Pharmaceutical quality by design a

practical approach First Edition Gibson

Visit to download the full and correct content document:
https://textbookfull.com/product/pharmaceutical-quality-by-design-a-practical-approac
h-first-edition-gibson/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Handbook of Pharmaceutical Wet Granulation Theory and

Practice in a Quality by Design Paradigm 1st Edition
Ajit S. Narang

https://textbookfull.com/product/handbook-of-pharmaceutical-wet-
granulation-theory-and-practice-in-a-quality-by-design-
paradigm-1st-edition-ajit-s-narang/

Pharmaceutical Drug Product Development and Process

Optimization-Effective Use of Quality by Design 1st
Edition Sarwar Beg (Editor)

https://textbookfull.com/product/pharmaceutical-drug-product-
development-and-process-optimization-effective-use-of-quality-by-
design-1st-edition-sarwar-beg-editor/

Pharmaceutical Preformulation and Formulation A

Practical Guide from Candidate Drug Selection to
Commercial Dosage Form Mark Gibson

https://textbookfull.com/product/pharmaceutical-preformulation-
and-formulation-a-practical-guide-from-candidate-drug-selection-
to-commercial-dosage-form-mark-gibson/

Quality assurance and quality control in the analytical

chemical laboratory : a practical approach Second
Edition Konieczka

https://textbookfull.com/product/quality-assurance-and-quality-
control-in-the-analytical-chemical-laboratory-a-practical-
approach-second-edition-konieczka/
Quality by Experimental Design Fourth Edition Thomas B.
Barker

https://textbookfull.com/product/quality-by-experimental-design-
fourth-edition-thomas-b-barker/

Reinforced Concrete Design A Practical Approach Chapter

12 3rd Edition Svetlana Brzev

https://textbookfull.com/product/reinforced-concrete-design-a-
practical-approach-chapter-12-3rd-edition-svetlana-brzev/

JavaScript Application Design A Build First Approach

1st Edition Nicolas Bevacqua

https://textbookfull.com/product/javascript-application-design-a-
build-first-approach-1st-edition-nicolas-bevacqua/

Water Distribution System Monitoring: A Practical

Approach for Evaluating Drinking Water Quality Second
Edition Abigail F. Cantor

https://textbookfull.com/product/water-distribution-system-
monitoring-a-practical-approach-for-evaluating-drinking-water-
quality-second-edition-abigail-f-cantor/

Statistics for Process Control Engineers A Practical

Approach First Edition Myke King

https://textbookfull.com/product/statistics-for-process-control-
engineers-a-practical-approach-first-edition-myke-king/
Pharmaceutical Quality by Design:
A Practical Approach
ADVANCES IN PHARMACEUTICAL TECHNOLOGY
A Wiley Book Series

Series Editors:
Dennis Douroumis, University of Greenwich, UK
Alfred Fahr, Friedrich–Schiller University of Jena, Germany
Jürgen Siepmann, University of Lille, France
Martin Snowden, University of Greenwich, UK
Vladimir Torchilin, Northeastern University, USA

Titles in the Series

Hot‐Melt Extrusion: Pharmaceutical Applications
Edited by Dionysios Douroumis
Drug Delivery Strategies for Poorly Water‐Soluble Drugs
Edited by Dionysios Douroumis and Alfred Fahr
Computational Pharmaceutics: Application of Molecular Modeling in Drug Delivery
Edited by Defang Ouyang and Sean C. Smith
Pulmonary Drug Delivery: Advances and Challenges
Edited by Ali Nokhodchi and Gary P. Martin
Novel Delivery Systems for Transdermal and Intradermal Drug Delivery
Edited by Ryan Donnelly and Raj Singh
Drug Delivery Systems for Tuberculosis Prevention and Treatment
Edited by Anthony J. Hickey
Continuous Manufacturing of Pharmaceuticals
Edited by Peter Kleinebudde, Johannes Khinast, and Jukka Rantanen
Pharmaceutical Quality by Design: A Practical Approach
Edited by Walkiria S. Schlindwein and Mark Gibson

Forthcoming Titles:
In Vitro Drug Release Testing of Special Dosage Forms
Edited by Nikoletta Fotaki and Sandra Klein
Characterization of Micro‐ and Nanosystems
Edited by Leena Peltonen
Therapeutic Dressings and Wound Healing Applications
Edited by Joshua Boateng
Process Analytics for Pharmaceuticals
Edited by Thomas de Beer, Jukka Rantanen and Clare Strachan
Pharmaceutical Quality
by Design
A Practical Approach

Edited by

WALKIRIA S. SCHLINDWEIN
De Montfort University
Leicester
United Kingdom
MARK GIBSON
A M PharmaServices Ltd
United Kingdom
This edition first published 2018
© 2018 John Wiley & Sons Ltd

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by
any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain
permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Walkiria S. Schlindwein and Mark Gibson to be identified as the authors of the editorial material in this work has
been asserted in accordance with law.

Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office
111 River Street, Hoboken, NJ 07030, USA
350 Main Street, Malden, MA 02148‐5020, USA
9600 Garsington Road, Oxford, OX4 2DQ, UK
The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Boschstr. 12, 69469 Weinheim, Germany
1 Fusionopolis Walk, #07‐01 Solaris South Tower, Singapore 138628

For details of our global editorial offices, customer services, and more information about Wiley products, visit us at
www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard
print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of
information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the
information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among
other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher
and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the
accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any
implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales
representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or
product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher
and authors endorse the information or services the organization, website, or product may provide or recommendations it may
make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice
and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate.
Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was
written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial
damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Schlindwein, Walkiria S., 1961– editor. | Gibson, Mark, 1957– editor.
Title: Pharmaceutical quality by design : a practical approach / edited by Dr. Walkiria S.
Schlindwein, Mark Gibson.
Description: First edition. | Hoboken, NJ : John Wiley & Sons, 2018. | Series: Advances in
pharmaceutical technology | Includes bibliographical references and index. |
Identifiers: LCCN 2017030338 (print) | LCCN 2017043153 (ebook) | ISBN 9781118895221 (pdf) |
ISBN 9781118895214 (epub) | ISBN 9781118895207 (cloth)
Subjects: LCSH: Drugs–Design. | Drugs–Quality control.
Classification: LCC RS420 (ebook) | LCC RS420 .P47 2018 (print) | DDC 615.1/9–dc23
LC record available at https://lccn.loc.gov/2017030338
Cover design by Wiley
Cover image: (Background) © ShutterWorx/Gettyimages;(Graph) Courtesy of Walkiria S. Schlindwein and Mark Gibson

Set in 10/12pt Times by SPi Global, Pondicherry, India

10 9 8 7 6 5 4 3 2 1
 Contents

List of Figures xiii

List of Tables xix
List of Contributors xxi
Series Preface xxiii
Preface xxv

1 Introduction to Quality by Design (QbD) 1

Bruce Davis and Walkiria S. Schlindwein
1.1 Introduction 1
1.2 Background 2
1.3 Science‐ and Risk‐Based Approaches 4
1.4 ICH Q8–Q12 5
1.5 QbD Terminology 6
1.6 QbD Framework 7
1.7 QbD Application and Benefits 7
1.8 Regulatory Aspects 8
1.9 Summary 9
1.10 References 9

2 Quality Risk Management (QRM) 11

Noel Baker
2.1 Introduction 11
2.2 Overview of ICH Q9 13
2.2.1 Start QRM Process 15
2.2.2 Risk Assessment 15
2.2.3 Risk Control 16
2.2.4 Risk Review 16
2.3 Risk Management Tools 17
vi Contents

2.4 Practical Examples of Use for QbD 22

2.4.1 Case Study 26
2.4.2 Pre‐work 26
2.4.3 Scoring Meeting 32
2.4.4 FMECA Tool 32
2.4.5 Risk Score 32
2.4.6 Detectability Score 34
2.4.7 Communication 35
2.5 Concluding Remarks 36
2.6 References 44

3 Quality Systems and Knowledge Management 47

Siegfried Schmitt
3.1 Introduction to Pharmaceutical Quality System 47
3.1.1 Knowledge Management – What Is It and
Why Do We Need It? 47
3.2 The Regulatory Framework 48
3.2.1 Knowledge Management in the Context of Quality by
Design (QbD) 48
3.2.2 Roles and Responsibilities for Quality System 49
3.2.3 Roles and Responsibilities for Knowledge Management 50
3.2.4 Implicit and Explicit Knowledge 50
3.3 The Documentation Challenge 51
3.4 From Data to Knowledge: An Example 56
3.5 Data Integrity 58
3.6 Quality Systems and Knowledge Management: Common
Factors for Success 58
3.7 Summary 59
3.8 References 60

4 Quality by Design (QbD) and the Development and Manufacture

of Drug Substance 61
Gerry Steele
4.1 Introduction 61
4.2 ICH Q11 and Drug Substance Quality 62
4.2.1 Enhanced Approach 63
4.2.2 Impurities 63
4.2.3 Physical Properties of Drug Substance 64
4.3 Linear and Convergent Synthetic Chemistry Routes 65
4.4 Registered Starting Materials (RSMs) 67
4.5 Definition of an Appropriate Manufacturing Process 68
4.5.1 Crystallization, Isolation and Drying of APIs 68
4.5.2 Types of Crystallization 69
4.5.3 Design of Robust Cooling Crystallization 70
4.6 In‐Line Process Analytical Technology and Crystallization Processes 78
4.6.1 Other Unit Operations 80
 Contents vii

4.7 Applying the QbD Process 82

4.7.1 Quality Risk Assessment (QRA) 83
4.8 Design of Experiments (DoE) 87
4.9 Critical Process Parameters (CPPs) 88
4.10 Design Space 88
4.11 Control Strategy 89
4.12 References 91

5 The Role of Excipients in Quality by Design (QbD) 97

Brian Carlin
5.1 Introduction 97
5.2 Quality of Design (QbD) 98
5.3 Design of Experiments (DoE) 100
5.4 Excipient Complexity 102
5.5 Composition 105
5.6 Drivers of Functionality or Performance 105
5.7 Limited Utility of Pharmacopoeial Attributes 106
5.8 Other Unspecified Attributes 107
5.9 Variability 107
5.10 Criticalities or Latent Conditions in the Finished Product 108
5.11 Direct or Indirect Impact of Excipient Variability 110
5.12 Control Strategy 111
5.13 Communication with Suppliers 112
5.14 Build in Compensatory Flexibility 113
5.15 Risk Assessment 113
5.16 Contingencies 114
5.17 References 114

6 Development and Manufacture of Drug Product 117

Mark Gibson, Alan Carmody, and Roger Weaver
6.1 Introduction 117
6.2 Applying QbD to Pharmaceutical Drug Product
Development119
6.3 Product Design Intent and the Target Product Profile (TPP) 120
6.4 The Quality Target Product Profile (QTPP) 126
6.5 Identifying the Critical Quality Attributes (CQAs) 128
6.6 Product Design and Identifying the Critical Material
Attributes (CMAs)133
6.7 Process Design and Identifying the Critical Process
Parameters (CPPs) 136
6.8 Product and Process Optimisation 139
6.9 Design Space 145
6.10 Control Strategy 150
6.11 Continuous Improvement 153
6.12 Acknowledgements 154
6.13 References 154
viii Contents

7 Design of Experiments 157

Martin Owen and Ian Cox
7.1 Introduction 157
7.2 Experimental Design in Action 158
7.3 The Curse of Variation 158
7.3.1 Signal‐to‐Noise Ratio 159
7.4 Fitting a Model 161
7.4.1 Summary of Fit 165
7.5 Parameter Estimates 165
7.6 Analysis of Variance 166
7.6.1 Reflection 168
7.7 ‘To Boldly Go’ – An Introduction to Managing Resource
Constraints using DoE 169
7.8 The Motivation for DoE 170
7.8.1 How Does the Workshop Exercise Work? 171
7.8.2 DoE Saves the Day! 172
7.9 Classical Designs 173
7.9.1 How Do Resource Constraints Impact the
Design Choice? 173
7.9.2 Resource Implications in Practice 173
7.10 Practical Workshop Design 174
7.10.1 Choice of Factors and Measurements 175
7.10.2 Data Collection and Choice of Design 175
7.10.3 Some Simple Data Visualization 175
7.10.4 Analysis of the Half Fraction 177
7.10.5 How to Interpret Prediction Profiles 177
7.10.6 Half Fraction and Alternate Half Fraction 178
7.10.7 Interaction Effects 178
7.10.8 Full Factorial 181
7.10.9 Central Composite Design 181
7.10.10 How Robust Is This DoE to Unexplained Variation? 181
7.11 How Does This Work? The Underpinning of Statistical
Models for Variation 184
7.12 DoE and Cycles of Learning 187
7.13 Sequential Classical Designs and Definitive Screening Designs 189
7.14 Building a Simulation 190
7.14.1 Sequential design, Part 1: Screening Design (10 Runs) 191
7.14.2 Sequential Design, Part II: Optimization Design (24 Runs) 191
7.14.3 Definitive Screening Design 194
7.14.4 Robustness Design 194
7.14.5 Additional Challenges 197
7.15 Conclusion 197
7.16 Acknowledgements 198
7.17 References 198
 Contents ix

8 Multivariate Data Analysis (MVDA) 201

Claire Beckett, Lennart Eriksson, Erik Johansson, and Conny Wikström
8.1 Introduction 201
8.2 Principal Component Analysis (PCA) 202
8.3 PCA Case Study: Raw Material Characterization using
Particle Size Distribution Curves 204
8.3.1 Dataset Description 204
8.3.2 Fitting a PCA Model to the 45 Training Set Batches 205
8.3.3 Classification of the 13 Test Set Batches 206
8.3.4 Added Value from DoE to Select Spanning Batches 208
8.4 Partial Least Squares Projections to Latent Structures (PLS) 208
8.5 PLS Case Study: A Process Optimization Model 210
8.5.1 Dataset Description 210
8.5.2 PLS Modeling of 85‐Samples SOVRING Subset 211
8.5.3 Looking into Cause‐and‐Effect Relationships 212
8.5.4 Making a SweetSpot Plot to Summarize the PLS Results 213
8.5.5 Using the PLS‐DoE Model as a Basis to Define a
Design Space and PARs for the SOVRING Process 215
8.5.6 Summary of SOVRING Application 217
8.6 Orthogonal PLS (OPLS® Multivariate Software) 217
8.7 Orthogonal PLS (OPLS® Multivariate Software) Case Study – Batch
Evolution Modeling of a Chemical Batch Reaction 218
8.7.1 Dataset Description 218
8.7.2 Batch Evolution Modeling 218
8.8 Discussion 220
8.8.1 The PAT Initiative 220
8.8.2 What Are the Benefits of Using DoE? 221
8.8.3 QbD and Design Space 222
8.8.4 MVDA/DoE Is Needed to Accomplish PAT/QbD in Pharma 223
8.8.5 MVDA: A Way to Power up the CPV Application 223
8.9 References 224

9 Process Analytical Technology (PAT) 227

Line Lundsberg‐Nielsen,Walkiria S. Schlindwein, and Andreas Berghaus
9.1 Introduction 227
9.2 How PAT Enables Quality by Design (QbD) 229
9.3 The PAT Toolbox 229
9.4 Process Sensors and Process Analysers 229
9.4.1 Process Sensors – Univariate 233
9.4.2 Process Analysers – Multivariate 233
9.4.3 Infrared (IR) 233
9.4.4 Near Infrared (NIR) 238
9.4.5 Tunable Diode Laser Spectroscopy (TDLS) 239
9.4.6 Ultraviolet‐Visible (UV‐Vis) 239
x Contents

9.4.7 Raman 239

9.4.8 Focused Beam Reflectance Measurements (FBRM)
and Laser Diffraction 239
9.4.9 Particle Vision and Measurement (PVM) 239
9.4.10 X‐Ray Fluorescence (XRF) 240
9.4.11 Imaging Technologies 240
9.5 Analyser Selection 240
9.6 Regulatory Requirements Related to PAT Applications 240
9.6.1 Europe 242
9.6.2 United States 242
9.7 PAT Used in Development 242
9.8 PAT Used in Manufacturing 243
9.9 PAT and Real Time Release Testing (RTRT) 245
9.10 PAT Implementation 245
9.11 Data Management 246
9.12 In‐Line Process Monitoring with UV‐Vis Spectroscopy: Case
Study Example 247
9.13 References 253

10 Analytical Method Design, Development, and Lifecycle Management 257

Joe de Sousa, David Holt, and Paul A. Butterworth
10.1 Introduction 257
10.2 Comparison of the Traditional Approach and the Enhanced
QbD Approach 258
10.3 Details of the Enhanced QbD Approach 260
10.4 Defining Method Requirements 262
10.5 Designing and Developing the Method 264
10.6 Understanding the Impact of Method Parameters on Performance 266
10.7 Defining the Method Control Strategy and Validating the Method 267
10.8 Monitoring Routine Method Performance for Continual
Improvement268
10.9 Summary 269
10.10 Example Case Studies 270
10.10.1 Case Study 1 – Establishment of Robust Operating
Ranges during Routine Method Use and Justifying the
Method Control Strategy (Including SST Criteria) 270
10.10.2 Risk Assessment and Definition of Ranges 270
10.10.3 Experimental Design 271
10.10.4 Evaluate the DoE 272
10.10.5 Documenting Method Performance 274
10.10.6 Case Study 2 – Evaluation of the Ruggedness of a
Dissolution Method for a Commercial Immediate Release
Tablet Product 274
10.10.7 Case Study Acknowledgements 278
10.11 References 278
 Contents xi

11 Manufacturing and Process Controls 281

Mark Gibson
11.1 Introduction to Manufacturing and Facilities 281
11.2 Validation of Facilities and Equipment 282
11.2.1 The International Society for Pharmaceutical Engineering
(ISPE) Baseline® Guide: Commissioning and Qualification 282
11.2.2 ASTM E2500‐07: Standard Guide for Specification, Design,
and Verification of Pharmaceutical and Biopharmaceutical
Manufacturing Systems and Equipment 284
11.2.3 Science‐Based Approach and Critical Aspects 285
11.2.4 Risk‐Based Approach 286
11.2.5 System and Component Impact Assessments 288
11.2.6 URSs for Systems 290
11.2.7 Specification and Design 290
11.2.8 Verification 290
11.3 Drug Product Process Validation: A Lifecycle Approach 292
11.3.1 Stage 1: Process Design/Product Development 295
11.3.2 Stage 2: Process Qualification 298
11.3.3 Stage 3: Continued Process Verification 299
11.4 The Impact of QbD on Process Equipment Design and
Pharmaceutical Manufacturing Processes 300
11.5 Introduction to Process Control in Pharmaceutical Manufacturing 302
11.6 Advanced Process Controls (APC) and Control Strategy 305
11.7 The Establishment of Continuous Manufacture 309
11.8 The Tablet Press as Part of a Continuous Tableting Line 312
11.9 Real‐Time Release Testing and Continuous Quality Verification 316
11.10 Acknowledgments 317
11.11 References 317

12 Regulatory Guidance 321

Siegfried Schmitt and Mustafa A. Zaman
12.1 Introduction 321
12.2 The Common Technical Document (CTD) Format 322
12.2.1 Quality Target Product Profile (QTPP) and Critical Quality
Attributes (CQAs) 324
12.2.2 Quality Risk Management (ICH Q9) 324
12.2.3 Product and Process Development (S.2.6 and P.2) 325
12.2.4 Control Strategy 326
12.2.5 Design Space (Optional) 327
12.3 Essential Reading 328
12.4 What Is Not Written, or Hidden, in the Guidance Documents? 329
12.5 Post‐Approval Change 330
12.6 Summary 331
12.7 References 332

Index 335
 List of Figures

Figure 1.1 A framework for QbD. 7

Figure 2.1 Medicinal product development flowchart. 14
Figure 2.2 Overview of a typical quality risk management process (ICH Q9). 15
Figure 2.3 Example of risk assessment tool: flowchart (basic tool). 21
Figure 2.4 Example of a cause and effect diagram (Ishikawa fishbone diagram). 22
Figure 2.5 Overall QRA process flow for the medicinal product. 27
Figure 2.6 Risk score approach (high level) based on current operating space. 34
Figure 2.7 Risk score approach (detailed level) based on current operating space. 35
Figure 2.8 Tacit benefits of QRM. 44
Figure 3.1 Through the ages. 52
Figure 3.2 ICH QbD timeline. 54
Figure 3.3 From data to knowledge. 56
Figure 4.1 Example of a convergent synthesis.  66
Figure 4.2 Temperature–solubility curve for AZD3342 in 1‐propanol:water and
IMS:water.71
Figure 4.3 Solubility–temperature and metastable zone limit curves. 72
Figure 4.4 Effect of cooling rate on the metastable zone width of sibenadet HCl. 73
Figure 4.5 Crystal16 data for AZD3342 polymorphs A and G. 74
Figure 4.6 Cooling crystallization scenarios: (a) natural cooling, (b) linear cooling,
(c) controlled cooling. 75
Figure 4.7 Effect of seed loading on particle size. 77
Figure 4.8 UV Data for an unseeded linear cooling crystallization of AZD3342. 79
Figure 4.9 FBRM Lasentec data and Photomicrograph for an unseeded, linear
cool crystallization of AZD3342. 80
Figure 4.10 UV Data for the dissolution and crystallization of AZD3342 using
seeding and a cubic cooling profile (controlled cooling). 81
Figure 4.11 FBRM Lasentec data and optical microscopy of AZD3342 crystals
using seeding and a cubic cooling profile (controlled cooling). 81
xiv List of Figures

Figure 4.12 Typical Ishikawa diagram for an anti‐solvent crystallization process. 84

Figure 5.1 Examples of “known and unknowns” to users and suppliers. 110
Figure 6.1 Process of determining how a quality attribute is deemed critical.
*A severity scale is used to assess relative magnitude.  129
Figure 6.2 Examples of where CQAs may be impacted in the manufacturing
process for an oral solid dosage form. 131
Figure 6.3 PDA decision tree for designating parameter criticality. 137
Figure 6.4 Fishbone diagram recognising material attributes (MAs) and process
­parameters (PPs) contributing to the quality attribute, ‘solubility’, of
a solid dispersion API/polymer product. 140
Figure 6.5 Use of risk assessment to determine what DoE studies to use. 142
Figure 6.6 Example of a process parameter criticality assessment
decision tree. 143
Figure 6.7 The manufacturing process flow for ACE tablets.  145
Figure 6.8 Critical parameters affecting blend uniformity for ACE tablets. 146
Figure 6.9 Schematic representation of the relationship between knowledge,
design, and control space. 147
Figure 6.10 Example of terminal sterilisation design space. 148
Figure 6.11 Comparison of traditional controls with advanced controls for the
real‐time release of an oral solid dosage form. 152
Figure 7.1 Constructing a variability gauge chart. 161
Figure 7.2 Impact of formulation of disintegration time with outlier. 162
Figure 7.3 Impact of formulation on disintegration with the corrected data point. 163
Figure 7.4 Prediction profilers showing the impact of formulation on
disintegration time. 164
Figure 7.5 Statistical analysis of the disintegration time data. 165
Figure 7.6 Comparison of the statistical analysis of Study A and Study B. 167
Figure 7.7 Identifying sources of variation. 168
Figure 7.8 The GlaxoSmithKline reaction simulator. 171
Figure 7.9 ‘Understanding and Optimizing Chemical Processes’ 2015
Workshop results. 172
Figure 7.10 Costs, risks and benefits of classical designs. 173
Figure 7.11 Data visualization. 177
Figure 7.12 Analysis of the half fraction. 178
Figure 7.13 Prediction profiles for the four different designs (shaded boxes
indicate statistically significant effects). 179
Figure 7.14 Interaction effects. 180
Figure 7.15 Explaining the experimental procedure to the ‘novice’ workshop
delegate.182
Figure 7.16 Comparison of ‘expert’ and ‘novice’ findings for the half
fraction design. 183
Figure 7.17 Theoretical model of a specific real‐world situation. 185
Figure 7.18 Cycles of learning iterate between the real world and our model of it. 188
Figure 7.19 Resource requirements for central composite design. 189
Figure 7.20 Simulation scenario schematic. 190
Figure 7.21 Analysis of the 10 run fractional factorial experiment. 192
 List of Figures xv

Figure 7.22 Prediction profiler for the 10 run fractional factorial. 192
Figure 7.23 Analysis of the central composite design. 194
Figure 7.24 Analysis of the 13 run DSD experiment. 195
Figure 7.25 Comparison of the central composite and definitive screening
design (DSD). 196
Figure 8.1 Notation used in principal components analysis (PCA). 202
Figure 8.2 PCA derives a model that fits the data as well as possible in the least
squares sense. Alternatively, PCA may be understood as maximizing
the variance of the projection coordinates. 203
Figure 8.3 The raw data curves of the 45 batches of the training set. 205
Figure 8.4 Scatter plot of the two principal components. Each point represents
one batch of raw material. The plot is color‐coded according
to supplier. 205
Figure 8.5 Loading line plots of the two principal components. 206
Figure 8.6 Scores and DModX plots for the particle size distribution data set.
Top left: Predicted scores for the test set batches. Top right: Scores
for the training set batches. Bottom left: Predicted DModX for
the test set batches. Bottom right: DModX for the training set batches. 207
Figure 8.7 Line plot of the power spectral density (PSD) curves of the six early
(red color) and three late (black color) batches of supplier L3. 208
Figure 8.8 Spanning batches can be selected using DoE. Such spanning batches
can be subjected to a thorough investigation in order to ensure
production robustness in all part of the PCA score space.  209
Figure 8.9 (Left) Scatter plot of HS_1 against HS_2, (right) PLS t1/t2 score
plot (note the resemblance to Figure 8.8). 210
Figure 8.10 (Left) Summary of fit plot of the PLS model of the SOVRING subset
with complete Y‐data. Five components were significant according to
cross‐validation. (Right) Individual R2Y‐ and Q2Y‐values of the six
responses. The most important responses are PAR, FAR, %Fe_FAR
and %P_FAR. 211
Figure 8.11 (Left) PLS w*c1/w*c2 loading weight plot. Ton_In is the most influential
factor for PAR and FAR. The model indicates that by increasing Ton_In,
the amounts of PAR and FAR increase. There is a significant quadratic
influence of HS_2 on the quality responses %Fe_FAR and %P_FAR.
This ­nonlinear dependence is better interpreted in a response contour
plot, or a response surface plot. (Right) PLS w*c3/w*c4 weight plot.  212
Figure 8.12 (Left) Response contour plot of PAR, where the influences of
Ton_In and HS_2 are seen. (Right) Response contour plot of FAR.  212
Figure 8.13 (Left) Response contour plot of %P_FAR. (Right) Response contour
plot of %Fe_FAR. 213
Figure 8.14 SweetSpot plot of the SOVRING example, suggesting the SweetSpot
should be located in the upper and right‐hand part.  214
Figure 8.15 A design space plot of the SOVRING example. The green area
corresponds to design space with a low risk, 1%, of failure. 214
Figure 8.16 A schematic representation of how the concepts knowledge space, design
space, and normal operation (control space) region relate to one another. 215
xvi List of Figures

Figure 8.17 The design space hypercube represents the largest regular geometrical
structure that can be inserted into the irregular design space. 216
Figure 8.18 Batch control chart showing predictions for two bad batches.
The BEM fitted using orthogonal PLS readily detects
the deviating batches. 218
Figure 8.19 Batch control chart showing predictions for two bad batches. The BEM
was fitted using classical PLS. 219
Figure 8.20 Contribution plot for batch DoE7 at 1.7 min showing which variables
are contributing to the process deviation. The process upset is caused
by a few of the process parameters. For example, the contribution plot
indicates that the reaction temperature (highlighted by red color)
is much higher than for a normal batch.  219
Figure 8.21 Line plot of the reaction temperature for the non‐NOC batch DoE7.
At the time point of 1.7 min, the reaction temperature is almost 15°
higher than the average temperature across the six NOC batches.
The ­deviation from normality of the reaction temperature for DoE7
increases further into the lifetime of the batch. 220
Figure 9.1 Elements of a PAT system (each element is further described
in Table 9.1). 230
Figure 9.2 Examples of PAT applications used for an oral solid dosage process
and benefits. 241
Figure 9.3 PAT applications used during processing in lieu of conventional
end‐product QC testing. 244
Figure 9.4 Elements to consider for inclusion in a PAT application, whether it is
for gaining process understanding or for process control. 246
Figure 9.5 Manufacturing process based on hot‐melt extrusion showing the main
unit operations from powders to final product, tablets. 248
Figure 9.6 Preliminary risk assessment (RA) analysis (Ishikawa diagram).
The factors highlighted were considered critical for the
extrusion process. 249
Figure 9.7 Representation of about 5000 UV‐Vis spectra from DoEs 1–3
and verification experiments.  250
Figure 9.8 Raw spectra from a DoE set of runs.  250
Figure 9.9 UV‐Vis spectra from the first screening design (DoE1) showing
a sample of the spectra collected.  251
Figure 10.1 Enhanced QbD lifecycle approach to analytical methods. 258
Figure 10.2 Example of the impact of method variability on overall variability
for drug product assay (LSL=95%LC and USL=105%LC). 261
Figure 10.3 Science‐ and risk‐based approach to analytical method design,
development and lifecycle management. 262
Figure 10.4 Ishikawa (fishbone) diagram for a drug product chromatographic
assay method. 265
Figure 10.5 Gage repeatability and reproducibility study. 266
Figure 10.6 I‐MR chart of drug product assay in manufacturing order –
produced using Minitab® Statistical Software. 269
Figure 10.7 Science and risk assessment process. 270
 List of Figures xvii

Figure 10.8 Centred and scaled coefficients for the resolution and separation
factor models. 273
Figure 10.9 Alternative statistical approach – main effect plot amount of TFA
in the mobile phase. 275
Figure 10.10 Centred and scaled coefficients for the signal‐to‐noise model. 276
Figure 10.11 Dissolution data from multivariate experimental study (coloured
by ­analyst) – individual tablet (open circles) and mean data
(solid circles) – produced using Minitab® Statistical Software.  277
Figure 10.12 Main effects plot from multivariate experimental study – produced
using Minitab® Statistical Software. 278
Figure 11.1 The “V” model concept of validation. 283
Figure 11.2 The ASTM E2500 system lifecycle and validation approach. 285
Figure 11.3 Sequence of activities for formulation, process design, and
optimization incorporating process validation activities according
to the lifecycle approach. 296
Figure 11.4 Stages of process validation showing potential changes. 300
Figure 11.5 The elements of a simple, single‐loop control system for
a liquid process. LT = sensor; LC = controller; LV = actuator. 304
Figure 11.6 Control strategy and design space for an IR tablet. 307
Figure 11.7 Typical batch process flow for a granulated tablet. 309
Figure 11.8 ConsiGma™ continuous tablet production line. 314
Figure 12.1 Summary of the intrinsic value of applying QbD. 322
Figure 12.2 The CTD triangle. 323
Figure 12.3 A typical example of an Ishikawa or fishbone diagram. Key: red =
potential high impact on CQA; Yellow = potential impact on CQA;
green = unlikely to have significant impact on CQA. Example:
roller compaction. 326
Figure 12.4 Types of post‐approval changes available in the EU relative to the
adopted level of risk during the evaluation of the procedure. 330
4 Pharmaceutical Quality by Design

As well as from the following industry parties:

●● Europe, the European Federation of Pharmaceutical Industries and Associations (EFPIA).
●● Japan, the Japan Pharmaceutical Manufacturers Association (JPMA).
●● The United States, the Pharmaceutical Research and Manufacturers of America (PhRMA).

ICH produces guidelines under headings of Safety, Quality and Efficacy. It has eminent
and broad‐ranging groups of experts involved in producing these guidelines, so these
­guidances are as near to global as one can obtain, even though they are neither global nor
mandatory, unless – as happens in some cases – regulatory agencies include them in their
national Good Manufacturing Practices (GMPs). Most can be considered, to all intents and
purposes, as internationally applicable.
It was ICH that first brought the term QbD to the pharmaceutical industry when in 2009
it published ICH Q8 (R2) [1], ‘Pharmaceutical Development’.
This was a watershed moment for the industry, as, after this, the importance of taking a
science‐ and risk‐based approach moved to front stage, and even terms like ‘manufacturing
science’ began to be heard.

1.3 Science‐ and Risk‐Based Approaches

Science, of course, has always been a fundamental element of the development of pharma-
ceuticals and, historically, innovative application of science has been core to producing the
many life‐saving and life‐enhancing drugs that the industry has produced over time.
So why all this supposedly new approach? What has changed?
Well, the fundamentals driving the need to understand pharmaceutical science remain
the same, but perhaps the following are factors that influenced a change in perspective:
●● The application of science is becoming more complex; for example, biotechnology‐
based drugs are more complicated to understand, make and analyse compared to small
molecules; specialised therapies such as advanced therapeutic medicinal products, gene
therapies, etc. are beginning to emerge.
●● The use and application of more sophisticated tools, for example, process analytical
technology (PAT), [6] has become more commonplace – although this tool is relatively
new for the pharmaceutical industry, it has been in use by other industries for many years.
●● More powerful data processing is now available to enable such tools to be used. An exam-
ple is design of experiments (DoE) (see Chapter 7) and multivariate analysis (MVA) (see
Chapter 8) can now be used for more sophisticated analysis than was possible previously.
●● The industry has become more global, often with many differing countries involved in
the supply chain, which has made it necessary to maintain quality across various interna-
tional boundaries and cultures.
●● The supply chain has become more fragmented and diverse, with many more parties
involved, including contract research organisations (CROs), contract manufacturing
organisations (CMOs) and external suppliers. ‘Virtual’ companies are now emerging, a
role that did not exist a few years ago.
●● Internal organisations are being re‐structured. An ‘over the wall’ attitude for technology
transfer, development and manufacturing is being heavily discouraged. Business benefits
are being seen in having closer working internal partnerships.
Another random document with
no related content on Scribd:
 Reference has already been made to Aurispa, who appears to
have been the most important manuscript-dealer of his time, not only
in Venice, but possibly in the world. Aurispa sent various agents to
Greece and to the farther East to collect manuscripts and kept
scribes busied in his work-shop in Venice in preparing authentic
copies of these texts. One of his travellers was Plantinerus, who was
sent to the Peloponnesus in 1415, and who succeeded in securing
there some valuable codices.[343] Plantinerus found, in executing his
commissions, that he had to come into competition with a traveller
sent out by Cosimo de’ Medici on a similar errand.
Venice possessed an advantage over the other Italian cities, not
only in the collection of texts, and in its facilities for manifolding
these, but in its position for securing wide sales for the same in the
cities outside of Italy, with which it was, in connection with its active
commerce, in regular relations. The lines of the Oxford printers,
Theo. Rood and Thomas Hunt, printed in their edition of the Letters
of Phalaris, give an indication of the relations of the English
university in the early part of the fifteenth century with the literary
marts of Southern Europe.

Celatos, Veneti, nobis transmittere libros

Cedite, nos aliis vendimus, O Veneti—[344]

(If you Venetians will send over to us the books which have been
hidden (i. e. difficult or rare books, or possibly books unearthed from
far off Eastern regions) we will find sale for the same.)
There is evidence in fact of a very active book-trade between
Venice and England for many years before the introduction into Italy
of the printing-press. The work of Aldus and of those who were
associated with him in carrying on printing and publishing
undertakings in Venice naturally very largely extended these
relations with the English scholars, but the channels for the same
had already been opened. The manuscript-dealers in Venice fixed
their place of business in the most frequented parts of the city—the
Bridge of the Rialto, and the Plaza of S. Mark.
 The trade of the Italian dealers in manuscripts was not brought to
an immediate close by the introduction of printing. The older scholars
still preferred the manuscript form for their books, and found it
difficult to divest themselves of the impression that the less costly
printed volumes were suited only for the requirements of the vulgar
herd. There are even, as Kirchhoff points out,[345] instances of
scribes preparing their manuscripts from printed “copy,” and there
are examples of these manuscript copies of printed books being
made with such literalness as to include the imprint of the printer.
The work of Aldus (continued with scholarly enterprise later by
such men as Froben of Basel and Estienne of Paris) in the printing of
Greek texts, although begun as early as 1495, and although
exercising a very wide influence upon the distribution of Greek
literature, was insufficient to supply the eager demand of the
scholars, while not many other printers were, in the early years of the
exercise of the art, prepared to incur the very considerable risk and
expense required for the production of Greek fonts of type. The risk
was, of course, by no means limited to the cost of the type; the
printers of the earlier Greek books had themselves but slight
familiarity with the literature of Greece, and they were obliged in
many cases to confide the selection and the editing of their texts to
editors to whom this literature was very largely still a novelty. The
printers hardly knew what books to select and they had no adequate
data upon which to base business calculations as to the extent of the
demand that could be looked for for any particular book. The feeling
that they were working in the dark was, therefore, a very natural one.
It was on this ground that, while printing-presses were, during the
century after 1450, multiplying rapidly through Europe, the printing of
Greek books continued to be for a large portion of the period an
exceptional class of undertakings, and work was still found for
scribes who could be trusted to make accurate transcripts of Greek
codices.
Kirchhoff gives the names of the following Italian manuscript-
dealers and scribes whose scholarly activity during the latter half of
the fifteenth century was especially important: Antonius Dazilas,
Cæsar Strategus, Constantius Librarius, Andreas Vergetius, and
Antonius Eparchus. The latter made various journeys to the East in
search of manuscripts. The fact that the dealers in manuscripts very
rarely placed their own names on the copies of the texts sent out
from their work-shops has, in a large number of cases, prevented
these names from being preserved for future record. The names that
have come into record are in the main such as have been referred to
in the correspondence of their scholarly friends and clients. I quote a
few of these references from the lists given by Kirchhoff:
In Bologna the oldest librarius whose work is referred to is Viliaric,
who was called an antiquarius, and whose shop was open in the
beginning of the thirteenth century. In a manuscript, previously
referred to, containing a treatise of Paul Orosius, originally written in
the seventh century, and from which this copy was transcribed early
in the thirteenth, there is at the end an inscription, as follows:

Confectus codex in statione magistri Viliaric antiquarii,

Ora pro me scriptore, sic dominum habeas protectorem.[346]

(This codex was completed in the stall of Master Viliaric,

bookseller; pray for (the soul of) me, the scribe, and you shall have
the Lord for your protector.)
This codex seems to have been prepared, according to the usual
university practice, for hire, as on the sixteenth page there is noted
the memorandum, “this quaternio has five sheets.”
In 1247, Nicolaus is recorded as being the stationarius
universitatis, and in the same year a certain Johannes Cambii is
recorded as a stationarius librorum; and Minghinus as a stationarius
peciarum. Here we have in one year record of three classes of
scribes being at work. They were all noted as being doctors of the
law, and they all appear on the list of persons exempt from military
service.
Later in the same century, a certain Cervotti, who had inherited
from a deceased brother a collection of books, undertook to use
these for profit by offering them for hire. The list of the books, drawn
up by the notary Noscimpax, has been preserved, and includes
twenty different works. Certain of these are collections of the
university lectures in the Faculty of law, and the others have also, in
the main, to do with the subject of jurisprudence. The first book on
the list is Diversitates Dominorum, and the last Margarita Gallacerti,
which latter does not appear properly to belong to the subject of
jurisprudence.
In the year 1400, there is reference to a scribe named Moses and
specified as a Jew, which, in view of the university regulations
previously referred to prohibiting the sale of books by Jews or to
Jews, is noteworthy.
The entry appears at the close of a manuscript of Bartholomæus
Brixiensis:
Emi hunc librum anno domini MCCCC die XXI. Mensis novembris
a Moysi Judeo pro viii. florenes.
Kirchhoff is of opinion that Moses must have been a travelling
pedlar, as it is difficult to believe that a Jew could have at that time
secured the post of a licensed university scribe.[347]
In Verona, there is reference to a certain Bonaventura, who is
recorded as a scriptor, and who seems to have occasionally utilised
for his manuscript work the hand of a woman. An inscription on one
of the manuscripts by Bonaventura, quoted by Endlicher, reads as
follows:

Dextra scriptoris careat gravitate doloris.

Detur pro penna scriptori pulchra puella.[348]

In Florence, the earliest librarius of note was probably Johannes

Aretinus, whose work continued during the years between 1375-
1417. Ambrosius Camaldulensis, who had so much to do with books
and with literature, takes pains, in a letter written in 1391, to send a
cordial greeting to the librarius Aretinus.[349] Bandini prints a letter of
Petrarch’s in which the latter refers to Aretinus as a friend for whom
he has a high regard and as a man of exceptional knowledge and
clearness of insight, and specifies, as works that he valued highly,
nine manuscripts which had been written by the hand of Aretinus.
These included Aristotle’s treatise on Ethics, several Essays of
Cicero, the Histories of Livy, Cicero’s Orations, Barbari on Marriage,
etc.
Kirchhoff gives a list of fourteen other Florentine librarii, whose
work extended over the years between 1410 and 1480. The latter
date is sixteen years later than the introduction of the printing-press
into Italy.
The most noteworthy by far of these manuscript-dealers of
Florence was Philippi Vespasiano, who has been previously referred
to, and who is to be ranked not only as the most important publisher
of the manuscript period, but as one of the great scholars of his time,
and as a man whose friendship was cherished by not a few of the
leaders of thought during the earlier period of the Renaissance. In
one of the Florentine collections has been preserved a number of
letters written to Vespasiano by his scholarly friends between the
years 1446 and 1463, and these letters show how honoured a
position he held in the generation of his time. He was, in fact, in
character and in ambition, as well as in the nature of his work, a
worthy predecessor of Aldus, and he lived long enough himself to
have seen some of the productions of the Aldine Press.
In his earlier years, Vespasiano was for a time secretary to
Cardinal Branda in Rome, and it is during this time that he devoted
himself earnestly to classic studies. It was while he was in Rome that
he began work upon a literary undertaking of his own, which
comprised a series of Memoirs of the noteworthy men of his time
with whom he had come into relations. The Medici, Duke Borso of
Ferrara, and other of the scholarly nobles made large use of
Vespasiano’s collections of manuscripts and facilities for producing
authentic transcripts.
He was one of the Italian dealers whose agents were actively at
work in Greece and in Asia Minor in the collecting of manuscripts,
and the clients to whom he supplied such manuscripts included
correspondents in Paris, Basel, Vienna, and Oxford.
 In the Bodleian Library in Oxford is a codex containing certain
works of Cyprian, on the first sheet of which is inscribed:
Vespasianus librarius Florentinus hunc librum Florentiæ
transcribendum curavit. (Vespasian, a Florentine librarius, had this
book transcribed at Florence.)
Another manuscript in the same collection, containing a
commentary on some comedies of Terence, is inscribed as follows:
Vespasianus librarius Florentinus fecit scribi Florentiæ.
(Vespasian, a Florentine librarius, had this book written in Florence.)
Both codices are beautiful examples of the best manuscript work
of the period.[350]
There are various references of the time showing that manuscripts
which bore the stamp of Vespasiano were not only beautiful in their
form, but possessed probably a higher authority than the work of any
other manuscript-dealer of the age for completeness and for
accuracy. He took contracts for the production of great libraries, and
it is recorded that, in preparing for Cosimo de’ Medici a collection of
two hundred works, he employed forty-five scribes for a term of
twenty-two months.[351] Vespasiano died in 1496, one year later
than the establishment in Venice of the Aldine Press.
Agnolo da Sandro is described as a bidellus, a manuscript-dealer,
in Florence as late as 1498, at which time the trade in manuscripts
must already have begun very seriously to diminish. Niccolo di
Giunta, who was active in the manuscript trade in Florence towards
the end of the fifteenth century, is famous as having been the
founder of the family of Giunta or Junta, which later took such an
important part in printing and publishing undertakings in Italy.
In Perugia, the first record of a manuscript-publisher bears date as
late as 1430. The name is Bontempo, and his inscription appears on
a parchment copy of an Infortiatum.
While there are various references to manuscript-dealers in Milan
of an early date, the first inscription bears date as late as 1452. The
name is Melchoir, who is described as a “dealer of note.” Filelfo
speaks of Melchoir as having copies of Cicero’s Letters for sale at
ten ducats each.[352]
Paolo Soardo, who was in business between 1470 and 1480, is
described as an apothecary and also as a dealer in delicatessen, but
he seems to have added to his employment that of a manifolder and
seller of manuscripts.
Jacobus Antiquarius speaks of having purchased from Paolo in
1480 a Roman history for the sum of one aureus. In Padua, Jacob, a
Jew, succeeded, notwithstanding the university regulations against
dealing in manuscripts by Jews, in carrying on between 1455 and
1460 a business in the sale of manuscripts. His inscription appears
on a number of classical codices of the time, and in a manuscript of
Horace, dating from the twelfth century, the owner makes reference
that he purchased the same in 1458 from Jacob, the Hebrew
librarius.[353]
The records of Ferrara give the names of Carnerio, bibliopola, and
of several others as doing business in manuscripts between 1440
and 1490.
In Rome the records of 1454 speak of Giovanni and Francisco as
cartolaji and librarii, that is to say, dealers in paper and also in
manuscripts. In that year these dealers had for sale among their
things, Letters of Cicero (without which work no well regulated
manuscript-dealer’s collection appears to have been complete) and
the works of Celsus. A copy of the latter was bought for Vespasiano
for the sum of twenty ducats. There is record during the same year
of a certain Spannocchia who also had Cicero’s Letters for sale.
In Genoa there were at this time one or two manuscript-dealers,
but, as before stated, the readers and scholars of Genoa appear for
the most part to have supplied themselves from Florence.
The most important trade in manuscripts during the fourteenth and
fifteenth centuries, as was the case during the fifteenth century with
the trade in printed books, was carried on in Venice and Florence. As
early as 1390 the inscription of Gabriel Ravenna, librarius, appears
in a copy of Seneca’s Tragedies.[354] Kirchhoff is of opinion that
Gabriel conducted, during the last fifteen years of the fourteenth
century, an important work-shop for the production of manuscripts.
A year or two later, occurs the name of Michael, a German
librarius, but it is possible that Michael’s work was more nearly that
of a secretary than of a manuscript-dealer. As Kirchhoff points out, it
is not always easy at this stage of the trade in manuscripts, to
distinguish between the inscriptions of the manuscript-dealers
certifying to the correctness of the copy sent out from their shops,
and the inscriptions of the scribes or secretaries who, having
completed for this or that employer specific copies of the works
required, added their names as a record on the final sheet.
Reference has already been made to Johannes Aurispa, by far the
most important of the manuscript-dealers of his time and possibly of
the entire Middle Ages. Aurispa was born in 1369 in Sicily. The
earlier years of his life were passed in Constantinople, where he
appears to have held a position of some importance in connection
with the Court. While in Constantinople, he began to make
collections of manuscripts, and he organised there a staff of skilled
scribes. In 1423, at the invitation of his friends, Ambrosius
Camaldulensis and Niccolo de’ Niccoli, he came to Florence,
bringing with him an invaluable collection of 238 manuscripts.
To this store he afterwards added, while in Florence, a further lot
of codices which he had had sent from Constantinople to Messina.
At this time, his interest in the collection of manuscripts appears to
have been a matter of scholarship merely and of sympathy with the
efforts of certain Florentine scholars whom he came to know, to
secure the material for their classical studies.
Later, however, in connection, doubtless, with the many
applications that came to him for transcripts of his codices, he
decided to organise a business as a bookseller and publisher. Before
taking this course, he had, it appears, sought a position as instructor,
first in Florence and afterwards in Bologna and in Ferrara, but had
not succeeded in finding the kind of a post that suited him.
Part of the evidence of his change of mind comes to us through
letters from Filelfo, whose keen scholarly interest brought him into
close relations with men having to do with literary production. Filelfo
writes to Aurispa, in 1440:
Totus es in librorum mercatura, sed in lectura mallem. Quid enim
prodest libros quotidie, nunc emere, nunc vendere, legere vere
nunquam! (You are completely absorbed in the occupation of trading
books, but I should choose that of reading them. For what does it
profit you to buy and sell books every day if you never have time for
their perusal.)
And again in 1441:
Sed ex tua ista taberna libraria nullus unquam prodit codex, nisi
cum quæstu.[355] (No book ever leaves your book-shop, except at a
profit to you.)
The publishing undertakings of Aurispa were devoted almost
entirely to works of classical literature. Among the authors whose
names appear either in the lists of books offered by him or in the
correspondence of his friends and clients, are as follows:
Philo Judæus, Strabo, Theophrastus, Demosthenes, Xenophon,
Proculus, Homer, Aristarchus, Athenæus, Sophocles, Æschylus,
Pindar, Oppian, Proclus, Eusebius, Gregory, Aristotle, Plutarch,
Plotinus, Lucian, Dio Cassio, Diodorus, and other Greek authors.
The Latin writers included Cicero (of necessity), Virgil, Pliny,
Quintilian, Macrobius, Apicius, and Antonius.
Aurispa seems to have enjoyed the confidence and friendship of
all the noted Italian scholars of his time, and the letters of his
correspondents speak with very cordial appreciation as well of the
importance of his services to literature, as of the extent of the
accuracy of his own scholarship. The only correspondent with whom
he appears to have had any trouble was Filelfi, but if Filelfi had not
managed to have friction with Aurispa, the bookseller would have
been an exception among the contemporaries of this irritable and
self-sufficient scholar.
In 1450, being then well advanced in years, Aurispa gave up his
business undertakings, took priestly orders, and lived thereafter as a
scriba apostolicus, dividing his time between Ferrara and Rome. He
declined tempting offers, made through his friend Panormita, to join
the literary circle of King Alphonso which had been brought together
about the Court in Naples.
After Aurispa’s death, Filelfo gave to his son-in-law, Sabbatinus, a
very cordial word of appreciation of the services and of the character
of the publisher. A portion of the manuscripts belonging to Aurispa’s
collection was purchased in 1461 by Duke Borso of Ferrara for two
hundred ducats.
A large collection of manuscripts was, however, in Aurispa’s
possession at the time of his death, and these were taken charge of
by Bartholomæus Facius, and, after various vicissitudes, many of
them have since found place in existing collections of Florence,
Venice, Vienna, Paris, and London. A selection of the letters
between Aurispa and his near friend Camaldulensis has also been
preserved.
Books in Spain.—At the time when the great manuscript-
dealers of Venice and Florence were carrying on business with the
literary centres of France, Germany, and England, they had some
dealings also with Spain; but their correspondence was practically
limited to the University of Salamanca, which had been founded
about 1220. The literary activities of Spain during the fifteenth
century were certainly much less important than those of either Italy
or France. They were of necessity seriously hampered by the long
series of wars with the Moors, while the final overthrow in 1492 of
the Moorish kingdom of Granada doubtless had, as one of many
results, a decidedly unfavourable influence upon the intellectual
development and the literary possibilities of the Peninsula. For two
centuries or more the scholars of the Moorish kingdom had busied
themselves in making collections of Arabic literature, while of not a
few of the more noteworthy works they caused to be prepared
versions in Latin, by means of which the books were made available
for the use of instructors and students in Salerno, Bologna, Padua,
and Paris. It was the case also that the first knowledge of certain
Greek authors came to the scholars of Europe through the Latin
translations which were produced in Cordova from the Arabic
versions. The Moorish scholars thus became a connecting link for
the transmission to the Western world of the philosophy and learning
of the East. Until its conquest and practical destruction by the
Spaniards in 1236, Cordova had been not only the political capital
but the centre of the intellectual life of the Moorish kingdom, so that it
was spoken of as the Athens of the West. At the close of the tenth
century it is said to have contained nearly one million inhabitants. In
connection with the work of its university and of the great library, a
large body of skilled scribes were busied with the manifolding of
manuscripts, and there appears to have been a regular exchange of
manuscripts between Cordova and Baghdad.
In the year 995, Thafar Al-baghdádé, the chief of the scribes of his
time, came from Baghdad and settled in Cordova. The Khalif Al-
hakem took him into his service and employed him in transcribing
books. The Khalif surpassed every one of his predecessors in the
love of literature and of the sciences, which he himself cultivated with
success and fostered in his dominions. Through his influence,
Andalusia became a great market to which the literary productions of
every clime were immediately brought for sale. He employed
merchants and agents to collect books for him in distant countries,
remitting for the purpose large sums of money from the treasury,
until, says the chronicler, “the number of books in Andalusia
exceeded all calculation.” The Khalif sent presents of money to
celebrated authors in the East with a view to encourage the
publication of works or to secure the first copies of these. Hearing,
for instance, that Abú-l-faraj of Ispahán had written a book entitled
Kitábu-l-aghani (The Book of Songs), he sent him a thousand dinars
of pure gold, in consideration of which he received a copy of the
work before it had been published in Persia. He did the same thing
with Abú Bekr Al-abhari, who had published a commentary on the
Mokhtassar.
Al-hakem also collected and employed in his own palace the most
skilful men of his time in the arts of transcribing, binding and
illuminating books. The great library that he brought together
remained in the palace of Cordova, until, during a siege of the city by
the Berbers, Hájib Wadheh, a freedman of Al-mansúr, ordered
portions of the books to be sold, the remainder being shortly
afterwards plundered and destroyed on the taking of the city. The
extent of the collection is indicated by the description of the
catalogue. In the Tekmílah, Ibun-l-abbáns is quoted by Al-Makkari as
saying that the catalogue comprised forty-four volumes, each volume
containing twenty sheets. Makkari estimates that the library
contained no less that four hundred thousand volumes. It is possible
that this number was over-estimated, at least, if we are to believe the
statement of Ibun-l-abbar that the Khalif Al-hakem had himself read
every book in the collection, writing on the fly-leaf the dates of his
perusal and details concerning the author.
Makkari gives a long list of famous authors who flourished in
Andalusia during the reign of Al-hakem, their productions including
works in law, medicine, history, topography, language, and poetry.
One of the historians, Al-tári-khí, was a paper merchant, and was
accordingly known by the name of Al-Warrak. I do not find record of
the names of any dealers in books or any account of the means
employed for their distribution.[356]
The Manuscript Trade in France.—While, in Italy, the
more important part of the trade in manuscripts was carried on
outside of the university circles, in France the university retained in
the hands of its own authorities the control and supervision of the
work of the manuscript-dealers; and the book-trade of the country,
not only during the manuscript period, but for many years after the
introduction of printing, was very directly associated with the
university organisation. The record of the production and of the trade
in books carried on by the stationarii, librarii, and the printer-
publishers of the university is presented in the chapter on the Making
of Books in the Universities.
During its earlier years, the trade in manuscripts was limited
practically to the city of Paris. The work of the official university
scribes in Paris was very similar to that which has already been
referred to for Bologna. It appears, however, that, in accordance with
the Parisian methods, there was less insistence upon the practice of
hiring manuscripts, either complete or in divisions, and there was an
earlier development of the practice of making an absolute sale of the
texts required.
Kirchhoff traces the beginning of the manuscript-trade back to the
second half of the eleventh century. He says that it is not clear
whether the earlier dealers were able to devote themselves
exclusively to the business of selling books, or whether, as he thinks
it more probable, they associated this business with some other
occupation. Jean de Garland, who compiled a kind of technological
directory or list of industries carried on in Paris in 1060, says:
Paravisus est locus ubi libri scholarium vendentur.[357] He is
apparently referring to the Place near the Cathedral Church, which
later became the centre of the Parisian book-trade. Peter of Blois,
writing, in the middle of the twelfth century, to an instructor in
jurisprudence in Paris, makes a more definite reference to the
Parisian manuscript-dealers. He speaks of the great collections of
valuable books which the Parisian dealers have for sale, and
laments the narrowness of his purse which prevents him from
purchasing many things which have tempted him.[358]
Bulæus, in his History of the University of Paris, published in 1665,
maintains that as early as 1174, the manuscript-dealers of Paris
formed a part of the organisation of the university, and that their work
had been brought fully under the regulation of the university
authorities. The university statistics, before the thirteenth century, do
not, however, appear fully to bear out this contention. The first
statutes which give detailed regulations concerning the book-trade
bear date as late as 1275. These statutes specify what texts and
what number of copies of each text the licensed booksellers should
keep in stock, and give a schedule, as was done in Bologna and
Padua, of the prices at which the loans and sales should be made.
Kirchhoff is of opinion that, prior to the middle of the thirteenth
century, the book-trade connected with the university, while it had
already assumed considerable proportions, had not been brought
thoroughly under university control. With this control came also as an
effect, the privileges which attached to the dealers as members of
the university body, and there is no evidence that the booksellers
enjoyed these privileges before 1250. Depping takes the ground,
that during the fifteenth century the sale of books in Paris was not
sufficient to constitute a business in itself, and that all dealers in
books had some other occupation or means of support, and
interested themselves in the sale of manuscripts only as an
additional occupation.[359]
It appears hardly likely, however, that manuscript-dealers should
be able to secure immunity from the general tax, which fell upon
nearly all other classes of dealers, on the ground of the importance
of their trade for education, unless they were able to show that they
were actively engaged in such trade. The regulation quoted by
Depping specifies among the free citizens of the city of Paris who
were not liable to the King’s tax,—libraires parcheminiers,
enlumineurs, escriipveins. It was evidently the intention of the
framers of the law to include under the exemption all dealers upon
whose trade the preparation and sale of manuscripts was directly
dependent. Under this heading were included, of necessity, the
scribes, the illuminators (who added to the text of the scribes the
artistic decorations and initial letters), and (most important of the
three) the dealers in parchment.
The fact that the booksellers are named in this schedule
separately from the scribes is an indication of the existence of a
bookselling trade of sufficient importance to call for the work of
capitalists employing in the preparation of their manuscripts the
services of the scribes and of the other workmen required. Work of
this kind can properly be classified as publishing.
The dealer was himself prohibited from making purchase of a
manuscript left in his hands until this had been offered for sale during
the term of not less than one month. Record was to be kept of the
name of the purchaser and of the price received.
The requirement that the price obtained for a manuscript should
be recorded, has secured the preservation, on a number of
manuscripts of the time, of a convenient record of their market value.
In a collection of sermons dating from the latter part of the fifteenth
century, for instance, is the record, “This book was sold for 20
Parisian sols.” In a text of Ovid of about the same time is noted
simply the price,—6 sols, Parisian.[360]
Newly prepared transcripts could not be licensed for renting until
they had been examined and passed as correct by the officials, and
until their renting prices had been placed on record. No new work
could be included in the lists of the stationarii until license for the
same had been secured. At this date, the usual term of rental of a
manuscript was one week, and an additional charge could be made
if the manuscript was held in excess of that time. In case a member
of the university had transcribed an incorrect or incomplete
manuscript, the stationarius was liable to him for damages to cover
his wasted labour. According to the general practice, the hirer of a
manuscript was obliged to deposit a pledge for the same, which
pledge could be disposed of by the stationarius after the term of one
year.
In the schedule presented by Chevillier of manuscripts licensed in
the early part of the thirteenth century, the prices specified cover only
the rates for renting. Chevillier points out that there is in this
schedule no indication of the division of the manuscripts into pecias,
the practice which was, as we have seen, the usual routine in the
Italian universities.[361]
An appraisal of the books contained in the library of the Sorbonne
in the year 1292 gives a value of 3812 livres, 10 sous, 8 deniers.[362]
The regulations concerning the sale of works on commission were
renewed in 1300, with provisions which must have rendered this
class of business not only unremunerative but peculiarly
troublesome. Such a sale could be made only in the presence of two
witnesses. No other bookseller was at liberty to purchase the book,
excepting with the permission and in the presence of the original
owner. Before a sale was made to a bookseller, the manuscript must
be allowed to remain exposed for sale not less than four days in the
library of the Dominican monastery.
Exceptions to the above regulations were permitted under the
express authority of the Rector of the university in case the original
possessor of the manuscript might be in immediate need of money, a
condition which probably obtained in a large number of cases.
The general purpose of these regulations appears to have been
the prevention of any undue increase in the market price or selling
value of manuscripts, or the “cornering of the market” on the part of
the manuscript-dealers in connection with texts which might be in
demand. Existing regulations of this kind tended, however, naturally
to fall into desuetude.
In 1411, an ordinance of Charles VI. made fresh reference to the
necessity of such supervision, mainly on the ground of the
convenience of tracing stolen manuscripts or unlicensed
manuscripts.
In 1342, the librarii were permitted to increase their selling
commission from four deniers to six deniers in the case of
manuscripts sold by them for clients who were not themselves
members of the university. Kirchhoff points out, however, that this
commission could by no means have represented the actual charges
made. The University of Paris claimed the authority to license its
librarii, and to carry on business not only in Paris but throughout
France. Librarii from without were, however, strictly prohibited from
carrying on business in Paris.
There were in Paris, in addition to the stationarii and librarii, a
certain number of unlicensed dealers who were not members of the
university, and who might be classed as book pedlars. While these
book pedlars enjoyed no university privileges, their business was
subjected to the supervision of the university authorities. It was the
purpose of the regulations to prevent dealers of this kind from taking
part in any higher grade book business. They were, for instance,
forbidden to sell any volume for a higher price than ten sous, which,
of necessity, limited their trade practically to chap-books, broadsides,
etc. They were also forbidden to trade in any covered shops, their
business being carried on in open booths. In case they were at any
time found to be trenching upon the business of the licensed or
certified book-dealers (libraires jurés), they forfeited promptly their
permits as book pedlars.
 In 1323, the Paris School was the most important in Europe for
theological studies, as that of Bologna was the authority on
jurisprudence, and that of Padua for medicine; and the trade of the
Paris booksellers was, therefore, largely devoted to theological
writings. It is partly on this ground that the records of the
monasteries in which there was scholarly and literary activity make
more frequent reference during this century to Paris as a book centre
than to any one of the Italian cities. When, for instance, King Wenzell
II. of Bohemia, at the time of the founding of the Cistercian abbey of
Königsaal, presented two hundred marks of silver for the
organisation of its library, the Abbot Conrad had, he reports, no other
course to take than to travel to Paris in order to purchase the books.
This was in the year 1327.[363] Johann Gerson, writing in 1395 to
Petrus de Alliaco, speaks of the wealth of the literary stores available
at this time in Paris. The list that he gives as an example of these
treasures is devoted exclusively to theological works.
While it is difficult to understand from the evidence available what
machinery may have been in existence during the thirteenth and
fourteenth centuries for the distribution of the books, there are
various references to indicate that such distribution took place
promptly over a very considerable territory. The anonymous author
of a polemical tract, written in order to point out the errors of some
heretical production, says:
Is autem erroneus liber positus fuit publice ad exemplandum
Parisius anno domini 1254. Unde certum est, quod jam publice
predicaretur, nisi boni prelati et predicatores impedirent.[364]
(In Paris in the year of our Lord 1254, this heretical book was
openly given to the scribes to be copied. Whence it is evident what
manner of doctrine would now be set forth to the public had not good
priests and preachers interfered.)
Kirchhoff is of the opinion that there began to be at this time in
connection with the work of the contemporary authors a kind of
publishing arrangement under which the author handed over to the
stationarii or to the librarii his literary production for multiplication and
for publication, either through renting, through sale, or in both
methods. He finds in the manuscript of a tract by Gerson, which was
given to the public in the year 1417, a notice to the effect that this
was published in Paris under the instructions of the author and under
the license of Magister Johannus, Cancellarius.[365]
The work of the manuscript-dealers was carried on in booths or
shops in various open places, but as a rule in the immediate
neighbourhood of the churches. Certain booths were to be found,
however, on the bridges and by the courts of justice; and a
neighbourhood particularly resorted to by the booksellers was the
Rue Neuve Notre Dame, where, in the year 1292, out of eight
licensed book-sellers, no less than three had their work-shops. On
the bridge Neuf Notre Dame, there were at the time of its falling, in
1499, a number of booksellers, three of whom are recorded as
having lost their stock through the accident. The places selected by
the earlier dealers in manuscripts became later the centre of the
Parisian trade in printed books.
As a result of their membership in the university, the dealers in
manuscripts shared in the exemption from the taxation enjoyed by
the university body. The royal tax collectors persisted, however, from
time to time in ignoring this right of exemption, and it was therefore
necessary at different periods to secure fresh enactments from the
royal ordinances in order to confirm the privilege. An example of
such an ordinance is that issued by Philip the Fair, in 1307. In the
cases in which the university placed an impost upon its members for
any special purpose, the manuscript dealers were, of course, obliged
to assume their share of such impost. At the time of their acceptance
as official or licensed dealers, they had to pay a fee, in the first place
of four sous, but after 1467 of eight sous. For the privilege of keeping
an open shop, the fee was twenty-four sous. A further fee of eight
sous was payable for each apprentice, and a weekly payment of
twelve deniers payable for each workman. These fees went into the
treasury of the booksellers’ corporation.
After 1456, under the enactment of the congregation of the
university, each manuscript dealer and paper dealer was called upon
to pay to the Rector of the university at the time of his acceptance
and license a scutum of gold.
 The four taxatores, the officials charged with the supervision of the
fees for the booksellers’ guild (usually the four senior or most
important members of the guild), were also charged with the
selection or approval of new members and with the supervision of
the proper carrying out of the various regulations controlling the
organisations of the guild. In the earlier period of the work, such
censorship as was found necessary concerning the books to be
published was exercised through these four taxators. They were also
the official representative body of the university guild.
In case any member of the guild suffered injury from unauthorised
competition, the guild had the power to suspend the business
operation with the person charged with committing the injury, until
the complaint could be passed upon. In case the rules of the
corporation had been broken, the corporation appears to have had
the power, at least up to the beginning of the fifteenth century, of
withdrawing the trade privilege or license.
The taxators or principales jurati, as they were sometimes called,
had power to proceed not only to supervise the business undertaking
of the members of the guild, but were also authorised to take
measures against the outside or unlicensed booksellers and to
proceed, if necessary, even to the point of seizure and confiscation
of their goods. In carrying out such measures, they were empowered
to call upon the university bedels for co-operation.
These unlicensed dealers or book pedlars, as they increased in
numbers, naturally attempted to withdraw themselves from the
jurisdiction and supervision of the university authorities. An
ordinance of Charles VI., dated June 20, 1411, confirms specifically
the right of control over the entire book-trade, and prohibits pedlars,
dealers, hucksters, etc., from taking part in the selling of
manuscripts, “of which business they could have no understanding.”
The edict went on to specify that the carrying on of the book
business by ignorant and irresponsible dealers not only caused
injury to the licensed book-dealers, but was a wrong upon the public,
in that it furthered the circulation of incorrect, incomplete, and
fraudulent manuscripts. This ordinance was doubtless issued at the
instance of the book-dealers’ guild, but it is evident that it was not
strictly carried out, as from year to year there are renewed
complaints of the competition of these ignorant and irresponsible
book pedlars.
It was considered important, in order to insure the proper control
by the university over the book-trade and the interests of the
scholars who depended upon the book-dealers for their text-books,
that the trade in the materials used in the manifolding of books
should also be strictly supervised. The special purpose of the
university authorities was to prevent any “cornering of the market” in
parchment, and to insure that the supply of this should be regular
and uniform in price.
Under the ordinance of 1291, the dealers in parchment were
forbidden to keep any secret stores of the same, but were obliged to
keep on file with the managers of the book guild the record of the
stock carried by them from month to month. The parchment-dealers
licensed to do business in Paris were forbidden to sell parchment to
dealers from outside of Paris. On the first day of the Trade Fair,
when foreign dealers brought parchment to Paris for sale, the
Parisian dealers were forbidden themselves to make purchases, this
day being reserved for such purchases as the university officials
might desire to make. In case, after the first day of the Fair, a foreign
dealer in parchment had before him more applications for his stock
than could be supplied, and among the applicants there should be
one representing the university, the latter was to be served first.
Outside of the time of the official Fair, the Paris dealers in parchment
were allowed to make purchases of their material only in the
monastery of S. Mathurin.
In case between the times of the Fair a foreign dealer or
manufacturer of parchment came to Paris, he was obliged to place
his stock in this same monastery and to give information concerning
this deposit to the Rector of the university. The Rector sent a
representative to examine and to schedule the parchment, and the
stock was priced by four of the licensed parchment-dealers
associated with the university. The university authorities had then for
twenty-four hours the first privilege of purchase. This regulation was