`
`
`
`ational DeSIg
`of Stable Protel
`Formulations
`Theoryand Practice
`
`
`
`
`
`Edited by
`John F. Carpenter
`'
`' and
`Mark C. Manning
`
`MAIA Exhibit 1011
`MAIA V. BRACCO
`
`,- IPR PETITION
`
`
`
`
`MAIA Exhibit 1011
`MAIA V. BRACCO
`IPR PETITION
`
`
`
`Rational Design of Stable
`Protein Formulations
`Theory and Practice
`
`
`
`
`
`
`Rational Design of Stable
`Protein Formulations
`Theory and Practice
`
`
`
`
`
`
`Pharmaceutical Biotechnology
`
`Series Editor: Ronald T. Borchardt
`The University of Kansas
`Lawrence, Kansas
`
`Recent volumes in this series:·
`
`Volume 7
`
`Volume 8
`
`PHYSICAL METHODS TO CHARACTERIZE
`PHARMACEUTICAL PROTEINS
`Edited by James N. Herron, Win Jiskoot,
`and Daan J. A. Crommelin
`
`MODELS FOR ASSESSING DRUG ABSORPTION
`AND METABOLISM
`Edited by Ronald T. Borchardt, Philip L. Smith,
`and Glynn Wilson
`
`Volume 9
`
`FORMULATION, CHARACTERIZATION, AND
`STABILITY OF PROTEIN DRUGS: Case Histories
`Edited by Rodney Pearlman and Y. John Wang
`
`Volume 10 PROTEIN DELIVERY: Physical Systems
`Edited by Lynda M. Sanders and R. Wayne Hendren
`
`Volume 11
`
`INTEGRATION OF PHARMACEUTICAL DISCOVERY AND
`DEVELOPMENT: Case Histories
`Edited by Ronald T. Borchardt, Roger M. Freidinger,
`Tomi K. Sawyer, and Philip L. Smith
`
`Volume 12 MEMBRANE TRANSPORTERS AS DRUG TARGETS
`Edited by Gordon L. Amidon and Wolfgang Sadee
`
`Volume 13 RATIONAL DESIGN OF STABLE PROTEIN
`FORMULATIONS: Theory and Practice
`Edited by John F. Carpenter and Mark C. Manning
`
`Volume 14 DEVELOPMENT AND MANUFACTURE OF PROTEIN
`PHARMACEUTICALS
`Edited by Steven L. Nail and Michael J. Akers
`
`A Chronological Listing of Volumes in this series appears at the back of this volume
`
`A Continuation Order Plan is available for this series. A continuation order will bring delivery of each
`new volume immediately upon publication. Volumes are billed only upon actual shipment. For further
`information please contact the publisher.
`
`
`
`
`
`
`Rational Design of Stable
`Protein Formulations
`Theory and Practice
`
`Edited by
`John F. Carpenter
`and
`Mark C. Manning
`University of Colorado Health Sciences Center
`'Denver, Colorado
`
`uwer Academic I Plenum Publishers
`w York, Boston, Dordrecht, London, Moscow
`
`~.:
`
`'1;'t
`
`
`
`
`
`
`Library of Congress Cataloging-in-Publication Data
`
`Rational design of stable protein formulations: theory and
`practice/edited by John F. Carpenter, Mark C. Manning.
`p.
`cm. -
`(Pharmaceutical biotechnology; v. 13)
`Includes bibliographical references and index.
`ISBN 0-306-46741-0
`1. Protein drugs-Stability. 2. Protein engineering. 3.
`I. Carpenter, John F.
`Drugs-Design.
`II. Manning, Mark C.
`Series.
`RS431.P75 R38 2002
`615' .19-dc21
`
`ill.
`
`2001057997
`
`ISBN: 0-306-46741-0
`
`© 2002 Kluwer Academic/.Plenum Publishers, New York
`233 Spring Street, New York, N.Y. 10013
`
`http://www.wkap.nl/
`
`109876543 21
`
`A C.I.P. record for this book is available from the Library of Congress
`
`All rights reserved
`
`No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form
`or by any means, electronic,_ mechanical, photocopying, microfilming, recording, or otherwise,
`without written permission from the Publisher
`
`Printed in the United States of America
`
`
`
`
`
`
`Contents
`
`Chapter 1
`
`Practical Approaches to Protein Formulation Development
`Byeong S. Chang and Susan Hershenson
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Preparation for Formulation Development ..................... .
`Resource Requirements for Formulation Development ... : ...... .
`Useful Information for Designing Formulations ............... .
`Preformulation Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Characterization of Protein Pharmaceuticals .................. .
`Accelerated Stability Studies ............................. .
`Development of Analytical Methods ....................... .
`Evaluation of the Significance of Problems .................. .
`Formulation Development ................................. .
`Formulation Options for Protein Pharmaceuticals ............... ·
`Typical Protein Stability Problems: Causes and Solutions ........ .
`Optimization of Formulation Variables ...................... .
`Necessary Studies for Formulation Development .............. .
`Strategies to Overcome Difficult Formulation Problems ......... .
`Fonnulation in Commercial Product Development ............... .
`Critical Formulation Decisions During Pharmaceutical
`·Development ........................................ .
`Formulation for Early Preclinical and Clinical Studies .......... .
`Commercial Formulation ................................ .
`Regulatory Issues in Formulation Development ............... .
`
`1
`3
`3
`4
`4
`5
`5
`6
`7
`10
`10
`13
`13
`15
`17
`18
`
`18
`19
`19
`20
`
`xiii
`
`I
`
`1,
`
`I
`:1
`
`
`
`
`
`
`xiv
`
`Contents
`
`. . . . . . . . . . . . . . . . . . . . . .
`Appendix: List of Regulatory Documents
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`22
`23
`
`Chapter 2
`
`Recombinant Production of Native Proteins from Escherichia coli
`Tsutomu Arakawa, Tiansheng Li, and Linda 0. Narhi
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Distribution of Expressed Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Cell Washing and Lysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Purification of Soluble, Folded Proteins . . . . . . . . . . . . . . . . . . . . . . . .
`Purification and Refolding of Soluble, Misfolded Proteins
`. . . . . . . . . .
`Purification and Refolding of Proteins from Inclusion Bodies . . . . . . . .
`Washing and Solubilization of Inclusion Bodies . . . . . . . . . . . . . . . .
`Purification of Expressed Proteins from Inclusion Bodies . . . . . . . . .
`Refolding Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Disulfide Bond Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Removal of Denaturant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Effects of Tag Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Effects of Excipients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Response Surface Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`High Pressure Disaggregation and Refolding
`. . . . . . . . . . . . . . . . . .
`Methods to Analyze Folded Structures . . . . . . . . . . . . . . . . . . . . . . . . . .
`Bioactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Binding to Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Dilsulfide Bond Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Spectroscopy
`Conformational Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Limited Proteolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`27
`28
`32
`34
`35
`36
`36
`36
`38
`41
`41
`44
`44
`47
`48
`48
`49
`49
`50
`50
`51
`51
`51
`
`Chapter 3
`
`Physical Stabilization of Proteins in Aqueous Solution
`Brent S. Kendrick, Tiansheng Li, and Byeong S. Chang
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Overview of Physical Stability
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`61
`62
`
`
`
`
`
`
`Contents
`
`Thermodynamic Control of Protein Stability ................. .
`Kinetic Control of Protein Stability ........................ .
`Interactions of Excipients with Proteins ....................... .
`Preferentially Excluded Cosolvents ........................ .
`Buffers/Salts ......................................... .
`Specific Binding of Ligands ............................. .
`Protein Self-Stabilization ................................ .
`Physical Factors Affecting Protein Stability .................... .
`Temperature .......................................... .
`Freeze-Thawing ...................................... .
`Agitation and Exposure to Denaturing Interfaces .............. .
`Pressure ..... · ....................................... .
`Conclusions ........................................... .
`Appendix: Derivation of the Wyman Linkage Function and
`Application to the Timasheff Preferential Exclusion
`Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`References
`
`xv
`
`62
`63
`65
`66
`67
`68
`69
`70
`70
`71
`71
`72
`73
`
`73
`78
`
`Chapter 4
`
`Effects of Conformation on the Chemical Stability of Pharmaceutically
`Relevant Polypeptides
`Jeffrey D. Meyer, Bert Ho, and Mark C. Manning
`
`Introduction ... ·. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Relationship Between Structure and Deamidation Rates . . . . . . . . . . . .
`Primary Structure Effects
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Secondary Structure Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Tertiary Structure Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Summary of Structure Effects on Deamidation . . . . . . . . . . . . . . . . .
`Role of Structure in Protein Oxidation . . . . . . . . . . . . . . . . . . . . . . . . .
`Types of Oxidation Processes
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Effects of Oxidation of Surface and Buried Methionines on
`Protein Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Limiting Solvent Accessibility of Residues . . . . . . . . . . . . . . . . . . . .
`Conformational Control of Oxidation in Aqueous Solution . . . . . . . .
`Structural Control of Oxidation in Lyophilized Products . . . . . . . . . .
`Summary of Structural Control of Oxidation
`. . . . . . . . . . . . . . . . . .
`Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`References . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`85
`86
`87
`89
`91
`92
`92
`93
`
`95
`96
`97 .
`99
`100
`101
`101
`
`
`
`
`
`
`xvi
`
`Chapter 5
`
`Contents
`
`Rational Design of Stable Lyophilized Protein Formulations:
`Theory and Practice
`John F. Carpenter, Beyong S. Chang, William Garzon-Rodriguez, and
`Theodore W. Randolph
`
`109
`111
`112
`
`113
`114
`114
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Minimal Criteria for a Successful Lyophilized Formulation
`. . . . . . . . .
`Inhibition of Lyophilization-Induced Protein Unfolding
`. . . . . . . . . .
`Storage at Temperatures Below Formulation Glass Transition
`Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`The Water Content is Relatively Low . . . . . . . . . . . . . . . . . . . . . . . .
`A Strong, Elegant Cake Structure is Obtained . . . . . . . . . . . . . . . . . .
`Steps Taken to Minimize Specific Routes of Protein Chemical
`Degradation ............................... : . . . . . . . .
`Rational Design of Stable Lyophiilized Formulations . . . . . . . . . . . . . .
`Choice of Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Specific Ligands/pH that Optimizes Thermodynamic Stability of
`Protein ................ , . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Trehalose or Sucrose to Inhibit Protein Unfolding and Provide
`Glassy Matrix ...... _. ;. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
`Bulldng Agent (e.g., Mannitol, Glycine or Hydroxyethyl Starch) . . . .
`126
`Nonionic Surfactant to Inhibit Aggregation. . . . . . . . . . . . . . . . . . . .
`127
`Acknowledgments
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`127
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`127
`
`116
`117
`118
`
`119
`
`Chapter 6
`
`Spray-Drying of Proteins
`Geoffrey Lee
`
`. . . . . . . . . . . . . . . . . . . . . . .
`Introduction: Why Spray-Dry a Protein?
`Developments in the Last 10 Years
`. . . . . . . . . . . . . . . . . . . . . . . . . . .
`The Practice of Spray-Drying Proteins . . . . . . . . . . . . . . . . . . . . . . . . .
`Type of Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Spray-Drying Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Influence of Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Pure Proteins
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Formulated Systems
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Use of Added Surface Active Substances . . . . . . . . . . . . . . . . . . . . .
`
`135
`136
`139
`139
`140
`147
`14 7
`149 ·
`151
`
`
`
`
`
`
`Contents
`
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Concluding Remarks
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`xvii
`
`156
`156
`
`Chapter 7
`
`Surfactant-Protein Interactions
`Theodore W Randolph and LaToya S. Jones
`
`Introduction .................................... ; . . . . . . .
`Proteins and Surfactants at Surfaces
`. . . . . . . . . . . . . . . . . . . . . . . . . .
`Protein-Surfactant Interactions irt Solution . . . . . . . . . . . . . . . . . . . . . .
`Surfactant Effects on Protein Assembly State . . . . . . . . . . . . . . . . . . . .
`Surfactant Effects-on Proteins During Freezing, Freeze-Drying
`and Reconstitution
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Enzymatic Degradation of Non-Ionic Surfactants . . . . . . . . . . . . . . . . .
`Recommendations for Protein Formulation
`. . . . . . . . . . . . . . . . . . . . .
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`159
`161
`166
`167
`
`169
`170
`170
`171
`
`Chapter 8
`
`·High Throughput Formulation: Strategies for Rapid Development of
`Stable Protein Products
`Rajiv Nayar and Mark C. Manning
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Overall Structure of the RTF Approach ....................... .
`Role of an Established Decision Tree for Formulation Design ....... .
`Constraints on a Pharmaceutically Acceptable Protein
`Formulation ....................................... .
`Proper Choice of Dosage Form ........................... .
`Preformulation Studies ................................. .
`Proper Choice of Excipients
`. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`Estimates of Resources Needed for Formulation Development .... .
`Use of Software and Databases to Assist in the RTF Process ....... .
`Essential Analytical Methods .............................. .
`Stability Protocols .................................... · · ·
`Unified Strategy for RTF ................................. .
`References . . . . . . . . . . . . . . . . . . . . . · · · · · · · · · · · · · · · · · · · · · · · ·
`
`hidex ................................................ .
`
`177
`179
`181
`
`182
`183
`185
`186
`188
`189
`191
`193
`194
`195
`
`199
`
`\\ ,,
`
`
`
`
`
`
`5
`
`Rational Design of Stable
`Lyophilized Protein Formulations:
`Theory and Practice
`
`John F. Carpenter1,2, Beyong S. Chanl,
`2
`William Garzon-Rodriguez1
`, and
`•
`Theodore W. Randolph1
`,4
`
`INTRODUCTION
`
`For ease of preparation and cost containment by the manufacturer, and ease of
`handling by the end user, an aqueous therapeutic protein formulation usually is
`preferred. However, with many proteins it is not possible-especially consider(cid:173)
`ing the time constraints for product development-to develop sufficiently stable
`aqueous formulations. Unacceptable de11aturation and aggregation can be induced
`readily by the numerous stresses to which a protein in aqueous solution is sensi(cid:173)
`tive; e.g., heating, agitation, freezing, pH changes, and exposure to interfaces or
`denaturants (Arakawa et al., 1993; Cleland et al., 1993; Brange, 2000; Bummer
`
`John F. Carpenter, William Garzon-Rodriguez, and Theodore W. Randdph
`• Center for Phar(cid:173)
`John F. Carpenter, William Garzon-Rodriguez, and Theodore W.
`maceutical Biotechnology.
`• Department of Pharmaceutical Sciences, School of Pharmacy, University of Colorado
`Randolph
`• Amgen, Inc., Thousand
`Beyong S. Chang
`Health Sciences Center, Denver, CO 80262.
`Theodore W. Ranqolph
`• Department of Chemical Engineering, Univer(cid:173)
`Oaks, CA 91320.
`sity of Colorado, Boulder, CO 80309.
`Rational Design of Stable Protein Formulations, edited by Carpenter and Manning. Kluwer Academic I Plenum
`Publishers, New York, 2002.
`
`109
`
`
`
`
`
`
`110
`
`John F. Carpenter et al.
`
`and Koppenol, 2000). Furthermore, even under conditions that thermodynami(cid:173)
`cally greatly favor the native state of proteins, aggregation can arise during
`months of storage in aqueous solution (e.g., Gu et al., 1991; Arakawa et al., 1993:
`Chen et al., 1994; Volkin and Middaugh, 1996; Chang et al., 1996a). In addition.
`several chemical degradation pathways (e.g., hydrolysis and deamidation) are
`mediated by water. In aqueous formulations, the rates of these and other (e.g ..
`oxidation) chemical degradation reactions can be unacceptably rapid on the time
`scale of storage (e.g., 18-24 months) for pharmaceutical products (Manning e1
`al., 1989; Cleland et al., 1993; Goolcharran et al., 2000; Bummer and Koppenol
`2000).
`In contrast, a properly lyophilized formulation can maintain adequate phys·
`ical and chemical stability of the protein during shipping and long-term storage
`even at ambient temperatures. As will be outlined in this chapter, developin§
`stable lyophilized protein formulations should be a rational, straightforwarc
`process, which for most proteins should be rapid. With liquid formulation devel·
`opment, it may only be possible to obtain adequate protein stability after length)
`studies. Furthermore, sometimes there are conflicting conditions (e.g., pH:
`needed to slow sufficiently multiple degradation pathways in aqueous solution
`Considering these issues plus the fact that formulation scientists now have to dea
`with numerous proteins and/or variants of a given protein, lyophilization shoulc
`be considered as a primary mode for product development. Only if a paralle
`effort to develop an aqueous formulation is successful, will a final lyophilizec
`product not be needed.
`Rapid formulation development has important financial ramifications. P
`drug product has a finite patent life, during which time the company has an exclu·
`sive market. Considering that even a moderately successful drug product ha~
`annual sales of hundreds of millions of dollars, potentially millions of dollars ir
`sales are lost for each day of delay in bringing a product to market. Unfortunately
`there are often delays because the formulations designed during early stage devel
`opment and clinical trials (e.g., frozen) were not adequate for the final product
`With a rational approach to formulation development, pharmaceutical scientist:
`and process engineers can minimize the risk of this problem and the time neede<
`to obtain a successful, final formulation. The success of such efforts depends or
`frank, open communication between the groups involved. For example, it is crit
`ical that the formulation scientists learn from the process engineers the issues fo:
`large-scale lyophilization tuns, which are usually conducted in units that do no
`have the capacity to match processing parameters obtained· in a small-seal<
`research lyophilizers.
`Despite the best efforts of the scientists and engineers, all too often delay:
`in formulation development arise because sufficient resources are not invested it
`product development. For example, sometimes, purchase of essential equipmen
`
`---------------- .
`
`
`
`
`
`
`II:
`I ,,
`
`Rational Design of Stable Lyophilized Protein Formulations
`
`111
`
`(e.g., a differential scanning calorimeter), which costs a minute fraction of a day's
`sale of product, is not allowed. To avoid unnecessary delays in product launch,
`which can have disastrous consequences for the company and for patients, it is
`essential that companies appreciate that product development is ultimately a key
`limiting factor in getting a therapeutic to market. Hence, development efforts need
`to be as well funded as the usually much more visible drug discovery research
`programs. If a product is not stable, it will not be marketed, no matter how dra(cid:173)
`matic an impact it can have on human health and the financial status of the
`company.
`
`MINIMAL CRITERIA FOR A SUCCESSFUL
`LYOPHILIZED FORMULATION
`
`Research over the past several years has demonstrated that five criteria
`define the minimal conditions necessary for a successful lyophilized protein for(cid:173)
`mulation (Table 1).
`The first four criteria can be met with use of the appropriate excipients and
`lyophilization cycle design. For information on the proper design of lyophiliza(cid:173)
`tion cycles, the reader is directed to the numerous previous reviews in this
`area (Franks, 1990; Pikal, 1990; Nail and Gatlin, 1993; Gatlin and Nail, 1994;
`Carpenter and Chang, 1996; Rey and May, 1999; Cappola, 2000). For the current
`chapter we will only consider cycle design in terms of the interplay between for(cid:173)
`mulation physical properties (e.g., collapse temperature) and process parameters
`(see below). The last criterion listed in Table 1 requires insight into the unique
`physicochemical properties of each therapeutic protein, which will be explained
`in more detail below. We will discuss in turn why each of these criteria is impor(cid:173)
`tant. Then we will present an explanation of how to design rationally a formula(cid:173)
`tion to meet these criteria.
`
`Table 1.
`Minimal Criteria for a Successful Lyophilized Protein Formulation
`
`1. Protein unfolding during freezing and drying is inhibited.
`2. The glass transition temperature of the product exceeds the planned storage temperature (e.g.,
`T, > 30°C).
`3. The water content is relatively low (e.g., 1% by mass).
`4. A strong, elegant cake structure is obtained (i.e., collapse and meltback are avoided).
`5. Steps are taken to minimize specific routes of protein chemical degradation (e.g., product vials
`are sealed under nitrogen to reduce the rate of methionine oxidation).
`
`
`
`
`
`
`112
`
`John F. Carpenter et al.
`
`Inhibition of Lyophilization-Induced Protein Unfolding
`
`The stresses of freezing and drying cause protein unfolding, and the fomm(cid:173)
`lation must be designed to inhibit unfolding at each step (Prestrelski et al.,
`1993a,b; Carpenter et al., 1993; Prestrelski et al., 1995; Constantino et al., 1995,
`1998; Griebenow and Klibanov, 1995; Allison et al., 1996, 1998, 1999, 2000;
`Chang et al., 1996b; Krielgaard et al., 1998a, 1999; Chen et al., 1999; Bell, 1999;
`Carrasquillo et al., 2000). Even if the formulation excipients and/or intrinsic ther(cid:173)
`modynamic stability of the protein prevent denaturation during freezing, unfold(cid:173)
`ing can arise during subsequent drying (Carpenter et al., 1993; Prestrelski et al.,
`1993b; Allison et al., 1998; Carrasquillo et al., 2000). Conversely, once a protein
`unfolds during freezing, it will not regain native structure during dehydration.
`For many proteins, unfolding during lyophilization leads to clinically
`unacceptable, non-native aggregates, even when samples are rehydrated immedi(cid:173)
`ately after lyophilization (Prestrelski et al., 1993a, 1995; Allison et al., 1996;
`Krielgaard et al., 1998a, 1999; Costantino et al., 1998). Aggregates are not nec(cid:173)
`essarily formed during freezing and drying. Rather, during rehydration refolding
`of structurally perturbed protein molecules competes with formation of non(cid:173)
`native protein aggregates (Prestrelski et al., 1993a). Aggregation can be mini(cid:173)
`mized by including stabilizing excipients (e.g., sucrose or trehalose) in the
`formulation to inhibit lyophilization-induced unfolding (Prestrelski et al., 1993a,
`1995; Allison et al., 1996; Krielgaard et al., 1998a, 1999; Costantino et al., 1998).
`Furthermore, fostering refolding during rehydration (e.g., with surfactants) can
`reduce aggregation (Chang et al., 1996c; Zhang et al., 1995, 1996).
`In addition to minimizing protein aggregation during lyophilization/rehy(cid:173)
`dration, maximizing retention of native protein structure in the dried solid is
`essential for optimizing long-term storage stability (Prestrelski et al., 1995; Chang
`et al., 1996b; Krielgaard et al., 1998a, 1999; Allison et al., 2000; Cleland et al.,
`2001). Both chemical and physical degradation in the dried solid can be acceler(cid:173)
`ated if protein unfolding is not inhibited during lyophilization. With chemical
`degradation, a non-native structure may provide an environment conducive to
`covalent modification of one or more residues. For example, exposure of a
`methionine, which is normally buried deep in the interior of the native protein,
`on the surface of an unfolded dried protein may foster oxidation. Increased levels
`of aggregates noted after storage and rehydration of unfolded proteins could be
`due to formation of non-native intermolecular contacts within the dried solid, per(cid:173)
`turbation of refolding during rehydration because of chemical degradation, and/ or
`other undefined processes.
`Infrared spectroscopy has been used routinely to compare the secondary
`structures of a protein in lyophilized formulations to that of the native protein in
`aqueous solution (Prestrelski et al., 1993a,b; Prestrelski et al., 1995; Dong et al.,
`
`
`
`
`
`
`Rational Design of Stable Lyophilized Protein Formulations
`
`113
`
`1995; Constantino et al., 1995, 1998; Griebenow and Klibanov, 1995; Allison et
`al., 1996, 1998, 1999, 2000; Chang et al., 1996b; Krielgaard et al., 1998a, 1999;
`Carpenter et al., 1998; Chen et al., 1999; Carrasquillo et al., 2000). This method
`should be considered essential in the development of stable lyophilized formula(cid:173)
`tions, because it allows one to assess rapidly the effectiveness of formulations at
`inhibiting protein unfolding. Technical details about how to employ infrared spec(cid:173)
`troscopy to design stable lyophilized protein formulations can be found in the
`papers cited above.
`
`Storage at Temperatures Below Formulation Glass .
`Transition Temperature
`
`In the dried powder, the protein is a component of an amorphous phase that
`includes amorphous excipients and water. If this glassy matrix is held below its
`characteristic glass transition temperature (T 8), the rate of diffusion-controlled
`reactions, including protein unfolding/aggregation and chemical degradation, are
`greatly reduced, relative to rates noted at temperatures >T8 (Roy et al., 1991;
`Franks, 1990; Franks et al., 1991; Pikal, 1994, 1999). T8 can be determined with
`differential scanning calorimetry (DSC) or other thermal scanning methods (Nail
`and Gatlin, 1993; Chang and Randall, 1992; Craig and Royall, 1998; Verdonck
`et al., 1999).
`Obtaining a formulation T8 in excess of the planned storage temperature
`(e.g., room temperature) is absolutely essential for optimal protein stability (e.g.,
`Franks et al., 1991; Pikal, 1994, 1999; Carpenter and Chang, 1996; Duddu and
`Dal Monte, 1997). The T8 of a given amorphous phase is dependent on the T8
`and mass percent of each component, including water (Angell, 1995; Franks
`et al., 1991; Levine and Slade, 1988, 1992; Pilcal, 1994, 1999). Compared to
`excipients, dried proteins have relatively high Tg's (e.g., >150°C; Angell, 1995).
`Thus, with all other factors being held constant, the formulation Tg varies directly
`with the mass fraction of protein. However, care must be taken that the mass frac(cid:173)
`tion of protein is not so high that there are not adequate levels of stabilizing excip(cid:173)
`ients to prevent protein unfolding during lyophilization (Cleland et al., 2001; and
`see below).
`Fortunately, sucrose and trehalose, which are the preferred excipients for
`inhibiting lyophilization-induced protein unfolding (see below), also provide a
`glassy matrix with acceptably high Tg values. For example, with water contents
`of 1% the T g for pure sucrose and trehalose are about 100 and 65°C, respectively
`(Crowe et al., 1998).
`It has now been documented with several proteins, that simply storing the
`formulation at temperature below Tg alone does not assure optimal stability. A
`
`
`
`
`
`
`114
`
`John F. Carpenter et al.
`
`native protein structure is also required. For example, proteins lyophilized in
`dextran alone are usually unfolded, but in a glassy matrix with a relatively high
`T 8 (e.g.,> 75° C). Yet they still degrade at relatively rapid rates compared to those
`for native protein molecules lyophilized with either sucrose or trehalose (Kriel(cid:173)
`gaard et al., 1998a, 1999; Lueckel et al., 1998; Allison et al., 2000; Yoshioka et
`al., 2000). On pharmaceutical time scales of several months of storage many
`degradative reactions are not coupled to the glass transition of a formulation. This
`is because on these times scales there is still significant molecular mobility, even
`at temperatures well below (e.g., more than 30° C) the T8 (Hancock et al., 1995;
`Duddu et al., 1997; Pikal, 1999; Yoshioka et al., 1999).
`
`The Water Content is Relatively Low
`
`Because of its very low T8 ( -135°C), wateris a potent plasticizerfor glasses;
`increasing water content in the dried formulation will greatly reduce T 8 • For
`example, increasing the water content of pure sucrose from 1 to about 3-4% (g
`H20/100 g dried powder) is sufficient to reduce the T 8 to below room temperature
`(Crowe et al., 1998). It is critical to achieve a suffiCiently low water level for a
`given formulation such that T8 exceeds the planned storage temperature. The
`lyophilization cycle dictates the initial water content (see reviews listed above).
`The most important parameter is the temperature for secondary drying, when the
`unfrozen water is desorbed (Pikal et al., 1990).
`Water can also be transferred to the product from the vial stoppers during
`storage (Pikal and Shah, 1992; DeGrazio and Flynn, 1992; Hora and Wolfe,
`1999). This effect can be dramatic. For example, let's consider a formulation con(cid:173)
`taining 10 mg of dried protein, 40 mg of sucrose and initial water content of 1%
`by weight. The total amorphous fraction containing protein and sucrose has 0.50
`mg of water. If l.Omg of water was transferred from the stopper to the product,
`the water content of this fraction would increase from 1% to 3.0%. This increase
`would be sufficient to lower the formulation T8 to below room temperature
`(Crowe et al., 1998). The risk of transfer of moisture from stoppers can be min(cid:173)
`imized by drying the stoppers before use, and, if acceptable for a given product,
`using stoppers coated with a material such as Teflon (see Hora and Wolfe, 1999).
`
`A Strong, Elegant Cake Structure is Obtained
`
`Often the most desired cake has strong, porous structure, without macro(cid:173)
`scopic collapse or meltback. This structure has a high surface area to volume
`
`
`
`
`
`
`Rational Design of Stable Lyophilized Protein Formulations
`
`115
`
`ratio, which aids in the rapid dissolution of product upon addition of water. A
`detailed account of how to obtain such a cake structure is beyond the scope of
`the current chapter, but is available in several previous reviews (see above). For
`the current purposes it is sufficient to focus on the impact of formulation com(cid:173)
`position on avoiding collapse or meltback. When a product is frozen, the protein
`and amorphous excipients (e.g., sucrose) are dispersed between ice crystals. and
`any excipient used as a crystalline bulking agent (e.g., glycine). To obtain an
`appropriate cake structure during lyophilization, the product temperature during
`primary drying, when the water in ice is sublimed, must be below the character(cid:173)
`istic collapse and eutectic melting temperatures of amorphous and crystalline
`solutes, respectively. Above the eutectic temperature, the melting of crystalline
`solutes leads to massive loss of porous structure and macroscopic dissolution of
`the frozen matrix into