. de Juan, E. Jr., Humayun, M. S., Hatchell, D. L., and Wilson, D. (1989).
Histopathology of experimental preretinal neovascularization, Invest.
Ophthalmol. Vis. Sci., 30:1495.

Copyright 2003 Marcel Dekker, Inc.

18
New Experimental Therapeutic
Approaches for Degenerative
Diseases of the Retina
Joyce Tombran-Tink
University of MissouriKansas City, Kansas City, Missouri, U.S.A.

I.

INTRODUCTION

The eye is made up of highly specialized and complex groups of tissues that
are important to vision. Visual impairment can result from severe damage to
the cornea, lens, nerve tissue of the eye, optic nerve, retina, or the brain.
Central visual disturbances usually involve macular atrophy and change in
the vitreous and aqueous humor, complications that often lead to a complete loss of vision. While cataracts and corneal opacity account for most
cases of blindness in developing countries, degenerative diseases of the retina
are the leading cause of blindness in the Western world.
Retinal degenerative diseases are etiologically complex disorders that,
for the most part, lack appropriate animal models, which are essential to
elucidate the progression of the disease or to test the ecacy of many
promising pharmacological agents. In this regard, the development of eective treatments for most forms of retinal degeneration has been slow, and
the intervention strategies currently available are aimed at preserving vision
only and not reversing the process of the disease. With the recent explosion
of sophisticated molecular technologies, the genetic and biochemical bases
of many retinal diseases are being better dened. This knowledge will accelerate experimental studies and facilitate the emergence of novel therapies for
ocular pathologies.

535

Copyright 2003 Marcel Dekker, Inc.

536

Tombran-Tink

The objectives of this chapter are: (a) to describe the clinically relevant
innovations for managing degenerative diseases of the retina, (b) to understand the biology, development, and unique characteristics of each
approach, (c) to discuss the clinical potential and limitations associated
with the developing therapeutics, and (d) to examine modications that
may potentiate the therapeutic benet of each application. A critical evaluation of basic and clinical research in the area of retinal and stem cell transplantation, retinal prosthesis, viral and nonviral DNA-based therapies, and
the use of soluble neurotrophic factors in managing retinal diseases will be
provided. These innovations are still in their infancy but promise enormous
therapeutic benets to blind individuals. However, because they are still
being dened, a true comparison between them is clearly beyond the
scope of this chapter. Before discussing developing therapeutics in the eye,
it will be pertinent to review a few relatively, well-dened degenerations of
the retina that are leading causes of blindness.

II.

DEGENERATIVE DISEASES OF THE RETINA

Diseases that aect the center of the retina are commonly categorized as
retinal degenerations. The death of photoreceptor cells, the light-sensitive
neurons of the retina, is the hallmark of these disorders. Loss of function of
the retinal pigment epithelium (RPE) and increased neovascularization are
often complicating factors in advanced stages of many retinal diseases.
While a relatively signicant body of information is available for some
forms of retinal degeneration, knowledge about specic mechanisms that
cause the death of photoreceptor cells is still obscure and hinders the progress of developing eective treatments. One feature that further complicates diagnosis and treatment is the high level of genetic heterogeneity
associated with inherited retinal diseases. For example, more than 150 mutations have been identied in rhodopsin- and peripherin-linked retinitis pigmentosa. In addition, the cellular diversity in ocular tissues allows a single
pathology to be associated with the dysfunction of more than one cell type.
This contributes to the disease progression and interferes with diagnosis in
advanced cases. In these conditions, end-stage pathologies are complex and
often involve multiple genes, many mutations, several dysfunctional cell
types, and the interruption of many biochemical pathways. Thus, it is virtually impossible for any single mutation-based or cell-based treatment to
emerge as an eective cure for late-state retinal degenerations. In this
regard, multimodal approaches and adjuvant therapies may prove to be
more eective in clinical attempts to impede the loss of vision in advanced
retinal pathologies. The genetic information currently available for retinitis

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

537

pigmentosa (RP), one of the better-characterized eye diseases, has been

useful in developing excellent natural and transgenic RP models for preclinical testing and for developing pharmaceutics. Progress, however, is slowed
for other degenerative diseases leading to blindness, such as macular degeneration, because of their less well-dened etiologies and lack of appropriate
animal models. Careful characterization of the genetic and biochemical
basis of these degenerations is essential for developing better animal models
to study disease progression and to evaluate new treatments.
A.

Macular Degeneration

The two most common degenerative diseases of the retina are macular
degeneration and retinitis pigmentosa. In both diseases, degeneration of
photoreceptor cells is the underlying cause of the pathology. Dysfunction
of the RPE cells and choroidal neovascularization are also major contributing factors in the development of macular degeneration.
Diseases that are grouped as macular degenerations aect central
vision. Damage to the macular region is the primary cause for the loss of
vision. Symptoms include blurred vision, distortion of lines and shapes,
blind spots, reading diculties, and an inability to recognize faces.
Peripheral vision is not severely aected in these individuals. There are
two groups of macular degeneration: age-related macular degeneration
(AMD) and juvenile inherited macular degeneration. AMD aects more
than 700,000 Americans each year and approximately 30 million individuals
worldwide. It is the leading cause of irreversible blindness in the Western
world in individuals over the age of 50. The disease aects the center of the
retina and is related to aging, light iris color, prolonged exposure to sunlight, smoking, and a family history. Aging mechanisms are the main factors
associated with degeneration of the RPE and photoreceptor cells and the
subsequent progression of AMD. Loss of RPE cell function compromises
the blood-retinal barrier, resulting in additional macula edema and further
degeneration of the few surviving photoreceptor cells. A loss of photoreceptor function results in the inability of the retina to convert light stimulus to
electrical signals.
Two forms of AMD are identied: wet and dry. In wet AMD, blood
vessels in the choroid undergo abnormal growth and invade the subretinal
space located in the retina and the choroid. The disease is referred to as
wet because the blood vessels leak their contents into the retina. The
accumulation of uid in the macula region subsequently leads to damage
of the retina. Leakage, bleeding, and scarring from choroidal neovascularization eventually result in irreversible blindness. Ninety percent of individuals with the wet form of AMD suer severe visual loss (14).

Copyright 2003 Marcel Dekker, Inc.

538

Tombran-Tink

Unfortunately, no curative treatments are currently available for wet AMD.

A new photodynamic therapy has shown potential to delay the progression
of the disease with early diagnosis. In this treatment, the drug visudyne
(verteporn) is activated by laser light after intravenous injection. At best,
however, laser-activated treatments can only delay visual loss, not reverse
the condition, restore vision, or prevent a loss of vision (58).
The dry form of AMD is milder and more common. It accounts for
approximately 90% of all AMD cases. It is not usually associated with
abnormal growth of blood vessels and uid leakage in the retina but, rather,
with aging, yellow spots called drusen. Drusen accumulate in a scattered
pattern in the aected retina region and interfere with the function of photoreceptor and RPE cells (9). There is still no eective FDA-approved treatment for dry AMD. Fortunately, visual acuity is not severely aected in dry
AMD, and the disease progression is relatively slow as compared to wet
AMD.
A small percentage of individuals have a genetic predisposition to
macular degeneration. This form of retinal degeneration is called juvenile
inherited macular degeneration and is diagnosed at a very young age. Two
examples of this disorder are vitelliform dystrophy, an autosomal dominant
disorder, and Stargardt disease, an autosomal recessive condition (10,11).
When developing a therapeutic approach for macular degeneration,
numerous factors are taken into consideration:
1.
2.
3.
4.

5.
6.
7.
8.
9.
10.

B.

Limiting further degeneration of the retina

Promoting survival of existing photoreceptors
Replacing photoreceptors that have degenerated
Restoring functional synaptic connections between the photoreceptors and inner neural retinal cells to maintain the appropriate
retinal circuitry
Replacing dysfunctional RPE cells
Regulating choroidal neovascularization
Repairing Bruchs membrane
Eliminating negative eects of surgical procedures and spontaneous biochemical changes
Restoring vision
Preventing recurrence of the condition.

Retinitis Pigmentosa

The term retinitis pigmentosa describes a group of inherited eye diseases

associated with the dysfunction of rods and cones, often characterized by
night blindness. The disease aects 4 million individuals worldwide and is

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

539

the leading cause of inherited blindness. RP is usually characterized by an

early onset and is diagnosed in children and early adolescents. Patients with
RP present with varying symptoms, such as diminished night vision (nyctalopia) and a gradual loss of peripheral vision. Central vision disturbances
are not generally evident until the later stages of the disease. In RP, the
aected photoreceptor cells are predominantly the light-sensitive rods.
Individuals suering from RP experience peripheral eld constriction and
a diminishing ability to see in dim light.
The rst signs of RP are marked by attenuation of retinal arterioles
and scattered brown-pigmented granules around the atrophied area. As the
disease progresses, degeneration of the RPE, choriocapillaries, and photoreceptor outersegments occurs. Metabolic by-products, such as lipofuscin,
accumulate and subsequently interfere with the visual transduction cascade.
Changes in the RPE cells compromise the integrity of the blood-retinal
barrier, and, as a consequence, subretinal leakage and macular edema ensue.
Dozens of mutations have been identied with RP in at least 10 different genes, including the genes for rhodopsin and peripherin. Given the
genetic complexity of this disease, it is not surprising that research has not
yet resulted in a successful mutation-based therapy for the condition. To
date, there is still no eective cure for any form of retinitis pigmentosa. RP
will likely benet the most from innovative DNA-based therapies due to the
current genetic information and excellent RP animal models available for
this disease (1215).
Usher syndrome is another type of retinitis pigmentosa. It is a recessively inherited condition accompanied by a loss of hearing. This condition
accounts for approximately 6% of hearing impairment in deaf individuals.
Suerers may experience both a severe loss of vision and loss of hearing, a
situation that increases social isolation of these individuals as auditory and
visual communication progressively decreases (16).
C.

Leber Congenital Amaurosis

RP is just one type of inherited disease that aects the retina. Leber congenital amaurosis (LCA) is an early-onset form of retinal degeneration
characterized by severe blindness from birth or early childhood.
Individuals with LCA have no detectable electroretinogram measurements
because the retina is unable to respond to light (17). Only recently, a mutation in a gene called RPE65 was identied as the cause of a specic form of
LCA (18). RPE65 is a product of the RPE cells and supports the function of
photoreceptor cells. The gene is involved in the biochemical cascade of
phototransduction and is important in converting light to neural signals.
Since the LCA mutation was dened, many experiments to replace the

Copyright 2003 Marcel Dekker, Inc.

540

Tombran-Tink

function of the mutated RPE65 gene have been carried out in animal models. These show promise in restoring some level of vision in individuals
suering from LCA.
Other less well-dened inherited diseases that result in degeneration of
the retina include:
1. Choroideremia, a disease with symptoms similar to RP but
which is also accompanied by degeneration of the choroid
2. Gyrate atrophy, a retinal disease with RP-like symptoms, cataract formation, and loss of peripheral vision
3. Bardet-Biedl syndrome, a form of RP with the additional complications of mental retardation, obesity, and kidney disorders
4. Refsum syndrome, a complex disease manifesting many RP
symptoms, as well as hearing impairment, neurological problems, abnormalities in red blood cells, and skin disorders
5. Bassen-Kornzweig syndrome, which involves RP complications
and accompanying neurological disturbances
There are no eective treatments for any of the RP or RP-like diseases.
Vitamin A supplements have been successful in slowing the progression of
RP without reversing the condition. The use of vitamin A, however, has not
gained wide acceptance as a treatment for retinitis pigmentosa. At the present time, advanced stages of retinal degenerations are irreparable by current pharmacological interventions. Patients suering from advanced retinal
degeneration may benet the most from developing technologies in molecular medicine where the benets outweigh the risks of the treatment. For
the remainder of this chapter, a few of the more innovative approaches in
the management of ocular diseases will be discussed, with emphasis on their
relevance in treating degenerative diseases of the retina.

III.

NEW THERAPEUTIC METHODS FOR RETINAL

DISEASES

A.

Retinal Transplants

Human retinal transplants are in the early stages of development as a therapeutic approach to retinal degenerations. The objective of this strategy is to
improve vision by replacing the lost function of retinal cells. Since degeneration of the RPE and photoreceptor cells is the underlying cause of several
retinal diseases, transplants of the neural retina, RPE, iris pigment epithelium (IPE), and photoreceptor cells are currently being evaluated for their
potential to integrate into the retina and to permanently substitute for the
function of cells they are replacing (1925). The technology has received

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

541

intense criticism as being unrealistic in its eort to improve vision. At the

same time, however, it has stimulated a rapid increase in experimental studies, with both retinal and stem cells transplants. The results obtained from
animal models and clinical trials of blind individuals are promising and
show potential for retinal transplants as an intervention strategy that can
improve and stabilize vision.
1.

Animal Studies

a. RPE Transplant. An animal model of hereditary retinal degeneration, often used to study the effects of transplants in the eyes, is the
Royal College of Surgeons (RCS) rats. These animals suffer from a wellcharacterized, early-onset form of photoreceptor cell degeneration. One of
the earliest indications that grafts of human fetal RPE cells are capable of
rescuing photoreceptor degeneration in RCS rats was found in a report
by Little and coworkers (26). In the study, sheets of healthy fetal human
RPE cells were transplanted into the subretinal space of dystrophic RCS
rats that have been immunosuppressed with cyclosporine. The thickness
of the outer nuclear layer (ONL), which contains the photoreceptor cells,
was measured 4 weeks after transplantation. A fourfold increase in thickness of the ONL, in the area where the grafts were placed, was observed
after the transplant period. Furthermore, the protected cells in the ONL
were morphologically similar to their normal counterparts in healthy retinas. ONL thickness remained unchanged in areas of the retina distant
from the site of transplant or in the retina of sham injected controls. This
preliminary study indicates that transplanted RPE cells are capable of rescuing a signicant percentage of photoreceptors from degenerating.
One question raised by many researchers in the ophthalmic eld is:
Are the rescued photoreceptor cells functional? Results from a follow-up
study suggested that morphological as well as functional rescue of photoreceptor cells can be achieved with RPE transplants (27). Data from the
study showed that a progressive central to peripheral loss of visual responsiveness occurs in pigmented dystrophic RCS rats. Recordings of single and
multiunit receptive elds across the surface of the superior colliculus were
used to determine the changes in visual responsiveness in the rat retinas. The
potential of RPE transplants to slow down progressive loss of vision was
subsequently tested in these animals by subretinal injections of healthy RPE
cells into the eyes of the dystrophic RCS rats. Recordings taken 85108 days
after the transplantation procedure indicated that photoreceptor rescue
occurred in the region of the graft and that the progressive central to peripheral loss of visual responsiveness in the animals was delayed by the
healthy donor RPE cells. Many RPE transplant studies have now been

Copyright 2003 Marcel Dekker, Inc.

542

Tombran-Tink

carried out in a number of animal models, and most have conrmed the
clinical potential of RPE transplants to integrate into the retina, to rescue
photoreceptor cells, and to improve vision.
b. Cograft Transplant: RPE-Retina Complex. For transplant procedures to become clinically feasible, it is important that the transplanted
tissue provide long-term benets to blind individuals. In a study that examined the long-term survival of retinal transplants, Sharma and coworkers observed that neuronal components of retinal grafts did not
survive as well as glial cells after transplant (28). They proposed that
photoreceptors require contact and trophic support of the adjacent RPE
cells to survive over an extended period of time. Cotransplantation of
these two tissues, therefore, may be clinically more advantageous in promoting tissue survival and retina repair for diseases in which both the
RPE and photoreceptor cells have degenerated. This hypothesis was supported by a recent study, which showed healthy morphology and normal
cell function of RPE-retina cografts after successful transplant into the
subretinal space of albino RCS rats (29). Morphological and biochemical
evaluations of the transplanted tissue suggest that the photoreceptor cells
in the cografts were well differentiated and capable of normal visual function (Fig. 1).
In support of the cograft study, another group demonstrated the benecial eects of RPE-retina cografts on photoreceptor cell morphology and
function in a study using a large number of dystrophic RCS rats (30).
Recordings in the superior colliculus, from areas of the retina restricted to
the site of transplant, indicated that visually evoked responses were restored
in the brain after cograft integration into the retina. These preliminary
results are encouraging and could have only been possible if normal retinal
circuitry was established by functional photoreceptor connections within the
cografts.
c. IPE Transplant. The need for readily available tissue alternatives
that can promote photoreceptor survival has led many laboratories to examine the transplant effects of the IPE in the retina. Several clinically appealing features are associated with the use of IPE transplants for retinal
degenerations. These features include: (a) the IPE contains pigmented cells
of the same embryological origin as the RPE cells; (b) IPE cells have a
high transdifferentiation potential; (c) they can be obtained with relative
ease by peripheral iridectomy; and, perhaps the most advantageous aspect
of this approach, (d) the promise of functionally autologous IPE transplants in patients with retinal degeneration.
A few studies have now demonstrated the clinical potential of IPE
transplants for retinal diseases. In one of the initial experiments, IPE cells,

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

543

obtained by iridectomy from Long-Evans rats, were transplanted into the

choroid and subretinal space of RCS rats. The benecial eects of the
allografts were evaluated 6 months after the procedure using computerassisted morphometric and microscopic analyses. In the transplanted rat
retinas, a remarkable eect on photoreceptor survival was observed even
though a high percentage of the grafted IPE cells were located in the choroid. The rescue appeared to be specic as photoreceptors were absent in the
nontransplanted group (31,32). These results conrmed earlier reports suggesting that the IPE cells are capable of delaying photoreceptor degeneration (22,35), and this may be a useful alternative approach to retinal
transplants (Fig. 2).
d. Autologous IPE Transplant. Recently, the successful transplantation of autologous IPE cells in animals has made the use of IPE cells for
retinal degeneration even more clinically appealing. In a study carried out
with a group of 25 rabbits, autologous IPE cells were harvested by iridectomy from one eye of each rabbit and transplanted into the subretinal
space of the contralateral eye. After a transplant period of 5 months, the
presence of healthy IPE was observed in the rabbits retina by light and
electron microscopic examination. The IPE grafts formed a polarized
monolayer above the retinal pigment epithelium and projected microvillous processes towards the photoreceptor cells. Furthermore, they appeared to be involved in the process of phagocytosis of shed
photoreceptor outer segments, a biological activity associated with the
RPE and one that is important to outersegment renewal. The lack of immunological side effects with the IPE transplants is of signicant clinical
importance (33). In addition to the experiments with rabbits, successful
transplant of autologous IPE was also shown in monkey retinas. Monkey
IPE cells, obtained by peripheral iridectomy, were rst cultured with autologous serum and labeled with DIL prior to the experimental procedure.
Labeled autologous IPE cells were subsequently transplanted into the
monkeys submacular region, and the eyes were observed regularly by
uorescein angiography and fundus examination. Analysis of histological
preparations, indicated that autologous IPE was present in the submacular region and that these appeared to interact with the RPE cells. Their
presence was determined by the intense uorescence of Dil-labeled IPE
cells in the region. The transplanted cells stayed in the area for at least 6
months after the experiment and did not mount an immunological response in the animals (35) (Fig. 3). Autologous IPE transplants may,
therefore, prove to be of signicant therapeutic benet in the treatment of
retinal diseases because they are easy to obtain, are capable of RPE function, and do not promote host-graft rejection.

Copyright 2003 Marcel Dekker, Inc.

544

Tombran-Tink

Figure 1 Albino RCS rat retinal degeneration model. This rat has a genetic defect
in the RPE and needs both photoreceptors and RPE replaced. The transplants
presented in B and C were achieved by transplanting sheets of fetal neural retina
together with its RPE. (A) Recipient RCS rat retina, close to the transplant, age 3
months. The photoreceptor layer is almost completely degenerated, and the inner
nuclear layer (IN) is immediately adjacent to the defective RPE. (Hematoxylin-Eosin
staining.) (B) Reconstructed area of RCS host retina (H) by a transplant (T) of
neural retina and RPE. Note the good integration between host and retinal transplant. (Toluidine blue staining.) Donor age E20, host age at time of transplantation,
2.2 months; age at time of death, 3.9 months. (From Ref. 29.)

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

545

Figure 1 (C) Overview of a co-transplant. The transplanted pigmented RPE sheet

can easily be seen because of the albino host. The parallel photoreceptor layer of the
transplant in contact with the supporting RPE sheet shows strong immunoreactivity
for the phototransduction protein rod
-transducin. Note that the photoreceptors in
rosettes (arrows) at the edges of the transplant show only weak staining. (No stain in
the host retina.) Host age at time of transplantation, 1.8 months; age at time of death,
3.2 months. The spaces in the section are tissue processing artifacts. (From Ref. 29.)

e. Genetically Modied Transplant. Trophic factors, released in a

sustained manner by transplanted RPE cells, are known to inuence the
progression of hereditary retinal degeneration in the RCS rats. Examples
of these are: basic broblast growth factor (bFGF), glial-derived neurotrophic factor (GDNF), and ciliary neurotrophic factor (CNTF), all of
which have been reported to promoted photoreceptor survival and delay
the progression of retinal degenerations (34).
An interesting variation in autologous IPE transplants involves the
genetic engineering of IPE cells to secrete high levels of specic cytokines
or growth factors, above their endogeneous levels, before transplanting
them into the retina. In a study designed to test whether genetically modied
cells were more potent in promoting photoreceptor survival, IPE cells were
transfected with the bFGF cDNA, which was cloned into the high-expression vector pCXN2 and subsequently transplanted into the subretinal space
of dystrophic RCS rats. An increase in the expression of bFGF in the retina
was observed as expected, and this level correlated with prolonged photoreceptor survival as compared to the control experiments (35). The increased
benecial eects on photoreceptor cells were attributed to the soluble
trophic factor secreted by the IPE cells. Genetically engineered transplants
may have an advantage in prolonging cell survival by transferring therapeutic genes that have the potential to augment the intrinsic pharmacological
ecacy of the transplanted cells. Gene transfer approaches for retinal
degenerative diseases are discussed in more detail in Sec. III.D.

Copyright 2003 Marcel Dekker, Inc.

546

Tombran-Tink

Figure 2 (A) IPE or bFGF transfected IPE cells were transplanted into the subretinal space of 21-day-old RCS rat eyes. Eyes were enucleated at 30 days after transplantation. (A) RCS rat retina; (B) IPE-transplanted retina; (C) bFGF-IPE
transplanted retina. Asterisk and arrows indicate debris and transplanted bFGFIPE, respectively. Bar 50 m. (D) Thickness of outer nuclear layer in bFGF-IPE
or IPE transplanted retina. (Courtesy of Dr. Toshiaki Aloe, Tohoku University,
Sendai, Japan.)

2. Clinical Trials
a. Human RPE Transplant. The success of the RPE and IPE transplants in animal models has prompted several pilot studies in AMD and
RP patients. A surgical team of ophthalmologists at the Johns Hopkins
Medical Institute recently attempted a Phase 1 clinical trial aimed at restoring vision with fetal human neural retinal transplants (36). Patients selected for the transplant had advanced cases of retinitis pigmentosa and
neovascular age-related macular degeneration. The objective of the pilot
study was to determine the safety of the surgical procedure, the patients
tolerance for the grafted tissue, and the visual outcome. The procedure involved surgically grafting small healthy fragments of fetal human neural
retina sheets into the subretinal space of eight patients with RP and one
with AMD. Recipients of the allografts were not given postoperative systemic immunosuppressive drugs. Several criteria were used to assess visual

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

547

improvement after the procedure. These included, pre- and postoperative

fundus photography, ourescein angiography, scanning laser ophthalmoscope (SLO), macular microperimetry, electroretinogram (ERG) recordings, and visual function testing. Safety of the procedure was
demonstrated in all nine subjects, and graft-versus-host complications
were not severe in any of the patients involved in the clinical trial. An improvement in light sensitivity was recorded for three of the patients in the
study. However, any positive changes in the quality of vision were shortlived and disappeared a few months after the transplant procedure. Only
one individual in the group, an RP patient, developed new blood vessels
at the graft site. It was difcult to determine a statistically signicant effect of the transplants from this initial trial since only a small cohort was
tested. However, a high tolerance for grafte retinal tissue in humans was
clearly demonstrated. The procedure is still in its experimental stages, but
with careful modications in surgical procedures and types of tissue selected for transplant, it may prove to be a feasible biopharmaceutic approach for preventing the loss of vision caused by degeneration of key
retinal cells (36).
Adjuvant therapy using RPE transplants is another approach currently being evaluated in retinal disorders requiring surgical removal of
neovascular membranes. Surgical removal of choroidal neovascular membranes is a technique used to slow down the progression of degeneration in
the retina, but it can also result in a loss of RPE cells and subsequent
disruption of the blood-retinal barrier. Under those conditions, vision is
further compromised by additional leakage and bleeding from the choroidal
blood vessels into the retinal tissue. Prior to the Johns Hopkins study, a
research team from Sweden reported that disruption of the blood-retinal
barrier in advanced AMD cases is highly likely to increase allograft rejection. They showed in their preliminary study that subretinal human RPE
transplants were rejected in 75% of the recipients treated. Tissue rejection
appeared to occur sooner (within 3 months after transplant procedures) in
all cases of neovascular AMD where the blood-retinal barrier was disrupted
by surgical removal of neovascular membranes. Long-term tolerance (620
months) of the grafted tissue was only seen in AMD patients with intact
blood-retinal barrier function (37). In 1994, the same group reported that
RPE transplants were well tolerated in the fovea or parafoveal regions of
AMD recipients who were subjected to surgical removal of subretinal neovascular membranes. While macular edema was observed, the fetal human
RPE cells survived and were capable of replacing lost RPE cells and repairing the damage caused to the retina by surgical intervention (19). While
these studies were preliminary, adjuvant transplant therapy is an interesting
concept worthy of more research eort. It is possible that administration of

Copyright 2003 Marcel Dekker, Inc.

548

Tombran-Tink

Figure 3 (A) Control: normal monkey retina; (B) autologous IPE transplant: light
microscopic examination. (Courtesy of Dr. Toshiaki Aloe, Tohoku University,
Sendai, Japan.)

immunosuppressive drugs after removal of CNV membranes may increase

the adjuvant potential of RPE grafts.
b. Human IPE Transplant. The clinical efcacy of autologous IPE
to improve vision after subretinal removal of neovascular membranes was
evaluated in AMD patients as well. Evidence accumulated from animal
studies indicated that iris pigment epithelium is easy to isolate, can integrate into the retina and substitute for RPE function, and is not associated with graft-versus-host complications. For these reasons, a pilot

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

549

Figure 3 (C) Autologous Dil-labeled IPE transplant seen in cryostat section.

Intensely uorescent IPE cells are seen between the photoreceptors and RPE layers.
(Courtesy of Dr. Tokashi Aloe, Tohoku University, Sendai, Japan.)

study was conducted using human autologous IPE to improve vision. A

cohort of 20 who had neovascular membranes removed was used in this
clinical trial. The transplant procedure was successful, and visual improvement was recorded for 5 patients. Thirteen individuals showed stable visual acuity, and 2 of the 20 patients experienced reduced vision with the
autologous IPE transplants. Rejection of the graft and other immunologically related complications were not observed in any of the AMD recipients (38). The clinical trial points to the use of IPE transplants as a more
feasible treatment option than RPE cells for retinal degenerations or for
those conditions associated with a loss of RPE cells after surgery.
So far, the therapeutic ecacy of retinal transplants to improve and
maintain vision remains nebulous. A signicant sustained positive eect on
visual function has not yet been demonstrated by any single transplantion
procedure. The results from animal and human studies indicate a reasonable
degree of success in graft tolerance, tissue integration, and function of the
transplanted tissue, but the data are still inconclusive and the procedure is
open for debate. Although these ndings are promising for future eorts in
the eld, several technical and biological concerns still need to be addressed
for the procedure to emerge as a successful and clinically relevant therapeutic approach for retinal degenerative diseases (20,23,36,39,40). Some of these
concerns include:
1. Choice of tissue: As yet, retinal tissue from other species cannot
be used for human transplantation purposes. Finding suitable

Copyright 2003 Marcel Dekker, Inc.

550

Tombran-Tink

human alternatives that are easily accessible and functional limits

current eorts in the eld. Human fetal retina tissue is, perhaps,
the most successful transplant treatment option for improving
human vision. The cells from these tissues are not well dierentiated and may contain stem cell populations with high survival
potential in the eye. However, there is much controversy over the
use of human fetal tissue for clinical applications, even though
they may provide the best therapeutic outcomes. Adult neurons,
on the other hand, are highly dierentiated, do not survive as
well as fetal neurons after transplant, and are not easily accessible. Autologous iris pigment epithelium has now emerged as a
very attractive alternative treatment approach for retinal degenerations, but more clinical trials using this tissue need to be
carried out. Neuronal cell lines may have potential because
they are easy to culture and manipulate, but their high tumorigenic potential eliminates them as a viable transplant option at
this time.
2. Surgical and mechanical considerations: During surgical implantation, tissues are directly grafted into the subretinal space
through a small retinotomy made in a posterior incision of the
sclera, underlying choroid, and peripheral retina. Retinal detachment and damage to the lens and vitreous can occur during this
procedure. Severe visual loss or other ocular complications can
result from the surgery. In this regard, the technique used should
allow for retinal reattachment in the area of transplant in a
relatively short period and small peripheral excisions that can
heal without suturing. Transplantation procedures are also
often followed by ocular infection, vasculitis, inammation,
and hemorrhaging in the eye (20). If substrates are used during
the procedure, as shown in some animal models, they should be
nontoxic, improve site-specic integration, and facilitate correct
orientation of the transplants. Thickness, degradability, malleability, and permeability of the substrates are some features
that aect the success of the transplant. For example, rigid substrates can introduce tears in the host tissue, leading to excessive
bleeding or retinal detachment. Thin biodegradable sheets of
polyglycolic and polylactic acid are reported to be optimal substrates for RPE grafts in rat retinas. They degrade easily, leaving
well-established sheets of RPE cells in the correct orientation in
the host retina. Transplant procedures, therefore, should seek to
minimize the risk of retinal detachment, ocular injuries, or other
complications that could have severe negative eect on vision.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

551

3. Graft rejection: Rejection of grafted tissue accounts for a signicant percentage of unsuccessful retinal transplants. Loss of
integrity of the RPE-retina complex and the blood-retinal barrier contributes to the rejection of transplants in the retina. It
is believed that nonclassical mechanisms of tissue rejection
may be operative in the retina, even though this tissue is
considered an immunologically privileged site. In severe retinal
degenerations, however, the retina may lose its immune protective status and elicit a strong response to allogeneic transplants, with the risks being higher in older patients.
Immunosuppressive therapy for nonneuronal allografts is
essential but is not a requirement for neural tissues. Where
autologous transplants are not possible, the use of immunosuppressive drugs in cases of severe retinal degenerations may
be essential to provide increased tissue tolerance.
4. Functional retinal circuitry: Proper host integration and orientation of transplants into target retinal sites are key factors in
predicting the ecacy of the transplant procedure.
Improvement in vision depends on correctly integrated grafts
that develop normal retinal lamination and functional synapses
with the host retina. Pigmented epithelial cells are polar in nature, with specic apical and basal characteristics that are important to their function of phagocytosis and trophic inuence.
Transplanted dissociated RPE cells often integrate randomly
into the host retina with little structural or functional polarity.
They frequently wander into the subretinal space where their
proliferation can promote retinal detachment or inhibit visual
transduction processes. Where degeneration of the retina is localized, transplantation procedures must be rened to promote
tissue integration with correct orientation and rapid reestablishment of the host retina to achieve positive visual outcome. It is
dicult to determine the success of transplant integration in
humans without histopathological information. So far, integration of grafted tissue and restoration of retinal circuitry have
only been examined and conrmed in histological preparations
of animal retinas. For morphological and histochemical evaluations, the eyes are monitored ophthalmoscopically after the
transplant procedure and eyes are enucleated at various time
points. Integration of tissue and expression of biological test
markers are visible in the preparations using light microscopy
and immunocytochemistry techniques. Lack of this information
results in somewhat speculative data regarding morphological

Copyright 2003 Marcel Dekker, Inc.

552

Tombran-Tink

and functional connections of human retinal transplant. Visual

improvement correlates of type, site, and number of transplants,
as well as sham eects on vision, are useful in determining the
success of the technology.
5. Trophic inuence: Growth factors, upregulated after mechanical
stress to ocular tissues, have resulted in a temporary delay of
retinal degenerations and a positive, transient eect on vision.
In this regard, surgically induced trophic inuence may account
for dierentiation and survival of the grafts and does not accurately reect the clinical ecacy of the transplanted tissue. In
addition, these factors may also promote neovascularization,
leading to complications that further compromise vision and
reduce the benecial eects of the transplant. Therefore, appropriate measures must be taken to evaluate the secondary eects
of trophic factors, induced by surgical manipulation, to obtain
conclusive results to validate the ecacy of a specic transplant
approach.
6. Visual improvement and stabilization: Transplanted tissues must
have the ability to improve vision. Grafts must integrate, restore
retinal circuitry, and perform functions required for long-term
visual therapy. Morphological or biochemical alterations in the
phenotype of the grafted cells can hinder the therapeutic eects
of the transplant on vision recovery. Long-term survival of the
grafted tissue must be ensured to maintain stabilized vision.
B.

Stem Cell Transplant

Stem cells are precursor cells that can give rise to dierent types of tissue in
an animal if placed in the right microenvironment. They have the potential
to undergo symmetrical as well as asymmetrical cell division. In symmetrical
division, both daughter cells retain stem cell characteristics. In asymmetrical
division, a daughter cell can either maintain the stem cell identity or initiate
specic dierentiation processes. The latter feature is of particular interest in
the eld of clinical medicine as the cells could be exploited to dierentiate
into a specic target cell type.
In humans, stem cells have an intrinsic potential to dierentiate into
any of the 210 tissues making up the body. Their potential to supply human
tissue is limitless and is a clinically attractive feature for tissue-replacement
therapy. Research in human developmental biology has led scientists to
isolate these precursor cells to study their multipotent characteristics and
application to replacing lost or nonfunctional tissues in the body. Although
stem cell transplantation studies in animal models have been in progress for

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

553

many years, it only has been recently that their application in human diseases has been investigated.
In November 1998, Thomas and Shamblott (41,42) led the pioneering
eld of utilizing human embryonic stem cells for clinical applications. Their
work was conducted independently using private funds because the cells
used in their studies were isolated from the inner cell mass of human blastocysts stage embryos. Five immortalized cell lines were developed from the
blastocysts, each with the potential to propagate indenitely or dierentiate
into any adult human cell. The cell lines met the requirements to be classied
as human embryonic stem cells. Since then, over 50 stem cell populations
have been isolated from the human blastocysts with the potential to develop
into useful cell lines.
Scientists have been encouraged by this breakthrough despite the contoversy surrounding the use of fetal tissue for developing biopharmaceuticals. The ethical dilemma stemming from embryonic stem cell study has led
researchers to focus on identifying similar precursor cells in adult human
tissues. Stem cell research initiatives now include the use of embryonic stem
cells (ES), embryonic germ cells (EG), as well as fetal and adult stem cells. In
1999, adult neural stem cells were isolated and shown to have the potential
to dierentiate in several blood cell lines. These lines can grow indenitely as
precursors in the lab for all other human cells (43) and may prove to be a
feasible option for fetal-derived stem cell populations.
The use of stem cells for tissue replacement or gene transfer in medicine is unlimited and oers hope for repopulating a tissue with cells that
have degenerated or are dysfunctional. For example, blood stem cells from
the umbilical cord show success in treatments for life-threatening diseases
such as leukemia or aplastic anemia (4446). Human bone marrow cells can
be induced to dierentiate into heart muscle cells, and this transdierentiation potential of marrow cells is signicant for developing new treatments
for millions suering from heart diseases (47). Stem cell transplant studies
are also underway to boost a patients immune response to infection (48).
The molecular mechanisms that drive stem cell plasticity are not clear.
It is obvious, though, that mechanisms must exist to balance the self-renewal
and dierentiation capacities of stem cells. Molecular technologies, such as
cDNA subtraction hybridization and microarray-based transcription proling, are now available and will allow us to dissect these intricate regenerative
and dierentiation processes in stem cells. In fact, several conserved subsets
of stem cell genes have been identied by these technologies, and they are
believed to be involved with proliferative and dierentiation mechanisms
(49). Structural and functional characterization of these genes will increase
our knowledge of how stem cells function and, perhaps, how they can be
generated from dierentiated tissue.

Copyright 2003 Marcel Dekker, Inc.

554

Tombran-Tink

Stem cell transplantation presents unprecedented opportunities for

medical advances in several life-threatening human diseases. It has the
potential to treat disorders such as heart disease, diabetes, and many
types of cancer. The possible use of adult stem cells to replace degenerating
or lost cells in the central nervous system could revolutionize neurobiology.
Neurodegenerative diseases, such as Parkinsons disease, Alzheimers disease, amyotrophic lateral sclerosis (ALS), and retinal pathologies, are prime
candidates for treatment approaches involving stem cells (50).
1. Stem Cells in the Retina
It has been generally accepted that the mature mammalian retina lacks
regenerative capacity. Contrary to this belief, several laboratories have
now reported that stem cell populations are present in the adult vertebrate
retina and that these cells can give rise to almost all of the cell types found in
the retina. Cell fatedetermination studies indicate several fundamental
principles associated with progression from progenitor to a specic adult
retinal cell. One emerging consensus model suggests that postmitotic retinal
neurons are produced during embryological development from a pool of
cycling progenitors in a process that is highly organized. Both intrinsic and
extrinsic regulators are believed to control cell fate choices during neurogenesis. The model proposes that progenitor retinal cells advance through
intrinsically determined states, during which time they are exposed to extrinsic cell fate modiers that inuence the cells decision to dierentiate into
morphologically and biochemically distinct cell types (51). Events that trigger mitotically active progenitor cells to acquire cell-specic traits will
remain an area of debate until currently available molecular tools are able
to construct seminal renewal and dierentiation pathways in stem cell function.
Studies identifying progenitor cells in the adult mammalian eye show
that they exhibit properties that classify them as stem cells. They are multipotent and have self-renewing properties. They have the capacity to dierentiate into several retinal specic cell types, including rod photoreceptors,
bipolar neurons, and Muller glia. Quiescent cells in the pigmented ciliary
margin of the adult mouse eye can clonally proliferate in vitro and form
neural spheres in the presence of broblast growth factor 2 (FGF2). The
resulting dierentiated cells express molecular markers corresponding to the
specic adult retinal cell type (52,53). In chicken, sh, and amphibians, a
zone of mitotically active proliferating cells is found in the peripheral margin
of the retina. These cells incorporate the thymidine analog BrdU and
express the cell cycle regulator proliferating cell nuclear antigen (PCNA),
indications of their proliferative capacity. The proliferating cells also express

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

555

homeodomain transcription factors, Pax6 and Chx-10, molecular marker

molecules expressed in multipotent progenitor cells. Dissociated embryonic
avian retina cells reaggregate in rotating cultures and form cellular spheres
with well-stratied retinal layers, a phenomenon that indicates the presence
of stem cell populations. The resulting dierentiated cells express markers
typical of their adult counterpart in the retina (5458). These discoveries are
biomedically relevant to transplantation strategies, in which retinal precursor cells are essential to replace degenerating retinal cells. Macular degeneration and retinitis pigmentosa are two diseases that could benet from
these ndings.
2.

Stem Cell Transplant Therapy in the Retina

With this recent knowledge, stem cell populations, derived from a number of
human organs, are being evaluated for their pluripotency in models of
ocular diseases. Already the transplantation of adult rat hippocampusderived neural stem cells (AHPCs), into ischemic retina has proven to be
successful. In this exciting study, AHPCs were shown to have the capacity to
dierentiate into cells with retina-specic characteristics. When hippocampal stem cells were injected into the eyes of adult rats subjected to ischemic
reperfusion injury, the cells were shown to integrate well into the injured
host retinas and expressed morphological and biochemical characteristics of
specic retinal cell types, as observed in histochemical preparations (59). A
similar experiment was performed to test the plasticity of retrovirally engineered AHPCs in the retina of adult and newborn rat eye. Integration into
the host retina was seen 4 weeks after intravitreal injection of the genetically
modied AHPCs. The cells developed characteristics of photoreceptors,
Mullers, amacrine, bipolar, horizontal, and astroglia at the site of injection
and were found in the correct spatial orientation in the retinal layer. They
expressed some, but not all, of the features of the specic adult retinal cell,
possibly because of the absence of early developmental cues of retinal neurogenesis (60). Long-term therapeutic benets of stem cells were also shown
for other eye tissues. For example, success in ocular surface reconstruction
and corneal wound healing was observed after limbal stem cell allotransplants and amniotic membrane transplants in patients with limbal stem cell
deciency or corneal damage (61,62).
There are limitations, however, to the use of stem cells for clinical
applications. One major disadvantage is that stem cells tend to spontaneously dierentiate in culture and lose their multipotent characteristics.
A homogeneous population of stem cells is dicult to maintain in vitro.
Cultures often contain multiple cell types because of spontaneous dierentiation in many biochemical directions. Some labs have transfected selection

Copyright 2003 Marcel Dekker, Inc.

556

Tombran-Tink

markers, such as the gene for green uorescent protein (GFP), into cultured
stem cells as an eective way to ensure that only undierentiated cells
expressing the marker will be selected for transplantation studies.
Undierentiated cells are easily identied with this approach, as only the
spontaneously dierentiated cells will downregulate the uorescent marker
molecule (63). These investigations oer unique conceptual and therapeutic
perspectives that can be developed clinically as alternate approaches to
interrupt the pathological consequences of genetic or acquired insults to
the retina.

C.

Retinal Prosthesis

The goal of retinal prosthesis is to restore vision by using highly sophisticated, miniaturized electronic devices (6478). In the normal retina, the
function of the photoreceptors is to initiate a neural signal in response to
light. The neural signal stimulates synapsing bipolar and horizontal cells
and is transmitted through the retina and output ganglion cells to the
optic nerve for interpretation in the brain. In retinal diseases in which
photoreceptor function is lost, the retina becomes insensitive to light and
visual signals cannot be transmitted. For this reason, RP patients may still
suer from blindness despite an intact optic nerve connection to the brain.
In the early stages of many degenerative retinal diseases, the inner retinal
layer is often spared, leaving healthy ganglion cells connected to the brain
through the optic nerve. Retinal prostheses are constructed to take advantage of these intact neural connections. The electronic devices respond to
light and generate signals that electronically stimulate the spared ganglion
cells, thereby activating the visual system. In this regard, the technology
aims to replace the function of the photoreceptor cells that have degenerated. Two types of retinal prosthetic devices, subretinal and the epiretinal
implants, have shown relative success as surgical implants capable of transmitting neural signals. A brief overview of the potential of these implants is
presented below.
1. Subretinal Implant
Subretinal prosthetics are surgically implanted in the subretinal space
located between the RPE and photoreceptor cells. In this location, it is
hoped that current generated by the implant in response to light stimulation
will alter the membrane potential of healthy inner retinal neurons and subsequently activate the visual system. The aim of this treatment is to mimic
photoreceptor cell function in the visual transduction cascade.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

557

Early eorts to use subretinal implants for this purpose have not
shown signicant success because of technical and microsurgical limitations
associated with the technology. Stability and biocompatibility of the
implants in the retina have also been factors limiting the development of
successful retinal prosthesis technology. In eorts to minimize these limitations, semiconductor technology consisting of thousands of silicon microphotodiode arrays has been introduced as a new subretinal prosthetic
approach (64,6871,73,74). Chow and coworkers (74) were able to show
reasonable success using semiconductor-based photodiodes to perform
photoreceptor function. In their initial studies, the device was surgically
implanted into the subretinal space of rabbits through a posterior retinotomy. The array was placed in close proximity to surviving horizontal and
bipolar cells in the retina and was sensitive to both visible and infrared light.
No external power supply was used with the device. Infrared stimuli were
used to distinguish between implant- and native photoreceptormediated
response. Recordings and histological examination of the retina indicated
that the implants maintained a stable, functional position in the subretinal
space. Data obtained from electrophysiological recordings showed that the
microphotodiode arrays generated current in response to light, and this was
transmitted through the output ganglion cells to the brain. A major setback
to the study was the signicant degeneration of retinal cells in the area of the
implant. This was due to the fact that nutrients from the choriocapillaris
were unable to reach the retina because of the nonpermeable nature of the
gold electrodebased implant. The remainder of the retina architecture
remained intact, and other side eects were not promoted by the implants.
The construction of devices that would minimize blockage of nutrients to
the retina is a current focus in the eld.
2.

Epiretinal Implant

Eorts to develop an independently functioning epiretinal device for

patients with outer retina degeneration are also promising for restoring
simple basic visual perception (67,72,7578). It is a similar concept to the
subretinal prosthesis. Epiretinal devices are constructed, however, to utilize
the output capacity of spared ganglion cells to transmit the signals they
generate to activate the visual system. They are designed to rest on the
inner surface of the retina rather than being implanted in the subretinal
space, where they can utilize the functional output potential of the ganglion
cells and optic nerve to stimulate the visual cortex (Fig. 4).
The advantage of the epiretinal approach over subretinal implants is
that the ganglion cells can be easily stimulated because of their accessibility
in the inner surface of the retina. These implants have shown success in an

Copyright 2003 Marcel Dekker, Inc.

558

Tombran-Tink

Figure 4 Patients with retinitis pigmentosa and macular degeneration become

blind when the rods and cones degenerate and no longer convert incoming light
into electrical impulses. While other parts of the retina remain healthy, the brain
stops receiving impulses required to provide vision. The retinal prosthesis is designed
to bypass the lost rods and cones and directly stimulate the surviving ganglion cells
that are connected to the brain through the optic nerve. Specially designed glasses
capture the visual scene and transmit this information into the eye through an
invisible laser beam. The laser strikes a solar panel (photodiode array) located within
the pupil to generate internal power. An ultra-thin electrode array carries the power
to the retinal surface, where it stimulates the ganglion cells. (Courtesy of Dr. Joseph
F. Rizzo, Harvard University, USA.)

initial study that used at platinum microelectrode arrays, embedded in a

thin lm of polyimide, as implants. The device was stabilized onto the inner
retina surface of cat eyes using cyanoacrylate adhesive. Recordings of neuronal activities in the visual cortex indicated electrical responses to retinal
stimulation experiments. Neither retinal detachment nor intraocular inammatory responses were observed after prolonged periods of implantation
(75). The retina microelectronic prosthetic eld has expanded considerably
in the last 5 years. Initial studies are exciting, and even though ambitious,
the method is proving to be a feasible approach to restoring vision. Several
types of retinal prosthesis have now been used to electrically stimulate the
visual system to produce ashes of visual perception in blind individuals.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

559

The technology is available for modication and microfabrication of less

complex, sophisticated devices that promise hope for mechanically restoring
vision. The challenges, however, that face the technology are numerous and
formidable. Major obstacles facing the visual prosthesis technology include
(6568,71,77):
1. Establishing a functional connection between the implant, the
neural retina and the brain
2. Detachment of the retina during microsurgical implantation
3. Inammatory reactions of the retina against the implanted
devices
4. Ocular infection following implantation procedures
5. Rejection of the electronic device
6. Wiring the retinal surface and maintaining mechanical stability
of an epiretinal or subretinal implant
7. Epiretinal devices may stimulate formation of retinal surface
membranes, which, in turn, could dislodge the implant and create an electrical resistance between the electrodes and the target
output cell
8. Subretinal implants could stimulate the RPE cells to migrate into
the retina, causing brosis, or impede the transport of nutrients
and trophic factors from the choriocapillaris to the retina, resulting in further degeneration of the retina
9 Surviving photoreceptor cells could degenerate at the site of subretinal implants because the devices may block choroidal circulation to the outer retina.
It is possible that photoreceptor function could be replaced by retinal
prosthesis? Could microelectronic devices establish a long-term functional
connection between the inner retina and the brain? Could vision be restored
to the blind? The preliminary studies have suggested that the use of retinal
prosthesis to restore visual loss is possible if the mechanical and surgical
limitations are reduced, precise electrical stimulation and integration with
the brain are achieved, and long-term biocompatibility of the implants with
the retina is attained. At the present time, however, the clinical usefulness of
visual microelectronic prosthesis still remains at a conceptual level.
D.

Gene Therapy

Millions of years of evolution have been dedicated to the development of

ecient viral delivery systems for transfer of genes within and across species.
Viruses have developed highly sophisticated mechanisms for cellular tropism and nuclear targeting. In the host cell they replicate by using the host

Copyright 2003 Marcel Dekker, Inc.

560

Tombran-Tink

transcription and translation machinery to express, replicate, package, and

release their genetic material to infect surrounding tissues. Many viruses
have developed strategies to recombine with and stably integrate their
genetic material into the host genome. Others do not integrate into the
host DNA but replicate their genetic material as episomes.
These viral properties have specic advantages in gene therapy, and
considerable interest has been expressed in utilizing the natural biological
properties of viruses in molecular medicine. The groundbreaking discovery
by Hershey and Chase in 1952 (79) that DNA is the genetic material, followed by the elucidation of the double helix and the genetic code (80,81),
were keys to the subsequent characterization of viral and cellular genomes.
These discoveries led to an increased understanding of the genetic disposition of many human diseases and an explosion in gene identication and
characterization. The genetic basis of many human diseases will continue to
expand with the completion of the Human Genome Project and, with it, a
rapid increase in the formulation of novel therapeutic approaches for many
diseases.
Several DNA-based therapies, using both viral and nonviral delivery
systems, have been developed and show varying degrees of success in clinical
trials. These represent a signicant percentage of current strategies in clinical
trials, especially in the area of cancers. It is interesting to read the earlier
literature on the possible application of gene therapy in humans and the
potential problems arising from the technology. For example, it was initially
thought that the success of gene therapy would be dependent on its potential
to correct the genetic defect in every cell associated with a particular disease.
This assumption is no longer valid, as studies have now shown that even
transient expression of some therapeutic genes in a small percentage of
target cells is sucient to slow the progression of a disease. Many concerns
still surround the technology, and many questions still remain unanswered,
but there is great optimism for the future of DNA-based therapies, especially for diseases that are currently incurable. In this section, I will present
an overview of gene therapy applications, with emphasis on this approach as
a practical treatment modality for retinal degenerations.
1. Gene Therapy Applications
Each individual carries a number of defective genes unknown to the carrier
unless the disease symptoms are expressed. This is because each of us carries
two copies of nearly all genesone inherited from our mother and the other
from our father. The exception to this are genes located on the Y chromosome, found only in males. In many cases if we carry only one defective copy
of a gene, the other, normal copy will assume the function of its defective

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

561

counterpart, and its expression alone will be sucient to mask the symptoms of the diseases. Such a gene and the associated disease are referred to
as recessive. An individual carrying such a gene is considered a carrier.
The disease phenotype will only develop if both copies of the gene are
defective in the same individual. The gene for sickle cell anemia is an example of a recessive gene. The disease aects the normal function of the red
blood cells and is inherited as an autosomal recessive trait. In other cases, a
single defective copy of a gene alone can cause expression of the disease
phenotype. These genes are referred to as dominant. An example of such
a disease is Huntingtons disease, which is inherited as an autosomal dominant disease aecting the nervous system.
Not all defective genes produce harmful eects. The environment in
which a gene operates contributes to its function. Some defective genes may
actually provide us with a selective advantage of survival in a particular
environment. For example, if both copies of the sickle cell gene are defective
in a person, a defective blood protein is produced and that individual will
suer from the disease. If only one copy of the gene is defective, the individual is a carrier of the disease but does not express the disease symptoms.
A remarkable feature of the sickle cell gene is that a person carrying one
defective copy is resistant to infection by malaria parasites and, therefore,
possesses a distinct survival advantage in environments where malaria is
endemic.
Approximately 1 in 10 individuals has or will develop an inherited
genetic disorder. Some diseases are caused by more than one defective
gene, while others are the result of a single defective gene. Hundreds of
specic disease conditions caused by mutations in only one gene have
been identied. These are referred to as single gene disorders, or monogenic
diseases, and are currently the best targets for gene therapy. Cystic brosis is
a monogenic disease in which a mutation in a single gene results in the
disease phenotype. Other diseases, such as retinitis pigmentosa, a retinal
degenerative disease causing blindness, are complex and are associated
with genetic defects in a number of genes. These are referred to as polygenic
disorders. Replacement of a single defective gene that contributes to this
phenotype may not necessarily result in curing the disease but may slow
down its progression.
Theoretically, if defective genes are prevented from producing defective proteins and a normal functioning protein can be put back into the
tissue, a disease phenotype should be reversed or the disease progression
slowed. This very simplistic assumption is the basis of gene therapythe
introduction of genetic material into an aected cell to correct an existing
disorder or to introduce a new function into cells with resulting therapeutic
benets.

Copyright 2003 Marcel Dekker, Inc.

562

Tombran-Tink

A disease may result from the production of a defective protein or the

overexpression, underproduction, or absence of a normal protein. Of all the
therapies currently on the horizon, perhaps the most tantalizing of them are
ones that will repair a single mutated gene or that will replace its defective
function. A target cell could also be provided with a new function that
allows it to ght o a disease or infection. The goal of gene therapy is not
only to introduce a therapeutic gene into diseased tissue, but also to control
its level of expression with a precision not likely to be achieved using other
methods and to cure or retard the progression of the disease after a single
administration of the gene. Current applications of gene transfer technology
aim to:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

Replace the defective gene in recessive diseases

Inactivate a gene in dominant diseases
Express a cell rescue factor
Express a protein that will inhibit the progression of a disease
Augment or downregulate the activity of an existing gene
Transfer suicide or toxic genes that mediate apoptotic
events that will selectively kill unwanted proliferating cells
Modulate the immune system in an eort to ward o the disease
Transfer drug-resistant genes
Confer sensitivity to an inert prodrug
Interrupt the life cycle of an infectious disease

At the present time, gene therapy approaches will probably be most

successful in diseases where patients are unable to synthesize a specic
protein and where the target gene is introduced into a single organ. If the
target tissue is somatic, the eects that result from the gene transfer intervention will only occur in the individual undergoing the therapy. In contrast, germ line therapy will not only aect the individual being treated, but
will also be evident in the ospring of that person. Germ line gene therapy
has the potential to eliminate many crippling hereditary disorders, however,
it is ethically unacceptable at this stage and may not be attempted until the
hurdles of somatic gene manipulations have been successfully dealt with.
While many limitations exist for gene therapy, a wide range of human
diseases, both familial and nonfamilial, are good candidates for this
approach.
2. The Retina as a Candidate Tissue for Gene Therapy
Ophthalmic involvement in the genetic basis of retinal disease has increasingly advanced due to the recent explosion in biotechnology. Ocular pathologies have been recognized for centuries, but many are now being

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

563

reclassied as their genetic etiologies undergo intense investigation at the

molecular level. This increases the potential for developing new DNA-based
treatments for inherited and acquired eye diseases.
The etiologies of blinding diseases, such as AMD, are often obscured
by secondary complications and therefore diagnosed at a late stage when
adjacent ocular tissues are also aected. The impact of gene transfer intervention may be signicant in the management of such diseases as we continue to understand early promoting mechanisms underlying their
progression. Prerequisites for eective DNA-based intervention in the eye
include:
1.
2.
3.
4.

Knowledge of the etiology of the disease

Characterization of the gene(s) associated with the disease
Eective delivery of the therapeutic gene
Low toxicity and immunological tolerance of the gene transfer
vector
5. Targeting transgenes to the specic aected area in the eye
6. Expression of the delivered gene at a therapeutic level
7. Long-term gene expression for reversing or correcting the disease
Diseases of the retina are good targets for gene transfer approaches.
Visual function hinges on two critical layers of the retina: the outer nulcear
layer, which contains the photoreceptor cells, and the retinal pigment epithelium, which lies adjacent to the photoreceptors. The relationships between
the RPE and photoreceptor cells is critical not only to normal function but
also to the pathology of several blinding hereditary diseases, including retinitis pigmentosa, Lebers congenital amaurosis, and macular degeneration.
Delivery of a therapeutic product to the RPE and photoreceptor cells
is likely to be most successful if the delivery system is designed to target
those specic cells. Because of the many compartments and structural complexity of the eye, it is easy for innoculated material to get trapped and
diused in any of the eye spaces and cell layers, reducing the availability
of the transgene to target sites. In addition, the expression of certain genes is
controlled through spatial and temporal regulation. Administration of high
vector titers is, therefore, necessary to achieve reasonable levels of gene
expression at the target site for attenuation of the disease. Attempts are
currently underway to minimize widespread eects of an exogenous gene
by using tissue- or cell-specic promoters to achieve regionally restricted
gene transfer.
Some of the most widely used vectors in ocular gene delivery include
the retrovirus, the adenovirus, and adeno-associated viruses. The choice of
viral vectors, the design of ecient delivery systems, and the development of

Copyright 2003 Marcel Dekker, Inc.

564

Tombran-Tink

well-characterized animal models of retinal diseases are key elements for

gene therapy to be successful in ocular pathologies.
3. Viral Vectors
Advances in genetic engineering and recombinant DNA technology have
made it possible to use viruses as clinical tools to transfer therapeutic genes
to treat human diseases. Viruses have developed eective mechanisms that
promote their survival and replication in a host cell. These intrinsic mechanisms allows them to (a) evade host immune surveillance, (b) attach to the cell
membrane, invade the cell, and enter the nucleus, and (c) direct the host cell
replication machinery to allow viral multiplication. They are armed with
features that meet gene therapy criteria. Research in the development of
gene transfer vectors has, thus, focused on the use of these viral properties
to facilitate the delivery of therapeutic genes to the nucleus of the cell where
the genes can be amplied and expressed.
Hundreds of viruses have been identied and their genome and structure well characterized. This knowledge has allowed researchers to select
those viruses with high tropism and low toxicity for the use in gene therapy.
Many viruses, however, can only infect a limited number of cell types. Some
viruses are pathogenic and will elicit a strong host immune response due to
the production of viral proteins required for infectivity, major disadvantages
to current gene therapy protocols. Other criteria essential to using a viralbased delivery system for therapeutic DNA include:
1. Stable gene transfer and expression
2. Long-term gene expression where a therapeutic product is
needed to attenuate the disease
3. The use of replication-decient viral vectors
4. Minimal intrinsic toxicity of the input vector and other associated pharmacological toxicity
5. High eciency of transfectiona critical number of the transgene must enter a sucient number of aected cells to elicit a
signicant biological response
6. Optimization of the delivery system to limit reinoculation
7. Use of vectors that do not result in insertional mutagenesis by
activating oncogenes or by inactivating of other essential genes
Genetic engineering designs to improve the safety and eciency of
viral gene transfer aim to manipulate the host immune system to reduce
immunological response to viral proteins, minimize infection by impairing
the viral replication machinery to limit the viral particle to a single infectious
cycle, and modify viral tropism and increase viral target specicity by alter-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

565

ing viral envelope proteins. Designing second-generation viruses with

decreased pathogenicity and high infectivity will increase the success of
gene therapy in biomedical research.
A recombinant viral vector can be constructed by cutting the viral
DNA and the gene of interest at specic sites with a restriction enzyme.
The cut pieces are subsequently ligated with a DNA ligase to create a
recombinant vector. Before the cutting and ligation procedure, the viral
genome is rst modied to remove key regulatory genes. This will render
the virus replication decient once it infects a host cell and create enough
space for ligation of the new gene. Other accessory genes may be removed
from the viral DNA to make the vector less immunogenic. Transfection of
the recombinant viral genome into trans-complementing cells lines, stably
transformed with the deleted regulatory coding cassettes from the viral
genome, allows for replication and packaging of the recombinant viral
DNA in vitro. Administration of a titer of recombinant virus obtained by
this method should result in viral infection of the target tissue and expression of the inserted gene without viral replication or production of infectious
progeny viruses in vivo (Fig. 5) (82,83).
a. Retrovirus. Retroviruses are RNA viruses that infect a wide
variety of eukaryotic cells. The retroviral genome codes for three core
genesgag, pol, and envwhich are anked on either side by a longterm repeat (LTR) sequence (Fig. 6). The gag gene, located in the 5 0 region of the genome, encodes the viral core proteins. The pol gene is the
second gene in the cassette and codes for three products: reverse transcriptase (RT), which allows the viral RNA to be transcribed into DNA
after entry into a host cell; viral integrase (Int), which facilitates integration of viral DNA into the host DNA; and viral protease, which acts
on the viral core proteins. The env gene is 3 0 of the pol gene and encodes the glycosylated viral envelope proteins, which determine the tropism of the virus.
The LTRs ank both ends of the genome and contain cis-acting
sequences that regulate viral replication, transcription, and integration
into the host genome. A packaging signal, which facilitates assembly of
viral particles, is found immediately 3 0 of the 5 0 LTR. When a retrovirus
infects a cell, its RNA is reversed-transcribed into DNA by RT. The DNA
subsequently integrates into the host chromosome as a provirus and directs
the synthesis of viral proteins by the host transcription and translation
machinery. While most retroviruses are nonpathogenic, a few have been
known to cause cancer. For example, the Rous sarcoma virus causes tumors
of the connective tissue in chickens. The tumorigenic properties of these
viruses have been associated with accessory genes called oncogenes.

Copyright 2003 Marcel Dekker, Inc.

566

Tombran-Tink

Figure 5 Viral packaging: (1) transfection of the vital genome into the packaging
cell to produce viral proteins; (2) transfection of recombinant viral DNA into the
trans complementing packaging cell line and replication of the viral DNA; (3) release
of infectious particles.

The most common retroviral vector used for gene transfer is the
Moloney murine leukemia virus (MoMuLV). The viral genome is modied
by deleting the three core genesgag, pol, and envbefore recombination
with the target gene. The deleted segment of the viral genome is then
replaced with the therapeutic gene of interest, and the recombinant vector
can be grown to therapeutic titers by transfecting it into a suitable packaging cell line that has been stably transformed with the deleted viral core
gene cassette. Replication and viral assembly proceed similar to infection
with a wild-type virus (8486). Recombinant viruses maintain their high
targeting eciency and stably integrate into the host genome. While long-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

567

Figure 6 Schematic representation of a basic retrovirus genome showing the age,

pol, and env genes, which are anked by the long-terminal repeats (LTR). A packaging signal ( ) is located next to the 5 0 LTR.

term gene expression is increased with stable integration, the potential risk
of activating protooncogenes or inactivating essential host genes at the site
of host genome integration are concerns when using this vector is gene
therapy approaches.
New engineering technologies are being developed to modify the tropism of retroviruses by altering the envelope proteins. Such modications are
designed to facilitate the cell-specic targeting by the virus. Retroviruses,
however, do not infect postmitotic cells. They require the active process of
mitosis for entry into the nucleus, a feature that does not make them a
suitable delivery system for post-mitotic tissues. The use of retroviral vectors
in gene transfer therapies is also limited because of the size of the transgene
insert that they can package and the low viral titers obtained in vitro.
Currently, therapeutic genes larger than 910 kb cannot be stably cloned
into the retroviral genome (Table 1). In addition to these limitations, murine
retroviruses are sensitive to inactivation by the human complement lysis
mediated pathway. Current research to minimize this sensitivity is focused
on the use of human packaging cells lines or modication of the viral envelope proteins to produce complement-resistant retroviral pseudotypes.
Despite these constraints, retroviruses are still the most widely used vectors
in current gene delivery systems and account for approximately 40% of all
gene therapyrelated clinical trials. Stable expression of many therapeutic
genes delivered by retroviruses has been observed for periods of over one
year.
Several studies have shown a clinical potential of retroviruses as a gene
delivery vehicle in the eye. In one report, modied retroviruses encoding
either a urokinase-type plasminogen activator (u-PA) or a tissue-type plasminogen activator (t-PA) gene was successfully transfected into human RPE
cell cultures. Ten weeks after infection, transfected cells were co-cultured
with umbilical vein endothelial cells (HUVECs) and their eects on cell
proliferation and wound healing assessed. Both transgenes were observed
to express large amounts of biologically active products in the RPE cells.
The expression of u-PA promoted HUVEC cell proliferation and wound
healing, while t-PA or the control, noninfected RPE cells in nonconuent
cultures had no signicant eect. Other retroviral constructs, designed to
carry an internal opsin promoter fragment, were able to specically direct

Copyright 2003 Marcel Dekker, Inc.

568

Table 1 Comparison of Vectors for Gene Therapy

Structure
Viral vectors
Retrovirus

Advantages
Moderate gene transfer eciency
Long-term gene expression

Adenovirus

Nonenveloped
Genome size  36 kb

Adeno-associated virus
(AAV)

Nonenveloped
Genome size  4:7 kb

High gene transfer eciency

Infects quiescent and
replicating cells
High titer  1011 1012 kb
High gene transfer eciency
Infects quiescent and
replicating cells
Long-term gene expression

Herpes simplex virus

(HSV)
Non viral vectors
Liposome
DNA complexes or naked
DNA

Enveloped
Genome size  152 kb

Copyright 2003 Marcel Dekker, Inc.

Large insert capacity  150 kb

Infects quiescent cells
Nonreplicating
Accommodate large therapeutic genes

Integrates into host DNA

titer  108 109
Insertional mutagenesis
Infect replicating cells only
Insert capacity  910 kb
Nonintegrating
Transient expression
Immunogenic
Insert capacity < 7 kb
Integrates into host DNA
Insertional mutagenesis
titer  109 1010
Contamination with
helper virus
Insert capacity  4:5 kb
Nonintegrating
titer  108 109
Low gene transfer eciency
Low gene transfer
eciency

Tombran-Tink

Enveloped
Genome size  10 kb

Disadvantages

Experimental Approaches to Retinal Diseases

569

gene expression in retinal photoreceptor cells after transduction of the cells

(87,88). The use of retroviruses to transfer genes that modify the angiogenic
potential of RPE cells or to rescue degenerating photoreceptor cells shows
potential for gene therapeutic applications in retinal diseases associated with
these processes.
Retrovirus-mediated transfer of suicide genes has also been tested in
several ocular disease models to determine transduction eciency of the
virus and feasibility of this approach as a treatment strategy. In one
study, an animal model that developed posterior capsule opacication
(PCO) was used to test the eect of the suicide gene, HSV-tk, and the
subsequent sensitivity of the transduced cells to ganciclovir (GCV). PCO
is a major complication of cataract surgery and develops from hyperplasia
of the lens epithelium following phacoemulsication. Retrovirus-mediated
transfer of the suicide gene, herpes simplex virus thymidine kinase (HSV-tk),
into rabbit lens epithelial cells resulted in ecient transduction of the gene
after in vitro and in vivo administration. While the transduced cells were
reported to be sensitive to GCV treatment after transduction, capsule opacication was not attenuated. The study, however, shows promise for the
use of retroviruses to transfect lens epithelial cells. In this case, modication
of the vector to increase sensitivity to GCV may be required for successful
alleviation of PCO (89).
A similar experiment was performed in vitro using human RPE cells
(HRPE), a major component of the membranes found in proliferative
vitreoretinopathy (PVR). In PVR, the RPE cells, broblasts, and other
proliferating cells form contractile membranes in the vitreous cavity of
the eye, which result in traction retinal detachment. The study was aimed
to test the transduction eciency of HSV-tk recombinant retroviral vector and the bystander eect after treatment with GCV as a possible
treatment strategy for PVR. Stable transfection of HRPE and broblast
cells and a strong bystander eect after GCV treatment were reported in
vitro (90). Transduction of the cells in experimental PVR-induced rabbits
was lower compared to the number of cells transduced in vitro, but a
strong bystander eect was observed for HSV-tkpositive cells along with
a signicant reduction in the severity of PVR in the animals (91,92).
Human and rabbit keratinocytes were also stably transduced by retroviral-mediated transfer of HSV-tk gene in a study that tested the ecacy
of this treatment in corneal wound healing after supercial keratectomy.
Topical application of supernatant containing the vector was administered after superical stromal keratectomy in rabbits. Transduction of
the rabbit keratinocytes was conrmed, with an eciency of approximately 2040%. Corneal stromal haze and scarring were signicantly
reduced after postoperative application of topical GCV (93).

Copyright 2003 Marcel Dekker, Inc.

570

Tombran-Tink

Polycations, such as polybrene and protamine sulfate, were reported to

signicantly improve the transduction eciency of the gene transfer into
keratocytes (93).
Other genes have been successfully transferred by retroviruses into
cells that are involved with retinal neovascularization. For example, retroviral-mediated transfer of tissue inhibitor of metalloproteinase-2 (TIMP-2)
into bovine choroidal cells resulted in the production of biologically active
TIMP-2 and a subsequent increase in angiogenic response to VEGF (94,95).
Adjuvant anti-angiogenic gene transfer in diabetic retinopathy, a major
cause of acquired blindness associated with retinal neovascularization and
subsequent traction retinal detachment, into photocoagulation sites of the
retina shows feasibility for retroviral-mediated gene therapy in this disorder
(96). While photocoagulation therapy delays visual loss, it does not prevent
progression to blindness. These innovative approaches indicate the versatile
potential of retrovirus-mediated gene transfer to attenuate PVR, to modulate corneal wound-healing mechanisms, and as possible treatments for
choroidal and retinal neovascularization.
b. Adenovirus. Increased clinical interest has focused on adenoviruses as vectors for gene transfer because of their broad host range infectivity in both postmitotic and proliferating cells. Forty-seven different
serotypes of human adenoviruses have been identied with a range of primary target site of infection, including the respiratory epithelium, eye,
gastrointestinal, and urinary tract cells. Adenoviruses are currently the viral gene delivery vehicle of choice for many retinal diseaserelated protocols.
Structurally, adenoviruses are not bound by an envelope like the retroviruses. They are encapsulated by a protein shell, which forms an icosahedral capsid structure around the DNA inner core. The capsid contains three
groups of proteins: (a) the hexons, which are structural proteins that associate as trimers called hexomers (trimeric hexons arrange on the planes of the
icosahedral structure of the virus); (b) the pentons, which associate as pentomers and penton complexes to form the 12 vertices of the icosahedrons; (c)
the bers, which assemble as trimeric bers and associate with the penton
vertices to form protruding spikes. The penton and ber subunits constitute
the infection machinery of the adenovirus. The protruding trimeric bers
mediate the initial attachment of the virus to cell surface receptors.
Subsequent internalization of the virus is mediated by its pentameric base
and the integrins on the cell surface (97). A new class of articial, silk-like
brous material is being evaluated in eorts to improve adenoviral tropism.
The adenovirus genome is a single linear double-stranded DNA molecule approximately 36 kb in length. The genome is divided into two func-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

571

tional noncontiguous, overlapping domains, which contain early (E1A,

E1B, E2, E3, E4) and late (L1L5) expressed genes. The regions are so
dened because of the onset time of transcription of the viral genes after
entry of the viral DNA into the host cell nucleus. The early and late genes
are anked by inverted terminal repeats (ITRs) that are important to replication of the viral genome and an encapsidation signal, located adjacent to
the 5 0 ITR, that is essential for viral packaging (98). Deletion of nonessential
viral genes, such as the E1 genes, generates a replication-decient viral
vector into which a passenger gene can be inserted. The size of foreign
DNA that can be packaged by the virus after E1 deletion is approximately
7.8 kb. The small DNA packaging size of the virus limits its use as vector for
the transfer of larger therapeutic genes. E1 genes are the rst to be transcribed upon entry of the viral DNA into the nucleus because their products
are essential to initiate expression of other key viral replication genes.
Deletion of the E1 gene cassette, therefore, interrupts the virus infectious
cycle by disabling the replication, assembly, and packaging machinery of the
virus. Furthermore, modication to the viral genome limits the recombinant
vector to one cycle of infection. Using recombinant DNA technology, the
deleted viral function can be rescued by transfecting the modied recombinant vector into transformed E1 complementing packaging cells that constitutively express E1 gene products. The recombinant replication-decient
virus can then be grown and concentrated to the high titers required for
therapeutic doses. After transfection into the target host cell, replication of
the adenovirus DNA is independent of the host cell replication. The viral
DNA rarely integrates into the host genome and resides as an episome in the
nucleus, a feature that makes adenoviruses a safe gene transfer vector
because oncogene activation and gene toxicity are minimized. Transgene
expression usually occurs within 24 hours of inoculation and can be determined by assays, such as polymerase chain reaction (PCR), ELISA,
Western, and Northern blotting technology (99,100).
The clinical advantages that makes this an attractive vehicle for gene
transfer include the relative ease with which the viral genome can be
manipulated to generate stable recombinants, low pathogenicity, and generation of high viral titers in vitro. Major concerns that limit the use of
adenovirus-based vectors include transient gene expression, limited size of
the transgene insert, and immunogenicity. Thus far, gene delivery using
retroviruses has proven to be very ecient, but long-term gene expression
is relatively poor (Table 1) and is often caused by the pathology at the site
of gene transfer, which could promote the loss of transduced cells in the
region. Repeated administration of the recombinant vector is not advisable
because of the high immunogenicity of the viral proteins. Therefore,
improvements in vector design to produce less immunogenic, second-gen-

Copyright 2003 Marcel Dekker, Inc.

572

Tombran-Tink

eration adenoviruses or ones that can evade the host immune response will
be signicant advancements in the use of this vector in broad-scale gene
therapy.
Adenoviral vectors can infect a few retinal cells, including RPE. They
have not been reported to be infectious agents for normal mature photoreceptor cells but successfully transduce developing or degenerating photoreceptor neurons. Adenoviruses are widely used in gene therapy protocols in
animal models of retinal degenerations. For example, a recombinant adenovirus carrying the wild-type b subunit of the cGMP gene has shown
eciency of transfection and transient rescue of degenerating photoreceptors in the rd mice. Protection of RPE cells against hemoglobin toxicity has
also been observed after infection with an adenoviral vector encoding the
human heme oxygenase (HO-1) gene. Heme oxygenase is an antioxident
produced by RPE cells. Transfected RPE cells overexpressed the HO-1
gene product more than threefold compared to nontransfected cells and
showed increased survival rates (93%) compared to controls (6575%)
when challenged with hemoglobin (101). Adenovirus-mediated transfer of
HO-1 may be of clinical relevance in retinal pathologies requiring cellular
antioxidant defense mechanisms.
In addition, adenoviral-mediated transfer of genes encoding trophic
factors has proved to be a useful strategy in promoting photoreceptor survival. Limitations of intraocular administration with therapeutic doses of
soluble trophic factors are often associated with the pharmacokinetics of the
gene product and the need for multiple injections for eective long-term
benecial changes to the retina. In this regard, transfer of neuroprotective
and antiangiogenic genes into the retina may be a more ecient approach
than direct injection of some soluble survival factors into the eye (102,103).
We have identied a 50 kDa neurotrophic protein, pigment epitheliumderived factor (PEDF), secreted by RPE cells, which is widely shown to
have neuroprotective and antiangiogenic activities (102,104,105).
Injections with the soluble protein have shown enormous therapeutic potential of PEDF to inhibit neovascularization and promote photoreceptor survival. However, a single administration of adenoviral-mediated transfer of
the PEDF gene in the eye has shown promise for the use of this approach in
retinal degenerations (106). In a recent study, a recombinant adenoviral
vector encoding the PEDF gene was injected into the vitreous of a mouse
angiogensis model, in which choroidal neovascularization was promoted by
laser-induced rupture of Bruchs membrane. A signicant correlation
between increased expression of the PEDF gene and reduction in choroidal
neovascularization was observed after transfection with the recombinant
adenovirus. Similar results were obtained with the PEDF gene in two
other mouse models of retinal neovascularizations: (a) transgenic mice

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

573

that overexpress vascular endothelial growth factor (VEGF) in photoreceptor cells, and (b) mice with oxygen-induced ischemic retinopathy. In both
cases, increased intraocular expression of PEDF, after transduction with the
recombinant vector, resulted in statistically signicant inhibition of retinal
neovascularization when compared to mice injected with null vectors.
Adenovirus-mediated transfer of the VEGF receptor shows a similar reduction in the growth of choroidal blood vessels in the retina. The recombinant
virus was constructed with the human VEGF receptor, t-1, which was
fused to the Fc portion of IgG (Adt-ExR vector). Subsequent transfection
of Adt-ExR into pigmented rats, in which subretinal neovascularization
(SNR) was induced by intense photocoagulation of the retina, resulted in
the inhibition of SNR with hybrid t-1 soluble receptor overexpression in
the subretinal space. The inhibition of neovascularization in the eye by
transfer of PEDF and the soluble VEGF receptor genes represents new
advances in the treatment of blinding diseases associated with apoptotic
and angiogenic mechanisms (107).
Adenoviral-mediated transfer of other neurotrophic genes also
demonstrates neuroprotective eects in the retina. Transfection with an
adenovirus encoding ciliary neurotrophic factor (CNTF), in retinal degeneration slow (rds/rds) or rd mice, resulted in constitutive expression of the
CNTF transgene. Promotion of photoreceptor outersegment formation,
increased expression of rhodopsin photopigment, decreased apoptotic
events, and an increase in the amplitude of a- and b-waves of scotopic
electroetinograms were some of the eects of the overexpressed CNTF
transgene (108,109). Similarly, administration of adenoviral vectors carrying
bPDE, basic broblast growth factor (bFGF), or the bcl-2 gene has shown
success of the approach as a therapeutic strategy to slow the course of
retinal degeneration (110113). Modications in the adenovirus genome to
create second-generation constructs appear to enhance the eectiveness of
the virus in ocular gene therapy. In one modication, the vector was constructed by deleting all of the essential viral genes, resulting in an encapsidated adenovirus mini-chromosome (EAM), which only contained the ITRs
for replication and the encapsidation signal necessary for packaging. EAMmediated delivery of bPDE not only eectively transduced retinal cells in the
rd mice, it also promoted prolonged transgene expression, increased survival
of photoreceptors, and resulted in decreased toxicity when compared to the
rst-generation viruses (114). Design of second-generation viruses with low
toxicity, eective gene transfer mechanisms, and long-term transgene
expression capabilities will be advantageous to the future of this vector in
ocular gene therapy.
In addition to transfection of photoreceptor and RPE cells with adenoviral vectors, retinal ganglion cells (RGCs), corneal epithelium, and con-

Copyright 2003 Marcel Dekker, Inc.

574

Tombran-Tink

junctival epithelium are also target tissues into which therapeutic genes can
be transferred with recombinant adenovirus. In an RGC model of degeneration, approximately 80% of RGCs could be induced to undergo apoptosis and degenerate following intraorbital transection of the optic nerve. A
single dose of adenovirus encoding brain-derived neurotrophic factor (AdBDNF) coinjected with a free radical scavenger, N-tert-butyl-(2-sulfophenyl)-nitrone (S-PBN), resulted in the survival of 63% axotomized RGCs,
indicating the clinical usefulness of the approach for treating RGCs following optic nerve transection (115,116). Similar strategies were tested in the
management of corneal and conjunctival abnormalities. Adenovirus type 5
(Ad 5) vector is reported to successfully deliver the reporter lacZ gene to
these tissues in humans and rats. Maximum lacZ expression occurred 27
days after inoculation. Moreover, the nonspecic upregulation of the
inammatory cytokines IL-6, IL-8, and ICAM-1 in the tissues, induced by
Ad5 infection, was suppressed with betamethasone, thereby allowing longerterm transgene expression (117). Some groups have found that other combinations, such as coinjection of E1-deleted AV vectors carrying the lacZ
reporter gene with a modied adenovirus encoding a secreted immunomodulatory molecule (CTLA4-Ig), could signicantly reduce the immunological consequences of gene transfer with adenoviruses and, thus, promote
prolonged transgene expression (118,119).
Corneal opacity, a condition associated with the expression of TGF-b,
is another serious cause of visual loss. The accessibility of the cornea has
encouraged many in the eld to turn to gene therapy alternatives to reduce
this condition. An example of the potential usefulness of gene therapy
approach for treating corneal opacity is shown in a study that used an
adenoviral vector encoding a fusion gene containing the human type II
TGF-b receptor and the Fc fragment of human IgG (AdTbeta-ExR).
Transfection with the recombinant vector has proven successful in the
expression of a soluble TGF-b receptor in Balb/c mice (120). High levels
of the soluble receptor were found in serum and ocular uids for at least 10
days after AdTbeta-ExR injection into the femoral muscle of the animals.
Furthermore, the overexpression of soluble TGF-b receptor inhibited TGFb signaling and may have resulted in the reduced corneal opacity observed in
mice subjected to silver nitrate-induced corneal injury. Angiogensis and
edema were also reduced in the injured corneas
Corneal opacity, a condition associated with the expression of TGF-,
is another serious cause of visual loss. the accessibility of the cornea has
encouraged many in the eld to turn to gene therapy alternatives to reduce
this condition. An example of the potential usefulness of gene therapy
approach for treating corneal opacity is sown in a study that used an adenoviral vector encoding a fusion gene containing the human type II TGF-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

575

receptor and the Fc fragment of human IgG (AdTbeta-ExR). Transfection

with the recombinant vector has proven successful in the expression of a
soluble TGF- receptor in Balb/c mice (120). High levels of the soluble
receptor were fond in serum and ocular uids for at least 10 days after
AdTbeta-EXR injection into the femoral muscle of the animals.
Furthermore, the overexpression of soluble TGF- receptor inhibited
TGF- signalling and may have resulted in the reduced corneal opacity
observed i mice subjected to silver nitrateinduced corneal injury.
Angiogenesis and edema were also reduced in the injured corneas with the
overexpression of TGF-b. Taken together, the results from these experiments suggest that adenoviral-mediated delivery of therapeutic genes is a
useful approach to attenuate visual loss.
c. Adeno-Associated Viruses. Adeno-associated viruses (AAVs) have
no known pathogenic effects in humans. It is estimated that 80% of the
population develop antibodies to these viruses. AAVs have a wide host
range, a high transduction frequency, and they preferentially integrate
into the human genome on chromosome 19q13.4. They do not require
host cell replication for integration. The AAV rep and cap gene cassettes
are deleted before packaging a passenger gene. However, the target gene
capacity of AAVs is limited to approximately 4.8 kb (Table 1). AAVs induce less host immune response than adenoviruses because a large percentage of their genome is deleted to accommodate the transgene. Thus, few
viral proteins are expressed in vivo. The viruses are naturally replication
incompetent and usually require the gene function of a coinfected adenovirus of herpes simplex, as well as trans-complementation of the deleted
rep and cap genes to generate viral progeny. Production of high AAV titers is difcult to obtain and toxicity to the host is increased because of
the contaminating helper virus (121,122). While this is a potentially
powerful gene transfer vehicle, the transgene capacity is a limiting feature
in the widespread utility of AAVs in gene transfer approaches.
Until recently, transduction of normal, mature photoreceptor cells has
been inecient and limited by toxicity and host immune response directed
against viral proteins. AAV transduction appears to obviate some of these
problems, but their use is constrained because of passenger gene size limitation and low titer. Ecient transduction using recombinant AAVs, however,
has been achieved in all retinal layers, the pigment epithelium, and the optic
nerve. Stable transgene expression, lower cytopathology, and reduced
immunogenicity have been reported after transduction with the recombinant
virus in some retinal studies (123129).
In one protocol, the AAV has shown considerable potential as an
eective system for delivering a functional active therapeutic gene in the

Copyright 2003 Marcel Dekker, Inc.

576

Tombran-Tink

treatment of Lebers congenital amaurosis (LCA). LCA is a clinically severe

retinal degeneration causing near total blindness in children. It is associated
with a mutation in the RPE65 gene. In a breakthrough study, a recombinant
AAV vector encoding the RPE65 gene was used to test the eectiveness of
wide-type RPE65 in a spontaneously occurring RPE65 canine model. The
dogs suer from an early onset of visual dysfunction similar to that seen in
humans aected with LCA. Intraocular innoculations with the recombinant
AAV-RPE65 construct resulted in eective transduction and expression of
the wild-type RPE65 gene product. Gene expression of the wild-type protein
correlated with a signicant improvement in visual acuity in the transfected
animals (130). AAV transfer of another gene, CNTF, was shown to improve
photoreceptor function in a rhodopsin knockout mouse model for retinitis
pigmentosa. After transfection into the subretinal space, long-term expression of the biologically active, secreted CNTF resulted in prolonged survival
of photoreceptors in the rhodopsin knockout (131). In the retinal degeneration slow (rds) mice, another mouse model for RP, the wild-type peripherin2 gene (Prph2), transferred by an AAV vector, promoted ultrastructure
stabilization of the photoreceptor layer in the retina. Prph2 is a photoreceptor-specic membrane glycoprotein found in the rims of the cells outer
segment discs. These discs contain photopigments essential for photon capture during visual transduction. Prph forms a complex with the rom-1 gene
product in the outersegments. The complex is essential to induce stable
generation of outer segments and formation of new stacks of discs.
During the study, it was noted that Prph2 transgene overexpression was
associated with the reestablishment of complex ultrastructure in the photoreceptor layer. This subsequently led to an electrophysiological correction of
the outer nuclear layer of the Prph2 transfected retinas (132). Ecient
transduction and long-term gene expression of the wild-type bPDE were
also reported to preserve photoreceptor cells after AAV transduction in
the retina (133). The results, although preliminary, suggest that AAV is
another potentially useful vector for transferring therapeutic genes to the
human retina.
d. Herpes Simplex Virus. Herpes simplex viruses (HSVs) are large
DNA pathogens that infect approximately 6090% of the worlds population. The co-evolution of this virus with humans is due, in part, to its
ability to evade host immune surveillance. It establishes a latent infection
in its host, a condition that allows the virus to remain unnoticed. These
viruses are predominantly neurotrophic pathogens that can infect both
quiescent and proliferating cells, an important feature in gene therapy
protocols for central nervous system (CNS) disorders. HSVs have the potential to establish a lifetime latent infection in cells of the nervous system.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

577

Acute infection or reactivation of latent HSV elicits a strong immune response from the host, often leading to corneal blindness or fatal sporadic
encephalitis. Chronic episodes of reactivation of latent HSV result in stromal keratitis and scarring corneal blindness. The viral genome encodes 81
known genes, 38 of which are important to in vitro viral replication. During the construction of a recombinant vector, several of the immediate
early genes are deleted to accommodate a passenger gene of > 150 kb
(134). Tropism, latent infective activity, large transgene capacity, and the
ability to evade the host immune system are some desirable features in the
use of this vector for gene therapy (Table 1).
Gene transfer experiments into the cornea, subconjunctiva and anterior chamber of the mouse eye with HSV-1 indicate that the virus is an
eective gene delivery vehicle in the eye (135,136). The eciency of HSVmediated transfer of the lacZ gene was also tested in monkey eyes, human
trabecular meshwork, and human ciliary muscle cells. Gene transfer was
reported to be successful after determination of the b-galactosidase activity
in the infected tissues. However, signicant inammation, mild vitritis, and
retinitis were observed in the eye after infection. Transgene delivery and
expression in RPE cells, optic nerve, retinal ganglion cells, and the iris
epithelium were also reported with HSV (137). The possibilities for HSV
as a gene delivery vehicle in retinal degenerative diseases are enormous
because the virus has a large gene transfer capacity and can infect a wide
range of retinal cells. However, its potential will only be fully realized with
modications that will decrease the vectors immunogenicity and reduce
packaging instability of the target gene.
The above-described four viruses are currently the most advanced gene
delivery system in clinical protocols, but others, including the lentivirus and
human immunodeciency viruses (HIVs), are being approached as possibilities for gene transfer therapy (138143). They are endowed with features
that could be advantageous to the gene therapy approach. Engineering
second-generation viruses with predictable biological properties and
reduced immunogenicity or developing chimeric vectors that combine the
advantageous properties of several delivery systems will enhance the use of
gene therapy and provide exibility in the treatment of many diseases.

4.

Non Viral Vectors

One of the greatest concerns that researchers are faced with in gene therapy
is the safety of viral vectors as gene delivery vehicles. Consequently, considerable eort has been devoted to evaluating and designing alternative
strategies for gene delivery. Nonviral approaches that are currently in devel-

Copyright 2003 Marcel Dekker, Inc.

578

Tombran-Tink

opment take into consideration the size of the therapeutic gene to be delivered, targeting specicity, immunogenicity, and toxicity.
a. Naked DNA. Perhaps the simplest nonviral gene delivery system
in use today is the transfer of naked DNA directly into cells. The overall
efciency of this method, however, is very poor when compared to viral
gene transfer. Without mechanical or chemical help, naked DNA will not
enter cells rapidly, and once inside, the nucleic acid is exposed and susceptible to enzymatic degradation. In addition, plasmid DNA carrying therapeutic gene does not usually integrate into the host genome, and gene
expression is transient in those cells that are successfully transfected. In
spite of these limitations, surprisingly high levels of gene expression have
been obtained in a few accessible tissues, such as skin and muscle, using
plasmid DNA. In such cases, treatment is carried out by directly injecting
the plasmid DNA into the tissue because the DNA is vulnerable to degradation in body uids. So far, the method is safe and nontoxic, but it lacks
the ability to transduce a large number of cells and requires surgical procedures to access internal tissues.
An improved strategy for delivery nucleic acid directly into cells is by
high velocity bombardment of the cells with DNA attached to gold particles
using a gene gun approach. Microparticle bombardment has shown some
impressive results in focal delivery of naked DNA to corneal cells with little
damage or irritation to the tissue (144). It is a method that is being developed for more widespread use and may be a solution to some of the problems encountered with viral vectors. Most gene transfer studies in the eye
are carried out using viral vector, but plasmid delivery of a few therapeutic
genes, such as tissue plasminogen activator and IFN-
, has been tested and
shows potential benets in treating corneal-related pathologies (145157). A
few studies have also reported successful gene expression in the retina using
plasmid DNA. In one case it was demonstrated that condensed plasmids
containing the human broblast growth factor genes were able to transduce
a small population of choroidal and RPE cells after subretinal injections
into RCS rat eyes. FGF gene expression in those tissues consequently
resulted in a delay in photoreceptor degeneration (148). While the current
methods of delivering naked DNA are still very inecient, the eye is in a
prime location to benet from improvements in mechanical delivery strategies that increase the therapeutic index of this approach.
b. Liposomes. Liposomes are probably the most widely known nonviral vectors used to transfer DNA into cells. The strategy involves encapsulation of plasmid DNA in lipid complexes that are capable of fusing
with the cell membrane and delivering the therapeutic genes intracellularly. Initially, this approach has encountered difculties because classical

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

579

liposomes are negatively charged lipids that do not interact spontaneously

with DNA. Charge limitations and the need to separate DNA-liposome
complexes after delivery have led to the development of positively charged
cationic lipids. These interact with DNA more readily and have proven to
be valuable tools that can compact and deliver DNA across the cell membrane with greater efcacy (149170). Cationic liposomes are typically formulated using a positively charged lipid and a co-lipid that will stabilize
the DNA complex. A commonly used formulation is a mixture of a cytofectin with a neutral lipid component such as DOPE. This combination
can be formulated into unilammellar vesicles by several methods, including reverse phase evaporation and microuidization. Stable complexes are
formed when DNA is combined with the vesicles. The DNA is subsequently condensed in the vesicles, forming nanometric particles that are
referred to as lipoplexes. These complexes protect the DNA, interact with
cell surface proteoglycans, and enter the cell by endocytosis (170) (Fig. 7).

Figure 7 Liposome-mediated transfer of nucleic acid. The nucleic acid is condensed in the liposome to form lipoplexes. These enter the cell by endocytosis.
The majority of lipoplexes are trapped in late endosome. A small percentage can
either be released into the cytoplasm (mRNA), where they are functional, or trac
non-specically to the nucleus, where they may form episomes.

Copyright 2003 Marcel Dekker, Inc.

580

Tombran-Tink

Currently, no more than 30 genes transfercompetent cationic liposomes

have been developed and are commercially available. Perhaps the most
widely used formulations are DOTAP and DOTMA, the latter of which
is sold as Lipofectin, an in vitro transfecting agent.
A disadvantage of current methods of liposome-mediated gene transfer is due to the large percentage of the DNA-bound complexes trapped in
late endosomes, where they undergo enzymatic degradation and are no
longer therapeutically useful. Only a small percentage of the bound nucleic
acid escapes systemic inactivation and endosome entrapment. Those that
manage to escape face yet another hurdle of getting to the nucleus and
maintaining their functional integrity. As with viral delivery, in vitro eectiveness of liposome-mediated gene transfer is often misleading and is a poor
guide for clinical ecacy. Conventional liposome formulations lack cell
specicity and can take hours for uptake into the cell. They are highly
susceptible to inactivation by a number of serum proteins that bind and
cause membrane destabilization, a major obstacle for systemic administration of liposomes. A current research focus in the pharmaceutical industry is
the development of sterically stable liposome formulations that are resistant
to serum disruption and that will not aggregate prior to delivery. One modication currently in development uses conventional liposome lipid membranes to covalently attach polymers such as polyethylene glycol (PEGlipid) to create stealth liposomes (152,155,156,162,169). Properly formulated
polymer-grafted liposomes are shown to be sterically stabilized compounds
that have long residence times in circulation, increased biodistribution, and
reduction in uptake by cells of the reticuloendothelial system. Other clinically advantageous features of pegylated liposome pharmacokinetics include
dose independence and increased ecacy as a slow release system for therapeutically active drugs. This is a fascinating technology that has the potential of being a tailor-made delivery system that will improve the therapeutic
index of a number of drugs.
Another modication strategy under active investigation is the manufacture of ligand-targeted liposomal drugs using combinatorial approaches.
Such molecular conjugates could potentially be more versatile than the
conventional systems (159164). In a recent study, transfection was
observed to be increased in hepatoma cells after the administration of modied lipoplexes containing triantennary galactosyl residues that specically
target hepatoma cells (172). Targeted delivery of doxorubicin to human
umbilical vein endothelial cells and subsequent decrease in the survival of
the cells were also achieved with immunoliposomes that were conjugated to
a monoclonal antibody against E-selectin, a surface marker of HUVECs
(173). While targeting will increase transfection eciency to specic tissues,
it does not address problems of DNA release encountered in the endosomes.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

581

Some researchers have shown that the association of amphiphilic peptides,

such as GALA, a pH-sensitive peptide, with cationic liposomes can induce
fusion and permeabilization at acidic pH values and improve release of the
DNA from endosomes. The peptides induce osmotic swelling and subsequent rupture of the endosomal membranes so that the DNA can escape
easily (174). These new modications, however, are not without problems.
Competition between ligand-mediated processes and nonspecic interactions with the cell membrane can hinder the ecacy of gene delivery and
must be resolved before ligand-modied liposomes are of clinical relevance.
In addition to engineering modications that result in specic cell
targeting and more ecient DNA release, mechanisms that will increase
nucleic acid condensation and promote nuclear targeting are promising
areas of research that will improve liposome-mediated gene delivery.
Modications using DOTAP liposome-protamine sulfate-DNA (LPD) formulations are shown to produce denser particles when bound to DNA and
result in consistently higher gene expression levels (175,176). Complexes
formed with polycations are also observed to be much smaller than those
formed with liposomes alone and have increased resistance to nuclease
degradation. Smaller-size complexes may allow for higher levels of gene
expression because of increased cellular internalization. An interesting variation to this hybrid concept is the use of UV-irradiated Japanese Sendai
virus (HVJ)-cationic liposome to facilitate nuclear targeting. The binding of
high mobility group 1 protein (HMG-1) increase the potency of the complex
and enhances nuclear targeting and stability of the DNA after delivery into
the nuclear envelope. The success of HVJ-liposome complexes in cancer
applications is thought to result from the ability of the complex to bypass
the endocytosis process, thereby minimizing the diculties encountered
when the DNA is released from the endosomes (171). The development of
synthetic chemical viruses that are capable of (a) extended blood circulation, (b) increased DNA microparticle condensation, (c) improved cellular
uptake, (d) exible tropism, (e) escaping enzymatic degradation, and (f)
nuclear targeting is an attractive challenge in the area of biopharmaceutics.
If realized, such compounds have enormous potential in gene therapy protocols and may surpass the clinical usefulness of viral vectors.
Liposome-based techniques have been optimized to successfully transfer functional genes into human primary RPE cells. In one study, dierences
in the eciency of transfection were observed between the types of liposomes used in the assay. Nontoxic transfer was achieved after each liposome
treatment, but the Tfx-50 formulation showed the most signicant results
when compared to transfection of the RPE cells with other liposome variations, including lipofectin, lipofectamine, Cellfectin, and DMRIE-C (177).
A fascinating variation of gene transfer by liposomes was achieved by a

Copyright 2003 Marcel Dekker, Inc.

582

Tombran-Tink

group of researchers who used liposome eye drops to transfer rat retinal
ganglion cells. Transfection was reported to be ecient and nontoxic to
ocular tissues. This approach represents an interesting development in nonsurgical gene delivery for retinal diseases (178). The use of another liposome
method, hemagglutinating virus of Japan liposomes, was tested for ecacy
in delivering tissue inhibitor of metalloproteinase-3 gene into rat RPE cells.
Not only was the transfection successful, but expression of the introduced
gene inhibited the development of experimental choroidal neovascularization induced by laser photocoagulation after transfection of the tissue (179).
These are only a few examples showing the feasibility of using, nonviral,
nontoxic synthetic DNA-complexing derivatives to transfer therapeutic
genes to the retina.
The development of innovative nonviral delivery system is still in its
infancy, but many advantages are associated with their use in gene transfer
applications: (a) they can package and deliver a transgene of any size; (b)
packaging cell lines are not required to generate high titers; (c) they are nonpathogenic and cannot replicate; (d) immunogenicity, toxicity, and inammation are minimized with their use; and (e) they can become completely
synthetic. While these are safe gene delivery systems, the disadvantages
currently lie in their overall ineciency of transfection and their inability
to achieve cell-specic and nuclear targeting. Modications that improve
these features will allow synthetic polymer-based gene vectors to be the
candidates of choice for pharmacological intervention in many diseases.
5. Gene Knockdown Therapy
a. Antisense Drugs. Antisense technology is a novel gene delivery
method that is increasingly applied to knock down the expression of a
specic target gene for therapeutic purposes or to study the function of
that gene. The fundamental principle of the antisense approach is to silence a gene using a short synthetic DNA or RNA sequence that is homologous to that contained within the target gene. Antisense
oligodeoxynucleotides (ODNs) are synthesized in the opposite direction of
the known complementary DNA sequence and are designed to hybridize
specically with their target sequences to interrupt the production of the
corresponding protein. Almost all human diseases are associated with a
dysfunctional protein. While most conventional drugs are designed to inhibit the disease-causing activity of a dysfunctional protein, antisense molecules are designed to inhibit the production of the protein. In principle,
gene silencing may be accomplished at the genetic level by inhibiting biological events, such as transcription, translation, or gene splicing (180
183). During the inhibition of transcription events, ODNs bind to double-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

583

stranded DNA to induce the formation of a short triple-helical structure.

This structure is mediated by Hoogsteen hydrogen bonds and sterically
hinders the transcription of a specic mRNA. In addition, translation of
RNA species can be interrupted by binding of ODNs to tRNA or preRNA to prevent their transport from the nucleus or by directly interacting
with target mRNA molecule after transcription. In cases where inhibition
occurs after the transcript is matured, antisense binding to RNA is intended to block ribosomal assembly or ribosomal sliding along the
mRNA during translation of the protein. ODNs can also be of therapeutic value if they are designed to target the intron-exon junctions of premature RNA. In this regard, they prevent splicing events that are essential
for maturation of the RNA transcript. Three regions of the RNA that are
considered the best targets when designing ODNs are the 5 0 Cap region,
the AUG translational initiation codon, and the 3 0 untranslated region.
The concept of disabling the function of a mRNA by hybridization of
antisense reagents is a simple one, but, like other gene-based therapies, the
technology has encountered diculties in the past. The technical problems
experienced in the early pioneering stages of antisense technology are only
now being elucidated and are the focus of active study. From these analyses,
several features are apparent in the design of eective antisense molecules:
determining the length of sequence with the greatest activity and specicity;
cellular uptake; specic targeting of the ODN; antisense stability; and toxicity. Other factors that have inuenced the eectiveness of antisense molecules are frequency of protein turnover, the intracellular environment of the
cell, and the extent of longevity of ODNs after administration.
Gene knockdown practices are still under development and will
require signicant modications before being clinically acceptable as a therapeutic modality. Introducing variations in antisense chemistry by subtle
changes in the phosphate or sugar moieties of the nucleic acid backbone
is one method that shows success in minimizing nuclease degradation of the
molecules. Replacing a nonbridging oxygen with a sulfur atom in the phosphodiester bond between nucleotides on the phosphate backbone generates
a phosphorothioate linkage, which is reported to be one of the most successful modication of antisense oligonucleotides to date (184,195).
Phosphorothioate compounds have shown ecacy in delivery and are less
vulnerable to intracellular nuclease degradation. A disadvantage of their
use, however, is that the constructs are chiral and form a racemix mixture
of ODN species that exhibit both desirable and undesirable properties in
vivo (186). Some ODNs are reported to be toxic, while others show nonspecic anity for proteins (1987). The technical progress in chemical modication of antisense has recently shifted from the rst-generation
phosphodiester oligonucleotides, which are still nuclease sensitive, to the

Copyright 2003 Marcel Dekker, Inc.

584

Tombran-Tink

more nuclease-resistant chimeric compounds that contain methoxyethyl

modications at the end of the ODNs. The development of oligonucleotide
conjugates with cell-penetrating and nuclear-targeting peptides and colloidal
antisense carriers that protect against degradation is emerging rapidly and
will signicantly improve cellular uptake, stability, subcellular tracking,
and increased in vivo activity (188192). Over 200 patents disclose antisense
sequences with therapeutic utility in the treatment of human diseases. It is a
powerful tool with exceptional clinical value and is being exploited to identify gene function and validate new drug targets.
Formivirsen (ISIS 2922) is the rst antisense oligonucleotide drug
approved for the treatment of cytomegalovirus (CMV)induced retinitis.
The 21-phosphorothioate oligonucleotide inhibits viral replication in the
human eye by binding to complementary sequences of early mRNA CMV
viral transcripts. In preliminary clinical trials, the progression of CMV retinitis in AIDS patients is signicantly delayed after intravitreal administration of formivirsen. Drug-clearance studies show that formivirsen exhibits
rst-order kinetics with a half-life of 62 hours in rabbits and 78 hours in
monkey. A mild and transient inammatory response and increase in intraocular pressure are observed after treatment with formivirsen. These appear
to be resolved spontaneously or reversed with topical steroid treatment
(193196).
Diseases characterized by retinal neovascularization are among the
principal candidates for antisense treatment. The use of antisense oligonucleotide against vascular endothelial growth factor (VEGF) has shown promising results for the treatment of proliferative retinopathy. After
intraocular administration in a murine model of retinal neovascularization,
phosphorothioate antisense molecules reduced VEGF protein synthesis and
the growth of new blood vessels in a dosage-dependent manner. The study
shows the therapeutic potential of ODNs in ischemia-induced proliferative
retinopathies (197). Proliferative vitreoretinopathy (PVR) is an ocular disorder often associated with proliferating RPE cells. Antisense knockdown of
c-myc, a protein active in the mitogenic pathway, inhibits the proliferation
of human retinal pigment epithelial cells, suggesting that c-myc ODNs may
be an exciting perspective in the treatment of PVR (198).
Retinal ganglion cell death is associated with increased expression of
the Bax protein after transection of the optic nerve. A phosphorothioate
Bax antisense oligonucleotide was reported to show therapeutic utility in
preserving ganglion cell following axotomy. Bax expression was reduced
and the number of surviving neurons increased after treatment with Bax
ODNs. This represents a novel approach for neurodegeneration due to
optic nerve injury (199). The use of ODNs to silence the expression of
another retinal gene GLAST, a glial glutamate transporter, showed sig-

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

585

nicant changes in normal retinal transmission and indicates the importance of GLAST in maintaining retinal function (200). Similarly, an antisense compound generated against the trkB receptor mRNA for brainderived neurotrophic factor (BDNF) alters the neurochemical phenotype
of retinal neurons (201). BDNF and its receptor are important to survival
and dierentiation of the retina and are potentially useful targets in retinal
degenerative diseases. Antisense targeting of bronectin transcripts was
also shown to reduce the expression of bronectin in retinal vascular
cells (202). The use of antisense oligonucleotides in these studies reects
the signicance of the technology in understanding the function and regulation of a specic protein and the potential therapeutic benets for
antisense-based ocular therapies.
b. Ribozymes. Ribozymes are naturally occurring catalytic RNA
and a new class of genetic tools used to inhibit gene expression. Designer
ribozymes are chemically designed to recognize and bind specic RNA
through complementary base-pair hybridization. Their value in human
therapeutics is dependent on their ability to distinguish between mutant
and wild-type RNA species and to act as molecular scissors to digest or
edit the target RNA in a way that will prevent translation of the corresponding protein (203205). There are developed as an alternate approach
to antisense drugs. Analysis of the physical, biochemical, and biological
properties of naturally occurring ribozymes has allowed researchers to
classify them according to their various catalytic functions:
1. Hammerhead ribozymes: These are approximately 30 nucleotides
long and the smallest ribozymes identied. They are found in
many viral DNA and are capable of site-specic cleavage of a
phosphodiester bond. Hammerhead ribozymes have been extensively studied, and many have been synthesized against RNA
targets. In recent years they have emerged as a potentially eective therapeutic measure in models of retinitis pigmentosa. In
areas of the brain, hammerhead ribozymes have been directed
against the amyloid peptide precursor (B-APP), which is associated with the pathogenesis of Alzheimers disease. Others, such
as angiozyme, have been synthesized against angiogenic processes involved in the progression of tumor metastasis.
2. Group 1 and Group 11 intron ribozymes: These species can selfsplice, digest, and ligate phosphodiester bonds. They are found in
lower eukaryotes and some bacteria. Group 1 intron ribozymes
mediate trans-splicing of RNA targets and is considered a useful
genetic tool in repairing mutations in defective genes.

Copyright 2003 Marcel Dekker, Inc.

586

Tombran-Tink

3. Ribonuclease P: Cleaves phosphodiester bonds of tRNA precursor molecules.

The catalytic activity of these molecules make them particularly interesting
in the treatment of dominantly inherited diseases. In autosomal dominant
retinitis pigmentosa (ADRP), a substitution of histidine for proline occurs
at codon 23 in the rhodopsin gene. This mutation is referred to as P23H and
is responsible for the synthesis of a mutant gene product that results in the
death of photoreceptor cells (206). Because the eld is relatively new, only a
few studies have been carried out in the retina to test the therapeutic eect of
ribozymes in ocular diseases. One research team has now shown that in vivo
expression and activity of hairpin and hammerhead ribozymes can be
achieved in a transgenic rat model of ADRP. Ecient transduction and
stable expression of the ribozymes were accomplished using an adeno-associated virus that contained a rod opsin promoter. The results suggested that
the expressed ribozymes discriminated between wild-type and mutant rhodopsin RNA and specically destroyed the P23H mutant specie. As a result,
translation of the P23H protein was inhibited and progression of photoreceptor degeneration in ADRP model was signicantly slowed down (206
212). Combining the advantages of current gene delivery strategies with
catalytic ribozymes has broad therapeutic implications for dominantly
expressed retinal diseases where the disease is already in progression.

E. Neurotrophic Factors
The neurotrophic approach to treating retinal diseases is of therapeutic
relevance in ophthalmology because trophic factors target apoptotic
mechanisms that are independent of the genetic mutation(s) for the disease.
Treatment with soluble neurotrophic factors has been shown to prevent the
death of retinal neurons in complex or dicult-to-treat ocular diseases
where the etiologies are not completely dened or where mutations in several genes are associated with the progression of the disease. The method of
delivering highly concentrated amounts of trophic factors to the eye is
straightforward and relatively simple to perform and bypasses the need
for complex viral or non viral delivery systems. Subretinal or intravitreal
injections are common routes of delivery to the aected area. Preparative
amounts of neurotrophic proteins can be easily puried from recombinant
expression systems, and combinations of several therapeutic proteins can be
administered simultaneously to the area of pathology. Another clinically
appealing feature of this approach is that the therapeutic ecacy of soluble
trophic factors is not hindered by the immunologic and toxic limitations
that are usually associated with vector-mediated delivery of DNA.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

587

Designing an eective treatment protocol, however, is based on adequate

knowledge of the pharmacokinetics of the trophic factor in a biological
system and establishing its ability to function in a physiological environment. A limitation associated with the use of trophic factors in retinal diseases is the need for multiple treatments to signicantly eect reversal of the
pathology. Unless these agents are administered topically to the eye or
packaged in a slow-release system, the method, while safe, is not convenient
for long-term management of retinal diseases.
One endogenous neurotrophic factor, which we have isolated and
characterized in our laboratory, is a 50 kDa protein, pigment epithelumderived factor (PEDF) (103,213), so named because it was initially isolated
from the retinal pigment epithelium. Functionally, there is a striking association between PEDF, or the lack thereof, and biological processes involving survival and death of retinal cells as well as angiogenic mechanisms in
the eye (35,215219,229). The PEDF gene is well characterized and is classied as a serine protease inhibitor because of its structural and sequence
homology with members of this group of genes (104,227). In addition,
PEDF maps to human chromosome 17p13.3 and is tightly linked to an
autosomal dominant retinitis pigmentosa locus in that region of the chromosome. Several polymorphisms have been identied in the gene, but none
has shown a direct correlation between PEDF and specic retinal pathologies (220224). However, in vivo and in vitro studies with the soluble protein
consistently demonstrate the neuroprotective and antiangiogenic activities
of PEDF, suggesting a promising future for this protein as a therapeutic that
can circumvent the eects of specic mutations or chemical stimulators that
cause the death of visual cells.
We rst identied the PEDF protein in the conditioned medium of
primary cultures of fetal human RPE cell and in the interphotoreceptor
matrix (IPM) located between the RPE and neural retina
(103,213,225,226). The protein is expressed in high concentration in fetal
and young adult RPE cells but appears to be severely downregulated in
senescing RPE cultures, a nding that suggests that it may play a role in
age-related retinal dysfunctions. In one of the rst studies, we showed that
PEDF inhibits the growth of a human retinoblastoma cell line (Y79) by
inducing dierentiation of the tumor cells into a phenotype that is reminiscent of matured neurons. In nontreated cultures, the Y79 cells grow as
clusters in suspension and do not spontaneously attach or dierentiate.
Treatment of these cells with a small dose of PEDF is eective in promoting
extensive neurite outgrowths from the tumor cells, upregulating neurolament proteins and neuron-specic enolase, and promoting connections
between the growing neurites of newly dierentiated cells (Fig. 8).
Approximately 90% of the cultures attach and dierentiate on poly-d-

Copyright 2003 Marcel Dekker, Inc.

588

Tombran-Tink

Figure 8 PEDF induces dierentiation in human Y79 retinoblastoma cells. (Left)

Proliferating nontreated cells are rounded in appearance. (Right) Cells treated with
50 ng/mL PEDF. Treated cells are nonproliferating, project long neurites, and stain
positive for the neurolament protein.

lysinecoated surfaces and maintain their dierentiated phenotype for

longer than 4 weeks in culture after a single PEDF treatment. This period
is extended by additional doses of PEDF to the cultures. From these initial
results, PEDF is implicated as a useful therapeutic for reducing ocular
tumor progression and prevent degeneration, associated with the growing
tumor, in the surrounding retinal tissue.
A role for PEDF in angiogenesis emerged when Dawson and colleagues
provided convincing evidence that PEDF, in addition to its neurotrophic
activity, functions as an angiogenic inhibitor (214). They showed that
PEDF inhibited the formation of new blood vessels in a rat model of corneal
neovascularization and also prevented the migration of endothelial cells in a
dose-dependant manner. The group also reported that PEDF inhibited
endothelial cell migration even in the presence of such angiogenic stimulators
as bFGF, VEGF, interleukin-8, aFGF, and lyopohosphatic acid.
Furthermore, when tested against other antiangiogenic agents, such as

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

589

angiostatin and endostatin, its ecacy was slightly more potent than those
inhibitors. In support of their study, we showed higher concentrations of
PEDF in the vitreous of patients with avascular proliferative vitreal retinopathy and diabetic retinopathy when compared to patients with retinal
pathologies associated with increased angiogenic activity (228). Based on
the clinical data, as well as vivo studies using animal models, it appears that
the concentration of PEDF in the eye is important to the vascular state of
ocular tissues. These results have stirred much interest in the ophthalmic eld
and have encouraged several groups to exploit the therapeutic potential of
PEDF in ocular diseases, such as age-related macular degeneration, where
both cell death and increased angiogenesis contribute to severe visual loss.
In several modes of induced retinal degeneration, convincing evidence
that photoreceptor cells survive in the presence of PEDF has been provided.
In an in vitro model of retinal damage, a large percentage of retinal neurons
undergo apoptosis and die after exposure to hydrogen peroxide (H2 O2 )
(215217). Hydrogen peroxide is a reactive oxygen species (ROS) found in
elevated concentration in light-damaged retinas. It is believed that ROS
contribute to degenerative and aging processes in the eye. To test the protective eects of PEDF in H2 O2 -damaged eyes, rat retinal cultures were
treated with PEDF before they were exposed to H2 O2 . In the presence of
PEDF, apoptotic mechanisms that led to cell death were inhibited, and
approximately 60% of the cells that would have otherwise degenerated
survived. Furthermore, a high percentage of the treated cells were rhodopsin
positive and, therefore, highly likely to be rod photoreceptors. In vivo, the
retina can also be damaged by exposure to constant light, in part because of
the generation of reactive oxygen species in a high-lipid-content region of
the retina. Photoreceptor degeneration is visible as early as the third day of
light exposure in the rat. In a study aimed at testing the eect of PEDF in
light-damaged rat eyes, we found that a single intravitreal injection of
PEDF, prior to chronic light exposure, was potent enough to inhibit the
light damage eects on photoreceptor. This was clearly seen in histological
preparations of the treated retina and electrophysical measurements of the
nuclei in the outer nuclear layer (ONL) (Fig. 9).
In a similar study, photoreceptor survival with PEDF treatment was
examined in two mutant mice types, homozygous retinal degeneration (rd/
rd) and retinal degeneration slow (rds/rds), in which photoreceptor loss is a
hallmark of the mutations. Intravitreal injections of PEDF resulted in a
transient but signicant delay in the death of photoreceptors in both
mutants (229). The ecacy of PEDF was also assessed in an embryonic
Xenopus model of retinal degeneration (218). In this model, mechanical
removal of the RPE cells from the Xenopus retina results in a distortion
of photoreceptor ultrastructure and disruption of outersegment formation,

Copyright 2003 Marcel Dekker, Inc.

590

Tombran-Tink

Figure 9 Neuroprotection of rat photoreceptor cells (ONL) by PEDF after 7 days

exposure to constant light. Severe damage to photoreceptor cells (ONL) is seen after
light exposure. (A) Unexposed normal rat retina; (B) retinal exposed to 7 days
constant light. (CF) Cross section of rat retina exposed to 7 days of constant
light after injections with PEDF, bFGF, or both: (C) PBS; (D) 1 g PEDF; (E)
1 g bFGF; (F) 1 g bFGF 1 g PEDF; (inset) electroretinograms (ERG) B-waves
amplitudes of the retina with each treatment. RPE, Retinal pigment epithelial cells;
OS, photoreceptor outersegment; IS photoreceptor inner segment; ONL, outer
nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer. (Courtesy of Dr.
James McGinnis and Wei Cao, Dean McGee Eye, Oklahoma, USA.)

a condition that eventually leads to the death of photoreceptors. In this

study it was found that administration of PEDF to the retina after RPE
detachment promoted features such as reorganization of the outersegments,
increased opsin synthesis by the photoreceptors, and a general correction of
the ONL ultrastructure. Moreover, the activity of PEDF in RPE detached

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

591

retinas was blocked by a neutralizing polyclonal antibody to the 50 kDa

native protein, suggesting that the rescuing eect was specic.
From these ndings it appears that the course of photoreceptor degeneration can be altered by neuroprotective agents, like PEDF, which can
prevent pathomorphological and apoptotic eects of neurodegenerative
promoters. Other trophic factors, such as bFGF, CNTF, and BDNF have
shown similar results in promoting photoreceptor survival in naturally
occurring inherited retinal degeneration models with genetic defects similar
to those in human inherited retinal degeneration (230). Survival factors are,
therefore, particularly attractive therapeutic tools that may prove to be
increasingly important in treating retinal degenerations if long-term, sustained delivery to the aected area is to be maintained.

REFERENCES
1. Adler, R., Curcio, C., Hicks, D., Price, D., and Wong, F. (1999). Cell death in
age-related macular degeneration. Mol. Vis., 5:31.
2. Votruba, M., and Gregor, Z. (2001). Neovascular age-related macular degeneration: present and future treatment options. Eye, 15(Pt 3):424429.
3. Sickenberg, M. (2001). Early detection, diagnosis and management of choroidal neovascularization in age-related macular degeneration: the role of
ophthalmologists. Ophthalmologica, 215(4):247253.
4. Rubin, G. S. (2001). Vision rehabilitation for patients with age-related macular degeneration. Eye, 15(Pt 3):430435.
5. Harding, S. (2001). Photodynamic therapy in the treatment of subfoveal choroidal neovascularization. Eye, 15(Pt 3):407412.
6. Schmidt-Erfurth U., and Hasan, T. (2000). Mechanisms of action of photodynamic therapy with verteporn for the treatment of age-related macular
degeneration. Surv. Ophthalmol., 45(3):195214.
7. Verma, L., Das, T., Binder, S., Heriot, W. J., Kirchhof, B., Venkatesh, P.,
Krebs, I., Stolba, U., Jahn, C., Feichtinger, H., Kellner, L., Krugluger, H.,
Pawelka, I., Frohner, U., Kruger, A., Li, W., and Tewari, H. K. (2000). New
approaches in the management of choroidal neovascular membrane in agerelated macular degeneration. Indian J. Ophthalmol., 48(4):263278.
8. Ho, Ac. (1999). Laser treatment in eyes with drusen. Curr. Opin. Ophthalmol.,
10(3):204208.
9. Hageman, G. S., and Mullins, R. F. (1999). Molecular composition of drusen
as related to substructural phenotype. Mol. Vis., 5:28.
10. Blodi, C. F., and Stone, E. M. (1990). Bests vitelliform dystrophy. Ophthal.
Paediatr. Genet., 11(1):4959.
11. Bither, P. P., and Berns, L. A. (1988). Stargardts disease: a review of the
literature. J. Am. Optom. Assoc., 59(2):106111.

Copyright 2003 Marcel Dekker, Inc.

592

Tombran-Tink

12. Phelan, J. K., and Bok, D. (2000). A brief review of retinitis pigmentosa and
the identied retinitis pigmentosa genes. Mol. Vis., 6:116124.
13. Baumgartner, W. A. (2000). Etiology, pathogenesis, and experimental treatment of retinitis pigmentosa. Med. Hypotheses, 54(5):814824.
14. van Soest, S., Westerveld, A., de Jong, P. T., Bleeker-Wagemakers, E. M., and
Bergen, A. A. (1999). Retinitis pigmentosa: dened from a molecular point of
view. Surv. Ophthalmol., 43(4):321334.
15. Shastry, B. S. (2000). Hereditary degenerative retinopathies: optimism for
somatic gene therapy. IUBMB Life, 49(6):479484.
16. Eudy, J. D., and Sumegi, J. (1999). Molecular genetics of Usher syndrome.
Cell Mol. Life Sci., 56(34):258267.
17. Perrault, I., Rozet, J. M., Gerber, S., Ghazi, I., Leowski, C., Ducroq, D.,
Souied, E., Duer, J. L., Munnich, A., and Kaplan, J. (1999). Leber congenital amaurosis. Mol. Genet. Metab., 68(2):200208.
18. Harris, E. W. (2001). Lebers congenital amaurosis and RPE65. Int.
Ophthalmol. Clin., 41(1):7382.
19. Algvere, P. V., Berglin, L., Gouras, P., and Shen, Y. (1994). Transplantation
of fetal retinal pigment epithelium in age-related macular degeneration with
subfoveal neovascularization. Graefes Arch. Clin. Exp. Ophthalmol., 232:707
716.
20. Mohand-Said, S., Hicks, D., Dreyfus, H., and Sashel, J. A. (2000). Selective
transplantation of rods delays cone loss in a retinitis pigmentosa model. Arch.
Ophthalmol., 118:807811.
21. Sheedlo, H. J., Li, L., and Turner, J. E. (1991). Photoreceptor cell rescue at
early and late RPE-cell transplantation periods during retinal disease in RCS
dystrophic rats. J. Neural Transplant Plast., 2:5563.
22. Rezai, K. A., Kohen, L., Wiedemann, P., and Heimann, K. (1997). Iris pigment epithelium transplantation. Graefes Arch. Clin. Exp. Ophthalmol.,
235:558562.
23. Bhatt, N. S., Newsome, D. A., and Fenech, T. (1994). Experimental transplantation of human retinal pigment epithelial cells on collagen substrates.
Am. J. Ophthalmol., 117:214221.
24. Jiang, L. Q., Jorquera, M., and Streilein, J. W. (1993). Subretinal space and
vitreous cavity as immunological privileged sites for retinal allografts. Invest.
Ophthalmol. Vis. Sci., 34:33473354.
25. Mohand-Said, S., Hicks, D., Dreyfus, H., and Sashel, J. A. (2000). Selective
transplantation of rods delays cone loss in a retinitis pigmentosa model. Arch.
Ophthalmol., 118:807811.
26. Little, C. W., Castillo, B., DiLoreto, D. A., Cox, C., Wyatt, J., del Cerro, C.,
and del Cerro, M. (1996). Transplantation of human fetal retinal pigment
epithelium rescues photoreceptor cells from degeneration in the Royal
College of Surgeons on rat retina. Invest. Ophthalmol. Vis. Sci., 37(1):204211.
27. Sauve, Y., Klassen, H., Whiteley, S. J., and Lund, R. D. (1998). Visual eld
loss in RCS rats and the eect of RPE cell transplantation. Exp. Neurol.,
152(2):243250.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

593

28. Sharma, R. K., Bergstrom, A., Zucker, C. L., Adolph, A. R., and Ehringer, B.
(2000). Survival of long-term retinal cell transplants. Acta. Ophthalmol.
Scand., 78:396402.
29. Aramant, R. B., Seiler, M. J., and Ball, S. L. (1999). Successful cotransplantation of intact sheets of fetal retinal with retinal pigment epithelium. Invest.
Ophthalmol. Vis. Sci., 40:15571564.
30. Woch, G., Aramant, R. B., Seiler, M. J., Sagdullaev, B. T., and McCall, M. A.
(2001). Retinal transplants restore visually evoked responses in rats with
photoreceptor degeneration. Invest. Ophthalmol. Vis. Sci., 42(7):16691676.
31. Schraermeyer, U., Kayatz, P., Thumann, G., Luther, T. T., Szurman, P.,
Kociok, N., and Bartz-Schmidt, K. U. (2000). Transplantation of iris pigment
epithelium into the choroid slows down the degeneration of photoreceptors in
the RCS rat. Graefes Arch. Clin. Exp. Ophthalmol., 238(12):979984.
32. Schraermeyer, U., Kociok, N., and Heimann, K. (1999). Rescue eect of IPE
transplants in RCS rats: short-term results. Invest. Ophthalmol. Vis. Sci.,
40(7):15451556.
33. Thumann, G., Bartz-Schmidt, K. U., El Bakri, H., Schraermeyer, U., Spee,
C., Cui, J. Z., Hinton, D. R., Ryan, S. J., and Heimann, K. (1999).
Transplantation of autologous iris pigment epithelium to the subretinal
space in rabbits. Transplantation, 68(2):195201.
34. Carwile, M. E., Culbert, R. B., Sturdivant, R. L., and Kraft, T. W. (1998).
Rod outer segment maintenance is enhanced in the presence of bFGF, CNTF
and GDNF. Exp. Eye Res., 66(6):791805.
35. Abe, T., Tomita, H., Kano, T., Yoshida, M., Ohashi, T., Nakamura, Y.,
Nishikawa, S., and Tamai, M. (2000). Autologous Iris pigment epithelial
cell transplantation in monkey subretinal region. Curr. Eye Res., 20:268275.
36. Humayun, M. S., de Juan, E., del Cerro, M., Dagnelie, G., Radner, W.,
Sadda, S. R., and del Cerro, C. (2000). Human neural retinal transplantation.
Invest. Ophthalmol. Vis. Sci., 41:31003106.
37. Algvere, P. V., Gouras, P., and Dafgard Kopp, E. (1999). Long-term outcome
of RPE allografts in non-immunosuppressed patients with AMD. Eur. J.
Ophthalmol., 9(3):217230.
38. Thumann, G., Aisenbrey, S., Schraermeyer, U., Laufaut, B., Esser, P., Walter,
P., and Bartz-Schmidt, K. U. (2000). Transplantation of autologous iris pigment epithelium after removal of choroidal neovascular membranes. Arch.
Ophthalmol., 118(10):13501355.
39. Bhatt, N. S., Newsome, Da., and French, T. (1994). Experimental transplantation of human retinal pigment epithelial cells on collagen substrates. Am. J.
Ophthalmol., 117:214221.
40. Fang, S. R., Kaplan, H. J., Del Piore, L. V., Liu, Y., Wang, X., Hornbect, R.,
Landgraf, M., Mason, G., and Silverman, M. S. (1993). Development of a
surgical procedure and instrument for transplantation of extended gelatin
sheets to the subretinal space. Invest. Ophthalmol. Vis. Sci., 34:1093.

Copyright 2003 Marcel Dekker, Inc.

594

Tombran-Tink

41. Thomas, J. A., Itskovitz-Eldor, J., Shapiro, S. S., Waknitz, M. A., Swiergiel, J.
J., Marshall, V. S., and Jones, J. M. (1998). Embryonic stem cell lines derived
from human blastocysts. Science, 282:11451147.
42. Shamblott, M. J., Axeman, J., Wang, S., Bugg, E. M., Littleeld, J. W.,
Donovan, P. J., Blumenthal, P. D., Huggins, G. R., and Gearhart, J. D.
(1998). Derivation of pluripotent system cells from cultured human primordial
germ cells. Proc. Natl. Acad. Sci. USA, 95:1372613731.
43. Bjornson, C. R., Rietze, R. L., Reynolds, B. A., Magli, M. C., Vescovi, A. L.
(1999). Turning brain into blood: a hematopoietic fate adopted by adult
neural stem cells in vivo. Science, 283(5401):534537.
44. Joshi, S. S., Tarantolo, S. R., Kuszynski, C. A., and Kessinger, A. (2000).
Antitumor therapeutic potential of activated human umbilical cord blood cells
against leukemia and breast cancer. Clin. Cancer Res., 6(11):43514358.
45. Gluckman, E. (2000). Current status of umbilical cord blood hematopoietic
stem cell transplantation. Exp. Hematol., 28(11):11971205.
46. Howery, R. P., Martin, P. L., Driscoll, T., Szabolcs, P., Kelly, T., Shpall, E. J.,
Bearman, S. I., Slat-Vasquez, V., Rubinstein, P., Stevens, C. E., and
Kurtzberg, J. (2000). Graft-versus-leukemia-induced complete remission following unrelated umbilical cord blood transplantation for acute leukemia.
Bone Marrow Transplant, 26(11):12511254.
47. Orlic, D., Kajstura, J., Chimenti, S., Jakoniuk, I., Anderson, S. M., Li, B.,
Pikel, J., McKay, R., Nadal-Ginard, B., Bodine, D. M., Leri, A., and
Anversa, P. (2001). Bone marrow cells regenerate infarcted myocardium.
Nature, 410(6829):701705.
48. Brown, J. M., Weissman, I. L., and Shizuru, J. A. (2001). Immunity to infections following hematopoietic cell transplantation. Curr. Opin. Immunol.,
13(4):451457.
49. Terskikh, A. V., Easterday, M. C., Li, L., Hood, L., Kornblum, H. I.,
Geschwind, D. H., and Weissman, I. L. (2001). From hematopoiesis to neuropoiesis: evidence of overlapping genetic programs. Proc. Natl. Acad. Sci.
USA, 98(14):79347939.
50. Gokhan, S., and Mehler, M. F. (2001). Basic and clinical neuroscience applications of embryonic stem cells. Anat. Rec., 265(3):142156.
51. Livesey, F. J., and Cepko, C. L. (2001). Vertebrate neural cell-fate determination: lessons from the retina. Nat. Rev. Neurosci., 2(2):109118.
52. Tropepe, V., Coles, B. L., Chiasson, B. J., Horsford, D. J., Elia, A. J.,
McInnes, R. R., and van der Kooy, D. (2000). Retinal stem cells in the
adult mammalian eye. Science, 287(5460):20322036.
53. Ahmed, I., Tang, L., and Pham, H. (2000). Identication of neural progenitors in the adult mammalian eye. Biochem. Biophys. Res. Commun.,
270(2):517521.
54. Henderson, T. R., Coster, D. J., and Williams, K. A. (2001). The long term
outcome of limbal allografts: the search for surviving cells. Br. J. Ophthalmol.,
85(5):604609.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

595

55. Perron, M., and Harris, W. A. (2000). Retinal stem cells in vertebrates.
Bioessays, 22(8):685688.
56. Tamalu, F., Chiba, C., Ishida, A. T., and Saito, T. (2000). Functional dierentiation of ganglion cells from multipotent progenitor cells in sliced retinal of
adult goldsh. J. Comp. Neurol., 419(3):297305.
57. Layer, P. G., Rothermel, A., and Willbold, E. (2001). From stem cells towards
neural layers: a lesson from re-aggregated embryonic retinal cells.
Neuroreport, 12(7):A3946.
58. Otteson, D. C., D. Costa, A. R., Hitchcock, P. F. (2001). Putative stem cells
and the lineage of rod photoreceptors in the mature retina of the goldsh. Dev.
Biol., 232(1):6276.
59. Kurimoto, Y., Shibuki, H., Kaneko, Y., Ichikawa, M., Kurokawa, T.,
Takahashi, M., and Yoshimura, N. (2001). Transplantation if adult rat hippocampus-derived neural stem cells into retina injured by transient ischemia.
Neurosci Lett., 306(12):5760.
60. Takahashi, M., Palmer, T. D., Takahashi, J., and Gage, F. H. (1998).
Widespread integration and survival of adult-derived neural progenitor cells
in the developing optic retina. Molec. Cell Neurosci., 12(6):340348.
61. Henderson, T. R., Coster, D. J., and Williams, K. A. (2001). The long term
outcome of limbal allografts: the search for surviving cells. Br. J. Ophthalmol.,
85(5):604609.
62. Anderson, D. F., Ellies, P., Pires, R. T., and Tseng, S. C. (2001). Amniotic
membrane transplantation for partial limbal stem cell deciency. Br. J.
Ophthalmol., 85(5):567575.
63. Eiges, R., Schuldiner, M., Drukker, M., Yanuka, O., Itskovitz-Eldor, J., and
Benvenisty, N. (2001). Establishment of human embryonic stem cell-transfected clones carrying a marker for undierentiated cells. Curr. Biol.,
11(7):514518.
64. Chow, A. Y. (1993). Electrical stimulation of the rabbit retinal with subretinal
electrodes and high density microphotodiode array implants. Invest.
Ophthalmol. Vis. Sci., 34:835.
65. Rizo, J. F. 3rd, Wyatt, J., Humayun, M., de Juan, E., Liu, W., Chow, A.,
Eckmiller, R., Zrenner, E., Yagi, T., and Abrams, G. (2001). Retinal prosthesis: an encouraging rst decade with major challenges ahead. Opthalmology,
108(1):1314.
66. Rizzo, J. F., and Wyatt, J. (1997). Prospects for a visual prosthesis.
Neuroscientists, 3:251262.
67. Eckmiller, R. (1997). Learning retina transplants with epiretinal contacts.
Ophthalmol. Res., 9:281289.
68. Peyman, G., Chow, A. Y., Liang, C., Chow, V. Y., Perlman, J. I., and
Peachey, N. S. (1998). Subretinal semiconductor microphotodiode array.
Ophthalmic Surg. Lasers, 29(3):234241.
69. Chow, A. Y., and Chow V. Y. (1997). Subretinal electrical stimulation of the
rabbit retina. Neurosci. Lett. 255:1316.

Copyright 2003 Marcel Dekker, Inc.

596

Tombran-Tink

70. Zrenner, E., Stett, A., Weiss, S., Aramant, R. B., Guenther, E., Kohler, K.,
Miliczek, K. D., Seiler, M. J. and Haemmerle, H. (1999). Can subretinal
microphotodiodes successfully replace degenerated photoreceptors? Vision
Res., 39(15):25552567.
71. Peachey, N. S., and Chow, A. Y. (1999). Subretinal implantation of semiconductor-based photodiodes: progress and challenges. J. Rehabil. Res. Dev.,
36(4):371376.
72. Humayun, M. S., de Juan, G., Dagnelie, R. G., Greenberg, R. H., Propst, D.
H., and Phillips, (1996). Visual perception elicited by electrical stimulation of
the retina in blind humans. Arch. Ophthalmol., 114:4046.
73. Zrenner, E., Miliczek, D. K., Gabel, V. P., Graf, H. G., Guenther, E.,
Haemmerle, H., Hoeinger, B., Kohler, K., Nish, W., Schubert, M., Stett,
A., and Weiss, (1997). The development of subretinal microphotodiodes for
replacement of degenerated photoreceptors. Ophthalmic Res., 29:269280.
74. Chow, A. Y., Pardue, M. T., Chow, V. Y., Peyman, G. A., Liang, C.,
Perlman, J. I., and Peachey, N. S. (2001). Implantation of silicon chip microphotodiode arrays into the cat subretinal space. IEEE Trans. Neural Syst.
Rehabil. Eng., 9(1):8695.
75. Hesse, L., Schanze, T., Wilms, M., and Eger, M. (2000). Implantation of
retina stimulation electrodes and recording of electrical stimulation responses
in the visual cortex of the cat. Graefes Arch. Clin. Exp. Ophthalmol.,
238(10):840845.
76. Walter, P., Szurman, P., Vobig, M., Berk, H., Ludtke-Handjery, H. C.,
Richter, H., Mittermayer, C., Heimann, K., and Sellhaus, B. (1999).
Successful long-term implantation of electrically inactive epiretinal microelectrode arrays in rabbits. Retina, 19(6):546542.
77. Nadig, M. N. (1999). Development of a silicone retinal implant: cortical
evoked potentials following focal stimulation of the rabbit retina with light
and electricity. Clin. Neurophysiol., 110(9):15451553.
78. Grumet, A. E., Wyatt, J. L. Jr, and Rizzo, J. F. 3rd. (2000). Multi-electrode
stimulation and recording in the isolated retina. J. Neurosci. Methods,
101(1):3142.
79. Hershey, A. D., and Chase, M. (1952). Independent functions of viral proteins
and nucleic acid in growth of bacteriophage. J. Gen. Physiol., 36:3956.
80. Watson, J. D., and Crick, F. H. C. (1953). Molecular structure of nucleic
acids: a structure for deoxyribose nucleic acid. Nature, 171:737738.
81. Crick, F. H. C., Barnett, L., Brenner, S., and Watts-Tobin, R. J. (1961).
General nature of the genetic code for proteins. Nature, 192:12271232.
82. Danos, O., and Mulligan, R. C. (1988). Safe and ecient generation of recombinant retroviruses with amphotropic and ecotropic host ranges. PNAS,
85:64606464.
83. Markowitz, D., Go, S., and Bank, A. (1988). A safe packaging line for gene
transfer: separating viral genes on two dierent plasmids. J. Virol. 62:1120
1124.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

597

84. Cone, R. D., and Mulligan, R. C. (1992). High-eciency gene transfer into
mammalian cells: generation of helper-free recombinant retrovirus with broad
mammalian host range. Biotechnology, 24:420424.
85. Con, J. M. (1996). Retroviridae: the viruses and their replication. In: Fields
Virology (B. N. Fields, D. M. Knippe, and P. Howley, eds.). LippincottRaven, Philadelphia, pp. 17671840.
86. Battini, J. L., Heard, J. M., and Danos, O. (1992). Receptor choice determinants in the envelope glycoproteins of amphotropic, xenotropic, and polytropic murine leukemia viruses. J. Virol, 66:14681475.
87. Sakamoto, T., Spee, C., Scuric, Z., Gordon, E. M., Hinton, D. R., Anderson,
W. F., and Ryan, S. J. (1998). Ability of retroviral transduction to modify the
angiogenic characteristics of RPE cells. Graefes Arch. Clin. Exp. Opthalmol.,
236(3):220229.
88. Kido, M., Rich, K. A., Yang, G., Barron, E., Kohn, D. B., al-Ubaidi, M. R.,
Blanks, J. C., and Lang, G. (1996). Use of retroviral vector with an internal
opsin promoter to direct gene expression to retinal photoreceptor cells. Curr.
Eye Res., 15(8):833844.
89. Couderec, B. C., de Neuville, S., Douin-Echinard, V., Serres, B., Manenti, S.,
Darbon, J. M., and Malecaze, F. (1999). Retrovirus-mediated transfer of a
suicide gene into lens epithelial cells in vitro and in an experimental model of
posterior capsule opacication. Curr. Eye Res., 19(6):472482.
90. Schubert, C. A., Kimura, H., Spee, C., Hinton, D. R., Gordon, E. M.,
Anderson, W. F., and Ryan, S. J. (1997). Retrovirus-mediated transfer of
the suicide gene into retinal pigment epithelial cells in vitro. Curr. Eye Res.,
16(7):656665.
91. Kimura, H., Sakamoto, T., Cardillo, J. A., Spee, C., Hinton, D. R, Gordon,
E. M., Anderson, W. F., and Ryan, S. J. (1996). Retrovirus-mediated suicide
gene transduction in the vitreous cavity of the eye: feasibility in prevention of
proliferative vitreoretinopathy. Hum. Gene Ther., 1:7(7):799808.
92. Sakamoto, T., Kimura, H., Scuric, Z., Spee, C., Gordon, E. M., Hinton, D.
R., Anderson, W. F., and Ryan, S. J. (1995). Inhibition of experimental proliferative vitreoretinopathy by retroviral vector-mediated transfer of suicide
gene. Can proliferative vitreoretinopathy be a target of gene therapy?
Ophthalmology, 102(10):14171424.
93. Seitz, B., Baktanian, E., Gordon, E. M., Anderson, W. F., LaBreed, L., and
McDonnell, P. J. (1998). Retroviral vector-mediated gene transfer into keratocytes: in vitro eects of polybrene and protamine sulfate. Graefes Arch. Clin.
Exp. Ophthalmol., 236(8):602612.
94. Bradshaw, J. J., Obritsch, W. F., Cho, B. J., Gregerson, D. S., and Holland, E.
J. (1999). Ex vivo transduction of corneal epithelial progenitor cells using a
retroviral vector. Invest. Ophthalmol. Vis. Sci., 40(1):230235.
95. Murata, T., Cui, J., Taba, K. E., Oh, J. Y., Spee, C., Hinton, D. R., and Ryan,
S. J. (2000). The possibility of gene therapy for the treatment of choroidal
neovascularization. Ophthalmology, 107(7):13641373.

Copyright 2003 Marcel Dekker, Inc.

598

Tombran-Tink

96. Murata, T., Homan, S., Ishibashi, T., Spee, C., Gordon, E. M., Anderson,
W. F., Hinton, D. R., and Ryan, S. J. (1998). Retrovirus-mediated gene
transfer targeted to retinal photocoagulation sites. Diabetologia, 41(5):500
506.
97. vanRaaij, M. J., Mitraki, A., Lavigne, G., and Cusack, S. (1999). A triple ?spiral in the adenovirus bre shaft reveals a new structural motif for a brous
protein. Nature 401:935938.
98. Shenk, T. (1996). Adenoviridae. The viruses and their replication. In: Fields
Virology (B. N. Fields, D. M. Knippe, and P. Howley, eds.). Lippincot-Raven,
Philadelphia, pp. 21112148.
99. Stratford-Perricaudet, L. D., Makeh, I., Perricaudet, M., and Briand, P.
(1992). Wide-spread long-term gene transfer to mouse skeletal muscles and
heart. J. Clin. Invest., 90:626630.
100. Bett, A. J., Haddara, W., Prevec, L., and Graham, F. L. (1994). An ecient
and exible system for construction of adenovirus vectors with insertions of
deletions in early regions 1 and 3. PNAS, 91:88028806.
101. Abraham, N. G., Da Silva, J. L., Dunn, M. W., Kigasawa, K., and Shibahara,
S. (1998). Retinal pigment epithelial cell-based gene therapy against hemoglobin toxicity. Int. J. Mol. Med., 1(4):657663.
102. Verma, L., Das, T., Binder, S., Heriot, W. J., Kirchhof, B., Venkatesh, P.,
Krebs, I., Stobla, U., Jahn, C., Feichtinger, H., Kellner, L., Krugluger, H.,
Pawelka, I., Frohner, U., Druger, A., Li, W., and Tewari, H. K. (2000). New
approaches in the management of choroidal neovascular membrane in agerelated macular degeneration. Indian J. Ophthalmol., 48:263278.
103. Tombran-Tink, J., Chader, G. G., and Johnson, L. V. (1991). PEDF: a pigment epithelium-derived factor with potent neuronal dierentiative activity.
Exp. Eye Res., 53(3):411414.
104. Steele, F. R., Chader, G. J., Johnson, L. V., and Tombran-Tink, J. (1993).
Pigment epithelium-derived factor: neurotrophic activity and identication as
a member of the serine protease inhibitor gene family. Proc. Natl. Acad. Sci.,
15;90(4):15261530.
105. Dawson, D. W., Volpert, O. V., Gillis, P., Crawford, S. E., Xu, H., Benedict,
W., and Bouck, N. P. (1999). Pigment epithelium-derived factor: a potent
inhibitor of angiogenesis. Science, 9;285(5425):245248.
106. Mori, K., Duh, E., Gehlbah, P., Ando, A., Takahashi, K., Pearlman, J., Mori,
K., Yang, H. S., Zack, D. J., Ettyreddy, D., Brough, D. E., Wei, L. L., and
Campochiaro, P. A. (2001). Pigment epithelium-derived factor inhibits retinal
and choroidal neovascularization. J. Cell Physiol., 188(2):253263.
107. Honda, M., Sakamoto, T., Ishibashi, T., Inomata, H., and Ueno, H. (2000).
Experimental subretinal neovascularization is inhibited by adenovirusmediated soluble VEGF/t-1 receptor gene transfection: a role of VEGF
and possible treatment for SRN in age-related macular degeneration. Gene
Ther., 7(11):978985.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

599

108. Cayouette, M., and Gravel, C. (1997). Adenovirus-mediated gene transfer of

ciliary neurotrophic factor can prevent photoreceptor degeneration in the
retinal degeneration (rd) mouse. Hum. Gene Ther., 8(4):423430.
109. Cayouette, M., Behn, D., Sendtner, M., Lachapelle, P., and Gravel, C. (1998).
Intraocular gene transfer of ciliary neurotrophic factor prevents death and
increases responsiveness of rod photoreceptors in the retinal degeneration
slow mouse. J. Neurosci., 18(22):92829293.
110. Bennett, J., Pakola, S., Zeng, Y., and Maguire, A. (1996). Humoral response
after administration of E1-deleted adenoviruses: immune privilege of the subretinal space. Hum. Gene Ther., 7(14):17631769.
111. Bennett, J., Tanabe, T., Sun, D., Zeng, Y., Kjeldbye, H., Gouras, P., and
Maguire, A. M. (1996). Photoreceptor cell rescue in retinal degeneration
(rd) mice by in vivo gene therapy. Nat. Med. 2(6):649654.
112. Akimoto, M., Miyatake, S., Kogishi, J., Hangai, M., Okazaki, K., Takahashi,
J. C., Saiki, M., Iwaki, M., and Honda, Y. (1999). Adenovirally expressed
basis broblast growth factor rescues photoreceptor cells in RCS rats. Invest.
Ophthalmol. Vis. Sci., 40(2):273279.
113. Bennett, J., Zehng, Y., Bajwa, R., Klatt, L., Li, Y., and Maguire, A. M.
(1998). Adenovirus-mediated delivery of rhodopsin-promoted bc1-2 results
in a delay in photoreceptor cell death in the rd/rd mouse.
114. Kumar-Singh, R., and Farber, D. B. (1998). Encapsidated adenovirus minichromosome-mediated delivery of genes to the retina: application to the rescue of photoreceptor degeneration. Hum. Mol. Genet., 7(12):18931900.
115. Isenmann, S., Klocker, N., Gravel, C., and Bahr, M. (1998). Short communication: protection of axotomized retinal ganglion cells by adenovirally delivered BDNF in vivo. Eur. J. Neurosci., 10(8):27512756.
116. Weise, J., Isenmann, S., Klocker, N., Kugler, S., Hirsch, S., Gravel, C., and
Bahr, M. (2000). Adenovirus-mediated expression of ciliary neurotrophic factor (CNTF) rescues axotomized rat retinal ganglion cells but does not support
axonal regeneration in vivo. Neurobiol. Dis., 7(3):212223.
117. Tsubota, K., Inoue, H., Ando, K., Ono, M., Yoshino, K., and Saito, I. (1998).
Adenovirus-mediated gene transfer to the ocular surface epithelium. Exp. Eye
Res., 67(5):531538.
118. Ali, R. R., Reichel, M. B., Byrnes, A. P., Stephens, C. J., Thrasher, A. J.,
Baker, D., Hunt, D. M., and Bhattacharya, S. S. (1998). Co-injection of
adenovirus expressing CTLA4-Ig prolongs adenovirally mediated lacZ reporter gene expression in the mouse retina. Gene Ther., 5(11):15611565.
119. Reichel, M. B., Ali, R. R., Thrasher, A. J., Hunt, D. M., Bhattacharya, S. S.,
and Baker, D. (1998). Immune response limit adenovirally mediated gene
expression in the adult mouse eye. Gene Ther., 5(8):10381046.
120. Sakamoto, T., Ueno, H., Sonoda, K., Hisatomi, T., Shimizu, K., Ohashi, H.,
Inomata, H. (2000). Blockage of TGF-beta by in vivo gene transfer of a
soluble TGF-beta type II receptor in the muscle inhibits corneal opacication,
edema and angiogenesis. Gene Ther., 7(22):19151924.

Copyright 2003 Marcel Dekker, Inc.

600

Tombran-Tink

121. Berns, K. L. (1996). Parvoviridae: the viruses and their replication. In: Fields
Virology (B. N. Fields, D. M. Knippe, and P. Howley, eds.). Lippincot-Raven,
Philadelphia, pp. 21732196.
122. McKeon, C., and Samulski, R. J. (1996). NIDDK Workshop on AAV
Vectors: Gene Transfer into Quiescent Cells. Hum. Gene Ther., 7:16151619.
123. Ali, R. R., Reichel, M. B., Thrasher, A. J., Levinsky, R. J., Kinnon, C.,
Kanuga, N., Hunt, D. M., and Bhattacharya, S. S. (1996). Gene transfer
into the mouse retina mediated by an adeno-associated viral vector. Hum.
Mol. Genet., 5(5):591594.
124. Flannery, J. G., Zolotukhin, S., Vaquero, M. I., LaVail, M. M., Muzyczka,
N., and Hauswirth, W. W. (1997). Ecient photoreceptor-targeted gene
expression in vivo by recombinant adeno-associated virus. Proc. Natl. Acad.
Sci. USA, 94(13):69166921.
125. Grant, C. A., Ponnazhagan, S., Wang, X. S., Srivastava, A., and Li, T. (1997).
Evaluation of recombinant adeno-associated virus as a gene transfer vector
for the retina. Curr. Eye Res., 16(9):949956.
126. Lai, Y. K., Rakoczy, P., Constable, I., and Rolling, F. (1998). Adeno-associated virus-mediated gene transfer into human retinal pigment epithelium
cells. Aust. NZ J. Ophthalmol., Suppl 1:S7779.
127. Guy, J., Qi, X., Muzyczka, N., and Hauswirth, W. W. (1999). Reporter
expression persists 1 year after adeno-associated virus-mediated gene transfer
to the optic nerve. Arch. Ophthalmol., 117(7):929937.
128. Dreyer, E. B., Vorwerk, C. K., Zurakowski, D., Simon, P. D., and Bennett, J.
(1999). Infection with adeno-associated virus may protect against excitotoxicity. Neuroreport, 10(14):28872890.
129. Bennett, J., Maguire, A. M., Cideciyan, A. V., Schnell, M., Glover, E., Anand,
V., Aleman, T. S., Chirmule, N., Gupta, A. R., Huang, Y., Gao, G. P.,
Nyberg, W. C., Tazelaar, J., Hughes, J., Wilson, J. M., and Jacobson, S. G.
(1999). Stable transgene expression in rod photoreceptors after recombinant
adeno-associated virus-mediated gene transfer to monkey retina. Proc. Natl.
Acad. Sci. USA, 96(17):99209925.
130. Acland, G. M., Aguirre, G. D., Ray, J., Zhang, Q., Aleman, T. S., Cideciyan,
A. V., Pearce-Kelling, S. E., Anand, V., Zeng, Y., Maguire, A. M., Jacobson,
S. G., Hauswirth, W. W., and Bennett, J. (2001). Gene therapy restores vision
in a canine model of childhood blindness. Nat. Genet., 28(1):9295.
131. Liang, F. Q., Dejneka, N. S., Cohen, D. R., Krasnoperova, N. V., Lem, J.,
Maguire, A. M., Dudus, L., Fisher, K. J., and Bennett, J. (2001). AAVmediated delivery of ciliary neurotrophic factor prolongs photoreceptor survival in the rhodopsin knockout mouse. Mol. Ther., 3(2):241248.
132. Ali, R. R., Sarra, G. M., Stephens, C., Alwis, M. D., Bainbridge, J. W.,
Munro, P. M., Fauser, S., Reichel, M. B., Kinnon, C., Hunt, D. M.,
Bhattacharya, S. S., and Thrasher, A. J. (2000). Restoration of photoreceptor
ultrastructure and function in retinal degeneration slow mice by gene therapy.
Nat. Genet., 25(3):245246.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

601

133. Jomary, C., Vincent, K. A., Grist, J., Neal, M. J., and Jones, S. E. (1997).
Rescue of photoreceptor function by AAV-mediated gene transfer in a mouse
mode of inherited retinal degeneration. Gene Ther., 4(7):683690.
134. Glorioso, J. C., DeLuca, N. A., and Fink, D. J. (1995). Development and
application of herpes simplex virus vectors for human gene therapy. An. Rev.
Microbiol., 49:675710.
135. Rodahl, E., and Haarr, L. (2000). A herpes simplex virus type I vector as
marker for retrograde neuronal tracing: characterization of lacZ transcription
and localization of labeled neuronal cells in sensory and autonomic ganglia
after inoculation of the anterior segment of the eye. Exp. Eye Res., 71(5):495
501.
136. Hudde, T., Rayner, S. A., De Alwis, M., Thrasher, A. J., Smith, J., Con, R.
S., George, A. J., and Larkin, D. F. (2000). Adeno-associated and herpes
simplex viruses as vectors for gene transfer to the corneal endothelium.
Cornea, 19(3):369373.
137. Spencer, B., Agarwala, S., Miskulin, M., Smith, M., and Brandt, C. R. (2000).
Herpes simplex virus-mediated gene delivery to the rodent visual system.
Invest. Ophthalmol. Vis. Sci., 41(6):13921401.
138. Wang, X., Appukuttan, B., Ott, S., Patel, R., Irvine, J., Song, J., Park, J. H.,
Smith, R., and Stout, J. T. (2000). Ecient and sustained transgene expression
in human corneal cells mediated by a lentiviral vector. Gene Ther., 7(3):196
200.
139. Galileo, D. S., Hunter, K., and Smith, S. B. (1999). Stable and ecient gene
transfer into the mutant retinal pigment epithelial cells of mitf(vit) mouse
using a lentiviral vector. Curr. Eye Res., 18(2):135142.
140. Yang, M., Wang, X. G., Stout, J. T., Chen, P., Hjelmeland, L. M.,
Appukuttan, B., and Fong, H. K. (2000). Expression of a recombinant
human RGR opsin in lentivirus-transduced cultured cells. Mol. Vis., 6:237
242.
141. Takahashi, M., Miyoshi, H., Verma, I. M., and Gage, F. H. (1999). Rescue
from photoreceptor degeneration in the rd mouse by human immunodeciency virus-mediated gene transfer. J. Virol., 73(9):78127816.
142. Miyoshi, H., Takahashi, M., Gage, F. H., and Verma, I. M. (1997). Stable and
ecient gene transfer into the retina using an HIV-based lentiviral vector.
Proc. Natl. Acad. Sci. USA, 94(19):1031910323.
143. Sikorski, R., and Peters, R. (1998). Gene therapy. Treating with HIV. Science,
282(5393):1438.
144. Tanelian, D. L., Barry, M. A., Johnston, S. A., Le, T., and Smith, G. (1997).
Controlled gene gun delivery and expression of DNA within the cornea.
Biotechniques, 23(3):484488.
145. Sakamota, T., Oshima, Y., Nakagawa, K., Ishibashi, T., Inomata, H., and
Sueishi, K. (1999). Target gene transfer of tissue plasminogen activator to
cornea by electric pulse inhibits intracameral brin formation and corneal
cloudiness. Hum. Gene Ther., 10(15):25512557.

Copyright 2003 Marcel Dekker, Inc.

602

Tombran-Tink

146. Oshima, Y., Sakamoto, T., Yamanaka, I., Nishi, T., Ishibashi, T., and
Inomata, H. (1998). Targeted gene transfer to corneal endothelium in vivo
by electric pulse. Gene Ther., 5(10):13471354.
147. Noisakran, S. J., and Carr, D. J. (2000). Therapeutic ecacy of DNA encoding IFN-alpha1 against corneal HSV-1 infection. Curr. Eye Res., 20(5):405
412.
148. Neuner-Jehl, M., Berghe, L. V., Bonnel, S., Uteza, Y., Benmezaine, F.,
Rouillot, J. S., Marchant, D., Kobetz, A., Duer, J. L., Menasche, M., and
Abitbol, M. (2000). Ocular cell transfection with human basic broblast
growth factor gene delays photoreceptor cell degeneration in RCS rats.
Hum. Gene Ther., 11(13):18751890.
149. Farhood, H., Gao, X., and Son, K. (1994). Cationic liposome for direct gene
transfer in therapy of cancer and other diseases. Ann. NY Acad. Sci., 716:23
34.
150. Felgner, P. L., Tsai, Y. J., Sukhu, L., Wheeler, C. J., Manthorpe, M.,
Marshall, J., and Cheng, S. H. (1995). Improved cationic lipid formulations
for in vivo gene therapy. Ann. NY Acad. Sci., 772:126139.
151. Bedu-Addo, F. K., Tang, P., Xu, Y., and Huang, L. (1996). Interaction of
poly-ethyleneglycol-phospholipid conjugates with cholesterol-phosphatidylcholine mixtures: sterically stabilized liposome formulations. Pharmacol.
Res., 13:718724.
152. Singh, M., Ferdous, A. J., Kanikkannan, N., and Faulkner, G. (2001). Stealth
monensin immunoliposomes as potentiator of immunotoxins in vitro. Eur. J.
Pharm. Biopharm., 52(1):1320.
153. Gabizon, A. (1993). Tailoring liposomes for cancer drug delivery: from the
bench to the clinic. A. Biol. Clin. (Paris), 51(9):811813.
154. Needham, D., McIntosh, T. J., and Lasic, D. D. (1992). Repulsive interactions
and mechanical stability of polymer-grafted lipid membranes. Biochim. Acta,
1108(1):4048.
155. Allen, T. M., and Hansen, C. (1991). Pharmacokinetics of stealth versus conventional liposomes: eect of dose. Biochim. Biophys. Acta, 1068(2):133141.
156. Allen, T. M., Mehra, T., Hansen, C., and Chin, Y. C. (1992). Stealth liposomes: an improved sustained release system for 1-beta-D-arabinofuranosylcytosine. Cancer Res., 52(9):24312439.
157. Gabizon, A., and Martin, F. (1997). Polyethylene glycol-coated (pegylated)
liposomal doxorubicin. Rationale for use in solid tumors. Suppl 4:1521.
158. Newman, M. S., Colbern, G. T., Working, P. K., Engbers, C., and Amantea,
M. A. (1999). Comparative pharmacokinetics, tissue distribution, and therapeutic eectiveness of cisplatin encapsulated in long-circulating, pegylated
liposomes (SPI-077) in tumor-bearing mice. Cancer Chemother. Pharmacol.,
43(6):524.
159. Coa, Y., and Suresh, M. R. (2000). Bispecic Mab aided liposomal drug
delivery. J. Drug Target, 8(4):257266.
160. Cristiano, R. J., and Roth, J. A. (1995). Molecular conjugates: a targeted gene
delivery vector for molecular medicine. J. Mol. Med., 73(10):479486.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

603

161. Gabizon, A. A. (2001). Pegylated liposomal doxorubicin: metamorphosis of

an old drug into a new form of chemotherapy. Cancer Invest., 19(4):424436.
162. Gabizon, A. A. (2001). Stealth liposomes and tumor targeting: one step
further in the quest for the magic bullet. Clin. Cancer Res., 7(2):223225.
163. Ishida, T., Iden, D. L., and Allen, T. M. (1999). A combinatorial approach to
producing sterically stabilized (Stealth) immunoliposomal drugs. FEBS Lett.,
460(1):129133.
164. Riveria, E., Valero, V., Syrewicz, L., Rahman, Z., Esteva, F. L., Theriault, R.
L., Rosales, M. M., Booser, D., Murray, J. L., Bast, R. C. Jr, and Hortobagyi,
G. N. (2001). Phase I study of stealth liposomal doxorubicin in combination
with gemcitabine in the treatment of patients with metastatic breast cancer. J.
Clin. Oncol., 19(6):17161722.
165. Schatzlein, A. G. (2001). Non-viral vectors in cancer gene therapy: principles
and progress. Anticancer Drugs, 12(4):275304.
166. Shimada, K., Matsuo, S., Sadzuka, Y., Miyagishima, A., Nozawa, Y., Hirota,
S., and Sonobe, T. (2000). Determination of incorporated amounts of
poly(ethylene glycol)-derivatized lipids in liposomes for the physicochemical
characterization of stealth liposomes. Int. J. Pharm., 203(12):255263.
167. Singh, M., Ferdous, A. J., Kanikkannan, N., and Faulkner, G. (2001). Stealth
monensin immunoliposomes as potentiator of immunotoxins in vitro. Eur. J.
Pharm. Biopharm., 52(1):1320.
168. Takeuchi, H., Kojima, H., Yamamoto, H., and Kawashima, Y. (2001).
Evaluation of circulation proles of liposomes coated with hydrophilic polymers having dierent molecular weights in rats. J. Control. Release, 75(1
2):8391.
169. Vaage, J., Donovan, D., Wip, E., Abra, R., Colbern, G., Uster, P., and
Working, P. (1999). Therapy of a xenografted human colonic carcinoma
using cisplatin or doxorubicin encapsulated in long-circulating pegylated
stealth liposomes. Int. J. Cancer, 80(1):134137.
170. Miller, A. D. (1999). Non-Viral Delivery Systems for Gene Therapy in
Understanding Gene Therapy. Bios Scientic Publisher, pp. 4360.
171. Yonemitsu, Y., Alton, E. W. F . W., Komori, K., Yoshizumi, T., Sugimachi,
K., and Kaneda, Y. (1998). HVJ (Sendai virus) liposome-mediated gene transfer: current status and future perspective. Int. J. Oncol., 12:12771285.
172. Remy, J. S., Kichler, A., Mordvinov, V., Schuber, F., and Behr, J. P. (1995).
Targeted gene transfer into hepatoma cells with lipopolyamine-condensed
DNA particles presenting galactose ligands: a stage toward articial viruses.
Proc. Natl. Acad. Sci. USA, 92(5):17441748.
173. Spragg, D. D., Alford, D. R., Greferath, R., Larsen, C. E., Lee, K. D.,
Gurtner, G. C., Cybulsky, M. I., Tosi, P. F., Nicolau, C., and Gimbrone,
M. A. Jr. (1997). Immunotargeting of liposomes to activated vascular
endothelial cells: a strategy for site-selective delivery in the cardiovascular
system. Proc. Natl. Acad. Sci. USA, 94(16):87958800.

Copyright 2003 Marcel Dekker, Inc.

604

Tombran-Tink

174. Simoes, S., Slepushkin, V., Gaspar, R., de Lima, M. C., and Duzgunes, N.
(1998). Gene delivery by negatively charges ternary complexes of DNA, cationic liposomes and transferrin of fusigenic peptides. Gene Ther. 5(7):955964.
175. Li, S., and Huang, L. (1997). In vivo gene transfer via intravenous administration of cationic lipid-protamine-DNA (LPD) complexes. Gene Ther.,
4(9):891900.
176. Li, S., Rizzo, M. A., Bhattachrya, S., and Huang, L. (1998). Characterization
of cationic lipid-protamine-DNA (LPD) complexes for intravenous gene
delivery. Gene Ther., 5(7):930937.
177. Abul-Hassan, K., Walmsley, R., and Boulton, M. (2000). Optimization of
non-viral gene transfer to human primary retinal pigment epithelial cells.
Curr. Eye Res., 20(5):361366.
178. Matsuo, T., Masuda, I., Yasuda, T., and Matsuo, N. (1996). Gene transfer to
the retina of rat by liposome eye drops. Biochem. Biophys. Res. Commun.,
219(3):947950.
179. Takahashi, T., Nakamura, T., Hayashi, A., Kamei, M., Nakabayashi, M.,
Okada, A. A., Tomita, N., Kaneda, Y., and Tano, Y. (2000). Inhibition of
experimental choroidal neovascularization by overexpression of tissue inhibitor of metalloproteinases-3 in retinal pigment epithelium cells. Am. J.
Ophthalmol., 130(6):774781.
180. Zamecnik, P. C., and Stephenson, M. L. (1978). Inhibition of Rous sarcoma
virus replication and cell transformation by a specic oligodeoxynucleotides.
Journal, 75(1):280284.
181. Crooke, S. T., and Lebleu, B. (1993). Antisense Research and Applications.
CRC Press, Boca Raton, FL.
182. Agrawal, S. (1996). Antisense oligonucleotides: towards clinical trials. Trends
Biotechnol., 14(10):376387.
183. Crooke, S. T. (1998). An overview of progress in antisense therapeutics.
Antisense Nucleic Acid Drug Dev., 8(2):115122.
184. Matsukura, M., Shinozuka, K., Zon, G., Mitsuya, H., Reitz, M., Cohen, J. S.,
and Broder, S. (1987). Phosphorothioate analogs of oligodeoxynucleotides:
inhibitors of replication and cytopathic eects of human immunodeciency
virus. Proc. Natl. Acad. Sci. USA, 84(21):77067710.
185. Stein, C. A., Subasinghe, C., Shinozuka, K., and Cohen, J. S. (1988).
Physicochemical properties of phosphorothioate oligodeoxynucleotides.
Nucleic Acids Res., 16(8):32093221.
186. LaPlanche, L. A., James, T. L., Powell, C., Wilson, W. D., Uznanski, B., Stec,
W. J., Summers, M. F., and Zon, G. (1996). Phosphorothioate-modied oligodeoxyribonucleotides, III. NMR and UV spectroscopic studies of the RpRp, Sp-SP, and Rp-Sp duplexes, [d(GGSAATTCC)]2, derived from diastereomeric O-ethyl phosphorothioates. Nucleic Acids Res., 14(22):90819093.
187. Srinivasan, S. K., and Iversen, P. (1995). Review of in vivo pharmacokinetics
and toxicology of phosphorothioate oligonucleotides. J. Clin. Lab. Anal.,
9(2):129137.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

605

188. Stepkowski, S. M. (2000). Development of antisense oligodeoxynucleotides

for transplantation. Curr. Opin. Mol. Ther., 2(3):304317.
189. Sohail, M., and Southern, E. M. (2000). Hybridization of antisense reagents to
RNA. Curr. Opin. Mol. Ther., 2(3):264271.
190. Srinivasan, S. K., and Iversen, P. (1995). Review of in vivo pharmacokinetics
and toxicology of phosphorothioate oligonucleotides. J. Clin. Lab. Anal.,
9(2):129137.
191. Lambert, G., Fattal, E., and Couvreur, P. (2001). Nanoparticulate systems for
the delivery of antisense oligonucleotides. Adv. Drug Deliv. Rev., 47(1):99112.
192. Lebedeva, I., and Stein, C. A. (2001). Antisense oligonucleotides: promise and
reality. Annu. Rev. Pharmacol. Toxicol., 41:403419.
193. de Smet, M. D., Meenken, C. J., and van den Horn, G. J. (1999).
Fomivirsena phosphorothioate oligonucleotide for the treatment of CMV
retinitis. Ocul. Immunol. Inamm., 7(34):189198.
194. Flores-Aguilar, M., Besen, G., Vuong, C., Tatebayashi, M., Munguia, D.,
Gangan, P., Wiley, C. A., and Freeman, W. R. (1997). Evaluation of retinal
toxicity and ecacy of anti-cytomegalovirus and anti-herpes simplex virus
antiviral phosphorothioate oligonucleotides ISIS 2922 and ISIS 4015. J.
Infect. Dis., 175(6):13081316.
195. Perry, C. M., and Balfour, J. A. (1999). Fomivirsen. Drugs, 57(3):375380.
196. Detrick, B., Nagineni, C. N., Grillone, L. R., Anderson, K. P., Henry, S. P.,
and Hooks, J. J. (2001). Inhibition of human cytomegalovirus replication in a
human retinal epithelial cell model by antisense oligonucleotides. Invest.
Ophthalmol. Vis. Sci., 42(1):163169.
197. Robinson, G. S., Pierce, E. A., Rook, S. L., Foley, E., Webb, R., and Smith,
L. E. (1996). Oligodeoxynucleotides inhibit retinal neovascularization in a
murine model of proliferative retinopathy. Proc. Natl. Acad. Sci. USA,
93(10):48514856.
198. Capeans, C., Pineiro, A., Dominguez, F., Loidi, L., Buceta, M., Carneiro, C.,
Garcia-Caballero, T., and Sanchez-Salorio, M. (1998). A c-myc antisense
oligonucleotide inhibits human retinal pigment epithelial cell proliferation.
Exp. Eye Res., 66(5):581589.
199. Isenmann, S., Engel, S., Gillardon, F., and Bahr, M. (1999). Bax antisense
oligonucleotides reduce axotomy-induced retinal ganglion cell death in vivo
by reduction of Bax protein expression. Cell Death Dier., 6(7):673682.
200. Barnett, N. L., and Pow, D. V. (2000). Antisense knockdown of GLAST, a
glial glutamate transporter, compromises retinal function. Invest. Ophthalmol.
Vis. Sci., 41(2):585591.
201. Rickman, D. W., and Rickman, C. B. (1996). Suppression of trkB expression
by antisense oligonucleotides alters a neuronal phenotype in the rod pathway
of the developing rat retina. Proc. Natl. Acad. Sci. USA, 93(22):1256412569.
202. Roy, S., Zhang, K., Roth, T., Vinogradov, S., Kao, R. S., and Kabanov, A.
(1999). Reduction of bronectin expression by intravitreal administration of
antisense oligonucleotides. Nat. Biotechnol., 17(5):476479.

Copyright 2003 Marcel Dekker, Inc.

606

Tombran-Tink

203. Goodchild, J. (2000). Hammerhead ribozymes: biochemical and chemical considerations. Curr. Opin. Mol. Ther., 2(3):272281.
204. Kiehntopf, M., Esquival, E. L., Brach, M. A., and Herrmann, F. (1995).
Ribozymes: biology, biochemistry, and implications for clinical medicine. J.
Mol. Med., 73(2):6571.
205. Michel F., and Ferat, J. L. (1995). Structure and activities of group II introns.
Annu. Rev. Biochem., 64:435461.
206. Lewin, A. S., Drenser, K. A., Hauswirth, W. W., Nishikawa, S., Yasumura,
D., Flannery, J. G., and LaVail, M. M. (1998). Ribozyme rescue of photoreceptor cells in a transgenic rat model of autosomal dominant retinitis pigmentosa. Nat. Med., 4(8):967971.
207. Hauswirth, W. W., and Lewin, A. S. (2000). Ribozyme uses in retinal gene
therapy. Prog. Retin. Eye Res., 19(6):689710.
208. Shaw, L. C., Skold, A., Wong, F., Petters, R., Hauswirth, W. W., and Lewin,
A. S. (2001). An allele-specic hammerhead ribozyme gene therapy for a
porcine model of autosomal dominant retinitis pigmentosa. Mol. Vis.,
26(7):613.
209. LaVail, M. M., Yasumura, D., Matthes, M. T., Drenser, K. A., Flannery, J.
G., Lewin, A. S., and Hauswirth, W. W. (2000). Ribozyme rescue of photoreceptor cells in P23H transgenic rats: long-term survival and late-state therapy. Proc. Natl. Acad. Sci. USA, 97(21):1148811493.
210. Hauswirth, W. W., LaVail, M. M., Flannery, J. G., and Lewin, A. S. (2000).
Ribozyme gene therapy for autosomal dominant retinal disease. Clin. Chem.
Lab. Med., 38(2):147153.
211. Shaw, L. C., Whalen, P. O., Drenser, K. A., Yan, W., Hauswirth, W. W., and
Lewin, A. S. (2000). Ribozymes in treatment of inherited retinal disease.
Methods Enzymol., 316:761776.
212. Drenser, K. A., Timmers, A. M., Hauswirth, W. W., and Lewin, A. S. (1998).
Ribozyme-targeted destruction of RNA associated with autosomal-dominant
retinitis pigmentosa. Invest. Ophthamol. Vis. Sci., 39(5):681689.
213. Tombran-Tink, J., and Johnson, L. V. (1989). Neuronal dierentiation of
retinoblastoma cells induced by medium conditioned by human RPE cells.
Invest. Opthalmol. Vis. Sci., 30:17001707.
214. Dawson, D. W., Volpert, O. V., Gillis, P., Crawford, S. E., Xu, H., Benedict,
W., and Bouck, N. P. (1999). Pigment epithelium-derived factor: a potent
inhibitor of angiogenesis. Science, 285(5425):245248.
215. McGinnis, J. G., Chen, W., Tombran-Tink, J., Mrazek, D. A., Lerious, V.,
and Cao, W. (1999). Retinal neurons in primary cell culture: inhibition of
apoptosis by pigment epithelial derived factor (PEDF). In: Retinal
Degenerative Diseases and Experimental Therapy, Kluwer Academic/Plenum
Publishers, New York, pp. 527537.
216. Cao, W., Tombran-Tink, J., Chen, W., Mrazek, D., Elias, R., and McGinnis,
J. F. (1999). Pigment Epithelium-Derived Factor protects cultured retinal
neurons against hydrogen peroxide-induced cell death. J. Neurosci. Res.,
57(6):789800.

Copyright 2003 Marcel Dekker, Inc.

Experimental Approaches to Retinal Diseases

607

217. Cao, W., Tombran-Tink, J., Elias, R., Sezates, Mrazek, D., and McGinnis, J.
F. (2001). In vivo protection of photoreceptors from light damage by pigment
epithelium-derived factor. Invest. Ophthalmol. Vis. Sci., 42:16461652.
218. Jablonski, M. M., Tombran-Tink, J., Mrazek, D. A., and Iannaccone, A.
(2000). Pigment Epithelium-Derived Factor supports normal development
of photoreceptor neurons and opsin expression after retinal pigment epithelium removal. J. Neurosci., 20(19):71497157.
219. Jablonski, M. M., Tombran-Tink, J., Mrazek, D. A., and Iannaccone, A.
(2001). Pigment Epithelium-Derived Factor supports normal Muller cell
development and Glutamine synthetase expression after removal of the retinal
pigment epithelium. Glia., 35:1425.
220. Tombran-Tink, J., Pawar, H., Swaroop, A., and Chader, G. J. (1994).
Localization of the gene for pigment epithelium-derived factor to chromosome 17p13.1 and expression in cultured retinoblastoma cells. Genomics,
19:266272.
221. Goliath, R., Tombran-Tink, J., Chader, G. J., Ramser, R., and Greenberg, J.
(1996). The gene for PEDF, a retinal growth factor is a prime candidate for
retinitis pigmentosa and is tightly linked to the RP13 locus on chromosome
17p13.3. Mol. Vis., 2:5.
222. Tombran-Tink, J., Chader, G., and Koenekoop, R. (1997). Molecular analysis
of the human PEDF gene, a candidate gene for retinal degeneration: localization to 17p13.3. Degen. Retin. Dis., 3:245254.
223. Greenberg, J., Goliath, R., Tombran-Tink, J., Chader, G. J., and Ramesar, R.
(1997). Growth factors in the retina: pigment epithelium-derived factor
(PEDF) now ne mapped to 17p13.3 and tightly linked to the RP13 locus.
Degen. Retin. Dis., 3:291294.
224. Koenekoop, R. K., Loyer, M., Pina, A. L., Davidson, J., Robitaille, J.,
Maumenee, I., and Tombran-Tink, J. (1999). Four polymorphic variations
in the PEDF gene identied during the mutation screening of patients with
leber congenital amaurosis. Mol. Vis., 5:10.
225. Tombran-Tink, J., Shivaram, S. M., Chader, G. J., Johnson, L. V., and Bok,
D. (1995). Expression, secretion and age-related downregulation of pigment
epithelium-derived factor, a serpin with neurotrophic activity. J. Neurosci.,
15(7):49925003.
226. Tombran-Tink, J., Li, A., Johnson, M. A., Johnson, L. V., and Chader, G. J.
(1992). Neurotrophic activity of interphotoreceptor matrix on human Y79
retinoblastoma cells. J. Comp. Neurol., 317:175186.
227. Tombran-Tink, J., Mazaruk, K., Chung, D., Linker, T., Chader, G. J., and
Rodriguez, I. (1996). Organization, evolutionary conservation, expression and
unusual Alu density of the human gene for pigment epithelium-derived factor,
a unique neurotrophic serpin. Mol. Vis., 2:11.
228. Ogata, N., Tombran-Tink, J., Nishikawa, M., Nishimura, T., Mitsuma, Y.,
Sakamoto, T., and Matsumura, M. (2001). Pigment Epithelium-derived factor
in the vitreous is low in diabetic retinopathy and high in rhegmatogenous
retinal detachment. Am. J. Ophthalmol., 132(3):378382.

Copyright 2003 Marcel Dekker, Inc.

608

Tombran-Tink

229. Cayouette, M., Smith, S. B., Becerra, S. P., and Gravel, C. (1999). Pigment
epithelium-derived factor delays the death of photoreceptors in mouse models
of inherited retinal degenerations. Neurobiol. Dis., 6:523532.
230. LaVail, M. M., Yasumura, D., Mathes, M. T., Lau-Villacorta, C., Unoki, K.,
Sung, C. H., and Steinberg, R. H. (1998). Protection of mouse photoreceptors
by survival factors in retinal degenerations. Invest. Ophthalmol. Vis. Sci.,
39:592602.

Copyright 2003 Marcel Dekker, Inc.

19
Gene, Oligonucleotide, and
Ribozyme Therapy in the Eye
Sudip K. Das
Idaho State University, Pocatello, Idaho, U.S.A.

Keith J. Miller
Bristol-Myers Squibb Company, Pennington, New Jersey, U.S.A.

I.

INTRODUCTION

Both antisense oligonucleotide as well as gene therapy aim at the transfer of

genetic material to the target cells for preventing or altering a particular
disease process. In gene therapy a good gene is inserted into the cell to repair
or replace the faulty gene to produce health-giving proteins, while an oligonucleotide can block the disease-causing protein. With the advances in the
identication of the major genes responsible for causing various diseases, an
increased focus on the use of therapeutic genes and oligonucleotides has
developed, including treatments for ocular diseases. Inherited retinal degeneration is one of the most common causes of blindness in the western world,
for which there is no therapy at all. The most researched form of inherited
retinal degeneration is retinitis pigmentosa (RP; damage to photoreceptors
of the retina by a single abnormal gene) with an occurrence up to 1:3000. A
number of important developments have been reported in last few years in
identifying genetic defects in RP. Gene and oligonucleotide therapy has been
successfully applied in retinal diseases, including the introduction of the
antisense oligonucleotide Vitravene1 for cytomegalovirus (CMV) retinitis
in patients with acquired immunodeciency syndrome (AIDS) and other
immuno suppressed states. Thus, an important milestone in antisense therapy for the eye has been reached. In this chapter we will discuss the princi609

Copyright 2003 Marcel Dekker, Inc.

610

Das and Miller

ples of gene, oligonucleotide, and ribozyme therapy with an emphasis on

delivery of these drugs to the eye.

II.

GENE THERAPY

The goal of gene therapy is to replace the defective or missing gene in a

human cell to correct its malfunction. The most extensively studied disease
in which gene therapy has been attempted is cystic brosis; treatment
involves the replacement of the mutant cystic brosis transmembrane conductance regulator (CFTR) gene (1). The main principle of gene therapy
involves transfection of cells with nonresident genes to accomplish in situ (a)
removal of a mutant gene and replacing it with a normal gene, (b) correction
of a mutant gene by site-directed recombination, and (c) improvement of a
gene function to enable production of sucient levels of product to supplement or replace normal expression (2,3). The progress of gene therapy,
based on systemic delivery, has been slowed recently by reports of adverse
events, including the death of a patient (4). On the other hand, a recent
study showed the ecacy of a gene therapy product that encodes clotting
factor VIII in two hemophilia patients who had no spontaneous bleeding
episodes for nearly one year (5).
The eye is an attractive target for gene therapy because of its accessibility and its immune privilege. In ophthalmology, the concept of gene
therapy has attracted interest for a number of retinal diseases, including
retinitis pigmentosa, age-related macular degeneration, and proliferative
vitreoretinopathy (6). In retinal disorders, specic genetic defects have
been found in 20 of the 87 mapped genes causing genetic eye disorders,
and most of the genetic defects result from single gene defects that are
both highly disabling and potentially correctable (7). The clinical transplantation of retinal pigmental epithelium with fetal RPE cells in the subretinal
space of patients with age-related macular degeneration could cause cystoidlike macular edema due to slow graft rejection (8). Therefore, in vivo administration of gene therapy drugs has high potential in treating those diseases.
The current status of gene therapy in the eye is restricted to somatic gene
therapy (the gene is introduced to somatic cells) performed by in vivo gene
transfer (administration of genes locally or systemically).
Early studies with gene therapy involved bombardment of cells ex vivo
with naked DNA. However, that process had several limitations, including
nonapplicability in clinical medicine and the absence of prolonged gene
expression in the target tissue. With the advances of molecular biology
and virology and better understanding of viral genome, it is now possible
to investigate a number of viral vectors for gene therapy. Hauswirth and

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

611

Beaufrere have listed four basic prerequisites that should be fullled for
ocular gene therapy (9):
1. An ecient and nontoxic gene delivery technique should be
available.
2. The genetic basis of the disease should be characterized so that
an appropriate therapeutic approach can be taken.
3. Expression of the therapeutic gene needs to be properly controlled in terms of the types of cells that do or do not support
expression.
4. An experimental animal model of the disease should be available
for preclinical testing of therapy.
Since the systemic delivery of genes does not reach, to a satisfactory
level, the ocular tissues, most gene therapy agents (except for gene delivery
to surface corneal epithelium) need to be administered through intravitreal
or intracameral injection. Because the ow of uids in the eye is from posterior to anterior segment, the delivery of therapeutic agent in the vitreal
chamber leads to distribution of the drug in the anterior cell linings and
also the general circulation in addition to distribution in the vitreal cell
linings.

A.

Viral-VectorMediated Gene Therapy

Gene delivery is the main obstacle preventing the full exploitation of the
genetic information at our disposal in this postgenomic era (10). The size,
charge, surface properties, and interaction with blood components determine the distribution of a colloidal system or macromolecule in the body
(11). Due to the large size of DNA, it does not diuse well through the
endothelium, keratinized tissues, and specialized barriers in the body. The
availability of large DNA to specied tissues is dependent mainly on the
blood ow. Moreover, those drugs administered through intravascular
routes are cleared from the body by rst-pass eect and by the mononuclear
phagocyte system. In addition, due to enzymatic action, DNA is prone to
degradation and fragmentation within a short time after vascular administration. Therefore systemic gene delivery to achieve sucient concentration in ocular tissues is extremely dicult. Normally two types of gene
delivery techniques are used in in vivo gene therapy: viral-mediated and
nonviral-mediated. Due to the natural ability to infect cells eciently, a
number of viruses, such as adenovirus, retrovirus, adeno-associated virus
and herpesvirus, have been widely studied for their ability to deliver genes to
specic tissues (12). Most viruses used in gene therapy are inactivated at the

Copyright 2003 Marcel Dekker, Inc.

612

Das and Miller

DNA replication level so they do not multiply outside the specic laboratory cell lines.
The hypothesis that genetic disorders can be treated using viralmediated gene transfer became reality in 1990, with the rst phase I
gene therapy clinical trial for the treatment of adenosine deaminase
(ADA) deciency (13). Also in 1990, retroviral-mediated gene therapy
was conducted in patients with melanoma with success (14). A comparison
of the viral vectors for gene delivery is illustrated, with an emphasis on
ocular targets, in Table 1. Of the gene transfer protocols listed, by viral
vectors as well as by number of patients in clinical trials, retroviral vectors
account for the largest number (38.3% by protocols and 51% by number
of patients). This is followed by adenovirus (25.6% by protocols and 18%
by number of patients), poxvirus (6.4% by protocols and 2.6% by number
of patients), adeno-associated virus (1.9% by protocols and 1% by number of patients), and herpes simplex virus (0.6% by protocols and 0.6% by
number of patients) (15).

B.

Retrovirus in Gene Therapy

A retrovirus is a positive strand, diploid, RNA virus that replicates through

a DNA intermediate (18). On infecting a cell, the RNA genome is copied
onto double-stranded DNA, which is integrated into the genome of the
infected cells. Murine leukemia virus (MLV) is the recombinant retrovirus
most widely used for gene delivery. For delivery, the plasmid carrying the
cloned gene is transfected into mouse cells that express the genes of the
retrovirus; also, the envelope protein of the MLV is changed so that it
infects human cells. Retroviruses can infect a wide variety of cells in a
number of species at a very slow rate, only a few copies of the retrovirus
integrating per genome. However, retroviruses can only infect actively dividing cells (19), as the breakdown of nuclear membrane is essential for the
transport of preintegration complex to the nucleoplasm. Therefore the lack
of infecting ability of retroviruses precludes their use in nondividing cells
like neuronal or some ocular tissues. Although the chance of mutagenesis is
small, the retroviruses can be carcinogenic as they integrate themselves in
the genome of the infected cells. A retrovirus binds to a new host cell by
virtue of the interaction of the Env glycoprotein (one of the three basic
polyproteins that retroviral genome codes for) with an appropriate cellular
receptor. This interaction triggers a series of events that ultimately lead to
the fusion of the lipid envelope surrounding the virus with the target cell
membrane. In addition to the limitation of foreign sequences in the viral
genome, the stability of engineered vectors could also be a concern.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

C.

613

Adenovirus in Gene Therapy

Human adenoviruses are a group of large, nonenveloped, double-stranded

DNA viruses that have been isolated from respiratory tract infections, conjunctivitis, and infant diarrhea (20,21). The pathology is primarily from
inammation and loss of infected epithelial cells (22). Adenoviruses replicate
in two main phases: early and late. In the early phase, the viruses adsorb and
penetrate (via receptor-mediated endocytosis), followed by transcription
and translation of an early set of genes. By removing the early region
genes it is possible to make the adenoviruses replication defective. The
late phase consists of the onset of DNA replication followed by expression
of viral genes and assembly of virions (23). These viruses are fairly stable
and responsive to purication and concentration by cesium chloride density
centrifugation and stored frozen until use. The DNA of adenoviruses rarely
integrates to the target genome and is therefore gradually lost after cell
division and thus is not oncogenic in humans. Their genome is completely
dened and modied, and recombinant viruses can be produced in large
quantities at a high concentration without compromising their ability to
infect cells (24,25). Moreover, adenoviruses can infect a wide variety of
cells and have a broad host range. However, they are highly immunogenic
and could originate inammatory and toxic reactions in the host (26). Lifelong treatment of genetic disorders requires periodic virus administration,
which is contraindicated due to humoral immune response. The route of
administration has signicant eect on transduction, being highest in the
immediate vicinity of the administration. In spite of several disadvantages,
adenoviruses are preferred over the retroviruses for gene therapy mainly due
to; (1) high titer values, (2) incorporation of bigger DNA molecule, maximum of 78 kb, possibly involving the complete deletion of the viral genome.

D.

Adeno-Associated Virus Vector

Adeno-associated virus (AAV) is a small, DNA-containing virus that

belongs to the family Parvoviridae. It obtained its name because once it
was mistakenly identied as an adenovirus and was found to be dependent
on adenovirus for its replication (27). AAV has not been associated with any
disease, but it has been isolated in humans in association with adenoviruses.
For replication, it requires co-infection with either an adenovirus or herpes
simplex virus as a helper virus (28). However, it replicates in many cell lines
with wide cell and tissue specicity (29) provided the appropriate helper
virus is present. Moreover, it can infect nondividing cells (30) and has a
property of site-specic integration of its genome (31). However, recombi-

Copyright 2003 Marcel Dekker, Inc.

614

Table 1 Comparison of Viral Vectors for Gene Delivery

Vectors
Retroviruses

Adenoviruses

Lentiviruses

Relatively high titers (106 107 cfu/mL)

Stable gene expression
Integrate to host genome
No toxic eects on infected cells
Only infect dividing cells
Insert capacity in the virion is in the
range of 10 kb
Very high titers (1010 pfu/mL)
Transiently high levels of gene
expression
Do not integrate to host genome
They can also infect nondividing cells
Large DNA inserts can be
accommodated in the vector (78 kb)
Stable gene expression
Can infect nondividing cells
Total insert capacity in the virion in
the range of 10 kb
Can be pseudotyped with retroviral
or VSV G envelopes leading to broad
tropism

Ocular target

Disadvantages

Does not infect

nondividing cells

Random insertion of viral

genome may possibly lead
to mutagenesis
Possibility of replication
competent virus formation by
homologous recombination

Corneal endothelium
Corneal epithelium
Iris
Muller cells
Retinal pigmental
epithelium
Photoreceptor
Trabecular meshwork
Photoreceptor
Retinal pigmental
epithelium

Not suitable for long-term

expression due to the lack of
integration into host genome
Host immune response
Complicated vector genome

Possible proviral insertional

mutagenesis in target cells
Serum conversion to HIV-1
Presence of HIV-1 accessory
proteins may represent
hazard in humans, in spite
of lack of HIV-1 infection

Das and Miller

Copyright 2003 Marcel Dekker, Inc.

Characteristics

Source: Refs. 16, 17.

Dicult to obtain high titers

Wide range of cells can be infected,
including ones that do not divide
Insert capacity is limited, 4.14.9 kb
Ability of the virus to establish latent
infection by viral genome integration
into cell genome
Nonpathogenic, nontoxic

Photoreceptor
Retinal pigmental
epithelium

Limited capacity for foreign

genes
Requires helper adeno- or
herpesvirus for replication
Lack of specic integration
for recombinant adenoassociated vectors (vira)
lined rod cell function for
a minimum of 8 months after
single

Gene, Oligonucleotide, and Ribozyme Therapy

Adeno-associated
viruses

615

Copyright 2003 Marcel Dekker, Inc.

616

Das and Miller

nant adeno-associated viruses do not share the integration site specicity of

wild-type AAV. Site-specic integration is especially advantageous as it
reduces the risks of transformation of the target cell through insertional
mutagenesis. The virus particles are heat stable and resistant to a variety
of chemicals such as chloroform and alcohol. The specic advantages of
using AAVthe lack of need for integration specicity in recombinant
AAV and its ability to infect nondividing cellshas made this vector very
popular for gene delivery. The main problem of AAV in gene therapy lies in
accommodating large foreign genes, the maximum range being 4.14.9 kb
(32) that can be replaced completely by the gene of interest, thus totally
avoiding any immunity problem associated with viral proteins upon transduction of the target cell.
E. Lentivirus
Due to the fact that lentivirus vectors are based on HIV or SIV (human and
simian immunodeciency viruses), a big hurdle in using these vectors for
gene delivery involves deleting substantial segments of HIV genome.
Moreover, since the integration of HIV-based vectors appears to be random, there is a possibility of integration-induced mutagenesis.
F. Current Status of Viral-VectorMediated Gene Therapy
in the Eye
Adenoviral vectors are able to enter a number of retinal tissues, including
retinal pigment epithelium, but they do not eciently transduce mature
wild-type photoreceptors. Recombinant adenovirus carrying wild-type
cDNA has been used for transient rescue of degenerating photoreceptors
in the rd mouse. However, one of the major problems is the inammatory
response caused by adenovirus (Ad5)-mediated gene transfer. Homan et al.
studied the cell-mediated immune response and stability of intraocular
transgene expression after adenovirus-mediated gene delivery (33). It has
been demonstrated that with the replication-decient recombinant adenovirus, vectors based on adenovirus type 5 can be used for direct gene transfer
to conjunctival and corneal epithelium in vitro and in vivo (34). However
there was nonspecic upregulation of inammatory cytokines IL-6 and IL-8
in conjunctival epithelium that was suppressed by betamethasone.
Adenoviral vectormediated transfer of the gene for catalase, the reactive
oxygen species scavenger that suppresses experimental optic neuritis, to the
optic nerve in animals with experimental optic neuritis showed signicant
reduction of demyelination and disruption of the blood-brain barrier (35).
Another problem with recombinant adenovirus is that transduction of cells

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

617

by adenovirus vectors results in only short-term transgene expression. Longterm eects can be obtained with transplantation of cornea infected ex vivo
with adenoviral vectors (36). Based on the fact that glaucomatous neurons
die by apoptotic cell death, the use of gene therapy agents targeted to the
ganglion cells via intravitreal or via brain delivery has been found to be
promising. Delivery of adenoviral vectors with the gene encoding for brainderived neurotropic factor (BDNF) to Muller cells of rats with axotomized
retinal ganglion cells (RGC) caused temporary prolongation of the life of
the cells (37).
Recombinant adeno-associated virus (rAAV) slowly integrates at random chromosomal sites over a period of weeks or months, which leads to
long-term photoreceptor transduction, in rodents, of about 1.5 years.
Unlike recombinant adenovirus, rAAV is able to transduce normal photoreceptor as well as retinal pigmental epithelium and retinal ganglion cells. A
drawback to rAAV is that it takes about 35 weeks before full rAAV gene
expression is noticed in rodent photoreceptors. rAAV have been used to
deliver the gene expressing neurturin (a potential neuronal survival factor in
central and peripheral nervous system) in vitro in retinal cells and in vivo for
intraocular delivery in mice (38). rAAV-mediated gene transfer of basic
broblast growth factor (FGF-2) in transgenic rat model of RP has shown
promise as a therapy for retinal degeneration (39). Lentivirus vectors containing the glycoprotein of vesicular stomatitis virus have been successfully
tested in rat retinas. A number of studies have been reported with lentiviral
vectors, mostly in rodents. In view of the potential of lentivirus and rAAV in
retinal gene therapy, more studies are necessary using these vectors.
Herpes virus vectors that have tissue-specic integration with high
expression and large payload of foreign gene can be used to deliver gene
therapy products to nondividing cells in the eye. Liu et al. (40) studied the
delivery of ribonucleotide reductase mutant (hrR3) expressing the E. coli
lacZ gene in primate ocular tissues. Transgene expression was detected in
retinal pigmented epithelium cells and sporadic retinal ganglion cells in
primate eyes receiving virus intracamerally and intravitreally, respectively.
Although replacement of a normal copy of the gene is one of the approaches
of gene therapy, in many conditions that are autosomal dominant, producing a normal copy of the gene would not correct the problem. Rather, a
replication-competent, virulent, ribonuclease reductasedecient herpes
simplex virus can provide the means to deliver therapeutic polypeptides
(like growth factors, neutrophins, and cytokines) in a continuous manner
to the cells (41).
Retinitis pigmentosa refers to a number of related inherited retinal
diseases with an incidence of 1 in 3000. An understanding of the genetic
causes of retinitis pigmentosa could lead to delivery of specic genes in

Copyright 2003 Marcel Dekker, Inc.

618

Das and Miller

RP. A number of retinal diseases in humans aect preferentially one

region of retina, commonly either fovea or peripheral retina, and gradually
progress to entire retina (42). The diseases include retinitis pigmentosa,
gyrate atrophy, choroideremia and rod-cone dystrophy. A recent study
by Bernstein and Wong showed that regional gene expression likely
plays a signicant but nonexclusive role in the development of regional
retinal disease (43). Retinal degeneration in the mouse due to high levels of
visible light has been shown to be dependent on the activation of transcription factor AP-1 leading to apoptotic cell death of photoreceptors
(44). Grimm et al. showed that expression of c-fos and c-jun mRNA is
transiently induced by exposure to damaging light (45). As only the caspase-1 (apoptosis-related genes) expression is upregulated, it could be possible that a gene that downregulates caspase-1 could be helpful in
progression of light-induced degeneration.
A recent report suggests that retroviral vectors can also be used in ex
vivo transfer of -galactosidase gene transduced to human retinal pigment
epithelial cells and transplanted under the rabbit retina. RPE cells at the site
of transplantation formed a monolayer and expressed -galactosidase for 14
days (46). This concept could be applied to gene therapy in choroidal neovascularization, which is responsible for most cases of severe visual loss in
age-related macular degeneration.
As far as the foreseeable future, there is no universal vector that can be
used for gene therapy. More clinical studies are needed to establish the
delivery of gene therapy products.
G.

Nonviral Vectors in Gene Therapy

Although viral vectors have been proven to be very eective in gene therapy, there are a number of drawbacks associated with them. Nonviral gene
delivery systems consist of a therapeutic gene and a synthetic gene delivery
system. The main functions of the synthetic carrier are to protect the
therapeutic gene from premature degradation and deliver the gene to the
nucleus of the target cell. Nonviral approaches to gene therapy have three
key elements: (a) a gene that codes for a specic therapeutic protein, (b) a
gene carrier system that is targeted to a specic tissue or cells, and (c) a
plasmid-based gene expression system that can modulate the duration and
expression levels of the therapeutic protein. The subcellular delivery of
DNA can occur in ve steps (47): membrane binding, internalization,
endosomal release, nuclear localization, and decomplexation (Fig. 1).
The transport of transfected DNA from the cytosol to the nucleus is the
most crucial point in successful gene transfer (48). It was also reported
that nondividing cells are poorly transfectable by nonviral gene transfer

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

619

Figure 1 Subcellular gene delivery using nonviral techniques.

systems, and based on the fact that DNA particles are bigger than 24 nm,
it is postulated that the nuclear uptake occurs preferentially in those cells
that are beginning mitosis consequent to breakdown of the nuclear membrane. Upon entering the cell, the genetic information is transcribed and
translated to a therapeutic protein that can produce a systemic eect, a cell
surface eect, and/or an intracellular eect (Fig. 2). The gene expression
can be modulated so that the therapeutic protein is secreted out of the cell
or stays inside the cell. Other applications of gene delivery involve the
creation of DNA vaccines. The nonviral vectors have gained popularity
mainly because (49):
1. They are nonimmunogenic (this may be a major advantage over
the viral vectors in the eye, where the blood-retinal barrier is
compromised in a number of diseases and during retinal surgery)
(50).
2. They can be designed based on thoroughly characterized chemical agents, and thus their physical interactions with the cells are
reproducible.
3. They can be designed based on the size of the DNA to be transported.
4. Large-scale production of plasmid DNA and transfection agents
is rather inexpensive and they can be produced in large quantity.
5. Testing of synthetic materials is less laborious.

Copyright 2003 Marcel Dekker, Inc.

620

Das and Miller

Figure 2

Eects following gene transfer in cells.

6. Normally the synthetic carriers are degraded in the body without

causing any toxic eect.
At the same time, nonviral vectors are associated with the major disadvantage of lower eciency than viral vectors in gene transfer and their transient
gene expression. Although nonviral gene delivery has aroused general
research interest in recent years, the literature suggests that application of
nonviral gene delivery to ocular tissues is limited. This may be due to the
fact that gene delivery in the ocular tissues needs ecient gene expression
that has not yet been achieved. For example, human retinal pigment epithelial cells have phagocytic activity and can internalize fairly large (2 m)
particles (51). The internalization is probably not the major obstacle, but
the escape of DNA from the endosomes or DNA entry into the nucleus may
be the major rate-limiting step (52). The serum components adhering to the
complexes may cause degradation of the complexes or prevent escape of
DNA from the endosomes (53). In the remainder of this chapter we will
briey overview the techniques of nonviral gene delivery with emphasis on
ocular delivery.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

1.

621

Naked DNA

Researchers became interested in naked or free DNA because DNA itself is

only weakly immunogenic and nonimmunogenic carriers could be delivered
safely. Although there are a number of techniques available for preparation
of puried plasmid DNA, there is no standardized method and most of the
techniques are proprietary. In addition, polymeric DNA is highly negatively
charged and can bind with a variety of agents by electrostatic and hydrophobic forces (54). Endotoxin and protein removal becomes dicult at
times (55). Naked DNA is useful in delivering the gene of interest to muscle
cells (56) directly or to tumors (57), where they induce transgene expression.
A gene for a particular protein is often ligated to a specically designed
plasmid (circular duplex DNA that is replicated extrachromosomally in
bacteria and sometimes in cells of other organisms). About 110 mg of
puried plasmid can be separated from 1 L of bacterial broth. While the
general DNA interaction is dependent on the primary, secondary, or tertiary
structure of DNA, the gene expression is dependent on the primary structure and the sequence of ligation of DNA into the plasmid (58). Vacik et al.
(59) reported an interesting model whereby transcription factors bind to the
DNA in the cytoplasm to create a protein-DNA complex that can enter the
nucleus using the protein import mechanism. Therefore, by using DNA
elements containing binding sites for transcription factors expressed in
unique cell types, it is possible to create plasmids that target the nucleus
in a cell-specic manner (59). A recent study with supercoiled plasmid DNA
injected into the cytoplasm of human corneal cells and keratocytes showed
that primary, nontransformed human corneal epithelial cells and keratocytes display sequence-specic nuclear import of plasmid DNA in the
absence of mitosis. The small sequence that mediates nuclear localization
of plasmids is active in both microinjection and cationic liposome-transfected cells. It was suggested that the inclusion of DNA sequence into
non-viral vectors should improve the eciency of ocular gene transfer in
vivo (60).
Nonviral vector-mediated transfections are classied under two subheadings: techniques suitable for in vitro transfection only and techniques
suitable for both in vitro and in vivo. Techniques for in vitro transfections in
cells involve calcium phosphate transfection, microinjection of plasmid
DNA into the cell nucleus, and electroporation, which involves application
of high voltage to a mixture of DNA and cells, forcing the DNA to move
through the pores formed. These techniques have been discussed elsewhere
and will not be discussed here as there is limited application in in vivo
systems.

Copyright 2003 Marcel Dekker, Inc.

622

Das and Miller

2. Gene Gun
Gene gun transfer of DNA has been successfully used in plant, microbial,
and mammalian cells (61). In addition to the advantage of nonimmunogenicity, it also provides good in vivo gene transfer eciency and focal gene
delivery. The process involves penetration of gold-coated DNA particles
into the target organ using an electric arc generated by high-voltage discharge that accelerates DNA-coated particles to high velocity (62). Plasmid
DNA encoding nontoxic green uorescent protein was delivered in rabbit
cornea, and there was no evidence of corneal or ocular damage (63). It has
also been shown that the ballistic transfer of genes for IL-4 and CTLA-4 in
orthotopic corneal grafting in BALB/c mice caused prolonged gene expression for 5 days (64). In vitro, gene expressing cornea specic keratin 12
protein in T antigen transformed rabbit corneal epithelium (65). Since the
target tissue is to be surgically exposed, this technique can be applied to the
surface of the eye only.
3. Liposomes
Liposomes are spherical particles composed of a lipid bilayer membrane
that encapsulates a part of the solvent. Depending on the nature of the
concentric layers, they are designated SUV (small unilamellar vesicle;
0.020:2 m), LUV (large unilamellar vesicle; 0.10:5 m), and LMV
(large multilamellar vesicle, 0.110 m). The surface characteristics of liposomes can be made neutral, negative, or positively charged depending on the
nature of the ligands on the surface. A number of lipids have been studied
for delivery of gene therapy products, but no single type has been identied
as the most suitable for all types of gene delivery (66). The use of plasmidbased gene expression is limited by their size (about 3000 kDa), hydrophilicity, and a large negative surface charge (30 to 70 mV) (67). These
properties inuence the distribution in the tissues, cellular uptake, and intracellular tracking of gene therapy drugs (68). It is now accepted that cationic lipids are necessary for development of a liposomal formulation for gene
therapy because the negative charge is a hindrance to cell membrane transport (69,70). The nature of the hydrophilic and hydrophobic parts and the
linkers plays an important role, although the structure-function relationships have not been established. On the other hand, cationic lipids (Table
2) cannot form liposomes alone and are normally accompanied by a neutral
lipid or a lipid-like compound, such as DOPE (dioleyl phosphatidylethanolamine) (71) or cholesterol. It has been shown that polycations that bind to
the negatively charged surface of the mammalian cell membranes can cause
charge neutralization, cell distortion, lysis, and agglutination (72). They can
also cause immune system stimulation (73) and can bind to the cell nuclei

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

Table 2

Commonly Used Cationic Lipids

Abbreviated
name
BGTC
DC-Chol
DMRIE
DODAB
DODAC
DOGS
DORI
DOSPA
DOSPER
DOTAP
DOTIM
DOTMA
DPPES
EDMPC
ELMPC
EOMPC
GAP-DLRIE
SAINT-n

623

Chemical name
Bis-guanidinium-tren-cholesterol
3 -[N-(N 0 ,N 0 -Dimethylaminoethane)carbamoyl] cholesterol
1,2-Dimyristoyloxypropyl-3-dimethylhydroxyethylammonium
bromide
Dioctadecyldimethylammonium bromide
Dioleoyldimethylammonium chloride
Dioctadecylamidoglycospermine
1,2-Dioleoyloxypropyl-3-dimethylhydroxyethylammonium
bromide
2,3-Dioleoyloxy-N-[2-sperminecarboxamido)ethyl]-N,N-dimethyl1-propanaminium
1,3-Dioleoyloxy-2 (6-carboxy spermyl) propylamide-4-acetate
1,2-Dioleoyloxy-3-(trimethylammonio)propane
1-[2-(Oleoyloxy)-ethyl]-2-oleoyl-3-(2-hydroxyethyl) imidazolinium
chloride
N-[1-(2,3-Dioleoyloxy)propyl]-N,N,N-trimethylammonium
chloride
Dipalmytoylphosphatidylethanolamidospermine
1,2-Dimyristoyl-sn-glycero-3-ethylphosphocholine chloride
1,2-Dilauroyl-sn-glycero-3-ethylphosphocholine chloride
1,2-Dioleoyl-sn-glycero-3-ethylphosphocholine chloride
()-N-(3-Aminopropyl)-N,N-dimethyl-2,3-bis(dodecyloxy)-1propanaminium bromide
A series of dialkyl pyridiniumalkyl halides

and internal membranes, interfering with the enzyme and cellular function
(74). Synthetic cationic phosphonolipids that are very similar to their natural counterparts may have lesser toxicity towards normal cells (75).
Synthetic polymers together with lipids have been found to be very eective
in transfection compared to lipid-DNA alone. Commercially available lipidbased transfection agents have been reviewed in a recent publication (76).
Polylysine Lipofectamine complexes have been found to be two to three
times more eective in luciferase expression than Lipofectamine alone
(77), with low molecular weight poly(l-lysine) being most eective. Block
copolymer micelles consisting of polyoxyethylene chains, being hydrophilic,
are considered a suitable vehicle for delivery of DNA. The biodistribution of
intravenously administered cationic liposomeplasmid DNA complexes is
not suitable because those complexes exhibiting strong zeta potential are
rapidly cleared from the circulation (78).

Copyright 2003 Marcel Dekker, Inc.

624

Das and Miller

Although liposomal gene delivery has been widely reviewed (79), there
are only a few ocular gene transfer studies using liposomes. Matsuo et al. (80)
demonstrated that instillation of liposomal eye drops of an expression plasmid vector for -galactosidase could transfer the gene to the retinal ganglion
cells of rat without causing any inammation (80). A recent study comparing liposome, calcium phosphate, and DEAE dextranbased delivery of
rey luciferase cDNA to human primary retinal pigment epithelial cells
in vitro shows that calcium phosphate and DEAE dextran techniques failed
to transfect the vector and led to high cytotoxicity, while liposome-based
methods successfully transferred the vectors to the RPE cells. The eciency
varied for dierent liposomes: Tfx-50 > Lipofectin > Lipofectamine >
Cellfectin > DMRIE-C (81). Cytomegalovirus-promoted LacZ genes and
nonhistone nuclear protein, high mobility group 1 (HMG1), were coencapsulated in liposomes, and the liposomes were then coated with the envelope
of inactivated hemagglutinating Sendai virus by fusion (HVJ liposome) for
in vivo transfer and expression of a reporter gene into adult mammalian
retina (82). A large array of literature suggests that there is no universal
delivery system that can be proposed for delivery of DNA-lipid complexes,
while it is a fact that cationic lipids aect the observable DNA expression.
Lipid-DNA complexes are referred to as lipoplex, polymer-DNA complexes as polyplex, and liposome-polymer-DNA complexes as lipopolyplex.
Lipoproteins that form natural biological emulsions have been proposed as carriers for gene transfer. Very-low-density lipoproteins (VLDL),
low-density lipoproteins (LDL), and high-density lipoproteins (HDL) are
now being considered potential delivery vehicles for gene transfer (83).
4. Peptides
A number of membrane-active peptides including bacterial proteins, pHspecic fusogenic, or lytic peptides have been studied for improvement of
polycationic-DNA based transfection. Peptides can act in two ways: (a)
anionic peptides prevent DNA degradation in endosomes by buering the
endosomal compartment; (b) cationic peptides bind to the DNA favoring
endocytosis. A series of natural or synthetic membrane-destabilizing peptides
have been identied as suitable for gene therapy (84). Peptides from viral
sequences such as the N-terminus of inuenza virus hemagglutinin HA-2,
the N-terminus of rhinovirus HRV2 VP-1 protein, and other synthetic and
natural sequences of amphipathic peptides have been found to increase
transfection eciency due to membrane- or vesicle-destabilizing activity
(85). The presence of endoosmolytic peptides signicantly improves the
cytoplasmic delivery. Two such peptides have been tested for their ability

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

625

to deliver the -galactosidase reporter gene to the corneal endothelial cells of

rabbit, pig, and human. Approximately 30% of corneal endothelial cells
were transfected with the integrin-binding moiety of the toxin from the
American pit viper (86).
5.

Polylysine and Polyethylenimine

Poly(l-lysine) and polyethylenimine (PE) have linear and branched structures, respectively. Polylysine-DNA complexes have been found to possess
good transfection ability at a high polycation:DNA ratio that is toxic to
normal cells. The complexes can be delivered eciently to the internal vesicles of the cell, but most often this does not improve the transfection rate,
possibly due to the fact that a membrane in the cell separates the complexes.
Chloroquine or glycerol often improves the transfection by causing lysosomal breakdown (87). PE contains ethyleneimine as the repeating units. A
large number of amino group nitrogens gives it a highly cationic charge.
Approximately 20% of amino nitrogens are protonated under normal physiological conditions, therefore, it has a better buer capacity than polylysine over the entire physiological pH range (88). A possible explanation for
the higher gene transfer eciency of PE is that the buering of endosomes
through PE provokes an enormous amount of proton accumulation followed by passive chloride inux. Such events could cause osmotic swelling
and subsequent endosome disruption and escape of contents of the endosome (89). RPE, which is involved in age-related macular degeneration, is
considered a potential gene therapy target because of its high intrinsic phagocytic function in vivo. It was reported that human RPE cell uptake and
expression of green uorescent protein and neomycin-resistant gene was
signicantly enhanced using polyplex. Although there was a rapid decline
in gene expression over 2 weeks, stable integration occurs at low frequency
(90). Toxicity data on polycations suggest that many polymers used for
transfections are most eective at concentrations that are just subtoxic,
and the cellular uptake of the complexes may be mediated by membrane
destabilization (91). Although polylysine could cause aggregation of erythrocytes at low doses and a wide variety of polycations have been reported
to be cytotoxic (92) or have other adverse eects in culture in vivo, there is
lack of information on all cationic polymers used in gene therapy.
6.

Dendrimers

Dendrimers are fractal polymers. Their formation begins with a core molecule with at least three chemically reactive chains. Further polymerization
occurs at the end of the chains to produce a spherical branched polymer.
They terminate at charged and uncharged amino groups and bind to the

Copyright 2003 Marcel Dekker, Inc.

626

Das and Miller

negatively charged DNA (93). On internalization, the dendrimers buer the

endosomal environment and inhibit pH-dependent lysosomal nucleases. The
eciency of the dendrimers can be increased severalfold by heat treatment
in water or isobutanol, which trims away parts of the molecule and makes it
more exible (94). Dendrimers have been found to possess low cytotoxicity
in vitro and in vivo (95). Polyamidoamine dendrimers are a new class of
nonviral agents that have been used for ex vivo gene transfer to corneal
epithelium. Whole thickness of rabbit or human corneas was transfected
with activated dendrimers and plasmids containing lacZ or TNFR-Ig
genes. The expression was restricted to endothelium for 3 days (96).
Dendrimers could be used as a potential delivery system for gene therapy
in the eye to prevent allograft rejection or in the treatment of other disorders
of corneal endothelium. Recently Urtti et al. (97) studied the delivery of
plasmids encoding -galactosidase (pCMVnLacZ) and luciferase (pCLuc4,
pRSVLuc, pSV2Luc) in human retinal pigment epithelial cells, and they
found that polyamidoamine starburst dendrimers and high molecular
weight polyethylenimine produce better transgene expression than cationic
lipids (DOTAP), despite the low percentage of transfected cells. It was also
shown that dendrimers and polyethylenimine combinations produce less
cellular toxicity in retinal pigment epithelial cell transfection.

7. Formulation Aspects of Nonviral Gene Delivery Products

A number of nonviral agents have been developed in recent years, but a
major limitation is the remarkable instability of the formulations (98). A
recent review by Pogocki and Schoneich discussed the factors aecting the
chemical stability of nucleic acidderived drugs (99). Normally the solutions
of the carrier and the DNA are mixed to form a nonuniform mixture and
administered. However, the transfection rate depends greatly on the physicochemical properties of the DNA-carrier complex (100). An ideal formulation would be a stable and transportable lyophilized powder or a frozen
sample that can be easily reconstituted into a homogenous isotonic solution
for ready administration. It was reported that agitation and freeze-thaw
process in shipping and storage of lipid-DNA complexes could signicantly
reduce the transfection rates, slow freezing being more detrimental than
quick freezing, while some sugars are able to retain transfection capability
and complex size during the freeze-thaw process (101,102). It is believed that
the same concept of preferential exclusion as applicable in frozen protein
formulations is also applicable in DNA-carrier complexes (103).

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

III.

627

OLIGONUCLEOTIDES

Oligonucleotide (ON) therapy is based on the principle of blocking the

synthesis of cellular proteins by interfering with either the transcription of
DNA to mRNA or the translation of mRNA to proteins. The expression of
a specic gene blocked on the translation level by a complementary oligonucleotide using Watson-Crick hybridization to targeted mRNA is called an
antisense oligonucleotide. Among several mechanisms by which antisense
molecules disrupt gene expression and inhibit protein synthesis, the ribonuclease H mechanisms is the most important. Ribonuclease H (RNaseH), an
enzyme present in most cells, recognizes the DNA-RNA or RNA-RNA
duplex and disrupts the base pairing by digesting the RNA part of the
double helix, rendering the digested mRNA incapable of translation.
When the single-stranded complementary DNA binds to portions of
DNA and forms a triple helix and inhibits protein synthesis, it is called
the triplex approach. The functions of dierent oligonucleotides are
depicted in Figure 3. Some researchers believe that the triplex approach is
not a true antisense approach. In August 1998, VitraveneTM (fomivirsen

Figure 3 Functions of oligonucleotides.

Copyright 2003 Marcel Dekker, Inc.

628

Das and Miller

sodium intravitreal injectable, ISIS 2922) became the worlds rst antisense
compound to be approved for marketing in the United States for cytomegalovirus retinitis (CMVR). CMVR occurs in an estimated 47% of AIDS
patients, with the incidence increasing as immunosuppression (decreased
CD4 count) becomes more severe. Fomivirsen sodium is a phosphorothioate
oligonucleotide, 21 nucleotides in length, with the following sequence: 5 0 GCG TTT GCT CTT CTT GCG-3 0 . Studies demonstrate that human retinal pigment epithelial (HRPE) cells were signicantly more sensitive than
broblasts to the antiviral actions of ISIS 2922 and ISIS 12212. The data
indicate that the anti-CMVR potency of the two oligonucleotides was similar. The enhanced potency of these oligonucleotides in HRPE cells may be
associated with a delay in viral gene transcription and slow viral replication
and spread in these cells (104). It was shown that following intravitreal
injection, antisense oligonucleotides preferentially accumulate in the retinal
pigment epithelium and remain present for an extended period of time. It
was also demonstrated that the presence of one antisense oligonucleotide
homologous to rat RPE cathepsin S (Cat 5) induced the accumulation of
photoreceptor-derived debris, which was reversible and temporary (105). A
number of antisense compounds have progressed to the level of Phase 3
clinical trials. Initial stability problems due to the activity of exo- and endonucleases have been largely overcome by chemical modication of the ON
structure. These modications, however, have resulted in a loss of selectivity
and the development of cellular toxicity in some instances. In addition,
antisense agents have been used in the diagnostic approaches to determine
whether the c-myc proto-oncogene is involved in cell proliferation and glycosaminoglycan synthesis in cultured orbital broblasts (106). Table 3 illustrates the characteristics of the major classes of DNA and RNA
oligonucleotides.
Table 3 Oligonucleotides Strategy
Type of
oligonucleotide

Strands

RNA

Single

DNA

Single
Single
Single
Double

Source: Ref. 107.

Copyright 2003 Marcel Dekker, Inc.

Target

Strategy

MRNA
MRNA
Cellular protein
DNA
mRNA
Cellular protein
Transcription factor

Antisense
Ribozyme
Combinatorial
Triple helix
Antisense
Combinatorial
Aptamer

Gene, Oligonucleotide, and Ribozyme Therapy

A.

629

Design of Oligonucleotides and Biological Stability

A number of factors have been determined to contribute to the ecacy of

antisense ON therapeutics, each of which must be carefully examined, but at
present there is no one method or rule for the successful design of an ON.
One primary consideration is the length of the ON species. Lengths of 1725
bases have been shown to be optimal, as longer ONs have the potential to
partially hybridize with nontarget RNA species (108). The G-C content of
the ON is an important factor, as a high percentage of these nucleotides may
result in the formation of secondary or tertiary structure within the oligonucleotide (109), which may aect the target binding.
Biological stability is the major barrier to consider when delivering
both DNA and RNA oligonucleotides to cells. Natural phosphodiester
(PO) oligonucleotides oer the opportunity to selectively intervene in gene
expression; however, they are rapidly degraded in biological uids (110,111)
and cells (112,113) by exo- and endonucleases, which hydrolyze the phosphodiester linkages in natural ONs. Thus, phosphodiester oligonucleotides,
in their native form, are not suitable in vitro, especially in serum containing
culture media, or in vivo following parenteral administration (114,115).
Intracerebroventricular applied phosphodiester ONs have somewhat longer
half-lives in the brain but still undergo extensive degradation.
Protection from nuclease action has been achieved by modication of
phosphate backbones, sugar moiety, and bases (Fig. 4). Substitution of one
of the nonbridging oxygen atoms on the phosphate groups with either a
methyl group or a sulfur atom produces methylphosphonate (116) and
phosphorothioates, respectively. Methylphosphonate oligodeoxynucleotides
(Table 4) are uncharged and enter the animal cells rapidly without distinguishing even a single mismatch (119). Phosphorothioates possess a negatively charged backbone and a sulfur-modied phosphate linkage that is less
sensitive to nucleases (120), but they display signicant nonsequence-specic background inhibition (121,122), which may result mainly from the
interaction with cellular proteins (123). Phosphorodithioate, which is prochiral like phosphodiester, still lacks antisense ecacy and specicity (124)
and may have potential toxicity in CNS applications (125128). Moreover,
methylphosphonate and phosphorothioate derivatives often exert lower
binding anity to their target sequences, possibly due to diastereomer formation (129,130). 2 0 -Methoxy, 2 0 -aminopropoxy, and 2 0 -uororibose are
examples of deoxyribose modications, and inosine and 5-methylcytosine
are examples of base modications. All of these chemical changes have been
shown to convey protection from exo- and endonuclease activity, thereby
increasing the stability and half-life of the antisense compounds. Each modication presents its own type of drawback when compared to phosphodie-

Copyright 2003 Marcel Dekker, Inc.

630

Das and Miller

Figure 4

Nucleotide bases and sugars.

ster ONs, including issues of toxicity, targets selectivity, and cellular uptake.
In spite of several disadvantages, these oligonucleotide modications have
thus far shown to be of utility (131).
In fact, the uptake of ONs in cells occurs in two steps: both charged
and uncharged ONs are initially taken up by the process of endocytosis and
accumulate in an endosomal-lysosomal compartment (132,133). In the second step, the ONs released from the endosomes enter the cytoplasm by an
unknown mechanism and the transfer of ON to nucleus follows rapidly
(134). One signicant advantage of using oligonucleotides over gene therapy
is that the transfer from cytoplasm to nucleus is faster with ONs than with
larger DNA in gene therapy.
B.

Cellular Uptake and Intracellular Fate

Cellular uptake of oligonucleotides has been a challenge, since only about 1

2% of the ONs may be taken up by cells (135). The uptake of ONs in the
cells is dependent on time, concentration, energy, and temperature, indicating it to be an active process (136). As ONs are negatively charged the
potential for diusion across the lipid bilayer is extremely low. The length
and sequence of ONs are determinants of their cellular uptake. Cell types,

Copyright 2003 Marcel Dekker, Inc.

Oligonucleotide

Nuclease stability

Phosphodiester (PO)
Phosphorothioate (PS)
Methylphosphonate (MP)

Poor
Better than PO
Better than PO

Charge

Cellular uptake

Polyanionic
Polyanionic
Neutral

Poor
Worse than PO
Better than PO

Solubility
Good
Better than PO
Worse than PO

Gene, Oligonucleotide, and Ribozyme Therapy

Table 4 Important properties of Major Classes of Oligonucleotides Based on Chemical Structures

631

Copyright 2003 Marcel Dekker, Inc.

632

Das and Miller

media, and culture conditions also aect uptake levels (137). A number of
mechanisms have been identied for internalization of ONs. Receptormediated endocytosis of ON sequences occurs in a number of cell types.
Proteins of 80 and 30 kDa have been identied to mediate ON uptake, while
a simple charge association with the membrane may also be sucient to
trigger endo- or pinocytosis (138,139).
The pharmacological eects of ONs are dependent on the entry of
ONs inside the cells and interaction with pre-mRNA or mRNA. The rate
of permeation of both charged ONs and uncharged methylphosphonate
compounds as well as alkyl-substituted phosphorothioate ONs across the
lipid bilayers was found to be extremely slow (140,141). It is believed that
there is binding of the oligonucleotide to the cell surface protein followed by
internalization into the endocytic compartment (142), mainly by adsorptive
endocytosis. It has been found that the oligonucleotides that adsorb best on
the cell surface have better antisense activity. The release of antisense drugs
from the endosomal compartment is the rate-limiting step for DNA transfection (143) as the antisense drug can be released by exocytosis or may be
partially digested (144).
Zelphati and Szoka (145) proposed that, in a majority of the cell
types, cationic lipids deliver oligonucleotides into the cell predominantly
via an endocytic pathway rather than by fusion with the plasma membrane. The oligonucleotide is released from the complex when anionic
lipids from the cytoplasm facing lipid monolayer of the cell ip into contact with the complex; the anionic lipids then laterally diuse into the
complex and form a charged neutralized ion pair with the cationic lipid,
which leads to displacement of the oligonucleotide from the cationic lipids
and its release into the cytoplasm (146,147) (Fig. 5). Recently, Wu-Pong
discussed transporters involved in the nucleic acid transport and presented
a hypothesis involving the porin-like transports in the internalization of
oligonucleotides (148).
Once inside the cell, ONs can be distributed in various regions of the
cell depending on the chemical modications made to the ON.
Phosphodiester and methylphosphonate ONs are sequestered within the
nucleus in most cell types examined (149). Phosphorothioate derivatives
remain in the cytoplasm in some cell lines but can enter the nucleus in others
(150). Most ONs seem dependent on endocytotic processes, indicating that
these molecules will enter the endosomal/lysosomal pathway. Interaction
with the lysosomes results in the degradation of all types of ONs (151).
Agents that disrupt endosomes can enhance the ecacy of some ONs
(152). Therefore, circumventing the endosomal/lysosomal pathway is one
possible mechanism to enhance the delivery and ecacy of antisense compounds.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

633

Figure 5 Oligonucleotide uptake pathway from cationic lipid complex. First the
cell surfaceassociated complex is internalized via an endosome (step 1). The complex initiates a destabilization of the endosomal membrane that results in a ip-op
of anionic lipids (step 2), which are predominantly located on the cytoplasm face of
the membrane. The anionic lipids laterally diuse into the complex and form a
charged neutralized ion pair with the cationic lipid (step 3). This displaces the oligonucleotide from the complex, and the oligonucleotide can diuse into the cytoplasm (step 4).

C.

Nonantisense Effects of Antisense Oligonucleotides

Although antisense technology has emerged as one of the latest advances in

the development of biotechnology in therapeutics, its success is hampered by
the nonspecic activity of antisense drugs. Based on the estimation of 34

Copyright 2003 Marcel Dekker, Inc.

634

Das and Miller

billion base pairs in the human genome, assuming a random distribution, it

is calculated that a minimum of 1215 bases is required to recognize a single
specic sequence of the genome. Interestingly, it has been pointed out that
increasing the length of oligonucleotides may threaten its specicity, as it
may hybridize to other mRNAs (153). It has been shown that changing the
nature of phosphate linkage and the sugar moiety without changing the
sequence of phosphorothioate and unmodied ONs may not aect the
hybridization to target mRNA, but it may have an eect on binding to
proteins (154). Phosphorothioate ONs (sulfated polyanions) have been
shown to bind with a number of growth factors, including vesicular
endothelial growth factor, and shown to inhibit protein kinase C or
RnaseH. The interactions can be nonspecic, sequence specic, or structure
specic. It was also shown that at a low concentration, these oligonucleotides could block their own uptake by the cells (155). The anticancer drugs
that function as DNA intercalators can bind to certain oligonucleotides
(156). A few motifs within the oligonucleotides could stimulate the immune
response, and it has been suggested to avoid unmethylated CpGs in antisense drugs (157).

D.

Delivery of Antisense Oligodeoxynucleotides

The major challenges in the delivery of antisense ONs are the stability of the
drug in the body and transport through intercellular membranes. Natural
phosphodiester oligonucleotides oer the opportunity to selectively intervene with gene expression, but they are rapidly degraded in biological uids
(158,159) and cells (160,161) by exo- or endonucleases, which hydrolyze the
phosphodiester linkages in natural ONs. Except for methylphosphonate
derivatives, ONs are polyanionic and hydrophilic in nature, which prevents
their passive diusion through the cellular and nuclear membranes (162
164). The extensively investigated phosphorothioate ONs may be more
stable to endonuclease action and show improved transport properties
(165).
One of the primary concerns of drug development is the ability to
provide the drug in an oral dosage form to enhance patient compliance.
Currently antisense ONs are primarily administered parenterally, with IV
administration being the most often used method. Parenteral administration
means that patients will likely be treated by a health care professional at a
health care facility or by a home infusion specialist: this requirement may
result in the cost of treatment becoming prohibitive. Such treatment regimens over the long term often result in a signicant decrease in compliance
as well. The development of delivery vehicles such as liposomes and micro-

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

635

particles may circumvent the problems of administration and distribution

(166,167).
1.

Polycations

As discussed in Sec. II, a number of polycations have been used to deliver

oligonucleotides. Most common is DEAE-D (diethyl aminoethyl dextran),
which complexes with RNA species (168) and enhances interferon induction
(169). Poly(l-lysine) has been used to enhance transport of several drugs,
including oligonucleotides (170,171). The rate and uptake of ONs were
increased as demonstrated using ow cytometric analysis and uorescentlabeled oligonucleotides (172). Highly specic delivery of ONs using positively charged poly(l-lysine) is attributed to its nonspecic interaction with
negatively charged molecules on the cell surface and subsequent internalization by absorptive endocytosis (173). However, the use of poly(l-lysine) is
limited due to its potential cytotoxicity, even at low concentrations (174),
and nonspecic conjugation with cell membrane components. The recently
introduced transfecting agent, polyethylenimine, has been showed to be
eective in studying neuronal ion channel function with antisense oligonucleotides (175).
2.

Lipid-Mediated Delivery

Cationic lipid-mediated transfection of DNA into cells is a well-documented

approach in gene therapy protocols (176,177). A highly ecient DNA transfection technique, lipofection, was introduced by Felgner et al. in 1987
(178). Commercially available LipofectinTM consists of a 1 : 1 weight mixture of positively charged quaternary amino lipid N-[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA) and neutral lipid
dioleylphosphatidylcholine (DOPC). Complexes of oligonucleotides and
the cationic lipid Lipofectin bind to and penetrate cellular membranes,
and ONs are subsequently released into the cytoplasm (179). Zelphati and
Szoka (180) reported that oligonucleotides are delivered by cationic lipids to
the cytoplasm at an early stage of the endocytotic pathway that leads to a
marked increase in antisense activity and oligonucleotide nuclear uptake. In
recent years, several types of cationic lipids have been introduced, including
quaternary ammonium compounds, lipid-conjugated polyamines, and cationic derivative of cholesterol-diacyl glycerol (for discussion and reviews, see
Refs. 181,182).
Typical uptake studies are conducted in serum-free medium (OPTIMEMTM) as Felgner et al. rst observed (183) and conrmed (184) that
inclusion of serum during incubation of cationic lipidnucleotide complex
with cells reduces the transfection eciency. In fact, the presence of only

Copyright 2003 Marcel Dekker, Inc.

636

Das and Miller

20% serum could reduce the cellular binding/uptake of cationic lipidoligonucleotide complex by 81% compared to that attained in serum-free medium. This eect may have been a result of negatively charged serum
components possibly attaching to the complex and preventing cellular interaction (185). However, the serum has little eect on some cationic lipidDNA formulations.
As carriers for the delivery of nucleic acids, liposomes oer a protective biocompatible and biodegradable delivery system that can enhance the
cellular uptake of nucleic acids (186). Liposomes made of cationic lipids
have been widely used to enhance cellular uptake of oligonucleotides.
Jaaskelainen et al. (187) showed that the anionic and cationic charge ratio
of liposomes is critical in fusion and aggregation of liposomes and that
negative charge on the cell membrane is responsible for fusion of the lipid
vesicles to cell membranes and the ability to transfer oligonucleotides (188).
Hartmann et al. (189) reported that cationic lipids that form complexes with
oligonucleotides markedly enhanced the amount of oligonucleotide uptake
in all cells types and were most eective at a 1 : 1 positive-to-negative molar
charge ratio.
pH-sensitive liposomes, consisting of dioleoyl phosphatidyl ethanolamine (DOPE) and a stabilizing acidic amphiphilic lipid, are another
unique vehicle to deliver ONs to the cells. pH-sensitive liposomes bind
to the cell membrane without any fusogenic peptide while the load is
shielded from nuclease action and is delivered from the endocytic compartment before reaching the lysosomes. These liposomes collapse in the endosomes where the negative charges of their lipids become protonated (pKa
5.76.5) (190). Milhaud et al. (191) encapsulated a phosphodiester oligodeoxynucleotide to inhibit vesicular stomatitis virus in murine L929 cells.
The liposomes were stable and remained pH sensitive for several hours;
however, the higher concentrations of lipids caused toxicity.
Immunoliposomes are specic means of delivering nucleic acids to targeted
cells with specic receptors. pH-sensitive immunoliposomes could promote
target-specic delivery of an exogenous gene in plasmid DNA to lymphoma cells (192,193) while reducing the degradation of the nucleic
acids by lysosomes (194). Ma and Wei (195) demonstrated that antiCD32 or anti-CD2 immunoliposomes improved the delivery of 18mer
anti-myb oligo to leukemia cells carrying the appropriate receptor for
the specic antibody-linked immunoliposomes and the uptake was twice
that of the liposomes or nonspecic immunoliposome encapsulated oligos.
A recent study by Brazeau et al. (196) demonstrated that cationic liposomes were less myotoxic than dendrimers and poly-(d,l-lysine).
Myotoxicity was dependent on the type of cationic lipid macromolecule,
concentration, molecular weight, and the presence of plasmid DNA, pos-

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

637

sibly due to the reduction of the net positive charge by the cationic lipidplasmid DNA complex.
Recent reports by Filion and Phillips (197,198) suggest that the cationic liposomes should be used with caution to deliver gene or antisense
oligonucleotides to mammalian cells. Irrespective of the DNA content,
cationic liposomes downregulate the synthesis of the protein kinase C
dependent mediators NO, TNF-
, and PGE2 by activated macrophages
after in vitro incubation under nontoxic conditions or after in vivo treatment. Prolonged incubation (> 3 h) of macrophages with cationic liposomes
induced high toxicity which was not observed with nonphagocytic T cells.
The rank order of toxicity was DOPE/DDAB(dimethyl dioctadecylammoniumbromide) >DOPE/DOTAP>DOPE/Dc-Chol >DOPE/dimyristoyltrimethylammonium propane (DMTAP). The replacement of DOPE by
dipalmitoylphosphatidylcholine reduced the toxicity. On the other hand,
anionic liposomes are safe and can be produced reproducibly. These liposomes are inactive in nature and are typically endocytosed, resulting in
degradation of DNA fragments in lysosomal endosomes (199). The pharmaceutical production of liposomes is still a challenge in view of its short
shelf life, scale-up problems, and absence of sucient safety data for these
carrier systems for chronic use (200).
Although there is growing interest in the ocular delivery of antisense
drugs using liposomes, extensive work has not been done in this area.
Couvreur et al. (201) reported the development and in vitro characterization
of a model oligonucleotide in liposome dispersed in Pluronic 407 (a thermosensitive gel). It was observed that the system could be eective in slow
delivery of antisense drugs to topical or intravitreal tissues. In addition to
charged liposomes, water-soluble block polycations consisting of polyoxyethylene (POE) and polyspermine (PS) have been reported to increase
the sequence-specic inhibition eect of oligonucleotide on the virus reproduction (202). Another interesting study shows that the use of hemagglutinating virus of fusogenic liposomes can transfer LacZ DNA and
phosphorothioate oligonucleotides to adult rat and primate trabecular
meshwork. This system may enable progress in glaucoma research and in
the development of nonviral somatic gene therapy of the trabecular meshwork to treat glaucoma (203).
3.

Biodegradable Microparticles

Recent cell culture and in vivo studies with antisense oligonucleotides have
promoted the need for safe and eective delivery systems for such drugs.
Microparticles and microspheres are monolithic, solid structures, distinguishable from uid and vesicular structures like liposomes. For parenteral

Copyright 2003 Marcel Dekker, Inc.

638

Das and Miller

delivery of bioactive agents, microparticles smaller than 1 m are used. A

number of biodegradable biopolymers (gelatin, albumin, polysaccharides,
etc.) and synthetic polymers [polyalkylcyanoacrylate, poly(lactide-co-glycolide), polycarprolactone, polyanhydride, polyorthoester, etc.] have been
used to formulate microparticles for delivery of a diverse range of bioactive
agents (204208). Biologically active agents could be formulated to be
entrapped in the matrix of the microparticles or simply be adsorbed on
the surface. Release of the therapeutic agent from the microparticles could
be based on diusion of the drug through the matrix of the polymer, surface, and/or bulk erosion of the polymer matrix or desorption of the active
agent from the surface. The use of biodegradable polymeric microparticles
for delivery of oligonucleotides is a viable means to deliver the drug more
eciently and could be useful in promoting cellular uptake through endocytosis, as demonstrated by studies employing cells such as broblasts and
macrophages (209,210).
The major advantage of biodegradable polyester poly(lactide-co-glycolide) (PLGA) polymer for parenteral oligonucleotide delivery systems is
its ability to control the time and/or rate at which the incorporated material
is released. This biodegradable system has been used to successfully deliver a
number of bioactive materials (211214), and the lack of toxicity of poly(lactide-co-glycolide) polymers has been established by FDA and by years of
their use as absorbable suture materials (215,216). PLGA undergoes random, nonenzymatic hydrolysis of its backbone ester linkages in vivo at a
rate that is inuenced by molecular weight, surface area, monomer stereoregularity, and the ratio of lactide to glycolide (217). The homopolymer
itself is degraded very slowly when compared with PLGA and is most
suitable for long-term delivery of bioactive agents (218). In addition to
PLGA nanoparticles, aminoalkylmethacrylate nanoparticles have been
extensively studied for stability and antisense ecacy in vitro (219).
In a recent review article, ODonnell and McGinity (220) comprehensively discussed the pros and cons of preparation of microparticles by the
widely used solvent evaporation technique. An interesting study by Akhtar
et al. (221) showed that ultrasound caused maximum degradation of ONs at
a lower pH, longer-chain-length ONs were more stable than the shorter
ones, and the phosphorothioate ONs were more stable than phosphodiester
during ultrasound exposure. This is an important parameter in the development of microparticle dosage forms as well as delivery of ON by phonophoresis. Akhtar et al. (222) incorporated 32 P-labeled oligonucleotides in
poly(l-lactide) matrices and observed that the degradation of ON due to
nuclease action is substantially reduced. They also observed that ON release
from matrices was dependent on the oligomer chemistry (S-oligos were
released more slowly than the D-oligos) and on the oligomer length (a 20mer oligo was released more slowly than a 7-mer). The same group of work-

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

639

ers (223) reported that cellular association of ONs entrapped within small
microparticles was improved 10-fold in murine macrophages compared with
free ONs. Uptake was enhanced when macrophages were activated with
INF- and lipolysaccharide treatment but decreased signicantly in the
presence of metabolic and phagocytosis inhibitors.
Polyalkylcyanoacrylate microparticles have been shown to be promising in protecting the adsorbed ONs from nuclease action (224). Adsorbed
ONs in such microparticles signicantly increased the ON uptake in the
human macrophagelike cell line U-937 as a result of the capture of microparticles by an endocytic/phagocytic pathway (225). Nakada et al. (226)
showed that ON-adsorbed poly(isobutylcyanoacrylate) microparticles protected the ONs against nuclease action in vivo and deliver them to the liver.
Godard et al. (227) observed that ONs covalently attached to a cholesterol
moiety and adsorbed onto poly(isohexylcyanoacrylate) nanoparticles are
more stable in biological media and better taken up by the eukaryotic
cells. However, the use of poly(isohexylcyanoacrylate) and quaternary
ammonium compounds as adsorbing agents has been criticized because of
the cytotoxicity of poly(hexylcyanoacrylate) and general toxicity of quaternary ammonium compounds (228).
Among the polyester polymers, our previous studies using an 18-base
deoxyribonucleic acid oligonucleotide (5 0 -CTGGTAGCATATGTAAGG3 0 ) of the rat serotonin transporter in the biodegradable poly(lactide-coglycolide) 85 : 15 microparticles retained the stability of oligonucleotide
and allowed a signicant reduction in the rate of uptake of 5-HT without
altering the anity of 5-HT for the transporter in rat basophilic leukemia
cells (229). The interaction of the particulate system with biological components, as well as its distribution and entry to the target cells, essentially
depends on the hydrodynamic size and the particle charge. As discussed
earlier, inherent negative charge of ONs signicantly retards its delivery,
and positively charged components in the matrix of the microparticles
would promote ecient uptake of ONs in the target cells.
The biodegradable microparticles can also be used for delivery of ONs
across the blood-brain barrier. Schroder and Sabel (230), and Kreuter (231)
have discussed the delivery of drug-loaded polyalkylcyanoacrylate microparticles to the brain tissues. An interesting study by Quong et al. (232)
discussed the encapsulation of DNA using alginate microparticles and the
eect of calcium ion.
4.

Physical Approaches

A more signicant and ecient transfer of plasmid DNA, polynucleotides,

and oligonucleotides can be achieved using the technique of electroporation.
This system provides higher eciency of transfer compared to other non-

Copyright 2003 Marcel Dekker, Inc.

640

Das and Miller

viral transfer systems, including cationic liposomes (233). Application of

electroporation to cultured cells is well established, but the use of electroporation in vivo has received little attention (234,235). Electroporation or a
change in the permeability of the cell membrane by electric current has
achieved popularity in recent years due to its high eciency of transfer of
gene or oligonucleotides (236). Recently, Flanagan and Wagner (237)
described ecient inhibition of gene using electroporation of antisense oligodeoxynucleotides. However, electroporation is not a viable technique for
delivery of antisense drugs to ocular tissues.

IV.

APTAMERS

Aptamers are articial nucleic acid ligands, typically composed of RNA,

single-stranded DNA, or a combination of these with nonnatural nucleotides that can be generated against amino acids, drugs, proteins, and other
molecules. They are isolated from complex libraries of synthetic nucleic acid
by an iterative process of adsorption, recovery, and reamplication. They
bind with the target molecules at a very low level with high specicity. One
of the earliest aptamers studied structurally was the 15 mer DNA aptamer
against thrombin, d(GGTTGGTGTGGTTGG) (238). RNA aptamers, or
nucleotides with predominantly RNA-like properties, are capable of adopting a greater variety of structures (239). The structures of protein-binding
aptamers aects the function of aptamers. For example, aptamers that bind
the HIV-1 regulatory protein, Rev, accommodate an alpha helix of their
target within a modied major groove that is widened by the presence of a
number of nonWatson-Crick base pairs (240). In contrast, a class of aptamers that bind the reverse transcriptase of HIV-1 possess a compact form of
pseudoknot structure (241). Aptamers can show a great deal of specicity.
For example,
-thrombin can be dierentiated from -thrombin (242); two
isozymes of protein kinase C diering in just 23 residues could be readily
distinguished (243), as could the CD4 glycoproteins of rat and mouse (73%
identity) (244). For the purpose of diverse application, it is necessary for the
aptamer to be small, typically 10 kDa. Interestingly, aptamers can bind with
small molecules, e.g., Zn2 (245), to bigger ones like ATP (246), neutral
disaccharides (247), dopamine (248), porphyrin (249), and aminoglycoside
antibiotics (250), spanning a size range of 65150 kDa or more.
Natural, 2 0 -OH RNA degrades very rapidly in human serum in vitro,
but this process can be slowed at least 1000-fold by 2 0 -amino modication
(135). In vivo studies on the t1=2 of single-stranded DNA aptamers give
values in the range of less than 2 to approximately 8 minutes (136138).
Conjugation of the aptamer to either lipids or polyethylene glycol has been

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

641

reported to improve the stability and distribution kinetics of DNA aptamers

suciently to produce therapeutic eects (139,140). Although there is no
evidence regarding aptamers in ocular therapy, there has been promising
research on thrombin aptamers (251252) and angiogenin (253) (a DNA
aptamer that prevents angiogenesis and cell proliferation) that indicates
the possible application of aptamers in ocular applications in future.
In a healthy adult, angiogenesis occurs during the menstrual cycle and
wound healing, but it is also a central feature of various disease states,
including cancer and diabetic retinopathy. Recently it has been reported
that an aptamer antagonist of vascular endothelial growth factor (VEGF)
can inhibit blood vessel growth, or angiogenesis, in vivo. Data from in vitro
experiments demonstrated that a liposomal formulation of VEGF aptamer
bound with high anity to VEGF and inhibited the growth of endothelial
cells and cancer cells. In all experiments, the liposomal formulation substantially enhanced the inhibitory activity of the aptamer.

V.

RIBOZYMES

An important development in the use of nucleic acidrelated drugs in ocular

tissues is the use of ribozymes. RNA enzymes or ribozymes are a relatively
new class of single-stranded RNA molecules capable of assuming threedimensional conformations and exhibiting catalytic activity that induces
site-specic cleavage, ligation, and polymerization of nucleotides involving
RNA or DNA (254). They function by binding to the target RNA moiety
through Watson-Crick base pairing and inactivate it by cleaving the phosphodiester backbone at a specic cutting site. Three classes of natural ribozymes have been described based on their unique characters in the sequence
and three-dimensional structure: (a) self-splicing introns of bacteria, mitochondria, (b) RNAase p, required for 5 0 trimming of tRNA precursors, and
(c) self-cleaving viral agents, including the hepatitis delta virus ribozyme.
Because of their small size, specicity for target sequence, and catalytic
activity (can be eective in a low concentration), ribozymes have great
potential for use in inhibiting viral replication or cell proliferation
(255,256). They may catalyze self-cleavage (intramolecular or in-cis catalysis) as well as the cleavage of external substrates (intermolecular or intrans catalysis) (257,258). The term hammerhead ribozyme reects its
structural similarity to a hammerhead. These are the best understood of all
ribozymes. Some ribonucleotides within the sequence selectively form
Watson-Crick base pairs with others to form a stem, while the rest stay in
a single-stranded state called a loop. These loops and stems can be predicted
at the secondary structure level using conformational energy analysis. Most

Copyright 2003 Marcel Dekker, Inc.

642

Das and Miller

ribozymes can be chemically engineered to cleave RNA in trans by separating the catalytic domain from ribozyme and attaching recognition (i.e.,
antisense) arms to the catalytic center in order to target a substrate. These
trans-acting ribozymes can be very useful in the study of molecular biology
and pharmaceutics.
The single-dose safety trial of the antihepatitis C ribozyme LY
466700 has been recently been completed in the rst cohort of normal
volunteers. Administration of LY466700 to chronic hepatitis C patients
has now been initiated in a clinical trial designed to study safety and to
assess the eect of the compound on HCV viral RNA levels following a 28day dose-response regimen.
About 1 in 3000 Americans has retinitis pigmentosa, a degenerative
disease, the symptoms of which generally rst appear during adolescence.
As the disorder progresses, night vision, peripheral vision, and ultimately all
sight can be lost. Currently, there is no way to halt the deterioration.
Autosomal dominated retinitis pigmentosa (ADRP) is caused by mutations
in genes that produce mutated proteins, leading to the apoptotic death of
photoreceptor cells (259). The pioneering work of Lewin and Hauswirth in
the delivery of ribozymes in ADRP in rats shows promise for ribozyme
therapy in many other autosomal dominant eye diseases, including glaucoma. They constructed ribozymes that would bind to the mutant forms
of rhodopsin messenger RNA but not to normal (wild-type) rhodopsin
mRNA. The most potent ribozymes were cloned (as DNA) in recombinant
adeno-associated viral vectors and injected behind the photoreceptor layer
of transgenic rats containing a mutated form of the rhodopsin gene with
defective P23H (proline residue to histidine at 23common form of ADRP
in North America) and S334ter (serine for premature termination at 334a
representative of whole class of ADRP mutations). The ribozyme successfully slowed the rod cell apoptosis and maintained rod cell function for a
minimum of 8 months after a single injection (260). In a recent study the
P347S porcine hammerhead ribozyme cloned in rAAV vector and packaged
into AAV was found to cleave the full length of P347S mRNA (261).
Moreover, ribozymes can discriminate between the mutant and wild-type
sequences of mRNA associated with autosomal dominant retinitis pigmentosa. The kinetics and specicity of ribozyme cleavage (Fig. 6) indicate that
they should reduce the amount of aberrant rhodopsin in the rod cells and
could have potential as therapeutic agents against genetic diseases. The most
active ribozymes against P23H target were rst hammerhead and then hairpin ribozyme (262). The same group also reported that ribozyme rescue
appears to be a potentially eective long-term therapy for autosomal dominant retinal degeneration and is highly eective even when the gene transfer
is done after signicant photoreceptor cell loss (263).

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

643

Figure 6 Hammerhead (A) and hairpin (B) ribozymes used to target P23H opsin
mRNA. The positions of cytosine nucleotide in wild-type rhodopsin are indicated.

There are several advantages of using ribozymes over antisense RNA

products (264):
1. Greater specicity of ribozymesribozymes inhibited viral replication 2- to 10-fold compared to a comparable antisense molecule (265).
2. Ribozymes are short, while antisense oligonucleotides could be
longribozymes are more selective than antisense oligonucleotides. A single base mismatch in a ribozyme could produce a
completely dierent eect, while the same type of mismatch in
an antisense oligonucleotide could have little impact.
Mutation-independent hammerhead ribozymes targeting rhodopsin
and peripherin have been screened in vitro and a number of extremely
ecient ribozymes identied subsequent to detailed kinetic analyses, suggesting that these ribozymes may provide mutation-independent methods
for treating ADRP. The reported ribozymes have potential for use in inherited retinopathies (266). The future of ribozymes in the management of

Copyright 2003 Marcel Dekker, Inc.

644

Das and Miller

wound healing in ocular diseases is still to be investigated. One group has

demonstrated that ribozymes against TGF-1 mRNA in vitro successfully
reduce expression of TGF-1 at both protein and mRNA levels within 3
days of application (267).

VI.

CONCLUSION

A great number of eye diseases are currently dicult to treat with conventional medications due to the lack of target specicity of small molecule
based therapies and the propensity of these compounds to enter the general
circulation. In addition, small molecules cannot restore the function of a
missing gene or protein. As more information about the genetic basis of
ocular disease is revealed, gene therapy and antisense oligonucleotidebased
treatment strategies may aord a mechanism by which to alter or correct
aberrant gene function in a highly selective manner. A great deal of progress
has been made in the engineering of vector systems, but the classic issue of
absorption and distribution, which exist with all drug development programs, remain a signicant hurdle to the use of genetic material in the
treatment of disease. Until eective mechanisms by which to deliver nucleotide-based materials are developed, the progress and utilities of these therapies will be limited. The ability to regain or specically inhibit gene function,
however, is a signicant force in driving the continued development of this
potentially powerful set of therapeutic tools.

REFERENCES
1. S. C. Hyde, D. R. Gill, C. F. Higgins, A. E. Trezise, L. J. MacVinish, A. W.
Cuthbert, R. Ratcli, M. J. Evans, and W. H. Colledge. (1993). Correction of
the ion transport defect in cystic brosis transgenic mice by gene therapy.
Nature, 362:250255.
2. J. A. Roth and R. J. Cristiano. (1997). Gene therapy for cancer. What have we
done and where are we going? J. Natl. Cancer Inst., 89:2139.
3. D. J. Weatherall. (1995). Scope and limitations of gene therapy. Br. Med.
Bull., 51:111.
4. N. Somia and I. M. Verma. (2000). Gene therapy: trials and tribulations. Nat.
Rev. Genet., 1:9199.
5. J. Stephenson. (2001). New therapies show promise for patients with leukemia, hemophilia and heart disease. J. Am. Med. Assoc., 285:153155.
6. T. Sakamoto, H. Kimura, Z. Scuric, and C. Spee et al. (1995). Inhibition of
experimental proliferative vitreoretinopathy by retroviral vector mediated
transfer of suicide gene. Ophthalmology, 102:14171424.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

645

7. A. F. Wright. (1997). Gene therapy for the eye. Br. J. Ophthalmol., 81:620
623.
8. P. V. Algvere, L. Berglin, P. Gouras, and Y. Sheng. (1994). Transplantation of
fetal retinal pigment epithelium in age-related macular degeneration with subfoveal neovascularization. Graefes Arch. Clin. Exp. Ophthalmol., 232:707
716.
9. W. W. Hauswirth and L. Beaufrere. (2000). Ocular gene therapy: Quo Vadis?
Invest. Ophthalmol. Vis. Sci., 41:28212826.
10. W. F. Anderson. (1998). Human gene therapy. Nature, 392:2530.
11. A. Rolland. (1993). Pharmaceutical Particulate Carriers: Therapeutic
Applications. Marcel Dekker, New York.
12. A. D. Miller. (1992). Human gene therapy comes of age. Nature, 357:455460.
13. W. F. Anderson, R. M. Blaese, and K. W. Culver. (1990). The ADA human
gene therapy clinical protocol. Hum. Gene Ther., 1:327362.
14. S. A. Rosenberg, P. Aebersold, K. Conetta, et al. (1990). Gene transfer into
humans: Immunotherapy of patients with advanced melanoma, using tumorinltrating lymphocytes modied by retroviral gene transduction. N. Engl. J.
Med., 323:570578.
15. The Journal of Gene Medicine, John Wiley & Sons, 2001, www.wiley.co.uk/
genmed.
16. G. Romano, C. Pacilio, and A. Giordano. (1999). Gene transfer technology in
therapy: Current applications and future goals. Stem. Cells, 17:191202.
17. J. Bennett, and A. M. Maguire. (2000). Gene therapy for ocular disease. Mol.
Ther., 1:501505.
18. J. M. Con. (1996). Retroviridae: The viruses and their replication. In
Virology, Vol. 2 (B. N. Fields, D. M. Knipe, and P. M. Howley, eds.).
Lippincott, Philadelphia, pp. 17671847.
19. D. G. Miller, M. A. Adam, and A. D. Miller. (1990). Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of
infection. Mol. Cell. Biol., 10:42394242.
20. P. L. Stewart, R. M. Burnett, M. Cyrla, et al. (1991). Image reconstruction
reveals the complex molecular organization of adenovirus. Cell, 67:145154.
21. P. L. Stewart, S. D. Fuller, and R. M. Burnett. (1993). Dierence imaging of
adenovirus: Bridging the resolution gap between X-ray crystallography and
electron microscopy. EMBO J., 12:25892599.
22. H. S. Ginsberg, and G. A. Prince. (1994). The molecular basis of adenovirus
pathogenesis. Infect. Agents Dis., 3:18.
23. T. J. Wickham, P. Mathias, D. A. Cheresh, and G. R. Nemerow. (1993).
Integrins alpha v beta 3 and alpha v beta 5 promote adenovirus internalization
but not virus attachment. Cell, 73:309319.
24. S. L. Brody and R. G. Crystal. (1994). Adenovirus-mediated in vivo gene
transfer. Ann. NY Acad. Sci., 716:90101.
25. M. Ali, N. R. Lemoine, and C. J. Ring. (1994). The use of DNA viruses as
vectors for gene therapy. Gene Ther., 1:367384.
26. S. Yei, N. Mittereder, K. Tang, C. OSullivan, and B. C. Trapnell. (1994).
Adenovirus-mediated gene transfer for cystic brosis: Quantitative evalua-

Copyright 2003 Marcel Dekker, Inc.

646

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Das and Miller

tion of repeated in vivo vector administration to the lung. Gene Ther.,
1:192200.
M. D. Hoggan, N. R. Blacklow, and W. P. Rowe. (1966). Studies of small
DNA viruses found in various adenovirus preparations: Physical, biological
and immunological characteristics. Proc. Natl. Acad. Sci. USA, 55:14671474.
H. Handa, K. Shiroki, and H. Shimojo. (1977). Establishment and characterization of KB cell lines latently infected with adeno-associated virus type 1.
Virology, 82:8492.
K. I. Berns, T. C. Pinkerton, G. F. Thomas, and M. D. Hoggan. (1975).
Detection of adeno-associated virus (AAV) specic nucleotide sequences in
DNA isolated from latently infected Detroit 6 cells. Virology, 68:556560.
G. Podsako, K. K. Wong, Jr., and S. Chatterjee (1994). Ecient gene transfer into nondividing cells by adeno-associated virus-based vectors. J. Virol.,
68:56565666.
R. M. Kotin, M. Siniscalco, R. J. Samulski, X. D. Zhu, L. Hunter, C. A.
Laughlin, S. Mclaughlin, N. Muzyczka, M. Rocchi, and K. I. Berns. (1990).
Site-specic integration by adeno-associated virus. Proc. Natl. Acad. Sci.
USA, 87:22112215.
J. Y. Dong, P. D. Fan, and R. A. Frizzell. (1996). Qualitative analysis of the
packaging capacity of recombinant adeno-associated virus. Hum. Gene Ther.,
7:21012112.
L. M. Homan, A. M. Maguire, and J. Bennett. (1997). Cell-mediated
immune response and stability of intraocular transgene expression after adenovirus-mediated delivery. Invest. Opthal. Vis. Sci., 38:22242233.
K. Tsubota, H. Inoue, K. Ando, M. Ono, K. Yoshino, and I. Saito. (1998).
Adenovirus-mediated gene transfer to the ocular surface epithelium. Exp. Eye
Res., 67:531538.
J. Guy, X. Qi, H. Wang, and W. W. Hauswirth. (1999). Adenoviral gene
therapy with catalase suppresses experimental optic neuritis. Arch.
Opthalmol., 117:15331539.
H. B. Oral, D. F. P. Larkin, Z. Fehervari, A. P. Byrnes, A. M. Rankin, D. O.
Haskard, M. J. A. Wood, M. J. Dallman, and A. J. T. George. (1997). Ex vivo
adenovirus mediated gene transfer and immunomodulatory protein production in human cornea. Gene Therapy, 4:639647.
A. Di Polo, L. J. Aigner, R. J. Dunn, G. M. Bray, and A. J. Aguayo. (1998).
Prolonged delivery of brain-derived neurotrophic factor by adenovirus
infected Muller cells temporarily rescues injured retinal ganglion cells. Proc.
Natl. Acad. Sci. USA, 95:39783983.
C. Jomary, J. Grist, J. Millbrandt, M. Neal, and S. E. Jones. (2001). Epitopetagged recombinant AAV vectors for expressing neurturin and its receptor in
retinal cells. Mol. Vis., 7: 3641.
D. Lau, L. H. McGee, S. Zhou, K. G. Rendahl, W. C. Manning, J. A.
Escobedo, and J. G. Flanner. (2000). Retinal degeneration is slowed in transgenic rats by AAV-mediated delivery of FGF-2. Invest. Ophthalmol. Vis. Sci.,
41:36223633.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

647

40. X. Liu, C. R. Brandt, B. T. Gabelt, P. J. Bryar, M. E. Smith, and P. L.

Kaufman. (1999). Herpes simplex virus mediated gene transfer to primate
ocular tissues. Exp. Eye. Res., 69:385395.
41. C. R. Brandt, R. E. Kalil, and S. Agarwala. (2000). Replication competent, a
virulent herpes simplex virus as a vector for neural and ocular gene therapy.
US Patent 6,106,826, August 22, 2000.
42. E. L. Berson. (1994). Retinitis pigmentosa and allied diseases. In: D. M.
Albert, F. A. Jakobiec, eds. Principles and Practice of Ophthalmology, Vol.
2. Saunders, Philadelphia, pp. 12141237.
43. S. L. Bernstein and P. Wong (1998). Regional expression of disease-related
genes in human and monkey retina. Mol. Vis., 4:24.
44. D. T. Organisciak and B. S. Winkler. (1994). Retinal light damage: Practical
and theoretical considerations. Prog. Retin. Eye Res., 13:129.
45. C. Grimm, A. Wenzel, F. Hafezi, and C. E. Reme. (2000). Gene expression in
the mouse retina: The eect of damaging light. Mol. Vis., 6:252260.
46. T. Murata, J. Cui, K. E. Taba, J.-Y. Oh, C. Spee, D. R. Hinton, and S. J.
Ryan. (2000). The possibility of gene therapy for the treatment of choroidal
neovascularization. Ophthalmology, 107:13641373.
47. R. Niven, J. Smith, and Y. Zhang. (1997). Toward development of a non-viral
gene therapeutics. Adv. Drug Deliv. Rev., 26:135150.
48. H. Pollard, J. S. Remy, G. Loussouarn, S. Demolombe, J. P. Behr, and D.
Escande. (1998). Polyethylenimine but not cationic lipids promotes transgene
delivery to the nucleus in mammalian cells. J. Biol. Chem., 273:75077511.
49. F. D. Ledley. (1995). Nonviral gene therapy: The promise of genes as pharmaceutical products. Hum. Gene Ther., 6:11291144.
50. J. W. Streilein. (1996). Ocular immune privilege and the Faustian dilemma.
Invest. Ophthalmol. Vis. Sci., 37:19401950.
51. H. Kimura, Y. Ogura, T. Moritera, Y. Honda, Y. Tabata, and Y. Ikada.
(1994). In vitro phagocytosis of polylactide microspheres by retinal pigment
epithelial cells and intracellular drug release. Curr. Eye Res., 13:353360.
52. J. Zabner, A. J. Fasbender, T. Moninger, K. A. Poellinger, and M. J. Welsh.
(1995). Cellular and molecular barriers to gene transfer by a cationic lipid. J.
Biol. Chem., 270:1899719007.
53. Y. Xu. and F. C. Szoka. (1996). Mechanism of DNA release from cationic
liposome/DNA complex used in cell transfection. Biochemistry, 35:56165623.
54. J. Flensburg, S., Eriksson, and H. Lindblom. (1988). Purication of supercoiled plasmid DNA by ion exchange chromatography. DNA Protein Eng.
Tech., 1:8590.
55. M. Cotten, A. Baker, M. Saltik, E. Wagner, and M. Buschle. (1994).
Lipopolysaccharide is a frequent contaminant of plasmid DNA preparations
and can be toxic to primary human cells in the presence of adenovirus. Gene
Ther., 1:239246.
56. M. A. Hickman, R. W. Malone, K. Lehmann-Buinsma, T. R. Sih, D. Knoell,
F. C. Szoka, R. Walzem, D. M. Carlson, and J. S. Powell. (1994). Gene

Copyright 2003 Marcel Dekker, Inc.

648

57.
58.
59.
60.
61.
62.

63.

64.

65.

66.
67.
68.

69.
70.

71.

72.

Das and Miller

expression following direct injection of DNA into liver. Human Gene Ther.,
5:14771483.
J. P. Yang and L. Huang. (1996). Direct gene transfer to mouse melanoma by
intratumor injection of free DNA. Gene Ther., 3:542548.
M. Kriegler. (1990). Gene transfer. In: Gene Transfer and Expression: A
Laboratory Manual. W. H. Freeman and Co., New York, pp. 38.
J. Vacik, B. S. Dean, W. E. Zimmer, and D. A. Dean. (1999). Cell-specic
nuclear import of plasmid DNA. Gene Ther., 6:10061014.
D. A. Dean, J. N. Byrd, and B. S. Dean. (1999). Nuclear targeting of plasmid
DNA in human corneal cells. Curr. Eye Res., 19:6675.
S. A. Johnston and D. C. Tang. (1994). Gene gun transfection of animal cells
and genetic immunization. Methods Cell Biol., 43:353365.
W. H. Sun, J. K. Burkholder, J. Sun, J. Culp, X. G. Lu, T. D. Pugh, W. B.
Ershler, and N. S. Yang. (1992). In vivo cytokine gene transfer by gene gun
reduces tumor growth in mice. Proc. Natl. Acad. Sci. USA, 89:1127711281.
D. L. Tanelian, M. A. Barry, S. A. Johnston, T. Le, and G. Smith (1997).
Controlled gene gun delivery and expression of DNA within the cornea. Bio
Techniques, 23:484488.
S. A. Konig Merediz, E. P. Zhang, B. Wittig, and F. Homann. (2000).
Ballistic transfer of minimalistic immunologically dened expression constructs for IL-4 and CTLA4 into the corneal epithelium in mice after orthotopic corneal allograft transplantation. Graefes Arch. Clin. Exp. Ophthalmol.,
238:701707.
A. Shiraishi, R. L., Converse, C. Y. Liu, F. Zhou, C. W. Kao, and W. W.
Kao. (1998). Identication of the cornea-specic keratin 12 promoter by in
vivo particle mediated gene transfer. Invest. Ophthalmol. Vis. Sci., 39:2554
2561.
P. L. Felgner. (1996). Improvements in cationic liposomes for in vivo gene
transfer. Hum. Gene Ther., 7:17911793.
E. Tomlinson and A. Rolland (1996). Controllable gene therapy:
Pharmaceutics of non-viral gene delivery systems. J. Control. Rel., 39:357372.
T. S. Ledley and F. D. Ledley. (1994). Multicompartment, numerical model of
cellular events in the pharmacokinetics of gene therapies. Hum. Gene Ther.,
5:679691.
S. T. Crooke. (1997). Advances in understanding the pharmacological properties of antisense oligonucleotides. Adv. Pharmacol., 40:149.
B. Tavitian, S., Terrazzino, B., Kuhnast, S. Marzabal, O. Stettler, F. Dolle,
J.-R. Deverre, A. Jobert, F Hinnen, B. Bendriem, C. Crouzel, and L. D.
Giamberardino. (1998). In vivo imaging of oligonucleotides with positron
emission tomography. Nat. Med., 4:467471.
H. Farhood, S. Serbina, and L. Huang. (1995). The role of dioleoylphosphatidylethanolamine in cationic liposome mediated gene transfer. Biochem.
Biophys. Acta, 1235:289295.
A. Katchalsky. (1964). Polyelectrolytes and their biological interactions.
Biophys. J., 4:941.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

649

73. H. Moroson. (1971). Polycation-treated tumor cells in vivo and in vitro.

Cancer Res., 31:373380.
74. E. Mayhew and S. J. Nordling. (1966). Electrophoretic mobility of mouse cells
and homologous isolated nuclei. J. Cell Physiol., 68:7580.
75. P. Delpine, C. Guillaume, V. Floch, S. Loisel, J. J. Yaouanc, J. C. Clement, H.
Des Abbayes, and C. Ferec. (2000). Cationic phosphonolipids as nonviral
vectors: in vitro and in vivo applications. J. Pharm. Sci., 89:629638.
76. D. L. Stull. (2000). New tools enable gene delivery: Companies improve existing technologies and oer new ones. Scientist, 14(24):30.
77. L. Vitiello, A. Chonn, J. D. Wasserman, C. Du, and R. G. Worton. (1996).
Condensation of plasmid DNA with polylysine improves liposome-mediated
gene transfer into established and primary muscle cells. Gene Ther., 3:396404.
78. G. Osaka, K. Carey, A. Cuthbertson, P. Godwoski, T. Patapo, A. Ryan, T.
Gadek, and J. Mordenti. (1996). Pharmacokinetics, tissue distribution, and
expression eciency of plasmid [P-33] DNA following intravenous administration of DNA/cationic lipid complexes in mice: Use of a novel radionuclide
approach. J. Pharm. Sci., 85:612618.
79. D. D. Lasic. (1997) Liposomes in Gene Delivery. CRC Press, Boca Raton, FL.
80. T. Matsuo, I. Masuda, T. Yasuda, and N. Matsuo. (1996). Gene transfer to
the retina of rat by liposome eye drops. Biochem. Biophys. Res. Commun.,
219:947950.
81. K. Abul-Hassan, R. Walmsley, and M. Boulton. (2000). Optimization of nonviral gene transfer to human primary retinal pigment epithelial cells. Curr. Eye
Res., 20:361366.
82. M. Hangai, Y. Kaneda, H. Tanihara, and Y. Honda. (1996). In vivo gene
transfer into the retina mediated by a novel liposomes system. Invest.
Ophthalmol. Vis. Sci., 37:26782685.
83. T. Hara, F. Liu, D. Liu, and L. Huang (1997). Emulsion formulations as a
vector for gene delivery in vitro and in vivo. Adv. Drug Deliv. Rev., 24:265
271.
84. C. Plank, W. Zauner, and E. Wagner. (1998). Application of membrane-active
peptides for drug and gene delivery across cellular membranes. Adv. Drug
Deliv. Rev., 34:2135.
85. E. Wagner. (1999). Application of membrane-active peptides for nonviral gene
delivery. Adv. Drug Deliv. Rev., 38:279289.
86. L. Shewring, L. Collins, S. L. Lightman, S. Hart, K. Gustafsson, and J. W.
Fabre. (1997). A nonviral vector system for ecient gene transfer to corneal
endothelial cells via membrane integrins. Transplantation, 64:763769.
87. E. Wagner (1999). Application of membrane-active peptides for nonviral gene
delivery. Adv. Drug Deliv. Rev., 38:279289.
88. M. X. Tang and F. C. Szoka. (1997). The inuence of polymer structure on
the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther., 4:823832.
89. O. Boussif, F. Lezoualch, M. A. Zanta, M. D. Mergny, D. Scherman, B.
Demeneix, and J. Behr. (1995). A versatile vector for gene and oligonucleotide

Copyright 2003 Marcel Dekker, Inc.

650

90.

91.

92.

93.

94.

95.

96.

97.

98.
99.
100.

101.

102.

103.

Das and Miller

transfer into cells in culture and in vivo: Polyethylenimine. Proc. Natl. Acad.
Sci. USA, 92:72977301.
E. Chaum, M. P. Hatton, and G. Stein. (1999). Polyplex-mediated gene transfer into human retinal pigment epithelial cells in vitro. J. Cell Biochemistry,
76:153160.
C. L. Bashford, G. M. Alder, G. Menestrina, K. J. Micklem, J. J. Murphy,
and C. A. Pasternak. (1986). Membrane damage by hemolytic viruses, toxins,
complement and other cytotoxic agents: A common mechanism blocked by
divalent cations. J. Biol. Chem., 261:93009308.
S. Choksakulnimitr, S. Masuda, H. Tokuda, Y. Takakura, and M. Hashida.
(1995). In vitro cytotoxicity of macromolecules in dierent cell culture systems. J. Control. Rel., 34:233241.
M. X. Tang and F. C. Szoka. (1997). The inuence of polymer structure on
the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther., 4:823832.
J. C. Roberts, M. K. Bhagat, and R. T. Zera. (1996). Preliminary biological
evaluation of polyamidoamine (PAMAM) Starburst dendrimers. J. Biomed.
Materials Res., 30:5365.
L. Qin, D. R. Pahud, Y. Ding, A. U. Bielinska, J. F. Kukowska-Latallo, J. R.
Baker, Jr., and J. Bromberg. (1998). Ecient transfer of genes into murine
cardiac grafts by Starburst polyamidoamine dendrimers. Hum. Gene Ther.,
9:553560.
T. Hudde, S. A. Rayner, R. M. Comer, M. Weber, J. D. Isaacs, H.
Waldmann, D. F. P. Larkin, and A. J. George. (1999). Activated polyamidoamine dendrimers, a non-viral vector for gene transfer to the corneal
endothelium. Gene Ther., 6:939943.
A. Urtti, J. Polansky, G. M. Lui, and F. C. Szoka. (2000). Gene delivery and
expression in human retinal pigment epithelial cells: eects of synthetic carriers, serum, extracellular matrix and viral promoters. J. Drug Target., 7:413
421.
M. A. Kay, D. Liu, and P. M. Hoogerbrugge. (1997). Gene therapy. Proc.
Natl. Acad. Sci. USA, 94:1274412746.
D. Pogocki and C. Schoneich. (2000). Chemical stability of nucleic acidderived drugs. J. Pharm. Sci., 89:443456.
I. Jaaskelainen, J. Monkkonen, and A. Urtti. (1994). Oligonucleotide cationic
liposome interactions. A physicochemical study. Biochem. Biophys. Acta,
1195:115123.
T. J. Anchordoquy, L. G. Girouard, J. F. Carpenter, and D. J. Kroll. (1998).
Stability of lipid/DNA complexes during agitation and freeze-thawing. J.
Pharm. Sci., 87:10461051.
S. D. Allison and T. J. Anchordoquy. (2000). Mechanisms of protection of
cationic lipid-DNA complexes during lyophilization. J. Pharm. Sci., 89:682
691.
L. M. Crowe et al. (1993). Does the preferential exclusion hypothesis apply to
hydrated phospholipid bilayers? Cryobiology, 30:224225.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

651

104. B. Detrick, C. N. Nagineni, L. R. Grillone, K. P. Anderson, S. P. Henry, and

J. J. Hooks (2001). Inhibition of human cytomegalovirus replication in a
human retinal epithelial cell model by antisense oligonucleotides. Invest.
Ophthalmol. Vis Sci., 42:163169.
105. P. E. Rakoczy, M. C. Lai, M. Watson, U. Seydel, and I. Constable. (1996).
Targeted delivery of an antisense oligonucleotide in the retina: Uptake, distribution, stability and eect. Antisense Nucleic Acid Drug Dev., 6:207213.
106. A. E. Heufelder and R. S. Bahn. (1995). Modulation of cellular functions in
retroorbital broblasts using antisense oligonucleotides targeting the c-myc
protooncogene. Invest. Ophthalmol. Vis. Sci., 36:14201432.
107. K. K. Jain. (1998). Antisense therapy. In: Textbook of Gene Therapy. Hogrefe
& Huber Publishers, Kirkland, WA, pp. 7399.
108. W. F. Lima, B. P. Monia, D. J. Ecker, and S. M. Freier. (1992). Implication of
RNA structure on antisense oligonucleotide hybridization kinetics.
Biochemistry, 31:1205512061.
109. J. R. Wyatt, T. A. Vickers, J. L. Roberson, R. W. Buckheit, Jr., T. Klimkait,
E. DeBaets, P. W. Davis, B. Rayner, J. L. Imbach, and D. J. Ecker. (1994).
Combinatorially selected guanosine-quartet structure is a potent inhibitor of
human immunodeciency virus envelope-medicated cell fusion. Proc. Natl.
Acad. Sci. USA, 91:13561360.
110. P. S. Eder, R. J. DeVine, J. M. Dagle, and J. A. Walder. (1991). Substrate
specicity and kinetics of degradation of antisense oligonucleotides by a 3 0
exonuclease in plasma. Antisense Res. Dev., 1:141151.
111. J. Goodchild, B. Kim, and P. C. Zamecnik. (1991). The clearance and degradation of oligodeoxynucleotides following intravenous injection into rabbits.
Antisense Res. Dev., 1:153160.
112. A. Teichman-Weinberg, U. Z. Littauer, and I. Ginzburg. (1988). The inhibition of neurite outgrowth in PC12 cells by tubulin antisense oligodeoxynucleotides. Gene, 72:297307.
113. T. Saison-Behmoaras, B. Tocque, I. Rey, M. Chassignol, N. T. Thuong, and
C. Helene. (1991). Short modied antisense oligonucleotides directed against
Ha-ras point mutation induce selective cleavage of the mRNA and inhibit T24
cells proliferation. EMBO J., 10:11111118.
114. Y. Rojanasakul. (1996). Antisense oligonucleotide therapeutics: Drug delivery
and targeting. Adv. Drug Delivery Rev., 18:115131.
115. S. T. Crooke. (1997). Advances in understanding the pharmacological properties of antisense oligonucleotides. Adv. Pharmacol., 40:149.
116. C. H. Agris, K. R. Blake, P. S. Miller, M. P. Reddy, and P. O. Tso. (1986).
Inhibition of vesicular stomatitis virus protein synthesis and infection by
sequence-specic oligodeoxyribonucleoside methylphosphonates. Biochemistry, 25:62686275.
117. F. Eckstein and G. Gish. (1989). Phosphorothioates in molecular biology.
Trends Biochem. Sci., 14:97100.
118. M. Matsukura, K. Shinozuka, G. Zon, H. Mitsuya, M. Reitz, J. S. Cohen,
and S. Broder. (1987). Phosphorothioate analogs of oligodeoxynucleotides:

Copyright 2003 Marcel Dekker, Inc.

652

119.

120.

121.
122.
123.
124.

125.

126.

127.

128.

129.

130.

131.

132.

Das and Miller

inhibitors of replication and cytopathic eects of human immunodeciency
virus. Proc. Natl. Acad. Sci. USA, 84:77067710.
E. H. Chang and P. S. Miller. (1991). Ras, an inner membrane transducer of
growth stimuli. In: Prospects for Antisense Nucleic Acid Therapeutics for
Cancer and AIDS (E. Wickstrom, ed). Wiley-Liss, New York, p. 115.
J. M. Campbell, T. A. Bacon, and E. Wickstrom. (1990).
Oligodeoxynucleoside phosphorothioate stability in subcellular extracts, culture media, sera and cerebrospinal uid. J. Biochem. Biophys. Methods,
20:259267.
R. P. Erickson and J. G. Izant, eds. (1991). Gene Regulation: Biology of
Antisense RNA and DNA. Raven Press, New York.
J. A. H. Murray, ed. (1992) Antisense RNA and DNA. Wiley-Liss, New York.
C. A. Stein (1996). Phosphorothioate antisense oligodeoxynucleotides:
Questions of specicity. Trends Biotechnol., 14:147149.
M. K. Ghosh, K. Ghosh, O. Dahl, and J. S. Cohen. (1993). Evaluation of
some properties of a phosphorodithioate oligodeoxyribonucleotide for antisense application. Nucleic Acids Res., 21:57615766.
P. Yaswen, M. R. Stampfer, K. Ghosh, and J. S. Cohen. (1993). Eects of
sequence of thioated oligonucleotides on cultured human mammary epithelial
cells. Antisense Res. Dev., 3:6777.
M. K. Ghosh, K. Ghosh, and J. S. Cohen. (1993). Phosphorothioate-phosphodiester oligonucleotide co-polymers: Assessment of antisense application.
Anticancer Drug Des., 8:1532.
C. A. Stein and A. M. Krieg. (1994). Problems in interpretation of data
derived from in vitro and in vivo use of antisense oligodeoxynucleotides.
Antisense Res. Dev., 4:6769.
C. Waheslstedt. (1997). Modulation of receptors. Practical approaches to the
regulation on antisense oligonucleotide gene knockout in the nervous system,
March 1619. Oxford University, UK.
R. S. Quartin and J. G. Wetmur. (1989). Eect of ionic strength on the
hybridization of oligodeoxynucleotides with reduced charge due to methylphosphonate linkages to unmodied oligodeoxynucleotides containing complementary sequence. Biochemistry, 28:10401047.
C. A. Stein, K. Mori, S. L. Loke, C. Subasinghe, K. Shinozuka, J. S. Cohen,
and L. M. Neckers. (1988). Phosphorothioate and normal oligodeoxyribonucleotides with 5 0 -linked acridine: Characterization and preliminary kinetics of
cellular uptake. Gene, 72:333341.
R. Zhang, Z. Lu, X. Zhang, H. Zhao, R. B. Diasio, T. Liu, Z. Jiang, and S.
Agrawal. (1995). In vivo stability and disposition of a self-stabilized oligodeoxynucleotide phosphorothiote in rats. Clin. Chem., 41:836843.
G. D. Gray, S. Basu, and E. Wickstrom. (1997). Transformed and immortalized cellular uptake of oligodeoxynucleoside phosphorothioate, 3 0 -alkylamino
oligodeoxynucleotides,
2 0 -O-methyl
oligoribonucleotides,
oligodeoxynucleoside and methylphosphonates, and peptide nucleic acids.
Biochem. Pharmacol., 53:14651476.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

653

133. Y. Shoji, S. Akhtar, A. Periasamy, B. Herman, and R. L. Juliano. (1991).

Mechanism of cellular uptake of modied oligodeoxynucleotides containing
methylphosphonate linkages. Nucleic Acids Res., 19:55435550.
134. T. L. Fisher, T. Terhorst, X. Cao, and R. W. Wagner (1993). Intracellular
disposition and metabolism of uorescently-labeled unmodied oligonucleotides microinjected into mammalian cells. Nucleic Acids Res., 21:38573865.
135. S. Wu-Pong, T. L. Weiss, and C. A. Hunt. (1992). Antisense c-myc oligodeoxyribonucleotide cellular uptake. Pharm. Res., 9:10101017.
136. R. M. Crooke. (1991). In vitro toxicology and pharmacokinetics of antisense
oligonucleotides. Anticancer Drug Des., 6:609646.
137. R. M. Crooke, M. J. Graham, M. E. Cooke, and S. T. Crooke. (1995). In vitro
pharmacokinetics of phosphorothioate antisense oligonucleotides. J.
Pharmacol. Exp. Ther., 275:462473.
138. L. A. Yakubov, E. A. Deeva, V. F. Zarytova, E. M. Ivanova, A. S. Ryte, L. V.
Yurchenko, and V. V. Vlassov. (1989). Mechanism of oligonucleotide uptake
by cells: Involvement of specic receptors? Proc. Natl. Acad. Sci. USA,
86:64546458.
139. R. M. Bennett, G. T. Gabor, and M. J. Merritt. (1985). DNA binding to
human leukocytes. Evidence for a receptor-mediated association, internalization, and degradation of DNA. J. Clin. Invest., 76:21822190.
140. S. Akhtar, S. Basu, E. Wickstrom, and R. L. Juliano. (1991). Interactions of
antisense DNA oligonucleotide analogs with phospholipid membranes (liposomes). Nucl. Acids Res., 19:55515559.
141. J. A. Hughes, C. F. Bennett, P. D. Cook, C. J. Guinosso, C. K. Mirabelli, and
R. L. Juliano. (1994). Lipid membrane permeability of 2 0 -modied derivatives
of phosphorothioate oligonucleotides. J. Pharm. Sci., 83:597600.
142. R. M. Crooke, M. J. Graham, M. E. Cooke, and S. T. Crooke. (1995). In vitro
pharmacokinetics of phosphorothioate antisense oligonucleotides. J.
Pharmacol. Exp. Ther., 275:462473.
143. J. Zabner, A. J. Fasbender, T. Moninger, K. A. Poellinger, and M. J. Welsh.
(1995). Cellular and molecular barriers to gene transfer by a cationic lipid. J.
Biol. Chem., 270: 1899719007.
144. J. L. Tonkinson and C. A. Stein (1994). Patterns of intracellular compartmentalization, tracking and acidication of 5 0 -uorescein labeled phosphodiester and phosphorothioate oilogodeoxynucleotides in HL60 cells. Nucleic
Acids Res., 22:42684275.
145. O. Zelphati and F. C. Szoka, Jr. (1997). Cationic liposomes as an oligonucleotide carrier: Mechanism of action. J. Liposome Res., 7:3149.
146. O. Zelphati and F. C. Szoka. (1996). Mechanism of oligonucleotide release
from cationic liposomes. Proc. Natl. Acad. Sci. USA, 93:1149311498.
147. F. C. Szoka, Y. Xu, and O. Zelphati. (1997). How are nucleic acids released in
cells from lipid-nucleic acid complexes? Adv. Drug Deliv. Rev., 24:291.
148. S. Wu-Pong. (2000). Alternative Interpretations of the oligonucleotide transport literature: Insights from nature. Adv. Drug Deliv. Rev., 44:5970.

Copyright 2003 Marcel Dekker, Inc.

654

Das and Miller

149. D. J. Chin, G. A. Green, G. Zon, F. C. Szoka, Jr., and R. M. Straubinger.

(1990). Rapid nuclear accumulation of injected oligodeoxyribonucleotides.
New Biol., 2:10911100.
150. M. Cerruzzi, K. Draper, and J. Schwartz. (1990). Nucleos. Nucleot., 9:679
695.
151. S. L. Loke, C. A. Stein, X. H. Zhang, K. Mori, M. Nakanishi, C. Subasinghe,
J. S. Cohen, and L. M. Neckers. (1989). Characterization of oligonucleotide
transport into living cells. Proc. Natl. Acad. Sci. USA, 86:34743478.
152. Y. Rojanaskul. (1996). Antisense oligonucleotide therapeutics: drug delivery
and targeting. Adv. Drug Delivery Rev., 18:115131.
153. T. M. Woolf, D. A. Melton, and C. G. B. Jennings. (1992). Specicity of
antisense oligonucleotides in vivo. Proc. Natl. Acad. Sci. USA, 89:73057309.
154. R. C. Bergan, E. Kyle, Y. Connell, and L. Neckers. (1995). Inhibition of
protein-tyrosine kinase activity in intact cells by the aptameric action of oligodeoxynucleotides. Antisense Res. Dev., 5:3338.
155. C. A. Stein, J. L. Tonkinson, L. M. Zhang, L. Yakubov, J. Gervasoni, R.
Traub, and S. A. Rotenberg. (1993). Dynamics of the internalization of phosphodiester oligodeoxynucleotides in HL60 cells. Biochemistry, 32:48554861.
156. R. A. Stull, G. Zon, and F. C. Szoka. (1993). Single-stranded phosphodiester
and phosphorothioate oligonucleotides bind actinomycin D and interfere with
tumor necrosis factor-induced lysis in the L929 cytotoxicity assay. Antisense
Res. Dev., 3:295300.
157. A. M. Krieg, A. K. Yi, S. Matson, T. J. Waldschmidt, G. A. Bishop, R.
Teasdale, G. A. Koretzky, and D. M. Klinman. (1995). CpG motifs in bacterial DNA trigger direct B-cell activation. Nature, 374:546549.
158. P. S. Eder, R. J. DeVine, J. M. Dagle, and J. A. Walder (1991). Substrate
specicity and kinetics of degradation of antisense oligonucleotides by a 3 0
exonuclease in plasma. Antisense Res. Dev., 1:141151.
159. J Goodchild, B. Kim, and P. C. Zamecnik. (1991). The clearance and degradation of oligodeoxynucleotides following intravenous injection into rabbits.
Antisense Res. Dev., 1:153160.
160. A. Teichman-Weinberg, U. Z. Littauer, and I. Ginzburg. (1988). The inhibition of neurite outgrowth in PC12 cells by tubulin antisense oligodeoxyribonucleotides. Gene, 72:297307.
161. T. Saison-Behmoaras, B. Tocque, I. Rey, M. Chassignol, N. T. Thuong, and
C. Helene. (1991). Short modied antisense oligonucleotides directed against
Ha-ras point mutation induce selective cleavage of the mRNA and inhibit T24
cells proliferation. EMBO J., 10:11111118.
162. C. A. Stein and Y. C. Cheng. (1993). Antisense oligonucleotides as therapeutic
agentsis the bullet really magical? Science, 261:10041012.
163. D. M. Tidd. (1990). A potential role for antisense oligonucleotide analogues in
the development of oncogene targeted cancer chemotherapy. Anticancer Res.,
10:11691182.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

655

164. C. A. Stein, and R. Narayanan (1996). Antisense oligodeoxynucleotides:

Internationalization, compartmentalization and nonsequence specicity.
Perspect. Drug Discov. Design, 4:4150.
165. R. M. Crooke. (1991). In vitro toxicity and pharmacokinetics of antisense
oligonucleotides. Anticancer Drug Des., 6:609646.
166. Y. Rojanasakul. (1996). Antisense oligonucleotide therapeutics: drug delivery
and targeting. Adv. Drug Delivery Rev., 18:115131.
167. J. W. Jaroszewski, and J. S. Cohen. (1991). Cellular uptake of antisense oligodeoxynucleotides. Adv. Drug Delivery Rev., 6:235250.
168. D. R. Tovell and J. S. Colter (1969). The interaction of tritium-labeled mengo
virus RNA and L cells: the eects of DMSO and DEAE-dextran. Virology,
37:624631.
169. F. Dianzani, S. Baron, C. E. Buckler, and H. B. Levy. (1971). Mechanism of
DEAE-D-dextran enhancement of polynucleotide induction of interferon.
Proc. Soc. Exp. Biol. Med., 136:11111114.
170. J. P. Leonetti, B. Rayner, M. Lemaitre, C. Gagnor, P. G. Milhaud, J. L.
Imbach, and B. Lebleu. (1988). Antiviral activity of conjugates between
poly(L-lysine) and synthetic oligodeoxyribonucleotides. Gene, 72:323332.
171. M. Stevenson and P. L. Iversen. (1989). Inhibition of human immunodeciency virus type 1-mediated cytopathic eect by poly(L-lysine)-conjugated
synthetic antisense oligodeoxyribonucleotides. J. Gen. Virol, 70:26732682.
172. J. P. Leonetti, G. Degols, and B. Lebleu. (1990). Biological activity of oligonucleotide poly(L-lysine) conjugates: Mechanism of cell uptake. Bioconj.
Chem., 1:149153.
173. H. J. P. Ryser and W. C. Shen. (1978). Conjugation of methotrexate to
poly(L-lysine) increases drug transport and overcomes drug resistance in cultured cells. Proc. Natl. Acad. Sci. USA, 75:38673870.
174. J. P. Leonetti, B. Reyner, M. Lemaitre, C. Gagnor, P. G. Milhaud, J. L.
Imbach, and B. Lebleu. (1988). Antiviral activity of conjugates between
poly(L-lysine) and synthetic oligodeoxyribonucleotides. Gene, 72:323332.
175. R. C. Lambert, Y. Maulet, J. L. Dupont, S. Mykita, P. Craig, S. Volsen, and
A. Feltz. (1996). Polyethylenimine-mediated DNA transfection of peripheral
and central neurons in primary culture: probing Ca2 channel structure and
function with antisense oligonucleotides. Mol. Cell. Neurosci., 7:239246.
176. G. J. Nabel, E. G. Nabel, Z. Y. Yang, B. A. Fox, G. E. Plautz, X. Gao, L.
Huang, S. Shu, D. Gordon, and A. E. Chang. (1993). Direct gene transfer with
DNA-liposome complexes in melanoma: Expression, biological activity, and
lack of toxicity in humans. Proc. Natl. Acad. Sci. USA, 90:1130711311.
177. N. J. Caplen, E. W. Alton, P. G. Middleton, J. R. Dorin, B. J. Stevenson, X.
Gao, S. R. Durham, P. K Jeery, M. E. Hodson, and C. Coutelle. (1995).
Liposome-mediated CFTR gene transfer to the nasal epithelium of patients
with cystic brosis. Nat. Med., 1:3946.
178. P. L. Felgner, Y. R. Gadek, M. Holm, R. Roman, H. W. Chan, M. Wenz, J.
P. Northop, G. M. Ringold, and M. Danielsen. (1987). Lipofection: a highly

Copyright 2003 Marcel Dekker, Inc.

656

179.

180.

181.
182.

183.

184.
185.

186.
187.

188.

189.

190.

191.

192.

Das and Miller

ecient lipid-mediated DNA transfection procedure. Proc. Natl. Acad. Sci.
USA, 84:74137417.
C. F. Bennett, M. Y. Chiang, H. Chan, J. E. Shoemaker, and C. K. Mirabelli.
(1992). Cationic lipids enhance cellular uptake and activity of phosphorothioate antisense oligonucleotides. Mol. Pharmacol. 41:10231033.
O. Zelphati and F. C. Szoka, Jr. (1996). Intracellular distribution and mechanism of delivery of oligonucleotides mediated by cationic lipids. Pharm. Res.,
13:13671372.
P. L. Felgner. (1990). Particulate systems and polymers for in vitro and in vivo
delivery of polynucleotides. Adv. Drug Del. Rev., 5:163187.
H. Farhood, X. Gao, K. Son, Y. Y. Yang, J. S. Lazo, L. Huang, J. Barsoum,
R. Bottega, and R. M. Epand. (1994). Cationic liposomes for direct gene
transfer in therapy of cancer and other diseases. Ann. NY Acad. Sci. USA,
716:2335.
P. L. Felgner, T. R. Gadek, M. Holm, R. Roman, H. W. Chan, M. Wenz, J. P.
Northrop, G. M. Ringold, and M. Danielsen. (1987). Lipofection: A highly
ecient, lipid-mediated DNA transfection procedure. Proc. Natl. Acad. Sci.
USA, 84:74137417.
P. L. Felgner and G. M. Ringold. (1989). Cationic liposome-mediated transfection. Nature, 337:387388.
D. C. Litzinger, J. M. Brown, I. Wala, S. A. Kaufman, G. Y. Van, C. L.
Farrell, and D. Collins. (1996). Fate of cationic liposomes and their complex
with oligonucleotide in vivo. Biochem. Biophys. Acta, 1281:139149.
R. L. Juliano and S. Akhtar. (1992). Liposomes as a drug delivery system for
antisense oligonucleotides. Antisense Res. Dev., 2:165176.
I. Jaaskelainen, J. Monkkonen, and A. Urtti. (1994). Oligonucleotide-cationic
liposome interactions. A physicochemical study. Biochem. Biophys. Acta,
1195:115123.
S. Capaccioli, G. Di Pasquale, E. Mini, T. Mazzei, and A. Quattrone. (1993).
Cationic lipids improve antisense oligonucleotide uptake and prevent degradation in cultured cells and in human serum. Biochem. Biophys. Res.
Commun., 197:818825.
G. Hartmann, A. Krug, M. Bidlingmaier, U. Hacker, A. Eigler, R. Albrecht,
C. J. Strasburger, and S. Endres. (1998). Spontaneous and cationic lipidmediated uptake of antisense oligonucleotides in human monocytes and lymphocytes. J. Pharmacol. Exp. Ther., 285:920928.
C. J. Chu, J. Dijkstra, M. Z. Lai, K. Hong, and F. C. Szoka. (1990). Eciency
of cytoplasmic delivery of pH-sensitive liposomes to cells in culture. Pharm.
Res., 7:824834.
P. G. Milhaud, J. P. Bongartz, B. Lebleu, and J. R. Philippot. (1990). pHsensitive liposomes and antisense oligonucleotide delivery. Drug Delivery,
3:6773.
C. Y. Wang and L. Huang. (1989). Highly ecient DNA delivery mediated by
pH-sensitive immunoliposomes. Biochemistry, 28:95089514.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

657

193. S. Akhtar, S. Basu, E. Wickstrom, and R. L. Juliano. (1991). Interactions of

antisense DNA oligonucleotide analogs with phospholipid membranes (liposomes). Nucl. Acid. Res., 19:55515559.
194. C. Ropert, M. Lavignon, C. Dubernet, P. Couvreur, and C. Malvy. (1992).
Oligonucleotides encapsulated in pH sensitive liposomes are ecient toward
Friend retrovirus. Biochem. Biophys. Res. Commun., 183:879885.
195. D. D. F. Ma and A. Q. Wei. (1996). Enhanced delivery of synthetic oligonucleotides to human leukemic cells by liposomes and immunoliposomes.
Leukemia Res., 20:925930.
196. G. A. Brazeau, S. Attia, S. Poxon, and J. A. Hughes. (1998). In vitro myotoxicity of selected cationic macromolecules used in non-viral gene delivery.
Pharm. Res., 15:680684.
197. M. C. Filion and N. C. Phillips. (1997). Toxicity and immunomodulatory
activity of liposomal vectors formulated with cationic lipids toward immune
eector cells. Biochem. Biophys. Acta, 1329:345356.
198. M. C. Filion and N. C. Phillips. (1998). Major limitations in the use of cationic liposomes for DNA delivery. Int. J. Pharm., 162:159170.
199. Lasic, D. D. (1997). Liposomes in Gene Delivery. CRC Press, New York,
p. 227.
200. G. Strom and D. J. A. Crommelin. (1998). Liposomes: Quo vadis? Pharm. Sci.
Technol. Today, 1:1931.
201. A. Bochot, E. Fattal, A. Gulik, G. Couarraze, and P. Couvreur. (1998).
Liposomes dispersed within a thermosensitive gel: a new dosage form for
ocular delivery of oligonucleotides. Pharm. Res., 15:13641369.
202. A. V. Kabanov, S. V. Vinogradov, Y. G. Suzdaltseva, and V. Yu Alakhov.
(1995). Water-soluble block polycations as carriers for oligonucleotide delivery. Bioconjugate Chem., 6:639643.
203. M. Hangai, H. Tanihara, Y. Honda, and Y. Kaneda. (1998). Introduction of
DNA into the rat and primate trabecular meshwork by fusogenic liposomes.
Invest. Ophthalmol. Vis. Sci., 39:509516.
204. J. Kreuter. (1991). Nanoparticles preparations and application. In:
Microcapsules and Nanoparticles in Medicine and Pharmacy (M. Donbrow,
ed.). CRC Press, London, pp. 125148.
205. J. Kreuter. (1978). Nanoparticles and nanocapsulesnew dosage forms in the
nanometer size range. Pharm. Acta Helv., 53:3339.
206. J. Heller. (1993). Polymers for controlled parenteral delivery of peptides and
proteins. Adv. Drug Del. Rev., 10:163204.
207. W. Lin, A. G. Coombes, M. C. Davies, S. S. Davis, and L. Illum. (1993).
Preparation of sub-100 nm human serum albumin nanospheres using a pHcoacervation method. J. Drug Target., 1:237243.
208. A. Maruyama, T. Ishihara, N. Adachi, and T. Akaike. (1994). Preparation of
nanoparticles bearing high density carbohydrate chains using carbohydratecarrying polymers as emulsier. Biomaterials, 15:10351042.

Copyright 2003 Marcel Dekker, Inc.

658

Das and Miller

209. V. Guise, P. Jaray, J. Delattre, F. Puisieux, M. Adolphe, and P. Couvreur.

(1987). Comparative cell uptake of propidium iodide associated with liposomes or nanoparticles. Cell. Mol. Biol., 33:397405.
210. P. Guiot and P. Couvreur. (1984). Quantitative study of the interaction
between polybutylcyanoacrylate nanoparticles and mouse peritoneal macrophages in culture. J. Pharm. Belg., 38:130134.
211. M. Singh, A. Singh, and G. P. Talwar. (1991). Controlled delivery of
diphtheria toxoid using biodegradable poly(D,L-lactide) microcapsules.
Pharm. Res., 8:958961.
212. A. M. Hazrati, D. H. Lewis, T. J. Atkins, R. C. Stohrer, and L. Meyer. (1992).
In vivo studies of controlled release tetanus vaccine. Proc. Int. Symp. Control.
Rel. Bioact. Mater., 19:114.
213. I. C. Bathurst, P. J. Barr, D. C. Kaslow, D. H. Lewis, T. J. Atkins, and M. E.
Rickey. (1992). Development of a single injection transmission blocking
malaria vaccine using biodegradable microspheres. Proc. Int. Symp. Control.
Rel. Bioact. Mater., 19:120.
214. J. H. Elridge, C. J. Hammond, J. A. Meulbroek, J. K. Staas, R. M. Gilley, and
T. R. Tice. (1990). Controlled vaccine release in the gut associated lymphoid
tissues I. Orally administered biodegradable microspheres target the Peyers
patches. J. Control. Rel., 11:205214.
215. D. K. Gilding and A. M. Reed. (1979). Biodegradable polymers for use in
surgerypolyglycolic/poly(lactic acid) homo- and copolymers. Polymer,
20:14591484.
216. D. L. Wise, T. D. Fellmann, J. E. Sanderson, and R. L. Wentworth. (1979).
In: Drug Carriers in Biology and Medicine. (G. Gregoriadis, ed.). Academic
Press, London.
217. M. Vert, S. M. Li, and H. Garreau. (1994). Attempts to map the structure and
degradation characteristics of aliphatic polyesters derived from lactic and
glycolic acids. J. Biomater. Sci. Polymer Ed., 6:639649.
218. C. G. Pitt and A. Schindler. (1984). Capronor: A biodegradable delivery
system for levonorgestrel. In: Long Acting Contraceptive Delivery Systems
(G. I. Zatuchni, A. Goldsmith, J. D. Shelton, and J. J. Sviarra, eds.).
Harper and Row, Philadelphia, pp. 4863.
219. H. P. Zobel, M. Junghans, V. Maienschein, D. Werner, M. Gilbert, H.
Zimmermann, C. Noe, J. Kreuter, and A. Zimmer. (2000). Enhanced antisense ecacy of oligonucleotides adsorbed to monomethylaminoethylmethacrylate methylmethacrylate copolymer nanoparticles. Eur. J. Pharm.
Biopharm., 49:203210.
220. J. W. McGinity and P. B. ODonnell. (1997). Preparation of microspheres by
the solvent evaporation technique. Adv. Drug Del. Rev., 28:2542.
221. V. M. Meidan, D. Dunnion, W. J. Irwin, and S. Akhtar. (1997). Eect of
ultrasound on the stability of oligodeoxynucleotides in vitro. Int. J. Pharm.,
152:121125.
222. K. J. Lewis, W. J. Irwin, and S. Akhtar. (1995). Biodegradable poly(L-lactic
acid) matrices for the sustained delivery of antisense oligonucleotides. J.
Control. Rel., 37:173183.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

659

223. S. Akhtar and K. J. Lewis. (1997). Antisense oligonucleotide delivery to cultured macrophages is improved by incorporation into sustained release biodegradable polymer microspheres. Int. J. Pharm., 151:5767.
224. C. Chavany, T. Le Doan, P. Couvreur, F. Puisieux, and C. Helene. (1992).
Polyalkylcyanoacrylate nanoparticles as polymeric carriers for antisense oligonucleotides. Pharm. Res., 9:441449.
225. C. Chavany, T. Saison-Behmoaras, T. Le Doan, F. Puisieux, P. Couvreur, and
C. Helene. (1994). Adsorption of oligonucleotides onto polyisohexylcyanoacrylate nanoparticles protects them against nucleases and increases their cellular uptake. Pharm. Res., 11:13701378.
226. Y. Nakada, E. Fattal, M. Foulquier, and P. Couvreur. (1996).
Pharmacokinetics and biodistribution of oligonucleotide adsorbed onto
poly(isobutylcyanoacrylate) nanoparticles after intravenous administration
in mice. Pharm. Res., 13:3843.
227. G. Godard, A. S. Boutorine, E. Saison-Behmoaras, and C. Helene. (1995).
Antisense eects of cholesterol-oligodeoxynucleotide conjugates associated
with poly(alkylcyanoacrylate) nanoparticles. Eur. J. Biochem., 232:404410.
228. I. Aynie, C. Vauthier, E. Fattal, M. Foulquier, and P. Couvreur. (1998).
Alginate nanoparticles as a novel carrier for antisense oligonucleotides. In:
Future Strategies for Drug Delivery with Particulate Systems (J. E. Diederichs
and R. H. Muller, eds.). MedPharm Scientic Publishers, Stuttgart, pp. 1116.
229. S. K. Das, K. J. Miller, and S. C. Chattaraj. (1998). Facilitated delivery of
oligonucleotides as inhibitor of serotonin reuptake. Proc. Inter. Symp.
Control. Rel. Bioact. Mater., 25:350351.
230. U. Schroder and B. A. Sabel. (1996). Dalargin loaded nanoparticles passed the
BBB. Proc. Int. Symp. Bioact. Mater., 23:611612.
231. J. Kreuter. (1996). Nanoparticles as potential drug delivery systems for the
brain. Proc. Int. Symp. Bioact. Mater., 23:8586.
232. D. Quong, R. J. Neufeld, G. Skjak-Braek, and D. Poncelet. (1998). External
versus internal source of calcium during the gelation of alginate beads for
DNA encapsulation. Biotechnol. Bioeng., 57:438446.
233. T. Nishi, K. Yoshizato, S. Yamashiro, H. Takeshima, K. Sato, K. Hamada, I.
Kitamura, T. Yoshimura, H. Saya, J. C. Kuratsu, and Y. Ushio. (1996). High
eciency in vivo gene transfer using intraarterial plasmid DNA injection
following in vivo electroporation. Cancer Res., 56:10501055.
234. L. M. Mir, S. Orlowski, J. J. Belehradek Jr., and C. Paoletti. (1991).
Electrochemotherapy potentiation of antitumor eect of bleomycin by local
electric pulses. Eur. J. Cancer, 27:6872.
235. A. V. Titomirov, S. Sukharev, and E. Kistanova. (1991). In vivo electroporation and stable transformation of skin cells of newborn mice by plasmid
DNA. Biochim. Biophys. Acta, 1088:131134.
236. K. E. Matthews, S. B. Dev, F. Toneguzzo, and A. Keating. (1995).
Electroporation for gene therapy, Methods Mol. Biol., 48:273280.
237. W. M. Flanagan and R. W. Wagner. (1997). Potent and selective gene inhibition using antisense oligodeoxynucleotides. Mol. Cell. Biochem., 172:213225.

Copyright 2003 Marcel Dekker, Inc.

660

Das and Miller

238. L. C. Bock, L. C. Grin, J. A. Latham, E. H. Vermaas, and J. J. Toole.

(1992). Selection of single-stranded DNA molecules that bind and inhibit
human thrombin. Nature, 355:564566.
239. A. R. Ferre-DAmare and J. A. Doudna. (1999). RNA FOLDs: Insights from
recent crystal structures. Annu. Rev. Biophys. Biomol. Struct., 28:5773.
240. X. Ye, A. Gorin, A. D. Ellington, and D. J. Patel. (1996). Deep penetration of
an alpha-helix into a widened RNA major groove in the HIV-1 rev peptideRNA aptamer complex. Nat. Struct. Biol., 3:10261033.
241. D. H. Burke, L. Scates, K. Andrews, and L. Gold. (1996). Bent pseudoknots
and novel RNA inhibitors of type 1 human immunodeciency virus (HIV-1)
reverse transcriptase. J. Mol. Biol., 264:650666.
242. L. R. Paborsky, S. N. McCurdy, L. C. Grin, J. J. Toole, and L. C. Leung.
(1993). The single-stranded DNA aptamer-binding site of human thrombin. J.
Biol. Chem., 268:2080820811.
243. R. Conrad, L. M. Keranen, A. D. Ellington, and A. C. Newton. (1994).
Isozyme-specic inhibition of protein kinase C by RNA aptamers. J. Biol.
Chem., 269:3205132054.
244. E. Kraus, W. James, and A. N. Barclay. (1998). Cutting edge: Novel RNA
ligands able to bind CD4 antigen and inhibition CD4+ T lymphocyte function. J. Immunol., 160:52095212.
245. J. Ciesiolka, J. Gorski, and M. Yarus. (1995). Selection of an RNA domain
that binds Zn2 , RNA, 1:538550.
246. M. Sassanfar and J. W. Szostak. (1993). An RNA motif that binds ATP.
Nature, 364:550553.
247. Q. Yang, I. J. Goldstein, H. Y. Mei, and D. R. Engelke. (1998). DNA ligands
that bind tightly and selectively to cellobiose. Proc. Natl. Acad. Sci. USA,
95:54625467.
248. C. Mannironi, A. DiNardo, P. Fruscoloni, and G. P. Tocchini-Valentini.
(1997). In vitro selection of dopamine RNA ligands. Biochemistry, 36:9726
9734.
249. Y. Li, C. R. Geyer, and D. Sen. (1996). Recognition of anionic porphyrins by
DNA aptamers. Biochemistry, 35:69116922.
250. M. Famulok and A. Huttenhofer. (1996). In vitro selection analysis of neomycin binding RNAs with a mutagenized pool of variants of the 16S rRNA
decoding region. Biochemistry, 35:42654270.
251. W. X. Li, A. V. Kaplan, G. W. Grant, J. J. Toole, and L. L. Leung. (1994). A
novel nucleotide-based thrombin inhibitor inhibits clot-bound thrombin and
reduces arterial platelet thrombus formation. Blood, 83:677682.
252. M. F. Kubik, A. W. Stephens, D. Schneider, R. A. Marlar, and D. Tasset.
(1994). High-anity RNA ligands to human alpha-thrombin. Nucleic Acids
Res., 22:26192626.
253. V. Nobile, N. Russo, G. F. Hu, and J. F. Riordan. (1998). Inhibition of
human angiogenin by DNA aptamers: Nuclear colocalization of an angiogenin-inhibitor complex. Biochemistry, 37:68576863.

Copyright 2003 Marcel Dekker, Inc.

Gene, Oligonucleotide, and Ribozyme Therapy

661

254. T. R. Cech and B. L. Bass. (1986). Biological catalysis by RNA. Annu. Rev.
Biochem., 55:599629.
255. A. Gervaix, L. Schwarz, P. Law, A. D. Ho, D. Looney, T. Lane, and F.
Wong-Staal. (1997). Gene therapy targeting peripheral blood CD34+ hematopoietic stem cells of HIV-infected individuals. Hum. Gene Ther., 8:2229
2238.
256. N. Sarver, E. M. Cantin, P. S. Chang, J. A. Zaia, P. A. Ladne, D. A. Stephens,
and J. J. Rossi. (1990). Ribozymes as potential anti-HIV-1 therapeutic agents.
Science, 247:12221225.
257. H. Kijima, H. Ishida, T. Ohkawa, M. Kashani-Sabet, and K. J. Scanlon.
(1995). Therapeutic applications of ribozymes. Pharmacol Ther., 68:247267.
258. J. Ohkawa, T. Koguma, T. Kohda, and K. Taira. (1995). Ribozymes: From
mechanistic studies to applications in vivo. J. Biochem. (Tokyo), 118:251258.
259. T. P. Dryja and T. Li. (1995). Molecular genetics of retinitis pigmentosa.
Hum. Mol. Genet., 4:17391743, 1995.
260. A. S. Lewin, K. A. Drenser, W. W. Hauswirth, S. Nishikawa, D. Yasumura, J.
G. Flannery, and M. M. LaVail. (1998). Ribozyme rescue of photoreceptor
cells in transgenic rat model of autosomal dominant retinitis pigmentosa. Nat.
Med., 4:967971.
261. L. C. Shaw, A. Skold, F. Wong, R. Petters, W. W. Hauswirth, and A. S.
Lewin. (2001). An allele-specic hammerhead ribozyme gene therapy for a
porcine model of autosomal dominant retinitis pigmentosa. Mol. Vis., 7:613.
262. K. A. Drenser, A. M. Timmers, W. W. Hauswirth, and A. S. Lewin. (1998).
Ribozyme-targeted destruction of RNA associated with autosomal dominant
retinitis pigmentosa. Invest. Ophthalmol. Vis. Sci., 39:681689.
263. M. M. LaVail, D. Yasumura, M. T. Matthes, K. A. Drenser, J. G. Flannery,
A. S. Lewin, and W. W. Hauswirth. (2000). Ribozyme rescue of photoreceptor
cells in P23H transgenic rats: Long-term survival and late-state therapy. Proc.
Natl. Acad. Sci. USA, 97:1148811493.
264. W. W. Hauswirth and A. S. Lewin. (2000). Ribozyme uses in retinal gene
therapy. Prog. Retin. Eye Res., 19:689710.
265. R. Hormes, M. Homann, I. Oelze, P. Marschall, M. Tabler, F. Eckstein, and
G. Sczakiel. (1997). The subcellular localization and length of hammerhead
ribozymes determine ecacy in human cells. Nucleic. Acids Res., 25:769775.
266. B. ONeill, S. Millington-Ward, M. OReilly, G. Tuohy, A. S. Kiang, P. F.
Kenna, P. Humphries, and G. J. Farrar. (2000). Ribozyme-based therapeutic
approaches for autosomal dominant retinitis pigmentosa. Invest. Opth. Vis.
Sci., 41:28632869.
267. X. Ren and G. Schultz. (1999). Reduction of transforming growth factor beta1 protein in cells transfected with plasmids expressing hammerhead and hairpin ribozymes. Invest. Ophthalmol. Vis. Sci., 40(suppl.):46.

Copyright 2003 Marcel Dekker, Inc.

20
Regulatory Considerations
Robert E. Roehrs
and D. Scott Krueger
Alcon Research, Ltd., Fort Worth, Texas, U.S.A.

I.

INTRODUCTION

The usual goal of ophthalmic drug delivery system research is to develop an

improved therapeutic regimen. Some form of performance testing is necessary to determine if the goal has been met, and such testing may involve
federal regulatory considerations. If the drug delivery researcher is only
interested in the in vitro performance of his or her system and/or its in
vivo performance in laboratory research animals for research and publication purposes, federal regulations can largely be ignored. However, if the
delivery system is being developed for testing and use in human and/or
veterinary medicine, a knowledge of the regulations governing animal and
human testing and ultimately the application to market such a pharmaceutical drug product will be essential.
The commercial consideration for development of an ophthalmic drug
delivery system is not limited to new therapeutic agents. Many existing
ophthalmic drugs have inherent limitations due to poor bioavailability or
short duration and are candidates for improved delivery systems. The current U.S. federal regulatory system oers some marketing incentives for
these new dosage form improvements through a period of market exclusivity
prior to generic competition. Obtaining a U.S. patent for the dosage form
improvement can also provide a market extension for the drug and require
competitors to delay market entry or develop a noninfringing improved
dosage form.

Retired

663

Copyright 2003 Marcel Dekker, Inc.

664

Roehrs and Krueger

Drug regulation is not limited to the United States, and most commercial development programs have the objective to obtain approval in the
major foreign markets as well as the United States. While regulatory
requirements vary considerably around the world, there are harmonization
eorts underway in the major countries, particularly between the United
States, the European Union, and Japan, that hopefully will lead one day to a
common marketing application for these countries if not mutually recognized approvals. This chapter will of necessity focus on the regulatory
requirements in the United States.

II.

OVERVIEW OF FEDERAL DRUG LAWS

Federal legislation regulating the importation of adulterated articles dates

back to the Import Drug Act of 1848. The rst signicant federal legislation
regulating the interstate shipment of food and drugs was enacted in 1906
and was known as the Pure Food and Drug Act. It prohibited the interstate
shipment of adulterated or misbranded foods or drugs. In 1912 it was
amended by Congress to include false statements or fraudulent claims as
part of the denition of a misbranded product (1).
A.

Federal Food, Drug & Cosmetic Act

In 1938, the Federal Food, Drug and Cosmetic Act (FD&C Act) was
enacted in response to the elixir of sulfanilamide disaster in which the manufacturer of the rst liquid form of a sulfa drug used diethylene glycol as the
solvent and over 100 deaths were attributed to its poisonous nature (2). The
1906 Act did not require premarket testing for safety and did not allow the
removal of unsafe drugs from the market. The elixir of sulfanilamide did
not contain alcohol and, only because of this technical violation of labeling,
was removed from the market as misbranded. The 1938 Act required drugs
to be tested for safety and to provide this information prior to marketing. It
contains a grandfather clause which exempts certain drugs on the market
at that time, and some of these drugs are still legally marketed under this
old drug provision of the Act.
The FD&C Act as amended is the primary federal law regulating
the interstate shipment of food, drugs, medical devices, and cosmetics
and is enforced by the U.S. Food and Drug Administration (FDA). It
has been amended numerous times to add new regulatory provisions,
and the most pertinent of these amendments are discussed below in
chronological order.

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

B.

665

Kefauver-Harris Amendments

The 1962 amendments required for the rst time that the proof of ecacy as
well as safety be submitted in a New Drug Application (NDA) for marketing approval. They also established the requirements for submission of a
clinical investigational application (IND) to the FDA prior to initiating
research on human subjects. These amendments also established Good
Manufacturing Practice (GMP) regulations (21 CFR 210 & 211).

C.

Environmental Policy Act

The National Environmental Policy Act of 1969, implemented by regulations of the Council on Environmental Quality, requires the FDA and other
federal agencies to assess the possible environmental eects of their actions.
As a result, FDA regulations (21 CFR 25) require that certain applications
to market drug products contain environmental assessments (EA). The
FDA reviews the EA information provided by the applicant as well as
other information available to the agency to determine if the requested
action will signicantly aect the human environment. If there is a nding
of no signicant impact (FONSI), the FDA is required to prepare and
publish the FONSI document. If there is a nding of possible signicant
impact, then a full environmental impact statement (EIS) is required of the
applicant. The Act and the implementing regulations dene certain low-risk
actions as categorical exclusions that do not require the submission of an
EA. The nal revised regulation was published on July 29, 1997 (62 FR
40569). The FDA has published a guidance document on the preparation of
environment assessments (3).

D.

Orphan Drug Act

The Orphan Drug Act of 1983 was enacted to provide incentives for the
research and development leading to market availability of drugs to treat
rare diseases. Only about 10 such products had been marketed in the decade
prior to the Act. The congressionally mandated R&D incentives include
research grants to investigators for the conduct of necessary clinical testing
to obtain FDA approval, tax credits for R&D, and signicant market exclusivity for the applicant who is the rst to obtain marketing approval for the
drug and rare disease. The Act also encourages early availability of orphan
drugs through open protocols, allowing patients to be added to ongoing
studies. Since 1983, more than 200 orphan products have been brought to
the market.

Copyright 2003 Marcel Dekker, Inc.

666

Roehrs and Krueger

An application is required to be submitted to the FDAs Oce of

Orphan Products Development (OOPD) for designation of a drug as an
orphan drug for a rare disease. A rare disease is dened as one where
there are fewer than 200,000 patients in the United States diagnosed with
the disease at the time of the application or one for which the company
developing the product cannot recover the R&D costs necessary to bring the
orphan product to the market. More than one drug can be designated as an
orphan drug for the same rare disease, and more than one applicant can
obtain designation for the same drug and rare disease. The drug must be
designated as an orphan drug prior to submission of the marketing application. A list of orphan drug designations and marketing approvals is published by the FDA monthly as is an annual cumulative update
(www.fda.gov/orphan).
The rst marketing application to obtain approval for the designated
drug and rare disease is awarded 7 years of marketing exclusivity as long as
an adequate supply of the drug to the market is maintained. However, the
nal regulation provides that when a drug, otherwise the same as the
approved orphan, is shown to be clinically superior for the rare disease, it
can also be approved and marketed. Therefore, new modications of the
same active moiety (salts, esters, etc.) or the same drug in an improved
delivery system which provides a signicant therapeutic advantage over
the approved orphan product can be approved prior to expiration of the
original market exclusivity period. FDA nal regulations (21 CFR 316) for
the designation and approval of orphan drugs were published in 1992 (57
FR 62076).
E. Drug Price Competition and Patent Restoration Act
This amendment, also known as the Waxman-Hatch Act, was passed in
1984 to allow marketing of generic equivalents of pioneer NDA drugs
approved since 1962 and thereby increase competition and lower drug
prices. An abbreviated NDA (ANDA) is required to be submitted for
approval and must demonstrate that the generic drug product is the
same as the pioneer NDA product (21 CFR Part 314 Subpart C). The
approval application is abbreviated in that the manufacturer does not have
to repeat the expensive and time-consuming animal safety and human clinical studies but must instead demonstrate that the generic product is bioequivalent to the pioneer drug product. However, the generic applicant is
required to meet the same FDA requirements for chemistry, manufacturing,
and quality control. The Act also modied the patent law such that it is no
longer an infringement to use the patented drug for experimental purposes
related to obtaining U.S. regulatory approval. Thus, the development of a

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

667

generic equivalent can be accomplished at any time prior to patent expiration.

The second part of the Act provides incentives to the pioneer industry
to continue the costly R&D programs for new therapeutic agents by extending, or in eect restoring, a limited portion of the patent term for certain
new drugs. The Act also established market exclusivity periods for new
drugs during which generic applications cannot be approved. These provisions will be discussed in more detail in Sec. VI of this chapter.
The Act also requires the FDA to publish a list of approved drug
products and update the list monthly. The FDA makes this list available
along with additional information in the publication Approved Drug
Products with Therapeutic Equivalence Evaluations, also known as the
Orange Book. The publication is an important information document
for the pharmacist in selecting multisource drug products considered by
the FDA as therapeutically equivalent when state law and the prescriber
allow generic substitution. It also provides the pharmaceutical industry with
therapeutic equivalence requirements as well as information on U.S. patents
that potentially could be infringed by generic applicants and the patent term
and/or market exclusivity period expirations. The Orange Book is also
available electronically via the FDA website at www.fda.gov/cder/ob/
default.htm.
F.

FDA Export Reform and Enhancement Act

Prior to 1986 only drugs approved by the FDA could be legally exported.
This placed the U.S. pharmaceutical industry at a competitive disadvantage
since new drugs may sometimes be rst approved overseas, requiring manufacturing plants to be located outside the United States to meet the need for
drug substances and drug products in these markets prior to FDA approval.
In 1986 Congress, recognizing the desire to retain jobs in the United States
that might otherwise be lost to oshore manufacturing, amended the export
law, allowing unapproved new drugs to be exported to 21 designated countries, under certain conditions, that have premarket approval systems comparable to the United States.
Ten years later, Congress amended the export act with further
enhancements to facilitate new drug exports, including investigational new
drugs for clinical testing overseas. It is now possible to export unapproved
human drugs to any country in the world if the drug complies with the laws
of the importing country, among other requirements, and it has been
approved for marketing in any of the currently designated countries of
Australia, Canada, Israel, Japan, New Zealand, Switzerland, South
African, and countries in the European Union and European Free Trade

Copyright 2003 Marcel Dekker, Inc.

668

Roehrs and Krueger

Association. The exporting company does not require prior FDA approval
but must provide a notication to the FDA. The company must also maintain records of all drugs exported and the countries to which they were
exported. There are additional requirements for good manufacturing practices and labeling.
Additional signicant new export enhancements include the ability to
export unapproved drugs to any of the designated countries to complete
manufacturing, packaging, and/or labeling processes in anticipation of marketing approval. This allows expedited market availability once ocial marketing authorization is obtained. Also, shipment of new drugs to the listed
countries for the purpose of clinical investigations may be made in accordance with the laws and requirements of the importing country, and such
shipments are exempt from U.S. IND regulations. The early phases of
human clinical research are sometimes conducted initially overseas, which
previously required a U.S. IND or other approval to export the clinical
supplies.
G.

Prescription Drug User Fee Act

In 1992 Congress, after consultation with the FDA and the pharmaceutical
industry, amended the FD&C Act to authorize the FDA to collect fees for
the review of certain human drug and biological applications and other
specic agency actions. Congress was reacting to the desire to speed
approval of safe and eective new human drugs and biologicals and the
need for additional resources at the FDA to accomplish this goal. The
prescription drug user fee act (PDUFA) was authorized for a 5-year period,
and during this period the agency was able to reduce the average review time
from 30 months to 15 months, made possible by FDA managerial reforms
and the addition of 700 employees nanced by collection of $329 million in
user fees from the pharmaceutical industry. Based on this success, PDUFA
was reauthorized in 1997 for 5 more years (PDUFA II), and the FDA goal
for review times for most new drug applications was shortened from 12 to 10
months. The one-time user fee for application review is now collected at the
time of submission. The fee is partially refunded if the application is not
accepted for ling and review. If accepted but not found approvable after a
complete review, there is no refund, but the FDA must provide a listing of
all deciencies, which must be overcome for approval.
In addition to one-time fees for review of new human drug applications, user fees are also required on an annual basis for prescription drug
manufacturing facilities and for approved prescription drug products prior
to approval of a generic version. During the rst 5 years of PDUFA, the
approximate average user fee charged was $200,000 for each new drug

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

669

application, $100,000 for each manufacturing facility, and $10,000 for each
product dosage form and strength.
Human drug applications that are exempt from user fees include those
for clinical investigations, generic drug approvals, over-the-counter drug
approvals, orphan drug approvals, and pediatric use approvals. Fees may
be waived in certain specied cases, including small businesses submitting
their rst approval application.
H.

FDA Modernization Act

In 1997, Congress passed major legislation focused on reforming the regulation of food, drugs, devices, and cosmetics. One of the major provisions of
the Act was the reauthorization of PDUFA for 5 years as described above.
A number of the reforms aecting drug products were already FDA and
industry initiatives to modernize and streamline the regulatory process for
approval of new drugs as well as the postapproval requirements for marketed drugs without lowering the standards by which these medical products
are introduced into the marketplace. These include measures to bring more
harmony to the regulation of biological and human drugs, eliminating the
batch certication procedures for insulin and antibiotics, eliminating the
separate regulations for antibiotics and drugs, streamlining the approval
process for biological and drug manufacturing changes, and reducing the
need for environmental assessments as part of product applications. Also,
the practice of allowing, in certain circumstances, one clinical investigation
as the basis for product approval for drugs is now codied. However, the
presumption that, as a general rule, two adequate and well-controlled studies are needed to prove the products safety and eectiveness is preserved in
the regulations.
The act also codied the FDAs regulations and practices to increase
patient access to experimental drugs and medical devices and to accelerate
the review of important new medicines. Additionally, the law provides for
an expanded database on clinical trials accessible by patients, and with
consent of the sponsor, the results of such trials will be included in the
database. Also, patients will receive advance notice when a manufacturer
plans to discontinue a drug on which they depend for life support or sustenance or for treatment of a serious or debilitating disease or condition.

III.

FOOD AND DRUG ADMINISTRATION

The Food and Drug Administration is the federal agency with statutory
authority to regulate the testing and marketing of new ophthalmic delivery

Copyright 2003 Marcel Dekker, Inc.

670

Roehrs and Krueger

systems based on the laws enacted by Congress. The FDA publishes the
proposed and nal regulations in the Federal Register (FR), and the implementing regulations are contained in Title 21 of the Code of Federal
Regulations (CFR). The FR and the CFR documents are available on the
Internet at www.access.gpo.gov.
The FDA is organized into various Centers, which have the primary
responsibility for reviewing clinical trial and marketing applications. There
is a Center for each major product category: Center for Drug Evaluation
and Research (CDER), Center for Biologics Evaluation and Research
(CBER), Center for Devices and Radiological Health (CDRH), Center
for Veterinary Medicine (CVM), and Center for Food Safety and Applied
Nutrition (CFSAN), which includes cosmetics and dietary supplements.
Within each Center are review divisions, usually organized by therapeutic
classes, which are staed by scientists and support sta who review applications and make recommendations for acceptance or rejection to Division
and Center management. Human ophthalmic drug products are reviewed
within the CDER Division, which is staed with ophthalmologists who
review the human clinical data, chemists who review the chemistry, manufacturing, and controls, and pharmacologists who review the animal studies.
Also included, as needed, in the application review team are microbiologists,
statisticians, and biopharmaceutics reviewers.
The FDA maintains an informative website on the Internet at
www.fda.gov, and each Center can be accessed through the site. The
CDER site can be accessed directly at www.fda.gov/cder. The FDA also
maintains a fax-on-demand system for access to guidance and information
documents.

IV.

REGULATORY CLASSIFICATION OF DELIVERY

SYSTEMS

The regulatory requirements for each legally dened class of medical products vary, and so it is important to know how a potential new ophthalmic
delivery system will be classied. Each FDA Center, in addition to statutory
requirements, diers in its rules and procedures for submission and review
of applications.

A.

Drug Versus Device

A drug is legally dened as:

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

671

1. Articles recognized in the ocial United States Pharmacopoeia

(USP), ocial Homeopathic Pharmacopoeia of the United States
or the National Formulary (NF) and their supplements.
2. Articles intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals.
3. Articles other than foods intended to aect the structure or any
function of the body of man or other animals.
4. Articles intended for use as a component of any article specied
in the above three clauses.
A device is dened as an instrument, apparatus, implement, machine,
contrivance, implant, in vitro reagent, or other similar or related article,
including any component, part, or accessory which is:
1. Recognized in the ocial USP or NF or any supplements
2. Intended for use in the diagnosis of disease or other conditions,
or in the cure, mitigation, treatment, or prevention of disease in
man or other animals
3. Intended to aect the structure or any function of the body of
man or other animals, and which does not achieve any of its
principal intended purpose through chemical action within or
on the body of man or other animals and which is not dependent
upon being metabolized for the achievement of any of its principal intended purposes
While there are some similarities in the two denitions, there are also
important dierences. The denition of a device lists specic types of articles
that are covered, and these are the articles that one would typically associate
with the literal denition of a device. A device is also an accessory of one of
these articles. For example, contact lens care products, which have compositions containing chemicals such as disinfectants, lubricant polymers, etc.,
are regulated as devices since they are considered necessary for the safe use
of another device, a contact lens.
Another ophthalmic example is seen with the regulatory history of the
Lacrisert,1 a sterile rod-shaped solid consisting entirely of a cellulosic polymer intended for use in the eye to slowly erode and dissolve in the tear lm
to provide lubrication for painful dry eye conditions. The FDA initially
approved it as a device and then changed its mind and reclassied it as a
drug (4). In doing so, the FDA explained that the term article in the denition of a drug is a broad category in contrast with the specic types of
articles listed in the device denition, and the Lacrisert is not one of those
specic device articles. The FDA also stated that a drug is a chemical or a
combination of chemicals in liquid, paste, powder, or other drug dosage

Copyright 2003 Marcel Dekker, Inc.

672

Roehrs and Krueger

form that is ingested, injected, or instilled into body orices or rubbed or

poured onto the body in order to achieve its intended medical purpose.
Also, note that the legal denition of a drug does not require it to achieve
its principal intended purpose through chemical action or by being metabolized.

B.

New Drug

A new drug is legally dened as one that is not generally recognized among
experts qualied by scientic training and experience as safe and eective for
use under the conditions prescribed, recommended, or suggested in its labeling (FD&C Act Section 201(p)). New drugs require INDs for conducting
clinical investigations and NDAs for marketing approval. The terms drug
and new drug are inclusive of the drug substance and the drug product.
It is important to understand that a new drug is not just a newly
discovered chemical or biological compound. This can best be illustrated
by several examples of when a drug can become a new drug for regulatory
purposes:
1. The drug is a new derivative of a known molecule such as a
prodrug of epinephrine.
2. A previously approved drug has been discovered to have a new
therapeutic use such as a nonsteroidal anti-inammatory agent
used to inhibit miosis during cataract surgery.
3. A component of a drug is new for drug use such as an EVA
polymer lm to control the release of pilocarpine in the eye or
a gel-forming polymer to extend the duration of IOP-lowering of
timolol maleate.
4. Two or more approved drugs are combined for use such as a
xed combination of tobramycin and dexamethasone.
5. A change is made in the route of administration such as a topical
ocular dosage form of acetazolamide for IOP reduction.
6. A change is made in the dosage or strength of an approved drug.
7. A change is made in the intended patient population such as the
use of a drug, approved to lower IOP in glaucoma patients, to be
used in normotensive patients prior to laser surgery to prevent
IOP spikes.
8. The addition or deletion of an inactive component changes the
risk-to-benet ratio for an approved drug.
9. Radiation sterilization is used for a drug product (21 CFR
200.30).

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

C.

673

Combination Drug and Device

It is possible to have as a new ophthalmic delivery system a combination of

a drug and a device. For example, the system could contain the drug in
aerosolized form, which is associated with a novel apparatus (device) to
instill the drug directly onto the surface of the eye in a manner that avoids
the blink response. The FDA product review jurisdiction would be determined by Intercenter agreements and usually assigned based on the primary
mode of action of the product.
In this chapter, we will discuss the requirements for ophthalmic delivery systems in which a drug is incorporated in a carrier for its pharmacological eect on the human eye and is reviewed as a new drug product by the
FDAs Drug Center (CDER).

V.

CLINICAL TESTING OF NEW OPHTHALMIC DRUG

DELIVERY SYSTEMS

A.

Human Versus Animal Testing

Human drug products are often tested during development in animals for
potential acute and chronic signs of toxicity as well as for their primary
and secondary pharmacological eects. If the new drug is shipped interstate for the purpose of clinical investigation in animals, an exemption
similar to a human IND is required, and the label must bear the following
statement (21 CFR 511.1b): Caution. Contains a new animal drug for use
in investigational animals in clinical trials. Not for use in humans. Edible
products of investigational animals are not to be used for food unless
authorization has been granted by the U.S. Food and Drug
Administration or by the U.S. Department of Agriculture. However, if
the interstate shipment is intended solely for use in animals used only for
laboratory research purposes, then it is exempted from the IND requirements if it is labeled as follows (21 CFR 312.160): Caution. Contains a
new drug for investigational use only in laboratory research animals or for
tests in vitro. Not for use in humans. The exemption also requires that
due diligence be used to assure that the consignee is regularly engaged in
conducting such tests and shipment will actually be used as stated in the
Caution. Records of the shipments must be kept for a period of 2 years
after shipment and delivery and made available to an FDA inspector if
requested.

Copyright 2003 Marcel Dekker, Inc.

674

B.

Roehrs and Krueger

Federal Versus State Regulation

Clinical investigations conducted solely intrastate do not escape federal

regulation if the drug or dosage form or any components of these articles
are obtained through interstate shipments.
C.

IND Exemption for Clinical Investigations in Humans

The FD&C Act provides for an exemption from prior approval for interstate shipment of new drugs if the shipment is for the purpose of clinical
testing in humans. The investigational new drug application (IND) is the
notice of exemption that must be submitted prior to the shipment (21 CFR
312). The applicant agrees not to begin clinical use for 30 days or longer if so
notied by the FDA. During the 30-day period, the FDA makes an initial
assessment of the clinical testing plans and the data supporting safe use in
human subjects. If the FDA has serious questions about the application, the
investigations may be put on a clinical hold until the applicant removes the
deciencies. Beyond the initial 30-day review period, the FDA will conduct
a more in-depth review of the data submitted and may from time to time
notify the sponsor of the application regarding deciencies that must be
corrected prior to additional clinical investigations being undertaken. The
sponsor is required to update the application with certain amendments to
ongoing investigation protocols and all new testing protocols. Also, the
sponsor is required to submit an annual progress report of the investigations
and immediate reports of serious and unexpected adverse reactions in
humans and certain serious ndings in animal safety tests.
D.

IND Application

FDA regulations specify the format and content requirements of an IND.

This is shown in Table 1.
1. Clinical Testing
The rst section of the IND informs the FDA as to what drug is to be tested,
the objectives of the clinical experiments, and the scientic rationale and
existing knowledge. The FDA uses this information to determine the adequacy of the technical data to support the proposed human experiments.
Clinical testing usually occurs in several phases, and an IND can be
led for one or more of these phases (Table 2). Typically, an IND would be
led with specic protocol(s) for Phase 1 and a general outline of the plan
for Phase 2. The IND can be amended as necessary to add additional protocols and a revised clinical plan included in the annual report. The clinical

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

675

Table 1

INDFormat and Content Outline

Part 1
Part 2
Part 3

Form FDA 1571

Table of Contents
Introductory Statementbrief introductory statement to include:
a. Name(s) of the drug
Drugs pharmacological class
Drugs structural formula
Dosage form and formulation
Route of administration
Broad objectives and planned duration of the proposed clinical
investigations
b. Brief summary of previous human experience with the drug:
Reference of other IND/NDAs, if pertinent
Investigational or marketing experience in other countries that may
be relevant to the safety of the proposed clinical investigation(s)
c. Identication of countries where drug has been withdrawn from
investigation or marketing for any reasons related to safety or
eectiveness and reasons for withdrawal
General Investigational Planbrief description of the overall plan for
investigating the drug product for the following year to include:
a. Rationale for drug study
b. Indication(s) to be studied
c. General approach to be followed in evaluating the drug
d. Kinds of clinical trials to be conducted in the rst year (indicate if
plans not developed for full year)
e. Estimate number of patients to be given the drug
f. Safety risks anticipatedany risks of particular severity or
seriousness anticipated based on toxicology data or prior human
experience with the drug or related drugs
Investigators Brochureto contain the following information:
a. Brief description of the drug (include structural formula) and the
formulation
b. Summary of the pharmacological and toxicological eects of the
drug in animals and, to the extent known, in man
c. Summary of pharmacokinetics and biological disposition of the drug
in animals and, if known, in man
d. Summary of information relating to safety and eectiveness in
humans from prior clinical studies (can append reprints when
pertinent and useful)
e. Description of possible risks and side eects to be anticipated based
on experience with the drug or with related drugs. Precautions or
special monitoring to be done as part of the investigations
Clinical Investigation
a. Protocol for each planned study

Part 4

Part 5

Part 6

Copyright 2003 Marcel Dekker, Inc.

676

Roehrs and Krueger

Table 1 Continued

Part 7

Part 8

b. Investigators
Name, address and CV of each investigator
Name of each subinvestigator
Name and address of research facilities
Name and address of IRB
c. Monitorname, title, and CV
(The person responsible for monitoring the conduct and progress of
the clinical investigations)
Safety Monitor(s)name, title and CV
(The person or persons responsible for review and evaluation of
information relevant to safety of the drug)
d. Contract Research Organizations (CRO)
i. Name and address of CRO used for any part of the clinical
studies
ii. Identify the studies and CRO monitor
iii. List sponsor obligations transferred to CRO, if any
e. Labeling for clinical supplies
Chemistry, Manufacturing, and Controls Information
a. Drug Substance
1. Description of physical and chemical characteristics
2. Name and address of manufacturer
3. Method of preparation
4. Reference standard
5. Specications
6. Methods of analysis
7. Stability
b. Drug Product
1. Components (reasonable alternatives)
i. Inactive componentstests and specications
2. Composition (reasonable variations)
3. Name and address of manufacturer
4. Manufacturing and packaging procedure
5. Specications
6. Methods of analysis
7. Packaging
8. Stability
9. Labeling for clinical supplies
10. Placebocomposition, manufacture, and control
11. Environmental analysisClaim for categorical exclusion
Pharmacology and Toxicology
a. Pharmacology and drug disposition
1. Section describing the pharmacological eects and mechanism(s)
of action of the drug in animals
2. Section describing the ADME of the drug, if known

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

Part 9

Part 10

677

b. Toxicology
1. ID and qualications of persons conducting and evaluating
results of studies concluding reasonably safe to begin purposed
investigations
2. Statement where studies conducted and where records available
for inspection
3. Integrated summary of the toxicological eects of the drug in
animals and in vitro
4. Detailed tox study reports with full tabulations of data for each
study primarily intended to support the safety of the proposed
clinical investigation
5. GLP Compliance Statement(s)
Previous Human Experience
Summary of known prior human experience with the investigational
drug to include:
a. If previously investigated or marketed (anywhere):
i. Detailed information about such experience relevant to safety of
proposed investigation or rationale
ii. If drug has been subject of controlled clinical trials, detailed
information on such trials relevant to an assessment of the
drugs eectiveness for the proposed investigational use
iii. Published material directly relevant to safety or eectiveness for
the proposed investigational useprovide full copies
Published material less directly relevantbibliography
b. For combination of drugsPart 9a information for each drug
c. Foreign marketing
i. List of countries where marketed
ii. List of countries where drug has been withdrawn from marketing
for reasons potentially related to safety or eectiveness
Pediatric Studiesplans for assessing pediatric safety and eectiveness

Source: Adapted from 21 CFR 312.23

protocol is a critical element of the investigational phase for a new drug.

FDA regulations establish requirements for the protocol (Table 3). The
FDA medical ocer reviewing the protocol will provide a critique as to
its scientic and regulatory acceptability as well as the acceptability of the
risk to human subjects. If the study is intended to be used as part of the
NDA to establish safety and eectiveness, it would be important to know if
the FDA has any serious questions about the protocol before proceeding.
Ophthalmic drug clinical development generally follows three phases,
particularly if the drug is a new molecule and this is its rst introduction into
humans. Phase 1 for ophthalmic drugs usually is focused on the potential for

Copyright 2003 Marcel Dekker, Inc.

678

Roehrs and Krueger

Table 2 Clinical Testing Phases

Phase 1: Initial introduction into manclosely monitored patients or normal
volunteers. Primarily for side eects with increasing doses, ADME and clinical
pharmacology and early readout on eectiveness. Used to design well-controlled,
scientically valid Phase 2 studies. Also used for structure-activity and mechanism
of action studies as well as using drugs as research tool to explore biology and
disease processes. Usually involve fewer than 100 subjects.
Phase 2: Controlled studies for eectiveness in patients and determination of
common short-term side eects and risks. Well-controlled and closely monitored
studies in usually no more than several hundred patients. Used to screen out
drugs with limited potential for safety and/or eectiveness. Data critical to
design Phase 3, particularly for dosage regimens and patient populations.
Phase 3: Expanded studies in patients both controlled and uncontrolled to
gather data on safety and eectiveness needed to evaluate overall benet-risk and
provide basis for labeling. Usually involve several hundred to several thousand
subjects.
Phase 4: Studies usually conducted after initial marketing. May be requested by
FDA as condition of approval. May be used to examine specic patient
subpopulations such as geriatric or pediatric or to increase exposure to further
dene benet-to-risk ratio.

ocular toxicity in healthy subjects. Because of the potential for systemic

absorption for ocular dosing, attention is also given to the possibility of
systemic eects, and concurrently, or in separate studies, the extent of systemic absorption is determined.
In Phase 2 of clinical testing, the drug is usually rst introduced into
patients with the disease and a dose-response relationship is investigated.

Table 3 Clinical Protocol Requirements

Objectives and purpose of study
Identication of each investigator and subinvestigator, research facilities, and
each IRB
Criteria for patient selection and exclusion and estimated number of patients
Study design including group (if any) and methods to minimize bias on part of
subjects, investigators, and analysts
Method for determining doses to be used, the maximum dose and duration of
patient exposure
Description of the observations and measurements to fulll objectives of study
Description of clinical procedures, lab tests, or other measures to monitor
drug eects in subjects and to minimize risk

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

679

For oral drugs, the dose-response testing is a crucial parameter; however, it

has not been a rigorous part of most topical ophthalmic drug-development
programs. The drug delivery researcher may nd that these data are missing
for his or her drug and needs to establish this relationship for optimization.
An example of this occurred during the development of the Ocusert containing pilocarpine in which patients were given multiple micro doses of pilocarpine topically to establish the required release rate to provide the desired
IOP-lowering response (5).
The nal Phase 3 testing is essential for providing the substantial
evidence from adequate and well-controlled studies required for proof of
safety and eectiveness. It is particularly important that the endpoints used
to measure the response of the delivery system be clinically relevant and that
enough patients are included to detect a signicant dierence if one exists.
The FDA has established standards for the conduct of clinical studies
in order to ensure the quality and integrity of the data on which the safety
and eectiveness decisions will be based and also, importantly, the protection of the rights and health of the participating subjects. These are commonly referred to as Good Clinical Practices (GCPs) but are not embodied
in one regulation. They are a combination of the Informed Consent regulation for clinical subjects (21 CFR 50), the Institutional Review Board (IRB)
regulations (21 CFR 56), and the obligations of sponsors and investigators
dened in the 1987 IND Rewrite regulations (21 CFR 312).
2.

Preclincal Testing

The nonclinical or preclinical section addresses the biological data that

support the pharmacological rationale for the intended use of the drug,
the animal toxicology data to assess the safety risks for human exposure,
and, if available, systemic and ocular pharmacokinetic data. These data may
necessarily be limited at this point, particularly if the drug is a new molecule.
If the delivery system is being developed with a known drug, then comparative tests will be useful to assess the risk. If the delivery system contains a
new component, such as a polymer or surfactant to prolong ocular residence
and/or enhance bioavailability, then additional safety testing may be
required for the new component if the supplier has not already provided
this information. In some cases this component may have already been used
in other drug applications or sometimes for food or cosmetic uses and have
established a generally recognized as safe (GRAS) status. The toxicologist
will have to assess the relevance of these data to the intended topical application.
FDA does not have specic requirements or guidelines to answer the
often asked question: How much animal safety data do I need for an

Copyright 2003 Marcel Dekker, Inc.

680

Roehrs and Krueger

ophthalmic IND? In general, to begin clinical testing, FDA will require at

least the same duration of testing in animals as proposed for human exposure. The requirements will vary with the particular drug and the novelty of
the particular delivery system. The approach of one company in establishing
the safety/toxicity prole of ophthalmic drugs and devices has recently been
published (6).
FDA implemented Good Laboratory Practice (GLP) regulations in
1976, which established standards for the conduct and reporting of all animal safety-related studies to be used in support of an IND or NDA. This
was a reaction to the discovery that some industrial and contract toxicology
testing labs were conducting studies in a sloppy manner and in instances had
falsied data that were submitted to FDA. FDA now routinely audits on a
periodic basis all labs conducting such studies. Therefore, if animal safety
data are generated in an academic institution and the results are to be used
in support of an IND, the studies must meet GLP regulations and any
deviations from these requirements must be explained. A certication of
GLP compliance is required in the IND for each safety study (21 CFR 58).
3. Chemistry, Manufacturing, and Controls
The next major section of the IND describes the chemistry of the drug
substance and the composition, manufacturing, packaging, quality control,
and stability of the dosage forms to be used in the clinical trials.
Inadequacies in this section can cause FDA to withhold the approval to
conduct the proposed clinical trials, particularly if the deciencies cause a
concern related to safety.
a. Drug Substance. The drug substance must be characterized as to
its structure and adequate analytical procedures must be specied to analyze routinely the identity and purity of the drug. If the drug is in an ofcial pharmacopeia, the monograph for the drug may be referenced;
however, the FDA is not bound by these requirements and may require
additional tests and specications. For example, the assay method for the
drug should be stability indicating, and some monographs may not meet
this standard. Also, FDA will be interested in the major impurities and require specic tests and specications for them, which may not be part of
a monograph.
The supplier of the drug substance is required to be identied as well
as the methods of synthesis and controls used by the manufacturer. This will
also be required for compendial drugs. Since the information on synthesis is
usually considered proprietary, FDA has established a mechanism by which
this condential information can be supplied for their review directly from
the manufacturer through a Drug Master File (DMF) (7). The manufac-

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

681

turer will send the FDA a letter authorizing the IND sponsor to access this
information in a specic DMF, and the sponsor is required to include a copy
of this letter of authorization in this section of the IND.
An authentic reference standard is required for each drug substance. If
not available from USP, it will have to be established independently, must
be of the highest purity available, and the method of synthesis and purication must be included.
b. Drug Product. The dosage form containing the active drug substance and its vehicle or delivery system must be described in detail:
ComponentsListing of all active and inactive ingredients that are
used in preparation of the nished product.
Inactive ComponentsThe quality standard which is used and, if
other than compendial items, the actual tests and specications.
CompositionThe quantitative composition for the entire formula
expressed in terms of percent, milligrams per milliliter, and a typical
batch quantity.
ManufactureThe name and address of each rm involved in the
manufacture, packaging, labeling, and testing of the drug product.
Method of ManufacturingThe method of manufacturing, packaging, and labeling the product and the controls used in these processes.
Packaging ComponentsThe packaging is identied and the components are described and specied. The USP species tests required
for suitability of plastics in ophthalmic containers, which are both
physicochemical and biological. The tests should be conducted on
the containers after they are cleaned and sterilized.
StabilitySucient stability data using stability-indicating methods
should be submitted to assure a stable product for the duration of
the clinical trials.
LabelingCopy of the labels to be applied to the containers of the
clinical supplies. These are usually multipart labels so as to provide
complete labeling information during shipment, which can be
removed before given to the patient to mask the identity of the
product from the patient and the physician. The label must bear
the statement: Caution: New Drug Limited by Federal Law to
Investigational Use.
PlaceboMany studies require the drug to be compared to a placebo,
which is usually the vehicle or delivery system itself. The same
information described above for the active drug product is provided
for the placebo dosage form.

Copyright 2003 Marcel Dekker, Inc.

682

Roehrs and Krueger

Environmental AssessmentThe environmental regulations provide

for a categorical exclusion from preparing an environmental assessment for clinical trials.
c. Sterilization. FDA regulation requires that all ophthalmic drug
products be manufactured and packaged in a sterile manner. This can be
accomplished in two ways: by terminal sterilization or by aseptic combination of sterile components. Terminal sterilization is preferred, since it
provides the greatest assurance of nal product sterility. Often this is not
feasible because of the heat lability of the ingredients or the packaging
system. Terminal sterilization is usually done by radiation or steam under
pressure. Many drug products are made sterile by sterilizing the individual
components, including the packaging materials, and then aseptically combining them in a sterile environment. The FDA has provided guidelines
for the proper validation of the aseptic process for sterile products (8).
Because of the much greater sterility assurance oered by a terminal
sterilization process, the FDA will require the applicant to justify the use of
an aseptic process for sterilization (9). Data, therefore, should be generated
during the development of the delivery system to determine the impact of a
terminal sterilization process on the nal packaged dosage form. This would
usually involve chemical and physical analyses of the product for degradation products and any change in the toxicology prole.
d. Preservation. If the delivery system is a liquid in a multiple dose
package, the requirement for addition of substance(s) to inhibit the
growth of microorganisms during use needs to be considered. The regulation states that these substances must be added or the product packaged
in such a manner that harmful contamination cannot reasonably occur
during use (21 CFR 200.50). Therefore, the regulation does not absolutely
require a preservative be used in all multidose packages. However, FDA
has not given any published guidance as to their interpretation of what
type of unpreserved multidose product would meet the requirements of
this regulation.
The regulation specically states a requirement for liquid products and
provides no guidance for multidose semisolids such as aqueous gels. The
drug delivery scientists should work closely with the microbiologist and
packaging scientist to determine the best means to accomplish the safe
administration of a sterile product.
New ophthalmic delivery systems in unit dose form oer the opportunity to improve the ability of the patient to comply with the prescribed
dosage regimen and also obviate the need for the addition of a preservative
agent.

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

683

e. Good Manufacturing Practices. Clinical supplies of the new

ophthalmic drug delivery system should be manufactured in conformance
with the GMP regulations. The pharmaceutical industry has recognized
that the ofcial GMP regulations are not always practical or suitable for
clinical supplies. The quantities are usually small, and since this is a research and development process, changes will necessarily occur as experience is gained and the scale of manufacture increases. FDA has also
recognized these facts and has issued guidelines that address the allowable
differences in GMP compliance for clinical trial manufacture (10).

VI.

MARKETING NEW OPHTHALMIC DRUG DELIVERY

SYSTEMS FOR HUMAN USE

Once a manufacturer has obtained sucient information from clinical trials

in humans, safety testing in animals, and chemistry and manufacturing
experience to establish that the new ophthalmic drug delivery system is
safe and eective for its intended use, a New Drug Application is submitted
to the FDA to obtain approval for distribution in interstate commerce.
Two types of NDAs for new drugs are recognized by the FD&C Act in
Section 505(b):
1. A new drug requires full reports of investigations for safety and
eectiveness among other requirements. For a new drug, rst
submitted for approval, the investigations providing the full
reports will ordinarily have been conducted by the applicants
or by others under contract with a full right of reference (21
CFR 314.3) to the results of the investigations for inclusion in
the marketing application. This is the typical NDA for a new
chemical entity (NCE) and is sometimes referred to as a Full
NDA.
2. The second type of NDA is one in which the applicant does not
have a full right of reference for some or all of the investigations
relied upon for demonstrating safety and eectiveness and is
relying on data generated by others, for example, publicly available information. This has been termed a paper NDA but is
now referred to as a 505(b)(2) NDA for the section of the Act.
The 505(b)(2) NDA is applicable to the development of a new ophthalmic
delivery system for an already approved drug because once the new drug is
approved, and any applicable patent protection and market exclusivity periods have expired, the safety and ecacy data become public information
and can be referenced by another applicant as part of a subsequent applica-

Copyright 2003 Marcel Dekker, Inc.

684

Roehrs and Krueger

tion for a modication of the approved drug (21 CFR 314.430). For the new
ophthalmic delivery system of an approved drug, the basic human clinical
and animal safety studies would not have to be repeated but only referenced
to the approved drugs NDA. Full reports would only be required for
demonstration of safety and ecacy of the new delivery system, i.e., the
changes made to the approved drug. A complete chemistry, manufacturing,
and controls section would also be required as this is not public information
(21 CFR 314.54). Additionally, for a Section 505(b)(2) NDA, a patent certication is required in the application for any patent that claims the drug,
drug product, or method of use for which prior investigations are relied upon
without a right of reference from the original applicant. In addition to any
patent protection, the new ophthalmic delivery system may be eligible for 3
years of Waxman-Hatch market exclusivity (21 CFR 314.108).
A.

NDA Content and Requirements

The format and content of an NDA is seen in Table 4. The FDA Form 356h
is the ocial signed application form.
An index to the entire application is essential to direct the reviewer to
the location of the contents by volume and page numbers.
A comprehensive summary of the entire application is written in the
style of a review article. A copy of the index and summary is provided to
each reviewer of the application. The summary begins with an annotated
copy of the proposed labeling, i.e., the products proposed package insert.
The applicant must be able to justify each statement in the labeling which is
annotated to the supporting information in the Summary and Technical
sections.
There are six major technical sections comprising the data generated to
support the approval for the claimed indication(s). FDA guidelines for the
technical sections are included on the FDAs web site at www.fda.gov/cder/
guidance/index.htm.
Chemistry, Manufacturing and Controldrug substance and drug
product
Microbiologyapplicable only if for an anti-infective drug
Human PharmacokineticsADME data from human studies
Pharmacologyanimal study data for pharmacology, toxicology, and
ADME
Clinicalhuman data for safety and ecacy
Statisticsmathematical analysis of the human clinical data
There are several sections ancillary to the technical sections. A section
containing documentation for the analytical methods validation, drug sub-

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations
Table 4

NDAFormat and Content Outline

FDA Form 356h

Index to Entire Application
Summary
a. Labeling text annotated to both Summary & Technical Sections
b. Pharmacological Class, Scientic Rationale, Intended Use(s), and
Potential Clinical Benets
c. Foreign Marketing HistoryApplicant and others if known
i.
Countries in which marketed
ii. Countries where withdrawn from market for safety or ecacy
iii. Countries where marketing applications pending
d. Summary of Chemistry, Manufacturing and Controls Section
e. Summary of Nonclinical Pharmacology & Toxicology Section
f. Summary of Human P-Kinetics & Bioavailability Section
g. Summary of Microbiology Section
h. Summary of Clinical Section including Statistical Analysis
i. Concluding DiscussionBenet/Risk Consideration
Chemistry, Manufacturing, and Controls Section
a. Drug Substance
i.
Description of Chemical & Physical Characteristics
ii. Stability
iii. Source & Location of Manufacturer(s)
iv. Synthesis, Purication, and Controls
v. Specications & Analytical Methods
vi. Reference Standard
b. Drug Product
i.
Components
ii. Specs and Test Methods for Inactives
iii. Composition
iv. Name and Location of Each Manufacturer
v. Manufacturing & Packaging Procedures & In-Process Controls
vi. Acceptance Specications for Each Batch
vii. Analytical and Microbiological Test Methods
viii. Packaging Container/Closure Systems
ix. Stability Data & Protocols
x. Proposed Expiring Dating & Storage Conditions
c. Environmental Assessment of Manufacturing Process & Ultimate Use
Samples, Methods Validation, Labeling
a. Description & assay results of samples for FDA validation
b. Methods Validation Package (Regulatory Specs & Methods)
c. Container Labels & Package Insert Labeling
Nonclinical Pharmacology & Toxicology
a. Pharmacology Studies
b. Toxicology Studies
c. Animal ADME Studies
d. GLP Compliance Statements

Copyright 2003 Marcel Dekker, Inc.

685

686

Roehrs and Krueger

Table 4 Continued
Human Pharmacokinetics & Bioavailability
a. Waiver for topical or injectable product bioavailability studies
b. For each human bio- or pharmacokinetic study:
i.
Study report including analytical and statistical methods
ii. IRB/Informed Consent Compliance Statements
c. For specs or methods to assure bioavailability of drug or product:
i.
Rationale for establishing spec or method
ii. Data and information supporting rationale
d. Summary Discussion & Analysis of:
i.
Pharmacokinetics & metabolism or active ingredient
ii. Bioavailability and/or bioequivalence of drug product
Microbiology (for Anti-Infectives Only)
Clinical Data
a. Clinical Pharmacology Studies
b. Controlled Clinical Studies
c. Uncontrolled Clinical Studies
d. All other data and information relevant to evaluation of the safety
and eectiveness of the drug product
e. Integrated Summary of Eectiveness
f. Integrated Summary of Safety
g. Studies related to abuse potential or over dosages
h. Integrated summary of Benets and Risks
i. Compliance Statements (IRB & IC) for each clinical study
j. Contract Research Organizations & Obligations Transferred
k. List of studies where original subject records were audited or reviewed
to verify accuracy of case reports
Reserved for Safety Update Reports
a. Required After NDA Filed at:
i.
4 months
ii. Approvable letter and whenever FDA requests update
b. New Safety Information from any source that may reasonably aect
the labeling
c. Case Report Forms for patients who died or were adverse event
dropouts
Statistical Section
a. Copy of the following Clinical Data Sections
i. Controlled Studies
ii. Integrated Summary of Eectiveness
iii. Integrated Summary of Safety
b. Documentation and Supporting Statistical Analysis used to Evaluate
each of the above Clinical Data Sections
Case Report Tabulations
Data on each patient from
a. Each adequate and well-controlled study (Phase 2 & 3)

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

687

b. Each Phase 1 Clinical Pharmacology study

c Safety data from other clinical studies
Case Report Forms
a. For patients who:
died during a clinical study or did not complete study due to adverse event
whether or not thought to be drug related including patients on reference
drug or placebo
b. Additional CRFs may be requested after submission and must be submitted
within 30 days of FDA request
Pediatric StudiesInformation on use in Pediatric Population unless
Waived or Deferred
Patent InformationListing of U.S. Patents that could be infringed
which claim
a. Drug
b. Drug product composition
c. Intended use
Patent CerticationFor 505(b)(2) Applications
Claimed Market Exclusivity
Financial Certication or Disclosure Statements
Source: Adapted from 21 CFR 314.50.

stance and drug product samples, and the container labels and package
insert. The samples are submitted to one or more FDA district laboratories
for validation of the regulatory analytical methods and to conrm the identity, purity, and strength of the drug and drug product. The FDA uses the
regulatory analytical methods to periodically test the drug product from
regular production batches after approval. A section is reserved for periodic
updates of new safety information that may be obtained after the NDA is
submitted while under review and just prior to approval.
The case report tabulations contain the clinical data from each
patients case report form (CRF). This is the raw data from the clinical
studies and is tabulated by entry into sophisticated relational databases
for evaluation and mathematical statistical analysis. A copy of the actual
CRFs is only required in the initial submission and periodic safety updates
for all deaths and patients discontinued due to adverse medical events.
However, the FDA may request additional patient CRFs during the review.
Studies in pediatric patients (birth to 16 years) are required for all new
drugs during development and for certain marketed drugs (21 CFR 314.55).
The applicant may request a waiver or deferment of the studies. A waiver
may be granted if the new drug may not provide a meaningful therapeutic
benet over existing pediatric treatments and is not likely to be used in a
substantial number of pediatric patients. The scope of the regulation

Copyright 2003 Marcel Dekker, Inc.

688

Roehrs and Krueger

includes new active ingredients, new indications, new dosage forms, new
dosage regimens, and new routes of administration. For marketed drugs,
FDA can require pediatric studies for those that are used in a substantial
number of pediatric patients for the claimed indications or would provide a
meaningful therapeutic benet over existing treatments and where the
absence of adequate labeling could pose signicant risks.
To provide adequate data to meet the requirements may not always
require controlled clinical trials in pediatric patients and may not require
separate studies. In some cases where the course of the disease and the
products eects in adults and children are similar, the FDA may accept
an extrapolation from adult studies with additional data such as dosing,
pharmacokinetic, and safety data in pediatric patients. Also, adequate data
may sometimes be obtained by including sucient pediatric patients as well
as adults in the original clinical studies where appropriate. The pediatric
data is only required for the indications claimed by the applicant.
Since the WaxmanHatch Act of 1984, NDAs must contain information of all unexpired U.S. patents for the drug substance, drug product, and
methods of use (21 CFR 314.53). The applicant certies the validity of the
patent information and the FDA, without verication, upon NDA approval
publishes the information in the Orange Book so that subsequent applicants
are put on notice as to the patents that may be infringed and delay their
entry into the marketplace.
For each patent that claims the drug, drug product, or method of use
for which prior investigations are relied upon without a right of reference
from the original applicant, a patent certication must be included (21 CFR
314.50(i)). The certication must show that there is no patent information
led with the FDA, that the patent has expired or the date the patent will
expire, or that the patent is invalid or will not be infringed. If the certication that the patent is invalid or not infringed is made, notice must be given
to the patent and NDA holder, who have 45 days to bring an infringement
action in court, which enjoins the FDA from approving the application for
30 months or until the case is resolved by the court, whichever is shorter (21
CFR 314.52).
The Act provides for periods of market exclusivity upon approval of
most NDAs. The exclusivity period must be requested by the applicant and
if granted by the FDA will be published in the Orange Book for notice to
generic applicants (21 CFR 314.50(j)). A 5-year period of market exclusivity
is provided for the rst approval of a new chemical entity and a 3-year
period for all others, such as new uses and new dosage forms. Certain
criteria must be met for the grant of market exclusivity, primarily the
requirement for the conduct of clinical studies essential to approval as
well as criteria for their nancial support. During the exclusivity period,

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

689

generic equivalent products cannot be marketed; however, during the 3-year

period a generic application may be submitted. For the 5-year period, a
generic application cannot be submitted during the rst 4 years and during
the last year of exclusivity can be submitted only if there is a patent challenge.
Congress included economic incentives for conducting pediatric studies in the Modernization Act of 1997 for drugs that are eligible for marketing exclusivity or patent protection under the Waxman-Hatch Act of 1984.
The legislation makes available a 6-month extension of existing market
exclusivity on drug products that the FDA has made a written request for
pediatric studies and the manufacturer has conducted such studies meeting
the requirements.
A long-standing requirement of clinical studies is that they are
designed, conducted, analyzed, and reported in a manner that minimizes
bias. One element that could introduce bias is the nancial arrangements
between the sponsor of the study and the clinical investigator. Since 1999,
the FDA has required certain nancial disclosure information for designated clinical studies be included in NDAs for their use in assessing data
reliability in making the nal approval decision (21 CFR Part 54).
The FDA is interested in certain nancial arrangements for clinical
studies that are relied on for proof of ecacy, including comparisons to
eective products, and those that make a signicant contribution to the
demonstration of safety. Specic nancial arrangements that must be disclosed for each such clinical investigator are those where the compensation
of the clinical investigator is aected by the outcome of the study, or where
there is signicant equity interest in the sponsoring company, or compensation is tied to product sales, or a proprietary interest in the product, or
receives signicant payments from the sponsor such as for grants to fund
ongoing research, compensation in the form of equipment, retainer for
ongoing consultation or honoraria, and any steps taken to minimize the
potential for bias due to these nancial arrangements. If none of the
above nancial arrangements were in eect for the clinical studies, then
the company may submit a certication attesting to this fact for each such
clinical investigator.
B.

Electronic Submission

The FDA has encouraged submission of parts or all of an NDA in electronic

form to expedite the review of large datasets such as clinical, animal safety,
biopharmaceutics, and chemistry sections. These computer-assisted NDAs
(CANDA) have been adopted by most pharmaceutical companies in the
anticipation that they will expedite the review and approval of their submis-

Copyright 2003 Marcel Dekker, Inc.

690

Roehrs and Krueger

sions. However, until recently, they could not be used to replace the ocial
paper copy version of the NDA, which can involve more than 100 volumes.
In March 1997, a new Electronic Records regulation was promulgated,
allowing acceptance of documents entirely in electronic format with a certifying electronic signature. The FDA has been active in providing guidance
to industry so that regulatory submissions can increasingly be made in
electronic format with the hope that entire regulatory submissions can be
paperless. The FDA provides updated information and guidance for electronic submissions on their website (www.fda.gov/cder/guidance/
index.htm).
C.

Patent Term Extension

The Waxman-Hatch Act makes available the possibility of extending the

term of one unexpired patent for the drug, drug product, or use but only
upon the rst NDA approval for a new chemical entity. This provision was
enacted to partially restore the eective patent life that is lost during the
time required to meet FDA approval requirements. The patent holder must
apply within 60 days of NDA approval to the U.S. Patent and Trademark
Oce (PTO) for extension of the selected unexpired patent. The PTO and
FDA cooperate to determine the eligibility of the patent and the period of
time for the extension. The extension is determined by a formula based on
the amount of time diligently pursuing development under the IND and the
amount of time from NDA ling until approval. The term extension is
subject to the limitations of a maximum of 5 years and a maximum of 14
years patent term remaining with the extension (21 CFR 60).
D.

Accelerated Availability of Certain New Drugs

The acquired immunodeciency syndrome (AIDS) crisis focused attention

on the drug development and regulatory processes for making new therapies
available for life-threatening and severely debilitating conditions, particularly where no approved therapy or suitable alternative therapies exist. FDA
instituted new procedures to make promising new drug therapies available
to treat patients as early in the drug development process as possible,
whereas at the same time obtaining data on the drugs safety and eectiveness in clinical trials.
The treatment IND regulations nalized in 1987 (21 CFR 312.34)
make it possible for patients to receive promising new drugs for certain
serious and life-threatening conditions as early as Phase 2 or 3 while controlled clinical trials are still underway to make a nal determination on the
new drugs safety and eectiveness. In some cases the manufacturer may

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

691

charge for the investigational new drug product if necessary to undertake or

continue the clinical trials. The sponsor of the new drug must apply to FDA
for treatment use and provide available data supporting the drugs potential
ecacy and safety. FDA may deny the request if the available scientic
evidence, taken as a whole, is insucient to support a reasonable conclusion
that the drug may be eective for its intended use or would expose patients
to unreasonable and signicant additional risk of illness or injury. Also,
clinical trials must be continuing to complete development and submission
of an application to market the drug.
FDA also instituted accelerated drug development and review programs for certain serious or life-threatening conditions so that approval
can be obtained for marketing at the earliest possible point that safety
and ecacy are reasonably established (21 CFR Part 312 Subpart E). The
FDAMA of 1997 essentially codied these programs for certain new drugs
designated as fast track products (FD&C Act Section 506). The fast track
programs of FDA are designed to facilitate the development and expedite
the review of new drugs that meet certain requirements as treatments for
serious or life-threatening conditions and that demonstrate potential to
address unmet medical needs. For these fast track products, FDA may
approve an NDA using a clinical endpoint or a surrogate endpoint that is
reasonably likely to predict clinical benet. Also, FDA can accept for review
portions of a marketing application prior to receipt of the complete application (21 CFR Part 314 Subpart H).
In addition to the fast track programs accelerating development and
market availability of certain new drugs, the FDA, working with industry
and Congress, has established NDA review time goals for all NDAs as part
of the quid pro quo for payment of user fees. Priority applications are to be
reviewed within 6 months of ling and all other applications (standard) are
to be reviewed within 10 months. The review classication as a priority or
standard NDA is established at the time the NDA is led and is based on
FDAs estimate of the drug products medical value, compared to already
available products, using the safety and eectiveness submitted in the original application. To be designated for a priority review, the FDA must
consider the drug product to be a signicant improvement compared to
marketed products as demonstrated by such elements as increased eectiveness, elimination or substantial reduction of an adverse drug reaction limiting use, documented enhancement of patient compliance, or evidence of
safety and eectiveness in a new subpopulation.

Copyright 2003 Marcel Dekker, Inc.

692

VII.

Roehrs and Krueger

INTERNATIONAL HARMONIZATION

The technical requirements for registration of human pharmaceuticals vary

throughout the world, which makes global development more expensive
and delays introduction of new products in many markets while studies are
being repeated to meet local requirements. In 1990, a unique project was
undertaken to bring together the regulatory authorities of the United
States, the European Union, and Japan along with scientic and regulatory experts from these three regions with the aim to produce a single set
of technical requirements for the registration of new drug products and
hence streamline the development process while maintaining the necessary
safeguards to protect public health. The International Conference on
Harmonization (ICH) is the driving force behind the project with representatives from the three major market regions and where most of the new
drug development originates. The output of ICH are tripartite guidelines
in the major areas of drug development: ecacy, safety, and quality. As of
January 2000, 37 guidelines have been nalized and adopted as tripartite
guidelines after rigorous scientic review and modication by the expert
working groups. The ecacy topics cover clinical testing programs and
safety monitoring, the safety topics cover preclinical toxicity and related
tests, and the quality topics are concerned with chemistry and control
specications. Harmonized guidelines are also being developed for products of biotechnology.
In addition to diering technical requirements in the three ICH
regions, the requirements for the technical sections of registration applications also dier, and a separate document must be prepared for each
region. The concept of a Common Technical Document (CTD) for each
of the three major topics has been adopted by ICH along with a standardized terminology for the reporting of adverse reactions in the Medical
Dictionary for Regulatory Activities (MeDRA). Also being explored by
ICH is harmonization of electronic information transfer standards
(ESTRI) so that the CTD can be established eventually as an electronic
document.
The ICH guidelines and additional information on the ICH process
can be obtained from the ICH Secretariat website at www.ifpma.org/
ich1.html. The FDA publishes the ICH guidelines in the Federal Register
at the draft stage for public comment and the nalized guidelines upon
adoption. They can be obtained from the FDA web site at www.fda.gov/
cder/guidance/index.htm.

Copyright 2003 Marcel Dekker, Inc.

Regulatory Considerations

693

REFERENCES
1. Nielsen, J. R. (1986). Handbook of Federal Drug Law. Philadelphia, Lea &
Febiger.
2. J. L. Fink, et al., eds. (1991). Pharmacy Law Digest. Facts and Comparisons,
Inc., p. DC-4.
3. Guidance for Industry, Environmental Assessment of Human Drugs and
Biologics Applications (Rev. 1), July 1998.
4. Fed. Reg., 47(200):46139, October 15, 1982.
5. Lerman, S., and Reininger, B. (1971). Can. J. Ophthalmol., 6:14.
6. Hackett, R. B. (1990). Non clinical study requirements for ophthalmic drugs
and devices in the United States. Lens & Eye Toxic. Res., 7(3&4):181.
7. Guideline for Drug Master Files (Sept. 1989). Center for Drug Evaluation and
Research, FDA.
8. Guidelines on Sterile Drug Products Produced by Aseptic Processing (1987).
June, FDA.
9. Fed. Reg., 56(198):51354, October 11, 1991.
10. Guideline on the Preparation of Investigational New Drug Products (Human &
Animal) (1991). March, FDA.

Copyright 2003 Marcel Dekker, Inc.

21
Patent Considerations
Robert E. Roehrs

Alcon Research, Ltd., Fort Worth, Texas, U.S.A.

I.

INTRODUCTION

A patent is the legal instrument by which inventors protect their substantial

investment in time, eort, and money to create new technology and new
products. Patents are a form of intellectual property with rights aorded by
the U.S. constitution in Article I, Section 8: The congress shall have the
power . . . to promote the progress of science and the useful arts by securing
for limited times to . . . inventors the exclusive right to their . . . discoveries.
Patents may be viewed as a contract between the inventor and the
government under which the inventor is granted a limited monopoly to
exclude others from making, using, selling, oering to sell the claimed invention in the United States, or importing it into the United States for a limited
period of years. In return for this grant, the inventor agrees to disclose to the
public via the patent the complete invention. Patents are an important
public source of technological information, and studies show that a majority
of patents contain information that is not published elsewhere. This public
disclosure promotes the progress of science by allowing others to begin
using the information to make their own improvements, which may also
be patentable and thus disclosed to the public. This is quite common in the
pharmaceutical eld, where many patents are improvements on earlier discovered drugs and their medical uses and dosage forms.
Patents are a synthesis of law and science. The legal issues involved in
protecting inventors rights can be as complex as the subject matter of the
invention. The pharmaceutical scientist needs legal expertise to fully protect

Retired

695

Copyright 2003 Marcel Dekker, Inc.

696

Roehrs

his or her inventions. Although patent law allows inventors to represent

themselves in obtaining a patent, it is advisable that they work closely with
a patent legal specialist, such as a registered patent attorney or agent, to
obtain an issued patent with the broadest legal protection that the law allows.
It is helpful if the inventor has some understanding of the basic legal requirements for patenting inventions to work eectively with the legal expert who
will prepare and prosecute the patent application. This overview refers only
to U.S. patents and patent law requirements unless otherwise indicated.

II.

UTILITY PATENTS

There are three types of patents: utility, design, and plant patents. Each has
its own legal requirements for obtaining patent protection. Pharmaceutical
patents for drugs and their delivery systems and uses are considered utility
patents. In this chapter we will be discussing the requirements and uses of
utility patents exclusively.
A patent consists of several parts that will be referred to in subsequent
sections. Two major parts of a patent that the drug delivery researcher
should become familiar with are the specication and the claims. The specication is the written description of the invention that concludes with one
or more claims that dene and delineate the scope of the rights granted to
the inventor. Patent claims are often compared to the description of real
property contained in a deed that sets the boundaries for the property. For a
patent, everything contained within the boundaries of the claims is the
exclusive right granted to the inventor.

III.

PATENT RIGHTS

A.

Right to Exclude

A U.S. patent gives the inventor the right to exclude others from making,
using, selling, or oering for sale his claimed invention within the United
States or importing the invention into the United States. If the U.S. patented
invention is a process, the right to exclude also prevents others from using,
oering for sale, or selling in the United States, or importing into the United
States, products made by that process. These infringing acts are sometimes
referred to as practicing the invention. In this chapter, the terms practice or use of the patented invention are used as shorthand for any and
all of the statutory infringing acts. Along with this exclusionary right is the
right to bring legal action in the courts when others practice the patented
invention prior to its expiration without permission from the patent owner.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

697

The right to exclude (a negative right) means that the inventor may in
some cases be excluded from practicing his own patented invention. This
occurs when there exists an unexpired U.S. patent that is broader in claims
than the inventors patent. The concept of freedom to practice or dominant/subservient patent claims is important for the drug delivery researcher
to understand, since he may consider incorporating patented drugs into his
new delivery system to improve their therapeutic benets. The owner of the
unexpired drug patent can exclude others from using the drug and therefore
has a dominant exclusionary right over subsequent patents that include the
drug as part of their claimed invention. Also, patentable improvements may
in some cases be subservient to the original inventions patent claims. On the
other hand, the inventor of the drug or earlier delivery system cannot use the
drug in the separately patented delivery system or use the improved version
without permission.
B.

Right to Assign

The inventor is the owner of the patent rights, and he or she may assign, i.e.,
transfer his or her interest in the patent, either the exclusive rights to the
whole patent or an undivided part or share of these rights. Employers may
require their employees (inventors) to assign their exclusive rights to inventions conceived as part of their employment. The assignment is recorded in
the United States Patent and Trademark Oce (USPTO). Upon assignment, the employer owns the property rights inherent in the patent.
C.

Right to License

The patent owner may license, i.e., transfer some or all of the patent rights;
however, title is retained by the patent owner. A license may be exclusive or
nonexclusive, and the licensee may be given the right to sublicense. The
commercial value of a drug delivery system patent often involves the ability
to license the technology to owners of patented drugs with limitations in
their pharmacokinetics or duration of eect. Companies such as Alza have
become successful using this corporate strategy.

IV.

PATENT TERM

The patent grant is for a limited term, which in the United States is now 20
years from the earliest eective ling date for applications led beginning
June 8, 1995. Prior to the new law, patents had a term of 17 years from issue.
In the transition to the new patent term, there are issued patents in eect

Copyright 2003 Marcel Dekker, Inc.

698

Roehrs

with expiration dates of 17 years from issue, or 20 years from the earliest
eective ling date, or various dates depending on certain patent term
extensions that have been enacted by Congress. Patent maintenance fees
are required at 3.5, 7.5, and 11.5 years after issuance (subject to a grace
period), and failure to pay the required fee will result in premature expiration of the patent term.
Patent expiration dates for FDA-approved drugs are published in the
FDA Orange Book. These dates are provided to FDA by the company
submitting the NDA and are not validated by FDA or the USPTO. (Patent
term extensions for FDA-approved drugs are discussed in Chapter 20.)
Patent terms may also be extended by the USPTO in certain cases where
delays in patent application review would provide the inventor with less
than a 17-year patent term. The new minimum patent term guarantee
applies only to delays by the USPTO during examination of the application
and not to delays caused by the applicant.

V. PATENTABLE INVENTIONS
Congress has dened patentable inventions in four broad categories:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement
thereof may obtain a patent therefor, subject to the conditions and requirements of this title.
Ophthalmic drug delivery systems (ODDS) typically are claimed in a
patent as compositions of matter and/or processes. Compositions of matter
include chemical compounds and physical mixtures. Drug delivery system
dosage forms are usually a combination or association of ingredients that
cooperate to produce a unitary result exhibiting properties dierent from
those possessed by the individual ingredients. One of the ingredients may be
a drug molecule, which, in association with the ingredients of the delivery
system (vehicle), is delivered in a new or improved manner to the eye. Drug
molecules sometimes are developed specically for ocular drug delivery
purposes, i.e., prodrugs such as dipivefrin or the carbonic anhydrase inhibitors dorzolamide and brinzolamide, and are claimed as compositions of
matter.
A patentable process consists of an act, operation, step, or series of
steps performed upon specied subject matter to produce a physical result.
A patentable process includes the method of making and method of using a
substance, e.g., a pharmacologically active ingredient. Method of use or
treatment claims often are recited in drug delivery patents. These so-called
use patents also can include new uses for old products.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

699

A patentable invention must contain claim(s) to statutory subject matter that are novel (new), useful (utility), and nonobvious among other requirements for specicity of the claims and adequacy of disclosure of the invention.
A.

Utility

This is perhaps the easiest of the three statutory requirements to satisfy for
patentability, particularly for most pharmaceutical inventions. Only useful
inventions may be patented, i.e., inventions providing a benecial use in
society. A useful invention does not have to provide an advantage or an
advance in the art to be considered to meet the utility requirement, nor does
it have to necessarily provide the best or even a commercially feasible product or process. Issues of utility may arise if the usefulness is not apparent or
credible. A patent usually states under the objectives of the invention one or
more intended purposes for the invention. These statements of specic utility are usually sucient unless there is a reason for one skilled in the art to
question the truth of the statement of utility or its scope. The invention must
be operable for at least one of the intended purposes to meet the utility
requirement.
For pharmaceutical inventions, issues can arise if the utility alleged has
never before been demonstrated or there are no in vitro tests that correlate
with a practical utility or animal tests with recognized relevance to human
utility. If human use is not alleged, it need not be proved, veterinary utility
being sucient. Also, FDA requirements for safety and ecacy are not
patent law requirements for utility. Indeed, patent protection is often
obtained prior to completion of human clinical trials, and it is the patent
rights that provide the economic incentive to complete FDAs marketing
requirements.
B.

Novelty and Loss of Right to Patent

Useful statutory subject matter may be patented if it meets the legal requirements of novelty and nonobviousness, and these two requirements essentially dene the limits of what may be protected for a particular invention.
Novelty is relatively straightforward, i.e., if the claimed subject matter has
been done before in exactly the same manner, then it is old (lack novelty)
and is not patentable. However, even if the invention is new, further inquiry
is required to determined whether it is nonobvious, i.e., not obvious to one
of ordinary skill in the art to which it pertains. The patent law also denes
certain circumstances where acts by the inventor or others may prevent a
patent from being obtained, i.e., so-called statutory bars, even though the
invention may meet all other requirements.

Copyright 2003 Marcel Dekker, Inc.

700

Roehrs

The novelty provisions of patent law prevent an inventor from obtaining a patent for subject matter that was known, used, patented, described, or
made by another prior to the applicants invention. The novelty requirement
is one of strict identity, i.e., a single reference must contain all the limitations
of the claimed subject matter and in the same order (steps) in which they are
arranged in the claim. This is termed anticipation, and for meeting the
novelty requirements, the USPTO cannot combine more than one reference
to show anticipation of the claimed subject matter.
Patent law prescribes the following conditions, which if are found to
have occurred prior to the invention prevent a patent from issuing for lack of
novelty:
1. The invention was known or used by others in this country, or
patented or described in a printed publication in this or a foreign
country
2. The invention was described in a patent granted on an application for patent by another led in the United States or on an
international application by another fullling certain requirements
3. The invention was made in this country by another who had not
abandoned, suppressed, or concealed it
Note that knowledge or use by others outside of this country does not
prevent a patent unless the invention has been published or patented anywhere in the world. Knowledge or use in this country refers to that which is
accessible to the public. The terms others and another refer to any
entity dierent from the so-called inventive entity, i.e., single inventor or
joint inventors.
Certain acts by the inventor may cause a loss of rights and bar the
issuance of a patent:
1. The invention was patented or described in a printed publication
in this or a foreign country or in public use or on sale in this
country, more than one year prior to the ling date of the U.S.
patent application.
2. The invention was abandoned.
3. The invention was rst patented or caused to be patented in a
foreign country prior to the ling date of the U.S. patent application on an application in the foreign country led more than 12
months before the ling of the U.S. patent application.
United States patent law provides a 1-year grace period for ling a
patent application in this country when the invention has been patented or
published anywhere in the world or for public use or sale in this country.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

701

Thus it is important that the inventor recognize when the 1-year grace
period may be triggered and keep the patent legal specialist informed of
all potential public disclosures, uses, and oers for sale of the invention so
that a timely application may be led to prevent a statutory bar. It is also
important that the inventor not try to second-guess the meaning of these
legal conditions since they have been and are subject to various legal interpretations by the courts based on specic set of facts and circumstances. For
example, what is considered printed is certainly now more complex than
when it was rst promulgated and has been interpreted to mean all material
accessible to the public in tangible form, including a single copy of a thesis
indexed and catalogued in a university library. The conduct of tests in the
public eye to perfect an invention or the sale of an invention for the sole
purpose of experimental use may or may not trigger the 1-year grace period,
and legal advice should be obtained in advance.

C.

Nonobviousness

A new and useful invention must also meet the statutory requirement for
nonobviousness. The patent law states that if the dierences between the
subject matter sought to be patented and the prior art are such that the
subject matter as a whole would have been obvious at the time the invention
was made to a person having ordinary skill in the art to which the subject
matter pertains, then the invention is considered obvious and a patent
cannot be granted.
The determination of whether a particular invention is or is not
obvious is largely subjective, although the courts have established some
guidelines and tests for its determination as a matter of law. Obviousness
does not hinge on the manner in which the invention was made, i.e., there is
no requirement for a ash of genius. Many inventions are made after
considerable experimentation or even by error or accident and rarely by
revolutionary imagination. The test of obviousness of an invention is not
whether it would have been obvious to try the combination of steps or
elements that led to the invention. In hindsight, it may seem like it would
have been obvious to try certain experiments leading to the claimed subject
matter, but the focus of the patent law is on the level of knowledge and skill
in the art at the time the invention was made.
To ascertain if the nonobviousness requirement has been met, the
USPTO and the courts examine the following:
1. The scope and content of the prior art
2. The dierences between the art and the claims at issue

Copyright 2003 Marcel Dekker, Inc.

702

Roehrs

3. The level of ordinary skill in the art

4. Whatever objective evidence may be present
The prior art is the body of knowledge or information that is publicly
available or accessible at the time of the invention. Patent law presumes
that the inventor and the person of ordinary skill have full and comprehensive knowledge of the prior art pertinent to the invention. For determining
obviousness, reference is made to the scope and content of the art relevant
to the subject matter of the invention as well as to so-called analogous art.
This is dierent from the requirement for novelty where anticipation is not
restricted to the prior art of the invention but includes all prior art.
Analogous prior art is that art that may be outside the eld of the invention
but that the inventor would examine to solve a problem or obtain a result
leading to the invention. Analogous art is said to be art that is reasonably
pertinent to the particular problem the inventor was involved with, and
there is a similarity of elements, problems, and purposes.
The content of the prior art is examined for what it teaches about the
claimed subject matter. Does the prior art teaching render it obvious to one
of ordinary skill in the art, or does it teach away, i.e., discourage one of
ordinary skill from following the path taken by the inventor or suggest that
it would be unlikely to be productive? Prior art that teaches away can be
evidence of nonobviousness.
If the invention is novel, then there must exist some dierence from
that found in the prior art, and it is this dierence that has sometimes been
termed the invention or inventive step. What matters is not necessarily
the degree of dierence, but whether or not the dierence is obvious to the
one ordinarily skilled in the art at the time the invention was made.
However, the degree of dierence can be a factor, particularly if it causes
disproportionate, unexpected, surprising, or unusual results. The dierence
from the prior art is examined for the inventions subject matter as a
whole. This means that the inherent properties of what is claimed and
disclosed in the patents specication are also part of the comparison to
the teachings of the prior art. If the invention produces certain advantages
or unexpected results and these might be crucial to a determination of
nonobviousness, they should be included in the patent specication.
The person of ordinary skill in the art is a hypothetical person and is
usually considered to be one who follows conventional wisdom in the art
rather than one who is an innovator in the art. The level of ordinary skill is
ascertained at the time the invention was made, not at a later time when
advances in the art could have made the original invention obvious, particularly if the advances were made possible by the claimed invention. Certain
factors may be considered in determining the level of ordinary skill, includ-

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

703

ing the education level of active workers in the art, the types of problems
encountered and the prior art solutions to those problems, the sophistication of the technology, and the rapidity with which innovations are made.
Objective evidence of nonobviousness usually comes as proof or evidence of what are called secondary considerations. The most common
type of secondary considerations include the commercial success of the
invention, the satisfaction of a long-felt need, the failure of others to nd
a solution to the problem the invention addresses, copying of the invention
by others, unexpected results, and expression of disbelief or praise by
experts. While the absence of such secondary considerations does not necessarily indicate that an invention is obvious, their presence in an invention is
not always conclusive of nonobviousness.

VI.

INVENTORSHIP AND PRIORITY OF INVENTION

Only the true and original inventor may obtain a U.S. patent. The inventor
cannot have derived the invention from someone else. If more than one
independent inventor applies for a patent on the same invention, it will be
awarded to the rst inventor. Although we speak of the inventor, it is more
correct to speak of the inventorship entity since there may be more than
one inventor. Where there is more than one inventor, the inventorship entity
is spoken of as joint inventors or co-inventors. Joint inventorship is quite
common in the pharmaceutical eld where several scientists collaborate on a
project. A patent can be invalidated if more or fewer than the correct number of co-inventors is named in an application, but the law presumes that the
listing of inventors in an issued patent is correct. Legal remedies are available to correct a mistake in the inventorship entity as long as it was made
without intent to deceive.
Each inventor owns an equal and undivided interest in the patent
rights and, absent an agreement to the contrary, can individually assign
or license their property rights to dierent parties. In most cases the joint
inventors work for the same organization and assign their patent rights to
their employer. However, if a project involves collaboration with scientists
outside of the organization and there is the possibility of joint inventorship,
the question of the ownership of property rights between the two organizations should to be addressed prior to applying for a patent.
A joint invention occurs when more than one person contributes
something meaningful to the conception of the claimed invention. The contribution must be more than a mere suggestion of a desired result or must do
more than merely follow the instructions of another. Someone preparing a
composition provided by his or her supervisor and submitting it for testing

Copyright 2003 Marcel Dekker, Inc.

704

Roehrs

does not rise to the level of an inventor. However, if in so doing he or she

makes a contribution, such as a modication in the composition that allows
it to achieve its intended purpose, then he or she could become a joint
inventor. According to U.S. patent law, joint inventors may obtain a patent
even though they did not physically work together or at the same time, each
did not make the same type or amount of contribution, or each did not
make a contribution to the subject matter of every claim of the patent.
United States patent law is a rst-to-invent system, as opposed to
other countries, which have rst-to-le systems. The priority of invention
is an issue when more than one inventorship entity submits a patent application for the same claimed subject matter. The priority contest (interference) is
invoked to determine who was the rst to invent. Invention is considered the
acts of conception and reduction to practice. The patent will be awarded to
the rst to conceive the invention even if last to reduce it to practice as long as
he or she was reasonably diligent in pursuing the reduction to practice from
the time prior to conception by the other inventorship entity. Conception is
considered the mental part of inventing. The inventor forms in his or her
mind the complete form of his or her part of the invention. In some cases,
conception may not be complete until after some experimentation has taken
place. Reduction to practice is the demonstration that the invention works
for its intended purpose, but the act of ling a patent application is recognized in patent law as a constructive reduction to practice. Scientists are
taught to keep records of their experimental work, and it is important to
record the conception of an invention in a timely manner and keep timely
records of the acts taken to reduce it to practice. The conception and reduction to practice records should be witnessed by someone other than another
co-inventor. The witness should have sucient knowledge or qualications
to understand the technical information regarding the invention records.

VII.
A.

PATENT APPLICATIONS
Provisional Application

Since 1995, a provisional patent application may be led with the USPTO
for the purpose of establishing priority of invention and the benet of an
earlier ling date but not starting the 20-year patent term. A patent does not
issue from ling a provisional application, and the application automatically
becomes abandoned after 1 year from ling. To obtain the benet of an
earlier ling date, a regular or nonprovisional patent application must be
led before the end of the 1-year period. The provisional application need
only contain a written description of the invention, drawings (if needed) and
the name of the inventor(s) along with some formal requirements and a fee.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

705

No claims are required in the application, but the invention must be fully
disclosed to support the claims of a subsequent regular application, otherwise the benet of the earlier ling date would only apply to the claims
supported by the disclosure in the provisional application.
B.

Nonprovisional Application

A regular application submitted to the USPTO to obtain a patent consists of

a number of parts prepared by a patent legal specialist with the aid of the
inventor(s). The following sections will focus on two major parts of the
application that are incorporated into the issued patent. These parts are
the patent specication and the claims.

VIII.

PATENT SPECIFICATION

The specication is the written description of the invention that concludes

with one or more claims that dene the exclusive rights granted to the
inventor. The specication is usually presented in several subparts:
1.
2.

Title of the Invention. A short and specic description of the

nature of the invention.
Cross-Reference to Related Applications, if any. These are prior
led applications of the same applicant and, if certain requirements are met, establish the earliest eective ling date for some
or all of the claimed subject matter. Several types that may
appear in the patent are:
a. Continuationfor the same invention but may be used to add
claims to subject matter disclosed but not claimed in the earlier application.
b. Continuation-in-partcontains some or all of the prior application plus new information disclosed (new matter) to support new claims.
c. Divisionalled for distinct or independent inventions than
cannot be claimed in the same application.
d. Provisionalpreviously discussed. For benet of earlier ling
date.
e. Foreign or international applicationEstablish priority (earlier) ling date.

3. Background of the Invention. Consists of two partsa brief statement of the eld of art to which subject matter pertains, and a

Copyright 2003 Marcel Dekker, Inc.

706

Roehrs

4.

5.

6.

7.

IX.

description of the relevant prior art including problems, limitations, disadvantages in the prior art and those that have been
solved or overcome by the applicants invention.
Summary of the Invention. A brief summary that presents the
substance or general idea of the invention and may include the
objectives of the invention and may point out advantages and
how it solves problems presented in the Background.
Brief Description of the Drawings. Drawings are not always
required unless they are needed for a full understanding of the
claimed subject matter. Drawings are more commonly used in
mechanical and electrical patents. However, graphical presentations of experimental data are also considered drawings and
sometimes are used in composition of matter patents.
Detailed Description. This section presents the invention
described in sucient detail to support the scope of the claimed
subject matter and distinguish it from other inventions and what
is old (prior art). It is also used to meet the patent law requirements to describe the method of making and using it in sucient
detail to allow a person skilled in the art to make and use the
invention without undue experimentation, and the best mode
(preferred embodiment) contemplated by the inventor for practicing the invention at the time of ling the application. If there are
drawings, each element should be described. For improvement
inventions, it points out specically the part or parts of the statutory subject matter such as a composition of matter that is the
improvement. It is understood in patent law that the inventor is
his own lexicographer and has latitude in choosing the terminology to describe and claim his invention, but there must be correspondence in terminology between the description, claims, and
drawings.
Abstract of the Invention. A brief statement of the technical disclosure in the patent so as to enable the patent oce and the
public to determine quickly from a cursory inspection the nature
and gist of the invention. The abstract appears on the front page
of the published patent.

PATENT CLAIMS

The patent concludes with one or more claims that dene the physical extent
or boundary of the exclusionary rights granted by the government to the
patent owner. The claims also serve notice on the public as to what acts will

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

707

violate the personal property rights (infringement) as well as the physical

embodiments of the new knowledge that will come into the public domain
upon expiration of the patents term.
The language of the claims follows certain prescribed rules. Each claim
is one sentence as the object of the introductory I (We) claim: and contains three parts: the preamble, the transition and the body. The preamble is
an introductory phrase that introduces the name of the subject matter and
may include the purpose, object, or intended use of the invention. The
preamble may also in some cases include elements of the prior art to segregate what is new or modied by the invention recited in the body of the
claim. The claimed invention in this Jepson-type claim consists of the preamble elements combined with the new or improved elements for operability. Ordinarily, the preamble does not limit the scope of the claims and most
patent practitioners avoid such limiting preambles. However, in some circumstances a preamble has been deemed by a court to be signicant to claim
scope where terms appearing in a preamble gives meaning to the claim and
properly denes the invention.
The second part of the claim, the transition, is the bridge between the
preamble and the body of the claim and plays a major role in the interpretation of the scope of the claim. Specic terminology for the transition has
been adopted in patent practice to indicate whether the body of the claim
should be read as open-ended or closed:
1. When the transition used is comprises (or comprising, including, containing), the claim is to be interpreted as open-ended,
i.e., the claim covers the elements (components, ingredients,
steps) recited in the body of the claim as well as additional elements. Thus the legal boundary is not limited only by the recited
elements in the claim. A competitor cannot merely add an additional element and avoid infringement. The term comprises can
be read as having at least.
2. The transitional phrase consisting of indicates a closed-ended
claim and the narrowest scope of coverage as the claim covers
only the recited elements. A competitor can add an additional
element and avoid literal infringement. This type of claim would
only be used if necessary for operability or to avoid a novelty or
obviousness rejection.
3. The transitional phrase consisting essentially of is intermediate
in scope and is intended to exclude additional unspecied elements that would aect the basis and novel characteristics of the
subject matter recited in the body of the claim.

Copyright 2003 Marcel Dekker, Inc.

708

Roehrs

The body of a claim recites the necessary elements required for the
invention to be operable and patentable over the prior art. Each element
usually includes a well-dened name distinguishing it from other elements,
distinctive features, or properties of the element necessary for operability
and patentability, such as concentration range, and nature of cooperation, if
any, with the other elements. The body of the claim can vary in scope as to
how the elements are claimed: the narrowest scope names one specic component (ingredient) per element; an intermediate scope names a group of
similar functionally equivalent components as the element; and the broadest
scope names only the function that the element performs. These variations
of course can be used only in light of their operability and patentability. The
use of alternative expressions (Markush groups) can be a dened genus such
as halogen or a homemade generic expression such as chlorine or bromine. The use of these groupings is considered closed-ended and is preceded by the use of consisting of. An example of a functional claim
element used in pharmaceutical compositions would be the element: a
pH-adjusting substance in an amount sucient to adjust the pH to a
value of from about 4 to 5.5. The specication would contain examples
describing a range of pH adjusting substances contemplated but not limiting
in the practice of the invention.
A patent usually includes multiple claims, some of which are independent and some dependent. The claims are numbered consecutively, and the
rst claim is an independent claim of the broadest scope of the claimed
invention, i.e., it contains the fewest limitations. It has the fewest elements
in the broadest language consistent with operability and patentability. The
dependent claims add an additional element or limitation and refers back to
another claim or claims that it so modies. If the invention covers a commercial product, there usually is a claim that is narrowed to recite all the
specic elements of the product.
There may be claims to more than one statutory class of subject matter. It is typical for drug delivery system patents to claim the delivery system
(vehicle) as a composition of matter, claim the vehicle plus a pharmacologically active drug, and also claim a method for using the delivery system to
administer the drug. For claims reciting a drug as an element of the composition of matter or method of use, it has become judicially acceptable to
refer to the concentration of the drug as a therapeutically eective
amount. For drug molecule patents, claims are usually included covering
the drug, the drug in dosage form(s) usually claimed broadly as pharmaceutically acceptable carriers, and methods for using the drug.
Composition of matter claims are preferred over method of use claims
since they would be in a dominant position to later patented new uses.
Also, the person infringing a method of use claim is usually a customer.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

X.

709

PATENT EXAMINATION AND PROSECUTION

Applications for U.S. patents are led with the USPTO and examined by a
patent examiner in the art class to which the invention pertains. The
examiner searches the prior art including U.S. patents, foreign patents,
printed publications as well as his or her own personal knowledge of the
art. The examiner is looking for references that show the identical invention or prior art that can be combined or modied to demonstrate that the
claimed subject matter would be obvious and thus unpatentable over the
prior art.
The patent application will most likely contain pertinent prior art
known to the inventor. In fact, each individual associated with the ling
and prosecution of a patent application has a duty to disclose to the
USPTO all information known to that individual to be material to
patentability with respect to each pending claim. This duty of disclosure
extends not only to the inventor(s) and the legal representative(s) but
also to every other person substantially involved in the preparation and
prosecution of the application and who is associated with the inventor,
with the assignee, or with anyone to whom there is an obligation to
assign. Individuals other than the inventor and legal representative can
comply by disclosing the information to the inventor or legal representative. The information may be disclosed in the patent specication or
in a separate Information Disclosure Statement (IDS). An IDS is not
construed as a representation that a search of the prior art was made or
that no better art exists. This duty to disclose is satised if all information known to be material to patentability of any claim issued in a
patent was cited by the patent oce or submitted to the patent oce.
The front page of an issued patent contains a listing of references cited
by the patent oce as to U.S. patents, foreign patents, and other publications. The front page of the patent also lists the classication of the
patents subject matter and the eld of search, i.e., the various subject
matter classications searched by the examiner.
The examiner will most likely issue an Oce Action (OA) that sets
forth any deciencies in the application and allow or reject some or all of the
claims. The examiner is required in the OA to set forth the basis and cite the
best art available and how it applies to the claims that are rejected. The
applicant can respond usually in the form of an amendment to traverse the
rejections. This give and take with the examiner is termed prosecution of
the application. The amendment may include an adavit or declaration to
remove a prior art reference used by the examiner to reject claims or to
provide data and information as objective evidence of invention. Often the
claims will be modied or narrowed to traverse a rejection and obtain an

Copyright 2003 Marcel Dekker, Inc.

710

Roehrs

allowance. The prosecution continues until claims are allowed or a nal

rejection is issued or the inventor abandons the application. In the case of
a nal rejection, the inventor may appeal within the patent oce and if
unsuccessful to the courts.
The prosecution of the application is conducted largely ex parte, i.e., in
condence between the USPTO and the inventor(s). The record of the prosecution, i.e., the examiners papers as well as the application and the inventors responses to the patent oce are kept in a folder that is termed a le
wrapper. This le wrapper does not come with the printed patent but may
be obtained separately from the patent oce. It is often useful to obtain to
determine what information the inventor provided that may not be a part of
the patent and what modications may have been made to the claims during
prosecution.
The rst public knowledge of a U.S. patent application, unless disclosed by the applicant or referred to in a published foreign patent application, had been the notication of an issued patent in the weekly Ocial
Gazette published by the USPTO. However, since March 15, 2001, and
subject to certain exceptions, regular nonprovisional utility patent applications led in the USPTO and also led in a foreign country are being
published 18 months after the eective U.S. ling date beginning with qualifying applications led on or after November 29, 2000, and those pending
on that date that request and qualify for voluntary publication. This change
in U.S. patent law brings U.S. practice in alignment with most European
countries and Japan that publish patent applications. To avoid publication
of a pending U.S. patent application, the applicant must specically request
that the application not be published and certify that no foreign counterpart
application will be led. Certain applications subject to a secrecy order will
not be published. The exception to publication of a patent application does
not exist in European countries and Japan.
Once an application is published, the application and the complete le
wrapper are publicly available. This provides the public with an early notice
of potentially patentable subject matter, although it is no guarantee that a
patent will be issued or the scope of the claims nally allowed. Also, it
allows members of the public such as competitors to monitor patent prosecution and to submit pertinent prior art for consideration by the patent
examiner. The benet to the inventor of having his or her pending patent
applications published is the additional grant of so-called provisional
rights to collect reasonable royalties for infringing acts prior to grant of
a patent as discussed in the next section.

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

XI.

711

PATENT INFRINGEMENT

The patent owners exclusive property rights may be violated (infringed) by

one or more of the acts of making, using, selling, oering to sell in the
United States, or importing the patented invention into the United States.
Some exceptions of interest to the pharmaceutical scientist are discussed
below. The act of making or using the subject matter of at least one claim
of a patent is sucient for infringement. There does not have to be commercialization or intent to commercialize to commit an infringing act, but
these are the usual reasons that a patent owner would sue for infringement.
What is infringed is one or more claims of the patent, i.e., an embodiment of
the invention and not the invention as a whole.
The question of whether an infringing act has been committed or
whether infringement can be avoided is answered by focusing on the patent
claims and their language in light of the patent specication, the prior art,
and the prosecution history (le wrapper) of the patent. Literal infringement
is determined by applying the claims to the suspect infringing product or
process. If each and every element (limitation) of a claim exists in the suspect
product or process, then there may be literal infringement. In patent jargon,
the claim reads on the suspect product or process. If there are more
elements in the suspect product or process and the claim is closed-ended
by the transitional phrase used, then there may not be literal infringement.
The apparently straightforward interpretation of infringement can be
complicated by the application of the judicially created Doctrine of
Equivalents. Courts have in some cases, where the accused infringing product or process does not contain a limitation exactly as stated in a claim,
extended the scope of the claim to interchangeable equivalents. Their reasoning is one of equity to the patent owner, who must nd language to
describe the various embodiments of the invention as best he or she can
while what is protected is the actual physical structure of the invention and
not merely the words used in the patent. This judicial doctrine is somewhat
controversial and can be legally complicated in its application. Courts
usually apply a tripartite test when applying the doctrine in a specic case
to determine if the element in issue performs substantially the same function
in substantially the same way to achieve substantially the same result as the
claimed element. The substantially equivalent element does not necessarily
have to be described in the patent but may be something that the hypothetical person skilled in the art would consider interchangeable. The range of
interchangeable equivalents may be determined from those available at the
time of the infringement and not just at the time of the invention.

Copyright 2003 Marcel Dekker, Inc.

712

Roehrs

When analyzing a patent for determining infringement or development

of a non-infringing product or process, the prosecution history in the patent
oce (le wrapper) should be examined. During prosecution of the patent
application, the claims may have been modied or certain arguments may
have been put forth to avoid a rejection of the claimed subject matter. In
subsequent enforcement of patent rights, the patent owner cannot try to
recapture subject matter relinquished during prosecution or change his position as to arguments made to avoid rejection. Subject matter relinquished
such as by adding further elements or limitations to a claim is not part of the
patent property and cannot be recaptured by the Doctrine of Equivalents.
With the publication of pending patent applications, provisional rights
are granted that enable the patent owner to collect reasonable royalties for
infringement during the period beginning on the date of publication of an
application until the date the patent issues. However, these provisional
rights are only available if the invention as claimed in the issued patent is
substantially similar to the invention as claimed in the published patent
application and proof is available that the alleged infringing party had
actual notice of the published patent application.
Several specic exceptions to the law of infringement have been
enacted by Congress. As discussed in Chapter 20, development and testing
to obtain data for submission to FDA for marketing approval is not an
act of infringement. This exception was made so that a drug patents
limited monopoly would not, in eect, be extended if development of a
generic product had to await patent expiration. The law makes the submission of the marketing application prior to patent expiration the infringement act.
Methods of diagnosis and treatment such as medical and surgical
procedures performed on a body can and are patented as they qualify
as statutory processes. Patent law was amended in 1996 (applies to patents
granted on or after September 30, 1996) to preclude any remedy for
infringement against a medical practitioner for his or her performance
of a medical or surgical procedure. However, the exempted medical activities do not include use of a patented machine, article of manufacture or
composition of matter in violation of the patent, or the practice of a
patented use of a composition of matter in violation of the patent, or
the practice of a process in violation of a biotechnology patent. Also,
the term patented use of a composition of matter does not include a
claim for a method of performing a medical or surgical procedure on a
body that recites the use of a composition of matter where the use of the
composition of matter does not directly contribute to achievement of the
objective of the claimed method. Thus, the exception is directed primarily
to pure medical and surgical procedures where it is in the public interest

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

713

not to limit the use of innovative procedures because of the health care
professionals concern for being sued for infringement.

XII.O

PATENT SEARCHING

In researching the patentability of an invention, it is important to know the

state of the prior art to which the invention pertains. Searching the prior art
includes patents that have issued as well as printed publications in the nonpatent prior art. While it is required to disclose to the USPTO material prior
art known at the time the application is led, it is not required by patent law
that a search of prior art be made. However, it is benecial to conduct as
thorough a search of prior art as time and resources will allow. There are
limitations on a patent search such that even the most thorough may not be
100% reliable in discovering all pertinent patented subject matter. Most
patent searches focus on U.S. patents but may miss pertinent foreign
patents, although the ability to search certain foreign patents using online
databases is improving. There may be patent applications that have not
been published, and patent databases that are searched may not be complete
or some patents may have been misclassied.
In addition to searching for patentability, patent searches are undertaken to ascertain validity of issued patents such as in defense of an infringement suit or to determine the right to use a prospective product (avoid
infringement). Patentability and validity searches focus on expired and
unexpired patents and published applications with a worldwide perspective
while a right-to-use search focuses on unexpired patents and published
applications in the United States and other countries of marketing interest.
Another type of patent search is a state-of-the-art or collection search in
which all patents and published applications in a particular area of technology are reviewed. The state-of-the-art search is particularly useful when
beginning research in a new product area and can aid in avoiding duplication of patented technology and focus eorts on improvements and new
approaches.
Patent searching is usually conducted manually or electronically and
increasingly by a combination of both methods using online patent databases. The availability of patent databases accessible through the Internet
make it possible for the inventor to conduct at least a preliminary search for
U.S. patents while in his or her lab or in the company or university library,
just as he or she would conduct a literature search. These databases make it
possible to keep up to date on new U.S. patents published on a weekly basis
and published patent applications as well as some international applications

Copyright 2003 Marcel Dekker, Inc.

714

Roehrs

and foreign patents. Specialized patent and research databases are being
developed in areas such as AIDS and biotechnology.
The traditional manual method involves determining the most likely
classication by the USPTO of the invention and then searching in these
classes and subclasses for pertinent patents. This search may be done at the
USPTO in Arlington, Virginia, where more than 6 million patents issued
since 1790 are led and can be manually examined. There is also a network
of more than 80 Patent and Trademark Depository Libraries (PTDL)
throughout the United States, most of which maintain a complete patent
image collection on microlm which is available for public patent searching.
A list of the locations of these PTDLs is available on the USPTO website at
www:uspto:gov=go=ptdl. These patent libraries have complete information
on U.S. patent classications including a computer search system known as
CASSIS on CD-ROMs from which a list of patents issued in each patent
classication can be obtained. The bibliographic data from the front page of
a patent can also be accessed from CASSIS for utility patents issued since
1969. The front page of each patent includes the title of the invention as well
as the patent abstract, which is often used to classify the subject matter of
the invention. The front page also contains the patent classication, both
U.S. and international, as well as references cited by the examiner. The
references include U.S. patents, foreign patent documents, and printed publications cited by the examiner as prior art. Some of the patent libraries also
have access to the APS (Automated Patent System) text search program.
The APS is connected by computer terminal to the patent database at the
USPTO and enables searching the entire text of patents issued since August
1971. Several of these libraries also have an enhanced version of APS that
allows viewing of patent images as well as text. Patent images are necessary
to view the drawings contained in an issued patent.
The basic method for using the U.S. patent classication system to
conduct a patent search involves compiling a list of classes and subclasses to
search. This is done by making lists of words (terms) describing the important structural and functional features of the invention and rst looking up
the terms in the Index to the U.S. Patent Classication Manual (Index). The
Index is an alphabetical listing of technical and common terms with corresponding U.S. patent classications used by the USPTO. The classes and
subclasses are numerical and written as, for example, 514/772, which signies Class 514 and its subclass 772. Second, once a list of potential classes
and subclasses has been compiled, the Manual of Classication (Manual) is
consulted, which lists all subclasses in an indented fashion under each main
class in numerical order. The Manual aids in narrowing the classication to
the most relevant subject matter. The Classication Denitions manual
should then be consulted to further determine the relevance of the chosen

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

715

classes and subclasses for searching as it contains a description of the subject

matter for each class and subclass used by the USPTO. It also contains a
cross-reference to other classication for similar subject matter and search
notes explaining the scope and limits of a particular class or subclass.
If an inventor already has one or more patents in the same subject area
as the prospective invention, he can begin classication searching using the
U.S. classication(s) and Field of Search information listed on the front
page of a patent. However, it is advisable to consult the USPTO classication documents referred to above, particularly for older patents, since the
classication numbers may have changed or new class/subclass combinations may have been established. If the new invention contains new or
dierent components, searching in additional classes and subclasses may
be required.
Once a list of the most relevant class/subclass combinations has been
obtained, the next step is to generate a list of patents issued in each class.
The patent lists can be obtained using the CASSIS system or from online
patent databases such as the USPTO. The lists contain the patent number
and the patent title. It is usually necessary to review more than just the
patent title to determine the most relevant patents that should be obtained
for further review. The online databases usually have a link to the full text of
the patent and a link to view the patent image (each page of the issued
patent). Another means to review each patent from the list that may be
relevant is the use of the Ocial Gazette that contains an exemplary claim
and a representative drawing (if any) for each issued patent.
In reviewing a number of ophthalmic drug delivery patents, U.S.
Classes 424 and 514 are frequently cited on the front page. The title of
both classes is identicalDrug, Bio-Aecting and Body Treating
Compositionsso these would be good places to begin a search.
However, we need to narrow the search to specic subclasses, and we nd
in the Manual that classication 424/78.04 is listed as ophthalmic preparations. Obtaining a list of patents in this classication indicates that of
January 2002 there are 238 U.S. patents cited. In class 514 in the Manual
we nd that subclass 912 contains ophthalmic subject matter involving the
treatment of an ophthalmic disorder, and a list of patents in classication
514/912 reveals 739 patents as of January 2002. Some of the ophthalmic
patents in both classications are obviously the same, but this example
points out that there is no single classication for all ophthalmic drug
delivery patents. For example, Class 427 is titled Coating Processes,
and a number of pharmaceutical patents are contained in this class including
ophthalmic drug and ocular devices. Also, Class 604, the Surgery classication, and subclass 890.1 include some patents directed to controlled release
therapeutic devices or systems, a number of which relate to ophthalmic

Copyright 2003 Marcel Dekker, Inc.

716

Roehrs

applications. Once relevant patent classications have been established, the

search can continually be updated by periodically obtaining new lists of
patents issued and published applications in these classications. New
patents are issued on a weekly basis, and these updates can easily be
obtained by online patent database, discussed in the next section.
Another use of the online patent databases is to monitor competitor
activity. Periodic updates of issued patents can be obtained for patent classications of research interest by searching in the current U.S. classication
eld. Monitoring can be enhanced by access to published foreign patent
applications and the U.S. applications that are published. If there are certain
companies and/or inventors that you want to monitor, this can be done by
searching for issued patents for the company in the assignee eld and the
inventor (full name) in the inventors eld. These elds are found on the
front page of a patent.
The USPTO maintains an online patent database that can be searched
without charge. The patent database contains full text of U.S. patents issued
since 1790 hyperlinked to their full-page images. The database can be
accessed from the Home page of the website at www:uspto:gov. The database can be searched in specic blocks of years or for all years (1790 to
present). For the years 17901976, the database can be searched only by
patent number and current U.S. classication. For patents issued from 1976
to present, keyword searches can be conducted on the full text. Advanced
Boolean searches can be done in selected patent elds as well as multiple
elds. Patent number searching is also available. Searching in specic class/
subclasses provides a list of issued patents with links to full-text and fullpage images. The access to the image of each page of a patent allows
examination of the drawings as well as the text. Special browser plug-ins
or applications may be necessary to view the page images. The database
contains a section that explains the contents and recent changes as well as
missing or withdrawn patent numbers. It has Help functions that assist in
most ecient use of the search functions and an Index to the site from which
additional information available on the site can be obtained, for example,
links to the three patent classication documents explained above. Patent
documents can be ordered online for delivery in various forms for a fee.
Since March 15, 2001, the USPTO has been publishing certain patent
applications, and these are available on their website in a separate database
from issued patents. The published application database can be searched in
essentially the same manner as the issued patent database. The published
application database contains full-text and full-page images.
The Delphion Intellectual Property Network (formerly IBM) at
www:delphion:com is a commercial online patent database. Due to a change
in their business model in 2001, searching free of charge is now limited to

Copyright 2003 Marcel Dekker, Inc.

Patent Considerations

717

U.S. granted patents for certain years, and the free search is only conducted
on the front page (bibliographic information) of the patent. More extensive
searching capabilities including published U.S. patent applications and certain foreign published applications and granted patents are available now
only by purchase of subscription packages. Copies of patent documents can
be ordered online and delivered in various forms for a fee. Many of the
searching features are the same as at the USPTO site, but Delphion and
other commercial feebased patent databases have an advantage because
they contain collections of certain foreign-published patent applications and
issued patents, and the searching can be done in certain cases across the
United States and international collections. The commercial sites may also
oer prior art searching capabilities using technical journals and magazines.
Commercial
online
patent
databases
include
Micropatent
(www:micropatent:com), LexPat (www:lexis  nexis:com), and QPAT
(www:qpat:com).

BIBLIOGRAPHY
The following publications were consulted in the preparation of the information presented in this chapter:
1. The Patent Act of 1952, as amended, codied in Title 35 United
States Code.
2. Title 37 Code of Federal Regulations. USPTO Rules and
Regulations.
3. Manual of Patent Examining Procedure, USPTO, 1997.
4. Patent Law Basics, Rosenberg, Peter D., West Group, Release
#9, July 2001.
5. The Law of Chemical and Pharmaceutical Invention,
Rosenstock, Jerome, Aspen Law and Business, Second Edition,
2001 Supplement.
6. Patent Practice Handbook, Canelias, Peter S., Aspen Law and
Business, 2001.

Copyright 2003 Marcel Dekker, Inc.